common : check for null getpwuid in hf-cache (#22550)

Signed-off-by: Adrien Gallouët <[email protected]>

.github/workflows/python-type-check.yml

@@ -31,7 +31,7 @@ jobs:
         uses: actions/setup-python@v6
         with:
           python-version: "3.11"
-          pip-install: -r requirements/requirements-all.txt ty==0.0.26
+          pip-install: -r requirements/requirements-all.txt ty==0.0.33
       # - name: Type-check with Pyright
       #   uses: jakebailey/pyright-action@v2
       #   with:

common/arg.h

@@ -25,7 +25,8 @@ struct common_arg {
     const char * value_hint_2 = nullptr; // for second arg value
     const char * env          = nullptr;
     std::string help;
-    bool is_sparam = false; // is current arg a sampling param?
+    bool is_sampling = false; // is current arg a sampling param?
+    bool is_spec = false; // is current arg a speculative decoding param?
     bool is_preset_only = false; // is current arg preset-only (not treated as CLI arg)
     void (*handler_void)   (common_params & params) = nullptr;
     void (*handler_string) (common_params & params, const std::string &) = nullptr;
@@ -74,7 +75,8 @@ struct common_arg {
     common_arg & set_examples(std::initializer_list<enum llama_example> examples);
     common_arg & set_excludes(std::initializer_list<enum llama_example> excludes);
     common_arg & set_env(const char * env);
-    common_arg & set_sparam();
+    common_arg & set_sampling();
+    common_arg & set_spec();
     common_arg & set_preset_only();
     bool in_example(enum llama_example ex);
     bool is_exclude(enum llama_example ex);

common/common.cpp

@@ -70,7 +70,7 @@ common_time_meas::~common_time_meas() {
 // CPU utils
 //
 
-int32_t cpu_get_num_physical_cores() {
+int32_t common_cpu_get_num_physical_cores() {
 #ifdef __linux__
     // enumerate the set of thread siblings, num entries is num cores
     std::unordered_set<std::string> siblings;
@@ -185,11 +185,11 @@ static int cpu_count_math_cpus(int n_cpu) {
 /**
  * Returns number of CPUs on system that are useful for math.
  */
-int32_t cpu_get_num_math() {
+int32_t common_cpu_get_num_math() {
 #if defined(__x86_64__) && defined(__linux__) && !defined(__ANDROID__)
     int n_cpu = sysconf(_SC_NPROCESSORS_ONLN);
     if (n_cpu < 1) {
-        return cpu_get_num_physical_cores();
+        return common_cpu_get_num_physical_cores();
     }
     if (is_hybrid_cpu()) {
         cpu_set_t affinity;
@@ -202,7 +202,7 @@ int32_t cpu_get_num_math() {
         }
     }
 #endif
-    return cpu_get_num_physical_cores();
+    return common_cpu_get_num_physical_cores();
 }
 
 // Helper for setting process priority
@@ -263,7 +263,7 @@ bool set_process_priority(enum ggml_sched_priority prio) {
 //
 
 
-void postprocess_cpu_params(cpu_params& cpuparams, const cpu_params* role_model) {
+void postprocess_cpu_params(common_cpu_params & cpuparams, const common_cpu_params * role_model) {
     int32_t n_set = 0;
 
     if (cpuparams.n_threads < 0) {
@@ -271,7 +271,7 @@ void postprocess_cpu_params(cpu_params& cpuparams, const cpu_params* role_model)
         if (role_model != nullptr) {
             cpuparams = *role_model;
         } else {
-            cpuparams.n_threads = cpu_get_num_math();
+            cpuparams.n_threads = common_cpu_get_num_math();
         }
     }
 
@@ -1521,7 +1521,7 @@ struct llama_context_params common_context_params_to_llama(const common_params &
     return cparams;
 }
 
-struct ggml_threadpool_params ggml_threadpool_params_from_cpu_params(const cpu_params & params) {
+struct ggml_threadpool_params ggml_threadpool_params_from_cpu_params(const common_cpu_params & params) {
     struct ggml_threadpool_params tpp;
 
     ggml_threadpool_params_init(&tpp, params.n_threads); // setup the defaults

common/common.h

@@ -54,7 +54,7 @@ struct common_control_vector_load_info;
 // CPU utils
 //
 
-struct cpu_params {
+struct common_cpu_params {
     int      n_threads                   = -1;
     bool     cpumask[GGML_MAX_N_THREADS] = {false}; // CPU affinity mask.
     bool     mask_valid                  = false;   // Default: any CPU
@@ -63,8 +63,8 @@ struct cpu_params {
     uint32_t poll                        = 50;      // Polling (busywait) level (0 - no polling, 100 - mostly polling)
 };
 
-int32_t cpu_get_num_physical_cores();
-int32_t cpu_get_num_math();
+int32_t common_cpu_get_num_physical_cores();
+int32_t common_cpu_get_num_math();
 
 //
 // Common params
@@ -297,34 +297,19 @@ struct common_params_model {
 
 struct common_ngram_mod;
 
-struct common_params_speculative {
-    common_speculative_type type = COMMON_SPECULATIVE_TYPE_NONE; // type of speculative decoding
+// draft-model-based speculative decoding parameters
+struct common_params_speculative_draft {
+    int32_t n_max = 16; // maximum number of tokens to draft during speculative decoding
+    int32_t n_min = 0;  // minimum number of draft tokens to use for speculative decoding
 
-    // general-purpose speculative decoding parameters
+    float p_split = 0.1f;  // speculative decoding split probability
+    float p_min   = 0.75f; // minimum speculative decoding probability (greedy)
 
-    int32_t n_max   = 16; // maximum number of tokens to draft during speculative decoding
-    int32_t n_min   = 0;  // minimum number of draft tokens to use for speculative decoding
-    float   p_split = 0.1f; // speculative decoding split probability
-    float   p_min   = 0.75f; // minimum speculative decoding probability (greedy)
+    common_params_model mparams;
 
-    // ngram-based speculative decoding
+    llama_model * model = nullptr; // a llama_model that can be shared by multiple speculative contexts
 
-    uint16_t ngram_size_n   = 12; // ngram size for lookup
-    uint16_t ngram_size_m   = 48; // mgram size for speculative tokens
-    uint16_t ngram_min_hits = 1; // minimum hits at ngram/mgram lookup for mgram to be proposed
-
-    std::shared_ptr<common_ngram_mod> ngram_mod;
-
-    std::string lookup_cache_static;  // path of static ngram cache file for lookup decoding           // NOLINT
-    std::string lookup_cache_dynamic; // path of dynamic ngram cache file for lookup decoding          // NOLINT
-
-    // draft-model speculative decoding
-
-    struct common_params_model mparams_dft;
-
-    llama_model * model_dft = nullptr; // a llama_model that can be shared by multiple speculative contexts
-
-    llama_context_params cparams_dft; // these are the parameters for the draft llama_context
+    llama_context_params cparams; // these are the parameters for the draft llama_context
 
     int32_t n_ctx        = 0;  // draft context size
     int32_t n_gpu_layers = -1; // number of layers to store in VRAM for the draft model (-1 - use default)
@@ -332,25 +317,60 @@ struct common_params_speculative {
     ggml_type cache_type_k = GGML_TYPE_F16; // KV cache data type for the K
     ggml_type cache_type_v = GGML_TYPE_F16; // KV cache data type for the V
 
-    struct cpu_params cpuparams;
-    struct cpu_params cpuparams_batch;
+    common_cpu_params cpuparams;
+    common_cpu_params cpuparams_batch;
 
     std::vector<ggml_backend_dev_t> devices; // devices to use for offloading
 
     std::vector<std::pair<std::string, std::string>> replacements; // main to speculative model replacements
     std::vector<llama_model_tensor_buft_override> tensor_buft_overrides;
+};
+
+struct common_params_speculative_ngram_mod {
+    int32_t n_match = 24;
+
+    int32_t n_max = 64;
+    int32_t n_min = 48;
+
+    // shared instance of the ngram container for all speculative decoding contexts
+    std::shared_ptr<common_ngram_mod> obj;
+};
+
+struct common_params_speculative_ngram_map {
+    uint16_t size_n   = 12; // ngram size for lookup
+    uint16_t size_m   = 48; // mgram size for speculative tokens
+    uint16_t min_hits = 1;  // minimum hits at ngram/mgram lookup for mgram to be proposed
+};
+
+struct common_params_speculative_ngram_cache {
+    std::string lookup_cache_static;  // path of static ngram cache file for lookup decoding
+    std::string lookup_cache_dynamic; // path of dynamic ngram cache file for lookup decoding
+};
+
+struct common_params_speculative {
+    // TODO: become a vector in order to support "chains of speculators"
+    common_speculative_type type = COMMON_SPECULATIVE_TYPE_NONE;
+
+    common_params_speculative_draft draft;
+
+    common_params_speculative_ngram_mod ngram_mod;
+    common_params_speculative_ngram_map ngram_simple;
+    common_params_speculative_ngram_map ngram_map_k;
+    common_params_speculative_ngram_map ngram_map_k4v;
+
+    common_params_speculative_ngram_cache ngram_cache;
 
     bool has_dft() const {
-        return !mparams_dft.path.empty() || !mparams_dft.hf_repo.empty();
+        return !draft.mparams.path.empty() || !draft.mparams.hf_repo.empty();
     }
 };
 
 struct common_params_vocoder {
     struct common_params_model model;
 
-    std::string speaker_file = ""; // speaker file path                                      // NOLINT
+    std::string speaker_file; // speaker file path
 
-    bool use_guide_tokens = false; // enable guide tokens to improve TTS accuracy            // NOLINT
+    bool use_guide_tokens = false; // enable guide tokens to improve TTS accuracy
 };
 
 struct common_params_diffusion {
@@ -433,8 +453,8 @@ struct common_params {
 
     enum llama_split_mode split_mode = LLAMA_SPLIT_MODE_LAYER; // how to split the model across GPUs
 
-    struct cpu_params cpuparams;
-    struct cpu_params cpuparams_batch;
+    common_cpu_params cpuparams;
+    common_cpu_params cpuparams_batch;
 
     ggml_backend_sched_eval_callback cb_eval = nullptr;
     void * cb_eval_user_data                 = nullptr;
@@ -678,7 +698,7 @@ std::string common_params_get_system_info(const common_params & params);
 
 bool parse_cpu_range(const std::string & range, bool(&boolmask)[GGML_MAX_N_THREADS]);
 bool parse_cpu_mask(const std::string & mask, bool(&boolmask)[GGML_MAX_N_THREADS]);
-void postprocess_cpu_params(cpu_params & cpuparams, const cpu_params * role_model = nullptr);
+void postprocess_cpu_params(common_cpu_params & cpuparams, const common_cpu_params * role_model = nullptr);
 bool set_process_priority(enum ggml_sched_priority prio);
 
 //
@@ -846,7 +866,7 @@ common_init_result_ptr common_init_from_params(common_params & params);
 
 struct llama_model_params     common_model_params_to_llama  (      common_params & params);
 struct llama_context_params   common_context_params_to_llama(const common_params & params);
-struct ggml_threadpool_params ggml_threadpool_params_from_cpu_params(const cpu_params & params);
+struct ggml_threadpool_params ggml_threadpool_params_from_cpu_params(const common_cpu_params & params);
 
 // clear LoRA adapters from context, then apply new list of adapters
 void common_set_adapter_lora(struct llama_context * ctx, std::vector<common_adapter_lora_info> & lora);

common/debug.cpp

@@ -1,9 +1,38 @@
 #include "debug.h"
 
+#include "common.h"
 #include "log.h"
 
 #include <cmath>
+#include <regex>
 #include <string>
+#include <vector>
+
+struct common_debug_cb_user_data::impl {
+    std::vector<uint8_t>    data;
+    std::vector<std::regex> tensor_filters;
+    bool                    abort_on_nan{false};
+};
+
+common_debug_cb_user_data::common_debug_cb_user_data() : pimpl(std::make_unique<impl>()) {}
+common_debug_cb_user_data::~common_debug_cb_user_data() = default;
+
+common_debug_cb_user_data::common_debug_cb_user_data(common_params & params, const std::vector<std::string> & filter_patterns, bool abort_on_nan)
+    : pimpl(std::make_unique<impl>())
+{
+    for (const auto & pattern : filter_patterns) {
+        try {
+            std::string anchored_pattern = "^" + pattern;
+            pimpl->tensor_filters.emplace_back(anchored_pattern, std::regex::optimize);
+        } catch (const std::regex_error & e) {
+            throw std::runtime_error("Invalid regex pattern '" + pattern + "': " + e.what());
+        }
+    }
+    pimpl->abort_on_nan = abort_on_nan;
+
+    params.cb_eval           = common_debug_cb_eval;
+    params.cb_eval_user_data = this;
+}
 
 static std::string common_ggml_ne_string(const ggml_tensor * t) {
     std::string str;
@@ -47,8 +76,7 @@ static float common_ggml_get_float_value(const uint8_t * data,
 
 #define INDENT "    "
 
-template <bool abort>
-void common_debug_print_tensor(uint8_t * data, ggml_type type, const int64_t * ne, const size_t * nb, int64_t n) {
+static void common_debug_print_tensor(uint8_t * data, ggml_type type, const int64_t * ne, const size_t * nb, int64_t n, bool abort_on_nan) {
     GGML_ASSERT(n > 0);
     float sum = 0;
     for (int64_t i3 = 0; i3 < ne[3]; i3++) {
@@ -94,7 +122,7 @@ void common_debug_print_tensor(uint8_t * data, ggml_type type, const int64_t * n
         LOG(INDENT "sum = %f\n", sum);
     }
 
-    if constexpr (abort) {
+    if (abort_on_nan) {
         if (std::isnan(sum)) {
             LOG("encountered NaN - aborting\n");
             exit(0);
@@ -112,8 +140,9 @@ void common_debug_print_tensor(uint8_t * data, ggml_type type, const int64_t * n
  * @param user_data user data to pass at each call back
  * @return true to receive data or continue the graph, false otherwise
  */
-template <bool abort_on_nan> bool common_debug_cb_eval(struct ggml_tensor * t, bool ask, void * user_data) {
-    auto * cb_data = (base_callback_data *) user_data;
+bool common_debug_cb_eval(struct ggml_tensor * t, bool ask, void * user_data) {
+    auto * cb_data = (common_debug_cb_user_data *) user_data;
+    auto * pimpl = cb_data->pimpl.get();
 
     const struct ggml_tensor * src0 = t->src[0];
     const struct ggml_tensor * src1 = t->src[1];
@@ -122,10 +151,10 @@ template <bool abort_on_nan> bool common_debug_cb_eval(struct ggml_tensor * t, b
         return true;  // Always retrieve data
     }
 
-    bool matches_filter = cb_data->tensor_filters.empty();
+    bool matches_filter = pimpl->tensor_filters.empty();
 
     if (!matches_filter) {
-        for (const auto & filter : cb_data->tensor_filters) {
+        for (const auto & filter : pimpl->tensor_filters) {
             if (std::regex_search(t->name, filter)) {
                 matches_filter = true;
                 break;
@@ -148,20 +177,14 @@ template <bool abort_on_nan> bool common_debug_cb_eval(struct ggml_tensor * t, b
 
     if (!is_host) {
         auto n_bytes = ggml_nbytes(t);
-        cb_data->data.resize(n_bytes);
-        ggml_backend_tensor_get(t, cb_data->data.data(), 0, n_bytes);
+        pimpl->data.resize(n_bytes);
+        ggml_backend_tensor_get(t, pimpl->data.data(), 0, n_bytes);
     }
 
     if (!ggml_is_quantized(t->type) && matches_filter) {
-        uint8_t * data = is_host ? (uint8_t *) t->data : cb_data->data.data();
-        common_debug_print_tensor<abort_on_nan>(data, t->type, t->ne, t->nb, 3);
+        uint8_t * data = is_host ? (uint8_t *) t->data : pimpl->data.data();
+        common_debug_print_tensor(data, t->type, t->ne, t->nb, 3, pimpl->abort_on_nan);
     }
 
     return true;
 }
-
-// Explicit template instantiations
-template bool common_debug_cb_eval<false>(ggml_tensor *, bool, void *);
-template bool common_debug_cb_eval<true>(ggml_tensor *, bool, void *);
-template void common_debug_print_tensor<false>(uint8_t *, ggml_type, const int64_t *, const size_t *, int64_t);
-template void common_debug_print_tensor<true>(uint8_t *, ggml_type, const int64_t *, const size_t *, int64_t);

common/debug.h

@@ -1,43 +1,31 @@
 #pragma once
-#include "common.h"
+
+#include <memory>
 #include <string>
 #include <vector>
-#include <regex>
 
 // common debug functions and structs
 
-// Print a tensor's detailed data
-// data - the tensor's data in byte format
-// type - the tensor's quantization type
-// ne   - the tensor dimensions array
-// nb   - the tensor strides array
-// n    - the number of rows/columns to fully print
-template <bool abort_on_nan> void common_debug_print_tensor(uint8_t * data, ggml_type type, const int64_t * ne, const size_t * nb, int64_t n);
+struct common_params;
 
 // Intended to use as callback for ggml_backend_sched_eval_callback
 // prints tensors that are processed in the computation graph
-// by default prints all tensors, but can be configured by creating a `base_callback_data` instance with
-// non-empty filter_patterns. See examples/debug.ccp for possible usage patterns
-// The template parameter determines whether an error should be thrown whenever a NaN is encountered
+// by default prints all tensors, but can be configured by creating a `common_debug_cb_user_data` instance with
+// non-empty filter_patterns. See examples/debug.cpp for possible usage patterns
+// `common_debug_cb_user_data` contains `abort_on_nan` flag that determines whether an error should be thrown whenever a NaN is encountered
 // in a tensor (useful for stopping debug sessions on first erroneous tensor)
 // The callback data will be passed as the third parameter (user_data)
-template <bool abort_on_nan> bool common_debug_cb_eval(struct ggml_tensor * t, bool ask, void * user_data);
-struct base_callback_data {
-    std::vector<uint8_t>    data;
-    std::vector<std::regex> tensor_filters;
+bool common_debug_cb_eval(struct ggml_tensor * t, bool ask, void * user_data);
 
-    base_callback_data() = default;
+struct common_debug_cb_user_data {
+    struct impl;
+    std::unique_ptr<impl> pimpl;
 
-    base_callback_data(common_params & params, const std::vector<std::string> & filter_patterns) {
-        for (const auto & pattern : filter_patterns) {
-            try {
-                std::string anchored_pattern = "^" + pattern;
-                tensor_filters.emplace_back(anchored_pattern, std::regex::optimize);
-            } catch (const std::regex_error & e) {
-                throw std::runtime_error("Invalid regex pattern '" + pattern + "': " + e.what());
-            }
-        }
-        params.cb_eval           = common_debug_cb_eval<false>;
-        params.cb_eval_user_data = this;
-    }
+    common_debug_cb_user_data();
+    ~common_debug_cb_user_data();
+
+    common_debug_cb_user_data(const common_debug_cb_user_data &) = delete;
+    common_debug_cb_user_data & operator=(const common_debug_cb_user_data &) = delete;
+
+    common_debug_cb_user_data(common_params & params, const std::vector<std::string> & filter_patterns, bool abort_on_nan = false);
 };

common/download.cpp

@@ -627,7 +627,7 @@ static hf_cache::hf_file find_best_model(const hf_cache::hf_files & files,
     if (!tag.empty()) {
         tags.push_back(tag);
     } else {
-        tags = {"Q4_K_M", "Q4_0"};
+        tags = {"Q4_K_M", "Q8_0"};
     }
 
     for (const auto & t : tags) {

common/fit.cpp

@@ -856,7 +856,7 @@ void common_memory_breakdown_print(const struct llama_context * ctx) {
         ggml_backend_dev_memory(dev, &free, &total);
 
         const size_t self = mb.model + mb.context + mb.compute;
-        const size_t unaccounted = total - self - free;
+        const int64_t unaccounted = static_cast<int64_t>(total) - static_cast<int64_t>(free) - static_cast<int64_t>(self);
 
         table_data.push_back({
             template_gpu,
@@ -867,7 +867,7 @@ void common_memory_breakdown_print(const struct llama_context * ctx) {
             std::to_string(mb.model / MiB),
             std::to_string(mb.context / MiB),
             std::to_string(mb.compute / MiB),
-            std::to_string(unaccounted / MiB)});
+            std::to_string(unaccounted / static_cast<int64_t>(MiB))});
     }
 
     // print memory breakdown for host:

common/hf-cache.cpp

@@ -57,7 +57,7 @@ static fs::path get_cache_directory() {
 #ifndef _WIN32
         const struct passwd * pw = getpwuid(getuid());
 
-        if (pw->pw_dir && *pw->pw_dir) {
+        if (pw && pw->pw_dir && *pw->pw_dir) {
             return fs::path(pw->pw_dir) / ".cache" / "huggingface" / "hub";
         }
 #endif

common/log.cpp

@@ -49,7 +49,7 @@ enum common_log_col : int {
 };
 
 // disable colors by default
-static std::vector<const char *> g_col = {
+static const char* g_col[] = {
     "",
     "",
     "",
@@ -247,7 +247,6 @@ public:
 
             entries = std::move(new_entries);
         }
-
         cv.notify_one();
     }
 
@@ -265,7 +264,6 @@ public:
                 {
                     std::unique_lock<std::mutex> lock(mtx);
                     cv.wait(lock, [this]() { return head != tail; });
-
                     cur = entries[head];
 
                     head = (head + 1) % entries.size();
@@ -301,7 +299,6 @@ public:
 
                 tail = (tail + 1) % entries.size();
             }
-
             cv.notify_one();
         }
 
@@ -338,7 +335,7 @@ public:
             g_col[COMMON_LOG_COL_CYAN]    = LOG_COL_CYAN;
             g_col[COMMON_LOG_COL_WHITE]   = LOG_COL_WHITE;
         } else {
-            for (size_t i = 0; i < g_col.size(); i++) {
+            for (size_t i = 0; i < std::size(g_col); i++) {
                 g_col[i] = "";
             }
         }
@@ -368,14 +365,20 @@ struct common_log * common_log_init() {
 }
 
 struct common_log * common_log_main() {
-    static struct common_log log;
+    // We intentionally leak (i.e. do not delete) the logger singleton because
+    // common_log destructor called at DLL teardown phase will cause hanging on Windows.
+    // OS will release resources anyway so it should not be a significant issue,
+    // though this design may cause logs to be lost if not flushed before the program exits.
+    // Refer to https://github.com/ggml-org/llama.cpp/issues/22142 for details.
+    static struct common_log * log;
     static std::once_flag    init_flag;
     std::call_once(init_flag, [&]() {
+        log = new common_log;
         // Set default to auto-detect colors
-        log.set_colors(tty_can_use_colors());
+        log->set_colors(tty_can_use_colors());
     });
 
-    return &log;
+    return log;
 }
 
 void common_log_pause(struct common_log * log) {

common/log.h

@@ -49,7 +49,11 @@ void common_log_default_callback(enum ggml_log_level level, const char * text, v
 struct common_log;
 
 struct common_log * common_log_init();
-struct common_log * common_log_main(); // singleton, automatically destroys itself on exit
+
+// Singleton, intentionally leaked to avoid Windows teardown hangs.
+// Call common_log_flush() before exit if you want to ensure all logs are flushed.
+struct common_log * common_log_main();
+
 void                common_log_pause (struct common_log * log); // pause  the worker thread, not thread-safe
 void                common_log_resume(struct common_log * log); // resume the worker thread, not thread-safe
 void                common_log_free  (struct common_log * log);

common/preset.cpp

@@ -43,7 +43,7 @@ static std::set<std::string> get_remote_preset_whitelist(const std::map<std::str
     for (const auto & it : key_to_opt) {
         const std::string & key = it.first;
         const common_arg & opt = it.second;
-        if (allowed_options.find(key) != allowed_options.end() || opt.is_sparam) {
+        if (allowed_options.find(key) != allowed_options.end() || opt.is_sampling) {
             allowed_keys.insert(key);
             // also add variant keys (args without leading dashes and env vars)
             for (const auto & arg : opt.get_args()) {

common/reasoning-budget.cpp

@@ -122,6 +122,20 @@ static void common_reasoning_budget_accept(struct llama_sampler * smpl, llama_to
             }
             break;
         case REASONING_BUDGET_DONE:
+            // Re-arm on a new start tag: some models emit multiple <think> blocks
+            // per response, and each should get a fresh budget window.
+            if (ctx->start_matcher.advance(token)) {
+                ctx->state = REASONING_BUDGET_COUNTING;
+                ctx->remaining = ctx->budget;
+                ctx->end_matcher.reset();
+                LOG_INF("reasoning-budget: re-activated on new start tag, budget=%d tokens\n", ctx->budget);
+
+                if (ctx->remaining <= 0) {
+                    ctx->state = REASONING_BUDGET_FORCING;
+                    ctx->force_pos = 0;
+                    LOG_INF("reasoning-budget: budget=0, forcing immediately\n");
+                }
+            }
             break;
     }
 }
@@ -218,34 +232,6 @@ static struct llama_sampler * common_reasoning_budget_init_state(
     );
 }
 
-struct llama_sampler * common_reasoning_budget_init(
-        const struct llama_vocab       * vocab,
-        const std::vector<llama_token> & start_tokens,
-        const std::vector<llama_token> & end_tokens,
-        const std::vector<llama_token> & forced_tokens,
-        int32_t                          budget,
-        const std::vector<llama_token> & prefill_tokens) {
-    // Determine initial state from prefill: COUNTING if the prefill begins with
-    // the start sequence but does not also contain the end sequence after it.
-    common_reasoning_budget_state initial_state = REASONING_BUDGET_IDLE;
-    if (!prefill_tokens.empty() && !start_tokens.empty() &&
-            prefill_tokens.size() >= start_tokens.size() &&
-            std::equal(start_tokens.begin(), start_tokens.end(), prefill_tokens.begin())) {
-        initial_state = REASONING_BUDGET_COUNTING;
-        // If the end sequence also follows the start in the prefill, reasoning
-        // was opened and immediately closed — stay IDLE.
-        if (!end_tokens.empty() &&
-                prefill_tokens.size() >= start_tokens.size() + end_tokens.size()) {
-            auto end_start = prefill_tokens.end() - (ptrdiff_t) end_tokens.size();
-            if (end_start >= prefill_tokens.begin() + (ptrdiff_t) start_tokens.size() &&
-                    std::equal(end_tokens.begin(), end_tokens.end(), end_start)) {
-                initial_state = REASONING_BUDGET_IDLE;
-            }
-        }
-    }
-    return common_reasoning_budget_init_state(vocab, start_tokens, end_tokens, forced_tokens, budget, initial_state);
-}
-
 struct llama_sampler * common_reasoning_budget_init(
         const struct llama_vocab       * vocab,
         const std::vector<llama_token> & start_tokens,

common/reasoning-budget.h

@@ -29,10 +29,7 @@ enum common_reasoning_budget_state {
 //   end_tokens     - token sequence for natural deactivation
 //   forced_tokens  - token sequence forced when budget expires
 //   budget         - max tokens allowed in the reasoning block
-//   prefill_tokens - tokens already present in the prompt (generation prompt);
-//                    used to determine the initial state: COUNTING if they begin
-//                    with start_tokens (but don't also end with end_tokens),
-//                    IDLE otherwise. COUNTING with budget <= 0 is promoted to FORCING.
+//   initial_state  - initial state
 //
 struct llama_sampler * common_reasoning_budget_init(
         const struct llama_vocab       * vocab,
@@ -40,16 +37,6 @@ struct llama_sampler * common_reasoning_budget_init(
         const std::vector<llama_token> & end_tokens,
         const std::vector<llama_token> & forced_tokens,
         int32_t                          budget,
-        const std::vector<llama_token> & prefill_tokens = {});
-
-// Variant that takes an explicit initial state (used by tests and clone).
-// COUNTING with budget <= 0 is promoted to FORCING.
-struct llama_sampler * common_reasoning_budget_init(
-        const struct llama_vocab       * vocab,
-        const std::vector<llama_token> & start_tokens,
-        const std::vector<llama_token> & end_tokens,
-        const std::vector<llama_token> & forced_tokens,
-        int32_t                          budget,
-        common_reasoning_budget_state    initial_state);
+        common_reasoning_budget_state    initial_state = REASONING_BUDGET_IDLE);
 
 common_reasoning_budget_state common_reasoning_budget_get_state(const struct llama_sampler * smpl);

common/sampling.cpp

@@ -260,32 +260,35 @@ struct common_sampler * common_sampler_init(const struct llama_model * model, st
         }
     }
 
+    // Compute prefill tokens from the generation prompt
+    std::vector<llama_token> prefill_tokens;
+    if (!params.generation_prompt.empty()) {
+        GGML_ASSERT(vocab != nullptr);
+        auto tokens = common_tokenize(vocab, params.generation_prompt, false, true);
+        for (size_t i = 0; i < tokens.size(); i++) {
+            std::string piece = common_token_to_piece(vocab, tokens[i], true);
+            if (i == 0 && std::isspace(piece[0]) && !std::isspace(params.generation_prompt[0])) {
+                // Some tokenizers will add a space before the first special token, need to exclude
+                continue;
+            }
+            LOG_DBG("%s: prefill token: %d = %s\n", __func__, tokens[i], piece.c_str());
+            prefill_tokens.push_back(tokens[i]);
+        }
+    }
+
     // Feed generation prompt tokens to the grammar sampler so it advances past
     // tokens the template already placed in the prompt.
     // Only applies to output-format and tool-call grammars; user-supplied grammars must not be prefilled.
-    std::vector<llama_token> prefill_tokens;
-    if (!params.generation_prompt.empty() && common_grammar_needs_prefill(params.grammar)) {
-        GGML_ASSERT(vocab != nullptr);
-        prefill_tokens = common_tokenize(vocab, params.generation_prompt, false, true);
-        if (!prefill_tokens.empty()) {
-            std::string first_token = common_token_to_piece(vocab, prefill_tokens[0], true);
-            if (std::isspace(first_token[0]) && !std::isspace(params.generation_prompt[0])) {
-                // Some tokenizers will add a space before the first special token, need to remove
-                prefill_tokens = std::vector<llama_token>(prefill_tokens.begin() + 1, prefill_tokens.end());
-            }
-        }
-
-        if (grmr && !params.grammar_lazy) {
-            try {
-                for (const auto & token : prefill_tokens) {
-                    llama_sampler_accept(grmr, token);
-                    LOG_DBG("%s: accepted prefill token (%d)\n", __func__, token);
-                }
-            } catch (std::exception &e) {
-                LOG_ERR("%s: error initializing grammar sampler for grammar:\n%s\n\nGeneration prompt:\n'%s'\n", __func__,
-                    common_grammar_value(params.grammar).c_str(), params.generation_prompt.c_str());
-                throw e;
+    if (grmr && !params.grammar_lazy && common_grammar_needs_prefill(params.grammar)) {
+        try {
+            for (const auto & token : prefill_tokens) {
+                llama_sampler_accept(grmr, token);
+                LOG_DBG("%s: grammar accepted prefill token (%d)\n", __func__, token);
             }
+        } catch (std::exception &e) {
+            LOG_ERR("%s: error initializing grammar sampler for grammar:\n%s\n\nGeneration prompt:\n'%s'\n", __func__,
+                common_grammar_value(params.grammar).c_str(), params.generation_prompt.c_str());
+            throw e;
         }
     }
 
@@ -296,8 +299,12 @@ struct common_sampler * common_sampler_init(const struct llama_model * model, st
             params.reasoning_budget_start,
             params.reasoning_budget_end,
             params.reasoning_budget_forced,
-            params.reasoning_budget_tokens < 0 ? INT_MAX : params.reasoning_budget_tokens,
-            prefill_tokens);
+            params.reasoning_budget_tokens < 0 ? INT_MAX : params.reasoning_budget_tokens);
+
+        for (const auto & token : prefill_tokens) {
+            llama_sampler_accept(rbudget, token);
+            LOG_DBG("%s: reasoning-budget accepted prefill token (%d)\n", __func__, token);
+        }
     }
 
     if (params.has_logit_bias()) {
@@ -431,7 +438,7 @@ static bool grammar_should_apply(struct common_sampler * gsmpl) {
     return true;
 }
 
-void common_sampler_accept(struct common_sampler * gsmpl, llama_token token, bool accept_grammar) {
+void common_sampler_accept(struct common_sampler * gsmpl, llama_token token, bool is_generated) {
     if (!gsmpl) {
         return;
     }
@@ -439,9 +446,11 @@ void common_sampler_accept(struct common_sampler * gsmpl, llama_token token, boo
     const auto tm = gsmpl->tm();
 
     // grammar_should_apply() checks the reasoning budget state, so calculate this before we accept
-    accept_grammar = accept_grammar && grammar_should_apply(gsmpl);
+    const auto accept_grammar = is_generated && grammar_should_apply(gsmpl);
 
-    llama_sampler_accept(gsmpl->rbudget, token);
+    if (gsmpl->rbudget && is_generated) {
+        llama_sampler_accept(gsmpl->rbudget, token);
+    }
 
     if (gsmpl->grmr && accept_grammar) {
         llama_sampler_accept(gsmpl->grmr, token);

common/sampling.h

@@ -41,8 +41,8 @@ struct common_sampler * common_sampler_init(const struct llama_model * model, st
 
 void common_sampler_free(struct common_sampler * gsmpl);
 
-// if accept_grammar is true, the token is accepted both by the sampling chain and the grammar
-void                    common_sampler_accept(struct common_sampler * gsmpl, llama_token token, bool accept_grammar);
+// if is_generated is true, the token is accepted by the sampling chain, the reasoning budget sampler, and the grammar sampler
+void                    common_sampler_accept(struct common_sampler * gsmpl, llama_token token, bool is_generated);
 void                    common_sampler_reset (struct common_sampler * gsmpl);
 struct common_sampler * common_sampler_clone (struct common_sampler * gsmpl);
 

common/speculative.cpp

@@ -151,6 +151,9 @@ struct common_speculative_state {
             llama_tokens & result) = 0;
 
     virtual void accept(uint16_t n_accepted) = 0;
+
+    virtual int32_t n_max(const common_params_speculative & params) const = 0;
+    virtual int32_t n_min(const common_params_speculative & params) const = 0;
 };
 
 struct common_speculative_checkpoint {
@@ -164,8 +167,6 @@ struct common_speculative_checkpoint {
     size_t size() const {
         return data.size();
     }
-
-    size_t ckpt_size   = 0;
 };
 
 struct common_speculative_state_draft : public common_speculative_state {
@@ -173,7 +174,7 @@ struct common_speculative_state_draft : public common_speculative_state {
     llama_context * ctx_dft;
 
     bool use_ckpt = false;
-    struct common_speculative_checkpoint ckpt;
+    common_speculative_checkpoint ckpt;
 
     common_sampler * smpl;
 
@@ -246,26 +247,16 @@ struct common_speculative_state_draft : public common_speculative_state {
         llama_batch_free(batch);
     }
 
-    void begin(const llama_tokens & prompt) override {
-        if (use_ckpt && ckpt.size() > 0) {
-            // delete checkpoint
-            LOG_DBG("%s: delete checkpoint, prompt.size=%zu, pos_min=%d, pos_max=%d, n_tokens=%" PRId64 ", size=%.3f MiB\n",
-                    __func__, prompt.size(), ckpt.pos_min, ckpt.pos_max, ckpt.n_tokens, (float) ckpt.data.size() / 1024 / 1024);
-            ckpt.pos_min   = 0;
-            ckpt.pos_max   = 0;
-            ckpt.n_tokens  = 0;
-            ckpt.ckpt_size = 0;
-            ckpt.data.clear();
-        }
+    void begin(const llama_tokens & /*prompt*/) override {
     }
 
-    size_t draft_create_checkpoint(int n_tokens_prompt, int n_tokens_batch) {
+    size_t create_checkpoint(int n_tokens_prompt) {
         int slot_id = 0;
         const size_t checkpoint_size = llama_state_seq_get_size_ext(ctx_dft, slot_id, LLAMA_STATE_SEQ_FLAGS_PARTIAL_ONLY);
 
         ckpt.pos_min  = llama_memory_seq_pos_min(llama_get_memory(ctx_dft), slot_id);
         ckpt.pos_max  = llama_memory_seq_pos_max(llama_get_memory(ctx_dft), slot_id);
-        ckpt.n_tokens = n_tokens_prompt - n_tokens_batch;
+        ckpt.n_tokens = n_tokens_prompt;
         ckpt.data.resize(checkpoint_size);
 
         const size_t n = llama_state_seq_get_data_ext(ctx_dft, ckpt.data.data(), checkpoint_size, slot_id, LLAMA_STATE_SEQ_FLAGS_PARTIAL_ONLY);
@@ -278,13 +269,13 @@ struct common_speculative_state_draft : public common_speculative_state {
         return n;
     }
 
-    size_t draft_restore_checkpoint(size_t ckpt_size_part_expected) {
+    size_t restore_checkpoint() {
         int slot_id = 0;
         LOG_DBG("%s: pos_min = %d, pos_max = %d\n", __func__, ckpt.pos_min, ckpt.pos_max);
         const size_t n = llama_state_seq_set_data_ext(ctx_dft, ckpt.data.data(), ckpt.size(), slot_id, LLAMA_STATE_SEQ_FLAGS_PARTIAL_ONLY);
-        if (n != ckpt_size_part_expected) {
-            GGML_ABORT("%s: failed to restore context checkpoint (pos_min=%d, pos_max=%d, size=%zu, get_data_ext->%zu, set_data_ext->%zu",
-                        __func__, ckpt.pos_min, ckpt.pos_max, ckpt.size(), ckpt_size_part_expected, n);
+        if (n != ckpt.size()) {
+            GGML_ABORT("%s: failed to restore context checkpoint (pos_min=%d, pos_max=%d, size=%zu",
+                        __func__, ckpt.pos_min, ckpt.pos_max, ckpt.size());
         }
         llama_memory_seq_rm(llama_get_memory(ctx_dft), slot_id, ckpt.pos_max + 1, -1);
 
@@ -296,6 +287,8 @@ struct common_speculative_state_draft : public common_speculative_state {
             const llama_tokens & prompt_tgt,
             llama_token id_last,
             llama_tokens & result) override {
+        const auto & sparams = params.draft;
+
         auto * spec = this;
 
         auto & batch      = spec->batch;
@@ -309,7 +302,7 @@ struct common_speculative_state_draft : public common_speculative_state {
         int reuse_i = 0; // index of part to be reused in prompt_dft
         int reuse_n = 0; // length of part to be reused in prompt_dft
 
-        const int n_ctx = llama_n_ctx(ctx_dft) - params.n_max;
+        const int n_ctx = llama_n_ctx(ctx_dft) - sparams.n_max;
 
         llama_tokens prompt_cnv;
         if (!spec->vocab_cmpt) {
@@ -341,13 +334,18 @@ struct common_speculative_state_draft : public common_speculative_state {
 
         const int i_start = std::max<int>(0, (int) prompt_cur.size() - n_ctx);
 
+        if (use_ckpt && i_start > 0) {
+            LOG_WRN("%s: context shift is not supported with checkpoint-based contexts - skipping\n", __func__);
+            return;
+        }
+
         // reuse as much as possible from the old draft context
         // ideally, the draft context should be as big as the target context and we will always reuse the entire prompt
         for (int i = 0; i < (int) prompt_dft.size(); ++i) {
             int cur = 0;
             while (i_start + cur < (int) prompt_cur.size() &&
-                    i       + cur < (int) prompt_dft.size() &&
-                    prompt_cur[i_start + cur] == prompt_dft[i + cur]) {
+                   i       + cur < (int) prompt_dft.size() &&
+                   prompt_cur[i_start + cur] == prompt_dft[i + cur]) {
                 cur++;
             }
 
@@ -355,21 +353,26 @@ struct common_speculative_state_draft : public common_speculative_state {
                 reuse_i = i;
                 reuse_n = cur;
             }
+
+            if (use_ckpt) {
+                break;
+            }
         }
 
         LOG_DBG("%s: reuse_i = %d, reuse_n = %d, #prompt_dft = %zu, #prompt_cur = %zu\n",
                 __func__, reuse_i, reuse_n, prompt_dft.size(), prompt_cur.size());
-        if (use_ckpt && ckpt.ckpt_size == 0 && reuse_n > 0) {
-            LOG_DBG("%s: no checkpoint available, no reuse, (reuse_i=%d, reuse_n=%d) -> (0, 0)\n",
-                    __func__, reuse_i, reuse_n);
+        if (use_ckpt && ckpt.n_tokens > reuse_n) {
+            LOG_DBG("%s: checkpoint (n_tokens = %d) is outdated -> delete it\n", __func__, (int) ckpt.n_tokens);
+
             reuse_i = 0;
             reuse_n = 0;
+
+            ckpt = {};
         }
 
         result.clear();
-        result.reserve(params.n_max);
+        result.reserve(sparams.n_max);
 
-        bool needs_ckpt = use_ckpt && prompt_dft.size() > 0;
         if (reuse_n == 0 || (use_ckpt && reuse_i > 0)) {
             llama_memory_clear(mem_dft, false);
             prompt_dft.clear();
@@ -380,7 +383,7 @@ struct common_speculative_state_draft : public common_speculative_state {
                 for (int i = reuse_i + reuse_n + 1; i < (int) prompt_dft.size(); ++i) {
                     result.push_back(prompt_dft[i]);
 
-                    if (params.n_max <= (int) result.size()) {
+                    if (sparams.n_max <= (int) result.size()) {
                         break;
                     }
                 }
@@ -388,50 +391,38 @@ struct common_speculative_state_draft : public common_speculative_state {
                 return;
             }
 
-            bool do_restore = false;
-            if (prompt_dft.size() > prompt_cur.size() && reuse_i + reuse_n < (int64_t) prompt_dft.size()) {
-                // This can happen after a partial acceptance (speculative decoding with checkpoints)
-                LOG_DBG("%s: #prompt_dft=%zu, #prompt_cur=%zu, shorten draft\n",
-                        __func__, prompt_dft.size(), prompt_cur.size());
-                prompt_dft.resize(prompt_cur.size());
-                do_restore = true;
-            }
-
             if (reuse_i > 0) {
+                GGML_ASSERT(!use_ckpt);
+
                 bool is_removed = llama_memory_seq_rm (mem_dft, 0, 0, reuse_i);
                 if (!is_removed) {
                     LOG_ERR("%s: llama_memory_seq_rm failed, reuse_i=%d\n", __func__, reuse_i);
+                    return;
                 }
                 llama_memory_seq_add(mem_dft, 0, reuse_i, -1, -reuse_i);
 
                 prompt_dft.erase(prompt_dft.begin(), prompt_dft.begin() + reuse_i);
             }
 
-            if (reuse_n < (int) prompt_dft.size() || do_restore) {
+            if (reuse_n < (int) prompt_dft.size()) {
                 if (use_ckpt) {
-                    if (ckpt.n_tokens > (int64_t) prompt_dft.size()) {
-                        LOG_INF("%s: checkpoint is too large, prompt_tgt.size=%zu, ckpt.n_tokens=%" PRId64 ", reuse_n=%d, prompt_dft.size=%zu\n",
-                                __func__, prompt_tgt.size(), ckpt.n_tokens, reuse_n, prompt_dft.size());
+                    if (ckpt.n_tokens > 0) {
+                        LOG_DBG("%s: restoring checkpoint, reuse_n=%d, prompt_dft.size=%zu\n", __func__, reuse_n, prompt_dft.size());
+                        restore_checkpoint();
+                        reuse_n = ckpt.n_tokens;
+                        prompt_dft.resize(reuse_n);
                     }
-                    draft_restore_checkpoint(ckpt.ckpt_size);
-                    reuse_n = ckpt.n_tokens;
-                    prompt_dft.resize(reuse_n);
-                    needs_ckpt = false;
                 } else {
-                    bool is_removed = llama_memory_seq_rm (mem_dft, 0, reuse_n, -1);
+                    const bool is_removed = llama_memory_seq_rm(mem_dft, 0, reuse_n, -1);
                     if (!is_removed) {
-                        LOG_ERR("%s: llama_memory_seq_rm failed, reuse_n=%d, prompt_dft.size=%zu\n",
-                                __func__, reuse_n, prompt_dft.size());
+                        LOG_ERR("%s: llama_memory_seq_rm failed, reuse_n=%d, prompt_dft.size=%zu\n", __func__, reuse_n, prompt_dft.size());
+                        return;
                     }
                     prompt_dft.erase(prompt_dft.begin() + reuse_n, prompt_dft.end());
                 }
             }
         }
 
-        if (needs_ckpt) {
-            ckpt.ckpt_size = draft_create_checkpoint(prompt_dft.size(), batch.n_tokens);
-        }
-
         // prepare a batch to evaluate any new tokens in the prompt
         common_batch_clear(batch);
 
@@ -445,12 +436,17 @@ struct common_speculative_state_draft : public common_speculative_state {
         // we should rarely end-up here during normal decoding
         if (batch.n_tokens > 0) {
             //LOG_DBG("%s: draft prompt batch: %s\n", __func__, string_from(ctx, batch).c_str());
+            LOG_DBG("%s: draft prompt batch: %d tokens\n", __func__, batch.n_tokens);
 
             int ret = llama_decode(ctx_dft, batch);
             if (ret != 0 && ret != 1) {
                 LOG_WRN("%s: llama_decode returned %d, prompt_cur.size=%zu\n",
                         __func__, ret, prompt_cur.size());
             }
+
+            if (use_ckpt) {
+                create_checkpoint(prompt_dft.size());
+            }
         }
 
         const llama_pos n_past = prompt_dft.size();
@@ -462,7 +458,7 @@ struct common_speculative_state_draft : public common_speculative_state {
 
         prompt_dft.push_back(id_last);
 
-        LOG_DBG("%s: draft prompt: %s\n", __func__, string_from(ctx_dft, prompt_dft).c_str());
+        //LOG_DBG("%s: draft prompt: %s\n", __func__, string_from(ctx_dft, prompt_dft).c_str());
 
         int ret = llama_decode(ctx_dft, batch);
         if (ret != 0 && ret != 1) {
@@ -473,7 +469,7 @@ struct common_speculative_state_draft : public common_speculative_state {
         common_sampler_reset(smpl);
 
         // sample n_draft tokens from the draft model
-        for (int i = 0; i < params.n_max; ++i) {
+        for (int i = 0; i < sparams.n_max; ++i) {
             common_batch_clear(batch);
 
             common_sampler_sample(smpl, ctx_dft, 0, true);
@@ -490,14 +486,14 @@ struct common_speculative_state_draft : public common_speculative_state {
 
             common_sampler_accept(smpl, id, true);
 
-            result.push_back(id);
-
-            if (params.n_max <= (int) result.size()) {
+            // only collect very high-confidence draft tokens
+            if (cur_p->data[0].p < sparams.p_min) {
                 break;
             }
 
-            // only collect very high-confidence draft tokens
-            if (cur_p->data[0].p < params.p_min) {
+            result.push_back(id);
+
+            if (sparams.n_max <= (int) result.size()) {
                 break;
             }
 
@@ -518,10 +514,14 @@ struct common_speculative_state_draft : public common_speculative_state {
             detokenized = replace_to_tgt(detokenized);
             LOG_DBG("draft->main detokenized string: '%s'\n", detokenized.c_str());
             result = common_tokenize(ctx_tgt, detokenized, false, true);
-            if (result.size() > (size_t)params.n_max) {
-                result.resize(params.n_max);
+            if (result.size() > (size_t) sparams.n_max) {
+                result.resize(sparams.n_max);
             }
         }
+
+        if (result.size() < (size_t) sparams.n_min) {
+            result.clear();
+        }
     }
 
     void accept(uint16_t n_accepted) override {
@@ -529,6 +529,14 @@ struct common_speculative_state_draft : public common_speculative_state {
         GGML_UNUSED(n_accepted);
     }
 
+    int32_t n_max(const common_params_speculative & params) const override {
+        return params.draft.n_max;
+    }
+
+    int32_t n_min(const common_params_speculative & params) const override {
+        return params.draft.n_min;
+    }
+
     std::string replace_to_dft(const std::string & input) const {
         std::string result = input;
 
@@ -581,6 +589,14 @@ struct common_speculative_state_eagle3 : public common_speculative_state {
         // noop
         GGML_UNUSED(n_accepted);
     }
+
+    int32_t n_max(const common_params_speculative & params) const override {
+        return params.draft.n_max;
+    }
+
+    int32_t n_min(const common_params_speculative & params) const override {
+        return params.draft.n_min;
+    }
 };
 
 // state of self-speculation (simple implementation, not ngram-map)
@@ -610,19 +626,27 @@ struct common_speculative_state_ngram_simple : public common_speculative_state {
         // noop
         GGML_UNUSED(n_accepted);
     }
+
+    int32_t n_max(const common_params_speculative & /*params*/) const override {
+        return config.size_mgram;
+    }
+
+    int32_t n_min(const common_params_speculative & /*params*/) const override {
+        return config.size_mgram;
+    }
 };
 
 struct common_speculative_state_ngram_map_k : public common_speculative_state {
     // draft ngram map for speculative decoding without draft model
-    common_ngram_map map;
+    common_ngram_map config;
 
     common_speculative_state_ngram_map_k(
             enum common_speculative_type type,
-            common_ngram_map map)
-        : common_speculative_state(type), map(std::move(map)) {}
+            common_ngram_map config)
+        : common_speculative_state(type), config(std::move(config)) {}
 
     void begin(const llama_tokens & prompt) override {
-        common_ngram_map_begin(map, prompt);
+        common_ngram_map_begin(config, prompt);
     }
 
     void draft(
@@ -630,12 +654,20 @@ struct common_speculative_state_ngram_map_k : public common_speculative_state {
             const llama_tokens & prompt_tgt,
             llama_token id_last,
             llama_tokens & result) override {
-        common_ngram_map_draft(map, prompt_tgt, id_last, result);
+        common_ngram_map_draft(config, prompt_tgt, id_last, result);
         GGML_UNUSED(params);
     }
 
     void accept(uint16_t n_accepted) override {
-        common_ngram_map_accept(map, n_accepted);
+        common_ngram_map_accept(config, n_accepted);
+    }
+
+    int32_t n_max(const common_params_speculative & /*params*/) const override {
+        return config.size_value;
+    }
+
+    int32_t n_min(const common_params_speculative & /*params*/) const override {
+        return config.size_value;
     }
 };
 
@@ -692,7 +724,7 @@ struct common_speculative_state_ngram_mod : public common_speculative_state {
             const llama_tokens & prompt_tgt,
             llama_token id_last,
             llama_tokens & result) override {
-        GGML_UNUSED(params);
+        const auto & sparams = params.ngram_mod;
 
         n_draft_last = 0;
 
@@ -712,16 +744,16 @@ struct common_speculative_state_ngram_mod : public common_speculative_state {
             i_last = cur_len - n;
         }
 
-        result.resize(n + params.n_max);
+        result.resize(n + sparams.n_max);
         for (size_t i = 0; i < n - 1; ++i) {
             result[i] = prompt_tgt[cur_len - n + 1 + i];
         }
         result[n - 1] = id_last;
 
-        for (int i = 0; i < params.n_max; ++i) {
+        for (int i = 0; i < sparams.n_max; ++i) {
             const llama_token token = mod.get(result.data() + i);
             if (token == common_ngram_mod::EMPTY) {
-                if (i < params.n_min) {
+                if (i < sparams.n_min) {
                     result.clear();
                     return;
                 }
@@ -743,17 +775,15 @@ struct common_speculative_state_ngram_mod : public common_speculative_state {
     }
 
     void accept(uint16_t n_accepted) override {
-        if (verbose) {
-            LOG_INF("%s: accepted %d tokens from %zu drafted tokens\n", __func__, n_accepted, n_draft_last);
-        }
-
         // compute acceptance fraction if we have a recorded draft length
         if (n_draft_last > 0) {
             const double f_acc = (double)n_accepted / (double)n_draft_last;
             if (f_acc < 0.5) {
                 n_low++;
                 if (n_low >= 3) {
-                    LOG_WRN("%s: low acceptance streak (%d) – resetting ngram_mod\n", __func__, n_low);
+                    if (verbose) {
+                        LOG_WRN("%s: low acceptance streak (%d) – resetting ngram_mod\n", __func__, n_low);
+                    }
 
                     mod.reset();
                     n_low = 0;
@@ -764,6 +794,14 @@ struct common_speculative_state_ngram_mod : public common_speculative_state {
             }
         }
     }
+
+    int32_t n_max(const common_params_speculative & params) const override {
+        return params.ngram_mod.n_max;
+    }
+
+    int32_t n_min(const common_params_speculative & params) const override {
+        return params.ngram_mod.n_min;
+    }
 };
 
 struct common_speculative_state_ngram_cache : public common_speculative_state {
@@ -857,6 +895,14 @@ struct common_speculative_state_ngram_cache : public common_speculative_state {
         // TODO: noop
         GGML_UNUSED(n_accepted);
     }
+
+    int32_t n_max(const common_params_speculative & /*params*/) const override {
+        return n_draft;
+    }
+
+    int32_t n_min(const common_params_speculative & /*params*/) const override {
+        return 0;
+    }
 };
 
 struct common_speculative {
@@ -865,11 +911,13 @@ struct common_speculative {
     common_speculative_state * curr_impl = nullptr; // current implementation in use (for stats)
 };
 
-static common_ngram_map get_common_ngram_map(const common_speculative_config & config) {
-    uint16_t size_key   = config.params.ngram_size_n;
-    uint16_t size_value = config.params.ngram_size_m;
-    bool     key_only   = (config.type == COMMON_SPECULATIVE_TYPE_NGRAM_MAP_K);
-    uint16_t min_hits   = config.params.ngram_min_hits;
+static common_ngram_map get_common_ngram_map(
+        common_speculative_type type,
+        const common_params_speculative_ngram_map & config) {
+    uint16_t size_key   = config.size_n;
+    uint16_t size_value = config.size_m;
+    bool     key_only   = type == COMMON_SPECULATIVE_TYPE_NGRAM_MAP_K;
+    uint16_t min_hits   = config.min_hits;
 
     return common_ngram_map(size_key, size_value, key_only, min_hits);
 }
@@ -927,8 +975,8 @@ common_speculative * common_speculative_init(
         common_params_speculative & params,
         llama_context             * ctx_tgt) {
     llama_context * ctx_dft = nullptr;
-    if (params.model_dft) {
-        ctx_dft = llama_init_from_model(params.model_dft, params.cparams_dft);
+    if (params.draft.model) {
+        ctx_dft = llama_init_from_model(params.draft.model, params.draft.cparams);
         if (ctx_dft == nullptr) {
             LOG_ERR("%s", "failed to create draft context\n");
             return nullptr;
@@ -938,7 +986,7 @@ common_speculative * common_speculative_init(
     // Compute the implementations to use based on the config and their order of preference
     std::vector<common_speculative_config> configs = {}; // list of speculative configs to try
     {
-        bool has_draft = !params.mparams_dft.path.empty();
+        bool has_draft = !params.draft.mparams.path.empty();
         bool has_draft_eagle3 = false; // TODO PR-18039: if params.speculative.eagle3
 
         bool has_ngram_cache   = (params.type == COMMON_SPECULATIVE_TYPE_NGRAM_CACHE);
@@ -961,16 +1009,17 @@ common_speculative * common_speculative_init(
             configs.push_back(common_speculative_config(COMMON_SPECULATIVE_TYPE_NGRAM_MAP_K4V, params));
         }
         if (has_ngram_mod) {
-            // shared instance for all speculative decoding contexts
-            if (!params.ngram_mod) {
-                params.ngram_mod = std::make_shared<common_ngram_mod>(params.ngram_size_n, 4*1024*1024);
+            auto & sparams = params.ngram_mod;
 
-                LOG_INF("%s: initialized ngram_mod with n=%d, size=%zu (%.3f MB)\n", __func__,
-                        params.ngram_size_n, params.ngram_mod->size(),
-                        (float)(params.ngram_mod->size_bytes())/1024/1024);
+            if (!sparams.obj) {
+                sparams.obj = std::make_shared<common_ngram_mod>(sparams.n_match, 4*1024*1024);
 
-                if (params.ngram_size_n < 16) {
-                    LOG_WRN("%s: ngram_mod n=%d is too small - poor quality is possible, see: https://github.com/ggml-org/llama.cpp/pull/19164\n", __func__, params.ngram_size_n);
+                LOG_INF("%s: initialized ngram_mod with n_match=%d, size=%zu (%.3f MB)\n", __func__,
+                        sparams.n_match, sparams.obj->size(), (float)(sparams.obj->size_bytes())/1024/1024);
+
+                if (sparams.n_match < 16) {
+                    LOG_WRN("%s: ngram_mod n_match=%d is too small - poor quality is possible, "
+                            "see: https://github.com/ggml-org/llama.cpp/pull/19164\n", __func__, sparams.n_match);
                 }
             }
 
@@ -1000,7 +1049,7 @@ common_speculative * common_speculative_init(
                 impls.push_back(std::make_unique<common_speculative_state_draft>(config.type,
                     /* .ctx_tgt      = */ ctx_tgt,
                     /* .ctx_dft      = */ ctx_dft,
-                    /* .replacements = */ params.replacements,
+                    /* .replacements = */ params.draft.replacements,
                     /* .use_ckpt     = */ use_ckpt
                 ));
                 break;
@@ -1010,18 +1059,18 @@ common_speculative * common_speculative_init(
                 break;
             }
             case COMMON_SPECULATIVE_TYPE_NGRAM_SIMPLE: {
-                common_ngram_map ngram_map = get_common_ngram_map(config);
+                common_ngram_map ngram_map = get_common_ngram_map(config.type, config.params.ngram_simple);
 
                 uint16_t ngram_size_key   = ngram_map.size_key;
                 uint16_t mgram_size_value = ngram_map.size_value;
 
                 auto config_simple = common_ngram_simple_config {
-                    /* .size_ngram      = */ ngram_size_key,
-                    /* .size_mgram      = */ mgram_size_value
+                    /* .size_ngram = */ ngram_size_key,
+                    /* .size_mgram = */ mgram_size_value
                 };
                 auto state = std::make_unique<common_speculative_state_ngram_simple>(
-                    /* .type            = */ config.type,
-                    /* .state           = */ config_simple
+                    /* .type  = */ config.type,
+                    /* .state = */ config_simple
                 );
                 impls.push_back(std::move(state));
                 break;
@@ -1030,18 +1079,17 @@ common_speculative * common_speculative_init(
             case COMMON_SPECULATIVE_TYPE_NGRAM_MAP_K4V: {
                 impls.push_back(std::make_unique<common_speculative_state_ngram_map_k>(
                     (config.type),
-                    get_common_ngram_map(config)
+                    get_common_ngram_map(config.type, config.params.ngram_map_k)
                 ));
                 break;
             }
             case COMMON_SPECULATIVE_TYPE_NGRAM_MOD: {
-                GGML_ASSERT(config.params.ngram_mod);
-                impls.push_back(std::make_unique<common_speculative_state_ngram_mod>(config.type, *config.params.ngram_mod));
+                GGML_ASSERT(config.params.ngram_mod.obj);
+                impls.push_back(std::make_unique<common_speculative_state_ngram_mod>(config.type, *config.params.ngram_mod.obj));
                 break;
             }
             case COMMON_SPECULATIVE_TYPE_NGRAM_CACHE: {
-                auto state = create_state_ngram_cache(
-                        params.lookup_cache_static, params.lookup_cache_dynamic, config);
+                auto state = create_state_ngram_cache(params.ngram_cache.lookup_cache_static, params.ngram_cache.lookup_cache_dynamic, config);
                 impls.push_back(std::make_unique<common_speculative_state_ngram_cache>(state));
                 break;
             }
@@ -1099,6 +1147,15 @@ llama_tokens common_speculative_draft(
             impl->n_call_draft++;
         }
 
+        {
+            const int n_min = impl->n_min(params);
+
+            if (!result.empty() && (int) result.size() < n_min) {
+                LOG_DBG("%s: ignoring small draft: %d < %d\n", __func__, (int) result.size(), n_min);
+                result.clear();
+            }
+        }
+
         if (!result.empty()) {
             LOG_DBG("%s: called impl %s, hist size = %zu, call_count = %zu, gen = %zu\n", __func__,
                     common_speculative_type_to_str(impl.get()->type).c_str(), prompt_tgt.size(),
@@ -1108,7 +1165,7 @@ llama_tokens common_speculative_draft(
             impl->n_gen_drafts++;
             impl->n_gen_tokens += result.size();
 
-            break; // We have a draft, so break out of the loop and return it.
+            break; // we have a draft, so break out of the loop and return it.
         }
     }
 
@@ -1136,6 +1193,32 @@ void common_speculative_accept(common_speculative * spec, uint16_t n_accepted) {
     }
 }
 
+int32_t common_speculative_n_max(const common_speculative * spec, const common_params_speculative & params) {
+    if (spec == nullptr) {
+        return 0;
+    }
+
+    int32_t n_max = 0;
+    for (const auto & impl : spec->impls) {
+        n_max = std::max(n_max, impl->n_max(params));
+    }
+
+    return n_max;
+}
+
+int32_t common_speculative_n_min(const common_speculative * spec, const common_params_speculative & params) {
+    if (spec == nullptr) {
+        return 0;
+    }
+
+    int32_t n_min = 0;
+    for (const auto & impl : spec->impls) {
+        n_min = std::max(n_min, impl->n_min(params));
+    }
+
+    return n_min;
+}
+
 void common_speculative_print_stats(const common_speculative * spec) {
     if (spec == nullptr) {
         return;

common/speculative.h

@@ -33,6 +33,9 @@ llama_tokens common_speculative_draft(
 // informs the speculative decoder that n_accepted tokens were accepted by the target model
 void common_speculative_accept(common_speculative * spec, uint16_t n_accepted);
 
