Daniel Han b8bdf73bb9 llama: fold the random-access prefetch into its own feature flag ...

LLAMA_MMAP_RANDOM_PREFETCH existed to measure the two halves of the feature
apart, and the measurement is done: on a cold cache over the same wikitext
run, MADV_RANDOM without the batched readahead takes 94.4 s against 36.7 s
for an untouched mapping, while the pair together take 34.1 s. Suppressing
the kernel's readahead only pays if we replace it, so the split let a user
select a 2.6x regression through a documented switch.

Keep the accessor, since the call site reads better than a mode comparison,
but derive it from the mode alone.

2026-08-27 10:01:47 +00:00

..

models

llama: narrow the random-access mmap advice to the gather table

2026-08-27 10:01:47 +00:00

CMakeLists.txt

llama: hold the qwen4exp indexer cache in a new llama_memory_hybrid_idx

2026-08-26 15:09:53 +00:00

llama-adapter.cpp

llama : check LoRA tensor data is within file bounds (#27056)

2026-08-16 09:38:01 +03:00

llama-adapter.h

llama : re-enable manual LoRA adapter free (#19983)

2026-03-18 12:03:26 +02:00

llama-arch.cpp

qwen4exp: hash the image placeholder for multimodal batches

2026-08-26 13:41:50 +00:00

llama-arch.h

qwen4exp: hash the image placeholder for multimodal batches

2026-08-26 13:41:50 +00:00

llama-batch.cpp

llama-batch: fix allowed decreasing pos in a seq (#25449)

2026-07-08 19:24:34 +03:00

llama-batch.h

llama-batch: add n_keep_tail in split_equal for recurrent models (#25278)

2026-07-08 15:55:19 +08:00

llama-chat.cpp

chat : add Granite 4.1 chat template (#23518)

2026-05-28 13:13:33 +02:00

llama-chat.h

chat : add Granite 4.1 chat template (#23518)

2026-05-28 13:13:33 +02:00

llama-context.cpp

llama: give qwen4exp a large-graph node budget

2026-08-27 00:45:01 +00:00

llama-context.h

llama: refactor fused ops (#24646)

2026-07-08 18:18:09 +08:00

llama-cparams.cpp

cparams : rename LLAMA_MAX_PARALLEL_SEQUENCES to LLAMA_MAX_SEQ (#14188)

2025-06-15 10:08:58 +03:00

llama-cparams.h

llama : support multi-output backend sampling (#25532)

2026-08-10 16:58:56 +03:00

llama-ext.h

mtmd: support Qwen3-TTS (note: breaking change to llama-tts binary) (#26254)

2026-08-04 17:26:15 +02:00

llama-grammar.cpp

grammar : parse \- in char classes as literal hyphen (#27591)

2026-08-25 09:05:24 +03:00

llama-grammar.h

common/grammar : replace problematic backtracking regex [\s\S]* (#18342)

2026-01-03 16:02:43 -06:00

llama-graph.cpp

llama: keep the qwen4exp top-k attention mask arch-local

2026-08-26 15:09:53 +00:00

llama-graph.h

llama: keep the qwen4exp top-k attention mask arch-local

2026-08-26 15:09:53 +00:00

llama-hparams.cpp

qwen4exp: give the PLE conv history its own mirrored recurrent row

2026-08-27 06:23:08 +00:00

llama-hparams.h

qwen4exp: shrink the PLE hparams storage

2026-08-27 10:01:47 +00:00

llama-impl.cpp

llama : correct platform-independent loading of BOOL metadata (#21428)

2026-04-06 01:40:38 +02:00

llama-impl.h

llama: refactor fused ops (#24646)

2026-07-08 18:18:09 +08:00

llama-io.cpp

server : avoid checkpoint data host copies (#22558)

2026-05-02 18:03:25 +03:00

llama-io.h

llama : add option to save memory in device buffers (#22679)

2026-05-05 06:35:07 +03:00

llama-kv-cache-dsa-iswa.cpp

model: add dots3-note (#27060)

2026-08-21 19:52:34 +02:00

llama-kv-cache-dsa-iswa.h

model: add dots3-note (#27060)

2026-08-21 19:52:34 +02:00

llama-kv-cache-dsa.cpp

llama : allocate indexer cache only in "full" indexer layers (#26474)

2026-08-03 14:56:30 +02:00

llama-kv-cache-dsa.h

llama : allocate indexer cache only in "full" indexer layers (#26474)

2026-08-03 14:56:30 +02:00

llama-kv-cache-dsv4.cpp

Deepseek 4: -sm tensor (#26490)

2026-08-24 09:20:25 +03:00

llama-kv-cache-dsv4.h

DeepseekV4 MTP + DSpark (#25784)

2026-08-02 20:55:34 +08:00

llama-kv-cache-iswa.cpp

llama-batch: add n_keep_tail in split_equal for recurrent models (#25278)

2026-07-08 15:55:19 +08:00

llama-kv-cache-iswa.h

DeepSeek V4 (#24162)

2026-06-