+int32_t common_speculative_n_max(const common_speculative * spec, const common_params_speculative & params);
+int32_t common_speculative_n_min(const common_speculative * spec, const common_params_speculative & params);
+
 // print statistics about the speculative decoding
 void common_speculative_print_stats(const common_speculative * spec);
 

convert_hf_to_gguf.py

@@ -272,6 +272,22 @@ class ModelBase:
 
         return tensors
 
+    @staticmethod
+    def _scale_is_trivial(scale: Tensor) -> bool:
+        return scale.numel() <= 1 and abs(float(scale.float().sum()) - 1.0) < 1e-6
+
+    def _write_scale_tensor(self, scale_name: str, scale: Tensor):
+        if not self._scale_is_trivial(scale):
+            scale_f32 = scale.float().numpy().flatten()
+            logger.info(f"  + {scale_name} (per-tensor scale, shape [{scale_f32.size}])")
+            self.gguf_writer.add_tensor(scale_name, scale_f32)
+
+    def _write_scales_tensor(self, scale_name: str, scales: list[float]):
+        if not np.allclose(scales, 1.0, atol=1e-6):
+            scale_vals = np.array(scales, dtype=np.float32)
+            logger.info(f"  + {scale_name} (per-expert scale, shape [{len(scales)}])")
+            self.gguf_writer.add_tensor(scale_name, scale_vals)
+
     def dequant_model(self):
         # If all quantized tensors were already handled (e.g. pure NVFP4), skip
         if self._is_nvfp4 and not any(k.endswith((".weight_scale", ".weight_scale_inv")) for k in self.model_tensors):
@@ -494,7 +510,7 @@ class ModelBase:
                         s = self.model_tensors[name]
                         self.model_tensors[weight_name] = lambda w=w, s=s: dequant_simple(w(), s(), None)
                         tensors_to_remove.append(name)
-                    if name.endswith((".k_scale", ".v_scale")):
+                    if name.endswith((".input_scale", ".k_scale", ".v_scale")):
                         tensors_to_remove.append(name)
             elif quant_method is not None:
                 raise NotImplementedError(f"Quant method is not yet supported: {quant_method!r}")
@@ -602,10 +618,6 @@ class ModelBase:
         raw = np.concatenate([d_grouped, qs_grouped], axis=-1).reshape(out_features, n_super * 36)
         return raw, [out_features, n_super * 64]
 
-    @staticmethod
-    def _nvfp4_scale2_is_trivial(scale2: Tensor) -> bool:
-        return scale2.numel() <= 1 and abs(float(scale2.float().sum()) - 1.0) < 1e-6
-
     def _repack_nvfp4(self, name: str, weight: Tensor, scale: Tensor, scale2: Tensor, input_scale: Tensor):
         if "language_model." in name:
             name = name.replace("language_model.", "")
@@ -616,19 +628,8 @@ class ModelBase:
         logger.info(f"Repacked {new_name} with shape {shape} and quantization NVFP4")
         self.gguf_writer.add_tensor(new_name, raw, raw_dtype=gguf.GGMLQuantizationType.NVFP4)
 
-        # Emit per-tensor scale2 as a separate F32 tensor when non-trivial
-        if not self._nvfp4_scale2_is_trivial(scale2):
-            scale2_f32 = scale2.float().numpy().flatten()
-            scale_name = new_name.replace(".weight", ".scale")
-            logger.info(f"  + {scale_name} (per-tensor NVFP4 scale2, shape [{scale2_f32.size}])")
-            self.gguf_writer.add_tensor(scale_name, scale2_f32)
-
-        # Emit per-tensor input_scale as a separate F32 tensor when non-trivial
-        if not self._nvfp4_scale2_is_trivial(input_scale):
-            input_scale_f32 = input_scale.float().numpy().flatten()
-            input_scale_name = new_name.replace(".weight", ".input_scale")
-            logger.info(f"  + {input_scale_name} (per-tensor NVFP4 input_scale, shape [{input_scale_f32.size}])")
-            self.gguf_writer.add_tensor(input_scale_name, input_scale_f32)
+        self._write_scale_tensor(new_name.replace(".weight", ".scale"), scale2)
+        self._write_scale_tensor(new_name.replace(".weight", ".input_scale"), input_scale)
 
     def _generate_nvfp4_tensors(self):
         # Per-layer expert merging to avoid holding all experts in memory
@@ -719,24 +720,17 @@ class ModelBase:
         logger.info(f"Repacked {new_name} with shape [{len(experts)}, {shape[0]}, {shape[1]}] and quantization NVFP4")
         self.gguf_writer.add_tensor(new_name, merged, raw_dtype=gguf.GGMLQuantizationType.NVFP4)
 
-        # Emit per-expert scale2 tensor if any expert has non-trivial scale2
         scales.sort(key=lambda x: x[0])
-        scale_vals = np.array([s[1] for s in scales], dtype=np.float32)
-        if not np.allclose(scale_vals, 1.0, atol=1e-6):
-            scale_name = new_name.replace(".weight", ".scale")
-            logger.info(f"  + {scale_name} (per-expert NVFP4 scale2, shape [{len(scales)}])")
-            self.gguf_writer.add_tensor(scale_name, scale_vals)
+        self._write_scales_tensor(new_name.replace(".weight", ".scale"), [s[1] for s in scales])
 
-        # Emit per-expert input_scale tensor if any expert has non-trivial input_scale
         input_scales.sort(key=lambda x: x[0])
-        input_scale_vals = np.array([s[1] for s in input_scales], dtype=np.float32)
-        if not np.allclose(input_scale_vals, 1.0, atol=1e-6):
-            input_scale_name = new_name.replace(".weight", ".input_scale")
-            logger.info(f"  + {input_scale_name} (per-expert NVFP4 input_scale, shape [{len(input_scales)}])")
-            self.gguf_writer.add_tensor(input_scale_name, input_scale_vals)
+        self._write_scales_tensor(new_name.replace(".weight", ".input_scale"), [s[1] for s in input_scales])
 
         del experts, merged
 
+    def _needs_nvfp4_processing(self) -> bool:
+        return True
+
     def prepare_tensors(self):
         # detect NVFP4 quantization (ModelOpt format)
         quant_algo = (self.hparams.get("quantization_config") or {}).get("quant_algo")
@@ -767,7 +761,7 @@ class ModelBase:
         # NVFP4 weights are repacked and written directly to gguf_writer.
         # This must run before dequant_model so NVFP4 tensors are removed
         # from model_tensors, leaving only non-NVFP4 (e.g. FP8) for dequant.
-        if self._is_nvfp4:
+        if self._is_nvfp4 and self._needs_nvfp4_processing():
             self._generate_nvfp4_tensors()
 
         self.dequant_model()
@@ -2199,6 +2193,10 @@ class MmprojModel(ModelBase):
                 # merge configs
                 self.preprocessor_config = {**self.preprocessor_config, **cfg}
 
+    def _needs_nvfp4_processing(self) -> bool:
+        # nvfp4 quantization applies to the text model only.
+        return False
+
     def get_vision_config(self) -> dict[str, Any] | None:
         config_name = "vision_config" if not self.is_mistral_format else "vision_encoder"
         return self.global_config.get(config_name)
@@ -4459,6 +4457,12 @@ class NemotronNanoV2VLModel(MmprojModel):
         }
         return vision_config
 
+    def dequant_model(self):
+        if self._is_nvfp4:
+            # Skip nvfp4 quantization for vision/audio model.
+            return
+        super().dequant_model()
+
     def set_gguf_parameters(self):
         if "image_mean" not in self.preprocessor_config:
             self.preprocessor_config["image_mean"] = [0.485, 0.456, 0.406]
@@ -4482,6 +4486,10 @@ class NemotronNanoV2VLModel(MmprojModel):
         if "input_conditioner" in name:
             return
 
+        # mtmd does not support video yet so skip tensors related to video.
+        if "radio_model.model.patch_generator.video_embedder" in name:
+            return
+
         # RADIO's pos_embed doesn't have .weight suffix, but clip.cpp expects it
         if "patch_generator.pos_embed" in name:
             if not name.endswith(".weight"):
@@ -6650,7 +6658,7 @@ class BertModel(TextModel):
 
         tokens: list[bytes] = [f"[PAD{i}]".encode("utf-8") for i in range(vocab_size)]
         scores: list[float] = [-10000.0] * vocab_size
-        toktypes: list[int] = [SentencePieceTokenTypes.UNUSED] * vocab_size  # ty: ignore[invalid-assignment]
+        toktypes: list[int] = [SentencePieceTokenTypes.UNUSED] * vocab_size
 
         if isinstance(tokenizer, SentencePieceProcessor):
             for token_id in range(tokenizer.vocab_size()):
@@ -10829,7 +10837,11 @@ class NemotronHModel(GraniteHybridModel):
         # uses self.model_arch to build the tensor name map, and all MoE-specific
         # mappings would be missed if it were called with the default non-MoE arch.
         hparams = ModelBase.load_hparams(args[0], self.is_mistral_format)
-        if "num_experts_per_tok" in hparams:
+        has_moe_params = (
+            "num_experts_per_tok" in hparams
+            or (isinstance(hparams.get("llm_config"), dict) and "num_experts_per_tok" in hparams["llm_config"])
+        )
+        if has_moe_params:
             self.model_arch = gguf.MODEL_ARCH.NEMOTRON_H_MOE
             self.is_moe = True
 
@@ -10976,6 +10988,11 @@ class NemotronHModel(GraniteHybridModel):
         if name.startswith(("vision_model.", "mlp1.")):
             return
 
+        if name.startswith(("sound_encoder.")):
+            return
+        if name.startswith(("sound_projection.")):
+            return
+
         # Strip language_model. prefix for VLM models (e.g., Nemotron Nano 12B v2 VL)
         if name.startswith("language_model."):
             name = name[len("language_model."):]

examples/debug/debug.cpp

@@ -202,10 +202,14 @@ static bool run(llama_context * ctx, const common_params & params) {
     print_tokenized_prompt(ctx, tokens, params.prompt);
 
     if (params.save_logits) {
-        output_data output {ctx, model, params};
-        std::filesystem::path model_path{params.model.path};
-        std::string model_name{model_path.stem().string()};
-        save_output_data(output, model_name, params.logits_output_dir);
+        try {
+            output_data output {ctx, model, params};
+            std::filesystem::path model_path{params.model.path};
+            std::string model_name{model_path.stem().string()};
+            save_output_data(output, model_name, params.logits_output_dir);
+        } catch (const std::exception & e) {
+            LOG_ERR("%s : error saving logits: %s\n", __func__, e.what());
+        }
     }
 
     return true;
@@ -223,7 +227,7 @@ int main(int argc, char ** argv) {
     llama_backend_init();
     llama_numa_init(params.numa);
 
-    std::optional<base_callback_data> cb_data;
+    std::optional<common_debug_cb_user_data> cb_data;
     if (!params.save_logits) {
         cb_data.emplace(params, params.tensor_filter);
     }

examples/eval-callback/eval-callback.cpp

@@ -3,7 +3,6 @@
 #include "debug.h"
 #include "log.h"
 #include "llama.h"
-#include "llama-cpp.h"
 
 #include <clocale>
 #include <string>
@@ -38,7 +37,7 @@ static bool run(llama_context * ctx, const common_params & params) {
 int main(int argc, char ** argv) {
     std::setlocale(LC_NUMERIC, "C");
 
-    base_callback_data cb_data;
+    common_debug_cb_user_data cb_data;
 
     common_params params;
 
@@ -53,7 +52,7 @@ int main(int argc, char ** argv) {
 
     // pass the callback to the backend scheduler
     // it will be executed for each node during the graph computation
-    params.cb_eval = common_debug_cb_eval<false>;
+    params.cb_eval = common_debug_cb_eval;
     params.cb_eval_user_data = &cb_data;
     params.warmup = false;
 

examples/gen-docs/gen-docs.cpp

@@ -73,12 +73,12 @@ static void write_help(std::ostringstream & ss, const md_file & md) {
     auto ctx_arg = common_params_parser_init(params, md.ex);
 
     std::vector<common_arg *> common_options;
-    std::vector<common_arg *> sparam_options;
+    std::vector<common_arg *> sampling_options;
     std::vector<common_arg *> specific_options;
     for (auto & opt : ctx_arg.options) {
         // in case multiple LLAMA_EXAMPLE_* are set, we prioritize the LLAMA_EXAMPLE_* matching current example
-        if (opt.is_sparam) {
-            sparam_options.push_back(&opt);
+        if (opt.is_sampling) {
+            sampling_options.push_back(&opt);
         } else if (opt.in_example(ctx_arg.ex)) {
             specific_options.push_back(&opt);
         } else {
@@ -93,7 +93,7 @@ static void write_help(std::ostringstream & ss, const md_file & md) {
     ss << "### Common params\n\n";
     write_table(ss, common_options);
     ss << "\n\n### Sampling params\n\n";
-    write_table(ss, sparam_options);
+    write_table(ss, sampling_options);
     ss << "\n\n### " << md.specific_section_header << "\n\n";
     write_table(ss, specific_options);
 

examples/lookup/lookup-create.cpp

@@ -37,9 +37,9 @@ int main(int argc, char ** argv){
 
     common_ngram_cache ngram_cache;
     common_ngram_cache_update(ngram_cache, LLAMA_NGRAM_STATIC, LLAMA_NGRAM_STATIC, inp, inp.size(), true);
-    fprintf(stderr, "%s: hashing done, writing file to %s\n", __func__, params.speculative.lookup_cache_static.c_str());
+    fprintf(stderr, "%s: hashing done, writing file to %s\n", __func__, params.speculative.ngram_cache.lookup_cache_static.c_str());
 
-    common_ngram_cache_save(ngram_cache, params.speculative.lookup_cache_static);
+    common_ngram_cache_save(ngram_cache, params.speculative.ngram_cache.lookup_cache_static);
 
     return 0;
 }

examples/lookup/lookup-stats.cpp

@@ -24,7 +24,7 @@ int main(int argc, char ** argv){
         return 1;
     }
 
-    const int n_draft = params.speculative.n_max;
+    const int n_draft = params.speculative.draft.n_max;
 
     // init llama.cpp
     llama_backend_init();
@@ -49,18 +49,18 @@ int main(int argc, char ** argv){
     {
         const int64_t t_start_draft_us = ggml_time_us();
 
-        if (!params.speculative.lookup_cache_static.empty()) {
+        if (!params.speculative.ngram_cache.lookup_cache_static.empty()) {
             try {
-                ngram_cache_static = common_ngram_cache_load(params.speculative.lookup_cache_static);
+                ngram_cache_static = common_ngram_cache_load(params.speculative.ngram_cache.lookup_cache_static);
             } catch (std::ifstream::failure const &) {
-                LOG_ERR("failed to open static lookup cache: %s", params.speculative.lookup_cache_static.c_str());
+                LOG_ERR("failed to open static lookup cache: %s", params.speculative.ngram_cache.lookup_cache_static.c_str());
                 exit(1);
             }
         }
 
-        if (!params.speculative.lookup_cache_dynamic.empty()) {
+        if (!params.speculative.ngram_cache.lookup_cache_dynamic.empty()) {
             try {
-                ngram_cache_dynamic = common_ngram_cache_load(params.speculative.lookup_cache_dynamic);
+                ngram_cache_dynamic = common_ngram_cache_load(params.speculative.ngram_cache.lookup_cache_dynamic);
             } catch (std::ifstream::failure const &) {} // if the file does not exist it will simply be created at the end of the program
         }
 

examples/lookup/lookup.cpp

@@ -25,7 +25,7 @@ int main(int argc, char ** argv){
     }
 
     // max. number of additional tokens to draft if match is found
-    const int n_draft = params.speculative.n_max;
+    const int n_draft = params.speculative.draft.n_max;
 
     // init llama.cpp
     llama_backend_init();
@@ -54,18 +54,18 @@ int main(int argc, char ** argv){
         const int64_t t_start_draft_us = ggml_time_us();
         common_ngram_cache_update(ngram_cache_context, LLAMA_NGRAM_MIN, LLAMA_NGRAM_MAX, inp, inp.size(), false);
 
-        if (!params.speculative.lookup_cache_static.empty()) {
+        if (!params.speculative.ngram_cache.lookup_cache_static.empty()) {
             try {
-                ngram_cache_static = common_ngram_cache_load(params.speculative.lookup_cache_static);
+                ngram_cache_static = common_ngram_cache_load(params.speculative.ngram_cache.lookup_cache_static);
             } catch (std::ifstream::failure const &) {
-                LOG_ERR("failed to open static lookup cache: %s", params.speculative.lookup_cache_static.c_str());
+                LOG_ERR("failed to open static lookup cache: %s", params.speculative.ngram_cache.lookup_cache_static.c_str());
                 exit(1);
             }
         }
 
-        if (!params.speculative.lookup_cache_dynamic.empty()) {
+        if (!params.speculative.ngram_cache.lookup_cache_dynamic.empty()) {
             try {
-                ngram_cache_dynamic = common_ngram_cache_load(params.speculative.lookup_cache_dynamic);
+                ngram_cache_dynamic = common_ngram_cache_load(params.speculative.ngram_cache.lookup_cache_dynamic);
             } catch (std::ifstream::failure const &) {} // if the file does not exist it will simply be created at the end of the program
         }
 
@@ -213,7 +213,7 @@ int main(int argc, char ** argv){
 
     // Update dynamic ngram cache with context ngram cache and save it to disk:
     common_ngram_cache_merge(ngram_cache_dynamic, ngram_cache_context);
-    common_ngram_cache_save(ngram_cache_dynamic, params.speculative.lookup_cache_dynamic);
+    common_ngram_cache_save(ngram_cache_dynamic, params.speculative.ngram_cache.lookup_cache_dynamic);
 
     LOG("\n\n");
 

examples/speculative-simple/speculative-simple.cpp

@@ -43,7 +43,7 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
-    if (params.speculative.mparams_dft.path.empty()) {
+    if (params.speculative.draft.mparams.path.empty()) {
         LOG_ERR("%s: --model-draft is required\n", __func__);
         return 1;
     }
@@ -77,7 +77,7 @@ int main(int argc, char ** argv) {
 
     // TODO: simplify this logic
     {
-        const auto & params_spec = params.speculative;
+        const auto & params_spec = params.speculative.draft;
 
         auto params_dft = params;
 
@@ -85,15 +85,15 @@ int main(int argc, char ** argv) {
         params_dft.n_ctx        = params_spec.n_ctx;
         params_dft.n_batch      = llama_n_ctx_seq(ctx_tgt);
         params_dft.devices      = params_spec.devices;
-        params_dft.model        = params_spec.mparams_dft;
+        params_dft.model        = params_spec.mparams;
         params_dft.n_gpu_layers = params_spec.n_gpu_layers;
 
         if (params_spec.cpuparams.n_threads > 0) {
-            params_dft.cpuparams.n_threads       = params.speculative.cpuparams.n_threads;
-            params_dft.cpuparams_batch.n_threads = params.speculative.cpuparams_batch.n_threads;
+            params_dft.cpuparams.n_threads       = params.speculative.draft.cpuparams.n_threads;
+            params_dft.cpuparams_batch.n_threads = params.speculative.draft.cpuparams_batch.n_threads;
         }
 
-        params_dft.tensor_buft_overrides = params.speculative.tensor_buft_overrides;
+        params_dft.tensor_buft_overrides = params.speculative.draft.tensor_buft_overrides;
 
         auto mparams_dft = common_model_params_to_llama(params_dft);
 
@@ -103,8 +103,8 @@ int main(int argc, char ** argv) {
             return 1;
         }
 
-        params.speculative.model_dft = model_dft.get();
-        params.speculative.cparams_dft = common_context_params_to_llama(params_dft);
+        params.speculative.draft.model = model_dft.get();
+        params.speculative.draft.cparams = common_context_params_to_llama(params_dft);
     }
 
     // Tokenize the prompt
@@ -187,16 +187,6 @@ int main(int argc, char ** argv) {
             // generate a new draft
             draft = common_speculative_draft(spec, params_spec, prompt_tgt, id_last);
 
-            if ((int) draft.size() > params_spec.n_max) {
-                LOG_WRN("draft size %zu exceeds max %d, truncating\n", draft.size(), params_spec.n_max);
-                draft.resize(params_spec.n_max);
-            }
-
-            if ((int) draft.size() < params_spec.n_min) {
-                LOG_DBG("ignoring small draft: %zu < %d\n", draft.size(), params_spec.n_min);
-                draft.clear();
-            }
-
             // save the original draft size
             n_draft = draft.size();
 
@@ -220,19 +210,12 @@ int main(int argc, char ** argv) {
             }
         }
 
-        GGML_ASSERT(n_draft > 0);
-
         // always have a token to evaluate from before - id_last
         common_batch_clear(batch_tgt);
         common_batch_add  (batch_tgt, id_last, n_past++, { 0 }, true);
 
         // evaluate the target model on [id_last, draft0, draft1, ..., draftN-1]
         {
-            // do not waste time on small drafts
-            if (draft.size() < (size_t) params_spec.n_min) {
-                draft.clear();
-            }
-
             for (size_t i = 0; i < draft.size(); ++i) {
                 common_batch_add(batch_tgt, draft[i], n_past + i, { 0 }, true);
             }
@@ -340,7 +323,7 @@ int main(int argc, char ** argv) {
     LOG_INF("decoded %4d tokens in %8.3f seconds, speed: %8.3f t/s\n", n_predict, (t_dec_end - t_dec_start) / 1e6f, n_predict  / ((t_dec_end - t_dec_start) / 1e6f));
 
     LOG_INF("\n");
-    LOG_INF("n_draft   = %d\n", params_spec.n_max);
+    LOG_INF("n_draft   = %d\n", params_spec.draft.n_max);
     LOG_INF("n_predict = %d\n", n_predict);
     LOG_INF("n_drafted = %d\n", n_drafted);
     LOG_INF("n_accept  = %d\n", n_accept);

examples/speculative/speculative.cpp

@@ -49,7 +49,7 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
-    if (params.speculative.mparams_dft.path.empty()) {
+    if (params.speculative.draft.mparams.path.empty()) {
         LOG_ERR("%s: --model-draft is required\n", __func__);
         return 1;
     }
@@ -58,7 +58,7 @@ int main(int argc, char ** argv) {
     const int n_seq_dft = params.n_parallel;
 
     // probability threshold for splitting a draft branch (only for n_seq_dft > 1)
-    const float p_draft_split = params.speculative.p_split;
+    const float p_draft_split = params.speculative.draft.p_split;
 
     std::default_random_engine rng(params.sampling.seed == LLAMA_DEFAULT_SEED ? std::random_device()() : params.sampling.seed);
     std::uniform_real_distribution<> u_dist;
@@ -80,15 +80,15 @@ int main(int argc, char ** argv) {
     ctx_tgt   = llama_init_tgt->context();
 
     // load the draft model
-    params.devices = params.speculative.devices;
-    params.model = params.speculative.mparams_dft;
-    params.n_gpu_layers = params.speculative.n_gpu_layers;
-    if (params.speculative.cpuparams.n_threads > 0) {
-        params.cpuparams.n_threads = params.speculative.cpuparams.n_threads;
+    params.devices = params.speculative.draft.devices;
+    params.model = params.speculative.draft.mparams;
+    params.n_gpu_layers = params.speculative.draft.n_gpu_layers;
+    if (params.speculative.draft.cpuparams.n_threads > 0) {
+        params.cpuparams.n_threads = params.speculative.draft.cpuparams.n_threads;
     }
 
-    params.cpuparams_batch.n_threads = params.speculative.cpuparams_batch.n_threads;
-    params.tensor_buft_overrides     = params.speculative.tensor_buft_overrides;
+    params.cpuparams_batch.n_threads = params.speculative.draft.cpuparams_batch.n_threads;
+    params.tensor_buft_overrides     = params.speculative.draft.tensor_buft_overrides;
 
     auto llama_init_dft = common_init_from_params(params);
 
@@ -110,13 +110,21 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
-    if (
-        llama_vocab_get_add_bos(vocab_tgt) != llama_vocab_get_add_bos(vocab_dft) ||
-        llama_vocab_get_add_eos(vocab_tgt) != llama_vocab_get_add_eos(vocab_dft) ||
-        llama_vocab_bos(vocab_tgt) != llama_vocab_bos(vocab_dft) ||
-        llama_vocab_eos(vocab_tgt) != llama_vocab_eos(vocab_dft)
-    ) {
-        LOG_ERR("%s: draft model special tokens must match target model to use speculation\n", __func__);
+    if (llama_vocab_get_add_bos(vocab_tgt) != llama_vocab_get_add_bos(vocab_dft) ||
+        (llama_vocab_get_add_bos(vocab_tgt) && llama_vocab_bos(vocab_tgt) != llama_vocab_bos(vocab_dft))) {
+        LOG_ERR("%s: draft model bos tokens must match target model to use speculation. add: %d - %d, id: %d - %d)\n",
+                __func__,
+                llama_vocab_get_add_bos(vocab_tgt), llama_vocab_get_add_bos(vocab_dft),
+                llama_vocab_bos(vocab_tgt), llama_vocab_bos(vocab_dft));
+        return 1;
+    }
+
+    if (llama_vocab_get_add_eos(vocab_tgt) != llama_vocab_get_add_eos(vocab_dft) ||
+        (llama_vocab_get_add_eos(vocab_tgt) && llama_vocab_eos(vocab_tgt) != llama_vocab_eos(vocab_dft))) {
+        LOG_ERR("%s: draft model eos tokens must match target model to use speculation. add: %d - %d, id: %d - %d)\n",
+                __func__,
+                llama_vocab_get_add_eos(vocab_tgt), llama_vocab_get_add_eos(vocab_dft),
+                llama_vocab_eos(vocab_tgt), llama_vocab_eos(vocab_dft));
         return 1;
     }
 
@@ -137,11 +145,12 @@ int main(int argc, char ** argv) {
         for (int i = SPEC_VOCAB_CHECK_START_TOKEN_ID; i < std::min(n_vocab_tgt, n_vocab_dft); ++i) {
             const char * token_text_tgt = llama_vocab_get_text(vocab_tgt, i);
             const char * token_text_dft = llama_vocab_get_text(vocab_dft, i);
+
             if (std::strcmp(token_text_tgt, token_text_dft) != 0) {
                 LOG_ERR("%s: draft model vocab must match target model to use speculation but ", __func__);
                 LOG_ERR("token %d content differs - target '%s', draft '%s'\n", i,
-                        common_token_to_piece(ctx_tgt, i).c_str(),
-                        common_token_to_piece(ctx_dft, i).c_str());
+                        common_token_to_piece(vocab_tgt, i).c_str(),
+                        common_token_to_piece(vocab_dft, i).c_str());
                 return 1;
             }
         }
@@ -183,7 +192,7 @@ int main(int argc, char ** argv) {
     //GGML_ASSERT(n_vocab == llama_vocab_n_tokens(model_dft));
 
     // how many tokens to draft each time
-    int n_draft = params.speculative.n_max;
+    int n_draft = params.speculative.draft.n_max;
 
     int n_predict = 0;
     int n_drafted = 0;

ggml/CMakeLists.txt

@@ -5,7 +5,7 @@ project("ggml" C CXX ASM)
 ### GGML Version
 set(GGML_VERSION_MAJOR 0)
 set(GGML_VERSION_MINOR 10)
-set(GGML_VERSION_PATCH 0)
+set(GGML_VERSION_PATCH 1)
 set(GGML_VERSION_BASE "${GGML_VERSION_MAJOR}.${GGML_VERSION_MINOR}.${GGML_VERSION_PATCH}")
 
 list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake/")

ggml/src/CMakeLists.txt

@@ -470,11 +470,10 @@ endforeach()
 
 target_link_libraries(ggml-base PRIVATE Threads::Threads)
 
-find_library(MATH_LIBRARY m)
-if (MATH_LIBRARY)
-    if (NOT WIN32 OR NOT DEFINED ENV{ONEAPI_ROOT})
-        target_link_libraries(ggml-base PRIVATE ${MATH_LIBRARY})
-    endif()
+if (DEFINED MATH_LIBRARY)
+    target_link_libraries(ggml-base PRIVATE ${MATH_LIBRARY})
+elseif (NOT WIN32 AND NOT DEFINED ENV{ONEAPI_ROOT})
+    target_link_libraries(ggml-base PRIVATE m)
 endif()
 
 if (CMAKE_SYSTEM_NAME MATCHES "Android")

ggml/src/ggml-backend-meta.cpp

@@ -1826,7 +1826,24 @@ static enum ggml_status ggml_backend_meta_graph_compute(ggml_backend_t backend,
                     continue;
                 }
 
-                i = get_i_delayed(i);
+                const int i_delayed = get_i_delayed(i);
+
+                // If we can delay the AllReduce we need to consider the interaction with zero-sized tensor slices.
+                // A backend with such a slice would normally have valid data after participating in the AllReduce with a node that has
+                //     its compute flag disabled and thus gets its data zeroed out.
+                // If the AllReduce is delayed then the nodes until that point also need to have their compute flag disabled.
+                if (i_delayed > i) {
+                    for (size_t j = 0; j < n_backends; j++) {
+                        auto & bcj = backend_ctx->backend_configs[j];
+                        if ((bcj.nodes[i]->flags & GGML_TENSOR_FLAG_COMPUTE) == 0) {
+                            for (int ii = i + 1; ii <= i_delayed; ii++) {
+                                bcj.nodes[ii]->flags &= ~GGML_TENSOR_FLAG_COMPUTE;
+                            }
+                        }
+                    }
+                }
+
+                i = i_delayed;
 
                 for (size_t j = 0; j < n_backends; j++) {
                     auto & bcj = backend_ctx->backend_configs[j];
@@ -2083,8 +2100,8 @@ static const ggml_backend_i ggml_backend_meta_i = {
     /* .free                    = */ ggml_backend_meta_free,
     /* .set_tensor_async        = */ ggml_backend_meta_set_tensor_async,
     /* .get_tensor_async        = */ ggml_backend_meta_get_tensor_async,
-    /* .get_tensor_2d_async     = */ nullptr,
     /* .set_tensor_2d_async     = */ nullptr,
+    /* .get_tensor_2d_async     = */ nullptr,
     /* .cpy_tensor_async        = */ nullptr,
     /* .synchronize             = */ ggml_backend_meta_synchronize,
     /* .graph_plan_create       = */ nullptr,

ggml/src/ggml-backend-reg.cpp

@@ -181,6 +181,12 @@ struct ggml_backend_registry {
             return;
         }
 
+        for (auto & entry : backends) {
+            if (entry.reg == reg) {
+                return;
+            }
+        }
+
 #ifndef NDEBUG
         GGML_LOG_DEBUG("%s: registered backend %s (%zu devices)\n",
             __func__, ggml_backend_reg_name(reg), ggml_backend_reg_dev_count(reg));
@@ -192,6 +198,12 @@ struct ggml_backend_registry {
     }
 
     void register_device(ggml_backend_dev_t device) {
+        for (auto & dev : devices) {
+            if (dev == device) {
+                return;
+            }
+        }
+
 #ifndef NDEBUG
         GGML_LOG_DEBUG("%s: registered device %s (%s)\n", __func__, ggml_backend_dev_name(device), ggml_backend_dev_description(device));
 #endif

ggml/src/ggml-blas/ggml-blas.cpp

@@ -262,9 +262,9 @@ static struct ggml_backend_i blas_backend_i = {
     /* .get_name                = */ ggml_backend_blas_get_name,
     /* .free                    = */ ggml_backend_blas_free,
     /* .set_tensor_async        = */ NULL,
-    /* .get_tensor_2d_async     = */ NULL,
-    /* .set_tensor_2d_async     = */ NULL,
     /* .get_tensor_async        = */ NULL,
+    /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ NULL,
     /* .synchronize             = */ NULL,
     /* .graph_plan_create       = */ NULL,

ggml/src/ggml-cann/aclnn_ops.cpp

@@ -25,6 +25,7 @@
 #include "ggml-impl.h"
 #include "ggml.h"
 
+
 #include <aclnnop/aclnn_add.h>
 #include <aclnnop/aclnn_add_rms_norm.h>
 #include <aclnnop/aclnn_addcdiv.h>
@@ -45,7 +46,9 @@
 #include <aclnnop/aclnn_fused_infer_attention_score_v2.h>
 #include <aclnnop/aclnn_ger.h>
 #include <aclnnop/aclnn_group_norm.h>
+#include <aclnnop/aclnn_gather_v2.h>
 #include <aclnnop/aclnn_grouped_matmul_v3.h>
+#include <aclnnop/aclnn_scatter.h>
 #include <aclnnop/aclnn_gt_scalar.h>
 #include <aclnnop/aclnn_im2col.h>
 #include <aclnnop/aclnn_index_copy.h>
@@ -62,6 +65,7 @@
 #include <aclnnop/aclnn_permute.h>
 #include <aclnnop/aclnn_pow.h>
 #include <aclnnop/aclnn_pow_tensor_tensor.h>
+#include <aclnnop/aclnn_recurrent_gated_delta_rule.h>
 #include <aclnnop/aclnn_reduce_sum.h>
 #include <aclnnop/aclnn_reflection_pad1d.h>
 #include <aclnnop/aclnn_repeat.h>
@@ -69,11 +73,15 @@
 #include <aclnnop/aclnn_rms_norm.h>
 #include <aclnnop/aclnn_roll.h>
 #include <aclnnop/aclnn_softmax.h>
+#include <aclnnop/aclnn_softmax_cross_entropy_with_logits.h>
 #include <aclnnop/aclnn_sub.h>
 #include <aclnnop/aclnn_sum.h>
 #include <aclnnop/aclnn_threshold.h>
 #include <aclnnop/aclnn_tril.h>
+#include <aclnnop/aclnn_triangular_solve.h>
 #include <aclnnop/aclnn_triu.h>
+#include <aclnnop/aclnn_logical_not.h>
+#include <aclnnop/aclnn_masked_fill_scalar.h>
 #include <aclnnop/aclnn_upsample_nearest_2d.h>
 #include <aclnnop/aclnn_weight_quant_batch_matmul_v2.h>
 #include <aclnnop/aclnn_zero.h>
@@ -151,6 +159,107 @@ void ggml_cann_op_unary_gated(std::function<void(ggml_backend_cann_context &, ac
     GGML_CANN_CALL_ACLNN_OP(ctx, InplaceMul, acl_dst.get(), acl_src1.get());
 }
 
+// Fused SwiGLU using aclnnSwiGlu: splits input along innermost dim, applies
+// SiLU to left half, multiplies by right half.
+//
+// Falls back to the generic two-kernel path when src[1] != nullptr (two
+// independent halves) or swapped != 0 (reversed activation order), as
+// aclnnSwiGlu only handles the single interleaved tensor in standard order.
+//
+// CANN tiling for SwiGlu requires (storageShapeDim + viewDims) to be even.
+// aclCreateTensor always uses storageShapeDim=1, so viewDims must be odd.
+// We use a 3D view (1+3=4, even) to satisfy this constraint while preserving
+// correct split semantics along the innermost (ne[0]) dimension.
+void ggml_cann_swiglu(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    auto silu_fn = [](ggml_backend_cann_context & ctx, aclTensor * acl_src, aclTensor * acl_dst) {
+        GGML_CANN_CALL_ACLNN_OP(ctx, Silu, acl_src, acl_dst);
+    };
+
+    const int32_t swapped = ggml_get_op_params_i32(dst, 1);
+    if (dst->src[1] != nullptr || swapped != 0) {
+        ggml_cann_op_unary_gated(silu_fn, ctx, dst);
+        return;
+    }
+
+    // aclnnSwiGlu requires the split dim (src->ne[0]) to be even; fall back otherwise.
+    if (dst->src[0]->ne[0] % 2 != 0) {
+        ggml_cann_op_unary_gated(silu_fn, ctx, dst);
+        return;
+    }
+
+    ggml_tensor * src0 = dst->src[0];
+    size_t elem_size = ggml_element_size(src0);
+
+    // src0 GGML: [2*ne0, ne1, ne2, ne3] → 3D view [2*ne0, ne1, ne2*ne3]
+    // CANN reversed: [ne2*ne3, ne1, 2*ne0], split along CANN dim 2 (last).
+    int64_t ne0_x2   = src0->ne[0];
+    int64_t ne1      = src0->ne[1];
+    int64_t ne23     = src0->ne[2] * src0->ne[3];
+    int64_t src3d_ne[] = { ne0_x2, ne1, ne23 };
+    size_t  src3d_nb[] = { (size_t)src0->nb[0], (size_t)src0->nb[1], (size_t)src0->nb[2] };
+    acl_tensor_ptr acl_src = ggml_cann_create_tensor(src0->data, ggml_cann_type_mapping(src0->type),
+                                                     elem_size, src3d_ne, src3d_nb, 3);
+
+    // dst GGML: [ne0, ne1, ne2, ne3] → 3D view [ne0, ne1, ne2*ne3]
+    int64_t ne0      = dst->ne[0];
+    int64_t dst3d_ne[] = { ne0, ne1, ne23 };
+    size_t  dst3d_nb[] = { (size_t)dst->nb[0], (size_t)dst->nb[1], (size_t)dst->nb[2] };
+    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst->data, ggml_cann_type_mapping(dst->type),
+                                                     elem_size, dst3d_ne, dst3d_nb, 3);
+
+    // CANN tensor [ne23, ne1, 2*ne0]: split along CANN dim 2 (last) = 2*ne0.
+    GGML_CANN_CALL_ACLNN_OP(ctx, SwiGlu, acl_src.get(), (int64_t)2, acl_dst.get());
+}
+
+// Fused GeGLU using aclnnGeGluV3: splits input along ne[0] (CANN last dim),
+// activates the LEFT half with GELU, multiplies by right half.
+// approximate: 0=tanh, 1=none(erf). activateLeft=true matches GGML convention.
+// outGelu is a required-but-discard output buffer.
+//
+// Falls back to the generic two-kernel path when src[1] != nullptr (two
+// independent halves) or swapped != 0 (reversed activation order), as
+// aclnnGeGluV3 only handles the single interleaved tensor in standard order.
+void ggml_cann_geglu(ggml_backend_cann_context & ctx, ggml_tensor * dst, int64_t approximate) {
+    auto gelu_fn = [](ggml_backend_cann_context & ctx, aclTensor * acl_src, aclTensor * acl_dst) {
+        GGML_CANN_CALL_ACLNN_OP(ctx, Gelu, acl_src, acl_dst);
+    };
+
+    const int32_t swapped = ggml_get_op_params_i32(dst, 1);
+    if (dst->src[1] != nullptr || swapped != 0) {
+        ggml_cann_op_unary_gated(gelu_fn, ctx, dst);
+        return;
+    }
+
+    // aclnnGeGluV3 requires the split dim (src->ne[0]) to be even; fall back otherwise.
+    if (dst->src[0]->ne[0] % 2 != 0) {
+        ggml_cann_op_unary_gated(gelu_fn, ctx, dst);
+        return;
+    }
+
+    ggml_tensor * src0 = dst->src[0];
+    acl_tensor_ptr acl_src = ggml_cann_create_tensor(src0);
+    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst);
+
+    // Allocate a temporary buffer for the required outGelu output (same shape as dst).
+    // Build contiguous strides since the pool allocation is a fresh buffer.
+    size_t  elem_size    = ggml_element_size(dst);
+    int64_t ne[GGML_MAX_DIMS] = { dst->ne[0], dst->ne[1], dst->ne[2], dst->ne[3] };
+    size_t  nb[GGML_MAX_DIMS];
+    nb[0] = elem_size;
+    for (int i = 1; i < GGML_MAX_DIMS; i++) {
+        nb[i] = nb[i - 1] * ne[i - 1];
+    }
+    size_t gelu_out_size = nb[GGML_MAX_DIMS - 1] * ne[GGML_MAX_DIMS - 1];
+    ggml_cann_pool_alloc gelu_out_alloc(ctx.pool(), gelu_out_size);
+
+    acl_tensor_ptr acl_gelu_out = ggml_cann_create_tensor(
+        gelu_out_alloc.get(), ggml_cann_type_mapping(dst->type), elem_size, ne, nb, GGML_MAX_DIMS);
+    // V3 adds activateLeft param; true → Gelu(left)*right, matching GGML convention.
+    // GGML dim 0 → CANN last dim (index GGML_MAX_DIMS-1 = 3 for 4D tensor).
+    GGML_CANN_CALL_ACLNN_OP(ctx, GeGluV3, acl_src.get(), (int64_t)(GGML_MAX_DIMS - 1), approximate, true,
+                             acl_dst.get(), acl_gelu_out.get());
+}
+
 /**
  * @brief Repeats elements of a tensor along each dimension according to the
  * specified repeat array.
@@ -445,28 +554,33 @@ void ggml_cann_l2_norm(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
     ggml_cann_pool_alloc temp_buffer_allocator(ctx.pool(), n_bytes);
     void *               buffer = temp_buffer_allocator.get();
 
-    int64_t div_ne[] = { 1, src->ne[1], src->ne[2], src->ne[3] };
-    size_t  div_nb[GGML_MAX_DIMS];
-    div_nb[0] = sizeof(float);
+    int64_t norm_ne[] = { 1, src->ne[1], src->ne[2], src->ne[3] };
+    size_t  norm_nb[GGML_MAX_DIMS];
+    norm_nb[0] = sizeof(float);
     for (int i = 1; i < GGML_MAX_DIMS; ++i) {
-        div_nb[i] = div_nb[i - 1] * div_ne[i - 1];
+        norm_nb[i] = norm_nb[i - 1] * norm_ne[i - 1];
     }
-    acl_tensor_ptr acl_div = ggml_cann_create_tensor(buffer, ACL_FLOAT, type_size, div_ne, div_nb, GGML_MAX_DIMS);
+    acl_tensor_ptr acl_norm = ggml_cann_create_tensor(buffer, ACL_FLOAT, sizeof(float), norm_ne, norm_nb, GGML_MAX_DIMS);
 
     std::vector<int64_t> norm_dims  = { 3 };
     acl_int_array_ptr    dims_array = ggml_cann_create_int_array(norm_dims.data(), norm_dims.size());
 
     float          p_value  = 2.0f;
     acl_scalar_ptr p_scalar = ggml_cann_create_scalar(&p_value, aclDataType::ACL_FLOAT);
-    GGML_CANN_CALL_ACLNN_OP(ctx, Norm, acl_src.get(), p_scalar.get(), dims_array.get(), true, acl_div.get());
+    GGML_CANN_CALL_ACLNN_OP(ctx, Norm, acl_src.get(), p_scalar.get(), dims_array.get(), true, acl_norm.get());
 
-    // Clamp norm to at least eps: scale = 1/fmaxf(norm, eps)
-    acl_scalar_ptr acl_min = ggml_cann_create_scalar(&eps, aclDataType::ACL_FLOAT);
-    float          flt_max = FLT_MAX;
-    acl_scalar_ptr acl_max = ggml_cann_create_scalar(&flt_max, aclDataType::ACL_FLOAT);
-    GGML_CANN_CALL_ACLNN_OP(ctx, Clamp, acl_div.get(), acl_min.get(), acl_max.get(), acl_div.get());
+    ggml_cann_pool_alloc clamp_buffer_allocator(ctx.pool());
+    acl_tensor_ptr       acl_clamped;
 
-    GGML_CANN_CALL_ACLNN_OP(ctx, Div, acl_src.get(), acl_div.get(), acl_dst.get());
+    if (eps > 0.0f) {
+        void *         clamp_buf  = clamp_buffer_allocator.alloc(n_bytes);
+        acl_clamped               = ggml_cann_create_tensor(clamp_buf, ACL_FLOAT, sizeof(float), norm_ne, norm_nb, GGML_MAX_DIMS);
+        acl_scalar_ptr eps_scalar = ggml_cann_create_scalar(&eps, aclDataType::ACL_FLOAT);
+        GGML_CANN_CALL_ACLNN_OP(ctx, ClampMin, acl_norm.get(), eps_scalar.get(), acl_clamped.get());
+    }
+
+    aclTensor * acl_div_input = acl_clamped ? acl_clamped.get() : acl_norm.get();
+    GGML_CANN_CALL_ACLNN_OP(ctx, Div, acl_src.get(), acl_div_input, acl_dst.get());
 }
 
 void ggml_cann_cross_entropy_loss(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
@@ -482,56 +596,30 @@ void ggml_cann_cross_entropy_loss(ggml_backend_cann_context & ctx, ggml_tensor *
     logits_nb[1]              = logits_nb[0] * logits_ne[0];
     acl_tensor_ptr acl_logits = ggml_cann_create_tensor(src0->data, ACL_FLOAT, sizeof(float), logits_ne, logits_nb, 2);
 
-    size_t               log_softmax_type_size = sizeof(float);
-    int64_t              log_softmax_n_bytes   = nr * nc * log_softmax_type_size;
-    ggml_cann_pool_alloc log_softmax_allocator(ctx.pool(), log_softmax_n_bytes);
-    void *               log_softmax_buffer = log_softmax_allocator.get();
-
-    int64_t log_softmax_ne[] = { nc, nr };
-    size_t  log_softmax_nb[2];
-    log_softmax_nb[0]              = log_softmax_type_size;
-    log_softmax_nb[1]              = log_softmax_nb[0] * log_softmax_ne[0];
-    acl_tensor_ptr acl_log_softmax = ggml_cann_create_tensor(log_softmax_buffer, ACL_FLOAT, log_softmax_type_size,
-                                                             log_softmax_ne, log_softmax_nb, 2);
-
-    GGML_CANN_CALL_ACLNN_OP(ctx, LogSoftmax, acl_logits.get(), 1, acl_log_softmax.get());
-
     int64_t labels_ne[] = { nc, nr };
     size_t  labels_nb[2];
     labels_nb[0]              = ggml_type_size(src1->type);
     labels_nb[1]              = labels_nb[0] * labels_ne[0];
     acl_tensor_ptr acl_labels = ggml_cann_create_tensor(src1->data, ACL_FLOAT, sizeof(float), labels_ne, labels_nb, 2);
 
-    size_t               mul_type_size = sizeof(float);
-    int64_t              mul_n_bytes   = nr * nc * mul_type_size;
-    ggml_cann_pool_alloc mul_allocator(ctx.pool(), mul_n_bytes);
-    void *               mul_buffer = mul_allocator.get();
+    size_t               loss_per_sample_type_size = sizeof(float);
+    int64_t              loss_per_sample_n_bytes   = nr * loss_per_sample_type_size;
+    ggml_cann_pool_alloc loss_per_sample_allocator(ctx.pool(), loss_per_sample_n_bytes);
+    void *               loss_per_sample_buffer = loss_per_sample_allocator.get();
 
-    int64_t mul_ne[] = { nc, nr };
-    size_t  mul_nb[2];
-    mul_nb[0]                     = mul_type_size;
-    mul_nb[1]                     = mul_nb[0] * mul_ne[0];
-    acl_tensor_ptr acl_mul_result = ggml_cann_create_tensor(mul_buffer, ACL_FLOAT, mul_type_size, mul_ne, mul_nb, 2);
+    int64_t loss_per_sample_ne[] = { nr };
+    size_t  loss_per_sample_nb[1];
+    loss_per_sample_nb[0] = loss_per_sample_type_size;
+    acl_tensor_ptr acl_loss_per_sample = ggml_cann_create_tensor(
+        loss_per_sample_buffer, ACL_FLOAT, loss_per_sample_type_size, loss_per_sample_ne, loss_per_sample_nb, 1);
 
-    GGML_CANN_CALL_ACLNN_OP(ctx, Mul, acl_log_softmax.get(), acl_labels.get(), acl_mul_result.get());
+    size_t               backprop_n_bytes = nr * nc * sizeof(float);
+    ggml_cann_pool_alloc backprop_allocator(ctx.pool(), backprop_n_bytes);
+    void *               backprop_buffer = backprop_allocator.get();
+    acl_tensor_ptr acl_backprop = ggml_cann_create_tensor(backprop_buffer, ACL_FLOAT, sizeof(float), logits_ne, logits_nb, 2);
 
-    size_t               sum_per_sample_type_size = sizeof(float);
-    int64_t              sum_per_sample_n_bytes   = nr * sum_per_sample_type_size;
-    ggml_cann_pool_alloc sum_per_sample_allocator(ctx.pool(), sum_per_sample_n_bytes);
-    void *               sum_per_sample_buffer = sum_per_sample_allocator.get();
-
-    int64_t sum_per_sample_ne[] = { nr };
-    size_t  sum_per_sample_nb[1];
-    sum_per_sample_nb[0]              = sum_per_sample_type_size;
-    acl_tensor_ptr acl_sum_per_sample = ggml_cann_create_tensor(
-        sum_per_sample_buffer, ACL_FLOAT, sum_per_sample_type_size, sum_per_sample_ne, sum_per_sample_nb, 1);
-
-    std::vector<int64_t> sum_dims   = { 1 };
-    acl_int_array_ptr    dims_array = ggml_cann_create_int_array(sum_dims.data(), sum_dims.size());
-    bool                 keep_dims  = false;
-
-    GGML_CANN_CALL_ACLNN_OP(ctx, ReduceSum, acl_mul_result.get(), dims_array.get(), keep_dims, ACL_FLOAT,
-                            acl_sum_per_sample.get());
+    GGML_CANN_CALL_ACLNN_OP(ctx, SoftmaxCrossEntropyWithLogits, acl_logits.get(), acl_labels.get(),
+                            acl_loss_per_sample.get(), acl_backprop.get());
 
     size_t               total_sum_type_size = sizeof(float);
     int64_t              total_sum_n_bytes   = 1 * total_sum_type_size;
@@ -547,11 +635,12 @@ void ggml_cann_cross_entropy_loss(ggml_backend_cann_context & ctx, ggml_tensor *
 
     std::vector<int64_t> total_sum_dims    = { 0 };
     acl_int_array_ptr total_sum_dims_array = ggml_cann_create_int_array(total_sum_dims.data(), total_sum_dims.size());
+    bool              keep_dims            = false;
 
-    GGML_CANN_CALL_ACLNN_OP(ctx, ReduceSum, acl_sum_per_sample.get(), total_sum_dims_array.get(), keep_dims, ACL_FLOAT,
+    GGML_CANN_CALL_ACLNN_OP(ctx, ReduceSum, acl_loss_per_sample.get(), total_sum_dims_array.get(), keep_dims, ACL_FLOAT,
                             acl_total_sum.get());
 
-    float          value        = -1.0f / static_cast<float>(nr);
+    float          value        = 1.0f / static_cast<float>(nr);
     acl_scalar_ptr scale_factor = ggml_cann_create_scalar(&value, aclDataType::ACL_FLOAT);
     acl_tensor_ptr acl_dst =
         ggml_cann_create_tensor(dst->data, ACL_FLOAT, sizeof(float), total_sum_ne, total_sum_nb, 1);
@@ -589,6 +678,33 @@ void ggml_cann_group_norm(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
                             acl_mean_out.get(), acl_rstd_out.get());
 }
 
+void ggml_cann_set(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    ggml_tensor * src0 = dst->src[0];
+    ggml_tensor * src1 = dst->src[1];
+
+    size_t nb1     = ((int32_t *) dst->op_params)[0];
+    size_t nb2     = ((int32_t *) dst->op_params)[1];
+    size_t nb3     = ((int32_t *) dst->op_params)[2];
+    size_t offset  = ((int32_t *) dst->op_params)[3];
+    bool   inplace = (bool) ((int32_t *) dst->op_params)[4];
+
+    size_t param_nb[] = { ggml_element_size(src0), nb1, nb2, nb3 };
+
+    // Create a view of dst at the target offset with src1's dimensions
+    acl_tensor_ptr acl_dst  = ggml_cann_create_tensor(dst, src1->ne, param_nb, GGML_MAX_DIMS, ACL_FORMAT_ND, offset);
+    acl_tensor_ptr acl_src1 = ggml_cann_create_tensor(src1);
+
+    if (!inplace) {
+        // First copy src0 to dst entirely
+        size_t cpy_size = ggml_nbytes(dst);
+        ACL_CHECK(
+            aclrtMemcpyAsync(dst->data, cpy_size, src0->data, cpy_size, ACL_MEMCPY_DEVICE_TO_DEVICE, ctx.stream()));
+    }
+
+    // Copy src1 into the target region of dst
+    GGML_CANN_CALL_ACLNN_OP(ctx, InplaceCopy, acl_dst.get(), acl_src1.get());
+}
+
 void ggml_cann_acc(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
     ggml_tensor * src0 = dst->src[0];
     ggml_tensor * src1 = dst->src[1];
@@ -652,6 +768,113 @@ void ggml_cann_sum(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
     aclnn_reduce_sum(ctx, dst, reduce_dims, 4);
 }
 
+void ggml_cann_cumsum(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    ggml_tensor * src = dst->src[0];
+    acl_tensor_ptr acl_src = ggml_cann_create_tensor(src);
+    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst);
+    // GGML cumsum operates along dim 0 (innermost / ne[0]).
+    // ggml_cann_create_tensor reverses dimensions to [ne3,ne2,ne1,ne0],
+    // so GGML dim 0 maps to CANN dim 3 (the last dim of the 4-D tensor).
+    GGML_CANN_CALL_ACLNN_OP(ctx, Cumsum, acl_src.get(), (int64_t)3,
+                            ggml_cann_type_mapping(dst->type), acl_dst.get());
+}
+
+void ggml_cann_solve_tri(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    ggml_tensor * src0 = dst->src[0];  // A: [N, N, B2, B3] lower triangular
+    ggml_tensor * src1 = dst->src[1];  // B: [K, N, B2, B3]
+
+    acl_tensor_ptr acl_a = ggml_cann_create_tensor(src0);
+    acl_tensor_ptr acl_b = ggml_cann_create_tensor(src1);
+    acl_tensor_ptr acl_x = ggml_cann_create_tensor(dst);
+
+    // mOut: triangular copy of A (required output), same shape as A.
+    const size_t a_bytes = ggml_nbytes(src0);
+    ggml_cann_pool_alloc m_alloc(ctx.pool(), a_bytes);
+    acl_tensor_ptr acl_m = ggml_cann_create_tensor(
+        m_alloc.get(), ggml_cann_type_mapping(src0->type),
+        ggml_type_size(src0->type), src0->ne, src0->nb, GGML_MAX_DIMS);
+
+    // Solve AX = B: upper=false (lower tri), transpose=false, unitriangular=false.
+    GGML_CANN_CALL_ACLNN_OP(ctx, TriangularSolve,
+        acl_b.get(), acl_a.get(), false, false, false,
+        acl_x.get(), acl_m.get());
+}
+
+void ggml_cann_diag(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    ggml_tensor * src = dst->src[0];
+
+    GGML_ASSERT(src->ne[1] == 1);
+
+    const int64_t N       = src->ne[0];
+    const int64_t n_batch = src->ne[2] * src->ne[3];
+    const size_t  nb_f32  = sizeof(float);
+
+    // Fill dst with zeros.
+    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst);
+    {
+        float          zero = 0.0f;
+        acl_scalar_ptr acl_zero = ggml_cann_create_scalar(&zero, ACL_FLOAT);
+        GGML_CANN_CALL_ACLNN_OP(ctx, InplaceFillScalar, acl_dst.get(), acl_zero.get());
+    }
+
+    // Copy src vector onto the diagonal of dst via strided views.
+    // src viewed as [N, n_batch], contiguous strides.
+    int64_t ne_vec[2]      = { N, n_batch };
+    size_t  nb_src_vec[2]  = { nb_f32, N * nb_f32 };
+    // dst diagonal view: stride (N+1)*4 steps along the diagonal.
+    size_t  nb_dst_diag[2] = { (N + 1) * nb_f32, N * N * nb_f32 };
+
+    acl_tensor_ptr acl_src_vec  = ggml_cann_create_tensor(src->data, ACL_FLOAT, nb_f32, ne_vec, nb_src_vec, 2);
+    acl_tensor_ptr acl_dst_diag = ggml_cann_create_tensor(dst->data, ACL_FLOAT, nb_f32, ne_vec, nb_dst_diag, 2);
+
+    GGML_CANN_CALL_ACLNN_OP(ctx, InplaceCopy, acl_dst_diag.get(), acl_src_vec.get());
+}
+
+void ggml_cann_fill(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    float c = ggml_get_op_params_f32(dst, 0);
+
+    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst);
+    acl_scalar_ptr acl_c   = ggml_cann_create_scalar(&c, ACL_FLOAT);
+    GGML_CANN_CALL_ACLNN_OP(ctx, InplaceFillScalar, acl_dst.get(), acl_c.get());
+}
+
+void ggml_cann_tri(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    ggml_tensor * src = dst->src[0];
+
+    const int64_t S       = src->ne[0];
+    const int64_t n_batch = src->ne[2] * src->ne[3];
+    const size_t  nb_f32  = sizeof(float);
+
+    int64_t ne3d[3] = { S, S, n_batch };
+    size_t  nb3d[3] = { nb_f32, S * nb_f32, S * S * nb_f32 };
+
+    const ggml_tri_type ttype = (ggml_tri_type) ggml_get_op_params_i32(dst, 0);
+
+    acl_tensor_ptr acl_src = ggml_cann_create_tensor(src->data, ACL_FLOAT, nb_f32, ne3d, nb3d, 3);
+    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst->data, ACL_FLOAT, nb_f32, ne3d, nb3d, 3);
+
+    switch (ttype) {
+        case GGML_TRI_TYPE_LOWER:
+            // Tril(-1): preserve row > col (strict lower), zero upper + diagonal.
+            GGML_CANN_CALL_ACLNN_OP(ctx, Tril, acl_src.get(), (int64_t)-1, acl_dst.get());
+            break;
+        case GGML_TRI_TYPE_UPPER_DIAG:
+            // Triu(0): preserve row <= col (upper + diagonal), zero strict lower.
+            GGML_CANN_CALL_ACLNN_OP(ctx, Triu, acl_src.get(), (int64_t)0, acl_dst.get());
+            break;
+        case GGML_TRI_TYPE_UPPER:
+            // Triu(1): preserve row < col (strict upper), zero lower + diagonal.
+            GGML_CANN_CALL_ACLNN_OP(ctx, Triu, acl_src.get(), (int64_t)1, acl_dst.get());
+            break;
+        case GGML_TRI_TYPE_LOWER_DIAG:
+            // Tril(0): preserve row >= col (lower + diagonal), zero strict upper.
+            GGML_CANN_CALL_ACLNN_OP(ctx, Tril, acl_src.get(), (int64_t)0, acl_dst.get());
+            break;
+        default:
+            GGML_ABORT("unsupported tri type");
+    }
+}
+
 void ggml_cann_upsample_nearest2d(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
     ggml_tensor *  src     = dst->src[0];
     acl_tensor_ptr acl_src = ggml_cann_create_tensor(src, nullptr, nullptr, 0, ACL_FORMAT_NCHW);
@@ -1695,152 +1918,90 @@ void ggml_cann_softmax(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
     aclnn_softmax(ctx, softmax_tensor.get(), 3, acl_dst.get());
 }
 
-/**
- * @brief Performs index select operation on a 4D tensor using the CANN backend.
- *
- * This function applies the `IndexSelect` operation along a specific dimension
- * of the source tensor (`src_buffer`) using the indices from the index tensor (`index`).
- * It iterates over the last two dimensions of the source tensor, creates the corresponding
- * CANN tensors for the source, index, and output slices, and executes the `IndexSelect`
- * operation for each slice.
- *
- * @param ctx The context for CANN backend operations.
- * @param src_buffer The source buffer containing the 4D input tensor data.
- * @param src_ne The dimensions of the source tensor.
- * @param src_nb The strides (byte offsets) of the source tensor.
- * @param dst_buffer The destination buffer where the output tensor data will be written.
- * @param dst_ne The dimensions of the destination tensor.
- * @param dst_nb The strides (byte offsets) of the destination tensor.
- * @param index The index tensor specifying the indices to select from the source tensor.
- * @param type The data type of the source and destination tensors.
- */
-static void aclnn_index_select_4d(ggml_backend_cann_context & ctx,
-                                  void *                      src_buffer,
-                                  int64_t *                   src_ne,
-                                  size_t *                    src_nb,
-                                  void *                      dst_buffer,
-                                  int64_t *                   dst_ne,
-                                  size_t *                    dst_nb,
-                                  ggml_tensor *               index,
-                                  ggml_type                   type) {
-    for (int64_t i = 0; i < src_ne[3]; i++) {
-        for (int64_t j = 0; j < src_ne[2]; j++) {
-            // src
-            acl_tensor_ptr acl_src_tensor =
-                ggml_cann_create_tensor((char *) src_buffer + i * src_nb[3] + j * src_nb[2],
-                                        ggml_cann_type_mapping(type), ggml_type_size(type), src_ne, src_nb, 2);
-
-            // index
-            acl_tensor_ptr acl_index = ggml_cann_create_tensor(
-                (char *) index->data + (i % index->ne[2]) * index->nb[2] + (j % index->ne[1]) * index->nb[1],
-                ggml_cann_type_mapping(index->type), ggml_element_size(index), index->ne, index->nb, 1);
-
-            // out
-            acl_tensor_ptr acl_out =
-                ggml_cann_create_tensor((char *) dst_buffer + i * dst_nb[3] + j * dst_nb[2],
-                                        ggml_cann_type_mapping(type), ggml_type_size(type), dst_ne, dst_nb, 2);
-            GGML_CANN_CALL_ACLNN_OP(ctx, IndexSelect, acl_src_tensor.get(), 0, acl_index.get(), acl_out.get());
-        }
-    }
-}
-
-/**
- * @brief Performs inplace index copy operation on a 4D tensor using the CANN backend.
- *
- * This function applies the `IndexCopy` operation along a specific dimension of the
- * destination tensor (`dst_buffer`) by copying elements from the source tensor (`src_buffer`)
- * to positions specified by the index tensor (`index`).
- * It iterates over the last two dimensions of the tensors, creates the corresponding
- * CANN tensors for source, index, and destination slices, and performs the index copy
- * operation for each slice.
- *
- * @param ctx The context for CANN backend operations.
- * @param src_buffer The source buffer containing the 4D input tensor data to be copied.
- * @param src_ne The dimensions of the source tensor.
- * @param src_nb The strides (byte offsets) of the source tensor.
- * @param dst_buffer The destination buffer where values will be copied to.
- * @param dst_ne The dimensions of the destination tensor.
- * @param dst_nb The strides (byte offsets) of the destination tensor.
- * @param index The index tensor specifying target positions in the destination tensor.
- * @param type The data type of the source and destination tensors.
- */
-static void aclnn_index_copy_4d(ggml_backend_cann_context & ctx,
-                                void *                      src_buffer,
-                                int64_t *                   src_ne,
-                                size_t *                    src_nb,
-                                void *                      dst_buffer,
-                                int64_t *                   dst_ne,
-                                size_t *                    dst_nb,
-                                ggml_tensor *               index,
-                                ggml_type                   type) {
-    for (int64_t i = 0; i < src_ne[3]; i++) {
-        for (int64_t j = 0; j < src_ne[2]; j++) {
-            // src
-            acl_tensor_ptr acl_src_tensor =
-                ggml_cann_create_tensor((char *) src_buffer + i * src_nb[3] + j * src_nb[2],
-                                        ggml_cann_type_mapping(type), ggml_type_size(type), src_ne, src_nb, 2);
-
-            // index
-            acl_tensor_ptr acl_index = ggml_cann_create_tensor(
-                (char *) index->data + (i % index->ne[2]) * index->nb[2] + (j % index->ne[1]) * index->nb[1],
-                ggml_cann_type_mapping(index->type), ggml_element_size(index), index->ne, index->nb, 1);
-
-            // out
-            acl_tensor_ptr acl_out =
-                ggml_cann_create_tensor((char *) dst_buffer + i * dst_nb[3] + j * dst_nb[2],
-                                        ggml_cann_type_mapping(type), ggml_type_size(type), dst_ne, dst_nb, 2);
-            GGML_CANN_CALL_ACLNN_OP(ctx, InplaceIndexCopy, acl_out.get(), 0, acl_index.get(), acl_src_tensor.get());
-        }
-    }
-}
 
 void ggml_cann_get_rows(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
-    ggml_tensor * src0 = dst->src[0];  // src
+    ggml_tensor * src0 = dst->src[0];  // weight
     ggml_tensor * src1 = dst->src[1];  // index
 
     GGML_ASSERT(dst->type == GGML_TYPE_F32 || dst->type == GGML_TYPE_F16
                 || dst->type == GGML_TYPE_BF16);
 
+    // n_idx: number of row indices per (i2, i3) batch slice.
+    // ggml guarantees: src0->ne[2] == src1->ne[1], src0->ne[3] == src1->ne[2], src1->ne[3] == 1.
+    const int64_t n_idx = src1->ne[0];
+
+    // Gather all (i2, i3) batch slices from src into dst.
+    // ggml_cann_create_tensor reverses dims, so ACL sees [ne1, ne0].
+    // GatherV2 with dim=0 gathers along ACL dim-0 == ggml ne[1] (the vocabulary / row axis).
+    // nb: the 4 strides of the source buffer (nb[0..1] for the 2D slice shape,
+    //     nb[2..3] for computing per-batch-slice base pointer offsets).
+    auto gather_batched = [&](void * src_base, aclDataType acl_type, size_t type_size,
+                              const size_t * nb) {
+        int64_t src_ne[2]  = { src0->ne[0], src0->ne[1] };
+        size_t  src_nb_2d[2] = { nb[0], nb[1] };
+        int64_t dst_ne[2]  = { src0->ne[0], n_idx };
+        size_t  dst_nb_2d[2] = { dst->nb[0], dst->nb[1] };
+        int64_t idx_ne[1]  = { n_idx };
+        size_t  idx_nb[1]  = { (size_t)ggml_element_size(src1) };
+
+        for (int64_t i3 = 0; i3 < src0->ne[3]; i3++) {
+            for (int64_t i2 = 0; i2 < src0->ne[2]; i2++) {
+                acl_tensor_ptr acl_src = ggml_cann_create_tensor(
+                    (char *)src_base + i3 * nb[3] + i2 * nb[2],
+                    acl_type, type_size, src_ne, src_nb_2d, 2);
+                acl_tensor_ptr acl_idx = ggml_cann_create_tensor(
+                    (char *)src1->data + i3 * src1->nb[2] + i2 * src1->nb[1],
+                    ggml_cann_type_mapping(src1->type), (size_t)ggml_element_size(src1),
+                    idx_ne, idx_nb, 1);
+                acl_tensor_ptr acl_dst = ggml_cann_create_tensor(
+                    (char *)dst->data + i3 * dst->nb[3] + i2 * dst->nb[2],
+                    acl_type, type_size, dst_ne, dst_nb_2d, 2);
+                GGML_CANN_CALL_ACLNN_OP(ctx, GatherV2, acl_src.get(), 0, acl_idx.get(), acl_dst.get());
+            }
+        }
+    };
+
     switch (src0->type) {
         case GGML_TYPE_BF16:
         case GGML_TYPE_F16:
         case GGML_TYPE_F32:
             if (src0->type == dst->type) {
-                aclnn_index_select_4d(ctx, src0->data, src0->ne, src0->nb, dst->data, dst->ne, dst->nb, src1,
-                                      dst->type);
+                gather_batched(src0->data,
+                               ggml_cann_type_mapping(src0->type), ggml_type_size(src0->type),
+                               src0->nb);
             } else {
-                acl_tensor_ptr       acl_src0 = ggml_cann_create_tensor(src0);
-                ggml_cann_pool_alloc src_buffer_allocator(ctx.pool(), ggml_nelements(src0) * ggml_element_size(dst));
-                void *               src_trans_buffer = src_buffer_allocator.get();
-                size_t               src_trans_nb[GGML_MAX_DIMS];
-                src_trans_nb[0] = dst->nb[0];
+                // Cast src0 to dst type, then gather.
+                ggml_cann_pool_alloc src_cast_allocator(ctx.pool(),
+                                                        ggml_nelements(src0) * ggml_element_size(dst));
+                size_t src_cast_nb[GGML_MAX_DIMS];
+                src_cast_nb[0] = ggml_type_size(dst->type);
                 for (int i = 1; i < GGML_MAX_DIMS; i++) {
-                    src_trans_nb[i] = src_trans_nb[i - 1] * src0->ne[i - 1];
+                    src_cast_nb[i] = src_cast_nb[i - 1] * src0->ne[i - 1];
                 }
-                acl_tensor_ptr src_trans_tensor =
-                    ggml_cann_create_tensor(src_trans_buffer, ggml_cann_type_mapping(dst->type),
-                                            ggml_type_size(dst->type), src0->ne, src_trans_nb, GGML_MAX_DIMS);
-                aclnn_cast(ctx, acl_src0.get(), src_trans_tensor.get(), ggml_cann_type_mapping(dst->type));
-                aclnn_index_select_4d(ctx, src_trans_buffer, src0->ne, src_trans_nb, dst->data, dst->ne, dst->nb, src1,
-                                      dst->type);
+                acl_tensor_ptr acl_src0     = ggml_cann_create_tensor(src0);
+                acl_tensor_ptr acl_src_cast = ggml_cann_create_tensor(
+                    src_cast_allocator.get(), ggml_cann_type_mapping(dst->type), ggml_type_size(dst->type),
+                    src0->ne, src_cast_nb, GGML_MAX_DIMS);
+                aclnn_cast(ctx, acl_src0.get(), acl_src_cast.get(), ggml_cann_type_mapping(dst->type));
+
+                gather_batched(src_cast_allocator.get(),
+                               ggml_cann_type_mapping(dst->type), ggml_type_size(dst->type),
+                               src_cast_nb);
             }
             break;
         case GGML_TYPE_Q8_0:
             {
-                // add 1 dim for bcast mul.
+                // Dequantize Q8_0 to dst type, then gather.
                 size_t  weight_nb[GGML_MAX_DIMS + 1], scale_nb[GGML_MAX_DIMS + 1], dequant_nb[GGML_MAX_DIMS + 1];
                 int64_t weight_ne[GGML_MAX_DIMS + 1], scale_ne[GGML_MAX_DIMS + 1], *dequant_ne;
-                int64_t scale_offset = 0;
-                // [3,4,5,64] -> [3,4,5,2,32]
-                weight_ne[0]         = QK8_0;
-                weight_ne[1]         = src0->ne[0] / QK8_0;
-                weight_nb[0]         = sizeof(int8_t);
-                weight_nb[1]         = weight_nb[0] * weight_ne[0];
+                weight_ne[0] = QK8_0;
+                weight_ne[1] = src0->ne[0] / QK8_0;
+                weight_nb[0] = sizeof(int8_t);
+                weight_nb[1] = weight_nb[0] * weight_ne[0];
                 for (int i = 2; i < GGML_MAX_DIMS + 1; i++) {
                     weight_ne[i] = src0->ne[i - 1];
                     weight_nb[i] = weight_nb[i - 1] * weight_ne[i - 1];
                 }
-                // [3,4,5,64] -> [3,4,5,2,1]
                 scale_ne[0] = 1;
                 scale_ne[1] = src0->ne[0] / QK8_0;
                 scale_nb[0] = sizeof(uint16_t);
@@ -1849,31 +2010,33 @@ void ggml_cann_get_rows(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
                     scale_ne[i] = src0->ne[i - 1];
                     scale_nb[i] = scale_nb[i - 1] * scale_ne[i - 1];
                 }
-                // [3,4,5,64] -> [3,4,5,2,32]
                 dequant_ne    = weight_ne;
                 dequant_nb[0] = ggml_type_size(dst->type);
                 for (int i = 1; i < GGML_MAX_DIMS + 1; i++) {
                     dequant_nb[i] = dequant_nb[i - 1] * dequant_ne[i - 1];
                 }
-                scale_offset = ggml_nelements(src0) * sizeof(int8_t);
-                ggml_cann_pool_alloc dequant_buffer_allocator(ctx.pool(),
-                                                              ggml_nelements(src0) * ggml_type_size(dst->type));
-                acl_tensor_ptr       acl_weight_tensor = ggml_cann_create_tensor(src0->data, ACL_INT8, sizeof(int8_t),
-                                                                                 weight_ne, weight_nb, GGML_MAX_DIMS + 1);
-                acl_tensor_ptr       acl_scale_tensor =
-                    ggml_cann_create_tensor(src0->data, ACL_FLOAT16, sizeof(uint16_t), scale_ne, scale_nb,
-                                            GGML_MAX_DIMS + 1, ACL_FORMAT_ND, scale_offset);
-                acl_tensor_ptr dequant_tensor =
-                    ggml_cann_create_tensor(dequant_buffer_allocator.get(), ggml_cann_type_mapping(dst->type),
-                                            ggml_type_size(dst->type), dequant_ne, dequant_nb, GGML_MAX_DIMS + 1);
-                aclnn_mul(ctx, acl_weight_tensor.get(), acl_scale_tensor.get(), dequant_tensor.get());
-                dequant_nb[0] = ggml_type_size(dst->type);
+                const int64_t scale_offset = ggml_nelements(src0) * sizeof(int8_t);
+                ggml_cann_pool_alloc dequant_allocator(ctx.pool(),
+                                                       ggml_nelements(src0) * ggml_type_size(dst->type));
+                acl_tensor_ptr acl_weight = ggml_cann_create_tensor(src0->data, ACL_INT8, sizeof(int8_t),
+                                                                     weight_ne, weight_nb, GGML_MAX_DIMS + 1);
+                acl_tensor_ptr acl_scale  = ggml_cann_create_tensor(
+                    src0->data, ACL_FLOAT16, sizeof(uint16_t), scale_ne, scale_nb,
+                    GGML_MAX_DIMS + 1, ACL_FORMAT_ND, scale_offset);
+                acl_tensor_ptr acl_dequant = ggml_cann_create_tensor(
+                    dequant_allocator.get(), ggml_cann_type_mapping(dst->type),
+                    ggml_type_size(dst->type), dequant_ne, dequant_nb, GGML_MAX_DIMS + 1);
+                aclnn_mul(ctx, acl_weight.get(), acl_scale.get(), acl_dequant.get());
+
+                // Reinterpret dequant buffer as 4D [src0->ne] with contiguous strides.
                 dequant_ne    = src0->ne;
+                dequant_nb[0] = ggml_type_size(dst->type);
                 for (int i = 1; i < GGML_MAX_DIMS; i++) {
                     dequant_nb[i] = dequant_nb[i - 1] * src0->ne[i - 1];
                 }
-                aclnn_index_select_4d(ctx, dequant_buffer_allocator.get(), dequant_ne, dequant_nb, dst->data, dst->ne,
-                                      dst->nb, src1, dst->type);
+                gather_batched(dequant_allocator.get(),
+                               ggml_cann_type_mapping(dst->type), ggml_type_size(dst->type),
+                               dequant_nb);
                 break;
             }
         default:
@@ -1883,31 +2046,70 @@ void ggml_cann_get_rows(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
 }
 
 void ggml_cann_set_rows(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
-    ggml_tensor * src0 = dst->src[0];  // src
-    ggml_tensor * src1 = dst->src[1];  // index
+    ggml_tensor * src0 = dst->src[0];  // source values
+    ggml_tensor * src1 = dst->src[1];  // row indices
+
+    // n_idx: number of source rows to scatter per batch slice.
+    // ggml guarantees: src0->ne[1] == src1->ne[0].
+    const int64_t n_idx = src1->ne[0];
+
+    // Copy n_idx rows of src [ne0, n_idx] into dst [ne0, ne1] at positions given by a 1D index.
+    // ggml_cann_create_tensor reverses dims, so ACL sees [ne1, ne0] for dst.
+    // InplaceIndexCopy with dim=0 copies along ACL dim-0 == ggml ne[1] (the row axis).
+    // src_nb: the 4 strides of the source buffer (nb[0..1] for the 2D slice shape,
+    //         nb[2..3] for computing per-batch-slice base pointer offsets).
+    auto scatter_batched = [&](void * src_base, aclDataType acl_type, size_t type_size,
+                               const size_t * src_nb) {
+        int64_t d_ne[2]    = { dst->ne[0], dst->ne[1] };
+        size_t  d_nb[2]    = { dst->nb[0], dst->nb[1] };
+        int64_t s_ne[2]    = { dst->ne[0], n_idx };
+        size_t  s_nb_2d[2] = { src_nb[0], src_nb[1] };
+        int64_t i_ne[1]    = { n_idx };
+        size_t  i_nb[1]    = { (size_t)ggml_element_size(src1) };
+
+        for (int64_t i3 = 0; i3 < dst->ne[3]; i3++) {
+            for (int64_t i2 = 0; i2 < dst->ne[2]; i2++) {
+                acl_tensor_ptr acl_dst = ggml_cann_create_tensor(
+                    (char *)dst->data + i3 * dst->nb[3] + i2 * dst->nb[2],
+                    acl_type, type_size, d_ne, d_nb, 2);
+                acl_tensor_ptr acl_idx = ggml_cann_create_tensor(
+                    (char *)src1->data + (i3 % src1->ne[2]) * src1->nb[2] + (i2 % src1->ne[1]) * src1->nb[1],
+                    ggml_cann_type_mapping(src1->type), (size_t)ggml_element_size(src1),
+                    i_ne, i_nb, 1);
+                acl_tensor_ptr acl_src = ggml_cann_create_tensor(
+                    (char *)src_base + i3 * src_nb[3] + i2 * src_nb[2],
+                    acl_type, type_size, s_ne, s_nb_2d, 2);
+                GGML_CANN_CALL_ACLNN_OP(ctx, InplaceIndexCopy, acl_dst.get(), 0, acl_idx.get(), acl_src.get());
+            }
+        }
+    };
 
     switch (dst->type) {
         case GGML_TYPE_F32:
-            {
-                aclnn_index_copy_4d(ctx, src0->data, src0->ne, src0->nb, dst->data, dst->ne, dst->nb, src1, dst->type);
-                break;
-            }
+            scatter_batched(src0->data,
+                            ggml_cann_type_mapping(dst->type), ggml_type_size(dst->type),
+                            src0->nb);
+            break;
         case GGML_TYPE_F16:
         case GGML_TYPE_BF16:
             {
-                acl_tensor_ptr       acl_src0 = ggml_cann_create_tensor(src0);
-                ggml_cann_pool_alloc src_buffer_allocator(ctx.pool(), ggml_nelements(src0) * sizeof(uint16_t));
-                void *               src_trans_buffer = src_buffer_allocator.get();
-                size_t               src_trans_nb[GGML_MAX_DIMS];
-                src_trans_nb[0] = sizeof(uint16_t);
+                // Cast src0 (F32) to dst type first.
+                ggml_cann_pool_alloc src_cast_allocator(ctx.pool(),
+                                                        ggml_nelements(src0) * ggml_type_size(dst->type));
+                size_t src_cast_nb[GGML_MAX_DIMS];
+                src_cast_nb[0] = ggml_type_size(dst->type);
                 for (int i = 1; i < GGML_MAX_DIMS; i++) {
-                    src_trans_nb[i] = src_trans_nb[i - 1] * src0->ne[i - 1];
+                    src_cast_nb[i] = src_cast_nb[i - 1] * src0->ne[i - 1];
                 }
-                acl_tensor_ptr src_trans_tensor = ggml_cann_create_tensor(
-                    src_trans_buffer, ggml_cann_type_mapping(dst->type), ggml_type_size(dst->type), src0->ne, src_trans_nb, GGML_MAX_DIMS);
-                aclnn_cast(ctx, acl_src0.get(), src_trans_tensor.get(), ggml_cann_type_mapping(dst->type));
-                aclnn_index_copy_4d(ctx, src_trans_buffer, src0->ne, src_trans_nb, dst->data, dst->ne, dst->nb, src1,
-                                    dst->type);
+                acl_tensor_ptr acl_src0     = ggml_cann_create_tensor(src0);
+                acl_tensor_ptr acl_src_cast = ggml_cann_create_tensor(
+                    src_cast_allocator.get(), ggml_cann_type_mapping(dst->type), ggml_type_size(dst->type),
+                    src0->ne, src_cast_nb, GGML_MAX_DIMS);
+                aclnn_cast(ctx, acl_src0.get(), acl_src_cast.get(), ggml_cann_type_mapping(dst->type));
+
+                scatter_batched(src_cast_allocator.get(),
+                                ggml_cann_type_mapping(dst->type), ggml_type_size(dst->type),
+                                src_cast_nb);
                 break;
             }
         default:
@@ -3268,29 +3470,50 @@ void ggml_cann_pad_reflect_1d(ggml_backend_cann_context & ctx, ggml_tensor * dst
     int64_t           paddingsArray[2] = { opts[0], opts[1] };
     acl_int_array_ptr paddings         = ggml_cann_create_int_array(paddingsArray, 2);
 
-    for (int64_t i = 0; i < src0->ne[3]; i++) {
-        acl_tensor_ptr acl_src =
-            ggml_cann_create_tensor((char *) src0->data + i * src0->ne[3], ggml_cann_type_mapping(src0->type),
-                                    ggml_element_size(src0), src0->ne, src0->nb, 3);
+    // Collapsing ne[2]*ne[3] into a single batch dimension requires that dim3
+    // is contiguous with respect to dim2 in both src and dst.
+    GGML_ASSERT(src0->nb[3] == src0->nb[2] * src0->ne[2]);
+    GGML_ASSERT(dst->nb[3]  == dst->nb[2]  * dst->ne[2]);
 
-        acl_tensor_ptr acl_dst =
-            ggml_cann_create_tensor((char *) dst->data + i * src0->ne[3], ggml_cann_type_mapping(dst->type),
-                                    ggml_element_size(dst), dst->ne, dst->nb, 3);
+    int64_t src_ne_3d[3] = { src0->ne[0], src0->ne[1], src0->ne[2] * src0->ne[3] };
+    int64_t dst_ne_3d[3] = { dst->ne[0],  dst->ne[1],  dst->ne[2]  * dst->ne[3]  };
 
-        GGML_CANN_CALL_ACLNN_OP(ctx, ReflectionPad1d, acl_src.get(), paddings.get(), acl_dst.get());
-    }
+    acl_tensor_ptr acl_src = ggml_cann_create_tensor(src0->data, ggml_cann_type_mapping(src0->type),
+                                                     ggml_element_size(src0), src_ne_3d, src0->nb, 3);
+
+    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst->data, ggml_cann_type_mapping(dst->type),
+                                                     ggml_element_size(dst), dst_ne_3d, dst->nb, 3);
+
+    GGML_CANN_CALL_ACLNN_OP(ctx, ReflectionPad1d, acl_src.get(), paddings.get(), acl_dst.get());
 }
 
 void ggml_cann_count_equal(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
     ggml_tensor * src0 = dst->src[0];
     ggml_tensor * src1 = dst->src[1];
 
+    // Write element-wise equality (0 or 1) into a temporary buffer to avoid
+    // modifying src0 in-place.  Use the same type as src0 so ReduceSum can
+    // consume it directly without a type cast.
+    ggml_cann_pool_alloc eq_alloc(ctx.pool(), ggml_nelements(src0) * ggml_element_size(src0));
+    size_t eq_nb[GGML_MAX_DIMS];
+    eq_nb[0] = ggml_element_size(src0);
+    for (int i = 1; i < GGML_MAX_DIMS; i++) {
+        eq_nb[i] = eq_nb[i - 1] * src0->ne[i - 1];
+    }
+    acl_tensor_ptr acl_eq = ggml_cann_create_tensor(
+        eq_alloc.get(), ggml_cann_type_mapping(src0->type), ggml_element_size(src0),
+        src0->ne, eq_nb, GGML_MAX_DIMS);
+
     acl_tensor_ptr acl_self  = ggml_cann_create_tensor(src0);
     acl_tensor_ptr acl_other = ggml_cann_create_tensor(src1);
+    GGML_CANN_CALL_ACLNN_OP(ctx, EqTensor, acl_self.get(), acl_other.get(), acl_eq.get());
 
-    GGML_CANN_CALL_ACLNN_OP(ctx, InplaceEqTensor, acl_self.get(), acl_other.get());
-
-    ggml_cann_sum(ctx, dst);
+    // Sum the 0/1 values into dst.
+    acl_tensor_ptr    acl_dst    = ggml_cann_create_tensor(dst);
+    int64_t           dims[4]    = { 0, 1, 2, 3 };
+    acl_int_array_ptr dims_arr   = ggml_cann_create_int_array(dims, 4);
+    GGML_CANN_CALL_ACLNN_OP(ctx, ReduceSum, acl_eq.get(), dims_arr.get(), true,
+                            ggml_cann_type_mapping(dst->type), acl_dst.get());
 }
 
 void ggml_cann_step(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
@@ -3306,6 +3529,27 @@ void ggml_cann_step(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
     GGML_CANN_CALL_ACLNN_OP(ctx, GtScalar, acl_src.get(), alpha.get(), acl_dst.get());
 }
 
+void ggml_cann_softplus(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    ggml_tensor * src0 = dst->src[0];
+
+    acl_tensor_ptr acl_src = ggml_cann_create_tensor(src0);
+    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst);
+
+    float          beta_val      = 1.0f;
+    float          threshold_val = 20.0f;
+    acl_scalar_ptr beta          = ggml_cann_create_scalar(&beta_val,      ACL_FLOAT);
+    acl_scalar_ptr threshold     = ggml_cann_create_scalar(&threshold_val, ACL_FLOAT);
+
+    GGML_CANN_CALL_ACLNN_OP(ctx, Softplus, acl_src.get(), beta.get(), threshold.get(), acl_dst.get());
+}
+
+void ggml_cann_geglu_quick(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
+    auto gelu_quick_fn = [](ggml_backend_cann_context & ctx, aclTensor * acl_src, aclTensor * acl_dst) {
+        GGML_CANN_CALL_ACLNN_OP(ctx, GeluV2, acl_src, 0, acl_dst);
+    };
+    ggml_cann_op_unary_gated(gelu_quick_fn, ctx, dst);
+}
+
 /**
  * @brief Performs expert-specific matrix multiplication (MoE) with
  * floating-point precision using the CANN backend.
@@ -3892,46 +4136,65 @@ void ggml_cann_flash_attn_ext(ggml_backend_cann_context & ctx, ggml_tensor * dst
 }
 
 static void ggml_cann_out_prod_fp(ggml_backend_cann_context & ctx, ggml_tensor * dst) {
-    ggml_tensor * src0 = dst->src[0];  // weight
-    ggml_tensor * src1 = dst->src[1];  // input
+    ggml_tensor * src0 = dst->src[0];  // weight  [ne00=m, ne01=K, ne02, ne03]
+    ggml_tensor * src1 = dst->src[1];  // input   [ne10=n, ne11=K, ne12, ne13]
     GGML_TENSOR_BINARY_OP_LOCALS
 
-    acl_tensor_ptr acl_dst = ggml_cann_create_tensor(dst);
-    GGML_CANN_CALL_ACLNN_OP(ctx, InplaceZero, acl_dst.get());
+    // dst[i,j] = sum_k src0[i,k] * src1[j,k]  i.e. dst = src0 @ src1^T.
+    //
+    // ggml_cann_create_tensor reverses dimension order, so ACL sees:
+    //   acl_src0 slice:   ggml[m,K]  ->  ACL[K,m]
+    //   acl_src1 slice:   ggml[n,K]  ->  ACL[K,n]
+    //   acl_dst  slice:   ggml[m,n]  ->  ACL[n,m]
+    //
+    // Build a transposed view of src1 by swapping ne[0]/ne[1]:
+    //   src1_t:  ggml[K,n] (swapped strides)  ->  ACL[n,K]
+    //
+    // Matmul(src1_t [n,K], src0 [K,m]) = [n,m] = acl_dst  ✓
+    //
+    // The outer batch loop is kept because src0 may have fewer batch slices than
+    // dst (ne02 <= ne2, ne03 <= ne3): this is a strided-broadcast not supported
+    // by standard CANN Matmul broadcasting.
+
+    const aclDataType src0_acl_type = ggml_cann_type_mapping(src0->type);
+    const aclDataType src1_acl_type = ggml_cann_type_mapping(src1->type);
+    const aclDataType dst_acl_type  = ggml_cann_type_mapping(dst->type);
+    const size_t      src0_type_sz  = ggml_type_size(src0->type);
+    const size_t      src1_type_sz  = ggml_type_size(src1->type);
+    const size_t      dst_type_sz   = ggml_type_size(dst->type);
 
     const int64_t dps2 = ne2 / ne02;
     const int64_t dps3 = ne3 / ne03;
+
     for (int64_t i3 = 0; i3 < ne3; i3++) {
         for (int64_t i2 = 0; i2 < ne2; i2++) {
             const int64_t i02 = i2 / dps2;
             const int64_t i03 = i3 / dps3;
 
-            const int64_t  i12 = i2;
-            const int64_t  i13 = i3;
-            acl_tensor_ptr accumulator =
-                ggml_cann_create_tensor((char *) dst->data + i2 * nb2 + i3 * nb3, ggml_cann_type_mapping(dst->type),
-                                        ggml_type_size(dst->type), dst->ne, dst->nb, 2);
+            // src0 2D slice at [i02, i03]: ggml [m, K] -> ACL [K, m]
+            int64_t src0_ne[2] = { ne00, ne01 };
+            size_t  src0_nb[2] = { nb00, nb01 };
+            acl_tensor_ptr acl_src0_s = ggml_cann_create_tensor(
+                (char *) src0->data + i02 * nb02 + i03 * nb03,
+                src0_acl_type, src0_type_sz, src0_ne, src0_nb, 2);
 
-            // The outer product needs to be accumulated in this dimension.
-            for (int64_t i1 = 0; i1 < ne11; i1++) {
-                acl_tensor_ptr acl_input = ggml_cann_create_tensor(
-                    (char *) src1->data + i1 * nb11 + i12 * nb12 + i13 * nb13, ggml_cann_type_mapping(src0->type),
-                    ggml_type_size(src0->type), src1->ne, src1->nb, 1);
+            // src1 transposed 2D slice at [i2, i3]: swap ne/nb -> ggml[K,n] -> ACL[n,K]
+            int64_t src1_t_ne[2] = { ne11, ne10 };
+            size_t  src1_t_nb[2] = { nb11, nb10 };
+            acl_tensor_ptr acl_src1_t = ggml_cann_create_tensor(
+                (char *) src1->data + i2 * nb12 + i3 * nb13,
+                src1_acl_type, src1_type_sz, src1_t_ne, src1_t_nb, 2);
 
-                acl_tensor_ptr acl_weight = ggml_cann_create_tensor(
-                    (char *) src0->data + i1 * nb01 + i02 * nb02 + i03 * nb03, ggml_cann_type_mapping(src0->type),
-                    ggml_type_size(src0->type), src0->ne, src0->nb, 1);
+            // dst 2D slice at [i2, i3]: ggml [m, n] -> ACL [n, m]
+            int64_t dst_ne[2] = { ne0, ne1 };
+            size_t  dst_nb[2] = { nb0, nb1 };
+            acl_tensor_ptr acl_dst_s = ggml_cann_create_tensor(
+                (char *) dst->data + i2 * nb2 + i3 * nb3,
+                dst_acl_type, dst_type_sz, dst_ne, dst_nb, 2);
 
-                ggml_cann_pool_alloc output_allocator(ctx.pool());
-                void *               output_buffer = output_allocator.alloc(ggml_nbytes(dst));
-                acl_tensor_ptr       acl_out = ggml_cann_create_tensor(output_buffer, ggml_cann_type_mapping(dst->type),
-                                                                       ggml_type_size(dst->type), dst->ne, dst->nb, 2);
-
-                GGML_CANN_CALL_ACLNN_OP(ctx, Ger, acl_input.get(), acl_weight.get(), acl_out.get());
-                float       alpha_value = 1.0f;
-                aclScalar * alpha       = aclCreateScalar(&alpha_value, ACL_FLOAT);
-                GGML_CANN_CALL_ACLNN_OP(ctx, InplaceAdd, accumulator.get(), acl_out.get(), alpha);
-            }
+            // Matmul(src1_t [n,K], src0 [K,m]) = [n,m] = acl_dst_s  ✓
+            GGML_CANN_CALL_ACLNN_OP(ctx, Matmul,
+                acl_src1_t.get(), acl_src0_s.get(), acl_dst_s.get(), (int8_t) 1);
         }
     }
 }
@@ -4170,3 +4433,4 @@ void ggml_cann_gated_linear_attn(ggml_backend_cann_context & ctx, ggml_tensor *
         }
     }
 }
+

ggml/src/ggml-cann/aclnn_ops.h

@@ -32,6 +32,9 @@
 #include <aclnnop/aclnn_cat.h>
 #include <aclnnop/aclnn_clamp.h>
 #include <aclnnop/aclnn_cos.h>
+#include <aclnnop/aclnn_cumsum.h>
+#include <aclnnop/aclnn_tril.h>
+#include <aclnnop/aclnn_triu.h>
 #include <aclnnop/aclnn_exp.h>
 #include <aclnnop/aclnn_gelu.h>
 #include <aclnnop/aclnn_gelu_v2.h>
@@ -47,6 +50,9 @@
 #include <aclnnop/aclnn_sign.h>
 #include <aclnnop/aclnn_silu.h>
 #include <aclnnop/aclnn_sin.h>
+#include <aclnnop/aclnn_softplus.h>
+#include <aclnnop/aclnn_swi_glu.h>
+#include <aclnnop/aclnn_geglu.h>
 #include <aclnnop/aclnn_slice.h>
 #include <aclnnop/aclnn_sqrt.h>
 #include <aclnnop/aclnn_tanh.h>
@@ -69,6 +75,9 @@
  */
 void ggml_cann_repeat(ggml_backend_cann_context & ctx, ggml_tensor * dst);
 
+void ggml_cann_swiglu(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+void ggml_cann_geglu(ggml_backend_cann_context & ctx, ggml_tensor * dst, int64_t approximate);
+
 /**
  * @brief   Applies the Leaky ReLU activation function to a tensor using the CANN
  *          backend.
@@ -325,6 +334,48 @@ void ggml_cann_sum_rows(ggml_backend_cann_context & ctx, ggml_tensor * dst);
 
 void ggml_cann_sum(ggml_backend_cann_context & ctx, ggml_tensor * dst);
 
+/**
+ * @brief   Computes the cumulative sum of a ggml tensor along dim 0 using the
+ *          CANN backend.
+ *
+ * @param ctx The CANN context used for operations.
+ * @param dst The destination tensor. dst->op is `GGML_OP_CUMSUM`.
+ */
+void ggml_cann_cumsum(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+
+/**
+ * @brief   Computes a triangular mask (tril/triu) of a square ggml tensor
+ *          using the CANN backend.
+ *
+ * @param ctx The CANN context used for operations.
+ * @param dst The destination tensor. dst->op is `GGML_OP_TRI`.
+ */
+void ggml_cann_tri(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+
+/**
+ * @brief   Solves a triangular linear system AX=B using the CANN backend.
+ *
+ * @param ctx The CANN context used for operations.
+ * @param dst The destination tensor. dst->op is `GGML_OP_SOLVE_TRI`.
+ */
+void ggml_cann_solve_tri(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+
+/**
+ * @brief   Creates a diagonal matrix from a vector using the CANN backend.
+ *
+ * @param ctx The CANN context used for operations.
+ * @param dst The destination tensor. dst->op is `GGML_OP_DIAG`.
+ */
+void ggml_cann_diag(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+
+/**
+ * @brief   Fills a tensor with a constant scalar value using the CANN backend.
+ *
+ * @param ctx The CANN context used for operations.
+ * @param dst The destination tensor. dst->op is `GGML_OP_FILL`.
+ */
+void ggml_cann_fill(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+
 /**
  * @brief   Upsamples a ggml tensor using nearest neighbor interpolation using
  *          the CANN backend.
@@ -461,6 +512,9 @@ void ggml_cann_timestep_embedding(ggml_backend_cann_context & ctx, ggml_tensor *
 // @see ggml_cann_dup.
 void ggml_cann_cpy(ggml_backend_cann_context & ctx, ggml_tensor * dst);
 
+// @see ggml_cann_acc, but copies src1 into dst instead of adding.
+void ggml_cann_set(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+
 /**
  * @brief   Computes the softmax activation with optional masking.
  *
@@ -813,6 +867,8 @@ void ggml_cann_count_equal(ggml_backend_cann_context & ctx, ggml_tensor * dst);
  *            dst->op is expected to be `GGML_OP_STEP`.
  */
 void ggml_cann_step(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+void ggml_cann_softplus(ggml_backend_cann_context & ctx, ggml_tensor * dst);
+void ggml_cann_geglu_quick(ggml_backend_cann_context & ctx, ggml_tensor * dst);
 
 /**
  * @brief   Performs the Flash Attention extended operator using the CANN backend.

ggml/src/ggml-cann/ggml-cann.cpp

@@ -1428,6 +1428,22 @@ static bool ggml_backend_cann_buffer_cpy_tensor(ggml_backend_buffer_t buffer,
     return false;
 }
 
+/**
+ * @brief Set a region of a tensor's device memory to a specified value.
+ *
+ * @param buffer The CANN buffer containing the tensor.
+ * @param tensor Pointer to the tensor whose memory will be set.
+ * @param value The value to which each byte in the region will be set.
+ * @param offset Byte offset within the tensor's data to start setting.
+ * @param size Number of bytes to set.
+ */
+static void ggml_backend_cann_buffer_memset_tensor(ggml_backend_buffer_t buffer, ggml_tensor * tensor, uint8_t value, size_t offset, size_t size) {
+    ggml_backend_cann_buffer_context * ctx = (ggml_backend_cann_buffer_context *) buffer->context;
+
+    ggml_cann_set_device(ctx->device);
+    ACL_CHECK(aclrtMemset((char *) tensor->data + offset, size, value, size));
+}
+
 /**
  * @brief Clear a CANN buffer by setting all its memory to a specified value.
  *
@@ -1454,7 +1470,7 @@ static const ggml_backend_buffer_i ggml_backend_cann_buffer_interface = {
     /* .free_buffer     = */ ggml_backend_cann_buffer_free_buffer,
     /* .get_base        = */ ggml_backend_cann_buffer_get_base,
     /* .init_tensor     = */ ggml_backend_cann_buffer_init_tensor,
-    /* .memset_tensor   = */ NULL,
+    /* .memset_tensor   = */ ggml_backend_cann_buffer_memset_tensor,
     /* .set_tensor      = */ ggml_backend_cann_buffer_set_tensor,
     /* .get_tensor      = */ ggml_backend_cann_buffer_get_tensor,
     /* .set_tensor_2d   = */ NULL,
@@ -1835,6 +1851,9 @@ static bool ggml_cann_compute_forward(ggml_backend_cann_context & ctx, struct gg
                 case GGML_UNARY_OP_STEP:
                     ggml_cann_step(ctx, dst);
                     break;
+                case GGML_UNARY_OP_SOFTPLUS:
+                    ggml_cann_softplus(ctx, dst);
+                    break;
                 default:
                     return false;
             }
@@ -1845,20 +1864,16 @@ static bool ggml_cann_compute_forward(ggml_backend_cann_context & ctx, struct gg
                     GGML_CANN_CALL_OP_UNARY_GATED(Relu);
                     break;
                 case GGML_GLU_OP_GEGLU:
+                    ggml_cann_geglu(ctx, dst, 0);  // approximate=0 → tanh
+                    break;
                 case GGML_GLU_OP_GEGLU_ERF:
-                    // aclnnGelu internally uses the erf-based approximation.
-                    GGML_CANN_CALL_OP_UNARY_GATED(Gelu);
+                    ggml_cann_geglu(ctx, dst, 1);  // approximate=1 → erf
                     break;
                 case GGML_GLU_OP_SWIGLU:
-                    GGML_CANN_CALL_OP_UNARY_GATED(Silu);
+                    ggml_cann_swiglu(ctx, dst);
                     break;
                 case GGML_GLU_OP_GEGLU_QUICK:
-                    {
-                        auto lambda = [](ggml_backend_cann_context & ctx, aclTensor * acl_src, aclTensor * acl_dst) {
-                            GGML_CANN_CALL_ACLNN_OP(ctx, GeluV2, acl_src, 0, acl_dst);
-                        };
-                        ggml_cann_op_unary_gated(lambda, ctx, dst);
-                    }
+                    ggml_cann_geglu_quick(ctx, dst);
                     break;
                 default:
                     return false;
@@ -1920,6 +1935,9 @@ static bool ggml_cann_compute_forward(ggml_backend_cann_context & ctx, struct gg
         case GGML_OP_CPY:
             ggml_cann_cpy(ctx, dst);
             break;
+        case GGML_OP_SET:
+            ggml_cann_set(ctx, dst);
+            break;
         case GGML_OP_CONT:
             ggml_cann_dup(ctx, dst);
             break;
@@ -1989,6 +2007,21 @@ static bool ggml_cann_compute_forward(ggml_backend_cann_context & ctx, struct gg
         case GGML_OP_SSM_CONV:
             ggml_cann_ssm_conv(ctx, dst);
             break;
+        case GGML_OP_CUMSUM:
+            ggml_cann_cumsum(ctx, dst);
+            break;
+        case GGML_OP_TRI:
+            ggml_cann_tri(ctx, dst);
+            break;
+        case GGML_OP_FILL:
+            ggml_cann_fill(ctx, dst);
+            break;
+        case GGML_OP_DIAG:
+            ggml_cann_diag(ctx, dst);
+            break;
+        case GGML_OP_SOLVE_TRI:
+            ggml_cann_solve_tri(ctx, dst);
+            break;
         default:
             return false;
     }
@@ -2324,6 +2357,7 @@ static enum ggml_status ggml_backend_cann_graph_compute(ggml_backend_t backend,
     if (use_cann_graph) {
         // If no matching graph is found, the graph needs to be recaptured.
         graph_capture_required = !cann_ctx->graph_lru_cache.find_and_move_to_front(cgraph);
+
         if (graph_capture_required) {
             // If no matching graph is found, add a new ACL graph.
             ggml_cann_graph * new_graph = ggml_cann_graph::create_from_cgraph(cgraph);
@@ -2382,6 +2416,7 @@ static bool ggml_backend_cann_supports_op(ggml_backend_dev_t dev, const ggml_ten
                 case GGML_UNARY_OP_SGN:
                 case GGML_UNARY_OP_STEP:
                 case GGML_UNARY_OP_GELU_ERF:
+                case GGML_UNARY_OP_SOFTPLUS:
                     return true;
                 default:
                     return false;
@@ -2572,6 +2607,7 @@ static bool ggml_backend_cann_supports_op(ggml_backend_dev_t dev, const ggml_ten
         case GGML_OP_SUM_ROWS:
         case GGML_OP_ARGSORT:
         case GGML_OP_ACC:
+        case GGML_OP_SET:
         case GGML_OP_GROUP_NORM:
             return true;
         case GGML_OP_PAD:
@@ -2649,6 +2685,16 @@ static bool ggml_backend_cann_supports_op(ggml_backend_dev_t dev, const ggml_ten
             }
         case GGML_OP_SSM_CONV:
             return true;
+        case GGML_OP_CUMSUM:
+            return op->src[0]->type == GGML_TYPE_F32;
+        case GGML_OP_TRI:
+            return op->src[0]->type == GGML_TYPE_F32;
+        case GGML_OP_FILL:
+            return op->src[0]->type == GGML_TYPE_F32;
+        case GGML_OP_DIAG:
+            return op->src[0]->type == GGML_TYPE_F32;
+        case GGML_OP_SOLVE_TRI:
+            return op->src[0]->type == GGML_TYPE_F32;
         default:
             return false;
     }
@@ -2700,8 +2746,8 @@ static const ggml_backend_i ggml_backend_cann_interface = {
     /* .free                    = */ ggml_backend_cann_free,
     /* .set_tensor_async        = */ ggml_backend_cann_set_tensor_async,
     /* .get_tensor_async        = */ ggml_backend_cann_get_tensor_async,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ ggml_backend_cann_cpy_tensor_async,
     /* .synchronize             = */ ggml_backend_cann_synchronize,
     /* .graph_plan_create       = */ NULL,

ggml/src/ggml-cpu/CMakeLists.txt

@@ -485,6 +485,13 @@ function(ggml_add_cpu_backend_variant_impl tag_name)
             if (GGML_RV_ZIHINTPAUSE)
                 string(APPEND MARCH_STR "_zihintpause")
             endif()
+            if (GGML_CPU_RISCV64_SPACEMIT)
+                # `xsmtvdotii' is only required for GCC >= 15.
+                if (CMAKE_C_COMPILER_ID STREQUAL "GNU" AND
+                    CMAKE_C_COMPILER_VERSION VERSION_GREATER_EQUAL 15)
+                    string(APPEND MARCH_STR "_xsmtvdotii")
+                endif()
+            endif()
 
             list(APPEND ARCH_FLAGS "-march=${MARCH_STR}" -mabi=lp64d)
         else()

ggml/src/ggml-cpu/amx/mmq.cpp

@@ -2005,12 +2005,12 @@ void tinygemm_kernel_amx(int M, int N, int KB, const void * RESTRICT _A, const v
     const int lda = KB * sizeof(TA);
     //const int ldb = KB * sizeof(TB);
 
-    static thread_local packed_B_t Tile0[TILE_N * TILE_K];
-    static thread_local packed_B_t Tile1[TILE_N * TILE_K];
-    static thread_local int8_t Tile23[TILE_M * TILE_K];
+    alignas(64) static thread_local packed_B_t Tile0[TILE_N * TILE_K];
+    alignas(64) static thread_local packed_B_t Tile1[TILE_N * TILE_K];
+    alignas(64) static thread_local int8_t Tile23[TILE_M * TILE_K];
 
-    static thread_local int32_t TileC0[TILE_M * TILE_N * 4];
-    static thread_local int32_t TileC1[TILE_M * TILE_N * 4];
+    alignas(64) static thread_local int32_t TileC0[TILE_M * TILE_N * 4];
+    alignas(64) static thread_local int32_t TileC1[TILE_M * TILE_N * 4];
 
     // double buffering C to interleave avx512 and amx
     int32_t * C_cur = TileC0;
@@ -2187,21 +2187,21 @@ void tinygemm_kernel_amx(int M, int N, int KB, const void * RESTRICT _A, const v
     const int m1 = std::max(M - TILE_M, 0);
     //const int lda = KB * sizeof(TA);
 
-    static thread_local int8_t Tile0[TILE_N * TILE_K];
-    static thread_local int8_t Tile1[TILE_N * TILE_K];
-    static thread_local int8_t Tile23[TILE_M * TILE_K];
+    alignas(64) static thread_local int8_t Tile0[TILE_N * TILE_K];
+    alignas(64) static thread_local int8_t Tile1[TILE_N * TILE_K];
+    alignas(64) static thread_local int8_t Tile23[TILE_M * TILE_K];
 
     // mat mul result for each group
-    static thread_local int32_t Tile4[TILE_M * TILE_N];
-    static thread_local int32_t Tile5[TILE_M * TILE_N];
-    static thread_local int32_t Tile6[TILE_M * TILE_N];
-    static thread_local int32_t Tile7[TILE_M * TILE_N];
+    alignas(64) static thread_local int32_t Tile4[TILE_M * TILE_N];
+    alignas(64) static thread_local int32_t Tile5[TILE_M * TILE_N];
+    alignas(64) static thread_local int32_t Tile6[TILE_M * TILE_N];
+    alignas(64) static thread_local int32_t Tile7[TILE_M * TILE_N];
 
     // sum of each QK_K block, contains 8 groups, int32
-    static thread_local int32_t Sumi4[TILE_M * TILE_N];
-    static thread_local int32_t Sumi5[TILE_M * TILE_N];
-    static thread_local int32_t Sumi6[TILE_M * TILE_N];
-    static thread_local int32_t Sumi7[TILE_M * TILE_N];
+    alignas(64) static thread_local int32_t Sumi4[TILE_M * TILE_N];
+    alignas(64) static thread_local int32_t Sumi5[TILE_M * TILE_N];
+    alignas(64) static thread_local int32_t Sumi6[TILE_M * TILE_N];
+    alignas(64) static thread_local int32_t Sumi7[TILE_M * TILE_N];
 
     const int k_group_size = std::is_same<TB, block_q6_K>::value ? 16 : 32;
     for (int i = 0; i < KB; ++i) {

ggml/src/ggml-cpu/arch/arm/repack.cpp

@@ -5023,6 +5023,71 @@ void ggml_gemm_q8_0_4x8_q8_0(int                        n,
     UNUSED(ncols_interleaved);
     UNUSED(blocklen);
 
+#if defined(__aarch64__) && defined(__ARM_FEATURE_SVE) && defined(__ARM_FEATURE_MATMUL_INT8)
+    if (svcntb() * 8 == 256) {
+        const block_q8_0x4 * b_ptr_base = (const block_q8_0x4 *) vx;
+
+        static const uint32_t idx_arr[8] = {0, 1, 4, 5, 2, 3, 6, 7};
+        svuint32_t idx = svld1(svptrue_b32(), idx_arr);
+        static const uint32_t idx_arr1[8] = {0, 1, 2, 3, 1, 2, 3, 0};
+        svuint32_t idx_sc1 = svld1(svptrue_b32(), idx_arr1);
+        static const uint32_t idx_arr2[8] = {0, 1, 2, 3, 0, 1, 2, 3};
+        svuint32_t idx_sc2 = svld1(svptrue_b32(), idx_arr2);
+
+        for (int y = 0; y < nr; y += 4) {
+            const block_q8_0x4 * a_ptr_base = (const block_q8_0x4 *) vy + (y / 4) * nb;
+
+            for (int x = 0; x < nc; x += ncols_interleaved) {
+                const block_q8_0x4 * b_ptr = b_ptr_base + (x / 4) * nb;
+                const block_q8_0x4 * a_ptr = a_ptr_base;
+
+                svfloat32_t acc_f32_01 = svdup_f32(0);
+                svfloat32_t acc_f32_23 = svdup_f32(0);
+
+                for (int b = 0; b < nb; b++) {
+
+                    svint32_t acc_01 = svdup_s32(0);
+                    svint32_t acc_23 = svdup_s32(0);
+
+                    // Process 4 chunks of 8 positions each
+                    for (int chunk = 0; chunk < 4; chunk++) {
+                        svint8_t s_a01 = svld1rq_s8(svptrue_b8(), a_ptr->qs + chunk * 32);
+                        svint8_t s_a23 = svld1rq_s8(svptrue_b8(), a_ptr->qs + chunk * 32 + 16);
+                        svint8_t s_b0123 = svld1_s8(svptrue_b8(), b_ptr->qs + chunk * 32);
+
+                        acc_01 = svmmla_s32(acc_01, s_a01, s_b0123);
+                        acc_23 = svmmla_s32(acc_23, s_a23, s_b0123);
+                    }
+
+                    // Reorder outputs from 2×2 tiles to row-major
+                    // acc[01] = [r0c0, r0c1, r1c0, r1c1, r0c2, r0c3, r1c2, r1c3]
+                    // acc[23] = [r2c0, r2c1, r3c0, r3c1, r2c2, r2c3, r3c2, r3c3]
+
+                    svint32_t row01 = svtbl_s32(acc_01, idx);
+                    svint32_t row23 = svtbl_s32(acc_23, idx);
+
+                    svfloat16_t temp1 = svld1_f16(svptrue_pat_b16(SV_VL4), (const __fp16 *) a_ptr->d);
+                    svfloat16_t temp2 = svld1_f16(svptrue_pat_b16(SV_VL4), (const __fp16 *) b_ptr->d);
+                    svfloat32_t sv_a_d = svtbl_f32(svcvt_f32_f16_x(svptrue_b32(), svzip1_f16(temp1, temp1)), idx_sc1);
+                    svfloat32_t sv_b_d = svtbl_f32(svcvt_f32_f16_x(svptrue_b32(), svzip1_f16(temp2, temp2)), idx_sc2);
+
+                    acc_f32_01 = svmla_f32_x(svptrue_b32(), acc_f32_01, svcvt_f32_s32_x(svptrue_b32(), row01), svmul_lane_f32(sv_b_d, sv_a_d, 0));
+                    acc_f32_23 = svmla_f32_x(svptrue_b32(), acc_f32_23, svcvt_f32_s32_x(svptrue_b32(), row23), svmul_lane_f32(sv_b_d, sv_a_d, 2));
+                    a_ptr++;
+                    b_ptr++;
+                }
+
+                svbool_t pg4 = svptrue_pat_b32(SV_VL4);
+                svst1_f32(pg4, s + (y+0) * bs + x, acc_f32_01);
+                svst1_f32(pg4, s + (y+1) * bs + x, svext_f32(acc_f32_01, acc_f32_01, 4));
+                svst1_f32(pg4, s + (y+2) * bs + x, acc_f32_23);
+                svst1_f32(pg4, s + (y+3) * bs + x, svext_f32(acc_f32_23, acc_f32_23, 4));
+            }
+        }
+        return;
+    }
+#endif  // SVE compile-time end
+
 #if defined(__aarch64__) && defined(__ARM_NEON) && defined(__ARM_FEATURE_MATMUL_INT8)
     const block_q8_0x4 * b_ptr_base = (const block_q8_0x4 *) vx;
 

ggml/src/ggml-cpu/ggml-cpu.cpp

@@ -195,8 +195,8 @@ static const struct ggml_backend_i ggml_backend_cpu_i = {
     /* .free                    = */ ggml_backend_cpu_free,
     /* .set_tensor_async        = */ NULL,
     /* .get_tensor_async        = */ NULL,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ NULL,
     /* .synchronize             = */ NULL,
     /* .graph_plan_create       = */ ggml_backend_cpu_graph_plan_create,

ggml/src/ggml-cpu/llamafile/sgemm.cpp

@@ -2321,6 +2321,9 @@ class tinyBLAS_Q0_PPC {
     }
 
     void matmul(int64_t m, int64_t n) {
+    #if defined(_AIX) || defined(__BIG_ENDIAN__)
+        mnpack(0, m, 0, n);
+    #else
         const int64_t mc = 64;
         const int64_t kc = 64;
         int64_t nc = 64;
@@ -2334,7 +2337,6 @@ class tinyBLAS_Q0_PPC {
         } else {
             n_aligned = (n / 64) * 64;
         }
-
         if (n_aligned > 0) {
             if (n_aligned % 64 == 0)      nc = 64;
             else if (n_aligned == n)      nc = n;
@@ -2352,6 +2354,7 @@ class tinyBLAS_Q0_PPC {
         } else {
             mnpack(0, m, 0, n);
         }
+    #endif
     }
 
   private:
@@ -3191,12 +3194,16 @@ class tinyBLAS_PPC {
     }
 
     void matmul(int64_t m, int64_t n) {
+    #if defined(_AIX) || defined(__BIG_ENDIAN__)
+        mnpack(0, m, 0, n);
+    #else
         int64_t mc = 256; int64_t nc = 256; int64_t kc = 256;
         if (m % mc == 0 && n % nc == 0 && k % kc == 0) {
             matmul_tiled(m, n, mc, nc, kc);
         } else {
             mnpack(0, m, 0, n);
         }
+    #endif
     }
 
   private:

ggml/src/ggml-cuda/common.cuh

@@ -830,6 +830,18 @@ static __device__ __forceinline__ float ggml_cuda_ue4m3_to_fp32(uint8_t x) {
 #endif // defined(GGML_USE_HIP) && defined(CDNA3) && defined(FP8_AVAILABLE) && HIP_VERSION >= 60200000
 }
 
+static __device__ __forceinline__ uint8_t ggml_cuda_fp32_to_ue4m3(float x) {
+#if defined(BLACKWELL_MMA_AVAILABLE) // This is used for NVFP4 subblock scale quantizations only
+    if (!(x > 0.0f)) {
+        return 0;
+    }
+    const __nv_fp8_e4m3 xf(x);
+    return xf.__x;
+#else
+     NO_DEVICE_CODE; // Used only for NVFP4 Scales for Activations, only for Blackwell
+#endif // defined(BLACKWELL_MMA_AVAILABLE)
+}
+
 __device__ __forceinline__ uint8_t ggml_cuda_float_to_fp4_e2m1(float x, float e) {
     const uint8_t sign_bit = (x < 0.0f) << 3;
     float         ax       = fabsf(x) * e;

ggml/src/ggml-cuda/fattn-mma-f16.cuh

@@ -66,6 +66,9 @@ static constexpr __host__ __device__ fattn_mma_config ggml_cuda_fattn_mma_get_co
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(256, 256, 32, 128, 2,  32, 128, 128, 128, 2, true);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(256, 256, 64, 128, 2,  32, 128, 128, 128, 2, true);
 
+    GGML_CUDA_FATTN_MMA_CONFIG_CASE(320, 256, 32, 128, 2,  32, 128, 128, 128, 1, false);
+    GGML_CUDA_FATTN_MMA_CONFIG_CASE(320, 256, 64, 256, 1,  32, 128, 128, 128, 1, false);
+
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512,  8,  64, 4,  32, 256, 256, 128, 1, false);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512, 16,  64, 4,  32, 256, 256, 128, 1, false);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512, 32, 128, 2,  32, 128, 128, 128, 1, false);
@@ -85,6 +88,9 @@ static constexpr __host__ __device__ fattn_mma_config ggml_cuda_fattn_mma_get_co
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(256, 256, 32, 128, 2,  64, 128, 128,  64, 2, true);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(256, 256, 64, 128, 2,  64, 128, 128,  64, 2, true);
 
+    GGML_CUDA_FATTN_MMA_CONFIG_CASE(320, 256, 32, 128, 2,  32, 128, 128, 128, 1, false);
+    GGML_CUDA_FATTN_MMA_CONFIG_CASE(320, 256, 64, 256, 1,  32, 128, 128, 128, 1, false);
+
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512,  8,  64, 4,  32,  96,  64, 128, 1, false);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512, 16,  64, 4,  32,  96,  64, 128, 1, false);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512, 32, 128, 2,  32, 128, 128, 128, 1, false);
@@ -118,6 +124,9 @@ static constexpr __host__ __device__ fattn_mma_config ggml_cuda_fattn_mma_get_co
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(256, 256, 32, 128, 2,  64, 128, 128,  64, 2, true);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(256, 256, 64, 128, 2,  64, 128, 128,  64, 2, true);
 
+    GGML_CUDA_FATTN_MMA_CONFIG_CASE(320, 256, 32, 128, 2,  64, 160, 128,  64, 2, true);
+    GGML_CUDA_FATTN_MMA_CONFIG_CASE(320, 256, 64, 128, 2,  64, 160, 128,  64, 2, false);
+
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512, 16,  64, 4,  32, 128, 128, 128, 1, false);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512, 32, 128, 2,  32, 128, 128, 128, 1, false);
     GGML_CUDA_FATTN_MMA_CONFIG_CASE(512, 512, 64, 256, 1,  32, 128, 128, 128, 1, false);
@@ -1217,7 +1226,7 @@ static __device__ __forceinline__ void flash_attn_ext_f16_process_tile(
         float KQ_max_scale[cols_per_thread];
 #pragma unroll
         for (int col = 0; col < cols_per_thread; ++col) {
-            const int jc = cols_per_warp == 8 ? T_C_KQ::get_j(col) : T_C_KQ::get_i(2*col);
+            const int jc = (threadIdx.y/np)*cols_per_warp + (cols_per_warp == 8 ? T_C_KQ::get_j(col) : T_C_KQ::get_i(2*col));
             const float sink = sinks_f[jc % ncols2];
 
             const float KQ_max_new = fmaxf(KQ_max[col], sink);
@@ -1825,6 +1834,10 @@ extern DECL_FATTN_MMA_F16_CASE(576, 512, 1, 16);
 extern DECL_FATTN_MMA_F16_CASE(576, 512, 2, 16);
 extern DECL_FATTN_MMA_F16_CASE(576, 512, 4, 16);
 
+// Mistral Small 4 (DKQ=320, DV=256), GQA=32-only build:
+extern DECL_FATTN_MMA_F16_CASE(320, 256,  1, 32);
+extern DECL_FATTN_MMA_F16_CASE(320, 256,  2, 32);
+
 // For GLM 4.7 Flash
 extern DECL_FATTN_MMA_F16_CASE(576, 512,  4,  4);
 extern DECL_FATTN_MMA_F16_CASE(576, 512,  8,  4);

ggml/src/ggml-cuda/fattn-tile.cu

@@ -38,6 +38,10 @@ void ggml_cuda_flash_attn_ext_tile(ggml_backend_cuda_context & ctx, ggml_tensor
             GGML_ASSERT(V->ne[0] == K->ne[0]);
             ggml_cuda_flash_attn_ext_tile_case<256, 256>(ctx, dst);
         } break;
+        case 320: {
+            GGML_ASSERT(V->ne[0] == 256);
+            ggml_cuda_flash_attn_ext_tile_case<320, 256>(ctx, dst);
+        } break;
         case 512: {
             GGML_ASSERT(V->ne[0] == K->ne[0]);
             ggml_cuda_flash_attn_ext_tile_case<512, 512>(ctx, dst);

ggml/src/ggml-cuda/fattn-tile.cuh

@@ -68,6 +68,8 @@ static constexpr __host__ __device__ uint32_t ggml_cuda_fattn_tile_get_config_nv
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 16, 256, 2,  64,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 32, 256, 2,  64,  64)
 
+    GGML_CUDA_FATTN_TILE_CONFIG_CASE(320, 256, 16, 256, 2,  64,  64)
+
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512,  4, 128, 2,  64,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512,  8, 256, 2,  64,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 16, 256, 2,  64,  64)
@@ -128,6 +130,8 @@ static constexpr __host__ __device__ uint32_t ggml_cuda_fattn_tile_get_config_nv
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 16, 256, 2,  32, 128)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 32, 256, 2,  32,  64)
 
+    GGML_CUDA_FATTN_TILE_CONFIG_CASE(320, 256, 16, 256, 2,  32,  64)
+
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512,  4, 128, 2,  32,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512,  8, 256, 2,  32,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 16, 256, 2,  32,  64)
@@ -195,6 +199,8 @@ static constexpr __host__ __device__ uint32_t ggml_cuda_fattn_tile_get_config_am
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 16, 256, 2,  32, 128)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 32, 256, 2,  32, 128)
 
+    GGML_CUDA_FATTN_TILE_CONFIG_CASE(320, 256, 32, 512, 1, 128,  64)
+
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512,  4, 128, 2,  64,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512,  8, 256, 2,  64,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 16, 256, 2,  64,  64)
@@ -264,6 +270,8 @@ static constexpr __host__ __device__ uint32_t ggml_cuda_fattn_tile_get_config_am
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 16, 256, 5,  32, 256)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 32, 256, 3,  64, 128)
 
+    GGML_CUDA_FATTN_TILE_CONFIG_CASE(320, 256, 32, 256, 2, 128,  64)
+
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512,  4, 128, 2,  64,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512,  8, 256, 2,  64,  64)
     GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 16, 256, 4,  64,  64)
@@ -1116,7 +1124,7 @@ static void launch_fattn_tile_switch_ncols1(ggml_backend_cuda_context & ctx, ggm
     constexpr size_t nbytes_shared = 0;
 
 #ifdef GGML_USE_HIP
-    if constexpr (DV <= 128) {
+    if constexpr (DKQ <= 128) {
         if (Q->ne[1] > 32/ncols2) {
             constexpr int cols_per_block = 64;
             const int nwarps    = ggml_cuda_fattn_tile_get_nthreads (DKQ, DV, cols_per_block, cc) / warp_size;
@@ -1130,7 +1138,7 @@ static void launch_fattn_tile_switch_ncols1(ggml_backend_cuda_context & ctx, ggm
 #endif // GGML_USE_HIP
 
 #ifndef GGML_USE_HIP
-    if constexpr (DV <= 256)
+    if constexpr (DKQ <= 256)
 #endif // GGML_USE_HIP
     {
         if (Q->ne[1] > 16/ncols2) {
@@ -1144,14 +1152,16 @@ static void launch_fattn_tile_switch_ncols1(ggml_backend_cuda_context & ctx, ggm
         }
     }
 
-    if (Q->ne[1] > 8/ncols2) {
-        constexpr int cols_per_block = 16;
-        const int nwarps    = ggml_cuda_fattn_tile_get_nthreads (DKQ, DV, cols_per_block, cc) / warp_size;
-        const int nbatch_fa = ggml_cuda_fattn_tile_get_nbatch_fa(DKQ, DV, cols_per_block, cc);
-        fattn_kernel_t fattn_kernel = flash_attn_tile<DKQ, DV, cols_per_block/ncols2, ncols2, use_logit_softcap>;
-        launch_fattn<DV, cols_per_block/ncols2, ncols2>
-            (ctx, dst, fattn_kernel, nwarps, nbytes_shared, nbatch_fa, true, true, false, warp_size);
-        return;
+    if constexpr (ncols2 <= 16) {
+        if (Q->ne[1] > 8/ncols2) {
+            constexpr int cols_per_block = 16;
+            const int nwarps    = ggml_cuda_fattn_tile_get_nthreads (DKQ, DV, cols_per_block, cc) / warp_size;
+            const int nbatch_fa = ggml_cuda_fattn_tile_get_nbatch_fa(DKQ, DV, cols_per_block, cc);
+            fattn_kernel_t fattn_kernel = flash_attn_tile<DKQ, DV, cols_per_block/ncols2, ncols2, use_logit_softcap>;
+            launch_fattn<DV, cols_per_block/ncols2, ncols2>
+                (ctx, dst, fattn_kernel, nwarps, nbytes_shared, nbatch_fa, true, true, false, warp_size);
+            return;
+        }
     }
 
     if constexpr (ncols2 <= 8) {
@@ -1210,6 +1220,25 @@ static void launch_fattn_tile_switch_ncols2(ggml_backend_cuda_context & ctx, ggm
     const int gqa_limit = nvidia && gqa_ratio <= 4 && DV <= 256 ? 16 : INT_MAX;
     const bool use_gqa_opt = mask && max_bias == 0.0f && Q->ne[1] <= gqa_limit && K->ne[1] % FATTN_KQ_STRIDE == 0;
 
+    if constexpr (DKQ == 320) {
+        // This branch is only used for Mistral Small 4 which has a GQA ratio of 32.
+        // On AMD, simply use that GQA ratio with 32 columns / block since we always have enough SRAM.
+        // On NVIDIA however, the tile kernel is only used for GPUs that can't use the mma kernel (Pascal and older).
+        // Therefore, use a GQA ratio of 16 with 16 columns / block to stay below 48 kiB of SRAM / block.
+#ifdef GGML_USE_HIP
+        if (use_gqa_opt && gqa_ratio % 32 == 0) {
+            launch_fattn_tile_switch_ncols1<DKQ, DV, 32, use_logit_softcap>(ctx, dst);
+            return;
+        }
+#else
+        if (use_gqa_opt && gqa_ratio % 16 == 0) {
+            launch_fattn_tile_switch_ncols1<DKQ, DV, 16, use_logit_softcap>(ctx, dst);
+            return;
+        }
+#endif // GGML_USE_HIP
+        GGML_ABORT("flash-attn tile (320/256): expected GQA ratio multiple of 32");
+    }
+
     if constexpr (DKQ == 576) {
         if (use_gqa_opt && gqa_ratio % 16 == 0) {
             launch_fattn_tile_switch_ncols1<DKQ, DV, 16, use_logit_softcap>(ctx, dst);
@@ -1221,7 +1250,7 @@ static void launch_fattn_tile_switch_ncols2(ggml_backend_cuda_context & ctx, ggm
         }
     }
 
-    if constexpr (DKQ <= 512) {
+    if constexpr (DKQ <= 512 && DKQ != 320) {
         if (use_gqa_opt && gqa_ratio % 8 == 0) {
             launch_fattn_tile_switch_ncols1<DKQ, DV, 8, use_logit_softcap>(ctx, dst);
             return;
@@ -1275,5 +1304,6 @@ extern DECL_FATTN_TILE_CASE( 96,  96);
 extern DECL_FATTN_TILE_CASE(112, 112);
 extern DECL_FATTN_TILE_CASE(128, 128);
 extern DECL_FATTN_TILE_CASE(256, 256);
+extern DECL_FATTN_TILE_CASE(320, 256);
 extern DECL_FATTN_TILE_CASE(512, 512);
 extern DECL_FATTN_TILE_CASE(576, 512);

ggml/src/ggml-cuda/fattn.cu

@@ -143,6 +143,22 @@ static void ggml_cuda_flash_attn_ext_mma_f16(ggml_backend_cuda_context & ctx, gg
             GGML_ASSERT(V->ne[0] == 256);
             ggml_cuda_flash_attn_ext_mma_f16_switch_ncols2<256, 256>(ctx, dst);
             break;
+        case 320:
+            // For Mistral Small 4, go straight to the ncols1 switch (ncols2=32-only build).
+            GGML_ASSERT(V->ne[0] == 256);
+            {
+                float max_bias = 0.0f;
+                memcpy(&max_bias, (const float *) KQV->op_params + 1, sizeof(float));
+
+                const bool use_gqa_opt = mask && max_bias == 0.0f;
+                GGML_ASSERT(use_gqa_opt);
+                GGML_ASSERT(Q->ne[2] % K->ne[2] == 0);
+                const int gqa_ratio = Q->ne[2] / K->ne[2];
+                GGML_ASSERT(gqa_ratio % 32 == 0);
+
+                ggml_cuda_flash_attn_ext_mma_f16_switch_ncols1<320, 256, 32>(ctx, dst);
+            }
+            break;
         case 512:
             GGML_ASSERT(V->ne[0] == 512);
             ggml_cuda_flash_attn_ext_mma_f16_switch_ncols2<512, 512>(ctx, dst);
@@ -352,6 +368,14 @@ static best_fattn_kernel ggml_cuda_get_best_fattn_kernel(const int device, const
                 return BEST_FATTN_KERNEL_NONE;
             }
             break;
+        case 320:
+            if (V->ne[0] != 256 || !gqa_opt_applies) {
+                return BEST_FATTN_KERNEL_NONE;
+            }
+            if (gqa_ratio % 32 != 0) {
+                return BEST_FATTN_KERNEL_NONE;
+            }
+            break;
         case 512:
             if (V->ne[0] != K->ne[0]) {
                 return BEST_FATTN_KERNEL_NONE;

ggml/src/ggml-cuda/ggml-cuda.cu

@@ -3556,6 +3556,9 @@ static bool ggml_cuda_can_fuse(const struct ggml_cgraph *                cgraph,
      && unary_ops.size() == 1 && unary_ops.begin()[0] == GGML_UNARY_OP_SILU) {
         const ggml_tensor * ssm_conv = cgraph->nodes[node_idx];
         const ggml_tensor * silu     = cgraph->nodes[node_idx+1];
+        if (ggml_get_unary_op(silu) != unary_ops.begin()[0]) {
+            return false;
+        }
 
         if (ssm_conv->type != GGML_TYPE_F32 || silu->type != GGML_TYPE_F32) {
             return false;
@@ -3564,6 +3567,31 @@ static bool ggml_cuda_can_fuse(const struct ggml_cgraph *                cgraph,
         return true;
     }
 
+    if (ops.size() == 3 && ops.begin()[0] == GGML_OP_SSM_CONV && ops.begin()[1] == GGML_OP_ADD
+     && ops.begin()[2] == GGML_OP_UNARY && unary_ops.size() == 1 && unary_ops.begin()[0] == GGML_UNARY_OP_SILU) {
+        const ggml_tensor * ssm_conv = cgraph->nodes[node_idx];
+        const ggml_tensor * add      = cgraph->nodes[node_idx+1];
+        const ggml_tensor * silu     = cgraph->nodes[node_idx+2];
+        if (ggml_get_unary_op(silu) != unary_ops.begin()[0]) {
+            return false;
+        }
+
+        if (ssm_conv->type != GGML_TYPE_F32 || add->type != GGML_TYPE_F32 || silu->type != GGML_TYPE_F32) {
+            return false;
+        }
+
+        // ADD must consume ssm_conv's output and broadcast a 1-D channel-wise bias.
+        const ggml_tensor * bias = (add->src[0] == ssm_conv) ? add->src[1] : add->src[0];
+        if (bias->type != GGML_TYPE_F32 || !ggml_is_contiguous(bias)) {
+            return false;
+        }
+        if (ggml_nelements(bias) != ssm_conv->ne[0] || bias->ne[0] != ssm_conv->ne[0]) {
+            return false;
+        }
+
+        return true;
+    }
+
     if (ops.size() == 2 && ops.begin()[0] == GGML_OP_UNARY && ops.begin()[1] == GGML_OP_MUL
      && unary_ops.size() == 1 && (unary_ops.begin()[0] == GGML_UNARY_OP_SILU || unary_ops.begin()[0] == GGML_UNARY_OP_SIGMOID || unary_ops.begin()[0] == GGML_UNARY_OP_SOFTPLUS)) {
         const ggml_tensor * unary = cgraph->nodes[node_idx];
@@ -3640,6 +3668,362 @@ static bool ggml_cuda_can_fuse(const struct ggml_cgraph *                cgraph,
     return false;
 }
 
+// try and fuse nodes and return the number of nodes to skip
+static int ggml_cuda_try_fuse(ggml_backend_cuda_context * cuda_ctx, ggml_cgraph * cgraph, int i) {
+
+    static bool disable_fusion = getenv("GGML_CUDA_DISABLE_FUSION") != nullptr && std::atoi(getenv("GGML_CUDA_DISABLE_FUSION"));
+    if (disable_fusion) {
+        return 0;
+    }
+
+    ggml_tensor * node = cgraph->nodes[i];
+
+    //topk-moe
+    if (cgraph->nodes[i]->op == GGML_OP_UNARY || cgraph->nodes[i]->op == GGML_OP_SOFT_MAX ||
+            cgraph->nodes[i]->op == GGML_OP_ARGSORT) {
+        ggml_cuda_topk_moe_args args;
+        const bool              can_fuse = ggml_cuda_topk_moe_fusion(cgraph, i, args);
+        std::vector<ggml_op>    ops;
+
+        if (can_fuse) {
+            const ggml_tensor * logits  = node->src[0];
+            ggml_tensor *       weights = nullptr;
+            ggml_tensor *       ids     = nullptr;
+            const ggml_tensor * bias    = nullptr;
+            const ggml_tensor * clamp   = nullptr;
+            const ggml_tensor * scale   = nullptr;
+
+            if (!args.delayed_softmax) {
+                ggml_op gating_op = args.sigmoid ? GGML_OP_UNARY : GGML_OP_SOFT_MAX;
+                int     out_nodes[2];  // nodes which can't be elided
+
+                if (args.prob_bias) {
+                    bias = cgraph->nodes[i + 2]->src[1];
+                    ops.insert(ops.end(), { gating_op, GGML_OP_RESHAPE, GGML_OP_ADD, GGML_OP_ARGSORT, GGML_OP_VIEW,
+                                            GGML_OP_GET_ROWS });
+                    out_nodes[0] = i + 4;
+                    ids          = cgraph->nodes[i + 4];
+                } else {
+                    ops.insert(ops.end(),
+                               { gating_op, GGML_OP_RESHAPE, GGML_OP_ARGSORT, GGML_OP_VIEW, GGML_OP_GET_ROWS });
+                    out_nodes[0] = i + 3;
+                    ids          = cgraph->nodes[i + 3];
+                }
+
+                if (args.norm) {
+                    ops.insert(ops.end(),
+                               { GGML_OP_RESHAPE, GGML_OP_SUM_ROWS, GGML_OP_CLAMP, GGML_OP_DIV, GGML_OP_RESHAPE });
+                    clamp = cgraph->nodes[i + ops.size() - 3];
+                }
+                if (args.scale) {
+                    ops.insert(ops.end(), { GGML_OP_SCALE });
+                    scale = cgraph->nodes[i + ops.size() - 1];
+                }
+
+                weights      = cgraph->nodes[i + ops.size() - 1];
+                out_nodes[1] = i + ops.size() - 1;
+
+                if (ggml_can_fuse_subgraph(cgraph, i, ops.size(), ops.data(), out_nodes, 2) &&
+                        ggml_cuda_should_use_topk_moe(node, logits, weights, ids) &&
+                        ggml_cuda_check_fusion_memory_ranges(cgraph, i, ops.size(), out_nodes, 2, /*is_topk_moe=*/true)) {
+                    ggml_cuda_op_topk_moe(*cuda_ctx, logits, weights, ids, clamp, scale, bias, args);
+                    return ops.size() - 1;
+                }
+            } else if (!args.norm && !args.prob_bias) {
+                //special case gpt-oss, no norm, no bias.
+                ops.insert(ops.end(), { GGML_OP_ARGSORT, GGML_OP_VIEW, GGML_OP_GET_ROWS, GGML_OP_RESHAPE,
+                                        GGML_OP_SOFT_MAX, GGML_OP_RESHAPE });
+                weights                     = cgraph->nodes[i + 5];
+                ids                         = cgraph->nodes[i + 1];
+                const ggml_tensor * softmax = cgraph->nodes[i + 4];
+
+                int out_nodes[2] = { i + 1, i + 5 };
+                if (ggml_can_fuse_subgraph(cgraph, i, ops.size(), ops.data(), out_nodes, 2) &&
+                        ggml_cuda_should_use_topk_moe(softmax, logits, weights, ids) &&
+                        ggml_cuda_check_fusion_memory_ranges(cgraph, i, ops.size(), out_nodes, 2, /*is_topk_moe=*/true)) {
+                    ggml_cuda_op_topk_moe(*cuda_ctx, logits, weights, ids, clamp, scale, bias, args);
+                    return ops.size() - 1;
+                }
+            }
+        }
+    }
+
+    //RoPE + view + set-rows
+    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_ROPE, GGML_OP_VIEW, GGML_OP_SET_ROWS }, {})) {
+        ggml_tensor * rope     = cgraph->nodes[i];
+        ggml_tensor * set_rows = cgraph->nodes[i + 2];
+
+        ggml_cuda_op_rope_fused(*cuda_ctx, rope, set_rows);
+        return 2;
+    }
+
+    // multi-(add or mul)
+    if (node->op == GGML_OP_ADD || node->op == GGML_OP_MUL) {
+        int     n_fuse = 0;
+        ggml_op ops[8];
+        std::fill(ops, ops + 8, node->op);
+
+        for (; n_fuse <= 6; ++n_fuse) {
+            if (!ggml_can_fuse(cgraph, i + n_fuse, ops + n_fuse, 2)) {
+                break;
+            }
+            if (cgraph->nodes[i + n_fuse] != cgraph->nodes[i + n_fuse + 1]->src[0]) {
+                break;
+            }
+            if (!ggml_are_same_layout(cgraph->nodes[i + n_fuse]->src[1], cgraph->nodes[i + n_fuse + 1]->src[1])) {
+                break;
+            }
+        }
+
+        n_fuse++;
+
+        if (n_fuse > 1) {
+            ggml_tensor fused_node;
+            memcpy(&fused_node, node, sizeof(ggml_tensor));
+            for (int j = 0; j < n_fuse - 1; ++j) {
+                fused_node.src[j + 2] = cgraph->nodes[i + j + 1]->src[1];
+            }
+            fused_node.data = cgraph->nodes[i + n_fuse - 1]->data;
+            if (node->op == GGML_OP_ADD) {
+                ggml_cuda_op_fused_add(*cuda_ctx, &fused_node, n_fuse);
+            } else {
+                ggml_cuda_op_fused_mul(*cuda_ctx, &fused_node, n_fuse);
+            }
+            return n_fuse - 1;
+        }
+    }
+
+    bool fused_mul_mat_vec = false;
+    int  fused_node_count  = 0;
+
+    // gate + glu + up
+    for (ggml_op op : { GGML_OP_MUL_MAT, GGML_OP_MUL_MAT_ID }) {
+        const ggml_op bias_op = op == GGML_OP_MUL_MAT ? GGML_OP_ADD : GGML_OP_ADD_ID;
+
+        if (ggml_cuda_can_fuse(cgraph, i, { op, bias_op, op, bias_op, GGML_OP_GLU }, {})) {
+            ggml_tensor * glu         = cgraph->nodes[i + 4];
+            ggml_tensor * gate_bias_n = glu->src[0];
+            ggml_tensor * up_bias_n   = glu->src[1];
+
+            //we don't assume the order for {gate, up}. Instead infer it from the bias tensor
+            ggml_tensor * gate_n = nullptr;
+            ggml_tensor * up_n   = nullptr;
+
+            if (gate_bias_n->src[0] == cgraph->nodes[i] || gate_bias_n->src[1] == cgraph->nodes[i]) {
+                gate_n = cgraph->nodes[i];
+                up_n   = cgraph->nodes[i + 2];
+            } else if (gate_bias_n->src[0] == cgraph->nodes[i + 2] || gate_bias_n->src[1] == cgraph->nodes[i + 2]) {
+                gate_n = cgraph->nodes[i + 2];
+                up_n   = cgraph->nodes[i];
+            } else {
+                continue;
+            }
+
+            auto get_bias_tensor = [](const ggml_tensor * bias_node, const ggml_tensor * mul_node, ggml_op op_bias) {
+                if (op_bias == GGML_OP_ADD) {
+                    if (bias_node->src[0] == mul_node) {
+                        return bias_node->src[1];
+                    }
+                    if (bias_node->src[1] == mul_node) {
+                        return bias_node->src[0];
+                    }
+                    return (ggml_tensor *) nullptr;
+                }
+                GGML_ASSERT(op_bias == GGML_OP_ADD_ID);
+                GGML_ASSERT(bias_node->src[0] == mul_node);
+                return bias_node->src[1];
+            };
+
+            ggml_tensor * up_bias_tensor   = get_bias_tensor(up_bias_n, up_n, bias_op);
+            ggml_tensor * gate_bias_tensor = get_bias_tensor(gate_bias_n, gate_n, bias_op);
+
+            if (!up_bias_tensor || !gate_bias_tensor) {
+                continue;
+            }
+
+            // we don't support repeating adds
+            if (bias_op == GGML_OP_ADD && (!ggml_are_same_shape(gate_bias_n->src[0], gate_bias_n->src[1]) ||
+                                           !ggml_are_same_shape(up_bias_n->src[0], up_bias_n->src[1]))) {
+                continue;
+            }
+
+            const ggml_tensor * src0 = up_n->src[0];
+            const ggml_tensor * src1 = up_n->src[1];
+            const ggml_tensor * ids  = up_n->src[2];
+
+            if (ggml_cuda_should_fuse_mul_mat_vec_f(up_n)) {
+                ggml_cuda_mm_fusion_args_host fusion_data{};
+                fusion_data.gate      = gate_n->src[0];
+                fusion_data.x_bias    = up_bias_tensor;
+                fusion_data.gate_bias = gate_bias_tensor;
+                fusion_data.glu_op    = ggml_get_glu_op(glu);
+
+                ggml_cuda_mul_mat_vec_f(*cuda_ctx, src0, src1, ids, glu, &fusion_data);
+                fused_mul_mat_vec = true;
+                fused_node_count  = 5;
+                break;
+            }
+
+            if (ggml_cuda_should_fuse_mul_mat_vec_q(up_n)) {
+                ggml_cuda_mm_fusion_args_host fusion_data{};
+                fusion_data.gate      = gate_n->src[0];
+                fusion_data.x_bias    = up_bias_tensor;
+                fusion_data.gate_bias = gate_bias_tensor;
+                fusion_data.glu_op    = ggml_get_glu_op(glu);
+
+                ggml_cuda_mul_mat_vec_q(*cuda_ctx, src0, src1, ids, glu, &fusion_data);
+                fused_mul_mat_vec = true;
+                fused_node_count  = 5;
+                break;
+            }
+        } else if (ggml_cuda_can_fuse(cgraph, i, { op, op, GGML_OP_GLU }, {})) {
+            ggml_tensor * glu  = cgraph->nodes[i + 2];
+            ggml_tensor * gate = glu->src[0];
+            ggml_tensor * up   = glu->src[1];
+
+            bool ok = (gate == cgraph->nodes[i] && up == cgraph->nodes[i + 1]) ||
+                      (gate == cgraph->nodes[i + 1] && up == cgraph->nodes[i]);
+
+            if (!ok) {
+                continue;
+            }
+
+            const ggml_tensor * src0 = up->src[0];
+            const ggml_tensor * src1 = up->src[1];
+            const ggml_tensor * ids  = up->src[2];
+
+            if (ggml_cuda_should_fuse_mul_mat_vec_f(up)) {
+                ggml_cuda_mm_fusion_args_host fusion_data{};
+                fusion_data.gate   = gate->src[0];
+                fusion_data.glu_op = ggml_get_glu_op(glu);
+
+                ggml_cuda_mul_mat_vec_f(*cuda_ctx, src0, src1, ids, glu, &fusion_data);
+                fused_mul_mat_vec = true;
+                fused_node_count  = 3;
+                break;
+            }
+
+            if (ggml_cuda_should_fuse_mul_mat_vec_q(up)) {
+                ggml_cuda_mm_fusion_args_host fusion_data{};
+                fusion_data.gate   = gate->src[0];
+                fusion_data.glu_op = ggml_get_glu_op(glu);
+
+                ggml_cuda_mul_mat_vec_q(*cuda_ctx, src0, src1, ids, glu, &fusion_data);
+                fused_mul_mat_vec = true;
+                fused_node_count  = 3;
+                break;
+            }
+        }
+    }
+
+    if (fused_mul_mat_vec) {
+        return fused_node_count - 1;
+    }
+
+    fused_mul_mat_vec = false;
+    fused_node_count  = 0;
+
+    // gate + add + glu + up + add
+    for (ggml_op op : { GGML_OP_MUL_MAT, GGML_OP_MUL_MAT_ID }) {
+        const ggml_op bias_op = op == GGML_OP_MUL_MAT ? GGML_OP_ADD : GGML_OP_ADD_ID;
+
+        if (!ggml_can_fuse(cgraph, i, { op, bias_op })) {
+            continue;
+        }
+
+        ggml_tensor * mm_node   = cgraph->nodes[i];
+        ggml_tensor * bias_node = cgraph->nodes[i + 1];
+
+        ggml_tensor * bias_tensor = nullptr;
+        if (bias_op == GGML_OP_ADD) {
+            if (bias_node->src[0] == mm_node) {
+                bias_tensor = bias_node->src[1];
+            } else if (bias_node->src[1] == mm_node) {
+                bias_tensor = bias_node->src[0];
+            } else {
+                continue;
+            }
+        } else {
+            if (bias_node->src[0] != mm_node) {
+                continue;
+            }
+            bias_tensor = bias_node->src[1];
+        }
+
+        const ggml_tensor * src0 = mm_node->src[0];
+        const ggml_tensor * src1 = mm_node->src[1];
+        const ggml_tensor * ids  = mm_node->src[2];
+
+        if (bias_op == GGML_OP_ADD_ID && bias_node->src[2] != ids) {
+            continue;
+        }
+
+        if (bias_op == GGML_OP_ADD && !ggml_are_same_shape(bias_node->src[0], bias_node->src[1])) {
+            continue;
+        }
+
+        ggml_cuda_mm_fusion_args_host fusion_data{};
+        fusion_data.x_bias = bias_tensor;
+
+        if (ggml_cuda_should_fuse_mul_mat_vec_f(mm_node)) {
+            ggml_cuda_mul_mat_vec_f(*cuda_ctx, src0, src1, ids, bias_node, &fusion_data);
+            fused_mul_mat_vec = true;
+            fused_node_count  = 2;
+            break;
+        }
+
+        if (ggml_cuda_should_fuse_mul_mat_vec_q(mm_node)) {
+            ggml_cuda_mul_mat_vec_q(*cuda_ctx, src0, src1, ids, bias_node, &fusion_data);
+            fused_mul_mat_vec = true;
+            fused_node_count  = 2;
+            break;
+        }
+    }
+
+    if (fused_mul_mat_vec) {
+        return fused_node_count - 1;
+    }
+
+    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_RMS_NORM, GGML_OP_MUL, GGML_OP_ADD }, {})) {
+        ggml_cuda_op_rms_norm_fused_add(*cuda_ctx, node, cgraph->nodes[i + 1], cgraph->nodes[i + 2]);
+        return 2;
+    }
+
+    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_RMS_NORM, GGML_OP_MUL }, {})) {
+        ggml_cuda_op_rms_norm_fused(*cuda_ctx, node, cgraph->nodes[i + 1]);
+        return 1;
+    }
+
+    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_SSM_CONV, GGML_OP_ADD, GGML_OP_UNARY }, { GGML_UNARY_OP_SILU })) {
+        ggml_cuda_op_ssm_conv(*cuda_ctx, node, cgraph->nodes[i + 1], cgraph->nodes[i + 2]);
+        return 2;
+    }
+
+    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_SSM_CONV, GGML_OP_UNARY }, { GGML_UNARY_OP_SILU })) {
+        ggml_cuda_op_ssm_conv(*cuda_ctx, node, /*bias_add_node=*/ nullptr, cgraph->nodes[i + 1]);
+        return 1;
+    }
+
+    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_UNARY, GGML_OP_MUL }, { GGML_UNARY_OP_SILU }) ||
+        ggml_cuda_can_fuse(cgraph, i, { GGML_OP_UNARY, GGML_OP_MUL }, { GGML_UNARY_OP_SIGMOID }) ||
+        ggml_cuda_can_fuse(cgraph, i, { GGML_OP_UNARY, GGML_OP_MUL }, { GGML_UNARY_OP_SOFTPLUS })) {
+        ggml_cuda_op_unary_mul(*cuda_ctx, node, cgraph->nodes[i + 1]);
+        return 1;
+    }
+
+    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_UNARY, GGML_OP_SQR }, { GGML_UNARY_OP_RELU })) {
+        ggml_cuda_op_relu_sqr(*cuda_ctx, node, cgraph->nodes[i + 1]);
+        return 1;
+    }
+
+    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_SCALE, GGML_OP_UNARY, GGML_OP_SCALE }, { GGML_UNARY_OP_TANH })) {
+        ggml_cuda_op_softcap(*cuda_ctx, cgraph->nodes[i + 2], node);
+        return 2;
+    }
+
+    return 0;
+}
+
 static void ggml_cuda_graph_evaluate_and_capture(ggml_backend_cuda_context * cuda_ctx, ggml_cgraph * cgraph, const bool use_cuda_graph, const bool cuda_graph_update_required, const void * graph_key) {
     bool graph_evaluated_or_captured = false;
 
@@ -3786,355 +4170,11 @@ static void ggml_cuda_graph_evaluate_and_capture(ggml_backend_cuda_context * cud
                     continue;
                 }
 
-                // start of fusion operations
-                static bool disable_fusion = (getenv("GGML_CUDA_DISABLE_FUSION") != nullptr);
-                if (!disable_fusion) {
-                    ggml_cuda_topk_moe_args args;
+                int nodes_to_skip = ggml_cuda_try_fuse(cuda_ctx, cgraph, i);
 
-                    if (cgraph->nodes[i]->op == GGML_OP_UNARY || cgraph->nodes[i]->op == GGML_OP_SOFT_MAX ||
-                        cgraph->nodes[i]->op == GGML_OP_ARGSORT) {
-                        const bool can_fuse = ggml_cuda_topk_moe_fusion(cgraph, i, args);
-
-                        std::vector<ggml_op> ops;
-
-                        if (can_fuse) {
-                            const ggml_tensor * logits  = node->src[0];
-                            ggml_tensor *       weights = nullptr;
-                            ggml_tensor *       ids     = nullptr;
-                            const ggml_tensor * bias    = nullptr;
-                            const ggml_tensor * clamp   = nullptr;
-                            const ggml_tensor * scale   = nullptr;
-
-                            if (!args.delayed_softmax) {
-                                ggml_op gating_op = args.sigmoid ? GGML_OP_UNARY : GGML_OP_SOFT_MAX;
-                                int     out_nodes[2];  // nodes which can't be elided
-
-                                if (args.prob_bias) {
-                                    bias = cgraph->nodes[i + 2]->src[1];
-                                    ops.insert(ops.end(), { gating_op, GGML_OP_RESHAPE, GGML_OP_ADD, GGML_OP_ARGSORT,
-                                                            GGML_OP_VIEW, GGML_OP_GET_ROWS });
-                                    out_nodes[0] = i + 4;
-                                    ids          = cgraph->nodes[i + 4];
-                                } else {
-                                    ops.insert(ops.end(), { gating_op, GGML_OP_RESHAPE, GGML_OP_ARGSORT, GGML_OP_VIEW,
-                                                            GGML_OP_GET_ROWS });
-                                    out_nodes[0] = i + 3;
-                                    ids          = cgraph->nodes[i + 3];
-                                }
-
-                                if (args.norm) {
-                                    ops.insert(ops.end(), { GGML_OP_RESHAPE, GGML_OP_SUM_ROWS, GGML_OP_CLAMP,
-                                                            GGML_OP_DIV, GGML_OP_RESHAPE });
-                                    clamp = cgraph->nodes[i + ops.size() - 3];
-                                }
-                                if (args.scale) {
-                                    ops.insert(ops.end(), { GGML_OP_SCALE });
-                                    scale = cgraph->nodes[i + ops.size() - 1];
-                                }
-
-                                weights      = cgraph->nodes[i + ops.size() - 1];
-                                out_nodes[1] = i + ops.size() - 1;
-
-                                if (ggml_can_fuse_subgraph(cgraph, i, ops.size(), ops.data(), out_nodes, 2) &&
-                                    ggml_cuda_should_use_topk_moe(node, logits, weights, ids) &&
-                                    ggml_cuda_check_fusion_memory_ranges(cgraph, i, ops.size(), out_nodes, 2, /*is_topk_moe=*/ true)) {
-                                    ggml_cuda_op_topk_moe(*cuda_ctx, logits, weights, ids, clamp, scale, bias, args);
-                                    i += ops.size() - 1;
-                                    continue;
-                                }
-                            } else if (!args.norm && !args.prob_bias) {
-                                //special case gpt-oss, no norm, no bias.
-                                ops.insert(ops.end(), { GGML_OP_ARGSORT, GGML_OP_VIEW, GGML_OP_GET_ROWS,
-                                                        GGML_OP_RESHAPE, GGML_OP_SOFT_MAX, GGML_OP_RESHAPE });
-                                weights                     = cgraph->nodes[i + 5];
-                                ids                         = cgraph->nodes[i + 1];
-                                const ggml_tensor * softmax = cgraph->nodes[i + 4];
-
-                                int out_nodes[2] = { i + 1, i + 5 };
-                                if (ggml_can_fuse_subgraph(cgraph, i, ops.size(), ops.data(), out_nodes, 2) &&
-                                    ggml_cuda_should_use_topk_moe(softmax, logits, weights, ids) &&
-                                    ggml_cuda_check_fusion_memory_ranges(cgraph, i, ops.size(), out_nodes, 2, /*is_topk_moe=*/ true)) {
-                                    ggml_cuda_op_topk_moe(*cuda_ctx, logits, weights, ids, clamp, scale, bias, args);
-                                    i += ops.size() - 1;
-                                    continue;
-                                }
-                            }
-                        }
-                    }
-
-                    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_ROPE, GGML_OP_VIEW, GGML_OP_SET_ROWS }, {})) {
-                        ggml_tensor * rope = cgraph->nodes[i];
-                        ggml_tensor * set_rows = cgraph->nodes[i + 2];
-
-                        ggml_cuda_op_rope_fused(*cuda_ctx, rope, set_rows);
-                        i += 2;
-                        continue;
-                    }
-
-                    if (node->op == GGML_OP_ADD || node->op == GGML_OP_MUL) {
-                        int n_fuse = 0;
-                        ggml_op ops[8];
-                        std::fill(ops, ops + 8, node->op);
-
-                        for (; n_fuse <= 6; ++n_fuse){
-                            if (!ggml_can_fuse(cgraph, i + n_fuse, ops + n_fuse, 2)) {
-                                break;
-                            }
-                            if (cgraph->nodes[i + n_fuse] != cgraph->nodes[i + n_fuse + 1]->src[0]) {
-                                break;
-                            }
-                            if (!ggml_are_same_layout(cgraph->nodes[i + n_fuse]->src[1], cgraph->nodes[i + n_fuse + 1]->src[1])) {
-                                break;
-                            }
-                        }
-
-                        n_fuse++;
-
-                        if (n_fuse > 1) {
-                            ggml_tensor fused_node;
-                            memcpy(&fused_node, node, sizeof(ggml_tensor));
-                            for (int j = 0; j < n_fuse - 1; ++j) {
-                                fused_node.src[j + 2] = cgraph->nodes[i + j + 1]->src[1];
-                            }
-                            fused_node.data = cgraph->nodes[i + n_fuse - 1]->data;
-                            if (node->op == GGML_OP_ADD) {
-                                ggml_cuda_op_fused_add(*cuda_ctx, &fused_node, n_fuse);
-                            } else {
-                                ggml_cuda_op_fused_mul(*cuda_ctx, &fused_node, n_fuse);
-                            }
-                            i += n_fuse - 1;
-
-                            continue;
-                        }
-                    }
-
-                    bool fused_mul_mat_vec = false;
-                    int fused_node_count = 0;
-
-                    for (ggml_op op : { GGML_OP_MUL_MAT, GGML_OP_MUL_MAT_ID }) {
-                        const ggml_op bias_op = op == GGML_OP_MUL_MAT ? GGML_OP_ADD : GGML_OP_ADD_ID;
-
-                        if (ggml_cuda_can_fuse(cgraph, i, { op, bias_op, op, bias_op, GGML_OP_GLU }, {})) {
-                            ggml_tensor * glu         = cgraph->nodes[i + 4];
-                            ggml_tensor * gate_bias_n = glu->src[0];
-                            ggml_tensor * up_bias_n   = glu->src[1];
-
-                            //we don't assume the order for {gate, up}. Instead infer it from the bias tensor
-                            ggml_tensor * gate_n      = nullptr;
-                            ggml_tensor * up_n        = nullptr;
-
-                            if (gate_bias_n->src[0] == cgraph->nodes[i] || gate_bias_n->src[1] == cgraph->nodes[i]) {
-                                gate_n = cgraph->nodes[i];
-                                up_n   = cgraph->nodes[i + 2];
-                            } else if (gate_bias_n->src[0] == cgraph->nodes[i + 2] || gate_bias_n->src[1] == cgraph->nodes[i + 2]) {
-                                gate_n = cgraph->nodes[i + 2];
-                                up_n   = cgraph->nodes[i];
-                            } else {
-                                continue;
-                            }
-
-                            auto get_bias_tensor = [](const ggml_tensor * bias_node, const ggml_tensor * mul_node, ggml_op op_bias) {
-                                if (op_bias == GGML_OP_ADD) {
-                                    if (bias_node->src[0] == mul_node) {
-                                        return bias_node->src[1];
-                                    }
-                                    if (bias_node->src[1] == mul_node) {
-                                        return bias_node->src[0];
-                                    }
-                                    return (ggml_tensor *) nullptr;
-                                }
-                                GGML_ASSERT(op_bias == GGML_OP_ADD_ID);
-                                GGML_ASSERT(bias_node->src[0] == mul_node);
-                                return bias_node->src[1];
-                            };
-
-                            ggml_tensor * up_bias_tensor   = get_bias_tensor(up_bias_n, up_n, bias_op);
-                            ggml_tensor * gate_bias_tensor = get_bias_tensor(gate_bias_n, gate_n, bias_op);
-
-                            if (!up_bias_tensor || !gate_bias_tensor) {
-                                continue;
-                            }
-
-                            // we don't support repeating adds
-                            if (bias_op == GGML_OP_ADD &&
-                                (!ggml_are_same_shape(gate_bias_n->src[0], gate_bias_n->src[1]) ||
-                                 !ggml_are_same_shape(up_bias_n->src[0], up_bias_n->src[1]))) {
-                                continue;
-                            }
-
-                            const ggml_tensor * src0 = up_n->src[0];
-                            const ggml_tensor * src1 = up_n->src[1];
-                            const ggml_tensor * ids  = up_n->src[2];
-
-                            if (ggml_cuda_should_fuse_mul_mat_vec_f(up_n)) {
-                                ggml_cuda_mm_fusion_args_host fusion_data{};
-                                fusion_data.gate      = gate_n->src[0];
-                                fusion_data.x_bias    = up_bias_tensor;
-                                fusion_data.gate_bias = gate_bias_tensor;
-                                fusion_data.glu_op    = ggml_get_glu_op(glu);
-
-                                ggml_cuda_mul_mat_vec_f(*cuda_ctx, src0, src1, ids, glu, &fusion_data);
-                                fused_mul_mat_vec = true;
-                                fused_node_count = 5;
-                                break;
-                            }
-
-                            if (ggml_cuda_should_fuse_mul_mat_vec_q(up_n)) {
-                                ggml_cuda_mm_fusion_args_host fusion_data{};
-                                fusion_data.gate      = gate_n->src[0];
-                                fusion_data.x_bias    = up_bias_tensor;
-                                fusion_data.gate_bias = gate_bias_tensor;
-                                fusion_data.glu_op    = ggml_get_glu_op(glu);
-
-                                ggml_cuda_mul_mat_vec_q(*cuda_ctx, src0, src1, ids, glu, &fusion_data);
-                                fused_mul_mat_vec = true;
-                                fused_node_count = 5;
-                                break;
-                            }
-                        } else if (ggml_cuda_can_fuse(cgraph, i, { op, op, GGML_OP_GLU }, {})) {
-                            ggml_tensor * glu  = cgraph->nodes[i + 2];
-                            ggml_tensor * gate = glu->src[0];
-                            ggml_tensor * up   = glu->src[1];
-
-                            bool ok = (gate == cgraph->nodes[i] && up == cgraph->nodes[i + 1])
-                                || (gate == cgraph->nodes[i + 1] && up == cgraph->nodes[i]);
-
-                            if (!ok) continue;
-
-                            const ggml_tensor * src0 = up->src[0];
-                            const ggml_tensor * src1 = up->src[1];
-                            const ggml_tensor * ids  = up->src[2];
-
-                            if (ggml_cuda_should_fuse_mul_mat_vec_f(up)) {
-                                ggml_cuda_mm_fusion_args_host fusion_data{};
-                                fusion_data.gate   = gate->src[0];
-                                fusion_data.glu_op = ggml_get_glu_op(glu);
-
-                                ggml_cuda_mul_mat_vec_f(*cuda_ctx, src0, src1, ids, glu, &fusion_data);
-                                fused_mul_mat_vec = true;
-                                fused_node_count = 3;
-                                break;
-                            }
-
-                            if (ggml_cuda_should_fuse_mul_mat_vec_q(up)) {
-                                ggml_cuda_mm_fusion_args_host fusion_data{};
-                                fusion_data.gate   = gate->src[0];
-                                fusion_data.glu_op = ggml_get_glu_op(glu);
-
-                                ggml_cuda_mul_mat_vec_q(*cuda_ctx, src0, src1, ids, glu, &fusion_data);
-                                fused_mul_mat_vec = true;
-                                fused_node_count = 3;
-                                break;
-                            }
-                        }
-                    }
-
-                    if (fused_mul_mat_vec) {
-                        i += fused_node_count - 1;
-                        continue;
-                    }
-
-                    fused_mul_mat_vec = false;
-                    fused_node_count = 0;
-
-                    for (ggml_op op : { GGML_OP_MUL_MAT, GGML_OP_MUL_MAT_ID }) {
-                        const ggml_op bias_op = op == GGML_OP_MUL_MAT ? GGML_OP_ADD : GGML_OP_ADD_ID;
-
-                        if (!ggml_can_fuse(cgraph, i, { op, bias_op })) {
-                            continue;
-                        }
-
-                        ggml_tensor * mm_node   = cgraph->nodes[i];
-                        ggml_tensor * bias_node = cgraph->nodes[i + 1];
-
-                        ggml_tensor * bias_tensor = nullptr;
-                        if (bias_op == GGML_OP_ADD) {
-                            if (bias_node->src[0] == mm_node) {
-                                bias_tensor = bias_node->src[1];
-                            } else if (bias_node->src[1] == mm_node) {
-                                bias_tensor = bias_node->src[0];
-                            } else {
-                                continue;
-                            }
-                        } else {
-                            if (bias_node->src[0] != mm_node) {
-                                continue;
-                            }
-                            bias_tensor = bias_node->src[1];
-                        }
-
-                        const ggml_tensor * src0 = mm_node->src[0];
-                        const ggml_tensor * src1 = mm_node->src[1];
-                        const ggml_tensor * ids  = mm_node->src[2];
-
-                        if (bias_op == GGML_OP_ADD_ID && bias_node->src[2] != ids) {
-                            continue;
-                        }
-
-                        if (bias_op == GGML_OP_ADD && !ggml_are_same_shape(bias_node->src[0], bias_node->src[1])) {
-                            continue;
-                        }
-
-                        ggml_cuda_mm_fusion_args_host fusion_data{};
-                        fusion_data.x_bias = bias_tensor;
-
-                        if (ggml_cuda_should_fuse_mul_mat_vec_f(mm_node)) {
-                            ggml_cuda_mul_mat_vec_f(*cuda_ctx, src0, src1, ids, bias_node, &fusion_data);
-                            fused_mul_mat_vec = true;
-                            fused_node_count = 2;
-                            break;
-                        }
-
-                        if (ggml_cuda_should_fuse_mul_mat_vec_q(mm_node)) {
-                            ggml_cuda_mul_mat_vec_q(*cuda_ctx, src0, src1, ids, bias_node, &fusion_data);
-                            fused_mul_mat_vec = true;
-                            fused_node_count = 2;
-                            break;
-                        }
-                    }
-
-                    if (fused_mul_mat_vec) {
-                        i += fused_node_count - 1;
-                        continue;
-                    }
-
-                    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_RMS_NORM, GGML_OP_MUL, GGML_OP_ADD}, {})) {
-                        ggml_cuda_op_rms_norm_fused_add(*cuda_ctx, node, cgraph->nodes[i+1], cgraph->nodes[i+2]);
-                        i += 2;
-                        continue;
-                    }
-
-                    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_RMS_NORM, GGML_OP_MUL}, {})) {
-                        ggml_cuda_op_rms_norm_fused(*cuda_ctx, node, cgraph->nodes[i+1]);
-                        i++;
-                        continue;
-                    }
-
-                    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_SSM_CONV, GGML_OP_UNARY }, { GGML_UNARY_OP_SILU })) {
-                        ggml_cuda_op_ssm_conv(*cuda_ctx, node, cgraph->nodes[i+1]);
-                        i++;
-                        continue;
-                    }
-
-                    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_UNARY, GGML_OP_MUL }, { GGML_UNARY_OP_SILU }) ||
-                        ggml_cuda_can_fuse(cgraph, i, { GGML_OP_UNARY, GGML_OP_MUL }, { GGML_UNARY_OP_SIGMOID }) ||
-                        ggml_cuda_can_fuse(cgraph, i, { GGML_OP_UNARY, GGML_OP_MUL }, { GGML_UNARY_OP_SOFTPLUS })) {
-                        ggml_cuda_op_unary_mul(*cuda_ctx, node, cgraph->nodes[i+1]);
-                        i++;
-                        continue;
-                    }
-
-                    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_UNARY, GGML_OP_SQR }, { GGML_UNARY_OP_RELU })) {
-                        ggml_cuda_op_relu_sqr(*cuda_ctx, node, cgraph->nodes[i+1]);
-                        i++;
-                        continue;
-                    }
-
-                    if (ggml_cuda_can_fuse(cgraph, i, { GGML_OP_SCALE, GGML_OP_UNARY, GGML_OP_SCALE }, { GGML_UNARY_OP_TANH })) {
-                        i += 2;
-                        ggml_cuda_op_softcap(*cuda_ctx, cgraph->nodes[i], node);
-                        continue;
-                    }
+                if (nodes_to_skip != 0) {
+                    i += nodes_to_skip;
+                    continue;
                 }
 #ifndef NDEBUG
                 assert(node->buffer->buft == ggml_backend_cuda_buffer_type(cuda_ctx->device));
@@ -4548,8 +4588,8 @@ static const ggml_backend_i ggml_backend_cuda_interface = {
     /* .free                    = */ ggml_backend_cuda_free,
     /* .set_tensor_async        = */ ggml_backend_cuda_set_tensor_async,
     /* .get_tensor_async        = */ ggml_backend_cuda_get_tensor_async,
-    /* .get_tensor_2d_async     = */ ggml_backend_cuda_set_tensor_2d_async,
-    /* .set_tensor_2d_async     = */ ggml_backend_cuda_get_tensor_2d_async,
+    /* .set_tensor_2d_async     = */ ggml_backend_cuda_set_tensor_2d_async,
+    /* .get_tensor_2d_async     = */ ggml_backend_cuda_get_tensor_2d_async,
     /* .cpy_tensor_async        = */ ggml_backend_cuda_cpy_tensor_async,
     /* .synchronize             = */ ggml_backend_cuda_synchronize,
     /* .graph_plan_create       = */ NULL,

ggml/src/ggml-cuda/mma.cuh

@@ -1015,25 +1015,35 @@ namespace ggml_cuda_mma {
 #endif // AMD_MFMA_AVAILABLE
     }
 
-    static __device__ __forceinline__ void mma_block_scaled(tile<16, 8, float> &     D,
-                                                            const tile<16, 8, int> & A,
-                                                            const tile<8, 8, int> &  B,
-                                                            uint32_t                 a_scale,
-                                                            uint32_t                 b_scale) {
+    template <ggml_type type>
+    static __device__ __forceinline__ void mma_block_scaled_fp4(tile<16, 8, float> &     D,
+                                                                const tile<16, 8, int> & A,
+                                                                const tile<8, 8, int> &  B,
+                                                                uint32_t                 a_scale,
+                                                                uint32_t                 b_scale) {
 #ifdef BLACKWELL_MMA_AVAILABLE
         const int * Axi = (const int *) A.x;
         const int * Bxi = (const int *) B.x;
         float *     Dxi = (float *) D.x;
 
-        asm volatile(
-            "mma.sync.aligned.kind::mxf4.block_scale.scale_vec::2X.m16n8k64.row.col.f32.e2m1.e2m1.f32.ue8m0 "
-            "{%0, %1, %2, %3}, {%4, %5, %6, %7}, {%8, %9}, {%0, %1, %2, %3}, "
-            "%10, {0, 0}, %11, {0, 0};"
-            : "+f"(Dxi[0]), "+f"(Dxi[1]), "+f"(Dxi[2]), "+f"(Dxi[3])
-            : "r"(Axi[0]), "r"(Axi[1]), "r"(Axi[2]), "r"(Axi[3]), "r"(Bxi[0]), "r"(Bxi[1]), "r"(a_scale), "r"(b_scale));
+        if constexpr (type == GGML_TYPE_MXFP4) {
+            asm volatile(
+                "mma.sync.aligned.kind::mxf4.block_scale.scale_vec::2X.m16n8k64.row.col.f32.e2m1.e2m1.f32.ue8m0 "
+                "{%0, %1, %2, %3}, {%4, %5, %6, %7}, {%8, %9}, {%0, %1, %2, %3}, "
+                "%10, {0, 0}, %11, {0, 0};"
+                : "+f"(Dxi[0]), "+f"(Dxi[1]), "+f"(Dxi[2]), "+f"(Dxi[3])
+                : "r"(Axi[0]), "r"(Axi[1]), "r"(Axi[2]), "r"(Axi[3]), "r"(Bxi[0]), "r"(Bxi[1]), "r"(a_scale), "r"(b_scale));
+        } else {
+            asm volatile(
+                "mma.sync.aligned.kind::mxf4nvf4.block_scale.scale_vec::4X.m16n8k64.row.col.f32.e2m1.e2m1.f32.ue4m3 "
+                "{%0, %1, %2, %3}, {%4, %5, %6, %7}, {%8, %9}, {%0, %1, %2, %3}, "
+                "%10, {0, 0}, %11, {0, 0};"
+                : "+f"(Dxi[0]), "+f"(Dxi[1]), "+f"(Dxi[2]), "+f"(Dxi[3])
+                : "r"(Axi[0]), "r"(Axi[1]), "r"(Axi[2]), "r"(Axi[3]), "r"(Bxi[0]), "r"(Bxi[1]), "r"(a_scale), "r"(b_scale));
+        }
 #else
         GGML_UNUSED_VARS(D, A, B, a_scale, b_scale);
-#endif  // BLACKWELL_MMA_AVAILABLE
+#endif // BLACKWELL_MMA_AVAILABLE
     }
 
     static __device__ __forceinline__ void mma(

ggml/src/ggml-cuda/mmq.cu

@@ -122,7 +122,7 @@ void ggml_cuda_mul_mat_q(
                             || GGML_CUDA_CC_IS_CDNA(cc);
 
     // TODO: tighter pool buffer size vs q8 path
-    const bool use_native_mxfp4 = blackwell_mma_available(cc) && src0->type == GGML_TYPE_MXFP4;
+    const bool use_native_fp4 = blackwell_mma_available(cc) && (src0->type == GGML_TYPE_MXFP4 || src0->type == GGML_TYPE_NVFP4);
 
     if (!ids) {
         const size_t nbytes_src1_q8_1 = ne13*ne12 * ne11*ne10_padded * sizeof(block_q8_1)/QK8_1 +
@@ -133,9 +133,9 @@ void ggml_cuda_mul_mat_q(
             const int64_t s11 = src1->nb[1] / ts_src1;
             const int64_t s12 = src1->nb[2] / ts_src1;
             const int64_t s13 = src1->nb[3] / ts_src1;
-            if (use_native_mxfp4) {
+            if (use_native_fp4) {
                 static_assert(sizeof(block_fp4_mmq) == 4 * sizeof(block_q8_1));
-                quantize_mmq_mxfp4_cuda(src1_d, nullptr, src1_q8_1.get(), src0->type, ne10, s11, s12, s13, ne10_padded,
+                quantize_mmq_fp4_cuda(src1_d, nullptr, src1_q8_1.get(), src0->type, ne10, s11, s12, s13, ne10_padded,
                                         ne11, ne12, ne13, stream);
 
             } else {
@@ -146,10 +146,8 @@ void ggml_cuda_mul_mat_q(
         }
 
         // Stride depends on quantization format
-        const int64_t s12 = use_native_mxfp4 ?
-                                ne11 * ne10_padded * sizeof(block_fp4_mmq) /
-                                    (8 * QK_MXFP4 * sizeof(int))  // block_fp4_mmq holds 256 values (8 blocks of 32)
-                                :
+        const int64_t s12 = use_native_fp4 ?
+                                ne11 * ne10_padded * sizeof(block_fp4_mmq) / (QK_K * sizeof(int)) :  // block_fp4_mmq holds 256 values
                                 ne11 * ne10_padded * sizeof(block_q8_1) / (QK8_1 * sizeof(int));
         const int64_t s13 = ne12*s12;
 
@@ -198,8 +196,8 @@ void ggml_cuda_mul_mat_q(
         const int64_t s12 = src1->nb[2] / ts_src1;
         const int64_t s13 = src1->nb[3] / ts_src1;
 
-        if (use_native_mxfp4) {
-            quantize_mmq_mxfp4_cuda(src1_d, ids_src1.get(), src1_q8_1.get(), src0->type, ne10, s11, s12, s13,
+        if (use_native_fp4) {
+            quantize_mmq_fp4_cuda(src1_d, ids_src1.get(), src1_q8_1.get(), src0->type, ne10, s11, s12, s13,
                                     ne10_padded, ne11_flat, ne12_flat, ne13_flat, stream);
         } else {
             quantize_mmq_q8_1_cuda(src1_d, ids_src1.get(), src1_q8_1.get(), src0->type, ne10, s11, s12, s13,
@@ -208,8 +206,9 @@ void ggml_cuda_mul_mat_q(
         CUDA_CHECK(cudaGetLastError());
     }
 
-    const int64_t s12 = use_native_mxfp4 ? ne11 * ne10_padded * sizeof(block_fp4_mmq) / (8 * QK_MXFP4 * sizeof(int)) :
-                                           ne11 * ne10_padded * sizeof(block_q8_1) / (QK8_1 * sizeof(int));
+    static_assert(QK_K == 8 * QK_MXFP4, "QK_K needs to be 8 * QK_MXFP4");
+    const int64_t s12 = use_native_fp4 ? ne11 * ne10_padded * sizeof(block_fp4_mmq) / (QK_K * sizeof(int)) :
+                                         ne11 * ne10_padded * sizeof(block_q8_1) / (QK8_1 * sizeof(int));
     const int64_t s13 = ne12*s12;
 
     // Note that ne02 is used instead of ne12 because the number of y channels determines the z dimension of the CUDA grid.

ggml/src/ggml-cuda/mmq.cuh

@@ -10,9 +10,9 @@
 using namespace ggml_cuda_mma;
 
 #define MMQ_DP4A_MAX_BATCH_SIZE 64 // Max. batch size to use for dp4a MMQ kernels when FP16 tensor cores are available.
-#define MMQ_ITER_K 256
-#define MMQ_ITER_K_MXFP4_FP4    512
-#define MMQ_NWARPS 8
+#define MMQ_ITER_K             256
+#define MMQ_ITER_K_FP4         512
+#define MMQ_NWARPS               8
 
 typedef void (*load_tiles_mmq_t)(const char * __restrict__ x, int * x_tile, const int kbx0, const int i_max, const int stride);
 typedef void (*vec_dot_mmq_t)(const int * __restrict__ x, const int * __restrict__ y, float * __restrict__ sum, const int k00);
@@ -46,9 +46,12 @@ struct block_q8_1_mmq {
     int8_t qs[4*QK8_1]; // 128 values quantized to 8 bit each
 };
 
+// this struct is used for fp4 data types (currently only used for Blackwell)
+// mxfp4 has block size 32, each int32 of d4 contains 2 e8m0 scales in the lower 16 bits
+// nvfp4 has block size 16, each int32 of d4 contains 4 ue4m3 scales
 struct block_fp4_mmq {
-    uint32_t d4[4];       // 8 E8M0 scales (1 per 32 values), 2 packed per uint32: d4[0]={s0,s1}, d4[1]={s2,s3}, etc.
-    int8_t   qs[4 * 32];  // 256 FP4 values packed as 4-bit pairs (2 per byte), 8 blocks of 32 values
+    uint32_t d4[4];
+    int8_t   qs[4 * 32];  // 256 FP4 values packed as 4-bit pairs (2 per byte)
 };
 
 static_assert(sizeof(block_q8_1_mmq) == 4*QK8_1 + 4*sizeof(half2), "Unexpected block_q8_1_mmq size");
@@ -143,10 +146,11 @@ static int get_mmq_y_host(const int cc) {
 
 static constexpr __device__ int get_iter_k([[maybe_unused]] const ggml_type type) {
 #if defined(BLACKWELL_MMA_AVAILABLE)
-    return type == GGML_TYPE_MXFP4 ? MMQ_ITER_K_MXFP4_FP4 : MMQ_ITER_K;
-#else
-    return MMQ_ITER_K;
+if (type == GGML_TYPE_NVFP4 || type == GGML_TYPE_MXFP4) {
+    return MMQ_ITER_K_FP4;
+}
 #endif // defined(BLACKWELL_MMA_AVAILABLE)
+    return MMQ_ITER_K;
 }
 
 static constexpr __device__ int get_mmq_y_device() {
@@ -213,8 +217,8 @@ static constexpr __host__ __device__ tile_x_sizes mmq_get_dp4a_tile_x_sizes(ggml
 }
 
 #define MMQ_MMA_TILE_X_K_Q8_0  (2*MMQ_TILE_NE_K + 2*MMQ_TILE_NE_K/QI8_0                   + 4)
-#define MMQ_MMA_TILE_X_K_FP4   (2*MMQ_TILE_NE_K + 8                                       + 4) // MXFP4
-#define MMQ_MMA_TILE_X_K_NVFP4 (2*MMQ_TILE_NE_K + MMQ_TILE_NE_K/2                         + 4) // NVFP4
+#define MMQ_MMA_TILE_X_K_FP4   (2*MMQ_TILE_NE_K + 8                                       + 4) // MXFP4 and NVFP4 Blackwell
+#define MMQ_MMA_TILE_X_K_NVFP4 (2*MMQ_TILE_NE_K + MMQ_TILE_NE_K/2                         + 4) // NVFP4 Generic
 #define MMQ_MMA_TILE_X_K_Q8_1  (2*MMQ_TILE_NE_K + 2*MMQ_TILE_NE_K/QI8_0                   + 4)
 #define MMQ_MMA_TILE_X_K_Q2_K  (2*MMQ_TILE_NE_K + MMQ_TILE_NE_K                           + 4)
 #define MMQ_MMA_TILE_X_K_Q3_K  (2*MMQ_TILE_NE_K + MMQ_TILE_NE_K/2                         + 4)
@@ -240,7 +244,11 @@ static constexpr __host__ __device__ int mmq_get_mma_tile_x_k(ggml_type type) {
         case GGML_TYPE_Q8_0:    return MMQ_MMA_TILE_X_K_Q8_0;
         // tile sizes are the same for Q8_1 and FP4 for blackwell
         case GGML_TYPE_MXFP4:   return MMQ_MMA_TILE_X_K_Q8_1;
+#if defined(BLACKWELL_MMA_AVAILABLE)
+        case GGML_TYPE_NVFP4:   return MMQ_MMA_TILE_X_K_FP4;
+#else
         case GGML_TYPE_NVFP4:   return MMQ_MMA_TILE_X_K_NVFP4;
+#endif // defined(BLACKWELL_MMA_AVAILABLE)
         case GGML_TYPE_Q2_K:    return MMQ_MMA_TILE_X_K_Q2_K;
         case GGML_TYPE_Q3_K:    return MMQ_MMA_TILE_X_K_Q3_K;
         case GGML_TYPE_Q4_K:    return MMQ_MMA_TILE_X_K_Q8_1;
@@ -934,6 +942,128 @@ static __device__ __forceinline__ void load_tiles_mxfp4_fp4(const char * __restr
     }
 }
 
+#ifdef BLACKWELL_MMA_AVAILABLE
+template <int mmq_y, bool need_check>
+static __device__ __forceinline__ void load_tiles_nvfp4_nvfp4(const char * __restrict__ x,
+                                                            int * __restrict__ x_tile,
+                                                            const int kbx0,
+                                                            const int i_max,
+                                                            const int stride) {
+    constexpr int nwarps = mmq_get_nwarps_device();
+    constexpr int warp_size = ggml_cuda_get_physical_warp_size();
+    constexpr int iter_k = get_iter_k(GGML_TYPE_NVFP4);
+    constexpr int threads_per_row = iter_k / QK_NVFP4; // each thread processes 1 block
+    constexpr int rows_per_warp = warp_size / threads_per_row;
+
+    uint32_t * x_u32 = (uint32_t *) x_tile;
+
+    const int txi = threadIdx.x;
+    const int kbx = txi % threads_per_row;
+    const int row_in_warp = txi / threads_per_row;
+
+    const block_nvfp4 * bxi_base = (const block_nvfp4 *) x + kbx0 + kbx;
+    uint32_t * x_u32_scale = x_u32 + 64 + kbx;
+
+#pragma unroll
+    for (int i0 = 0; i0 < mmq_y; i0 += rows_per_warp * nwarps) {
+        int i = i0 + threadIdx.y * rows_per_warp + row_in_warp;
+
+        if constexpr (need_check) {
+            i = min(i, i_max);
+        }
+
+        const block_nvfp4 * bxi = bxi_base + i * stride;
+        const int row_base = i * MMQ_MMA_TILE_X_K_FP4;
+        const int q_base = row_base + 8 * kbx;
+
+        const uint32_t * src_qs = reinterpret_cast<const uint32_t *>(bxi->qs);
+
+#pragma unroll
+        for (int sub = 0; sub < QK_NVFP4 / QK_NVFP4_SUB; ++sub) {
+            x_u32[q_base + 2 * sub + 0] = src_qs[2 * sub + 0];
+            x_u32[q_base + 2 * sub + 1] = src_qs[2 * sub + 1];
+        }
+
+        x_u32_scale[row_base] = get_int_b4(bxi->d, 0);
+    }
+}
+
+// Shared MMA kernel for MXFP4 and NVFP4 on Blackwell.
+// Both quantizations encode values as e2m1 (FP4) and produce one uint32 scale per
+// m16n8k64 MMA call; only the PTX kind (scale_vec::2X ue8m0 vs scale_vec::4X ue4m3)
+// and the per-type stride constant differ.
+template <int mmq_x, int mmq_y, ggml_type type>
+static __device__ __forceinline__ void vec_dot_fp4_fp4_mma(const int * __restrict__ x,
+                                                           const int * __restrict__ y,
+                                                           float * __restrict__ sum,
+                                                           const int k00) {
+    static_assert(type == GGML_TYPE_MXFP4 || type == GGML_TYPE_NVFP4,
+                  "vec_dot_fp4_fp4_mma: type must be MXFP4 or NVFP4");
+
+    typedef tile<16, 8, int>   tile_A;
+    typedef tile<8, 8, int>    tile_B;
+    typedef tile<16, 8, float> tile_C;
+
+    constexpr int stride        = MMQ_MMA_TILE_X_K_FP4;
+    constexpr int granularity   = mmq_get_granularity_device(mmq_x);
+    constexpr int rows_per_warp = 2 * granularity;
+    constexpr int ntx           = rows_per_warp / tile_C::I;
+    constexpr int nfrags        = MMQ_TILE_NE_K / tile_A::J;
+
+    y += (threadIdx.y % ntx) * (tile_C::J * MMQ_TILE_Y_K);
+
+    const int *      x_qs = (const int *) x;
+    const uint32_t * x_sc = (const uint32_t *) (x_qs + 2 * MMQ_TILE_NE_K);
+    const int *      y_qs = (const int *) y + 4;
+    const uint32_t * y_sc = (const uint32_t *) y;
+
+    // 2 threads per quad supply the packed scale register to the block_scale MMA,
+    // see https://docs.nvidia.com/cuda/parallel-thread-execution/#warp-level-block-scaling
+    const int tidx_A = threadIdx.x / 4 + (threadIdx.x % 2) * 8;
+    const int tidx_B = threadIdx.x / 4;
+    const int i0     = (threadIdx.y / ntx) * rows_per_warp;
+
+    tile_A   A[ntx][nfrags];
+    uint32_t scaleA[ntx][nfrags];
+
+#pragma unroll
+    for (int n = 0; n < ntx; ++n) {
+#pragma unroll
+        for (int frag = 0; frag < nfrags; ++frag) {
+            const int k0 = k00 + frag * tile_A::J;
+            load_ldmatrix(A[n][frag], x_qs + (i0 + n * tile_A::I) * stride + k0, stride);
+            scaleA[n][frag] = x_sc[(i0 + n * tile_A::I + tidx_A) * stride + k0 / tile_A::J];
+        }
+    }
+
+#pragma unroll
+    for (int j0 = 0; j0 < mmq_x; j0 += ntx * tile_C::J) {
+        tile_B   B[nfrags];
+        uint32_t scaleB[nfrags];
+
+#pragma unroll
+        for (int frag = 0; frag < nfrags; ++frag) {
+            const int k0 = frag * tile_B::J;
+            load_generic(B[frag], y_qs + j0 * MMQ_TILE_Y_K + k0, MMQ_TILE_Y_K);
+            scaleB[frag] = y_sc[(j0 + tidx_B) * MMQ_TILE_Y_K + frag];
+        }
+
+#pragma unroll
+        for (int n = 0; n < ntx; ++n) {
+#pragma unroll
+            for (int frag = 0; frag < nfrags; ++frag) {
+                tile_C C = {};
+                mma_block_scaled_fp4<type>(C, A[n][frag], B[frag], scaleA[n][frag], scaleB[frag]);
+#pragma unroll
+                for (int l = 0; l < tile_C::ne; ++l) {
+                    sum[(j0 / tile_C::J + n) * tile_C::ne + l] += C.x[l];
+                }
+            }
+        }
+    }
+}
+#endif // BLACKWELL_MMA_AVAILABLE
+
 
 template <int mmq_y, bool need_check>
 static __device__ __forceinline__ void load_tiles_nvfp4(const char * __restrict__ x,
@@ -1163,77 +1293,6 @@ static __device__ __forceinline__ void vec_dot_q8_0_q8_1_mma(
 #endif // defined(AMD_MFMA_AVAILABLE) || defined(AMD_WMMA_AVAILABLE)
 }
 
-template <int mmq_x, int mmq_y>
-static __device__ __forceinline__ void vec_dot_mxfp4_mxfp4_mma(const int * __restrict__ x,
-                                                               const int * __restrict__ y,
-                                                               float * __restrict__ sum,
-                                                               const int k00) {
-    typedef tile<16, 8, int>   tile_A;
-    typedef tile<8, 8, int>    tile_B;
-    typedef tile<16, 8, float> tile_C;  // Output is float for native scaled MMA
-
-    constexpr int granularity   = mmq_get_granularity_device(mmq_x);
-    constexpr int rows_per_warp = 2 * granularity;
-    constexpr int ntx           = rows_per_warp / tile_C::I;  // Number of x minitiles per warp.
-
-    y += (threadIdx.y % ntx) * (tile_C::J * MMQ_TILE_Y_FP4_K);
-
-    // Match layout from load_tiles_mxfp4_fp4
-    const int *      x_qs = (const int *) x;
-    const uint32_t * x_sc = (const uint32_t *) (x_qs + 2 * MMQ_TILE_NE_K);
-    const int *      y_qs = (const int *) y + 4;
-    const uint32_t * y_sc = (const uint32_t *) y;
-
-    // tile_A has a length of 64 logical values vs. 32 values in block_mxfp4
-    tile_A   A[ntx][MMQ_TILE_NE_K / (2 * QI_MXFP4)];
-    uint32_t scaleA[ntx][MMQ_TILE_NE_K / (2 * QI_MXFP4)];
-
-    // Block scale
-    // Each thread has to point to a 4 byte scale value
-    // https://docs.nvidia.com/cuda/parallel-thread-execution/#warp-level-block-scaling
-
-    const int i0 = (threadIdx.y / ntx) * rows_per_warp;
-
-#pragma unroll
-    for (int n = 0; n < ntx; ++n) {
-#pragma unroll
-        for (int k01 = 0; k01 < MMQ_TILE_NE_K; k01 += 2 * QI_MXFP4) {
-            const int k0 = k00 + k01;
-
-            load_ldmatrix(A[n][k01 / (2 * QI_MXFP4)], x_qs + (i0 + n * tile_A::I) * MMQ_MMA_TILE_X_K_FP4 + k0,
-                          MMQ_MMA_TILE_X_K_FP4);
-
-            // based on block-scaling document, 2 threads in each quad need to supply to the scale value
-            const int tidx         = threadIdx.x / 4 + (threadIdx.x % 2) * 8;
-            scaleA[n][k01 / (2 * QI_MXFP4)] =
-                *(x_sc + (i0 + n * tile_A::I + tidx) * MMQ_MMA_TILE_X_K_FP4 + k0 / (2 * QI_MXFP4));
-        }
-    }
-
-#pragma unroll
-    for (int j0 = 0; j0 < mmq_x; j0 += ntx * tile_C::J) {
-#pragma unroll
-        for (int k01 = 0; k01 < MMQ_TILE_NE_K; k01 += 2 * QI_MXFP4) {
-            tile_B   B;
-            uint32_t scaleB;  // 2xN scales
-
-            load_generic(B, y_qs + j0 * MMQ_TILE_Y_FP4_K + k01, MMQ_TILE_Y_FP4_K);
-
-            scaleB = y_sc[(j0 + threadIdx.x / 4) * MMQ_TILE_Y_FP4_K + k01 / (2 * QI_MXFP4)];
-
-#pragma unroll
-            for (int n = 0; n < ntx; ++n) {
-                tile_C C;
-
-                mma_block_scaled(C, A[n][k01 / (2 * QI_MXFP4)], B, scaleA[n][k01 / (2 * QI_MXFP4)], scaleB);
-#pragma unroll
-                for (int l = 0; l < tile_C::ne; ++l) {
-                    sum[(j0 / tile_C::J + n) * tile_C::ne + l] += C.x[l];
-                }
-            }
-        }
-    }
-}
 
 template <int mmq_x, int mmq_y>
 static __device__ __forceinline__ void vec_dot_q8_1_q8_1_dp4a(
@@ -3259,7 +3318,7 @@ struct mmq_type_traits<mmq_x, mmq_y, need_check, GGML_TYPE_MXFP4> {
     static constexpr int              vdr          = VDR_MXFP4_Q8_1_MMQ;
 #ifdef BLACKWELL_MMA_AVAILABLE
     static constexpr load_tiles_mmq_t load_tiles  = load_tiles_mxfp4_fp4<mmq_y, need_check>;
-    static constexpr vec_dot_mmq_t    vec_dot_mma = vec_dot_mxfp4_mxfp4_mma<mmq_x, mmq_y>;
+    static constexpr vec_dot_mmq_t    vec_dot_mma = vec_dot_fp4_fp4_mma<mmq_x, mmq_y, GGML_TYPE_MXFP4>;
 #else
     static constexpr load_tiles_mmq_t load_tiles   = load_tiles_mxfp4<mmq_y, need_check>;
     static constexpr vec_dot_mmq_t    vec_dot_mma  = vec_dot_q8_0_q8_1_mma<mmq_x, mmq_y, MMQ_Q8_1_DS_LAYOUT_D4>;
@@ -3270,8 +3329,13 @@ struct mmq_type_traits<mmq_x, mmq_y, need_check, GGML_TYPE_MXFP4> {
 template <int mmq_x, int mmq_y, bool need_check>
 struct mmq_type_traits<mmq_x, mmq_y, need_check, GGML_TYPE_NVFP4> {
     static constexpr int              vdr          = VDR_NVFP4_Q8_1_MMQ;
+#ifdef BLACKWELL_MMA_AVAILABLE
+    static constexpr load_tiles_mmq_t load_tiles   = load_tiles_nvfp4_nvfp4<mmq_y, need_check>;
+    static constexpr vec_dot_mmq_t    vec_dot_mma  = vec_dot_fp4_fp4_mma<mmq_x, mmq_y, GGML_TYPE_NVFP4>;
+#else
     static constexpr load_tiles_mmq_t load_tiles   = load_tiles_nvfp4<mmq_y, need_check>;
     static constexpr vec_dot_mmq_t    vec_dot_mma  = vec_dot_q8_0_16_q8_1_mma<mmq_x, mmq_y>;
+#endif // BLACKWELL_MMA_AVAILABLE
     static constexpr vec_dot_mmq_t    vec_dot_dp4a = vec_dot_q8_0_16_q8_1_dp4a<mmq_x, mmq_y>;
 };
 
@@ -3406,7 +3470,7 @@ static __device__ __forceinline__ void mul_mat_q_process_tile(
 
 #if defined(BLACKWELL_MMA_AVAILABLE)
     // FP4 tile stores 8 blocks
-    constexpr int ne_block = (type == GGML_TYPE_MXFP4) ? 8 * QK_MXFP4 : 4 * QK8_1;
+    constexpr int ne_block = (type == GGML_TYPE_MXFP4 || type == GGML_TYPE_NVFP4) ? QK_K : 4 * QK8_1;
 #else
     constexpr int ne_block = 4 * QK8_1;
 #endif  // defined(BLACKWELL_MMA_AVAILABLE)

ggml/src/ggml-cuda/mmvq.cu

@@ -115,6 +115,7 @@ static constexpr __host__ __device__ int get_mmvq_mmid_max_batch_pascal_older(gg
         case GGML_TYPE_IQ4_NL:  return 6;
         case GGML_TYPE_IQ4_XS:  return 5;
         case GGML_TYPE_MXFP4:   return 4;
+        case GGML_TYPE_NVFP4:   return 4;
         case GGML_TYPE_Q2_K:    return 4;
         case GGML_TYPE_Q3_K:    return 4;
         case GGML_TYPE_Q4_0:    return 6;
@@ -135,6 +136,7 @@ static constexpr __host__ __device__ int get_mmvq_mmid_max_batch_turing_plus(ggm
         case GGML_TYPE_IQ3_S:   return 6;
         case GGML_TYPE_IQ3_XXS: return 7;
         case GGML_TYPE_MXFP4:   return 7;
+        case GGML_TYPE_NVFP4:   return 8;
         case GGML_TYPE_Q2_K:    return 7;
         case GGML_TYPE_Q3_K:    return 5;
         default:                return MMVQ_MAX_BATCH_SIZE;
@@ -221,6 +223,7 @@ static constexpr __host__ __device__ int get_mmvq_mmid_max_batch_rdna4(ggml_type
         case GGML_TYPE_IQ4_NL:  return 7;
         case GGML_TYPE_IQ4_XS:  return 5;
         case GGML_TYPE_MXFP4:   return 5;
+        case GGML_TYPE_NVFP4:   return 5;
         case GGML_TYPE_Q3_K:    return 4;
         case GGML_TYPE_Q4_0:    return 7;
         case GGML_TYPE_Q4_1:    return 7;

ggml/src/ggml-cuda/quantize.cu

@@ -70,6 +70,102 @@ __device__ __forceinline__ uint8_t compute_e8m0_scale(float amax) {
     return static_cast<uint8_t>(biased);
 }
 
+
+static __global__ void quantize_mmq_nvfp4(
+        const float * __restrict__ x, const int32_t * __restrict__ ids, void * __restrict__ vy,
+        const int64_t ne00, const int64_t s01, const int64_t s02, const int64_t s03,
+        const int64_t ne0, const int64_t ne1, const int64_t ne2) {
+#if defined(BLACKWELL_MMA_AVAILABLE)
+
+    const int64_t i0_base = ((int64_t) blockDim.x * blockIdx.y + threadIdx.x) * QK_NVFP4_SUB;
+    if (i0_base >= ne0) {
+        return;
+    }
+
+    const int64_t i1 = blockIdx.x;
+    const int64_t i2 = blockIdx.z % ne2;
+    const int64_t i3 = blockIdx.z / ne2;
+    const int64_t i01 = ids ? ids[i1] : i1;
+    const int64_t k_block = i0_base / QK_K;
+    const int64_t blocks_per_col = (ne0 + QK_K - 1) / QK_K;
+    if (k_block >= blocks_per_col) {
+        return;
+    }
+
+    const int64_t ib = blockIdx.z * ((int64_t) blocks_per_col * ne1) + k_block * ne1 + blockIdx.x;
+    block_fp4_mmq * y = (block_fp4_mmq *) vy;
+    block_fp4_mmq * yb = y + ib;
+
+    const int sub = (i0_base % QK_K) / QK_NVFP4_SUB;
+
+    float vals_raw[QK_NVFP4_SUB];
+    float amax_raw = 0.0f;
+    const int64_t base_idx = i3 * s03 + i2 * s02 + i01 * s01;
+#pragma unroll
+    for (int k = 0; k < QK_NVFP4_SUB; k++) {
+        const int64_t i00 = i0_base + k;
+        if (i00 < ne00) {
+            const float v = x[base_idx + i00];
+            vals_raw[k] = v;
+            amax_raw = fmaxf(amax_raw, fabsf(v));
+        } else {
+            vals_raw[k] = 0.0f;
+        }
+    }
+
+    static constexpr int test_offsets[5] = { 0, -1, 1, -2, 2};
+    const int first_fp8_code = (int) ggml_cuda_fp32_to_ue4m3(amax_raw / 6.0f);
+
+    float best_err = FLT_MAX;
+    uint8_t fp8_code = 0;
+    float subblock_scale = 0.0f;
+
+#pragma unroll // Check +/- 2 to find best code to reduce NVFP4 activation loss. Negligible overhead on Blackwell.
+    for (int i = 0; i < 5; i++) {
+        const int test_code = first_fp8_code + test_offsets[i];
+        if (test_code < 0 || test_code > 0x7e) {
+            continue;
+        }
+        const uint8_t code = (uint8_t) test_code;
+        const float test_scale = ggml_cuda_ue4m3_to_fp32(code);
+        const float test_inv_scale = test_scale > 0.0f ? 0.5f / test_scale : 0.0f;
+        float cur_err = 0.0f;
+#pragma unroll
+        for (int k = 0; k < QK_NVFP4_SUB; ++k) {
+            const float v = vals_raw[k];
+            const uint8_t q = ggml_cuda_float_to_fp4_e2m1(v, test_inv_scale);
+            const float err_diff = fabsf(v) - fabsf(kvalues_mxfp4[q & 0x7]) * test_scale;
+            cur_err = fmaf(err_diff, err_diff, cur_err);
+        }
+
+        if (cur_err < best_err) {
+            best_err = cur_err;
+            fp8_code = test_code;
+            subblock_scale = test_scale;
+        }
+    }
+
+    const float inv_scale = subblock_scale > 0.0f ? 0.5f / subblock_scale : 0.0f;
+    uint32_t q0 = 0;
+    uint32_t q1 = 0;
+#pragma unroll // this is faster than the previous __nv_fp4x4_e2m1
+    for (int k = 0; k < QK_NVFP4_SUB / 4; ++k) {
+        q0 |= (uint32_t) ggml_cuda_float_to_fp4_e2m1(vals_raw[k +  0], inv_scale) << (8 * k);
+        q0 |= (uint32_t) ggml_cuda_float_to_fp4_e2m1(vals_raw[k +  8], inv_scale) << (8 * k + 4);
+        q1 |= (uint32_t) ggml_cuda_float_to_fp4_e2m1(vals_raw[k +  4], inv_scale) << (8 * k);
+        q1 |= (uint32_t) ggml_cuda_float_to_fp4_e2m1(vals_raw[k + 12], inv_scale) << (8 * k + 4);
+    }
+
+    uint32_t * yqs = reinterpret_cast<uint32_t *>(yb->qs);
+    yqs[2 * sub + 0] = q0;
+    yqs[2 * sub + 1] = q1;
+    reinterpret_cast<uint8_t *>(yb->d4)[sub] = fp8_code;
+#else
+    NO_DEVICE_CODE; // This is for Blackwell NVFP4 activations only.
+#endif // defined(BLACKWELL_MMA_AVAILABLE)
+
+}
+
 // quantize values in the format mxfp4 is stored which is interleaved nibbles
 // i.e. a block a0-a31 is represented as a0a16,a1a17 ...a15a31
 static __global__ void quantize_mmq_mxfp4(const float * __restrict__ x,
@@ -316,28 +412,32 @@ void quantize_mmq_q8_1_cuda(
     }
 }
 
-void quantize_mmq_mxfp4_cuda(const float *                    x,
-                             const int32_t *                  ids,
-                             void *                           vy,
-                             [[maybe_unused]] const ggml_type type_src0,
-                             const int64_t                    ne00,
-                             const int64_t                    s01,
-                             const int64_t                    s02,
-                             const int64_t                    s03,
-                             const int64_t                    ne0,
-                             const int64_t                    ne1,
-                             const int64_t                    ne2,
-                             const int64_t                    ne3,
-                             cudaStream_t                     stream) {
-    GGML_ASSERT(ne0 % (2 * QK_MXFP4) == 0);
+void quantize_mmq_fp4_cuda(
+        const float * x, const int32_t * ids, void * vy, const ggml_type type_src0,
+        const int64_t ne00, const int64_t s01, const int64_t s02, const int64_t s03,
+        const int64_t ne0, const int64_t ne1, const int64_t ne2, const int64_t ne3, cudaStream_t stream) {
+    GGML_ASSERT(type_src0 == GGML_TYPE_MXFP4 || type_src0 == GGML_TYPE_NVFP4);
+    GGML_ASSERT(ne0 > 0);
 
-    constexpr int nwarps = 8;
-    constexpr int vals_per_warp  = 2 * QK_MXFP4;
-    constexpr int vals_per_block = nwarps * vals_per_warp;
+    if (type_src0 == GGML_TYPE_NVFP4) {
+        GGML_ASSERT(ne00 % QK_NVFP4 == 0);
+        constexpr int nvfp4_block_size = 128;
+        const int64_t block_num_y = (ne0 + QK_NVFP4_SUB * nvfp4_block_size - 1) / (QK_NVFP4_SUB * nvfp4_block_size);
+        const dim3 block_size(nvfp4_block_size, 1, 1);
+        const dim3 num_blocks(ne1, block_num_y, ne2 * ne3);
+        quantize_mmq_nvfp4<<<num_blocks, block_size, 0, stream>>>(
+            x, ids, vy, ne00, s01, s02, s03, ne0, ne1, ne2);
+    } else {
+        GGML_ASSERT(ne0 % (2 * QK_MXFP4) == 0);
 
-    const int64_t block_num_y = (ne0 + vals_per_block - 1) / vals_per_block;
-    const dim3    num_blocks(ne1, block_num_y, ne2 * ne3);
-    const dim3    block_size(WARP_SIZE, nwarps, 1);
+        constexpr int nwarps = 8;
+        constexpr int vals_per_warp  = 2 * QK_MXFP4;
+        constexpr int vals_per_block = nwarps * vals_per_warp;
 
-    quantize_mmq_mxfp4<<<num_blocks, block_size, 0, stream>>>(x, ids, vy, ne00, s01, s02, s03, ne0, ne1, ne2);
+        const int64_t block_num_y = (ne0 + vals_per_block - 1) / vals_per_block;
+        const dim3    num_blocks(ne1, block_num_y, ne2 * ne3);
+        const dim3    block_size(WARP_SIZE, nwarps, 1);
+
+        quantize_mmq_mxfp4<<<num_blocks, block_size, 0, stream>>>(x, ids, vy, ne00, s01, s02, s03, ne0, ne1, ne2);
+    }
 }

ggml/src/ggml-cuda/quantize.cuh

@@ -26,7 +26,7 @@ void quantize_mmq_q8_1_cuda(
         ggml_type type_src0, int64_t ne00, int64_t s01, int64_t s02, int64_t s03,
         int64_t ne0, int64_t ne1, int64_t ne2, int64_t ne3, cudaStream_t stream);
 
-void quantize_mmq_mxfp4_cuda(const float *   x,
+void quantize_mmq_fp4_cuda(const float *   x,
                              const int32_t * ids,
                              void *          vy,
                              ggml_type       type_src0,

ggml/src/ggml-cuda/ssm-conv.cu

@@ -3,6 +3,7 @@
 
 template <bool apply_silu, size_t split_d_inner, size_t d_conv>
 static __global__ void ssm_conv_f32(const float * __restrict__ src0, const float * __restrict__ src1,
+                                    const float * __restrict__ bias,
                                     const int src0_nb0, const int src0_nb1, const int src0_nb2, const int src1_nb1,
                                     float * __restrict__ dst, const int dst_nb0, const int dst_nb1, const int dst_nb2,
                                     const int64_t n_t) {
@@ -27,6 +28,8 @@ static __global__ void ssm_conv_f32(const float * __restrict__ src0, const float
         w[j] = w_block[tid * stride_w + j];
     }
 
+    float b = bias != nullptr ? bias[bidy * split_d_inner + tid] : 0.0f;
+
     for (int64_t i = 0; i < n_t; i++) {
         float sumf = 0.0f;
 
@@ -42,12 +45,14 @@ static __global__ void ssm_conv_f32(const float * __restrict__ src0, const float
         for (size_t j = 0; j < d_conv; j++) {
             sumf += x[(i + j) % d_conv] * w[j];
         }
+        sumf += b;
         y_block[i * stride_y + tid] = apply_silu ? ggml_cuda_op_silu_single(sumf) : sumf;
     }
 }
 
 template <bool apply_silu, size_t split_d_inner, size_t d_conv, int64_t split_n_t>
 static __global__ void ssm_conv_long_token_f32(const float * __restrict__ src0, const float * __restrict__ src1,
+                                               const float * __restrict__ bias,
                                                const int src0_nb0, const int src0_nb1, const int src0_nb2,
                                                const int src1_nb1, float * __restrict__ dst, const int dst_nb0,
                                                const int dst_nb1, const int dst_nb2, const int64_t n_t) {
@@ -97,6 +102,8 @@ static __global__ void ssm_conv_long_token_f32(const float * __restrict__ src0,
         w[j] = w_block[tid * stride_w + j];
     }
 
+    float b = bias != nullptr ? bias[bidy * split_d_inner + tid] : 0.0f;
+
     // Compute from shared memory
     for (int64_t i = 0; i < local_n_t; i++) {
         float sumf = 0.0f;
@@ -104,12 +111,13 @@ static __global__ void ssm_conv_long_token_f32(const float * __restrict__ src0,
         for (size_t j = 0; j < d_conv; j++) {
             sumf += smem[tid * n_cols + i + j] * w[j];
         }
+        sumf += b;
         y_block[i * stride_y + tid] = apply_silu ? ggml_cuda_op_silu_single(sumf) : sumf;
     }
 }
 
 template <bool apply_silu>
-static void ssm_conv_f32_cuda(const float * src0, const float * src1, const int src0_nb0, const int src0_nb1,
+static void ssm_conv_f32_cuda(const float * src0, const float * src1, const float * bias, const int src0_nb0, const int src0_nb1,
                               const int src0_nb2, const int src1_nb1, float * dst, const int dst_nb0, const int dst_nb1,
                               const int dst_nb2, const int64_t nc, const int64_t nr, const int64_t n_t,
                               const int64_t n_s, cudaStream_t stream) {
@@ -120,14 +128,14 @@ static void ssm_conv_f32_cuda(const float * src0, const float * src1, const int
         constexpr int kNC = decltype(NC)::value;
         if (n_t <= 32) {
             const dim3 blocks(n_s, (nr + threads - 1) / threads, 1);
-            ssm_conv_f32<apply_silu, threads, kNC><<<blocks, threads, 0, stream>>>(src0, src1, src0_nb0, src0_nb1, src0_nb2, src1_nb1,
+            ssm_conv_f32<apply_silu, threads, kNC><<<blocks, threads, 0, stream>>>(src0, src1, bias, src0_nb0, src0_nb1, src0_nb2, src1_nb1,
                                                                        dst, dst_nb0, dst_nb1, dst_nb2, n_t);
         } else {
             const int64_t split_n_t = 32;
             dim3          blocks(n_s, (nr + threads - 1) / threads, (n_t + split_n_t - 1) / split_n_t);
             const size_t  smem_size = threads * (kNC - 1 + split_n_t) * sizeof(float);
             ssm_conv_long_token_f32<apply_silu, threads, kNC, split_n_t><<<blocks, threads, smem_size, stream>>>(
-                src0, src1, src0_nb0, src0_nb1, src0_nb2, src1_nb1, dst, dst_nb0, dst_nb1, dst_nb2, n_t);
+                src0, src1, bias, src0_nb0, src0_nb1, src0_nb2, src1_nb1, dst, dst_nb0, dst_nb1, dst_nb2, n_t);
         }
     };
 
@@ -140,11 +148,18 @@ static void ssm_conv_f32_cuda(const float * src0, const float * src1, const int
     }
 }
 
-void ggml_cuda_op_ssm_conv(ggml_backend_cuda_context & ctx, ggml_tensor * dst, ggml_tensor * silu_dst) {
+void ggml_cuda_op_ssm_conv(ggml_backend_cuda_context & ctx, ggml_tensor * dst, ggml_tensor * bias_add_node, ggml_tensor * silu_dst) {
     const struct ggml_tensor * src0 = dst->src[0];  // conv_x
     const struct ggml_tensor * src1 = dst->src[1];  // conv1d.weight
+    const bool fuse_bias = bias_add_node != nullptr;
     const bool fuse_silu = silu_dst != nullptr;
 
+    // bias always comes with silu.
+    GGML_ASSERT(!fuse_bias || fuse_silu);
+
+    // The bias (when fused) is the non-conv operand of the ADD node.
+    const struct ggml_tensor * bias = fuse_bias ? (bias_add_node->src[0] == dst ? bias_add_node->src[1] : bias_add_node->src[0]) : nullptr;
+
     // When fusing, write to silu_dst (the node downstream references).
     const struct ggml_tensor * out = fuse_silu ? silu_dst : dst;
 
@@ -160,16 +175,23 @@ void ggml_cuda_op_ssm_conv(ggml_backend_cuda_context & ctx, ggml_tensor * dst, g
 
     const float * src0_d = (const float *) src0->data;
     const float * src1_d = (const float *) src1->data;
+    const float * bias_d = fuse_bias ? (const float *) bias->data : nullptr;
     float *       dst_d  = (float *) out->data;
     cudaStream_t  stream = ctx.stream();
 
     GGML_ASSERT(src0->type == GGML_TYPE_F32);
     GGML_ASSERT(out->type == GGML_TYPE_F32);
+    if (fuse_bias) {
+        GGML_ASSERT(bias->type == GGML_TYPE_F32);
+        GGML_ASSERT(ggml_is_contiguous(bias));
+        GGML_ASSERT(ggml_nelements(bias) == nr);
+    }
+
     if (fuse_silu) {
-        ssm_conv_f32_cuda<true>(src0_d, src1_d, src0->nb[0], src0->nb[1], src0->nb[2], src1->nb[1], dst_d, out->nb[0], out->nb[1],
+        ssm_conv_f32_cuda<true>(src0_d, src1_d, bias_d, src0->nb[0], src0->nb[1], src0->nb[2], src1->nb[1], dst_d, out->nb[0], out->nb[1],
                           out->nb[2], nc, nr, n_t, n_s, stream);
     } else {
-        ssm_conv_f32_cuda<false>(src0_d, src1_d, src0->nb[0], src0->nb[1], src0->nb[2], src1->nb[1], dst_d, out->nb[0], out->nb[1],
+        ssm_conv_f32_cuda<false>(src0_d, src1_d, bias_d, src0->nb[0], src0->nb[1], src0->nb[2], src1->nb[1], dst_d, out->nb[0], out->nb[1],
                           out->nb[2], nc, nr, n_t, n_s, stream);
     }
 }

ggml/src/ggml-cuda/ssm-conv.cuh

@@ -1,3 +1,3 @@
 #include "common.cuh"
 
-void ggml_cuda_op_ssm_conv(ggml_backend_cuda_context & ctx, ggml_tensor * dst, ggml_tensor * silu_dst = nullptr);
+void ggml_cuda_op_ssm_conv(ggml_backend_cuda_context & ctx, ggml_tensor * dst, ggml_tensor * bias_add_node = nullptr, ggml_tensor * silu_dst = nullptr);

ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_1-ncols2_32.cu

@@ -2,4 +2,5 @@
 
 #include "../fattn-mma-f16.cuh"
 
+DECL_FATTN_MMA_F16_CASE(320, 256, 1, 32);
 DECL_FATTN_MMA_F16_CASE(576, 512, 1, 32);

ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_2-ncols2_32.cu

@@ -2,4 +2,5 @@
 
 #include "../fattn-mma-f16.cuh"
 
+DECL_FATTN_MMA_F16_CASE(320, 256, 2, 32);
 DECL_FATTN_MMA_F16_CASE(576, 512, 2, 32);

ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq320-dv256.cu

@@ -0,0 +1,5 @@
+// This file has been autogenerated by generate_cu_files.py, do not edit manually.
+
+#include "../fattn-tile.cuh"
+
+DECL_FATTN_TILE_CASE(320, 256);

ggml/src/ggml-cuda/template-instances/generate_cu_files.py

@@ -3,7 +3,7 @@
 from glob import glob
 import os
 
-HEAD_SIZES_KQ = [40, 64, 72, 80, 96, 112, 128, 256, 512, 576]
+HEAD_SIZES_KQ = [40, 64, 72, 80, 96, 112, 128, 256, 320, 512, 576]
 
 TYPES_KV = ["GGML_TYPE_F16", "GGML_TYPE_Q4_0", "GGML_TYPE_Q4_1", "GGML_TYPE_Q5_0", "GGML_TYPE_Q5_1", "GGML_TYPE_Q8_0", "GGML_TYPE_BF16"]
 
@@ -62,7 +62,7 @@ for filename in glob("*.cu"):
     os.remove(filename)
 
 for head_size_kq in HEAD_SIZES_KQ:
-    head_size_v = head_size_kq if head_size_kq != 576 else 512
+    head_size_v = 256 if head_size_kq == 320 else (head_size_kq if head_size_kq != 576 else 512)
     with open(f"fattn-tile-instance-dkq{head_size_kq}-dv{head_size_v}.cu", "w") as f:
         f.write(SOURCE_FATTN_TILE.format(head_size_kq=head_size_kq, head_size_v=head_size_v))
 
@@ -84,13 +84,16 @@ for ncols in [8, 16, 32, 64]:
                     continue
                 if head_size_kq == 72:
                     continue
-                if head_size_kq == 512 and ncols2 not in (4, 8):
+                # Skip compilation of unused ncols2 values for niche head sizes:
+                if head_size_kq == 320 and ncols2 != 32: # Mistral Small 4
                     continue
-                if head_size_kq != 576 and ncols2 in (16, 32):
+                if head_size_kq == 512 and ncols2 not in (4, 8): # Gemma 4
                     continue
-                if head_size_kq == 576 and ncols2 not in (4, 16, 32):
+                if head_size_kq == 576 and ncols2 not in (4, 16, 32): # Deepseek, GLM 4.7 Flash
                     continue
-                head_size_v = head_size_kq if head_size_kq != 576 else 512
+                if head_size_kq not in (320, 576) and ncols2 in (16, 32):
+                    continue
+                head_size_v = 256 if head_size_kq == 320 else (head_size_kq if head_size_kq != 576 else 512)
                 f.write(SOURCE_FATTN_MMA_CASE.format(ncols1=ncols1, ncols2=ncols2, head_size_kq=head_size_kq, head_size_v=head_size_v))
 
 for type in TYPES_MMQ:

ggml/src/ggml-hexagon/ggml-hexagon.cpp

@@ -48,14 +48,16 @@ using intvec  = std::vector<int>;
 using uintvec = std::vector<unsigned int>;
 using u32vec  = std::vector<uint32_t>;
 
-static size_t opt_ndev         = 1;
-static size_t opt_nhvx         = 0; // use all
-static int    opt_arch         = 0; // autodetect
-static int    opt_etm          = 0;
-static int    opt_verbose      = 0;
-static int    opt_profile      = 0; // profiling mode (0-disabled, 1-basic, 2-pmu)
-static int    opt_hostbuf      = 1; // hostbuf ON by default
-static int    opt_use_hmx      = 1; // when set, enable HMX; when 0, use HVX only
+static int    opt_arch    = 0; // autodetect
+static size_t opt_ndev    = 1;
+static size_t opt_nhvx    = 0; // use all
+static int    opt_use_hmx = 1; // when set, enable HMX; when 0, use HVX only
+static size_t opt_vmem    = HTP_OP_MAX_VMEM_DEFAULT;  // max available va space for buffer mappings
+static size_t opt_mbuf    = 1ul * 1024 * 1024 * 1024; // max buffer size
+static int    opt_etm     = 0;
+static int    opt_verbose = 0;
+static int    opt_profile = 0; // profiling mode (0-disabled, 1-basic, 2-pmu)
+static int    opt_hostbuf = 1; // hostbuf ON by default
 
 // Default PMU events, if profiling with PMU (mode=2) is enabled
 // See https://docs.qualcomm.com/doc/80-N2040-60/topic/pmu-events.html
@@ -66,6 +68,7 @@ static u32vec opt_pmu_evt { 0x3, 0x111, 0x100, 0x105, 0x240, 0x256, 0x7D, 0x8C }
 static int opt_opstage  = HTP_OPSTAGE_QUEUE | HTP_OPSTAGE_COMPUTE;
 static int opt_opbatch  = 1024; // max number of ops in a batch
 static int opt_opqueue  = 16;   // max number of pending batches
+
 static std::regex* opt_opfilter = NULL; // regex of ops to not claim
 
 #define HEX_VERBOSE(...) \
@@ -110,7 +113,7 @@ static void ggml_hexagon_dump_op_supp(const std::string &sess_name, const struct
     if (!opt_verbose) return;
 
     op_desc desc(op);
-    GGML_LOG_DEBUG("ggml-hex: %s supports-op %s : %s : %s : %s : %s : %s : %s\n", sess_name.c_str(),
+    GGML_LOG_DEBUG("ggml-hex: %s supports-op %s: %s : %s : %s : %s : %s : %s\n", sess_name.c_str(),
                 ggml_op_desc(op), desc.names, desc.dims, desc.types, desc.strides, desc.buffs, supp ? "yes" : "no");
 }
 
@@ -118,8 +121,6 @@ static void ggml_hexagon_dump_op_prof(const std::string &sess_name, const ggml_t
                                       uint32_t op_usec, uint32_t op_cycles, const uint32_t pmu[]) {
     if (!opt_profile) return;
 
-    op_desc desc(op);
-
     char pmu_str[256] = "";
     if (opt_profile > 1) {
         static_assert(HTP_PROF_PMU_NCNT == 8, "current implementation assumes 8 PMU counters");
@@ -127,6 +128,7 @@ static void ggml_hexagon_dump_op_prof(const std::string &sess_name, const ggml_t
                 pmu[0], pmu[1], pmu[2], pmu[3], pmu[4], pmu[5], pmu[6], pmu[7]);
     }
 
+    op_desc desc(op);
     GGML_LOG_DEBUG("ggml-hex: %s profile-op %s: %s : %s : %s : %s : usec %u cycles %u%s\n", sess_name.c_str(),
             ggml_op_desc(op), desc.names, desc.dims, desc.types, desc.strides, op_usec, op_cycles, pmu_str);
 }
@@ -191,33 +193,30 @@ struct ggml_hexagon_shared_buffer {
     bool                   mapped;
     bool                   pinned;
 
-    void mmap(bool pinned = false) {
-        int err = fastrpc_mmap(sess->domain_id, this->fd, (void *) this->base, 0, this->size, FASTRPC_MAP_FD_DELAYED);
+    void mmap() {
+        fastrpc_map_flags flags = this->pinned ? FASTRPC_MAP_FD : FASTRPC_MAP_FD_DELAYED;
+
+        int err = fastrpc_mmap(sess->domain_id, this->fd, (void *) this->base, 0, this->size, flags);
         if (err != 0) {
             GGML_LOG_ERROR("ggml-hex: %s buffer mapping failed : domain_id %d size %zu fd %d error 0x%08x\n", sess->c_name(),
                     sess->domain_id, this->size, this->fd, (unsigned) err);
             throw std::runtime_error("ggml-hex: fastrpc_mmap failed (see log for details)");
         }
 
-        if (pinned) {
-            err = htp_iface_mmap(sess->handle, this->fd, this->size, pinned);
-            if (err != 0) {
-                GGML_LOG_ERROR("ggml-hex: %s buffer pinning failed : domain_id %d size %zu fd %d error 0x%08x\n", sess->c_name(),
-                        sess->domain_id, this->size, this->fd, (unsigned) err);
-                throw std::runtime_error("ggml-hex: htp_iface_mmap failed (see log for details)");
-            }
-        }
-
-        this->mapped = true;
-        this->pinned = pinned;
         HEX_VERBOSE("ggml-hex: %s mapped buffer: base %p size %zu fd %d pinned %u\n",
                 sess->c_name(), (void *) this->base, this->size, this->fd, pinned);
+
+        this->mapped = true;
     }
 
     void unmap() {
         if (!this->mapped) return;
 
-        htp_iface_munmap(sess->handle, this->fd);
+        if (!this->pinned) {
+            // HTP might still hold a reference, tell it drop it
+            htp_iface_munmap(sess->handle, this->fd);
+        }
+
         fastrpc_munmap(sess->domain_id, this->fd, (void *) this->base, this->size);
 
         HEX_VERBOSE("ggml-hex: %s unmapped buffer: base %p size %zu fd %d\n", sess->c_name(),
@@ -227,7 +226,7 @@ struct ggml_hexagon_shared_buffer {
         this->fd     = -1;
     }
 
-    void alloc(size_t size, bool pinned = false) {
+    void alloc(size_t size) {
         if (this->base) return;
 
         this->base = (uint8_t *) rpcmem_alloc2(RPCMEM_HEAP_ID_SYSTEM, RPCMEM_DEFAULT_FLAGS, size);
@@ -245,8 +244,7 @@ struct ggml_hexagon_shared_buffer {
 
         HEX_VERBOSE("ggml-hex: %s allocated buffer: base %p size %zu fd %d pinned %d\n", sess->c_name(),
                     (void *) this->base, this->size, this->fd, (int) pinned);
-
-        mmap(pinned);
+        mmap();
     }
 
     void free() {
@@ -262,15 +260,14 @@ struct ggml_hexagon_shared_buffer {
     }
 
     ggml_hexagon_shared_buffer(ggml_hexagon_session * sess, size_t size, bool pinned = false) {
-        size += 4 * 1024;  // extra page for padding
-
         this->sess   = sess;
         this->size   = 0;
         this->base   = nullptr;
         this->fd     = -1;
         this->mapped = false;
+        this->pinned = pinned;
 
-        alloc(size, pinned);
+        alloc(size);
     }
 
     ~ggml_hexagon_shared_buffer() {
@@ -1475,6 +1472,7 @@ static ggml_backend_buffer_t ggml_backend_hexagon_buffer_type_alloc_buffer(
             ggml_backend_buffer_type_t buffer_type, size_t size) {
     auto sess = static_cast<ggml_backend_hexagon_buffer_type_context *>(buffer_type->context)->sess;
     try {
+        size += 4 * 1024;  // guard page
         ggml_hexagon_shared_buffer * sbuf = new ggml_hexagon_shared_buffer(sess, size);
         return ggml_backend_buffer_init(buffer_type, ggml_backend_hexagon_buffer_interface, sbuf, size);
     } catch (const std::exception & exc) {
@@ -1487,6 +1485,7 @@ static ggml_backend_buffer_t ggml_backend_hexagon_repack_buffer_type_alloc_buffe
             ggml_backend_buffer_type_t buffer_type, size_t size) {
     auto sess = static_cast<ggml_backend_hexagon_buffer_type_context *>(buffer_type->context)->sess;
     try {
+        size += 4 * 1024;  // guard page
         ggml_hexagon_shared_buffer * sbuf = new ggml_hexagon_shared_buffer(sess, size);
         return ggml_backend_buffer_init(buffer_type, ggml_backend_hexagon_buffer_interface, sbuf, size);
     } catch (const std::exception & exc) {
@@ -1505,7 +1504,7 @@ static size_t ggml_backend_hexagon_buffer_type_get_alloc_size(ggml_backend_buffe
 }
 
 static size_t ggml_backend_hexagon_buffer_type_get_max_size(ggml_backend_buffer_type_t buffer_type) {
-    return 1UL * 1024 * 1024 * 1024;  // 1GB per buffer
+    return opt_mbuf; // typically 1GB per buffer
     GGML_UNUSED(buffer_type);
 }
 
@@ -1573,14 +1572,14 @@ struct ggml_hexagon_opbatch {
         d_map.clear();
     }
 
-    ggml_hexagon_opbatch(ggml_hexagon_session *sess, size_t batch_size) {
+    ggml_hexagon_opbatch(ggml_hexagon_session *sess, size_t batch_size, size_t max_vmem) {
         this->sess = sess;
 
         n_bufs_max = HTP_OP_MAX_BUFS;
         n_ops_max  = batch_size;
         n_tens_max = n_ops_max + n_ops_max * HTP_OP_MAX_INPUTS;
 
-        b_vmem_max = HTP_OP_MAX_VMEM;
+        b_vmem_max = max_vmem;
 
         ops.resize(n_ops_max);
 
@@ -1592,6 +1591,9 @@ struct ggml_hexagon_opbatch {
         t_map.reserve(n_tens_max);
         d_map.reserve(n_tens_max);
 
+        GGML_LOG_INFO("ggml-hex: %s op batching: n-bufs %u n-tensors %u n-ops %u vmem %zu\n",
+                sess->c_name(), n_bufs_max, n_tens_max, n_ops_max, b_vmem_max);
+
         reset();
     }
 
@@ -1925,6 +1927,8 @@ void ggml_hexagon_session::flush_batch() {
     // Bump pending flag (cleared in the session::flush once we get the response)
     this->op_pending++;  // atomic inc
 
+    HEX_VERBOSE("ggml-hex: %s queue-opbatch: %p size %u\n", this->c_name(), dbuf.ptr, dbuf.size);
+
     int err = dspqueue_write(this->queue, 0, 1, &dbuf, sizeof(req), (const uint8_t*) &req, DSPQUEUE_TIMEOUT);
     if (err != 0) {
         GGML_ABORT("ggml-hex: %s dspqueue_write failed: 0x%08x\n", this->c_name(), (unsigned) err);
@@ -1944,6 +1948,35 @@ void ggml_hexagon_session::flush(bool all) {
     flush_pending(all);
 }
 
+static size_t ggml_hexagon_measure_max_vmem(ggml_hexagon_session *sess) {
+    // Allocate a bunch pinned buffers till failure.
+    // This is kind of expensive but handy for figuring out exactly how much we can mmap on a specific device.
+    // Typically we're going to allocate all/most of these buffers anyway for the model weights.
+
+    std::vector<ggml_hexagon_shared_buffer *> sbufs;
+
+    const size_t MiB = 1024 * 1024;
+    const size_t GiB = MiB  * 1024;
+
+    size_t vmem = 0;
+    size_t step = 256u * MiB;
+
+    try {
+        sbufs.push_back(new ggml_hexagon_shared_buffer(sess, GiB, true)); vmem += GiB;
+        sbufs.push_back(new ggml_hexagon_shared_buffer(sess, GiB, true)); vmem += GiB;
+        sbufs.push_back(new ggml_hexagon_shared_buffer(sess, GiB, true)); vmem += GiB;
+
+        while (1) {
+            sbufs.push_back(new ggml_hexagon_shared_buffer(sess, step, true));
+            vmem += step;
+        }
+    } catch (...) { }
+
+    for (auto b : sbufs) { delete b; }
+
+    return vmem - step; // backoff to account for overhead from internal mappings
+}
+
 void ggml_hexagon_session::allocate(int dev_id) noexcept(false) {
     this->valid_session = false;
     this->valid_handle  = false;
@@ -1957,7 +1990,7 @@ void ggml_hexagon_session::allocate(int dev_id) noexcept(false) {
 
     this->op_pending  = 0;
 
-    GGML_LOG_INFO("ggml-hex: allocating new session: %s\n", this->name.c_str());
+    GGML_LOG_DEBUG("ggml-hex: %s allocating new session\n", this->name.c_str());
 
     domain * my_domain = get_domain(this->domain_id);
     if (my_domain == NULL) {
@@ -2033,9 +2066,6 @@ void ggml_hexagon_session::allocate(int dev_id) noexcept(false) {
 
     this->valid_handle = true;
 
-    GGML_LOG_INFO("ggml-hex: new session: %s : session-id %d domain-id %d uri %s handle 0x%lx\n", this->name.c_str(),
-                  this->session_id, this->domain_id, session_uri, (unsigned long) this->handle);
-
     // Enable FastRPC QoS mode
     {
         struct remote_rpc_control_latency l;
@@ -2047,6 +2077,9 @@ void ggml_hexagon_session::allocate(int dev_id) noexcept(false) {
         }
     }
 
+    GGML_LOG_INFO("ggml-hex: %s new session : session-id %d domain-id %d uri %s handle 0x%lx\n", this->c_name(),
+                  this->session_id, this->domain_id, session_uri, (unsigned long) this->handle);
+
     const size_t req_q_size = (sizeof(htp_opbatch_req) * opt_opqueue * 2) + 1024;
     const size_t rsp_q_size = (sizeof(htp_opbatch_rsp) * opt_opqueue * 2) + 1024;
 
@@ -2091,13 +2124,19 @@ void ggml_hexagon_session::allocate(int dev_id) noexcept(false) {
     }
 
     // Allocate buffers and state for op batching
-    this->op_batch = new ggml_hexagon_opbatch(this, opt_opbatch);
     this->op_queue = new ggml_hexagon_opqueue(this, opt_opbatch, opt_opqueue);
 
-    // Start processing op batch requests
-    err = htp_iface_start(this->handle, dev_id, this->queue_id, opt_nhvx, opt_use_hmx);
+    if (!opt_vmem) {
+        opt_vmem = ggml_hexagon_measure_max_vmem(this);
+        GGML_LOG_INFO("ggml-hex: %s measured max vmem %zu\n", this->c_name(), opt_vmem);
+    }
+
+    this->op_batch = new ggml_hexagon_opbatch(this, opt_opbatch, opt_vmem);
+
+    // Start dspqueue/opbatch processing
+    err = htp_iface_start(this->handle, dev_id, this->queue_id, opt_nhvx, opt_use_hmx, opt_vmem);
     if (err != 0) {
-        GGML_LOG_ERROR("ggml-hex: failed to start session: 0x%08x\n", (unsigned) err);
+        GGML_LOG_ERROR("ggml-hex: %s failed to start session: 0x%08x\n", this->c_name(), (unsigned) err);
         throw std::runtime_error("ggml-hex: iface start failed (see log for details)");
     }
     this->valid_iface = true;
@@ -2108,17 +2147,17 @@ void ggml_hexagon_session::release() noexcept(true) {
 
     int err;
 
-    delete this->op_batch;
-    delete this->op_queue;
-
-    // Stop the DSP-side service and close the queue
     if (this->valid_iface) {
+        // Stop dspqueue/opbatch processing
         err = htp_iface_stop(this->handle);
         if (err != 0) {
             GGML_ABORT("ggml-hex: htp_iface_stop failed: 0x%08x\n", (unsigned) err);
         }
     }
 
+    delete this->op_batch;
+    delete this->op_queue;
+
     if (opt_etm) {
         err = htp_iface_etm(this->handle, 0);
         if (err != 0) {
@@ -2997,8 +3036,8 @@ static struct ggml_backend_i hexagon_backend_i = {
     /* .free                    = */ ggml_backend_hexagon_free,
     /* .set_tensor_async        = */ NULL,
     /* .get_tensor_async        = */ NULL,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ NULL,
     /* .synchronize             = */ ggml_backend_hexagon_synchronize,
     /* .graph_plan_create       = */ NULL,
@@ -3380,21 +3419,6 @@ struct ggml_hexagon_registry {
 ggml_hexagon_registry::ggml_hexagon_registry(ggml_backend_reg_t reg) {
     GGML_LOG_INFO("ggml-hex: Hexagon backend (experimental) : allocating new registry : ndev %zu\n", opt_ndev);
 
-    if (!opt_arch) {
-        int err = get_hex_arch_ver(CDSP_DOMAIN_ID, &opt_arch);
-        if (err != 0) {
-            GGML_LOG_ERROR("ggml-hex: failed to query HTP version (err %d) defaulting to v73\n", err);
-            opt_arch = 73;
-        }
-    }
-
-#if defined(__ANDROID__)
-    if (opt_arch < 75) {
-        opt_ndev = 1;
-        GGML_LOG_WARN("ggml-hex: forcing ndev to 1 for SoCs archs lower than v75.\n");
-    }
-#endif
-
     GGML_LOG_INFO("ggml-hex: Hexagon Arch version v%d\n", opt_arch);
 
     // Create devices / sessions
@@ -3480,32 +3504,67 @@ static void ggml_hexagon_init(ggml_backend_reg * reg) {
     static_assert((unsigned int) HTP_TYPE_IQ4_NL == (unsigned int) GGML_TYPE_IQ4_NL,
                   "please update hexagon_type to match ggml_type");
 
-    const char * str_verbose = getenv("GGML_HEXAGON_VERBOSE");
-    const char * str_hostbuf = getenv("GGML_HEXAGON_HOSTBUF");
-    const char * str_opstage = getenv("GGML_HEXAGON_OPSTAGE");
-    const char * str_opbatch = getenv("GGML_HEXAGON_OPBATCH");
-    const char * str_opqueue = getenv("GGML_HEXAGON_OPQUEUE");
-    const char * str_opfilter= getenv("GGML_HEXAGON_OPFILTER");
-    const char * str_profile = getenv("GGML_HEXAGON_PROFILE");
-    const char * str_etm     = getenv("GGML_HEXAGON_ETM");
-    const char * str_nhvx    = getenv("GGML_HEXAGON_NHVX");
-    const char * str_use_hmx = getenv("GGML_HEXAGON_USE_HMX");
-    const char * str_ndev    = getenv("GGML_HEXAGON_NDEV");
-    const char * str_arch    = getenv("GGML_HEXAGON_ARCH");
+    const char * str_verbose  = getenv("GGML_HEXAGON_VERBOSE");
+    const char * str_hostbuf  = getenv("GGML_HEXAGON_HOSTBUF");
+    const char * str_opstage  = getenv("GGML_HEXAGON_OPSTAGE");
+    const char * str_opbatch  = getenv("GGML_HEXAGON_OPBATCH");
+    const char * str_opqueue  = getenv("GGML_HEXAGON_OPQUEUE");
+    const char * str_opfilter = getenv("GGML_HEXAGON_OPFILTER");
+    const char * str_profile  = getenv("GGML_HEXAGON_PROFILE");
+    const char * str_etm      = getenv("GGML_HEXAGON_ETM");
+    const char * str_nhvx     = getenv("GGML_HEXAGON_NHVX");
+    const char * str_use_hmx  = getenv("GGML_HEXAGON_USE_HMX");
+    const char * str_ndev     = getenv("GGML_HEXAGON_NDEV");
+    const char * str_arch     = getenv("GGML_HEXAGON_ARCH");
+    const char * str_vmem     = getenv("GGML_HEXAGON_VMEM");
+    const char * str_mbuf     = getenv("GGML_HEXAGON_MBUF");
+
+    // Init Arch first since it affects other defaults
+    if (!str_arch) {
+        int err = get_hex_arch_ver(CDSP_DOMAIN_ID, &opt_arch);
+        if (err != 0) {
+            GGML_LOG_ERROR("ggml-hex: failed to query HTP version (err %d) defaulting to v73\n", err);
+            opt_arch = 73;
+        }
+    } else {
+        if (str_arch[0] == 'v' || str_arch[0] == 'V') {
+            str_arch++;
+        }
+        opt_arch = strtoul(str_arch, NULL, 0);
+    }
+
+    size_t MiB = 1024 * 1024;
+
+    // Update vmem default
+    opt_vmem = opt_arch >= 75 ? HTP_OP_MAX_VMEM_DEFAULT : 3000 * MiB;
 
     auto RE_ICASE = std::regex_constants::icase;
 
-    opt_opfilter     = str_opfilter ? new std::regex(str_opfilter, RE_ICASE) : NULL;
-    opt_verbose      = str_verbose  ? atoi(str_verbose)             : 0;
-    opt_hostbuf      = str_hostbuf  ? atoi(str_hostbuf)             : opt_hostbuf;
-    opt_opstage      = str_opstage  ? strtoul(str_opstage, NULL, 0) : opt_opstage;
-    opt_opbatch      = str_opbatch  ? strtoul(str_opbatch, NULL, 0) : opt_opbatch;
-    opt_opqueue      = str_opqueue  ? strtoul(str_opqueue, NULL, 0) : opt_opqueue;
-    opt_etm          = str_etm      ? atoi(str_etm)                 : 0;
-    opt_nhvx         = str_nhvx     ? strtoul(str_nhvx, NULL, 0)    : opt_nhvx;
-    opt_use_hmx      = str_use_hmx  ? atoi(str_use_hmx)             : opt_use_hmx;
-    opt_ndev         = str_ndev     ? strtoul(str_ndev, NULL, 0)    : opt_ndev;
-    opt_hostbuf      = str_hostbuf  ? atoi(str_hostbuf)             : opt_hostbuf;
+    opt_opfilter  = str_opfilter ? new std::regex(str_opfilter, RE_ICASE) : NULL;
+    opt_verbose   = str_verbose  ? atoi(str_verbose)                      : 0;
+    opt_hostbuf   = str_hostbuf  ? atoi(str_hostbuf)                      : opt_hostbuf;
+    opt_opstage   = str_opstage  ? strtoul(str_opstage, NULL, 0)          : opt_opstage;
+    opt_opbatch   = str_opbatch  ? strtoul(str_opbatch, NULL, 0)          : opt_opbatch;
+    opt_opqueue   = str_opqueue  ? strtoul(str_opqueue, NULL, 0)          : opt_opqueue;
+    opt_profile   = str_profile  ? atoi(str_profile)                      : 0;
+    opt_etm       = str_etm      ? atoi(str_etm)                          : 0;
+    opt_nhvx      = str_nhvx     ? strtoul(str_nhvx, NULL, 0)             : opt_nhvx;
+    opt_use_hmx   = str_use_hmx  ? atoi(str_use_hmx)                      : opt_use_hmx;
+    opt_ndev      = str_ndev     ? strtoul(str_ndev, NULL, 0)             : opt_ndev;
+    opt_hostbuf   = str_hostbuf  ? atoi(str_hostbuf)                      : opt_hostbuf;
+    opt_mbuf      = str_mbuf     ? strtoul(str_mbuf, NULL, 0) * MiB       : opt_mbuf;
+    opt_vmem      = str_vmem     ? strtoul(str_vmem, NULL, 0) * MiB       : opt_vmem;
+
+    if (opt_ndev > GGML_HEXAGON_MAX_SESSIONS) {
+        opt_ndev = GGML_HEXAGON_MAX_SESSIONS;
+    }
+
+#if defined(__ANDROID__)
+    if (opt_arch < 75) {
+        opt_ndev = 1;
+        GGML_LOG_WARN("ggml-hex: forcing ndev to 1 for SoCs archs lower than v75.\n");
+    }
+#endif
 
     if (str_profile) {
         opt_pmu_evt = [&]() -> std::vector<uint32_t> {
@@ -3520,17 +3579,6 @@ static void ggml_hexagon_init(ggml_backend_reg * reg) {
                 vec_to_str<uint32_t, 16>(opt_pmu_evt).c_str());
     }
 
-    if (opt_ndev > GGML_HEXAGON_MAX_SESSIONS) {
-        opt_ndev = GGML_HEXAGON_MAX_SESSIONS;
-    }
-
-    if (str_arch) {
-        if (str_arch[0] == 'v') {
-            str_arch++;
-        }
-        opt_arch = strtoul(str_arch, NULL, 0);
-    }
-
     reg->context = new ggml_hexagon_registry(reg);
 }
 

ggml/src/ggml-hexagon/htp/htp-ctx.h

@@ -20,7 +20,7 @@ struct htp_mmap {
     uint64_t size;
     uint64_t base;
     uint32_t fd;
-    uint32_t pinned;
+    uint32_t reserved;
 };
 
 // Scratchpad state
@@ -77,6 +77,8 @@ struct htp_context {
     atomic_bool            vtcm_valid;
     atomic_bool            vtcm_needs_release;
 
+    uint64_t               max_vmem;
+
     struct htp_ops_context octx;
 
 #ifdef HTP_HAS_HMX

ggml/src/ggml-hexagon/htp/htp-ops.h

@@ -90,15 +90,11 @@ enum htp_op_code {
 #define HTP_OP_MAX_INPUTS  6    // aka GGML_MAX_SRCS
 #define HTP_OP_MAX_PARAMS  16   // aka GGML_MAX_OP_PARAMS
 
-#define HTP_OP_MAX_BUFS    8
+#define HTP_OP_MAX_BUFS    16
 #define HTP_OP_MAX_REQS    256
 #define HTP_OP_MAX_TENSORS (HTP_OP_MAX_REQS * HTP_OP_MAX_INPUTS + HTP_OP_MAX_REQS)
 
-#if __HVX_ARCH__ < 75
-#define HTP_OP_MAX_VMEM    (3167538380u)
-#else
-#define HTP_OP_MAX_VMEM    (3221225472u)
-#endif
+#define HTP_OP_MAX_VMEM_DEFAULT (3355443200u)
 
 #define HTP_MMAP_MAX_VMEM  (2147483648u)
 

ggml/src/ggml-hexagon/htp/htp_iface.idl

@@ -11,9 +11,9 @@ struct htp_iface_pmu_conf {
 };
 
 interface htp_iface : remote_handle64 {
-    AEEResult start(in uint32 sess_id, in uint64 dsp_queue_id, in uint32 n_hvx, in uint32 use_hmx);
+    AEEResult start(in uint32 sess_id, in uint64 dsp_queue_id, in uint32 n_hvx, in uint32 use_hmx, in uint64 max_vmem);
     AEEResult stop();
-    AEEResult mmap(in uint32 fd, in uint32 size, in uint32 pinned);
+    AEEResult mmap(in uint32 fd, in uint32 size);
     AEEResult munmap(in uint32 fd);
     AEEResult profiler(in uint32 mode, in htp_iface_pmu_conf pmu);
     AEEResult etm(in uint32 enable);

ggml/src/ggml-hexagon/htp/main.c

@@ -210,7 +210,7 @@ AEEResult htp_iface_close(remote_handle64 handle) {
     return AEE_SUCCESS;
 }
 
-AEEResult htp_iface_mmap(remote_handle64 handle, uint32 fd, uint32 size, uint32 pinned) {
+AEEResult htp_iface_mmap(remote_handle64 handle, uint32_t fd, uint32_t size) {
     struct htp_context * ctx = (struct htp_context *) handle;
     if (!ctx) {
         return AEE_EBADPARM;
@@ -220,7 +220,6 @@ AEEResult htp_iface_mmap(remote_handle64 handle, uint32 fd, uint32 size, uint32
     for (uint32_t i=0; i<HTP_MAX_MMAPS; i++) {
         struct htp_mmap *m = &ctx->mmap[i];
         if (m->fd == fd) {
-            m->pinned = pinned;
             return AEE_SUCCESS;
         }
     }
@@ -229,7 +228,7 @@ AEEResult htp_iface_mmap(remote_handle64 handle, uint32 fd, uint32 size, uint32
     for (uint32_t i=0; i<HTP_MAX_MMAPS; i++) {
         struct htp_mmap *m = &ctx->mmap[i];
         if (!m->size) {
-            FARF(HIGH, "mmap : fd %u size %u pinned %u", fd, size, pinned);
+            FARF(HIGH, "mmap : fd %u size %u", fd, size);
 #if __HVX_ARCH__ > 73
             void *va = HAP_mmap2(NULL, size, HAP_PROT_READ | HAP_PROT_WRITE, 0, fd, 0);
 #else
@@ -248,7 +247,6 @@ AEEResult htp_iface_mmap(remote_handle64 handle, uint32 fd, uint32 size, uint32
             m->base   = (uint64_t) va;
             m->fd     = fd;
             m->size   = size;
-            m->pinned = pinned;
 
             return AEE_SUCCESS;
         }
@@ -275,7 +273,6 @@ AEEResult htp_iface_munmap(remote_handle64 handle, uint32 fd) {
             m->size   = 0;
             m->base   = NULL;
             m->fd     = -1;
-            m->pinned = 0;
         }
     }
 
@@ -358,7 +355,7 @@ static void vtcm_free(struct htp_context * ctx) {
 static void htp_packet_callback(dspqueue_t queue, int error, void * context);
 static void htp_error_callback(dspqueue_t queue, int error, void * context);
 
-AEEResult htp_iface_start(remote_handle64 handle, uint32 sess_id, uint64 dsp_queue_id, uint32 n_hvx, uint32 use_hmx) {
+AEEResult htp_iface_start(remote_handle64 handle, uint32 sess_id, uint64 dsp_queue_id, uint32 n_hvx, uint32 use_hmx, uint64_t max_vmem) {
     struct htp_context * ctx = (struct htp_context *) handle;
 
     if (!ctx) {
@@ -376,12 +373,12 @@ AEEResult htp_iface_start(remote_handle64 handle, uint32 sess_id, uint64 dsp_que
                               htp_error_callback,   // Error callback; no errors expected on the DSP
                               (void *) ctx,         // Callback context
                               &ctx->queue);
-
     if (err) {
         FARF(ERROR, "Queue import failed with 0x%08x", (unsigned) err);
         return err;
     }
 
+    ctx->max_vmem    = max_vmem;
     ctx->thread_id   = qurt_thread_get_id();
     ctx->thread_prio = qurt_thread_get_priority(ctx->thread_id);
 
@@ -622,8 +619,8 @@ static inline bool reuse_buf(struct htp_context *ctx, uint32_t *m_reuse, struct
 }
 
 static inline void drop_mmap(struct htp_context *ctx, struct htp_mmap *m) {
-    if (m->size && !m->pinned) {
-        FARF(HIGH, "unmap : fd %u base %p size %u pinned %u", m->fd, (void*) m->base, (uint32_t) m->size, m->pinned);
+    if (m->size) {
+        FARF(HIGH, "unmap : fd %u base %p size %u", m->fd, (void*) m->base, (uint32_t) m->size);
 #if __HVX_ARCH__ > 73
         HAP_munmap2((void *) m->base, m->size);
 #else
@@ -660,9 +657,8 @@ static inline void mmap_buf(struct htp_context *ctx, struct htp_buf_desc *b) {
             m->base   = b->base = (uint64_t) va;
             m->fd     = b->fd;
             m->size   = b->size;
-            m->pinned = 0;
 
-            FARF(HIGH, "mmap : fd %u base %p size %u pinned %u", m->fd, (void*) m->base, (uint32_t) m->size, m->pinned);
+            FARF(HIGH, "mmap : fd %u base %p size %u", m->fd, (void*) m->base, (uint32_t) m->size);
             return;
         }
     }
@@ -672,8 +668,8 @@ static void prep_op_bufs(struct htp_context *ctx, struct htp_buf_desc *bufs, uin
     uint32_t m_reuse = 0; // mmap reuse mask (index from ctx->mmap array)
     uint32_t b_reuse = 0; // buf reuse count
 
-    size_t   m_vmem  = 0; // mapped vmem
-    size_t   e_vmem  = 0; // extra  vmem
+    uint64_t m_vmem  = 0; // mapped vmem
+    uint64_t e_vmem  = 0; // extra  vmem
 
     // See what we can reuse
     for (uint32_t i=0; i < n_bufs; i++) {
@@ -687,9 +683,10 @@ static void prep_op_bufs(struct htp_context *ctx, struct htp_buf_desc *bufs, uin
     // See how much vmem we have mmaped right now
     for (uint32_t i=0; i<HTP_MAX_MMAPS; i++) { m_vmem += ctx->mmap[i].size; }
 
-    FARF(HIGH, "prep-bufs : pass1 mmap-vmem %zu extra-vmem %zu n-bufs %u b-reuse %u", m_vmem, e_vmem, n_bufs, b_reuse);
+    FARF(HIGH, "prep-bufs : pass1 mmap-vmem %zu extra-vmem %zu max-vmem %zu : n-bufs %u b-reuse %u",
+            (size_t) m_vmem, (size_t) e_vmem, (size_t) ctx->max_vmem, n_bufs, b_reuse);
 
-    if ((m_vmem + e_vmem) > HTP_OP_MAX_VMEM) {
+    if ((m_vmem + e_vmem) > ctx->max_vmem) {
         // Drop unused mappings
         for (uint32_t i=0; i < HTP_MAX_MMAPS; i++) {
             bool used = m_reuse & (1<<i);

ggml/src/ggml-metal/ggml-metal.cpp

@@ -166,8 +166,8 @@ static ggml_backend_buffer_i ggml_backend_metal_buffer_private_i = {
     /* .memset_tensor           = */ ggml_backend_metal_buffer_private_memset_tensor,
     /* .set_tensor              = */ ggml_backend_metal_buffer_private_set_tensor,
     /* .get_tensor              = */ ggml_backend_metal_buffer_private_get_tensor,
-    /* .get_tensor_2d_async     = */ NULL,
-    /* .set_tensor_2d_async     = */ NULL,
+    /* .set_tensor_2d           = */ NULL,
+    /* .get_tensor_2d           = */ NULL,
     /* .cpy_tensor              = */ ggml_backend_metal_buffer_private_cpy_tensor,
     /* .clear                   = */ ggml_backend_metal_buffer_private_clear,
     /* .reset                   = */ NULL,
@@ -567,8 +567,8 @@ static ggml_backend_i ggml_backend_metal_i = {
     /* .free                    = */ ggml_backend_metal_free,
     /* .set_tensor_async        = */ ggml_backend_metal_set_tensor_async,
     /* .get_tensor_async        = */ ggml_backend_metal_get_tensor_async,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ ggml_backend_metal_cpy_tensor_async, // only needed for multi-GPU setups
     /* .synchronize             = */ ggml_backend_metal_synchronize,
     /* .graph_plan_create       = */ NULL,

ggml/src/ggml-opencl/ggml-opencl.cpp

@@ -4343,9 +4343,9 @@ static ggml_backend_i ggml_backend_opencl_i = {
     /* .free                    = */ ggml_backend_opencl_free,
     /* .set_tensor_async        = */ NULL,  /* ggml_backend_opencl_set_tensor_async */
     /* .get_tensor_async        = */ NULL,  /* ggml_backend_opencl_get_tensor_async */
-    /* .cpy_tensor_async        = */ NULL,  /* ggml_backend_opencl_cpy_tensor_async */
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
+    /* .cpy_tensor_async        = */ NULL,  /* ggml_backend_opencl_cpy_tensor_async */
     /* .synchronize             = */ ggml_backend_opencl_synchronize,
     /* .graph_plan_create       = */ NULL,
     /* .graph_plan_free         = */ NULL,

ggml/src/ggml-rpc/ggml-rpc.cpp

@@ -740,9 +740,9 @@ static ggml_backend_i ggml_backend_rpc_interface = {
     /* .free                    = */ ggml_backend_rpc_free,
     /* .set_tensor_async        = */ NULL,
     /* .get_tensor_async        = */ NULL,
-    /* .cpy_tensor_async        = */ NULL,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
+    /* .cpy_tensor_async        = */ NULL,
     /* .synchronize             = */ ggml_backend_rpc_synchronize,
     /* .graph_plan_create       = */ NULL,
     /* .graph_plan_free         = */ NULL,
@@ -1101,7 +1101,7 @@ bool rpc_server::set_tensor(const std::vector<uint8_t> & input) {
         fs::path cache_file = fs::path(cache_dir) / hash_str;
         std::ofstream ofs(cache_file, std::ios::binary);
         ofs.write((const char *)data, size);
-        GGML_LOG_INFO("[%s] saved to '%s'\n", __func__, cache_file.c_str());
+        GGML_LOG_INFO("[%s] saved to '%s'\n", __func__, cache_file.string().c_str());
     }
     ggml_backend_tensor_set(tensor, data, offset, size);
     return true;

ggml/src/ggml-sycl/ggml-sycl.cpp

@@ -4700,8 +4700,8 @@ static ggml_backend_i ggml_backend_sycl_interface = {
     /* .free                    = */ ggml_backend_sycl_free,
     /* .set_tensor_async        = */ ggml_backend_sycl_set_tensor_async,
     /* .get_tensor_async        = */ ggml_backend_sycl_get_tensor_async,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ NULL, // ggml_backend_sycl_cpy_tensor_async,
                                            // // TODO: update for the new
                                            // interface

ggml/src/ggml-virtgpu/ggml-backend.cpp

@@ -34,8 +34,8 @@ static ggml_backend_i ggml_backend_remoting_interface = {
     /* .free                    = */ ggml_backend_remoting_free,
     /* .set_tensor_async        = */ NULL,  // ggml_backend_remoting_set_tensor_async,
     /* .get_tensor_async        = */ NULL,  // ggml_backend_remoting_get_tensor_async,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ NULL,  // ggml_backend_remoting_cpy_tensor_async,
     /* .synchronize             = */ NULL,  // ggml_backend_remoting_synchronize,
     /* .graph_plan_create       = */ NULL,

ggml/src/ggml-vulkan/ggml-vulkan.cpp

@@ -20,12 +20,19 @@ DispatchLoaderDynamic & ggml_vk_default_dispatcher();
 #define VULKAN_HPP_DEFAULT_DISPATCHER ggml_vk_default_dispatcher()
 
 #include <vulkan/vulkan.hpp>
-// SPIRV-Headers: LunarG Windows SDK uses Include/spirv-headers/spirv.hpp (not spirv/unified1/). MinGW/MSYS2 and
-// Linux packages use Khronos layout spirv/unified1/spirv.hpp. See docs/build.md#vulkan.
-#if defined(_WIN32) && !defined(__MINGW32__)
-#include <spirv-headers/spirv.hpp>
+
+// SPIR-V Headers: different SDK installations expose different include paths.
+// LunarG Vulkan SDK on Windows typically provides <spirv-headers/spirv.hpp>.
+// Linux packages, MSYS2 and MinGW often use the Khronos layout <spirv/unified1/spirv.hpp>.
+#if __has_include(<spirv/unified1/spirv.hpp>)
+#    include <spirv/unified1/spirv.hpp>
+#elif __has_include(<spirv-headers/spirv.hpp>)
+#    include <spirv-headers/spirv.hpp>
+#elif __has_include(<spirv.hpp>)
+#    include <spirv.hpp>
 #else
-#include <spirv/unified1/spirv.hpp>
+     // Fallback to let the compiler throw a standard "file not found" error
+#    include <spirv/unified1/spirv.hpp>
 #endif
 
 #include <algorithm>
@@ -6838,7 +6845,7 @@ static void ggml_vk_buffer_write_nc_async(ggml_backend_vk_context * ctx, vk_cont
     }
 }
 
-static bool ggml_vk_buffer_write_2d_async(vk_context subctx, vk_buffer& dst, size_t offset, const void * src, size_t spitch, size_t width, size_t height, bool sync_staging = false) {
+static bool ggml_vk_buffer_write_2d_async(vk_context subctx, vk_buffer& dst, size_t offset, const void * src, size_t spitch, size_t dpitch, size_t width, size_t height, bool sync_staging = false) {
     VK_LOG_DEBUG("ggml_vk_buffer_write_2d_async(" << width << ", " << height << ")");
     // Check if src is pinned memory
     vk_buffer buf = nullptr;
@@ -6848,7 +6855,7 @@ static bool ggml_vk_buffer_write_2d_async(vk_context subctx, vk_buffer& dst, siz
     if (buf != nullptr) {
         // Memory is pinned, use as staging buffer
         std::vector<vk::BufferCopy> slices(1);
-        if (width == spitch) {
+        if (width == spitch && width == dpitch) {
             // Only do single write if stride is equal
             slices[0].srcOffset = buf_offset;
             slices[0].dstOffset = offset;
@@ -6857,7 +6864,7 @@ static bool ggml_vk_buffer_write_2d_async(vk_context subctx, vk_buffer& dst, siz
             slices.resize(height);
             for (size_t i = 0; i < height; i++) {
                 slices[i].srcOffset = buf_offset + i * spitch;
-                slices[i].dstOffset = offset + i * width;
+                slices[i].dstOffset = offset + i * dpitch;
                 slices[i].size = width;
             }
         }
@@ -6874,21 +6881,30 @@ static bool ggml_vk_buffer_write_2d_async(vk_context subctx, vk_buffer& dst, siz
     }
 
     // Staging buffer required
-    const size_t copy_size = width*height;
-    ggml_vk_ensure_sync_staging_buffer(dst->device, copy_size);
+    const size_t staging_size = width * height;
+    ggml_vk_ensure_sync_staging_buffer(dst->device, staging_size);
 
     vk_buffer& staging_buffer = dst->device->sync_staging;
 
-    VkBufferCopy buf_copy = {
-        0,
-        offset,
-        copy_size};
+    std::vector<vk::BufferCopy> slices(1);
+    if (width == dpitch) {
+        slices[0].srcOffset = 0;
+        slices[0].dstOffset = offset;
+        slices[0].size = staging_size;
+    } else {
+        slices.resize(height);
+        for (size_t i = 0; i < height; i++) {
+            slices[i].srcOffset = i * width;
+            slices[i].dstOffset = offset + i * dpitch;
+            slices[i].size = width;
+        }
+    }
 
     ggml_vk_sync_buffers(nullptr, subctx);
-    vkCmdCopyBuffer(subctx->s->buffer->buf, (VkBuffer)staging_buffer->buffer, (VkBuffer)dst->buffer, 1, &buf_copy);
+    subctx->s->buffer->buf.copyBuffer((VkBuffer)staging_buffer->buffer, (VkBuffer)dst->buffer, slices);
 
     if (width == spitch) {
-        deferred_memcpy((uint8_t *)staging_buffer->ptr, src, width * height, &subctx->in_memcpys);
+        deferred_memcpy((uint8_t *)staging_buffer->ptr, src, staging_size, &subctx->in_memcpys);
     } else {
         for (size_t i = 0; i < height; i++) {
             deferred_memcpy((uint8_t *)staging_buffer->ptr + i * width, (const uint8_t *) src + i * spitch, width, &subctx->in_memcpys);
@@ -6899,24 +6915,24 @@ static bool ggml_vk_buffer_write_2d_async(vk_context subctx, vk_buffer& dst, siz
 
 static bool ggml_vk_buffer_write_async(vk_context subctx, vk_buffer& dst, size_t offset, const void * src, size_t size, bool sync_staging = false) {
     VK_LOG_DEBUG("ggml_vk_buffer_write_async(" << size << ")");
-    return ggml_vk_buffer_write_2d_async(subctx, dst, offset, src, size, size, 1, sync_staging);
+    return ggml_vk_buffer_write_2d_async(subctx, dst, offset, src, size, size, size, 1, sync_staging);
 }
 
-static void ggml_vk_buffer_write_2d(vk_buffer& dst, size_t offset, const void * src, size_t spitch, size_t width, size_t height) {
+static void ggml_vk_buffer_write_2d(vk_buffer& dst, size_t offset, const void * src, size_t spitch, size_t dpitch, size_t width, size_t height) {
     VK_LOG_DEBUG("ggml_vk_buffer_write_2d(" << width << ", " << height << ")");
     // Buffer is already mapped
     if(dst->memory_property_flags & vk::MemoryPropertyFlagBits::eHostVisible) {
         GGML_ASSERT(dst->memory_property_flags & vk::MemoryPropertyFlagBits::eHostCoherent);
 
         for (size_t i = 0; i < height; i++) {
-            memcpy((uint8_t *)dst->ptr + offset + i * width, (const uint8_t *) src + i * spitch, width);
+            memcpy((uint8_t *)dst->ptr + offset + i * dpitch, (const uint8_t *) src + i * spitch, width);
         }
     } else {
         std::lock_guard<std::recursive_mutex> guard(dst->device->mutex);
 
         vk_context subctx = ggml_vk_create_temporary_context(dst->device->transfer_queue.cmd_pool);
         ggml_vk_ctx_begin(dst->device, subctx);
-        bool ret = ggml_vk_buffer_write_2d_async(subctx, dst, offset, src, spitch, width, height, true);
+        bool ret = ggml_vk_buffer_write_2d_async(subctx, dst, offset, src, spitch, dpitch, width, height, true);
         GGML_ASSERT(ret);
         ggml_vk_ctx_end(subctx);
 
@@ -6937,7 +6953,7 @@ static void ggml_vk_buffer_write_2d(vk_buffer& dst, size_t offset, const void *
 
 static void ggml_vk_buffer_write(vk_buffer& dst, size_t offset, const void * src, size_t size) {
     VK_LOG_DEBUG("ggml_vk_buffer_write(" << size << ")");
-    ggml_vk_buffer_write_2d(dst, offset, src, 0, size, 1);
+    ggml_vk_buffer_write_2d(dst, offset, src, size, size, size, 1);
 }
 
 static bool ggml_vk_buffer_read_2d_async(vk_context subctx, vk_buffer& src, size_t offset, void * dst, size_t spitch, size_t dpitch, size_t width, size_t height, bool sync_staging = false) {
@@ -6983,15 +6999,35 @@ static bool ggml_vk_buffer_read_2d_async(vk_context subctx, vk_buffer& src, size
     }
 
     // Fall back to staging buffer
-    const size_t copy_size = dpitch * height;
-    ggml_vk_ensure_sync_staging_buffer(src->device, copy_size);
+    const size_t staging_size = width * height;
+    ggml_vk_ensure_sync_staging_buffer(src->device, staging_size);
 
     vk_buffer& staging_buffer = src->device->sync_staging;
 
-    ggml_vk_sync_buffers(nullptr, subctx);
-    subctx->s->buffer->buf.copyBuffer(src->buffer, staging_buffer->buffer, slices);
+    std::vector<vk::BufferCopy> staging_slices(1);
+    if (width == spitch) {
+        staging_slices[0].srcOffset = offset;
+        staging_slices[0].dstOffset = 0;
+        staging_slices[0].size = staging_size;
+    } else {
+        staging_slices.resize(height);
+        for (size_t i = 0; i < height; i++) {
+            staging_slices[i].srcOffset = offset + i * spitch;
+            staging_slices[i].dstOffset = i * width;
+            staging_slices[i].size = width;
+        }
+    }
 
-    deferred_memcpy(dst, staging_buffer->ptr, copy_size, &subctx->out_memcpys);
+    ggml_vk_sync_buffers(nullptr, subctx);
+    subctx->s->buffer->buf.copyBuffer(src->buffer, staging_buffer->buffer, staging_slices);
+
+    if (width == dpitch) {
+        deferred_memcpy(dst, staging_buffer->ptr, staging_size, &subctx->out_memcpys);
+    } else {
+        for (size_t i = 0; i < height; i++) {
+            deferred_memcpy((uint8_t *) dst + i * dpitch, (const uint8_t *) staging_buffer->ptr + i * width, width, &subctx->out_memcpys);
+        }
+    }
     return true;
 }
 
@@ -6999,8 +7035,8 @@ static bool ggml_vk_buffer_read_async(vk_context subctx, vk_buffer& src, size_t
     return ggml_vk_buffer_read_2d_async(subctx, src, offset, dst, size, size, size, 1, sync_staging);
 }
 
-static void ggml_vk_buffer_read(vk_buffer& src, size_t offset, void * dst, size_t size) {
-    VK_LOG_DEBUG("ggml_vk_buffer_read(" << src->buffer << ", " << offset << ", " << size << ")");
+static void ggml_vk_buffer_read_2d(vk_buffer& src, size_t offset, void * dst, size_t spitch, size_t dpitch, size_t width, size_t height) {
+    VK_LOG_DEBUG("ggml_vk_buffer_read_2d(" << src->buffer << ", " << offset << ", " << width << ", " << height << ")");
 
     // If the device is not an UMA device the memory is host-accessible through rebar. While writing
     // through PCIe is sufficient fast reading back data from PCIe is slower than going through
@@ -7008,18 +7044,20 @@ static void ggml_vk_buffer_read(vk_buffer& src, size_t offset, void * dst, size_
     if(src->memory_property_flags & vk::MemoryPropertyFlagBits::eHostVisible && src->device->uma) {
         GGML_ASSERT(src->memory_property_flags & vk::MemoryPropertyFlagBits::eHostCoherent);
 
-        memcpy(dst, (uint8_t *) src->ptr + offset, size);
+        for (size_t i = 0; i < height; i++) {
+            memcpy((uint8_t *) dst + i * dpitch, (const uint8_t *) src->ptr + offset + i * spitch, width);
+        }
     } else {
         std::lock_guard<std::recursive_mutex> guard(src->device->mutex);
 
         vk_context subctx = ggml_vk_create_temporary_context(src->device->transfer_queue.cmd_pool);
         ggml_vk_ctx_begin(src->device, subctx);
-        bool ret = ggml_vk_buffer_read_async(subctx, src, offset, dst, size, true);
+        bool ret = ggml_vk_buffer_read_2d_async(subctx, src, offset, dst, spitch, dpitch, width, height, true);
         GGML_ASSERT(ret);
         ggml_vk_ctx_end(subctx);
 
         ggml_vk_submit(subctx, src->device->fence);
-        VK_CHECK(src->device->device.waitForFences({ src->device->fence }, true, UINT64_MAX), "vk_buffer_read waitForFences");
+        VK_CHECK(src->device->device.waitForFences({ src->device->fence }, true, UINT64_MAX), "vk_buffer_read_2d waitForFences");
         src->device->device.resetFences({ src->device->fence });
         ggml_vk_queue_command_pools_cleanup(src->device);
 
@@ -7029,6 +7067,11 @@ static void ggml_vk_buffer_read(vk_buffer& src, size_t offset, void * dst, size_
     }
 }
 
+static void ggml_vk_buffer_read(vk_buffer& src, size_t offset, void * dst, size_t size) {
+    VK_LOG_DEBUG("ggml_vk_buffer_read(" << src->buffer << ", " << offset << ", " << size << ")");
+    ggml_vk_buffer_read_2d(src, offset, dst, size, size, size, 1);
+}
+
 static void ggml_vk_buffer_copy_async(vk_context& ctx, vk_buffer& dst, size_t dst_offset, vk_buffer& src, size_t src_offset, size_t size) {
     VK_LOG_DEBUG("ggml_vk_buffer_copy_async(" << size << ")");
     // Make sure both buffers are on same device
@@ -7060,7 +7103,7 @@ static void ggml_vk_buffer_copy(vk_buffer& dst, size_t dst_offset, vk_buffer& sr
         // Copy to src staging buffer
         ggml_vk_buffer_copy(src->device->sync_staging, 0, src, src_offset, size);
         // Copy to dst buffer
-        ggml_vk_buffer_write_2d(dst, dst_offset, src->device->sync_staging->ptr, 0, size, 1);
+        ggml_vk_buffer_write(dst, dst_offset, src->device->sync_staging->ptr, size);
     }
 }
 
@@ -13007,6 +13050,7 @@ static bool ggml_vk_build_graph(ggml_backend_vk_context * ctx, ggml_cgraph * cgr
             if (vk_perf_logger_enabled && vk_perf_logger_concurrent) {
                 ctx->query_node_idx[ctx->query_idx] = node_idx;
                 compute_ctx->s->buffer->buf.writeTimestamp(vk::PipelineStageFlagBits::eAllCommands, ctx->query_pool, ctx->query_idx++);
+                ggml_vk_sync_buffers(ctx, compute_ctx);
             }
         }
         // Add all fused nodes to the unsynchronized lists.
@@ -13607,6 +13651,20 @@ static void ggml_backend_vk_buffer_set_tensor(ggml_backend_buffer_t buffer, ggml
     ggml_vk_buffer_write(buf, vk_tensor_offset(tensor) + tensor->view_offs + offset, data, size);
 }
 
+static void ggml_backend_vk_buffer_set_tensor_2d(ggml_backend_buffer_t buffer, ggml_tensor * tensor, const void * data, size_t offset,
+                                                 size_t size, size_t n_copies, size_t stride_tensor, size_t stride_data) {
+    VK_LOG_DEBUG("ggml_backend_vk_buffer_set_tensor_2d(" << buffer << ", " << tensor << ", " << data << ", " << offset << ", " << size << ", " <<
+                 n_copies << ", " << stride_tensor << ", " << stride_data << ")");
+    ggml_backend_vk_buffer_context * buf_ctx = (ggml_backend_vk_buffer_context *)buffer->context;
+    vk_buffer buf = buf_ctx->dev_buffer;
+
+    if (size == 0) {
+        return;
+    }
+
+    ggml_vk_buffer_write_2d(buf, vk_tensor_offset(tensor) + tensor->view_offs + offset, data, stride_data, stride_tensor, size, n_copies);
+}
+
 static void ggml_backend_vk_buffer_get_tensor(ggml_backend_buffer_t buffer, const ggml_tensor * tensor, void * data, size_t offset, size_t size) {
     VK_LOG_DEBUG("ggml_backend_vk_buffer_get_tensor(" << buffer << ", " << tensor << ", " << data << ", " << offset << ", " << size << ")");
     ggml_backend_vk_buffer_context * buf_ctx = (ggml_backend_vk_buffer_context *)buffer->context;
@@ -13620,6 +13678,21 @@ static void ggml_backend_vk_buffer_get_tensor(ggml_backend_buffer_t buffer, cons
     ggml_vk_buffer_read(buf, vk_tensor_offset(tensor) + tensor->view_offs + offset, data, size);
 }
 
+static void ggml_backend_vk_buffer_get_tensor_2d(ggml_backend_buffer_t buffer, const ggml_tensor * tensor, void * data, size_t offset,
+                                                 size_t size, size_t n_copies, size_t stride_tensor, size_t stride_data) {
+    VK_LOG_DEBUG("ggml_backend_vk_buffer_get_tensor_2d(" << buffer << ", " << tensor << ", " << data << ", " << offset << ", " << size << ", " <<
+                 n_copies << ", " << stride_tensor << ", " << stride_data << ")");
+    ggml_backend_vk_buffer_context * buf_ctx = (ggml_backend_vk_buffer_context *)buffer->context;
+
+    if (size == 0) {
+        return;
+    }
+
+    vk_buffer buf = buf_ctx->dev_buffer;
+
+    ggml_vk_buffer_read_2d(buf, vk_tensor_offset(tensor) + tensor->view_offs + offset, data, stride_tensor, stride_data, size, n_copies);
+}
+
 static bool ggml_backend_vk_buffer_cpy_tensor(ggml_backend_buffer_t buffer, const ggml_tensor * src, ggml_tensor * dst) {
     if (ggml_nbytes(src) == 0) {
         return true;
@@ -13654,8 +13727,8 @@ static ggml_backend_buffer_i ggml_backend_vk_buffer_interface = {
     /* .memset_tensor   = */ ggml_backend_vk_buffer_memset_tensor,
     /* .set_tensor      = */ ggml_backend_vk_buffer_set_tensor,
     /* .get_tensor      = */ ggml_backend_vk_buffer_get_tensor,
-    /* .set_tensor_2d   = */ NULL,
-    /* .get_tensor_2d   = */ NULL,
+    /* .set_tensor_2d   = */ ggml_backend_vk_buffer_set_tensor_2d,
+    /* .get_tensor_2d   = */ ggml_backend_vk_buffer_get_tensor_2d,
     /* .cpy_tensor      = */ ggml_backend_vk_buffer_cpy_tensor,
     /* .clear           = */ ggml_backend_vk_buffer_clear,
     /* .reset           = */ NULL,
@@ -13811,8 +13884,9 @@ static ggml_backend_buffer_type_t ggml_backend_vk_get_default_buffer_type(ggml_b
     return &ctx->device->buffer_type;
 }
 
-static void ggml_backend_vk_set_tensor_async(ggml_backend_t backend, ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
-    VK_LOG_DEBUG("ggml_backend_vk_set_tensor_async(" << size << ")");
+static void ggml_backend_vk_set_tensor_2d_async(ggml_backend_t backend, ggml_tensor * tensor, const void * data, size_t offset,
+                                                size_t size, size_t n_copies, size_t stride_tensor, size_t stride_data) {
+    VK_LOG_DEBUG("ggml_backend_vk_set_tensor_2d_async(" << size << ", " << n_copies << ")");
     ggml_backend_vk_context * ctx = (ggml_backend_vk_context *)backend->context;
     GGML_ASSERT((tensor->buffer->buft == ggml_backend_vk_get_default_buffer_type(backend) || tensor->buffer->buft == ggml_backend_vk_host_buffer_type()) && "unsupported buffer type");
 
@@ -13826,7 +13900,6 @@ static void ggml_backend_vk_set_tensor_async(ggml_backend_t backend, ggml_tensor
 
     if (ctx->device->async_use_transfer_queue) {
         if (ctx->transfer_ctx.expired()) {
-            // Initialize new transfer context
             cpy_ctx = ggml_vk_create_context(ctx, ctx->transfer_cmd_pool);
             ctx->transfer_ctx = cpy_ctx;
             ggml_vk_ctx_begin(ctx->device, cpy_ctx);
@@ -13841,25 +13914,48 @@ static void ggml_backend_vk_set_tensor_async(ggml_backend_t backend, ggml_tensor
 
     auto dst_offset = vk_tensor_offset(tensor) + tensor->view_offs + offset;
 
-    bool ret = ggml_vk_buffer_write_async(cpy_ctx, buf, dst_offset, data, size);
+    bool ret = ggml_vk_buffer_write_2d_async(cpy_ctx, buf, dst_offset, data, stride_data, stride_tensor, size, n_copies);
 
     if (!ret) {
-        ggml_vk_ensure_sync_staging_buffer(ctx, size);
+        const size_t staging_size = size * n_copies;
+        ggml_vk_ensure_sync_staging_buffer(ctx, staging_size);
         ggml_vk_sync_buffers(nullptr, cpy_ctx);
 
-        vk::BufferCopy buffer_cpy;
-        buffer_cpy.srcOffset = 0;
-        buffer_cpy.dstOffset = dst_offset;
-        buffer_cpy.size = size;
+        std::vector<vk::BufferCopy> slices(1);
+        if (size == stride_tensor) {
+            slices[0].srcOffset = 0;
+            slices[0].dstOffset = dst_offset;
+            slices[0].size = staging_size;
+        } else {
+            slices.resize(n_copies);
+            for (size_t i = 0; i < n_copies; i++) {
+                slices[i].srcOffset = i * size;
+                slices[i].dstOffset = dst_offset + i * stride_tensor;
+                slices[i].size = size;
+            }
+        }
 
-        cpy_ctx->s->buffer->buf.copyBuffer(ctx->sync_staging->buffer, buf->buffer, { buffer_cpy });
-        deferred_memcpy(ctx->sync_staging->ptr, data, size, &cpy_ctx->in_memcpys);
+        cpy_ctx->s->buffer->buf.copyBuffer(ctx->sync_staging->buffer, buf->buffer, slices);
+
+        if (size == stride_data) {
+            deferred_memcpy(ctx->sync_staging->ptr, data, staging_size, &cpy_ctx->in_memcpys);
+        } else {
+            for (size_t i = 0; i < n_copies; i++) {
+                deferred_memcpy((uint8_t *)ctx->sync_staging->ptr + i * size, (const uint8_t *)data + i * stride_data, size, &cpy_ctx->in_memcpys);
+            }
+        }
         ggml_vk_synchronize(ctx);
     }
 }
 
-static void ggml_backend_vk_get_tensor_async(ggml_backend_t backend, const ggml_tensor * tensor, void * data, size_t offset, size_t size) {
-    VK_LOG_DEBUG("ggml_backend_vk_get_tensor_async(" << size << ")");
+static void ggml_backend_vk_set_tensor_async(ggml_backend_t backend, ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
+    VK_LOG_DEBUG("ggml_backend_vk_set_tensor_async(" << size << ")");
+    ggml_backend_vk_set_tensor_2d_async(backend, tensor, data, offset, size, 1, size, size);
+}
+
+static void ggml_backend_vk_get_tensor_2d_async(ggml_backend_t backend, const ggml_tensor * tensor, void * data, size_t offset,
+                                                size_t size, size_t n_copies, size_t stride_tensor, size_t stride_data) {
+    VK_LOG_DEBUG("ggml_backend_vk_get_tensor_2d_async(" << size << ", " << n_copies << ")");
     ggml_backend_vk_context * ctx = (ggml_backend_vk_context *)backend->context;
     GGML_ASSERT((tensor->buffer->buft == ggml_backend_vk_get_default_buffer_type(backend) || tensor->buffer->buft == ggml_backend_vk_host_buffer_type()) && "unsupported buffer type");
 
@@ -13874,24 +13970,45 @@ static void ggml_backend_vk_get_tensor_async(ggml_backend_t backend, const ggml_
     vk_buffer buf = buf_ctx->dev_buffer;
 
     auto src_offset = vk_tensor_offset(tensor) + tensor->view_offs + offset;
-    bool ret = ggml_vk_buffer_read_async(compute_ctx, buf, src_offset, data, size);
+    bool ret = ggml_vk_buffer_read_2d_async(compute_ctx, buf, src_offset, data, stride_tensor, stride_data, size, n_copies);
 
-    // If that failed, copy synchronously through a staging buffer
     if (!ret) {
-        ggml_vk_ensure_sync_staging_buffer(ctx, size);
+        const size_t staging_size = size * n_copies;
+        ggml_vk_ensure_sync_staging_buffer(ctx, staging_size);
         ggml_vk_sync_buffers(nullptr, compute_ctx);
 
-        vk::BufferCopy buffer_cpy;
-        buffer_cpy.srcOffset = src_offset;
-        buffer_cpy.dstOffset = 0;
-        buffer_cpy.size = size;
+        std::vector<vk::BufferCopy> slices(1);
+        if (size == stride_tensor) {
+            slices[0].srcOffset = src_offset;
+            slices[0].dstOffset = 0;
+            slices[0].size = staging_size;
+        } else {
+            slices.resize(n_copies);
+            for (size_t i = 0; i < n_copies; i++) {
+                slices[i].srcOffset = src_offset + i * stride_tensor;
+                slices[i].dstOffset = i * size;
+                slices[i].size = size;
+            }
+        }
 
-        compute_ctx->s->buffer->buf.copyBuffer(buf->buffer, ctx->sync_staging->buffer, { buffer_cpy });
-        deferred_memcpy(data, ctx->sync_staging->ptr, size, &compute_ctx->out_memcpys);
+        compute_ctx->s->buffer->buf.copyBuffer(buf->buffer, ctx->sync_staging->buffer, slices);
+
+        if (size == stride_data) {
+            deferred_memcpy(data, ctx->sync_staging->ptr, staging_size, &compute_ctx->out_memcpys);
+        } else {
+            for (size_t i = 0; i < n_copies; i++) {
+                deferred_memcpy((uint8_t *)data + i * stride_data, (const uint8_t *)ctx->sync_staging->ptr + i * size, size, &compute_ctx->out_memcpys);
+            }
+        }
         ggml_vk_synchronize(ctx);
     }
 }
 
+static void ggml_backend_vk_get_tensor_async(ggml_backend_t backend, const ggml_tensor * tensor, void * data, size_t offset, size_t size) {
+    VK_LOG_DEBUG("ggml_backend_vk_get_tensor_async(" << size << ")");
+    ggml_backend_vk_get_tensor_2d_async(backend, tensor, data, offset, size, 1, size, size);
+}
+
 static bool ggml_backend_vk_cpy_tensor_async(ggml_backend_t backend_src, ggml_backend_t backend_dst, const ggml_tensor * src, ggml_tensor * dst) {
     VK_LOG_DEBUG("ggml_backend_vk_cpy_tensor_async(" << src << " -> " << dst << ", size=" << ggml_nbytes(src) << ")");
     ggml_backend_vk_context * ctx = (ggml_backend_vk_context *)backend_dst->context;
@@ -14496,6 +14613,7 @@ static ggml_status ggml_backend_vk_graph_compute(ggml_backend_t backend, ggml_cg
         compute_ctx = ggml_vk_get_compute_ctx(ctx);
         ctx->query_idx = 0;
         compute_ctx->s->buffer->buf.writeTimestamp(vk::PipelineStageFlagBits::eAllCommands, ctx->query_pool, ctx->query_idx++);
+        ggml_vk_sync_buffers(ctx, compute_ctx);
     }
 
     ctx->prealloc_y_last_pipeline_used = nullptr;
@@ -14732,6 +14850,7 @@ static ggml_status ggml_backend_vk_graph_compute(ggml_backend_t backend, ggml_cg
                 ctx->query_nodes[ctx->query_idx] = cgraph->nodes[i];
                 ctx->query_fusion_names[ctx->query_idx] = fusion_string;
                 compute_ctx->s->buffer->buf.writeTimestamp(vk::PipelineStageFlagBits::eAllCommands, ctx->query_pool, ctx->query_idx++);
+                ggml_vk_sync_buffers(ctx, compute_ctx);
             } else {
                 // track a fusion string and number of fused ops for the current node_idx
                 ctx->query_fusion_names[i] = fusion_string;
@@ -15113,8 +15232,8 @@ static ggml_backend_i ggml_backend_vk_interface = {
     /* .free                    = */ ggml_backend_vk_free,
     /* .set_tensor_async        = */ ggml_backend_vk_set_tensor_async,
     /* .get_tensor_async        = */ ggml_backend_vk_get_tensor_async,
-    /* .get_tensor_2d_async     = */ NULL,
-    /* .set_tensor_2d_async     = */ NULL,
+    /* .set_tensor_2d_async     = */ ggml_backend_vk_set_tensor_2d_async,
+    /* .get_tensor_2d_async     = */ ggml_backend_vk_get_tensor_2d_async,
     /* .cpy_tensor_async        = */ ggml_backend_vk_cpy_tensor_async,
     /* .synchronize             = */ ggml_backend_vk_synchronize,
     /* .graph_plan_create       = */ NULL,

ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vecq_funcs.glsl

@@ -296,13 +296,22 @@ vec2 get_dm_scale(uint ib, uint iqs) {
     const uint ib_k = ib / 8;
     const uint iqs_k = (ib % 8) * 8 + iqs;
     const uint is = iqs_k / 8;
-    u8vec2 scale_dm;
-    if (is < 4) {
-        scale_dm = u8vec2(data_a[ib_k].scales[is] & 0x3F, data_a[ib_k].scales[is + 4] & 0x3F);
-    } else {
-        scale_dm = u8vec2((data_a[ib_k].scales[is+4] & 0xF) | ((data_a[ib_k].scales[is-4] & 0xC0) >> 2),
-                          (data_a[ib_k].scales[is+4] >>  4) | ((data_a[ib_k].scales[is  ] & 0xC0) >> 2));
-    }
+
+    const uvec3 scales = uvec3(data_a_packed32[ib_k].scales[0],
+                               data_a_packed32[ib_k].scales[1],
+                               data_a_packed32[ib_k].scales[2]);
+    const uint scalesoffs = (is & 3) * 8;
+
+    const uint scidx0 = (is < 4) ? 0 : 2;
+    const uint scidxshift0 = scalesoffs;
+    const uint scidxshift1 = (is < 4) ? scalesoffs : scalesoffs + 2;
+    const uint mbidx0 = (is < 4) ? 1 : 2;
+    const uint mbidxshift0 = (is < 4) ? scalesoffs : scalesoffs + 4;
+    const uint mbidxshift1 = (is < 4) ? scalesoffs : scalesoffs + 2;
+
+    const uint8_t sc    = uint8_t(((scales[scidx0] >> scidxshift0) & 0xF) | ((scales[0] >> scidxshift1) & 0x30));
+    const uint8_t mbyte = uint8_t(((scales[mbidx0] >> mbidxshift0) & 0xF) | ((scales[1] >> mbidxshift1) & 0x30));
+    u8vec2 scale_dm = u8vec2(sc, mbyte);
 
     return FLOAT_TYPEV2(data_a_packed32[ib_k].dm) * FLOAT_TYPEV2(scale_dm);
 }

ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_funcs.glsl

@@ -201,19 +201,20 @@ void load_a_to_shmem(const uint pos_a, const uint row, const uint col, const uin
 
             const vec2 loadd = vec2(data_a[ib].dm);
 
-            const uint scidx0 = (is < 4) ? is : (is + 4);
-            const uint scidx1 = (is < 4) ? is : (is - 4);
-            const uint scidxmask1 = (is < 4) ? 0x30 : 0xC0;
-            const uint scidxshift1 = (is < 4) ? 0 : 2;
-            const uint mbidx0 = is + 4;
-            const uint mbidx1 = (is < 4) ? is + 4 : is;
-            const uint mbidxmask0 = (is < 4) ? 0xF : 0xF0;
-            const uint mbidxshift0 = (is < 4) ? 0 : 4;
-            const uint mbidxmask1 = (is < 4) ? 0x30 : 0xC0;
-            const uint mbidxshift1 = (is < 4) ? 0 : 2;
+            const uvec3 scales = uvec3(data_a_packed32[ib].scales[0],
+                                       data_a_packed32[ib].scales[1],
+                                       data_a_packed32[ib].scales[2]);
+            const uint scalesoffs = (is & 3) * 8;
 
-            const uint8_t sc = uint8_t((data_a[ib].scales[scidx0] & 0xF) | ((data_a[ib].scales[scidx1] & scidxmask1) >> scidxshift1));
-            const uint8_t mbyte = uint8_t((data_a[ib].scales[mbidx0] & mbidxmask0) >> mbidxshift0 | ((data_a[ib].scales[mbidx1] & mbidxmask1) >> mbidxshift1));
+            const uint scidx0 = (is < 4) ? 0 : 2;
+            const uint scidxshift0 = scalesoffs;
+            const uint scidxshift1 = (is < 4) ? scalesoffs : scalesoffs + 2;
+            const uint mbidx0 = (is < 4) ? 1 : 2;
+            const uint mbidxshift0 = (is < 4) ? scalesoffs : scalesoffs + 4;
+            const uint mbidxshift1 = (is < 4) ? scalesoffs : scalesoffs + 2;
+
+            const uint8_t sc    = uint8_t(((scales[scidx0] >> scidxshift0) & 0xF) | ((scales[0] >> scidxshift1) & 0x30));
+            const uint8_t mbyte = uint8_t(((scales[mbidx0] >> mbidxshift0) & 0xF) | ((scales[1] >> mbidxshift1) & 0x30));
 
             const float d = loadd.x * sc;
             const float m = -loadd.y * mbyte;
@@ -237,19 +238,20 @@ void load_a_to_shmem(const uint pos_a, const uint row, const uint col, const uin
 
             const vec2 loadd = vec2(data_a[ib].dm);
 
-            const uint scidx0 = (is < 4) ? is : (is + 4);
-            const uint scidx1 = (is < 4) ? is : (is - 4);
-            const uint scidxmask1 = (is < 4) ? 0x30 : 0xC0;
-            const uint scidxshift1 = (is < 4) ? 0 : 2;
-            const uint mbidx0 = is + 4;
-            const uint mbidx1 = (is < 4) ? is + 4 : is;
-            const uint mbidxmask0 = (is < 4) ? 0xF : 0xF0;
-            const uint mbidxshift0 = (is < 4) ? 0 : 4;
-            const uint mbidxmask1 = (is < 4) ? 0x30 : 0xC0;
-            const uint mbidxshift1 = (is < 4) ? 0 : 2;
+            const uvec3 scales = uvec3(data_a_packed32[ib].scales[0],
+                                       data_a_packed32[ib].scales[1],
+                                       data_a_packed32[ib].scales[2]);
+            const uint scalesoffs = (is & 3) * 8;
 
-            const uint8_t sc    = uint8_t((data_a[ib].scales[scidx0] & 0xF)                         | ((data_a[ib].scales[scidx1] & scidxmask1) >> scidxshift1));
-            const uint8_t mbyte = uint8_t(((data_a[ib].scales[mbidx0] & mbidxmask0) >> mbidxshift0) | ((data_a[ib].scales[mbidx1] & mbidxmask1) >> mbidxshift1));
+            const uint scidx0 = (is < 4) ? 0 : 2;
+            const uint scidxshift0 = scalesoffs;
+            const uint scidxshift1 = (is < 4) ? scalesoffs : scalesoffs + 2;
+            const uint mbidx0 = (is < 4) ? 1 : 2;
+            const uint mbidxshift0 = (is < 4) ? scalesoffs : scalesoffs + 4;
+            const uint mbidxshift1 = (is < 4) ? scalesoffs : scalesoffs + 2;
+
+            const uint8_t sc    = uint8_t(((scales[scidx0] >> scidxshift0) & 0xF) | ((scales[0] >> scidxshift1) & 0x30));
+            const uint8_t mbyte = uint8_t(((scales[mbidx0] >> mbidxshift0) & 0xF) | ((scales[1] >> mbidxshift1) & 0x30));
 
             const float d = loadd.x * sc;
             const float m = -loadd.y * mbyte;

ggml/src/ggml-webgpu/ggml-webgpu-shader-lib.hpp

@@ -26,21 +26,21 @@
 // Matrix multiplication parameters
 
 // Register tiling parameters
-#define WEBGPU_MUL_MAT_TILE_M    4
-#define WEBGPU_MUL_MAT_TILE_N    4
-#define WEBGPU_MUL_MAT_WG_SIZE_M 8
-#define WEBGPU_MUL_MAT_WG_SIZE_N 8
+#define WEBGPU_MUL_MAT_TILE_M           4
+#define WEBGPU_MUL_MAT_TILE_N           4
+#define WEBGPU_MUL_MAT_WG_SIZE_M        8
+#define WEBGPU_MUL_MAT_WG_SIZE_N        8
 #define WEBGPU_MUL_MAT_REG_TILE_K_FLOAT 8
 #define WEBGPU_MUL_MAT_REG_TILE_K_QUANT 32
 
 // Subgroup matrix parameters
 // The number of subgroups in the M dimension
-#define WEBGPU_MUL_MAT_SUBGROUP_M        2
+#define WEBGPU_MUL_MAT_SUBGROUP_M            2
 // The number of subgroups in the N dimension
-#define WEBGPU_MUL_MAT_SUBGROUP_N        4
+#define WEBGPU_MUL_MAT_SUBGROUP_N            4
 // The number of subgroup matrices each subgroup accumulates over
-#define WEBGPU_MUL_MAT_SUBGROUP_MATRIX_M 4
-#define WEBGPU_MUL_MAT_SUBGROUP_MATRIX_N 2
+#define WEBGPU_MUL_MAT_SUBGROUP_MATRIX_M     4
+#define WEBGPU_MUL_MAT_SUBGROUP_MATRIX_N     2
 #define WEBGPU_MUL_MAT_SUBGROUP_TILE_K_FLOAT 32
 #define WEBGPU_MUL_MAT_SUBGROUP_TILE_K_QUANT 32
 
@@ -59,19 +59,32 @@ template <typename T> inline void ggml_webgpu_hash_combine(size_t & seed, const
     seed ^= std::hash<T>{}(value) + 0x9e3779b9 + (seed << 6) + (seed >> 2);
 }
 
+// Calculates base address of a tensor ignoring the fake base pointer
+inline uintptr_t ggml_webgpu_tensor_addr(const ggml_tensor * tensor) {
+    const ggml_tensor * base_tensor = tensor->view_src ? tensor->view_src : tensor;
+    return (uintptr_t) base_tensor->data + tensor->view_offs;
+}
+
+inline bool ggml_webgpu_tensor_equal(const ggml_tensor * a, const ggml_tensor * b) {
+    return a->buffer == b->buffer && ggml_webgpu_tensor_addr(a) == ggml_webgpu_tensor_addr(b);
+}
+
+inline bool ggml_webgpu_tensor_overlap(const ggml_tensor * a, const ggml_tensor * b) {
+    return a->buffer == b->buffer && ggml_webgpu_tensor_addr(a) < ggml_webgpu_tensor_addr(b) + ggml_nbytes(b) &&
+           ggml_webgpu_tensor_addr(b) < ggml_webgpu_tensor_addr(a) + ggml_nbytes(a);
+}
+
 struct ggml_webgpu_shader_lib_context {
     ggml_tensor * src0;
     ggml_tensor * src1;
     ggml_tensor * src2;
     ggml_tensor * src3;
     ggml_tensor * src4;
+    ggml_tensor * src5;
     ggml_tensor * dst;
 
     uint32_t max_wg_size;
     size_t   wg_mem_limit_bytes       = 0;
-    bool     inplace                  = false;
-    bool     overlap                  = false;
-    bool     src_overlap              = false;
     bool     supports_subgroups       = false;
     bool     supports_subgroup_matrix = false;
     uint32_t sg_mat_m                 = 0;
@@ -88,6 +101,14 @@ struct webgpu_pipeline {
 
 struct ggml_webgpu_generic_shader_decisions {
     uint32_t wg_size = 0;
+    bool     inplace = false;
+};
+
+struct ggml_webgpu_binary_shader_decisions {
+    uint32_t wg_size     = 0;
+    bool     inplace     = false;
+    bool     overlap     = false;
+    bool     src_overlap = false;
 };
 
 struct ggml_webgpu_processed_shader {
@@ -102,11 +123,12 @@ struct ggml_webgpu_ssm_conv_shader_decisions {
 };
 
 struct ggml_webgpu_ssm_scan_pipeline_key {
-    int type;
-    int d_state;
+    int  type;
+    int  d_state;
+    bool xbc_overlap;
 
     bool operator==(const ggml_webgpu_ssm_scan_pipeline_key & other) const {
-        return type == other.type && d_state == other.d_state;
+        return type == other.type && d_state == other.d_state && xbc_overlap == other.xbc_overlap;
     }
 };
 
@@ -115,6 +137,7 @@ struct ggml_webgpu_ssm_scan_pipeline_key_hash {
         size_t seed = 0;
         ggml_webgpu_hash_combine(seed, key.type);
         ggml_webgpu_hash_combine(seed, key.d_state);
+        ggml_webgpu_hash_combine(seed, key.xbc_overlap);
         return seed;
     }
 };
@@ -122,6 +145,7 @@ struct ggml_webgpu_ssm_scan_pipeline_key_hash {
 struct ggml_webgpu_ssm_scan_shader_decisions {
     uint32_t wg_size;
     uint32_t tokens_per_tile;
+    bool     xbc_overlap = false;
 };
 
 /** Argsort **/
@@ -242,6 +266,13 @@ struct ggml_webgpu_rms_norm_mul_pipeline_key_hash {
     }
 };
 
+struct ggml_webgpu_rms_norm_mul_shader_decisions {
+    uint32_t wg_size     = 0;
+    bool     inplace     = false;
+    bool     overlap     = false;
+    bool     src_overlap = false;
+};
+
 /** Pad **/
 struct ggml_webgpu_pad_pipeline_key {
     bool circular;
@@ -463,9 +494,10 @@ struct ggml_webgpu_unary_pipeline_key_hash {
 /** FlashAttention */
 
 enum ggml_webgpu_flash_attn_path : uint32_t {
-    GGML_WEBGPU_FLASH_ATTN_PATH_SUBGROUP_MATRIX = 0u,
-    GGML_WEBGPU_FLASH_ATTN_PATH_TILE            = 1u,
-    GGML_WEBGPU_FLASH_ATTN_PATH_VEC             = 2u,
+    GGML_WEBGPU_FLASH_ATTN_PATH_NONE            = 0u,
+    GGML_WEBGPU_FLASH_ATTN_PATH_SUBGROUP_MATRIX = 1u,
+    GGML_WEBGPU_FLASH_ATTN_PATH_TILE            = 2u,
+    GGML_WEBGPU_FLASH_ATTN_PATH_VEC             = 3u,
 };
 
 struct ggml_webgpu_flash_attn_pipeline_key {
@@ -503,11 +535,12 @@ struct ggml_webgpu_flash_attn_pipeline_key_hash {
 };
 
 struct ggml_webgpu_flash_attn_decisions {
-    uint32_t path      = GGML_WEBGPU_FLASH_ATTN_PATH_SUBGROUP_MATRIX;
-    uint32_t q_tile    = 0;
-    uint32_t kv_tile   = 0;
-    uint32_t wg_size   = 0;
-    bool     kv_direct = false;
+    uint32_t path       = GGML_WEBGPU_FLASH_ATTN_PATH_NONE;
+    uint32_t q_tile     = 0;
+    uint32_t kv_tile    = 0;
+    uint32_t wg_size    = 0;
+    bool     kv_direct  = false;
+    bool     kv_overlap = false;
 };
 
 inline constexpr uint32_t GGML_WEBGPU_FLASH_ATTN_TILE_KV_VEC_WIDTH = 4u;
@@ -552,7 +585,7 @@ inline ggml_webgpu_flash_attn_pipeline_key ggml_webgpu_flash_attn_make_pipeline_
     key.head_dim_qk                         = (uint32_t) context.src0->ne[0];
     key.head_dim_v                          = (uint32_t) context.src2->ne[0];
     key.kv_direct                           = kv_direct;
-    key.kv_overlap                          = context.src_overlap;
+    key.kv_overlap                          = ggml_webgpu_tensor_overlap(context.src1, context.src2);
     key.has_mask                            = has_mask;
     key.has_sinks                           = has_sinks;
     key.uses_logit_softcap                  = ggml_get_op_params_f32(context.dst, 2) != 0.0f;
@@ -677,19 +710,29 @@ inline ggml_webgpu_flash_attn_decisions ggml_webgpu_flash_attn_get_decisions(
                           (context.src0->ne[0] % GGML_WEBGPU_FLASH_ATTN_TILE_KV_VEC_WIDTH == 0) &&
                           (context.src2->ne[0] % GGML_WEBGPU_FLASH_ATTN_TILE_KV_VEC_WIDTH == 0) && !use_vec;
 
-    decisions.path = use_vec  ? GGML_WEBGPU_FLASH_ATTN_PATH_VEC :
-                     use_tile ? GGML_WEBGPU_FLASH_ATTN_PATH_TILE :
-                                GGML_WEBGPU_FLASH_ATTN_PATH_SUBGROUP_MATRIX;
+    decisions.path = use_vec                          ? GGML_WEBGPU_FLASH_ATTN_PATH_VEC :
+                     use_tile                         ? GGML_WEBGPU_FLASH_ATTN_PATH_TILE :
+                     context.supports_subgroup_matrix ? GGML_WEBGPU_FLASH_ATTN_PATH_SUBGROUP_MATRIX :
+                                                        GGML_WEBGPU_FLASH_ATTN_PATH_NONE;
+
+    if (decisions.path == GGML_WEBGPU_FLASH_ATTN_PATH_NONE) {
+        return decisions;
+    }
 
     const ggml_webgpu_flash_attn_pipeline_key key = ggml_webgpu_flash_attn_make_pipeline_key(context, decisions.path);
     decisions.kv_direct                           = key.kv_direct;
+    const uint32_t max_kv_tile                    = ggml_webgpu_flash_attn_max_kv_tile(context, key);
+    // invalidate if even the smallest kv_tile doesn't fit in shared memory
+    if (max_kv_tile == 0) {
+        decisions.path = GGML_WEBGPU_FLASH_ATTN_PATH_NONE;
+        return decisions;
+    }
 
     if (decisions.path == GGML_WEBGPU_FLASH_ATTN_PATH_VEC) {
-        const uint32_t min_kv_tile = ggml_webgpu_flash_attn_max_kv_tile(context, key);
-        decisions.q_tile           = 1u;
-        decisions.kv_tile          = std::max(8u, std::min(32u, min_kv_tile));
-        decisions.kv_tile          = (decisions.kv_tile / 8u) * 8u;
-        decisions.wg_size          = std::max(1u, std::min<uint32_t>(32u, context.max_subgroup_size));
+        decisions.q_tile  = 1u;
+        decisions.kv_tile = std::max(8u, std::min(32u, max_kv_tile));
+        decisions.kv_tile = (decisions.kv_tile / 8u) * 8u;
+        decisions.wg_size = std::max(1u, std::min<uint32_t>(32u, context.max_subgroup_size));
         if (decisions.kv_direct) {
             decisions.kv_tile = std::min(decisions.kv_tile, GGML_WEBGPU_KV_SEQ_PAD);
             while (GGML_WEBGPU_KV_SEQ_PAD % decisions.kv_tile != 0) {
@@ -702,9 +745,8 @@ inline ggml_webgpu_flash_attn_decisions ggml_webgpu_flash_attn_get_decisions(
     decisions.q_tile =
         decisions.path == GGML_WEBGPU_FLASH_ATTN_PATH_TILE ? GGML_WEBGPU_FLASH_ATTN_TILE_Q_TILE : context.sg_mat_m;
     decisions.kv_tile = decisions.path == GGML_WEBGPU_FLASH_ATTN_PATH_TILE ?
-                            std::min(64u, ggml_webgpu_flash_attn_max_kv_tile(context, key)) :
-                            std::min(ggml_webgpu_flash_attn_max_kv_tile(context, key),
-                                     context.sg_mat_n * GGML_WEBGPU_FLASH_ATTN_PREFERRED_KV_SG_TILES);
+                            std::min(64u, max_kv_tile) :
+                            std::min(max_kv_tile, context.sg_mat_n * GGML_WEBGPU_FLASH_ATTN_PREFERRED_KV_SG_TILES);
     decisions.wg_size = decisions.path == GGML_WEBGPU_FLASH_ATTN_PATH_TILE ?
                             GGML_WEBGPU_FLASH_ATTN_PREFERRED_WG_SIZE :
                             std::max(context.max_subgroup_size, GGML_WEBGPU_FLASH_ATTN_PREFERRED_WG_SIZE);
@@ -723,7 +765,6 @@ inline ggml_webgpu_flash_attn_decisions ggml_webgpu_flash_attn_get_decisions(
                                      context.sg_mat_n;
         }
     }
-
     return decisions;
 }
 
@@ -1021,7 +1062,7 @@ class ggml_webgpu_shader_lib {
     webgpu_pipeline get_row_norm_pipeline(const ggml_webgpu_shader_lib_context & context) {
         ggml_webgpu_row_norm_pipeline_key key = {};
         key.op                                = context.dst->op;
-        key.inplace                           = context.inplace;
+        key.inplace                           = ggml_webgpu_tensor_equal(context.src0, context.dst);
 
         auto it = row_norm_pipelines.find(key);
         if (it != row_norm_pipelines.end()) {
@@ -1051,8 +1092,12 @@ class ggml_webgpu_shader_lib {
         const uint32_t row_norm_wg_size = 128u;
         uint32_t       wg_size          = std::min(context.max_wg_size, row_norm_wg_size);
         defines.push_back(std::string("WG_SIZE=") + std::to_string(wg_size));
-        auto processed          = preprocessor.preprocess(wgsl_row_norm, defines);
-        row_norm_pipelines[key] = ggml_webgpu_create_pipeline(device, processed, variant);
+        auto processed                  = preprocessor.preprocess(wgsl_row_norm, defines);
+        auto decisions                  = std::make_shared<ggml_webgpu_generic_shader_decisions>();
+        decisions->wg_size              = wg_size;
+        decisions->inplace              = key.inplace;
+        row_norm_pipelines[key]         = ggml_webgpu_create_pipeline(device, processed, variant);
+        row_norm_pipelines[key].context = decisions;
         return row_norm_pipelines[key];
     }
 
@@ -1127,7 +1172,7 @@ class ggml_webgpu_shader_lib {
     webgpu_pipeline get_set_pipeline(const ggml_webgpu_shader_lib_context & context) {
         ggml_webgpu_set_pipeline_key key = {};
         key.type                         = context.dst->type;
-        key.inplace                      = context.inplace;
+        key.inplace                      = ggml_webgpu_tensor_equal(context.src0, context.dst);
 
         auto it = set_pipelines.find(key);
         if (it != set_pipelines.end()) {
@@ -1160,6 +1205,7 @@ class ggml_webgpu_shader_lib {
         auto processed           = preprocessor.preprocess(wgsl_set, defines);
         auto decisions           = std::make_shared<ggml_webgpu_generic_shader_decisions>();
         decisions->wg_size       = context.max_wg_size;
+        decisions->inplace       = key.inplace;
         webgpu_pipeline pipeline = ggml_webgpu_create_pipeline(device, processed, variant);
         pipeline.context         = decisions;
         set_pipelines[key]       = pipeline;
@@ -1287,6 +1333,7 @@ class ggml_webgpu_shader_lib {
                     std::transform(type_upper.begin(), type_upper.end(), type_upper.begin(), ::toupper);
 
                     switch (key.src_type) {
+                        case GGML_TYPE_Q1_0:
                         case GGML_TYPE_Q4_0:
                         case GGML_TYPE_Q5_0:
                         case GGML_TYPE_Q8_0:
@@ -1323,8 +1370,10 @@ class ggml_webgpu_shader_lib {
 
                     defines.push_back("DST_TYPE=f32");
 
-                    if ((key.src_type >= GGML_TYPE_Q4_0 && key.src_type <= GGML_TYPE_Q8_1) ||
-                        key.src_type == GGML_TYPE_IQ4_NL) {
+                    if (key.src_type == GGML_TYPE_Q1_0) {
+                        defines.push_back("BLOCK_SIZE=128u");
+                    } else if ((key.src_type >= GGML_TYPE_Q4_0 && key.src_type <= GGML_TYPE_Q8_1) ||
+                               key.src_type == GGML_TYPE_IQ4_NL) {
                         defines.push_back("BLOCK_SIZE=32u");
                     } else if (key.src_type >= GGML_TYPE_Q2_K) {
                         defines.push_back("BLOCK_SIZE=256u");
@@ -1352,7 +1401,7 @@ class ggml_webgpu_shader_lib {
 
     webgpu_pipeline get_scale_pipeline(const ggml_webgpu_shader_lib_context & context) {
         ggml_webgpu_scale_pipeline_key key = {};
-        key.inplace                        = context.inplace;
+        key.inplace                        = ggml_webgpu_tensor_equal(context.src0, context.dst);
 
         auto it = scale_pipelines.find(key);
         if (it != scale_pipelines.end()) {
@@ -1372,6 +1421,7 @@ class ggml_webgpu_shader_lib {
         auto processed           = preprocessor.preprocess(wgsl_scale, defines);
         auto decisions           = std::make_shared<ggml_webgpu_generic_shader_decisions>();
         decisions->wg_size       = context.max_wg_size;
+        decisions->inplace       = key.inplace;
         webgpu_pipeline pipeline = ggml_webgpu_create_pipeline(device, processed, variant);
         pipeline.context         = decisions;
         scale_pipelines[key]     = pipeline;
@@ -1465,6 +1515,8 @@ class ggml_webgpu_shader_lib {
         ggml_webgpu_ssm_scan_pipeline_key key = {};
         key.type                              = context.dst->type;
         key.d_state                           = (int) context.src0->ne[0];
+        key.xbc_overlap                       = ggml_webgpu_tensor_overlap(context.src1, context.src4) &&
+                          ggml_webgpu_tensor_overlap(context.src1, context.src5);
 
         auto it = ssm_scan_pipelines.find(key);
         if (it != ssm_scan_pipelines.end()) {
@@ -1496,12 +1548,17 @@ class ggml_webgpu_shader_lib {
             variant += "_wg_reduce";
         }
 
+        if (key.xbc_overlap) {
+            defines.push_back("XBC_OVERLAP");
+        }
+
         variant += "_d" + std::to_string(key.d_state);
 
         auto processed             = preprocessor.preprocess(wgsl_ssm_scan, defines);
         auto decisions             = std::make_shared<ggml_webgpu_ssm_scan_shader_decisions>();
         decisions->wg_size         = wg_size;
         decisions->tokens_per_tile = tokens_per_tile;
+        decisions->xbc_overlap     = key.xbc_overlap;
         webgpu_pipeline pipeline   = ggml_webgpu_create_pipeline(device, processed, variant);
         pipeline.context           = decisions;
         ssm_scan_pipelines[key]    = pipeline;
@@ -1615,6 +1672,24 @@ class ggml_webgpu_shader_lib {
                     defines.push_back("MUL_ACC_" + type_upper);
                     defines.push_back("U32_DEQUANT_HELPERS");
                     defines.push_back("SRC0_INNER_TYPE=u32");
+                    switch (context.src0->type) {
+                        case GGML_TYPE_IQ1_S:
+                        case GGML_TYPE_IQ1_M:
+                        case GGML_TYPE_IQ2_S:
+                        case GGML_TYPE_IQ3_S:
+                        case GGML_TYPE_IQ4_NL:
+                        case GGML_TYPE_IQ4_XS:
+                            defines.push_back(type_upper + "_GRID");
+                            break;
+                        case GGML_TYPE_IQ2_XXS:
+                        case GGML_TYPE_IQ2_XS:
+                        case GGML_TYPE_IQ3_XXS:
+                            defines.push_back(type_upper + "_GRID");
+                            defines.push_back(type_upper + "_TABLES");
+                            break;
+                        default:
+                            break;
+                    }
                     break;
                 }
         }
@@ -1639,7 +1714,9 @@ class ggml_webgpu_shader_lib {
         uint32_t wg_size        = WEBGPU_MUL_MAT_VEC_WG_SIZE;
         uint32_t outputs_per_wg = WEBGPU_MUL_MAT_VEC_FLOAT_OUTPUTS_PER_WG;
 
-        if (key.src0_type >= GGML_TYPE_Q2_K) {
+        if (key.src0_type == GGML_TYPE_Q1_0) {
+            outputs_per_wg = WEBGPU_MUL_MAT_VEC_LEGACY_Q_OUTPUTS_PER_WG;
+        } else if (key.src0_type >= GGML_TYPE_Q2_K) {
             outputs_per_wg = WEBGPU_MUL_MAT_VEC_K_Q_OUTPUTS_PER_WG;
         } else if (key.src0_type >= GGML_TYPE_Q4_0) {
             outputs_per_wg = WEBGPU_MUL_MAT_VEC_LEGACY_Q_OUTPUTS_PER_WG;
@@ -1729,6 +1806,25 @@ class ggml_webgpu_shader_lib {
                     defines.push_back("U32_DEQUANT_HELPERS");
                     defines.push_back("SRC0_INNER_TYPE=u32");
 
+                    switch (context.src0->type) {
+                        case GGML_TYPE_IQ1_S:
+                        case GGML_TYPE_IQ1_M:
+                        case GGML_TYPE_IQ4_NL:
+                        case GGML_TYPE_IQ4_XS:
+                            defines.push_back(type_upper + "_GRID");
+                            break;
+                        case GGML_TYPE_IQ2_XXS:
+                        case GGML_TYPE_IQ2_XS:
+                        case GGML_TYPE_IQ2_S:
+                        case GGML_TYPE_IQ3_XXS:
+                        case GGML_TYPE_IQ3_S:
+                            defines.push_back(type_upper + "_GRID");
+                            defines.push_back(type_upper + "_TABLES");
+                            break;
+                        default:
+                            break;
+                    }
+
                     variant += std::string("_") + src0_name;
                     break;
                 }
@@ -1741,11 +1837,9 @@ class ggml_webgpu_shader_lib {
 
         uint32_t tile_k;
         if (key.use_subgroup_matrix) {
-            tile_k = is_quant ? WEBGPU_MUL_MAT_SUBGROUP_TILE_K_QUANT
-                              : WEBGPU_MUL_MAT_SUBGROUP_TILE_K_FLOAT;
+            tile_k = is_quant ? WEBGPU_MUL_MAT_SUBGROUP_TILE_K_QUANT : WEBGPU_MUL_MAT_SUBGROUP_TILE_K_FLOAT;
         } else {
-            tile_k = is_quant ? WEBGPU_MUL_MAT_REG_TILE_K_QUANT
-                              : WEBGPU_MUL_MAT_REG_TILE_K_FLOAT;
+            tile_k = is_quant ? WEBGPU_MUL_MAT_REG_TILE_K_QUANT : WEBGPU_MUL_MAT_REG_TILE_K_FLOAT;
         }
 
         // Tiles
@@ -1978,9 +2072,8 @@ class ggml_webgpu_shader_lib {
         defines.push_back("SCALAR");
 
         // mul_mat_id is register-tile only.
-        const uint32_t tile_k = ggml_is_quantized(context.src0->type)
-                                    ? WEBGPU_MUL_MAT_REG_TILE_K_QUANT
-                                    : WEBGPU_MUL_MAT_REG_TILE_K_FLOAT;
+        const uint32_t tile_k =
+            ggml_is_quantized(context.src0->type) ? WEBGPU_MUL_MAT_REG_TILE_K_QUANT : WEBGPU_MUL_MAT_REG_TILE_K_FLOAT;
 
         // Tiles
         defines.push_back("TILE_M=" + std::to_string(WEBGPU_MUL_MAT_TILE_M) + "u");
@@ -2016,8 +2109,8 @@ class ggml_webgpu_shader_lib {
         key.type                                = context.dst->type;
         key.op                                  = op;
         key.is_unary                            = is_unary;
-        key.inplace                             = context.inplace;
-        key.ttype                               = (ggml_tri_type) ggml_get_op_params_i32(context.dst, 0);
+        key.inplace = ggml_webgpu_tensor_equal(context.src0, context.dst) || context.dst->op == GGML_OP_FILL;
+        key.ttype   = (ggml_tri_type) ggml_get_op_params_i32(context.dst, 0);
 
         auto it = unary_pipelines.find(key);
         if (it != unary_pipelines.end()) {
@@ -2075,6 +2168,7 @@ class ggml_webgpu_shader_lib {
         auto processed           = preprocessor.preprocess(wgsl_unary, defines);
         auto decisions           = std::make_shared<ggml_webgpu_generic_shader_decisions>();
         decisions->wg_size       = context.max_wg_size;
+        decisions->inplace       = key.inplace;
         webgpu_pipeline pipeline = ggml_webgpu_create_pipeline(device, processed, variant);
         pipeline.context         = decisions;
         unary_pipelines[key]     = pipeline;
@@ -2083,9 +2177,9 @@ class ggml_webgpu_shader_lib {
 
     webgpu_pipeline get_rms_norm_mul_pipeline(const ggml_webgpu_shader_lib_context & context) {
         ggml_webgpu_rms_norm_mul_pipeline_key key = {};
-        key.inplace                               = context.inplace;
-        key.overlap                               = context.overlap;
-        key.src_overlap                           = context.src_overlap;
+        key.inplace                               = ggml_webgpu_tensor_equal(context.src0, context.dst);
+        key.overlap                               = ggml_webgpu_tensor_equal(context.src1, context.dst);
+        key.src_overlap                           = ggml_webgpu_tensor_overlap(context.src0, context.src1);
 
         auto it = rms_norm_mul_pipelines.find(key);
         if (it != rms_norm_mul_pipelines.end()) {
@@ -2109,12 +2203,15 @@ class ggml_webgpu_shader_lib {
 
         defines.push_back(std::string("WG_SIZE=") + std::to_string(context.max_wg_size));
 
-        auto processed              = preprocessor.preprocess(wgsl_rms_norm_mul, defines);
-        auto decisions              = std::make_shared<ggml_webgpu_generic_shader_decisions>();
-        decisions->wg_size          = context.max_wg_size;
-        webgpu_pipeline pipeline    = ggml_webgpu_create_pipeline(device, processed, variant);
-        pipeline.context            = decisions;
-        rms_norm_mul_pipelines[key] = pipeline;
+        auto processed                  = preprocessor.preprocess(wgsl_rms_norm_mul, defines);
+        auto pipeline_decisions         = std::make_shared<ggml_webgpu_rms_norm_mul_shader_decisions>();
+        pipeline_decisions->wg_size     = context.max_wg_size;
+        pipeline_decisions->inplace     = key.inplace;
+        pipeline_decisions->overlap     = key.overlap;
+        pipeline_decisions->src_overlap = key.src_overlap;
+        webgpu_pipeline pipeline        = ggml_webgpu_create_pipeline(device, processed, variant);
+        pipeline.context                = pipeline_decisions;
+        rms_norm_mul_pipelines[key]     = pipeline;
         return rms_norm_mul_pipelines[key];
     }
 
@@ -2122,9 +2219,9 @@ class ggml_webgpu_shader_lib {
         ggml_webgpu_binary_pipeline_key key = {};
         key.type                            = context.dst->type;
         key.op                              = context.dst->op;
-        key.inplace                         = context.inplace;
-        key.overlap                         = context.overlap;
-        key.src_overlap                     = context.src_overlap;
+        key.inplace                         = ggml_webgpu_tensor_equal(context.src0, context.dst);
+        key.overlap                         = ggml_webgpu_tensor_equal(context.src1, context.dst);
+        key.src_overlap                     = ggml_webgpu_tensor_overlap(context.src0, context.src1);
 
         auto it = binary_pipelines.find(key);
         if (it != binary_pipelines.end()) {
@@ -2163,11 +2260,15 @@ class ggml_webgpu_shader_lib {
 
         defines.push_back(std::string("WG_SIZE=") + std::to_string(context.max_wg_size));
 
-        auto processed           = preprocessor.preprocess(wgsl_binary, defines);
-        auto decisions           = std::make_shared<ggml_webgpu_generic_shader_decisions>();
-        decisions->wg_size       = context.max_wg_size;
+        auto processed                  = preprocessor.preprocess(wgsl_binary, defines);
+        auto pipeline_decisions         = std::make_shared<ggml_webgpu_binary_shader_decisions>();
+        pipeline_decisions->wg_size     = context.max_wg_size;
+        pipeline_decisions->inplace     = key.inplace;
+        pipeline_decisions->overlap     = key.overlap;
+        pipeline_decisions->src_overlap = key.src_overlap;
+
         webgpu_pipeline pipeline = ggml_webgpu_create_pipeline(device, processed, variant);
-        pipeline.context         = decisions;
+        pipeline.context         = pipeline_decisions;
         binary_pipelines[key]    = pipeline;
         return binary_pipelines[key];
     }
@@ -2252,6 +2353,7 @@ class ggml_webgpu_shader_lib {
                                             size_t                                 storage_offset_alignment) {
         const ggml_webgpu_flash_attn_decisions decisions =
             ggml_webgpu_flash_attn_get_decisions(context, storage_offset_alignment);
+        GGML_ASSERT(decisions.path != GGML_WEBGPU_FLASH_ATTN_PATH_NONE);
         ggml_webgpu_flash_attn_pipeline_key key = ggml_webgpu_flash_attn_make_pipeline_key(context, decisions.path);
         auto                                it  = flash_attn_pipelines.find(key);
         if (it != flash_attn_pipelines.end()) {
@@ -2328,7 +2430,8 @@ class ggml_webgpu_shader_lib {
             defines.push_back(std::string("SG_MAT_K=") + std::to_string(context.sg_mat_k));
         }
 
-        auto pipeline_decisions = std::make_shared<ggml_webgpu_flash_attn_decisions>(decisions);
+        auto pipeline_decisions        = std::make_shared<ggml_webgpu_flash_attn_decisions>(decisions);
+        pipeline_decisions->kv_overlap = key.kv_overlap;
         defines.push_back(std::string("Q_TILE=") + std::to_string(decisions.q_tile));
         defines.push_back(std::string("KV_TILE=") + std::to_string(decisions.kv_tile));
         defines.push_back(std::string("WG_SIZE=") + std::to_string(decisions.wg_size));
@@ -2520,7 +2623,7 @@ class ggml_webgpu_shader_lib {
     webgpu_pipeline get_rope_pipeline(const ggml_webgpu_shader_lib_context & context) {
         ggml_webgpu_rope_pipeline_key key = {};
         key.type                          = context.dst->type;
-        key.inplace                       = context.inplace;
+        key.inplace                       = ggml_webgpu_tensor_equal(context.src0, context.dst);
         key.has_ff                        = (context.src2 != nullptr);
 
         auto it = rope_pipelines.find(key);
@@ -2559,6 +2662,7 @@ class ggml_webgpu_shader_lib {
         auto processed           = preprocessor.preprocess(wgsl_rope, defines);
         auto decisions           = std::make_shared<ggml_webgpu_generic_shader_decisions>();
         decisions->wg_size       = context.max_wg_size;
+        decisions->inplace       = key.inplace;
         webgpu_pipeline pipeline = ggml_webgpu_create_pipeline(device, processed, variant);
         pipeline.context         = decisions;
         rope_pipelines[key]      = pipeline;
@@ -2570,7 +2674,7 @@ class ggml_webgpu_shader_lib {
         key.mask_type                         = context.src1 ? context.src1->type : GGML_TYPE_F32;
         key.has_mask                          = (context.src1 != nullptr);
         key.has_sink                          = (context.src2 != nullptr);
-        key.inplace                           = context.inplace;
+        key.inplace                           = ggml_webgpu_tensor_equal(context.src0, context.dst);
 
         auto it = soft_max_pipelines.find(key);
         if (it != soft_max_pipelines.end()) {
@@ -2611,6 +2715,7 @@ class ggml_webgpu_shader_lib {
         auto processed           = preprocessor.preprocess(wgsl_soft_max, defines);
         auto decisions           = std::make_shared<ggml_webgpu_generic_shader_decisions>();
         decisions->wg_size       = context.max_wg_size;
+        decisions->inplace       = key.inplace;
         webgpu_pipeline pipeline = ggml_webgpu_create_pipeline(device, processed, variant);
         pipeline.context         = decisions;
         soft_max_pipelines[key]  = pipeline;

ggml/src/ggml-webgpu/ggml-webgpu.cpp

@@ -108,12 +108,9 @@ static inline uint32_t ggml_webgpu_u32_from_f32(float value) {
 // their locations.
 static void * const webgpu_ptr_base = (void *) (uintptr_t) 0x1000;  // NOLINT
 
-// Always returns the base offset of a tensor, regardless of views.
-static uint64_t webgpu_tensor_offset(const ggml_tensor * tensor) {
-    if (tensor->view_src) {
-        return (uint8_t *) tensor->view_src->data - (uint8_t *) webgpu_ptr_base;
-    }
-    return (uint8_t *) tensor->data - (uint8_t *) webgpu_ptr_base;
+static size_t ggml_webgpu_tensor_offset(const ggml_tensor * tensor) {
+    const ggml_tensor * base_tensor = tensor->view_src ? tensor->view_src : tensor;
+    return (size_t) ((uintptr_t) base_tensor->data - (uintptr_t) webgpu_ptr_base) + tensor->view_offs;
 }
 
 /* Struct definitions */
@@ -375,10 +372,6 @@ static void ggml_webgpu_create_buffer(wgpu::Device &    device,
     buffer = device.CreateBuffer(&buffer_desc);
 }
 
-static size_t ggml_webgpu_tensor_offset(const ggml_tensor * tensor) {
-    return webgpu_tensor_offset(tensor) + tensor->view_offs;
-}
-
 static wgpu::Buffer ggml_webgpu_tensor_buf(const ggml_tensor * tensor) {
     ggml_backend_webgpu_buffer_context * ctx = (ggml_backend_webgpu_buffer_context *) tensor->buffer->context;
     return ctx->buffer;
@@ -398,34 +391,31 @@ static size_t ggml_webgpu_tensor_binding_size(webgpu_context & ctx, ggml_tensor
     return ROUNDUP_POW2(ggml_nbytes(t) + ggml_webgpu_tensor_misalignment(ctx, t), WEBGPU_STORAGE_BUF_BINDING_MULT);
 }
 
-// Used to determine if two tensors are the same for in-place operations
-static bool ggml_webgpu_tensor_equal(ggml_tensor * a, ggml_tensor * b) {
-    return (ggml_webgpu_tensor_buf(a).Get() == ggml_webgpu_tensor_buf(b).Get()) &&
-           (ggml_webgpu_tensor_offset(a) == ggml_webgpu_tensor_offset(b));
-}
-
-// Used to determine if two tensors share the same buffer and their byte ranges overlap,
-static bool ggml_webgpu_tensor_overlap(ggml_tensor * a, ggml_tensor * b) {
-    return (ggml_webgpu_tensor_buf(a).Get() == ggml_webgpu_tensor_buf(b).Get()) &&
-           ggml_webgpu_tensor_offset(a) < (ggml_webgpu_tensor_offset(b) + ggml_nbytes(b)) &&
-           ggml_webgpu_tensor_offset(b) < (ggml_webgpu_tensor_offset(a) + ggml_nbytes(a));
-}
-
-struct binary_overlap_flags {
-    bool inplace;  // src0 == dst
-    bool overlap;  // src1 == dst
-    bool src_overlap;
+struct ggml_webgpu_merged_binding_range {
+    size_t offset;
+    size_t size;
 };
 
-static binary_overlap_flags ggml_webgpu_detect_binary_overlap(ggml_tensor * src0,
-                                                              ggml_tensor * src1,
-                                                              ggml_tensor * dst) {
-    binary_overlap_flags flags = {};
-    flags.inplace              = ggml_webgpu_tensor_equal(src0, dst);
-    flags.overlap              = ggml_webgpu_tensor_overlap(src1, dst);
-    flags.src_overlap          = ggml_webgpu_tensor_overlap(src0, src1);
+static ggml_webgpu_merged_binding_range ggml_webgpu_tensor_merged_binding_range(
+    webgpu_context &                     ctx,
+    std::initializer_list<ggml_tensor *> tensors) {
+    size_t merged_offset = SIZE_MAX;
+    size_t merged_end    = 0;
 
-    return flags;
+    for (ggml_tensor * tensor : tensors) {
+        const size_t bind_offset = ggml_webgpu_tensor_align_offset(ctx, tensor);
+        const size_t bind_end    = bind_offset + ggml_webgpu_tensor_binding_size(ctx, tensor);
+
+        merged_offset = std::min(merged_offset, bind_offset);
+        merged_end    = std::max(merged_end, bind_end);
+    }
+
+    return { merged_offset, merged_end - merged_offset };
+}
+
+static uint32_t ggml_webgpu_tensor_merged_element_offset(const ggml_tensor *                      tensor,
+                                                         const ggml_webgpu_merged_binding_range & merged_range) {
+    return (uint32_t) ((ggml_webgpu_tensor_offset(tensor) - merged_range.offset) / ggml_type_size(tensor->type));
 }
 
 static wgpu::BindGroupEntry ggml_webgpu_make_bind_group_entry(uint32_t     binding,
@@ -753,18 +743,16 @@ static webgpu_encoded_op ggml_webgpu_set(webgpu_context & ctx,
                                          ggml_tensor *    src0,
                                          ggml_tensor *    src1,
                                          ggml_tensor *    dst) {
-    const bool inplace = ggml_webgpu_tensor_equal(src0, dst);
-
     ggml_webgpu_shader_lib_context shader_lib_ctx = {};
     shader_lib_ctx.src0                           = src0;
     shader_lib_ctx.src1                           = src1;
     shader_lib_ctx.dst                            = dst;
     shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.inplace     = inplace;
 
     webgpu_pipeline pipeline = ctx->shader_lib->get_set_pipeline(shader_lib_ctx);
 
-    auto * decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
+    auto *     decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
+    const bool inplace   = decisions->inplace;
 
     const uint32_t ne            = inplace ? (uint32_t) ggml_nelements(src1) : (uint32_t) ggml_nelements(dst);
     const uint32_t dst_type_size = (uint32_t) ggml_type_size(dst->type);
@@ -1126,19 +1114,39 @@ static webgpu_encoded_op ggml_webgpu_ssm_scan(webgpu_context & ctx,
                                               ggml_tensor *    dst) {
     ggml_webgpu_shader_lib_context shader_lib_ctx = {};
     shader_lib_ctx.src0                           = src0;
+    shader_lib_ctx.src1                           = src1;
+    shader_lib_ctx.src4                           = src4;
+    shader_lib_ctx.src5                           = src5;
     shader_lib_ctx.dst                            = dst;
     shader_lib_ctx.max_wg_size        = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
     shader_lib_ctx.supports_subgroups = ctx->global_ctx->capabilities.supports_subgroups;
 
-    webgpu_pipeline pipeline = ctx->shader_lib->get_ssm_scan_pipeline(shader_lib_ctx);
+    webgpu_pipeline pipeline    = ctx->shader_lib->get_ssm_scan_pipeline(shader_lib_ctx);
+    auto *          decisions   = static_cast<ggml_webgpu_ssm_scan_shader_decisions *>(pipeline.context.get());
+    const bool      xbc_overlap = decisions->xbc_overlap;
+
+    uint32_t offset_x        = (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src1) / ggml_type_size(src1->type));
+    uint32_t offset_B        = (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src4) / ggml_type_size(src4->type));
+    uint32_t offset_C        = (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src5) / ggml_type_size(src5->type));
+    size_t   xbc_bind_offset = 0;
+    size_t   xbc_bind_size   = 0;
+    if (xbc_overlap) {
+        const ggml_webgpu_merged_binding_range merged_range =
+            ggml_webgpu_tensor_merged_binding_range(ctx, { src1, src4, src5 });
+        xbc_bind_offset = merged_range.offset;
+        xbc_bind_size   = merged_range.size;
+        offset_x        = ggml_webgpu_tensor_merged_element_offset(src1, merged_range);
+        offset_B        = ggml_webgpu_tensor_merged_element_offset(src4, merged_range);
+        offset_C        = ggml_webgpu_tensor_merged_element_offset(src5, merged_range);
+    }
 
     std::vector<uint32_t> params = {
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src0) / ggml_type_size(src0->type)),
-        (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src1) / ggml_type_size(src1->type)),
+        offset_x,
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src2) / ggml_type_size(src2->type)),
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src3) / ggml_type_size(src3->type)),
-        (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src4) / ggml_type_size(src4->type)),
-        (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src5) / ggml_type_size(src5->type)),
+        offset_B,
+        offset_C,
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src6) / ggml_type_size(src6->type)),
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, dst) / ggml_type_size(dst->type)),
 
@@ -1174,11 +1182,24 @@ static webgpu_encoded_op ggml_webgpu_ssm_scan(webgpu_context & ctx,
     };
 
     std::vector<wgpu::BindGroupEntry> entries = {
-        ggml_webgpu_make_tensor_bind_group_entry(ctx, 0, src0), ggml_webgpu_make_tensor_bind_group_entry(ctx, 1, src1),
-        ggml_webgpu_make_tensor_bind_group_entry(ctx, 2, src2), ggml_webgpu_make_tensor_bind_group_entry(ctx, 3, src3),
-        ggml_webgpu_make_tensor_bind_group_entry(ctx, 4, src4), ggml_webgpu_make_tensor_bind_group_entry(ctx, 5, src5),
-        ggml_webgpu_make_tensor_bind_group_entry(ctx, 6, src6), ggml_webgpu_make_tensor_bind_group_entry(ctx, 7, dst),
+        ggml_webgpu_make_tensor_bind_group_entry(ctx, 0, src0),
     };
+    if (xbc_overlap) {
+        entries.push_back(
+            ggml_webgpu_make_bind_group_entry(1, ggml_webgpu_tensor_buf(src1), xbc_bind_offset, xbc_bind_size));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 2, src2));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 3, src3));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 4, src6));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 5, dst));
+    } else {
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 1, src1));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 2, src2));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 3, src3));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 4, src4));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 5, src5));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 6, src6));
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 7, dst));
+    }
 
     const uint32_t total_wg       = (uint32_t) (src0->ne[1] * src0->ne[2] * src1->ne[3]);
     const uint32_t max_wg_per_dim = ctx->global_ctx->capabilities.limits.maxComputeWorkgroupsPerDimension;
@@ -1389,6 +1410,18 @@ static webgpu_encoded_op ggml_webgpu_mul_mat(webgpu_context & ctx,
                 case GGML_TYPE_Q5_K:
                 case GGML_TYPE_Q3_K:
                 case GGML_TYPE_Q2_K:
+                case GGML_TYPE_Q1_0:
+                    use_fast = true;
+                    break;
+                case GGML_TYPE_IQ1_S:
+                case GGML_TYPE_IQ1_M:
+                case GGML_TYPE_IQ2_XXS:
+                case GGML_TYPE_IQ2_XS:
+                case GGML_TYPE_IQ2_S:
+                case GGML_TYPE_IQ3_XXS:
+                case GGML_TYPE_IQ3_S:
+                case GGML_TYPE_IQ4_NL:
+                case GGML_TYPE_IQ4_XS:
                     use_fast = true;
                     break;
                 default:
@@ -1641,23 +1674,38 @@ static webgpu_encoded_op ggml_webgpu_flash_attn(webgpu_context & ctx,
     float m0          = powf(2.0f, -(max_bias) / n_head_log2);
     float m1          = powf(2.0f, -(max_bias / 2.0f) / n_head_log2);
 
+    ggml_webgpu_shader_lib_context shader_lib_ctx = {};
+    shader_lib_ctx.src0                           = Q;
+    shader_lib_ctx.src1                           = K;
+    shader_lib_ctx.src2                           = V;
+    shader_lib_ctx.src3                           = mask;
+    shader_lib_ctx.src4                           = sinks;
+    shader_lib_ctx.dst                            = dst;
+    shader_lib_ctx.supports_subgroups             = ctx->global_ctx->capabilities.supports_subgroups;
+    shader_lib_ctx.supports_subgroup_matrix       = ctx->global_ctx->capabilities.supports_subgroup_matrix;
+    shader_lib_ctx.max_wg_size        = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
+    shader_lib_ctx.wg_mem_limit_bytes = ctx->global_ctx->capabilities.limits.maxComputeWorkgroupStorageSize;
+    shader_lib_ctx.sg_mat_m           = ctx->global_ctx->capabilities.sg_mat_m;
+    shader_lib_ctx.sg_mat_n           = ctx->global_ctx->capabilities.sg_mat_n;
+    shader_lib_ctx.sg_mat_k           = ctx->global_ctx->capabilities.sg_mat_k;
+    shader_lib_ctx.max_subgroup_size  = ctx->global_ctx->capabilities.max_subgroup_size;
+    webgpu_pipeline pipeline          = ctx->shader_lib->get_flash_attn_pipeline(
+        shader_lib_ctx, ctx->global_ctx->capabilities.limits.minStorageBufferOffsetAlignment);
+    auto *     decisions  = static_cast<ggml_webgpu_flash_attn_decisions *>(pipeline.context.get());
     const int  has_mask   = (mask != nullptr);
     const int  has_sinks  = (sinks != nullptr);
-    const bool kv_overlap = ggml_webgpu_tensor_overlap(K, V) && K->type == V->type;
+    const bool kv_overlap = decisions->kv_overlap;
 
     uint32_t offset_k       = (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, K) / ggml_type_size(K->type));
     uint32_t offset_v       = (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, V) / ggml_type_size(V->type));
     size_t   kv_bind_offset = 0;
     size_t   kv_bind_size   = 0;
     if (kv_overlap) {
-        const size_t k_bind_offset = ggml_webgpu_tensor_align_offset(ctx, K);
-        const size_t v_bind_offset = ggml_webgpu_tensor_align_offset(ctx, V);
-        const size_t k_bind_end    = k_bind_offset + ggml_webgpu_tensor_binding_size(ctx, K);
-        const size_t v_bind_end    = v_bind_offset + ggml_webgpu_tensor_binding_size(ctx, V);
-        kv_bind_offset             = std::min(k_bind_offset, v_bind_offset);
-        kv_bind_size               = std::max(k_bind_end, v_bind_end) - kv_bind_offset;
-        offset_k = (uint32_t) ((ggml_webgpu_tensor_offset(K) - kv_bind_offset) / ggml_type_size(K->type));
-        offset_v = (uint32_t) ((ggml_webgpu_tensor_offset(V) - kv_bind_offset) / ggml_type_size(V->type));
+        const ggml_webgpu_merged_binding_range merged_range = ggml_webgpu_tensor_merged_binding_range(ctx, { K, V });
+        kv_bind_offset                                      = merged_range.offset;
+        kv_bind_size                                        = merged_range.size;
+        offset_k                                            = ggml_webgpu_tensor_merged_element_offset(K, merged_range);
+        offset_v                                            = ggml_webgpu_tensor_merged_element_offset(V, merged_range);
     }
 
     std::vector<uint32_t> params = {
@@ -1708,26 +1756,6 @@ static webgpu_encoded_op ggml_webgpu_flash_attn(webgpu_context & ctx,
     }
     entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, binding_index++, dst));
 
-    ggml_webgpu_shader_lib_context shader_lib_ctx = {};
-    shader_lib_ctx.src0                           = Q;
-    shader_lib_ctx.src1                           = K;
-    shader_lib_ctx.src2                           = V;
-    shader_lib_ctx.src3                           = mask;
-    shader_lib_ctx.src4                           = sinks;
-    shader_lib_ctx.dst                            = dst;
-    shader_lib_ctx.src_overlap                    = kv_overlap;
-    shader_lib_ctx.supports_subgroups             = ctx->global_ctx->capabilities.supports_subgroups;
-    shader_lib_ctx.supports_subgroup_matrix       = ctx->global_ctx->capabilities.supports_subgroup_matrix;
-    shader_lib_ctx.max_wg_size        = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.wg_mem_limit_bytes = ctx->global_ctx->capabilities.limits.maxComputeWorkgroupStorageSize;
-    shader_lib_ctx.sg_mat_m           = ctx->global_ctx->capabilities.sg_mat_m;
-    shader_lib_ctx.sg_mat_n           = ctx->global_ctx->capabilities.sg_mat_n;
-    shader_lib_ctx.sg_mat_k           = ctx->global_ctx->capabilities.sg_mat_k;
-    shader_lib_ctx.max_subgroup_size  = ctx->global_ctx->capabilities.max_subgroup_size;
-    webgpu_pipeline pipeline          = ctx->shader_lib->get_flash_attn_pipeline(
-        shader_lib_ctx, ctx->global_ctx->capabilities.limits.minStorageBufferOffsetAlignment);
-    auto * decisions = static_cast<ggml_webgpu_flash_attn_decisions *>(pipeline.context.get());
-
     if (decisions->path != GGML_WEBGPU_FLASH_ATTN_PATH_VEC) {
         uint32_t wg_per_head = CEIL_DIV(Q->ne[1], decisions->q_tile);
         uint32_t wg_x        = wg_per_head * Q->ne[2] * Q->ne[3];  // wg per head * number of heads * number of batches
@@ -1909,18 +1937,17 @@ static webgpu_encoded_op ggml_webgpu_flash_attn(webgpu_context & ctx,
 
 static webgpu_encoded_op ggml_webgpu_unary_op(webgpu_context & ctx, ggml_tensor * src, ggml_tensor * dst) {
     bool is_unary = dst->op == GGML_OP_UNARY;
-    bool inplace  = ggml_webgpu_tensor_equal(src, dst) || (dst->op == GGML_OP_FILL);
 
     ggml_webgpu_shader_lib_context shader_lib_ctx = {};
     shader_lib_ctx.src0                           = src;
     shader_lib_ctx.src1                           = nullptr;
     shader_lib_ctx.dst                            = dst;
     shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.inplace     = inplace;
 
     webgpu_pipeline pipeline = ctx->shader_lib->get_unary_pipeline(shader_lib_ctx);
 
-    auto * decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
+    auto *     decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
+    const bool inplace   = decisions->inplace;
 
     uint32_t ne = (uint32_t) ggml_nelements(dst);
 
@@ -1982,41 +2009,38 @@ static webgpu_encoded_op ggml_webgpu_binary_op(webgpu_context & ctx,
                                                ggml_tensor *    src0,
                                                ggml_tensor *    src1,
                                                ggml_tensor *    dst) {
-    binary_overlap_flags flags = ggml_webgpu_detect_binary_overlap(src0, src1, dst);
-
     ggml_webgpu_shader_lib_context shader_lib_ctx = {};
     shader_lib_ctx.src0                           = src0;
     shader_lib_ctx.src1                           = src1;
     shader_lib_ctx.dst                            = dst;
     shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.inplace     = flags.inplace;
-    shader_lib_ctx.overlap     = flags.overlap;
-    shader_lib_ctx.src_overlap = flags.src_overlap;
 
-    webgpu_pipeline pipeline = ctx->shader_lib->get_binary_pipeline(shader_lib_ctx);
-
-    auto * decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
+    webgpu_pipeline pipeline  = ctx->shader_lib->get_binary_pipeline(shader_lib_ctx);
+    auto *          decisions = static_cast<ggml_webgpu_binary_shader_decisions *>(pipeline.context.get());
 
     uint32_t ne = (uint32_t) ggml_nelements(dst);
 
     size_t src0_webgpu_tensor_align_offset = ggml_webgpu_tensor_align_offset(ctx, src0);
     size_t src1_webgpu_tensor_align_offset = ggml_webgpu_tensor_align_offset(ctx, src1);
 
-    uint32_t offset_merged_src0 = 0;
-    uint32_t offset_merged_src1 = 0;
-    if (flags.src_overlap) {
-        size_t min_off     = std::min(src0_webgpu_tensor_align_offset, src1_webgpu_tensor_align_offset);
-        offset_merged_src0 = (uint32_t) ((src0_webgpu_tensor_align_offset - min_off) / ggml_type_size(src0->type));
-        offset_merged_src1 = (uint32_t) ((src1_webgpu_tensor_align_offset - min_off) / ggml_type_size(src0->type));
+    uint32_t offset_src0   = (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src0) / ggml_type_size(src0->type));
+    uint32_t offset_src1   = (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src1) / ggml_type_size(src1->type));
+    size_t   merged_offset = 0;
+    size_t   merged_size   = 0;
+    if (decisions->src_overlap) {
+        const ggml_webgpu_merged_binding_range merged_range =
+            ggml_webgpu_tensor_merged_binding_range(ctx, { src0, src1 });
+        merged_offset = merged_range.offset;
+        merged_size   = merged_range.size;
+        offset_src0   = ggml_webgpu_tensor_merged_element_offset(src0, merged_range);
+        offset_src1   = ggml_webgpu_tensor_merged_element_offset(src1, merged_range);
     }
 
     std::vector<uint32_t> params = {
         ne,
-        (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src0) / ggml_type_size(src0->type)),
-        (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src1) / ggml_type_size(src1->type)),
+        offset_src0,
+        offset_src1,
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, dst) / ggml_type_size(dst->type)),
-        offset_merged_src0,
-        offset_merged_src1,
         (uint32_t) (src0->nb[0] / ggml_type_size(src0->type)),
         (uint32_t) (src0->nb[1] / ggml_type_size(src0->type)),
         (uint32_t) (src0->nb[2] / ggml_type_size(src0->type)),
@@ -2036,12 +2060,9 @@ static webgpu_encoded_op ggml_webgpu_binary_op(webgpu_context & ctx,
 
     std::vector<wgpu::BindGroupEntry> entries;
 
-    if (flags.src_overlap) {
-        size_t merged_offset = std::min(src0_webgpu_tensor_align_offset, src1_webgpu_tensor_align_offset);
-        size_t merged_end    = std::max(src0_webgpu_tensor_align_offset + ggml_webgpu_tensor_binding_size(ctx, src0),
-                                        src1_webgpu_tensor_align_offset + ggml_webgpu_tensor_binding_size(ctx, src1));
-        entries.push_back(ggml_webgpu_make_bind_group_entry(0, ggml_webgpu_tensor_buf(src0), merged_offset,
-                                                            merged_end - merged_offset));
+    if (decisions->src_overlap) {
+        entries.push_back(
+            ggml_webgpu_make_bind_group_entry(0, ggml_webgpu_tensor_buf(src0), merged_offset, merged_size));
         entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 1, dst));
     } else {
         entries.push_back(ggml_webgpu_make_bind_group_entry(0, ggml_webgpu_tensor_buf(src0),
@@ -2050,7 +2071,7 @@ static webgpu_encoded_op ggml_webgpu_binary_op(webgpu_context & ctx,
         entries.push_back(ggml_webgpu_make_bind_group_entry(1, ggml_webgpu_tensor_buf(src1),
                                                             src1_webgpu_tensor_align_offset,
                                                             ggml_webgpu_tensor_binding_size(ctx, src1)));
-        if (!flags.inplace && !flags.overlap) {
+        if (!decisions->inplace && !decisions->overlap) {
             entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 2, dst));
         }
     }
@@ -2156,29 +2177,15 @@ static std::optional<webgpu_encoded_op> ggml_webgpu_rms_norm_mul(webgpu_context
         GGML_ABORT("rms_norm must be equal to the one of mul_src0 and mul_src1");
     }
 
-    bool overlap = (ggml_webgpu_tensor_equal(rn_dst, mul_src0) && ggml_webgpu_tensor_equal(mul_src1, dst)) ||
-                   (ggml_webgpu_tensor_equal(rn_dst, mul_src1) && ggml_webgpu_tensor_equal(mul_src0, dst));
-    bool inplace     = ggml_webgpu_tensor_equal(rn_src, dst);
-    bool src_overlap = ggml_webgpu_tensor_overlap(rn_src, mul_src);
-
-    uint32_t offset_merged_rn_src               = 0;
-    uint32_t offset_merged_mul_src              = 0;
-    size_t   rn_src_webgpu_tensor_align_offset  = ggml_webgpu_tensor_align_offset(ctx, rn_src);
-    size_t   mul_src_webgpu_tensor_align_offset = ggml_webgpu_tensor_align_offset(ctx, mul_src);
-
-    if (src_overlap) {
-        size_t min_offset = std::min(rn_src_webgpu_tensor_align_offset, mul_src_webgpu_tensor_align_offset);
-        offset_merged_rn_src =
-            (uint32_t) ((rn_src_webgpu_tensor_align_offset - min_offset) / ggml_type_size(rn_src->type));
-        offset_merged_mul_src =
-            (uint32_t) ((mul_src_webgpu_tensor_align_offset - min_offset) / ggml_type_size(mul_src->type));
-    }
+    uint32_t offset_rn_src = (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, rn_src) / ggml_type_size(rn_src->type));
+    uint32_t offset_mul_src =
+        (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, mul_src) / ggml_type_size(mul_src->type));
+    size_t merged_offset = 0;
+    size_t merged_size   = 0;
 
     std::vector<uint32_t> params = {
-        (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, rn_src) / ggml_type_size(rn_src->type)),
-        (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, mul_src) / ggml_type_size(mul_src->type)),
-        offset_merged_rn_src,
-        offset_merged_mul_src,
+        offset_rn_src,
+        offset_mul_src,
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, dst) / ggml_type_size(dst->type)),
         (uint32_t) (rn_src->nb[1] / ggml_type_size(rn_src->type)),
         (uint32_t) (rn_src->nb[2] / ggml_type_size(rn_src->type)),
@@ -2202,16 +2209,32 @@ static std::optional<webgpu_encoded_op> ggml_webgpu_rms_norm_mul(webgpu_context
 
     std::vector<wgpu::BindGroupEntry> entries;
 
-    if (inplace || overlap) {
+    ggml_webgpu_shader_lib_context shader_lib_ctx = {};
+    shader_lib_ctx.src0                           = rn_src;
+    shader_lib_ctx.src1                           = mul_src;
+    shader_lib_ctx.dst                            = dst;
+    shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
+
+    webgpu_pipeline pipeline  = ctx->shader_lib->get_rms_norm_mul_pipeline(shader_lib_ctx);
+    auto *          decisions = static_cast<ggml_webgpu_rms_norm_mul_shader_decisions *>(pipeline.context.get());
+
+    if (decisions->src_overlap) {
+        const ggml_webgpu_merged_binding_range merged_range =
+            ggml_webgpu_tensor_merged_binding_range(ctx, { rn_src, mul_src });
+        merged_offset  = merged_range.offset;
+        merged_size    = merged_range.size;
+        offset_rn_src  = ggml_webgpu_tensor_merged_element_offset(rn_src, merged_range);
+        offset_mul_src = ggml_webgpu_tensor_merged_element_offset(mul_src, merged_range);
+        params[0]      = offset_rn_src;
+        params[1]      = offset_mul_src;
+    }
+
+    if (decisions->inplace || decisions->overlap) {
         entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 0, rn_src));
         entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 1, mul_src));
-    } else if (src_overlap) {
-        size_t merged_offset = std::min(rn_src_webgpu_tensor_align_offset, mul_src_webgpu_tensor_align_offset);
-        size_t merged_end =
-            std::max(rn_src_webgpu_tensor_align_offset + ggml_webgpu_tensor_binding_size(ctx, rn_src),
-                     mul_src_webgpu_tensor_align_offset + ggml_webgpu_tensor_binding_size(ctx, mul_src));
-        entries.push_back(ggml_webgpu_make_bind_group_entry(0, ggml_webgpu_tensor_buf(rn_src), merged_offset,
-                                                            merged_end - merged_offset));
+    } else if (decisions->src_overlap) {
+        entries.push_back(
+            ggml_webgpu_make_bind_group_entry(0, ggml_webgpu_tensor_buf(rn_src), merged_offset, merged_size));
         entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 1, dst));
     } else {
         entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 0, rn_src));
@@ -2219,20 +2242,10 @@ static std::optional<webgpu_encoded_op> ggml_webgpu_rms_norm_mul(webgpu_context
         entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 2, dst));
     }
 
-    ggml_webgpu_shader_lib_context shader_lib_ctx = {};
-    shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.inplace     = inplace;
-    shader_lib_ctx.overlap     = overlap;
-    shader_lib_ctx.src_overlap = src_overlap;
-
-    webgpu_pipeline pipeline = ctx->shader_lib->get_rms_norm_mul_pipeline(shader_lib_ctx);
-
     return ggml_backend_webgpu_build(ctx, pipeline, params, entries, ggml_nrows(dst));
 }
 
 static webgpu_encoded_op ggml_webgpu_row_norm(webgpu_context & ctx, ggml_tensor * src, ggml_tensor * dst) {
-    bool inplace = ggml_webgpu_tensor_equal(src, dst);
-
     std::vector<uint32_t> params = {
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src) / ggml_type_size(src->type)),
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, dst) / ggml_type_size(dst->type)),
@@ -2249,18 +2262,18 @@ static webgpu_encoded_op ggml_webgpu_row_norm(webgpu_context & ctx, ggml_tensor
         ggml_webgpu_u32_from_f32(ggml_get_op_params_f32(dst, 0))  // epsilon, treated as f32 in the shader
     };
 
-    std::vector<wgpu::BindGroupEntry> entries = { ggml_webgpu_make_tensor_bind_group_entry(ctx, 0, src) };
-    if (!inplace) {
-        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 1, dst));
-    }
-
     ggml_webgpu_shader_lib_context shader_lib_ctx = {};
     shader_lib_ctx.src0                           = src;
     shader_lib_ctx.dst                            = dst;
     shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.inplace     = inplace;
 
-    webgpu_pipeline pipeline = ctx->shader_lib->get_row_norm_pipeline(shader_lib_ctx);
+    webgpu_pipeline pipeline  = ctx->shader_lib->get_row_norm_pipeline(shader_lib_ctx);
+    auto *          decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
+
+    std::vector<wgpu::BindGroupEntry> entries = { ggml_webgpu_make_tensor_bind_group_entry(ctx, 0, src) };
+    if (!decisions->inplace) {
+        entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 1, dst));
+    }
     return ggml_backend_webgpu_build(ctx, pipeline, params, entries, ggml_nrows(src));
 }
 
@@ -2275,14 +2288,13 @@ static webgpu_encoded_op ggml_webgpu_rope(webgpu_context & ctx,
     shader_lib_ctx.src2                           = src2;
     shader_lib_ctx.dst                            = dst;
     shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.inplace     = ggml_webgpu_tensor_equal(src0, dst);
 
     webgpu_pipeline pipeline = ctx->shader_lib->get_rope_pipeline(shader_lib_ctx);
 
     auto * decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
 
-    const int inplace         = ggml_webgpu_tensor_equal(src0, dst);
-    const int has_freq_factor = (src2 != nullptr);
+    const bool inplace         = decisions->inplace;
+    const int  has_freq_factor = (src2 != nullptr);
 
     const int n_dims     = ((int32_t *) dst->op_params)[1];
     const int mode       = ((int32_t *) dst->op_params)[2];
@@ -2409,14 +2421,11 @@ static webgpu_encoded_op ggml_webgpu_glu(webgpu_context & ctx,
 }
 
 static webgpu_encoded_op ggml_webgpu_scale(webgpu_context & ctx, ggml_tensor * src, ggml_tensor * dst) {
-    bool inplace = ggml_webgpu_tensor_equal(src, dst);
-
     ggml_webgpu_shader_lib_context shader_lib_ctx = {};
     shader_lib_ctx.src0                           = src;
     shader_lib_ctx.src1                           = nullptr;
     shader_lib_ctx.dst                            = dst;
     shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.inplace     = inplace;
 
     webgpu_pipeline pipeline  = ctx->shader_lib->get_scale_pipeline(shader_lib_ctx);
     auto *          decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
@@ -2442,7 +2451,7 @@ static webgpu_encoded_op ggml_webgpu_scale(webgpu_context & ctx, ggml_tensor * s
     // bindgroups unchanged
     std::vector<wgpu::BindGroupEntry> entries = { ggml_webgpu_make_tensor_bind_group_entry(ctx, 0, src) };
 
-    if (!inplace) {
+    if (!decisions->inplace) {
         entries.push_back(ggml_webgpu_make_tensor_bind_group_entry(ctx, 1, dst));
     }
 
@@ -2461,17 +2470,17 @@ static webgpu_encoded_op ggml_webgpu_soft_max(webgpu_context & ctx,
     shader_lib_ctx.src2                           = src2;
     shader_lib_ctx.dst                            = dst;
     shader_lib_ctx.max_wg_size = ctx->global_ctx->capabilities.limits.maxComputeInvocationsPerWorkgroup;
-    shader_lib_ctx.inplace     = ggml_webgpu_tensor_equal(src0, dst);
 
-    webgpu_pipeline pipeline = ctx->shader_lib->get_soft_max_pipeline(shader_lib_ctx);
+    webgpu_pipeline pipeline  = ctx->shader_lib->get_soft_max_pipeline(shader_lib_ctx);
+    auto *          decisions = static_cast<ggml_webgpu_generic_shader_decisions *>(pipeline.context.get());
 
-    const int inplace     = ggml_webgpu_tensor_equal(src0, dst);
-    const int has_mask    = (src1 != nullptr);
-    const int has_sink    = (src2 != nullptr);
-    float     max_bias    = ggml_get_op_params_f32(dst, 1);
-    float     n_head_log2 = float(1u << (uint32_t) floor(log2(src0->ne[2])));
-    float     m0          = powf(2.0f, -(max_bias) / n_head_log2);
-    float     m1          = powf(2.0f, -(max_bias / 2.0f) / n_head_log2);
+    const bool inplace     = decisions->inplace;
+    const int  has_mask    = (src1 != nullptr);
+    const int  has_sink    = (src2 != nullptr);
+    float      max_bias    = ggml_get_op_params_f32(dst, 1);
+    float      n_head_log2 = float(1u << (uint32_t) floor(log2(src0->ne[2])));
+    float      m0          = powf(2.0f, -(max_bias) / n_head_log2);
+    float      m1          = powf(2.0f, -(max_bias / 2.0f) / n_head_log2);
 
     std::vector<uint32_t> params = {
         (uint32_t) (ggml_webgpu_tensor_misalignment(ctx, src0) / ggml_type_size(src0->type)),
@@ -3067,7 +3076,7 @@ static void ggml_backend_webgpu_set_tensor_async(ggml_backend_t backend,
                                                  size_t         size) {
     GGML_UNUSED(backend);
     auto * buf_ctx      = (ggml_backend_webgpu_buffer_context *) tensor->buffer->context;
-    size_t total_offset = webgpu_tensor_offset(tensor) + tensor->view_offs + offset;
+    size_t total_offset = ggml_webgpu_tensor_offset(tensor) + offset;
 
     // Write aligned portion
     buf_ctx->global_ctx->queue.WriteBuffer(buf_ctx->buffer, total_offset, data, (size / 4) * 4);
@@ -3098,8 +3107,8 @@ static ggml_backend_i ggml_backend_webgpu_i = {
     /* .free                    = */ ggml_backend_webgpu_free,
     /* .set_tensor_async        = */ ggml_backend_webgpu_set_tensor_async,
     /* .get_tensor_async        = */ NULL,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ NULL,
     /* .synchronize             = */ ggml_backend_webgpu_synchronize,
     /* .graph_plan_create       = */ NULL,
@@ -3149,7 +3158,7 @@ static void ggml_backend_webgpu_buffer_memset_tensor(ggml_backend_buffer_t buffe
     WEBGPU_LOG_DEBUG("ggml_backend_webgpu_buffer_memset_tensor(" << buf_ctx->label << ", " << tensor << ", " << value
                                                                  << ", " << offset << ", " << size << ")");
 
-    size_t total_offset = webgpu_tensor_offset(tensor) + tensor->view_offs + offset;
+    size_t total_offset = ggml_webgpu_tensor_offset(tensor) + offset;
 
     // This is a trick to set all bytes of a u32 to the same 1 byte value.
     uint32_t val32 = (uint32_t) value * 0x01010101;
@@ -3168,7 +3177,7 @@ static void ggml_backend_webgpu_buffer_set_tensor(ggml_backend_buffer_t buffer,
     WEBGPU_LOG_DEBUG("ggml_backend_webgpu_buffer_set_tensor(" << buf_ctx->label << ", " << tensor << ", " << data
                                                               << ", " << offset << ", " << size << ")");
 
-    size_t total_offset = webgpu_tensor_offset(tensor) + tensor->view_offs + offset;
+    size_t total_offset = ggml_webgpu_tensor_offset(tensor) + offset;
 
     buf_ctx->global_ctx->queue.WriteBuffer(buf_ctx->buffer, total_offset, data, (size / 4) * 4);
 
@@ -3200,7 +3209,7 @@ static void ggml_backend_webgpu_buffer_get_tensor(ggml_backend_buffer_t buffer,
                                                               << ", " << offset << ", " << size << ")");
     wgpu::Device device = buf_ctx->global_ctx->device;
 
-    size_t total_offset = webgpu_tensor_offset(tensor) + tensor->view_offs + offset;
+    size_t total_offset = ggml_webgpu_tensor_offset(tensor) + offset;
 
     size_t final_size = size;
     if (size % 4 != 0) {
@@ -3725,6 +3734,7 @@ static bool ggml_backend_webgpu_device_supports_buft(ggml_backend_dev_t dev, ggm
 
 static bool ggml_webgpu_supported_qtype(ggml_type type) {
     switch (type) {
+        case GGML_TYPE_Q1_0:
         case GGML_TYPE_Q4_0:
         case GGML_TYPE_Q4_1:
         case GGML_TYPE_Q5_0:
@@ -3819,6 +3829,7 @@ static bool ggml_backend_webgpu_device_supports_op(ggml_backend_dev_t dev, const
                         switch (src0->type) {
                             case GGML_TYPE_F32:
                             case GGML_TYPE_F16:
+                            case GGML_TYPE_Q1_0:
                             case GGML_TYPE_Q4_0:
                             case GGML_TYPE_Q4_1:
                             case GGML_TYPE_Q5_0:
@@ -3857,6 +3868,7 @@ static bool ggml_backend_webgpu_device_supports_op(ggml_backend_dev_t dev, const
                     switch (src0->type) {
                         case GGML_TYPE_F32:
                         case GGML_TYPE_F16:
+                        case GGML_TYPE_Q1_0:
                         case GGML_TYPE_Q4_0:
                         case GGML_TYPE_Q4_1:
                         case GGML_TYPE_Q5_0:
@@ -3906,6 +3918,10 @@ static bool ggml_backend_webgpu_device_supports_op(ggml_backend_dev_t dev, const
                     shader_lib_ctx, ctx->webgpu_global_ctx->capabilities.limits.minStorageBufferOffsetAlignment);
                 const size_t limit_bytes = ctx->webgpu_global_ctx->capabilities.limits.maxComputeWorkgroupStorageSize;
                 const bool   has_mask    = op->src[3] != nullptr;
+                if (decisions.path == GGML_WEBGPU_FLASH_ATTN_PATH_NONE) {
+                    supports_op = false;
+                    break;
+                }
                 if (decisions.path == GGML_WEBGPU_FLASH_ATTN_PATH_VEC) {
                     const size_t min_bytes =
                         ggml_webgpu_flash_attn_wg_mem_bytes(decisions.q_tile, decisions.kv_tile, (uint32_t) src0->ne[0],

ggml/src/ggml-webgpu/wgsl-shaders/binary.wgsl

@@ -7,8 +7,6 @@ struct Params {
     offset_src0: u32,
     offset_src1: u32,
     offset_dst: u32,
-    offset_merged_src0: u32,
-    offset_merged_src1: u32,
 
     stride_src0_0: u32,
     stride_src0_1: u32,
@@ -134,8 +132,8 @@ fn update(dst_i: u32, src0_i: u32, src1_i: u32) {
 @compute @workgroup_size(WG_SIZE)
 fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
     if (gid.x < params.ne) {
-        let src0_i = params.offset_src0 + params.offset_merged_src0 + src0_index(gid.x);
-        let src1_i = params.offset_src1 + params.offset_merged_src1 + src1_index(gid.x);
+        let src0_i = params.offset_src0 + src0_index(gid.x);
+        let src1_i = params.offset_src1 + src1_index(gid.x);
         update(params.offset_dst + gid.x, src0_i, src1_i);
     }
 }

ggml/src/ggml-webgpu/wgsl-shaders/get_rows.wgsl

@@ -27,6 +27,24 @@ fn copy_elements(src_base: u32, dst_base: u32, offset: u32) {
 }
 #endif
 
+#ifdef Q1_0
+fn copy_elements(src_base: u32, dst_base: u32, offset: u32) {
+    let block_byte_base = (src_base + offset) * 18;
+    let d = load_f16_as_f32_at_src(block_byte_base);
+    for (var j: u32 = 0u; j < 4u; j++) {
+        let q_packed = load_u32_at_src(block_byte_base + 2u + j * 4u);
+        let dst_base128 = dst_base + offset * 128u + j * 32u;
+        for (var k: u32 = 0; k < 4u; k++) {
+            let q_byte = get_byte(q_packed, k);
+            for (var bit: u32 = 0; bit < 8u; bit++) {
+                let w = select(-d, d, ((q_byte >> bit) & 1u) != 0u);
+                dst[dst_base128 + k * 8u + bit] = w;
+            }
+        }
+    }
+}
+#endif
+
 #ifdef Q4_0
 fn copy_elements(src_base: u32, dst_base: u32, offset: u32) {
     let block_byte_base = (src_base + offset) * 18; // Block stride: 18 bytes

ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_decls.tmpl

@@ -61,6 +61,39 @@ fn init_shmem_src1(thread_id: u32, batch_offset: u32, offset_n: u32, k_outer: u3
 #endif // INIT_SRC1_SHMEM_FLOAT
 #endif
 
+#ifdef INIT_SRC0_SHMEM_Q1_0
+const BLOCK_SIZE = 128u;
+const BLOCK_SIZE_BYTES = 18u;
+const NQ = 8u; // 8 weights (1 byte of qs) per thread per iteration
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var i = thread_id * NQ; i < TILE_SRC0_SHMEM; i += TOTAL_WORKGROUP_SIZE * NQ) {
+        let tile_m = i / TILE_K;
+        let tile_k_start = i % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k_start = k_outer + tile_k_start;
+
+        if (global_m >= params.m) {
+            break;
+        }
+
+        let block_k = global_k_start / BLOCK_SIZE;
+        let byte_in_block = (global_k_start % BLOCK_SIZE) / 8u;
+        let src0_idx = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+        let d = load_f16_at_src0(block_byte_base);
+        let q_byte = load_u32_at_src0(block_byte_base + 2u + byte_in_block) & 0xFFu;
+
+        for (var bit = 0u; bit < NQ; bit++) {
+            let global_k = global_k_start + bit;
+            if (global_k < params.k) {
+                shmem[i + bit] = select(-d, d, ((q_byte >> bit) & 1u) != 0u);
+            }
+        }
+    }
+}
+#endif // INIT_SRC0_SHMEM_Q1_0
+
 #ifdef INIT_SRC0_SHMEM_Q4_0
 const BLOCK_SIZE = 32u;
 const BLOCK_SIZE_BYTES = 18u;
@@ -707,3 +740,426 @@ fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u3
     }
 }
 #endif // INIT_SRC0_SHMEM_Q6_K
+
+#ifdef INIT_SRC0_SHMEM_IQ4_NL
+const BLOCK_SIZE = 32u;
+const BLOCK_SIZE_BYTES = 18u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+        let d               = load_f16_at_src0(block_byte_base);
+
+        let pos       = k_in_block % 16u;
+        let nib_shift = (k_in_block / 16u) * 4u;
+        let q_packed  = load_u32_at_src0(block_byte_base + 2u + (pos / 4u) * 4u);
+        let nib       = (get_byte(q_packed, pos % 4u) >> nib_shift) & 0xFu;
+
+        shmem[elem_idx] = d * f16(kvalues_iq4nl[nib]);
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ4_NL
+
+#ifdef INIT_SRC0_SHMEM_IQ4_XS
+const BLOCK_SIZE = 256u;
+const BLOCK_SIZE_BYTES = 136u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+
+        let d_scales_h = load_u32_at_src0(block_byte_base);
+        let d          = bitcast<vec2<f16>>(d_scales_h).x;
+        let scales_h   = d_scales_h >> 16u;
+
+        let ib  = k_in_block / 32u;
+        let pos = k_in_block % 32u;
+
+        let scales_l_word = load_u32_at_src0(block_byte_base + 4u);
+        let ls_lo         = (get_byte(scales_l_word, ib / 2u) >> ((ib & 1u) * 4u)) & 0xFu;
+        let ls_hi         = ((scales_h >> (2u * ib)) & 3u) << 4u;
+        let dl            = d * f16(i32(ls_lo | ls_hi) - 32);
+
+        let iqs       = ib * 16u + (pos % 16u);
+        let nib_shift = (pos / 16u) * 4u;
+        let q_packed  = load_u32_at_src0(block_byte_base + 8u + (iqs / 4u) * 4u);
+        let nib       = (get_byte(q_packed, iqs % 4u) >> nib_shift) & 0xFu;
+
+        shmem[elem_idx] = dl * f16(kvalues_iq4nl[nib]);
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ4_XS
+
+#ifdef INIT_SRC0_SHMEM_IQ1_S
+const BLOCK_SIZE = 256u;
+const BLOCK_SIZE_BYTES = 50u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+        let d               = load_f16_as_f32_at_src0(block_byte_base);
+
+        let ib  = k_in_block / 32u;
+        let pos = k_in_block % 32u;
+        let l   = pos / 8u;
+        let j   = pos % 8u;
+
+        let qh    = load_u32_at_src0(block_byte_base + 34u + ib * 2u) & 0xFFFFu;
+        let dl    = d * (2.0 * f32((qh >> 12u) & 7u) + 1.0);
+        let delta = select(IQ1_DELTA, -IQ1_DELTA, (qh & 0x8000u) != 0u);
+
+        let qs_w = load_u32_at_src0(block_byte_base + 2u + ib * 4u);
+        let ig   = (get_byte(qs_w, l) | (((qh >> (3u * l)) & 7u) << 8u)) * 8u;
+
+        let gw = iq1_grid[(ig + j) / 16u];
+        let g  = (gw >> (((ig + j) % 16u) * 2u)) & 3u;
+        let gs = bitcast<i32>(g << 30u) >> 30u;
+
+        shmem[elem_idx] = f16(dl * (f32(gs) + delta));
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ1_S
+
+#ifdef INIT_SRC0_SHMEM_IQ1_M
+const BLOCK_SIZE = 256u;
+const BLOCK_SIZE_BYTES = 56u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+
+        let scales0 = load_u32_at_src0(block_byte_base + 48u);
+        let scales1 = load_u32_at_src0(block_byte_base + 52u);
+        let scale_packed = ((scales0 >> 12u) & 0xFu) |
+                           ((scales0 >> 24u) & 0x00F0u) |
+                           ((scales1 >>  4u) & 0x0F00u) |
+                           ((scales1 >> 16u) & 0xF000u);
+        let d = f32(bitcast<vec2<f16>>(scale_packed).x);
+
+        let ib  = k_in_block / 32u;
+        let pos = k_in_block % 32u;
+        let l   = pos / 8u;
+        let j   = pos % 8u;
+
+        let scales = select(scales0, scales1, ib >= 4u);
+        let sw = (scales >> (16u * ((ib / 2u) % 2u))) & 0xFFFFu;
+        let s_pair = (sw >> (6u * (ib % 2u) + 3u * (l / 2u))) & 0x7u;
+        let dl     = d * f32(2u * s_pair + 1u);
+
+        let qh_word = load_u32_at_src0(block_byte_base + 32u + (ib / 2u) * 4u);
+        let qh      = qh_word >> (16u * (ib % 2u));
+        let qh_nib  = (qh >> (4u * l)) & 0xFu;
+
+        let qs_w = load_u32_at_src0(block_byte_base + ib * 4u);
+        let idx  = get_byte(qs_w, l) | ((qh_nib & 7u) << 8u);
+        let delta = select(IQ1_DELTA, -IQ1_DELTA, (qh_nib & 0x8u) != 0u);
+
+        let ig = idx * 8u;
+        let gw = iq1_grid[(ig + j) / 16u];
+        let g  = (gw >> (((ig + j) % 16u) * 2u)) & 3u;
+        let gs = bitcast<i32>(g << 30u) >> 30u;
+
+        shmem[elem_idx] = f16(dl * (f32(gs) + delta));
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ1_M
+
+#ifdef INIT_SRC0_SHMEM_IQ2_XXS
+const BLOCK_SIZE = 256u;
+const BLOCK_SIZE_BYTES = 66u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+        let d               = load_f16_as_f32_at_src0(block_byte_base);
+
+        let entry_idx = k_in_block / 8u;
+        let j         = k_in_block % 8u;
+
+        let ib = entry_idx & ~3u;
+        let l  = entry_idx & 3u;
+
+        let aux0 = load_u32_at_src0(block_byte_base + 2u + ib * 2u);
+        let aux1 = load_u32_at_src0(block_byte_base + 2u + (ib + 2u) * 2u);
+        let db   = d * (0.5 + f32(aux1 >> 28u)) * 0.25;
+
+        let ig    = get_byte(aux0, l) * 8u;
+        let is    = (aux1 >> (7u * l)) & 127u;
+        let signs = get_byte(ksigns_iq2xs[is / 4u], is % 4u);
+
+        let g = get_byte(iq2xxs_grid[(ig + j) / 4u], (ig + j) % 4u);
+        let m = select(1.0, -1.0, (get_byte(kmask_iq2xs[j / 4u], j % 4u) & signs) != 0u);
+
+        shmem[elem_idx] = f16(db * f32(g) * m);
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ2_XXS
+
+#ifdef INIT_SRC0_SHMEM_IQ2_XS
+const BLOCK_SIZE = 256u;
+const BLOCK_SIZE_BYTES = 74u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+        let d               = load_f16_as_f32_at_src0(block_byte_base);
+
+        let entry_idx = k_in_block / 8u;
+        let j         = k_in_block % 8u;
+
+        let ib = entry_idx & ~3u;
+        let l  = entry_idx & 3u;
+
+        let scales_word = load_u32_at_src0(block_byte_base + 66u + (ib / 16u) * 4u);
+        let s           = get_byte(scales_word, (ib % 16u) / 4u);
+        let s_nib       = select(s & 0xFu, (s >> 4u) & 0xFu, (l / 2u) != 0u);
+        let dl          = d * (0.5 + f32(s_nib)) * 0.25;
+
+        let qs_word = load_u32_at_src0(block_byte_base + 2u + (ib + l) * 2u);
+        let qs_val  = qs_word & 0xFFFFu;
+        let ig      = (qs_val & 511u) * 8u;
+        let is      = qs_val >> 9u;
+        let signs   = get_byte(ksigns_iq2xs[is / 4u], is % 4u);
+
+        let g = get_byte(iq2xs_grid[(ig + j) / 4u], (ig + j) % 4u);
+        let m = select(1.0, -1.0, (get_byte(kmask_iq2xs[j / 4u], j % 4u) & signs) != 0u);
+
+        shmem[elem_idx] = f16(dl * f32(g) * m);
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ2_XS
+
+#ifdef INIT_SRC0_SHMEM_IQ2_S
+const BLOCK_SIZE = 256u;
+const BLOCK_SIZE_BYTES = 82u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+        let d               = load_f16_as_f32_at_src0(block_byte_base);
+
+        let ib = k_in_block / 32u;
+        let l  = (k_in_block % 32u) / 8u;
+        let j  = k_in_block % 8u;
+
+        let scales_word = load_u32_at_src0(block_byte_base + 74u + (ib / 4u) * 4u);
+        let s           = get_byte(scales_word, ib % 4u);
+        let s_nib       = select(s & 0xFu, (s >> 4u) & 0xFu, (l / 2u) != 0u);
+        let dl          = d * (0.5 + f32(s_nib)) * 0.25;
+
+        let qs_word = load_u32_at_src0(block_byte_base + 2u + ib * 4u);
+        let qh_word = load_u32_at_src0(block_byte_base + 66u + (ib / 4u) * 4u);
+        let qh_b    = (get_byte(qh_word, ib % 4u) << (8u - 2u * l)) & 0x300u;
+        let ig      = (get_byte(qs_word, l) | qh_b) * 8u;
+
+        let signs_word = load_u32_at_src0(block_byte_base + 34u + ib * 4u);
+        let signs      = get_byte(signs_word, l);
+
+        let g = get_byte(iq2s_grid[(ig + j) / 4u], (ig + j) % 4u);
+        let m = select(1.0, -1.0, (get_byte(kmask_iq2xs[j / 4u], j % 4u) & signs) != 0u);
+
+        shmem[elem_idx] = f16(dl * f32(g) * m);
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ2_S
+
+#ifdef INIT_SRC0_SHMEM_IQ3_XXS
+const BLOCK_SIZE = 256u;
+const BLOCK_SIZE_BYTES = 98u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+        let d               = load_f16_as_f32_at_src0(block_byte_base);
+
+        let ib_pair = k_in_block / 32u;
+        let in_pair = k_in_block % 32u;
+        let l       = in_pair / 8u;
+        let in_l    = in_pair % 8u;
+        let k2      = in_l / 4u;
+        let j       = in_l % 4u;
+
+        let ib            = ib_pair * 2u;
+        let sc_sign_off   = block_byte_base + 2u + (ib + 32u) * 2u;
+        let sc_sign       = load_u32_at_src0(sc_sign_off);
+        let db            = d * (0.5 + f32(sc_sign >> 28u)) * 0.5;
+        let is            = (sc_sign >> (7u * l)) & 127u;
+        let signs         = get_byte(ksigns_iq2xs[is / 4u], is % 4u);
+
+        let ig_word = load_u32_at_src0(block_byte_base + 2u + (ib * 2u + l) * 2u) & 0xFFFFu;
+        let ig_byte = get_byte(ig_word, k2);
+        let g       = get_byte(iq3xxs_grid[ig_byte], j);
+        let m       = select(1.0, -1.0, (get_byte(kmask_iq2xs[k2], j) & signs) != 0u);
+
+        shmem[elem_idx] = f16(db * f32(g) * m);
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ3_XXS
+
+#ifdef INIT_SRC0_SHMEM_IQ3_S
+const BLOCK_SIZE = 256u;
+const BLOCK_SIZE_BYTES = 110u;
+
+fn init_shmem_src0(thread_id: u32, batch_offset: u32, offset_m: u32, k_outer: u32) {
+    for (var elem_idx = thread_id; elem_idx < TILE_SRC0_SHMEM; elem_idx += TOTAL_WORKGROUP_SIZE) {
+        let tile_m = elem_idx / TILE_K;
+        let tile_k = elem_idx % TILE_K;
+        let global_m = offset_m + tile_m;
+        let global_k = k_outer + tile_k;
+
+        if (global_m >= params.m || global_k >= params.k) {
+            shmem[elem_idx] = f16(0.0);
+            continue;
+        }
+
+        let block_k    = global_k / BLOCK_SIZE;
+        let k_in_block = global_k % BLOCK_SIZE;
+
+        let src0_idx        = batch_offset + global_m * params.stride_01 + block_k;
+        let block_byte_base = src0_idx * BLOCK_SIZE_BYTES;
+        let d               = load_f16_as_f32_at_src0(block_byte_base);
+
+        let ib   = k_in_block / 64u;
+        let rest = k_in_block % 64u;
+        let k    = rest / 32u;
+        let in_k = rest % 32u;
+        let l    = in_k / 8u;
+        let in_l = in_k % 8u;
+        let k2   = in_l / 4u;
+        let j    = in_l % 4u;
+
+        let scales_word = load_u32_at_src0(block_byte_base + 106u);
+        let s           = get_byte(scales_word, ib);
+        let s_nib       = select(s & 0xFu, (s >> 4u) & 0xFu, k != 0u);
+        let dl          = d * (1.0 + 2.0 * f32(s_nib));
+
+        let qh_word = load_u32_at_src0(block_byte_base + 66u + (ib / 2u) * 4u);
+        let qh_byte = get_byte(qh_word, (ib % 2u) * 2u + k);
+
+        let ig_word = load_u32_at_src0(block_byte_base + 2u + (ib * 8u + k * 4u + l) * 2u) & 0xFFFFu;
+        let ig_lo   = get_byte(ig_word, 0u) | ((qh_byte << (8u - 2u * l)) & 256u);
+        let ig_hi   = get_byte(ig_word, 1u) | ((qh_byte << (7u - 2u * l)) & 256u);
+        let ig      = select(ig_lo, ig_hi, k2 != 0u);
+
+        let signs_word = load_u32_at_src0(block_byte_base + 74u + (ib * 2u + k) * 4u);
+        let signs      = get_byte(signs_word, l);
+
+        let g = get_byte(iq3s_grid[ig], j);
+        let m = select(1.0, -1.0, (get_byte(kmask_iq2xs[k2], j) & signs) != 0u);
+
+        shmem[elem_idx] = f16(dl * f32(g) * m);
+    }
+}
+#endif // INIT_SRC0_SHMEM_IQ3_S

ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_vec.wgsl

@@ -128,6 +128,38 @@ fn main(
     }
 #endif
 
+#ifdef MUL_ACC_Q1_0
+#define BLOCK_SIZE 128
+#define BLOCK_SIZE_BYTES 18
+#define THREADS_PER_BLOCK 16
+#define ELEMS_PER_THREAD (BLOCK_SIZE/THREADS_PER_BLOCK)
+
+    let num_blocks = params.k / BLOCK_SIZE;
+    let thread_within_block = thread_id % THREADS_PER_BLOCK;
+    for (var block = thread_id / THREADS_PER_BLOCK; block < num_blocks; block += WG_SIZE / THREADS_PER_BLOCK) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + thread_within_block * ELEMS_PER_THREAD;
+        var x_block: array<f32, ELEMS_PER_THREAD>;
+        for (var i = 0u; i < ELEMS_PER_THREAD; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+                let d = f32(load_f16_at_src0(block_byte_base));
+                let q_byte = load_u32_at_src0(block_byte_base + 2u + thread_within_block) & 0xFFu;
+                var row_sum = 0.0;
+                for (var bit = 0u; bit < 8u; bit++) {
+                    let w = select(-d, d, ((q_byte >> bit) & 1u) != 0u);
+                    row_sum += w * x_block[bit];
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
 #ifdef MUL_ACC_Q4_0
 #define BLOCK_SIZE 32
 #define BLOCK_SIZE_BYTES 18
@@ -812,6 +844,520 @@ fn main(
     }
 #endif
 
+#ifdef MUL_ACC_IQ1_S
+#define BLOCK_SIZE 256
+#define BLOCK_SIZE_BYTES 50
+#define THREADS_PER_BLOCK 16
+
+    let tid = thread_id % THREADS_PER_BLOCK;
+    let block_group = thread_id / THREADS_PER_BLOCK;
+    let num_block_groups: u32 = WG_SIZE / THREADS_PER_BLOCK;
+
+    let sub_blk = tid / 2u;
+    let half    = tid % 2u;
+    let slot0   = half * 2u;
+    let y_offset = sub_blk * 32u + slot0 * 8u;
+
+    let num_blocks = params.k / BLOCK_SIZE;
+
+    for (var block = block_group; block < num_blocks; block += num_block_groups) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + y_offset;
+        var x_block: array<f32, 16>;
+        for (var i = 0u; i < 16u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+
+                let d     = f32(load_f16_at_src0(block_byte_base));
+                let qh    = load_u32_at_src0(block_byte_base + 34u + sub_blk * 2u) & 0xFFFFu;
+                let dl    = d * f32(2u * ((qh >> 12u) & 7u) + 1u);
+                let delta = select(IQ1_DELTA, -IQ1_DELTA, (qh & 0x8000u) != 0u);
+                let qs_w  = load_u32_at_src0(block_byte_base + 2u + sub_blk * 4u);
+
+                var row_sum = 0.0;
+                for (var ll = 0u; ll < 2u; ll++) {
+                    let l       = slot0 + ll;
+                    let qs_byte = get_byte(qs_w, l);
+                    let ig      = (qs_byte | (((qh >> (3u * l)) & 7u) << 8u)) * 8u;
+                    let gw      = iq1_grid[ig / 16u];
+                    let bit_base = (ig % 16u) * 2u;
+                    for (var j = 0u; j < 8u; j++) {
+                        let g  = (gw >> (bit_base + j * 2u)) & 3u;
+                        let gs = select(f32(g), f32(g) - 4.0, (g & 2u) != 0u);
+                        row_sum += dl * (gs + delta) * x_block[ll * 8u + j];
+                    }
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
+#ifdef MUL_ACC_IQ1_M
+#define BLOCK_SIZE 256
+#define BLOCK_SIZE_BYTES 56
+#define THREADS_PER_BLOCK 16
+
+    let tid = thread_id % THREADS_PER_BLOCK;
+    let block_group = thread_id / THREADS_PER_BLOCK;
+    let num_block_groups: u32 = WG_SIZE / THREADS_PER_BLOCK;
+
+    let sub_blk = tid / 2u;
+    let half    = tid % 2u;
+    let slot0   = half * 2u;
+    let y_offset = sub_blk * 32u + slot0 * 8u;
+
+    let num_blocks = params.k / BLOCK_SIZE;
+
+    for (var block = block_group; block < num_blocks; block += num_block_groups) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + y_offset;
+        var x_block: array<f32, 16>;
+        for (var i = 0u; i < 16u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+
+                let sc_lo = load_u32_at_src0(block_byte_base + 48u);
+                let sc_hi = load_u32_at_src0(block_byte_base + 52u);
+                let sc0 = sc_lo & 0xFFFFu;
+                let sc1 = (sc_lo >> 16u) & 0xFFFFu;
+                let sc2 = sc_hi & 0xFFFFu;
+                let sc3 = (sc_hi >> 16u) & 0xFFFFu;
+                let d_bits = (sc0 >> 12u) | ((sc1 >> 8u) & 0xF0u) | ((sc2 >> 4u) & 0xF00u) | (sc3 & 0xF000u);
+                let d = f32(bitcast<vec2<f16>>(d_bits)[0]);
+
+                let sc_u16 = select(select(sc2, sc3, sub_blk >= 6u),
+                                    select(sc0, sc1, sub_blk >= 2u),
+                                    sub_blk < 4u);
+
+                let qs_w = load_u32_at_src0(block_byte_base + sub_blk * 4u);
+                let qh = load_u32_at_src0(block_byte_base + 32u + sub_blk * 2u) & 0xFFFFu;
+                let qh_lo = qh & 0xFFu;
+                let qh_hi = (qh >> 8u) & 0xFFu;
+
+                var row_sum = 0.0;
+                for (var ll = 0u; ll < 2u; ll++) {
+                    let l = slot0 + ll;
+                    let bit_off = 6u * (sub_blk % 2u) + 3u * (l / 2u);
+                    let sub_scale = (sc_u16 >> bit_off) & 0x7u;
+                    let dl = d * f32(2u * sub_scale + 1u);
+                    let qh_byte = select(qh_lo, qh_hi, l >= 2u);
+                    let ll2 = l % 2u;
+                    let grid_idx = get_byte(qs_w, l) | (((qh_byte >> (4u * ll2)) & 7u) << 8u);
+                    let delta = select(IQ1_DELTA, -IQ1_DELTA, ((qh_byte >> (3u + 4u * ll2)) & 1u) != 0u);
+                    let ig = grid_idx * 8u;
+                    let gw = iq1_grid[ig / 16u];
+                    let bit_base = (ig % 16u) * 2u;
+                    for (var j = 0u; j < 8u; j++) {
+                        let g  = (gw >> (bit_base + j * 2u)) & 3u;
+                        let gs = select(f32(g), f32(g) - 4.0, (g & 2u) != 0u);
+                        row_sum += dl * (gs + delta) * x_block[ll * 8u + j];
+                    }
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
+#ifdef MUL_ACC_IQ2_XXS
+#define BLOCK_SIZE 256
+#define BLOCK_SIZE_BYTES 66
+#define THREADS_PER_BLOCK 16
+
+    let tid = thread_id % THREADS_PER_BLOCK;
+    let block_group = thread_id / THREADS_PER_BLOCK;
+    let num_block_groups: u32 = WG_SIZE / THREADS_PER_BLOCK;
+
+    let sub_blk = tid / 2u;
+    let half    = tid % 2u;
+    let slot0   = half * 2u;
+    let y_offset = sub_blk * 32u + slot0 * 8u;
+
+    let num_blocks = params.k / BLOCK_SIZE;
+
+    for (var block = block_group; block < num_blocks; block += num_block_groups) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + y_offset;
+        var x_block: array<f32, 16>;
+        for (var i = 0u; i < 16u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+                let d = f32(load_f16_at_src0(block_byte_base));
+                let aux_lo = load_u32_at_src0(block_byte_base + 2u + sub_blk * 8u);
+                let aux_hi = load_u32_at_src0(block_byte_base + 2u + sub_blk * 8u + 4u);
+                let ls = aux_hi >> 28u;
+                let db = d * (0.5 + f32(ls)) * 0.25;
+
+                var row_sum = 0.0;
+                for (var ll = 0u; ll < 2u; ll++) {
+                    let l = slot0 + ll;
+                    let grid_idx = (aux_lo >> (8u * l)) & 0xFFu;
+                    let signs_idx = (aux_hi >> (7u * l)) & 0x7Fu;
+                    let signs = (ksigns_iq2xs[signs_idx / 4u] >> ((signs_idx % 4u) * 8u)) & 0xFFu;
+                    let gw_lo = iq2xxs_grid[grid_idx * 2u];
+                    let gw_hi = iq2xxs_grid[grid_idx * 2u + 1u];
+                    for (var j = 0u; j < 8u; j++) {
+                        let gw = select(gw_hi, gw_lo, j < 4u);
+                        let b = f32((gw >> ((j & 3u) * 8u)) & 0xFFu);
+                        let s = select(1.0, -1.0, ((signs >> j) & 1u) != 0u);
+                        row_sum += db * b * s * x_block[ll * 8u + j];
+                    }
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
+#ifdef MUL_ACC_IQ2_XS
+#define BLOCK_SIZE 256
+#define BLOCK_SIZE_BYTES 74
+#define THREADS_PER_BLOCK 16
+
+    let tid = thread_id % THREADS_PER_BLOCK;
+    let block_group = thread_id / THREADS_PER_BLOCK;
+    let num_block_groups: u32 = WG_SIZE / THREADS_PER_BLOCK;
+
+    let sub_blk = tid / 2u;
+    let half    = tid % 2u;
+    let slot0   = half * 2u;
+    let y_offset = sub_blk * 32u + slot0 * 8u;
+
+    let num_blocks = params.k / BLOCK_SIZE;
+
+    for (var block = block_group; block < num_blocks; block += num_block_groups) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + y_offset;
+        var x_block: array<f32, 16>;
+        for (var i = 0u; i < 16u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+                let d = f32(load_f16_at_src0(block_byte_base));
+                let qs_lo = load_u32_at_src0(block_byte_base + 2u + sub_blk * 8u);
+                let qs_hi = load_u32_at_src0(block_byte_base + 2u + sub_blk * 8u + 4u);
+                let scales_word = load_u32_at_src0(block_byte_base + 66u + (sub_blk / 4u) * 4u);
+                let scales_byte = get_byte(scales_word, sub_blk % 4u);
+
+                var row_sum = 0.0;
+                for (var ll = 0u; ll < 2u; ll++) {
+                    let l = slot0 + ll;
+                    let qs_word = select(qs_hi, qs_lo, l < 2u);
+                    let half2 = (l % 2u) * 16u;
+                    let qs_val = (qs_word >> half2) & 0xFFFFu;
+                    let grid_idx = qs_val & 0x1FFu;
+                    let signs_idx = (qs_val >> 9u) & 0x7Fu;
+                    let sub_scale = (scales_byte >> (4u * (l / 2u))) & 0xFu;
+                    let db = d * (0.5 + f32(sub_scale)) * 0.25;
+                    let signs = (ksigns_iq2xs[signs_idx / 4u] >> ((signs_idx % 4u) * 8u)) & 0xFFu;
+                    let gw_lo = iq2xs_grid[grid_idx * 2u];
+                    let gw_hi = iq2xs_grid[grid_idx * 2u + 1u];
+                    for (var j = 0u; j < 8u; j++) {
+                        let gw = select(gw_hi, gw_lo, j < 4u);
+                        let b = f32((gw >> ((j & 3u) * 8u)) & 0xFFu);
+                        let s = select(1.0, -1.0, ((signs >> j) & 1u) != 0u);
+                        row_sum += db * b * s * x_block[ll * 8u + j];
+                    }
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
+#ifdef MUL_ACC_IQ2_S
+#define BLOCK_SIZE 256
+#define BLOCK_SIZE_BYTES 82
+#define THREADS_PER_BLOCK 16
+
+    let tid = thread_id % THREADS_PER_BLOCK;
+    let block_group = thread_id / THREADS_PER_BLOCK;
+    let num_block_groups: u32 = WG_SIZE / THREADS_PER_BLOCK;
+
+    let sub_blk = tid / 2u;
+    let half    = tid % 2u;
+    let slot0   = half * 2u;
+    let y_offset = sub_blk * 32u + slot0 * 8u;
+
+    let num_blocks = params.k / BLOCK_SIZE;
+
+    for (var block = block_group; block < num_blocks; block += num_block_groups) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + y_offset;
+        var x_block: array<f32, 16>;
+        for (var i = 0u; i < 16u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+                let d = f32(load_f16_at_src0(block_byte_base));
+                let qs_w = load_u32_at_src0(block_byte_base + 2u + sub_blk * 4u);
+                let sg_w = load_u32_at_src0(block_byte_base + 34u + sub_blk * 4u);
+                let qh_word = load_u32_at_src0(block_byte_base + 66u + (sub_blk / 4u) * 4u);
+                let qh_byte = get_byte(qh_word, sub_blk % 4u);
+                let sc_word = load_u32_at_src0(block_byte_base + 74u + (sub_blk / 4u) * 4u);
+                let scales_byte = get_byte(sc_word, sub_blk % 4u);
+
+                var row_sum = 0.0;
+                for (var ll = 0u; ll < 2u; ll++) {
+                    let l = slot0 + ll;
+                    let qs_byte = get_byte(qs_w, l);
+                    let sign_byte = get_byte(sg_w, l);
+                    let grid_idx = qs_byte | (((qh_byte >> (2u * l)) & 3u) << 8u);
+                    let sub_scale = (scales_byte >> (4u * (l / 2u))) & 0xFu;
+                    let db = d * (0.5 + f32(sub_scale)) * 0.25;
+                    let gw_lo = iq2s_grid[grid_idx * 2u];
+                    let gw_hi = iq2s_grid[grid_idx * 2u + 1u];
+                    for (var j = 0u; j < 8u; j++) {
+                        let gw = select(gw_hi, gw_lo, j < 4u);
+                        let b = f32((gw >> ((j & 3u) * 8u)) & 0xFFu);
+                        let s = select(1.0, -1.0, ((sign_byte >> j) & 1u) != 0u);
+                        row_sum += db * b * s * x_block[ll * 8u + j];
+                    }
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
+#ifdef MUL_ACC_IQ3_XXS
+#define BLOCK_SIZE 256
+#define BLOCK_SIZE_BYTES 98
+#define THREADS_PER_BLOCK 16
+
+    let tid = thread_id % THREADS_PER_BLOCK;
+    let block_group = thread_id / THREADS_PER_BLOCK;
+    let num_block_groups: u32 = WG_SIZE / THREADS_PER_BLOCK;
+
+    let sub_blk = tid / 2u;
+    let half    = tid % 2u;
+    let slot0   = half * 2u;
+    let y_offset = sub_blk * 32u + slot0 * 8u;
+
+    let num_blocks = params.k / BLOCK_SIZE;
+
+    for (var block = block_group; block < num_blocks; block += num_block_groups) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + y_offset;
+        var x_block: array<f32, 16>;
+        for (var i = 0u; i < 16u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+                let d = f32(load_f16_at_src0(block_byte_base));
+                let qs_lo = load_u32_at_src0(block_byte_base + 2u + sub_blk * 8u);
+                let qs_hi = load_u32_at_src0(block_byte_base + 2u + sub_blk * 8u + 4u);
+                let aux = load_u32_at_src0(block_byte_base + 66u + sub_blk * 4u);
+                let ls = aux >> 28u;
+                let db = d * (0.5 + f32(ls)) * 0.5;
+
+                var row_sum = 0.0;
+                for (var ll = 0u; ll < 2u; ll++) {
+                    let l = slot0 + ll;
+                    let qs_word = select(qs_hi, qs_lo, l < 2u);
+                    let byte_pos = (l % 2u) * 2u;
+                    let grid_idx_0 = (qs_word >> (byte_pos * 8u)) & 0xFFu;
+                    let grid_idx_1 = (qs_word >> ((byte_pos + 1u) * 8u)) & 0xFFu;
+                    let signs_idx = (aux >> (7u * l)) & 0x7Fu;
+                    let signs = (ksigns_iq2xs[signs_idx / 4u] >> ((signs_idx % 4u) * 8u)) & 0xFFu;
+                    let grid1 = iq3xxs_grid[grid_idx_0];
+                    let grid2 = iq3xxs_grid[grid_idx_1];
+                    for (var j = 0u; j < 4u; j++) {
+                        let b1 = f32((grid1 >> (j * 8u)) & 0xFFu);
+                        let b2 = f32((grid2 >> (j * 8u)) & 0xFFu);
+                        let s1 = select(1.0, -1.0, ((signs >> j) & 1u) != 0u);
+                        let s2 = select(1.0, -1.0, ((signs >> (j + 4u)) & 1u) != 0u);
+                        row_sum += db * b1 * s1 * x_block[ll * 8u + j];
+                        row_sum += db * b2 * s2 * x_block[ll * 8u + j + 4u];
+                    }
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
+#ifdef MUL_ACC_IQ3_S
+#define BLOCK_SIZE 256
+#define BLOCK_SIZE_BYTES 110
+#define THREADS_PER_BLOCK 16
+
+    let tid = thread_id % THREADS_PER_BLOCK;
+    let block_group = thread_id / THREADS_PER_BLOCK;
+    let num_block_groups: u32 = WG_SIZE / THREADS_PER_BLOCK;
+
+    let sub_blk = tid / 2u;
+    let half    = tid % 2u;
+    let slot0   = half * 2u;
+    let y_offset = sub_blk * 32u + slot0 * 8u;
+
+    let num_blocks = params.k / BLOCK_SIZE;
+
+    for (var block = block_group; block < num_blocks; block += num_block_groups) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + y_offset;
+        var x_block: array<f32, 16>;
+        for (var i = 0u; i < 16u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+                let d = f32(load_f16_at_src0(block_byte_base));
+                let qs_lo = load_u32_at_src0(block_byte_base + 2u + sub_blk * 8u);
+                let qs_hi = load_u32_at_src0(block_byte_base + 2u + sub_blk * 8u + 4u);
+                let qh_word = load_u32_at_src0(block_byte_base + 66u + (sub_blk / 4u) * 4u);
+                let qh_byte = get_byte(qh_word, sub_blk % 4u);
+                let sg_w = load_u32_at_src0(block_byte_base + 74u + sub_blk * 4u);
+                let sc_word = load_u32_at_src0(block_byte_base + 106u);
+                let scales_byte = get_byte(sc_word, sub_blk / 2u);
+                let sub_scale = (scales_byte >> (4u * (sub_blk % 2u))) & 0xFu;
+                let db = d * (1.0 + 2.0 * f32(sub_scale));
+
+                var row_sum = 0.0;
+                for (var ll = 0u; ll < 2u; ll++) {
+                    let l = slot0 + ll;
+                    let qs_word = select(qs_hi, qs_lo, l < 2u);
+                    let byte_pos = (l % 2u) * 2u;
+                    let qs0 = (qs_word >> (byte_pos * 8u)) & 0xFFu;
+                    let qs1 = (qs_word >> ((byte_pos + 1u) * 8u)) & 0xFFu;
+                    let grid_idx_1 = qs0 | (((qh_byte >> (2u * l)) & 1u) << 8u);
+                    let grid_idx_2 = qs1 | (((qh_byte >> (2u * l + 1u)) & 1u) << 8u);
+                    let sign_byte = get_byte(sg_w, l);
+                    let grid1 = iq3s_grid[grid_idx_1];
+                    let grid2 = iq3s_grid[grid_idx_2];
+                    for (var j = 0u; j < 4u; j++) {
+                        let b1 = f32((grid1 >> (j * 8u)) & 0xFFu);
+                        let b2 = f32((grid2 >> (j * 8u)) & 0xFFu);
+                        let s1 = select(1.0, -1.0, ((sign_byte >> j) & 1u) != 0u);
+                        let s2 = select(1.0, -1.0, ((sign_byte >> (j + 4u)) & 1u) != 0u);
+                        row_sum += db * b1 * s1 * x_block[ll * 8u + j];
+                        row_sum += db * b2 * s2 * x_block[ll * 8u + j + 4u];
+                    }
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
+#ifdef MUL_ACC_IQ4_NL
+#define BLOCK_SIZE 32
+#define BLOCK_SIZE_BYTES 18
+#define THREADS_PER_BLOCK 4
+#define ELEMS_PER_THREAD (BLOCK_SIZE/THREADS_PER_BLOCK)
+
+    let num_blocks = params.k / BLOCK_SIZE;
+    let thread_within_block = thread_id % THREADS_PER_BLOCK;
+    for (var block = thread_id / THREADS_PER_BLOCK; block < num_blocks; block += WG_SIZE / THREADS_PER_BLOCK) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + thread_within_block * 4u;
+        var x_block: array<f32, ELEMS_PER_THREAD>;
+        for (var i = 0u; i < ELEMS_PER_THREAD / 2u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+            x_block[i + 4u] = f32(src1[x_base + i + 16u]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+                let d = f32(load_f16_at_src0(block_byte_base));
+                var row_sum = 0.0;
+
+                let q_packed = load_u32_at_src0(block_byte_base + 2u + 4u * thread_within_block);
+                for (var byte_idx = 0u; byte_idx < 4u; byte_idx++) {
+                    let q_byte = get_byte(q_packed, byte_idx);
+                    let q_lo = f32(kvalues_iq4nl[q_byte & 0xFu]) * d;
+                    let q_hi = f32(kvalues_iq4nl[(q_byte >> 4u) & 0xFu]) * d;
+                    row_sum += q_lo * x_block[byte_idx];
+                    row_sum += q_hi * x_block[byte_idx + 4u];
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
+#ifdef MUL_ACC_IQ4_XS
+#define BLOCK_SIZE 256
+#define BLOCK_SIZE_BYTES 136
+#define THREADS_PER_BLOCK 16
+
+    let tid = thread_id % THREADS_PER_BLOCK;
+    let block_group = thread_id / THREADS_PER_BLOCK;
+    let num_block_groups: u32 = WG_SIZE / THREADS_PER_BLOCK;
+
+    let sub_blk = tid / 2u;
+    let half    = tid % 2u;
+    let y_offset = sub_blk * 32u + half * 16u;
+
+    let num_blocks = params.k / BLOCK_SIZE;
+
+    for (var block = block_group; block < num_blocks; block += num_block_groups) {
+        let x_base = src1_idx_base + block * BLOCK_SIZE + y_offset;
+        var x_block: array<f32, 16>;
+        for (var i = 0u; i < 16u; i++) {
+            x_block[i] = f32(src1[x_base + i]);
+        }
+
+        for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
+            let output_row = row_base + row;
+            if (output_row < params.m) {
+                let block_byte_base = (src0_batch_offset + output_row * params.stride_01 + block) * BLOCK_SIZE_BYTES;
+                let d = f32(load_f16_at_src0(block_byte_base));
+                let scales_h = load_u16_at_src0(block_byte_base + 2u);
+                let scales_l_word = load_u32_at_src0(block_byte_base + 4u);
+                let sl_byte = get_byte(scales_l_word, sub_blk / 2u);
+                let sl = (sl_byte >> (4u * (sub_blk % 2u))) & 0xFu;
+                let sh_bits = (scales_h >> (2u * sub_blk)) & 3u;
+                let ls = i32(sl | (sh_bits << 4u));
+                let dl = d * f32(ls - 32);
+
+                let qs_byte_off = 8u + sub_blk * 16u;
+                let q_w0 = load_u32_at_src0(block_byte_base + qs_byte_off);
+                let q_w1 = load_u32_at_src0(block_byte_base + qs_byte_off + 4u);
+                let q_w2 = load_u32_at_src0(block_byte_base + qs_byte_off + 8u);
+                let q_w3 = load_u32_at_src0(block_byte_base + qs_byte_off + 12u);
+
+                var row_sum = 0.0;
+                for (var i = 0u; i < 16u; i++) {
+                    let q_word = select(
+                        select(q_w0, q_w1, i >= 4u),
+                        select(q_w2, q_w3, i >= 12u),
+                        i >= 8u);
+                    let q_byte = get_byte(q_word, i % 4u);
+                    let nib = select(q_byte & 0xFu, (q_byte >> 4u) & 0xFu, half == 1u);
+                    row_sum += f32(kvalues_iq4nl[nib]) * dl * x_block[i];
+                }
+                acc[row] += row_sum;
+            }
+        }
+    }
+#endif
+
 #ifdef USE_SUBGROUP_REDUCTION
     for (var row = 0u; row < OUTPUTS_PER_WG; row++) {
         let subgroup_total = subgroupAdd(acc[row]);

ggml/src/ggml-webgpu/wgsl-shaders/rms_norm_mul.wgsl

@@ -66,8 +66,6 @@ fn update(rn_src_offset: u32, dst_offset: u32, scale: f32, mul_src_offset: u32)
 struct Params {
     offset_rn_src: u32,
     offset_mul_src: u32,
-    offset_merged_rn_src: u32,
-    offset_merged_mul_src: u32,
     offset_dst: u32,
 
     stride_rn_src1: u32,
@@ -107,8 +105,8 @@ fn main(@builtin(workgroup_id) wid: vec3<u32>,
     i = i % (params.ne2 * params.ne1);
     let i2 = i / params.ne1;
     let i1 = i % params.ne1;
-    let i_rn_src_row = params.offset_rn_src + params.offset_merged_rn_src + i3 * params.stride_rn_src3 + i2 * params.stride_rn_src2 + i1 * params.stride_rn_src1;
-    let i_mul_src_row = params.offset_mul_src + params.offset_merged_mul_src + (i3 % params.mul_src_ne3) * params.stride_mul_src3 + (i2 % params.mul_src_ne2) * params.stride_mul_src2 + (i1 % params.mul_src_ne1) * params.stride_mul_src1;
+    let i_rn_src_row = params.offset_rn_src + i3 * params.stride_rn_src3 + i2 * params.stride_rn_src2 + i1 * params.stride_rn_src1;
+    let i_mul_src_row = params.offset_mul_src + (i3 % params.mul_src_ne3) * params.stride_mul_src3 + (i2 % params.mul_src_ne2) * params.stride_mul_src2 + (i1 % params.mul_src_ne1) * params.stride_mul_src1;
     let i_dst_row = params.offset_dst + i3 * params.stride_dst3 + i2 * params.stride_dst2 + i1 * params.stride_dst1;
 
     let elems = (params.ne0 + WG_SIZE - 1) / WG_SIZE;

ggml/src/ggml-webgpu/wgsl-shaders/ssm_scan.wgsl

@@ -45,6 +45,14 @@ struct Params {
 };
 
 @group(0) @binding(0) var<storage, read_write> s_in: array<f32>;
+#ifdef XBC_OVERLAP
+@group(0) @binding(1) var<storage, read_write> x_B_C_merged: array<f32>;
+@group(0) @binding(2) var<storage, read_write> dt: array<f32>;
+@group(0) @binding(3) var<storage, read_write> A: array<f32>;
+@group(0) @binding(4) var<storage, read_write> ids: array<i32>;
+@group(0) @binding(5) var<storage, read_write> dst: array<f32>;
+@group(0) @binding(6) var<uniform> params: Params;
+#else
 @group(0) @binding(1) var<storage, read_write> x: array<f32>;
 @group(0) @binding(2) var<storage, read_write> dt: array<f32>;
 @group(0) @binding(3) var<storage, read_write> A: array<f32>;
@@ -53,6 +61,7 @@ struct Params {
 @group(0) @binding(6) var<storage, read_write> ids: array<i32>;
 @group(0) @binding(7) var<storage, read_write> dst: array<f32>;
 @group(0) @binding(8) var<uniform> params: Params;
+#endif
 
 var<workgroup> shared_x_dt: array<f32, TOKENS_PER_TILE>;
 var<workgroup> shared_dtsp: array<f32, TOKENS_PER_TILE>;
@@ -98,7 +107,11 @@ fn main(
                 let dt0 = dt[dt_idx];
                 let dtsp = select(log(1.0 + exp(dt0)), dt0, dt0 > 20.0);
                 shared_dtsp[tid] = dtsp;
+#ifdef XBC_OVERLAP
+                shared_x_dt[tid] = x_B_C_merged[x_idx] * dtsp;
+#else
                 shared_x_dt[tid] = x[x_idx] * dtsp;
+#endif
             }
         }
 
@@ -116,16 +129,28 @@ fn main(
 
             let b_idx = params.offset_B + tid + g * params.stride_B1 + token * params.stride_B2 + i3 * params.stride_B3;
             let c_idx = params.offset_C + tid + g * params.stride_C1 + token * params.stride_C2 + i3 * params.stride_C3;
+#ifdef XBC_OVERLAP
+            let s = s_prev * dA + x_B_C_merged[b_idx] * x_dt;
+#else
             let s = s_prev * dA + B[b_idx] * x_dt;
+#endif
             s_prev = s;
 
 #ifdef USE_SUBGROUP_REDUCTION
+#ifdef XBC_OVERLAP
+            let subgroup_partial = subgroupAdd(s * x_B_C_merged[c_idx]);
+#else
             let subgroup_partial = subgroupAdd(s * C[c_idx]);
+#endif
             if (subgroup_invocation_id == 0u) {
                 shared_reduce[reduce_idx - tid + subgroup_id] = subgroup_partial;
             }
+#else
+#ifdef XBC_OVERLAP
+            shared_reduce[reduce_idx] = s * x_B_C_merged[c_idx];
 #else
             shared_reduce[reduce_idx] = s * C[c_idx];
+#endif
 #endif
 
             workgroupBarrier();

ggml/src/ggml-zdnn/ggml-zdnn.cpp

@@ -423,8 +423,8 @@ static ggml_backend_i ggml_backend_zdnn_i = {
     /* .free                   = */ ggml_backend_zdnn_free,
     /* .set_tensor_async       = */ NULL,
     /* .get_tensor_async       = */ NULL,
-    /* .get_tensor_2d_async    = */ NULL,
     /* .set_tensor_2d_async    = */ NULL,
+    /* .get_tensor_2d_async    = */ NULL,
     /* .cpy_tensor_async       = */ NULL,
     /* .synchronize            = */ NULL,
     /* .graph_plan_create      = */ NULL,

ggml/src/ggml-zendnn/ggml-zendnn.cpp

@@ -407,8 +407,8 @@ static struct ggml_backend_i ggml_backend_zendnn_i = {
     /* .free                    = */ ggml_backend_zendnn_free,
     /* .set_tensor_async        = */ NULL,
     /* .get_tensor_async        = */ NULL,
-    /* .get_tensor_2d_async     = */ NULL,
     /* .set_tensor_2d_async     = */ NULL,
+    /* .get_tensor_2d_async     = */ NULL,
     /* .cpy_tensor_async        = */ NULL,
     /* .synchronize             = */ NULL,
     /* .graph_plan_create       = */ NULL,

scripts/jinja/jinja-tester.py

@@ -20,6 +20,7 @@ from PySide6.QtCore import Qt, QRect, QSize
 from jinja2 import TemplateSyntaxError
 from jinja2.sandbox import ImmutableSandboxedEnvironment
 from datetime import datetime
+from typing import Callable
 
 
 def format_template_content(template_content):
@@ -395,7 +396,7 @@ class JinjaTester(QMainWindow):
                 ensure_ascii=ensure_ascii,
             )
         )
-        env.globals["strftime_now"] = lambda format: datetime.now().strftime(format)  # ty: ignore[invalid-assignment]
+        env.globals["strftime_now"]: Callable[[str], str] = lambda format: datetime.now().strftime(format)
         env.globals["raise_exception"] = raise_exception  # ty: ignore[invalid-assignment]
         try:
             template = env.from_string(template_str)

scripts/pr2wt.sh

@@ -68,11 +68,19 @@ dir=$(basename $(pwd))
 git branch -D pr/$PR 2> /dev/null
 git worktree add -b pr/$PR ../$dir-pr-$PR pr/$PR/$head_ref 2> /dev/null
 
+og_path=$(pwd)
 wt_path=$(cd ../$dir-pr-$PR && pwd)
 
 echo "git worktree created in $wt_path"
 
 cd $wt_path
+
+# pi agent setup in the worktree
+if [[ -f "$og_path/.pi/SYSTEM.md" && ! -f ".pi/SYSTEM.md" ]]; then
+    mkdir -p .pi
+    ln -sfn "$og_path/.pi/SYSTEM.md" .pi/SYSTEM.md
+fi
+
 git branch --set-upstream-to=pr/$PR/$head_ref
 git pull   --ff-only || {
     echo "error: failed to pull pr/$PR"

scripts/snapdragon/adb/run-cli.sh

@@ -54,13 +54,23 @@ opqueue=
 opflt=
 [ "$OF" != "" ] && opflt="GGML_HEXAGON_OPFILTER=$OF"
 
+vmem=
+[ "$VM" != "" ] && opflt="GGML_HEXAGON_VMEM=$VM"
+
+mbuf=
+[ "$MB" != "" ] && opflt="GGML_HEXAGON_MBUF=$MB"
+vmem=
+[ "$VM" != "" ] && vmem="GGML_HEXAGON_VMEM=$VM"
+
+mbuf=
+[ "$MB" != "" ] && mbuf="GGML_HEXAGON_MBUF=$MB"
 set -x
 
 adb $adbserial $adbhost shell " \
   cd $basedir; ulimit -c unlimited;        \
     LD_LIBRARY_PATH=$basedir/$branch/lib   \
     ADSP_LIBRARY_PATH=$basedir/$branch/lib \
-    $verbose $sched $opmask $profile $nhvx $hmx $ndev $hb $opbatch $opqueue $opflt \
+    $verbose $sched $opmask $profile $nhvx $hmx $ndev $hb $opbatch $opqueue $opflt $vmem $mbuf \
       ./$branch/bin/llama-cli --no-mmap -m $basedir/../gguf/$model \
          --poll 1000 -t 6 --cpu-mask 0xfc --cpu-strict 1           \
          --ctx-size 8192 --ubatch-size 256 -fa on                  \

scripts/snapdragon/adb/run-completion.sh

@@ -54,13 +54,19 @@ opqueue=
 opflt=
 [ "$OF" != "" ] && opflt="GGML_HEXAGON_OPFILTER=$OF"
 
+vmem=
+[ "$VM" != "" ] && vmem="GGML_HEXAGON_VMEM=$VM"
+
+mbuf=
+[ "$MB" != "" ] && mbuf="GGML_HEXAGON_MBUF=$MB"
+
 set -x
 
 adb $adbserial $adbhost shell " \
   cd $basedir; ulimit -c unlimited;        \
     LD_LIBRARY_PATH=$basedir/$branch/lib   \
     ADSP_LIBRARY_PATH=$basedir/$branch/lib \
-    $verbose $sched $opmask $profile $nhvx $hmx $ndev $hb $opbatch $opqueue $opflt \
+    $verbose $sched $opmask $profile $nhvx $hmx $ndev $hb $opbatch $opqueue $opflt $vmem $mbuf \
       ./$branch/bin/llama-completion --no-mmap -m $basedir/../gguf/$model \
          --poll 1000 -t 6 --cpu-mask 0xfc --cpu-strict 1                  \
          --ctx-size 8192 --ubatch-size 256 -fa on                         \

scripts/sync-ggml.last

@@ -1 +1 @@
-1c40d85a4dcfcd62176f649b8682433bb1a6caef
+387fa29fbbf3149f06a631c7850b6c35c24b0232

scripts/sync_vendor.py

@@ -5,7 +5,7 @@ import os
 import sys
 import subprocess
 
-HTTPLIB_VERSION = "refs/tags/v0.43.1"
+HTTPLIB_VERSION = "refs/tags/v0.43.2"
 
 vendor = {
     "https://github.com/nlohmann/json/releases/latest/download/json.hpp":     "vendor/nlohmann/json.hpp",

scripts/wc2wt.sh

@@ -0,0 +1,58 @@
+#!/usr/bin/env bash
+
+# initialize a new worktree from a branch name:
+#
+# - creates a new branch from current HEAD
+# - creates a new worktree in a parent folder, suffixed with the branch name
+#
+# sample usage:
+#   ./scripts/wc2wt.sh gg/new-feature-foo-bar
+#   ./scripts/wc2wt.sh gg/new-feature-foo-bar opencode
+#   ./scripts/wc2wt.sh gg/new-feature-foo-bar "cmake -B build && cmake --build build"
+#   ./scripts/wc2wt.sh gg/new-feature-foo-bar "bash -l"
+
+function usage() {
+    echo "usage: $0 <branch_name> [cmd]"
+    exit 1
+}
+
+# check we are in the right directory
+if [[ ! -f "scripts/wc2wt.sh" ]]; then
+    echo "error: this script must be run from the root of the repository"
+    exit 1
+fi
+
+if [[ $# -lt 1 || $# -gt 2 ]]; then
+    usage
+fi
+
+BRANCH=$1
+
+if [[ -z "$BRANCH" ]]; then
+    echo "error: branch name must not be empty"
+    exit 1
+fi
+
+dir=$(basename $(pwd))
+# sanitize branch name for directory name (replace / with -)
+dir_suffix=$(echo "$BRANCH" | tr '/' '-')
+
+git worktree add -b "$BRANCH" "../$dir-$dir_suffix" HEAD
+
+og_path=$(pwd)
+wt_path=$(cd "../$dir-$dir_suffix" && pwd)
+
+echo "git worktree created in $wt_path"
+
+cd "$wt_path"
+
+# pi agent setup in the worktree
+if [[ -f "$og_path/.pi/SYSTEM.md" && ! -f ".pi/SYSTEM.md" ]]; then
+    mkdir -p .pi
+    ln -sfn "$og_path/.pi/SYSTEM.md" .pi/SYSTEM.md
+fi
+
+if [[ $# -eq 2 ]]; then
+    echo "executing: $2"
+    eval "$2"
+fi

src/models/llama.cpp

@@ -72,9 +72,6 @@ llm_build_llama<embed>::llm_build_llama(const llama_model & model, const llm_gra
             cur = build_attn(inp_attn,
                     model.layers[il].wo, model.layers[il].wo_b, model.layers[il].wo_s,
                     Qcur, Kcur, Vcur, nullptr, nullptr, nullptr, kq_scale, il);
-            if (model.layers[il].wo_s) {
-                cur = ggml_mul(ctx0, cur, model.layers[il].wo_s);
-            }
             cb(cur, "attn_out", il);
         }
         if (il == n_layer - 1 && inp_out_ids) {

src/models/qwen3.cpp

@@ -58,9 +58,6 @@ llm_build_qwen3::llm_build_qwen3(const llama_model & model, const llm_graph_para
             cur = build_attn(inp_attn,
                     model.layers[il].wo, model.layers[il].wo_b, model.layers[il].wo_s,
                     Qcur, Kcur, Vcur, nullptr, nullptr, nullptr, 1.0f/sqrtf(float(n_embd_head)), il);
-            if (model.layers[il].wo_s) {
-                cur = ggml_mul(ctx0, cur, model.layers[il].wo_s);
-            }
         }
         if (il == n_layer - 1 && inp_out_ids) {
             cur   = ggml_get_rows(ctx0,   cur, inp_out_ids);

src/models/qwen3moe.cpp

@@ -58,9 +58,6 @@ llm_build_qwen3moe::llm_build_qwen3moe(const llama_model & model, const llm_grap
             cur = build_attn(inp_attn,
                     model.layers[il].wo, model.layers[il].wo_b, model.layers[il].wo_s,
                     Qcur, Kcur, Vcur, nullptr, nullptr, nullptr, 1.0f/sqrtf(float(n_embd_head)), il);
-            if (model.layers[il].wo_s) {
-                cur = ggml_mul(ctx0, cur, model.layers[il].wo_s);
-            }
         }
         if (il == n_layer - 1 && inp_out_ids) {
             cur   = ggml_get_rows(ctx0,   cur, inp_out_ids);

tests/test-arg-parser.cpp

@@ -40,8 +40,12 @@ int main(void) {
                     }
                 }
 
+                // exclude spec args from this check
+                // ref: https://github.com/ggml-org/llama.cpp/pull/22397
+                const bool skip = opt.is_spec;
+
                 // ensure shorter argument precedes longer argument
-                if (opt.args.size() > 1) {
+                if (!skip && opt.args.size() > 1) {
                     const std::string first(opt.args.front());
                     const std::string last(opt.args.back());
 
@@ -124,9 +128,9 @@ int main(void) {
     assert(params.n_batch == 9090);
 
     // --draft cannot be used outside llama-speculative
-    argv = {"binary_name", "--draft", "123"};
+    argv = {"binary_name", "--spec-draft-n-max", "123"};
     assert(true == common_params_parse(argv.size(), list_str_to_char(argv).data(), params, LLAMA_EXAMPLE_SPECULATIVE));
-    assert(params.speculative.n_max == 123);
+    assert(params.speculative.draft.n_max == 123);
 
     // multi-value args (CSV)
     argv = {"binary_name", "--lora", "file1.gguf,\"file2,2.gguf\",\"file3\"\"3\"\".gguf\",file4\".gguf"};

tests/test-backend-ops.cpp

@@ -2984,7 +2984,7 @@ struct test_bin_bcast : public test_case {
     bool run_whole_graph() override { return nf > 1; }
 
     std::string vars() override {
-        return VARS_TO_STR5(type, ne, nr, nf, perm1);
+        return VARS_TO_STR6(type, ne, nr, nf, perm1, src_overlap);
     }
 
     size_t op_size(ggml_tensor * t) override {
@@ -3579,6 +3579,49 @@ struct test_ssm_conv : public test_case {
     }
 };
 
+// GGML_OP_SSM_CONV + GGML_OP_ADD (channel-wise bias, optional) + GGML_OP_UNARY(SILU) (fused operation)
+struct test_ssm_conv_bias_silu : public test_case {
+    const ggml_type type;
+    const std::array<int64_t, 4> ne_a;
+    const std::array<int64_t, 4> ne_b;
+    const bool fuse_bias;
+
+    std::string op_desc(ggml_tensor * t) override {
+        GGML_UNUSED(t);
+        return "SSM_CONV_BIAS_SILU";
+    }
+
+    bool run_whole_graph() override { return true; }
+
+    std::string vars() override {
+        return VARS_TO_STR4(type, ne_a, ne_b, fuse_bias);
+    }
+
+    test_ssm_conv_bias_silu(ggml_type type, std::array<int64_t, 4> ne_a, std::array<int64_t, 4> ne_b,
+            bool fuse_bias)
+        : type(type), ne_a(ne_a), ne_b(ne_b), fuse_bias(fuse_bias) {}
+
+    ggml_tensor * build_graph(ggml_context * ctx) override {
+        ggml_tensor * a = ggml_new_tensor(ctx, type, 4, ne_a.data());
+        ggml_tensor * b = ggml_new_tensor(ctx, type, 4, ne_b.data());
+        ggml_set_name(a, "a");
+        ggml_set_name(b, "b");
+
+        ggml_tensor * out = ggml_ssm_conv(ctx, a, b);
+
+        if (fuse_bias) {
+            ggml_tensor * bias = ggml_new_tensor_1d(ctx, type, out->ne[0]);
+            ggml_set_name(bias, "bias");
+            out = ggml_add(ctx, out, bias);
+        }
+
+        out = ggml_silu(ctx, out);
+
+        ggml_set_name(out, "out");
+        return out;
+    }
+};
+
 // GGML_OP_SSM_SCAN
 struct test_ssm_scan : public test_case {
     const ggml_type type;
@@ -3589,9 +3632,10 @@ struct test_ssm_scan : public test_case {
     const int64_t n_group;
     const int64_t n_seq_tokens;
     const int64_t n_seqs;
+    const bool    xbc_overlap;
 
     std::string vars() override {
-        return VARS_TO_STR7(type, d_state, head_dim, n_head, n_group, n_seq_tokens, n_seqs);
+        return VARS_TO_STR8(type, d_state, head_dim, n_head, n_group, n_seq_tokens, n_seqs, xbc_overlap);
     }
 
     test_ssm_scan(ggml_type type = GGML_TYPE_F32,
@@ -3600,16 +3644,31 @@ struct test_ssm_scan : public test_case {
             int64_t n_head  = 32,
             int64_t n_group = 1,
             int64_t n_seq_tokens = 32,
-            int64_t n_seqs = 32)
-        : type(type), d_state(d_state), head_dim(head_dim), n_head(n_head), n_group(n_group), n_seq_tokens(n_seq_tokens), n_seqs(n_seqs) {}
+            int64_t n_seqs = 32,
+            bool xbc_overlap = false)
+        : type(type), d_state(d_state), head_dim(head_dim), n_head(n_head), n_group(n_group), n_seq_tokens(n_seq_tokens), n_seqs(n_seqs), xbc_overlap(xbc_overlap) {}
 
     ggml_tensor * build_graph(ggml_context * ctx) override {
         ggml_tensor * s   = ggml_new_tensor_4d(ctx, type, d_state,  head_dim,     n_head,       n_seqs);
-        ggml_tensor * x   = ggml_new_tensor_4d(ctx, type, head_dim, n_head,       n_seq_tokens, n_seqs);
         ggml_tensor * dt  = ggml_new_tensor_3d(ctx, type, n_head,   n_seq_tokens, n_seqs);
         ggml_tensor * A   = ggml_new_tensor_2d(ctx, type, (head_dim > 1) ? 1 : d_state, n_head);
-        ggml_tensor * B   = ggml_new_tensor_4d(ctx, type, d_state,  n_group,      n_seq_tokens, n_seqs);
-        ggml_tensor * C   = ggml_new_tensor_4d(ctx, type, d_state,  n_group,      n_seq_tokens, n_seqs);
+        ggml_tensor * x;
+        ggml_tensor * B;
+        ggml_tensor * C;
+
+        if (xbc_overlap) {
+            ggml_tensor * xbc = ggml_new_tensor_4d(ctx, type, d_state, n_head, n_seq_tokens, 2 * n_seqs);
+            x = ggml_view_4d(ctx, xbc, head_dim, n_head, n_seq_tokens, n_seqs,
+                             xbc->nb[1], xbc->nb[2], xbc->nb[3], xbc->nb[3]);
+            B = ggml_view_4d(ctx, xbc, d_state, n_group, n_seq_tokens, n_seqs,
+                             xbc->nb[1], xbc->nb[2], xbc->nb[3], 0);
+            C = ggml_view_4d(ctx, xbc, d_state, n_group, n_seq_tokens, n_seqs,
+                             xbc->nb[1], xbc->nb[2], xbc->nb[3], 2 * xbc->nb[3]);
+        } else {
+            x = ggml_new_tensor_4d(ctx, type, head_dim, n_head, n_seq_tokens, n_seqs);
+            B = ggml_new_tensor_4d(ctx, type, d_state,  n_group, n_seq_tokens, n_seqs);
+            C = ggml_new_tensor_4d(ctx, type, d_state,  n_group, n_seq_tokens, n_seqs);
+        }
         ggml_tensor * ids = ggml_new_tensor_1d(ctx, GGML_TYPE_I32,  n_seqs);
         ggml_tensor * out = ggml_ssm_scan(ctx, s, x, dt, A, B, C, ids);
         return out;
@@ -3799,7 +3858,7 @@ struct test_mul_mat : public test_case {
 
     double max_nmse_err(ggml_backend_t backend) override {
         // for blackwell we quantize activations to mxfp4 instead of q8_1 so we add higher tolerance
-        if (type_a == GGML_TYPE_MXFP4 && backend_has_feature(backend, "BLACKWELL_NATIVE_FP4")) {
+        if ((type_a == GGML_TYPE_MXFP4 || type_a == GGML_TYPE_NVFP4) && backend_has_feature(backend, "BLACKWELL_NATIVE_FP4")) {
             return 2e-2;
         }
         return max_nmse_err();
@@ -3935,7 +3994,7 @@ struct test_mul_mat_id : public test_case {
 
     double max_nmse_err(ggml_backend_t backend) override {
         // for blackwell we quantize activations to mxfp4 instead of q8_1 so we add higher tolerance
-        if (type_a == GGML_TYPE_MXFP4 && backend_has_feature(backend, "BLACKWELL_NATIVE_FP4")) {
+        if ((type_a == GGML_TYPE_MXFP4 || type_a == GGML_TYPE_NVFP4) && backend_has_feature(backend, "BLACKWELL_NATIVE_FP4")) {
             return 2e-2;
         }
         return max_nmse_err();
@@ -7961,9 +8020,31 @@ static std::vector<std::unique_ptr<test_case>> make_test_cases_eval() {
         }
     }
 
+    // fused ssm_conv + (optional) bias_add + silu. The bias-only graph (no silu) is intentionally
+    // not tested since there's no fusion for that pattern in ggml_cuda_can_fuse.
+    for (int64_t d_conv : {3, 4, 9}) {
+        for (int64_t d_inner : {1024, 1536, 2048}) {
+            for (bool fuse_bias : {false, true}) {
+                // short token path (n_t <= 32)
+                test_cases.emplace_back(new test_ssm_conv_bias_silu(
+                    GGML_TYPE_F32, {d_conv, d_inner, 1, 1}, {d_conv, d_inner, 1, 1}, fuse_bias));
+                test_cases.emplace_back(new test_ssm_conv_bias_silu(
+                    GGML_TYPE_F32, {2 * d_conv, d_inner, 1, 1}, {d_conv, d_inner, 1, 1}, fuse_bias));
+                test_cases.emplace_back(new test_ssm_conv_bias_silu(
+                    GGML_TYPE_F32, {d_conv, d_inner, 4, 1}, {d_conv, d_inner, 1, 1}, fuse_bias));
+                // long token path (n_t > 32)
+                test_cases.emplace_back(new test_ssm_conv_bias_silu(
+                    GGML_TYPE_F32, {d_conv - 1 + 64, d_inner, 1, 1}, {d_conv, d_inner, 1, 1}, fuse_bias));
+                test_cases.emplace_back(new test_ssm_conv_bias_silu(
+                    GGML_TYPE_F32, {d_conv - 1 + 64, d_inner, 4, 1}, {d_conv, d_inner, 1, 1}, fuse_bias));
+            }
+        }
+    }
+
     test_cases.emplace_back(new test_ssm_scan(GGML_TYPE_F32, 16, 1, 1024, 1, 32, 4)); // Mamba-1
     test_cases.emplace_back(new test_ssm_scan(GGML_TYPE_F32, 128, 64, 16, 2, 32, 4)); // Mamba-2
     test_cases.emplace_back(new test_ssm_scan(GGML_TYPE_F32, 256, 64,  8, 2, 32, 4)); // Falcon-H1
+    test_cases.emplace_back(new test_ssm_scan(GGML_TYPE_F32, 128, 128, 4, 4, 16, 2, true)); // x/B/C overlap
 
     test_cases.emplace_back(new test_rwkv_wkv6(GGML_TYPE_F32, 32, 64, 1, 1));
     test_cases.emplace_back(new test_rwkv_wkv6(GGML_TYPE_F32, 32, 64, 32, 1));
@@ -8976,6 +9057,8 @@ static std::vector<std::unique_ptr<test_case>> make_test_cases_perf() {
     // Examples from granite-4.0-h-1b/ggml-model-Q8_0.gguf
     test_cases.emplace_back(new test_ssm_conv(GGML_TYPE_F32, {515, 3328, 1, 1}, {4, 3328, 1, 1})); // prefill
     test_cases.emplace_back(new test_ssm_conv(GGML_TYPE_F32, {4,   3328, 1, 1}, {4, 3328, 1, 1})); // generate
+    test_cases.emplace_back(new test_ssm_conv_bias_silu(GGML_TYPE_F32, {515, 3328, 1, 1}, {4, 3328, 1, 1}, true));  // prefill
+    test_cases.emplace_back(new test_ssm_conv_bias_silu(GGML_TYPE_F32, {4,   3328, 1, 1}, {4, 3328, 1, 1}, true));  // generate
     test_cases.emplace_back(new test_ssm_scan(GGML_TYPE_F32, 128, 64, 48, 1, 512, 1)); // prefill
     test_cases.emplace_back(new test_ssm_scan(GGML_TYPE_F32, 128, 64, 48, 1, 1,   1)); // generate
 

tests/test-chat.cpp

@@ -2249,6 +2249,46 @@ static void test_template_output_peg_parsers(bool detailed_debug) {
             .reasoning_format(COMMON_REASONING_FORMAT_AUTO)
             .expect(message_assist)
             .run();
+
+        {
+            // additional tests for https://github.com/ggml-org/llama.cpp/pull/21760
+            auto tmpls = read_templates("models/templates/google-gemma-4-31B-it.jinja");
+
+            common_chat_msg tool_call_msg = simple_assist_msg(
+                "Let me check.", "", "special_function", "{\"arg1\": 1}","c0");
+
+            common_chat_msg tool_msg;
+            tool_msg.role         = "tool";
+            tool_msg.tool_name    = "special_function";
+            tool_msg.tool_call_id = "c0";
+            tool_msg.content      = "{\"r\":\"ok\"}";
+
+            {
+                common_chat_templates_inputs inputs;
+                inputs.messages              = { message_user, tool_call_msg, tool_msg };
+                inputs.tools                 = { special_function_tool };
+                inputs.add_generation_prompt = true;
+
+                auto params = common_chat_templates_apply(tmpls.get(), inputs);
+
+                if (!string_ends_with(params.prompt, "<turn|>\n<|turn>model\n")) {
+                    throw std::runtime_error("Missing generation prompt for Gemma 4");
+                }
+            }
+
+            {
+                common_chat_templates_inputs inputs;
+                inputs.messages              = { message_user, tool_call_msg, tool_msg };
+                inputs.tools                 = { special_function_tool };
+                inputs.add_generation_prompt = false;
+
+                auto params = common_chat_templates_apply(tmpls.get(), inputs);
+
+                if (string_ends_with(params.prompt, "<|turn>model\n")) {
+                    throw std::runtime_error("Gemma 4: generation prompt was modified despite add_generation_prompt=false");
+                }
+            }
+        }
     }
 
     {

tests/test-log.cpp

@@ -35,5 +35,9 @@ int main() {
         threads[i].join();
     }
 
+    common_log_flush(common_log_main());
+    // We explicitly free the logger singleton to avoid hanging on Windows
+    // related to timing issues of thread startup and DLL teardown
+    common_log_free(common_log_main());
     return 0;
 }

tests/test-reasoning-budget.cpp

@@ -227,7 +227,30 @@ int main(void) {
             3);     // forcing continues through i=3
     }
 
-    printf("OK (5 tests passed)\n");
+    // Test 6: Multi-block thinking. First block ends naturally at i=2, second
+    // start tag at i=3 re-arms the budget, which then exhausts at i=5.
+    // Regression: before this fix, DONE absorbed all subsequent tokens and a
+    // second <think> block ran unbudgeted.
+    // Flow: i=0 accept(100)->COUNTING rem=2; i=1 accept(50)->rem=1;
+    //       i=2 accept(101)->end_matcher matches, DONE;
+    //       i=3 accept(100)->re-arm, COUNTING rem=2;
+    //       i=4 accept(60)->rem=1; i=5 accept(61)->rem=0->FORCING;
+    //       i=6 apply()->forces token[0]=102, accept(62)->force_pos=1, stay FORCING;
+    //       i=7 apply()->forces token[1]=101, accept(63)->force_pos=2->DONE.
+    {
+        const std::vector<llama_token> start = {100};
+        const std::vector<llama_token> end = {101};
+        const std::vector<llama_token> forced = {102, 101};
+        const std::vector<llama_token> sequence = {100, 50, 101, 100, 60, 61, 62, 63};
+
+        test_reasoning_budget("multi-block re-arms budget after DONE", sequence, start, end, forced,
+            2,      // budget of 2 tokens (per block)
+            REASONING_BUDGET_IDLE,
+            6,      // forcing starts at i=6 (after second block exhausts at i=5)
+            7);     // forcing continues through i=7
+    }
+
+    printf("OK (6 tests passed)\n");
 
     printf("Testing UTF-8 boundary detection... ");
     test_utf8_boundary_detection();

tools/llama-bench/llama-bench.cpp

@@ -372,7 +372,7 @@ static const cmd_params cmd_params_defaults = {
     /* n_ubatch             */ { 512 },
     /* type_k               */ { GGML_TYPE_F16 },
     /* type_v               */ { GGML_TYPE_F16 },
-    /* n_threads            */ { cpu_get_num_math() },
+    /* n_threads            */ { common_cpu_get_num_math() },
     /* cpu_mask             */ { "0x0" },
     /* cpu_strict           */ { false },
     /* poll                 */ { 50 },

tools/mtmd/debug/mtmd-debug.cpp

@@ -72,7 +72,7 @@ int main(int argc, char ** argv) {
 
     mtmd::context_ptr ctx_mtmd;
     common_init_result_ptr llama_init;
-    base_callback_data cb_data;
+    common_debug_cb_user_data cb_data;
 
     llama_init = common_init_from_params(params);
     {
@@ -89,7 +89,7 @@ int main(int argc, char ** argv) {
         {
             // always enable debug callback
             mparams.cb_eval_user_data = &cb_data;
-            mparams.cb_eval = common_debug_cb_eval<false>;
+            mparams.cb_eval = common_debug_cb_eval;
         }
         ctx_mtmd.reset(mtmd_init_from_file(clip_path, model, mparams));
         if (!ctx_mtmd.get()) {

tools/mtmd/mtmd-cli.cpp

@@ -90,7 +90,7 @@ struct mtmd_cli_context {
     int n_threads    = 1;
     llama_pos n_past = 0;
 
-    base_callback_data cb_data;
+    common_debug_cb_user_data cb_data;
 
     mtmd_cli_context(common_params & params) : llama_init(common_init_from_params(params)) {
         model = llama_init->model();
@@ -145,7 +145,7 @@ struct mtmd_cli_context {
         mparams.image_max_tokens = params.image_max_tokens;
         if (std::getenv("MTMD_DEBUG_GRAPH") != nullptr) {
             mparams.cb_eval_user_data = &cb_data;
-            mparams.cb_eval = common_debug_cb_eval<false>;
+            mparams.cb_eval = common_debug_cb_eval;
         }
         ctx_vision.reset(mtmd_init_from_file(clip_path, model, mparams));
         if (!ctx_vision.get()) {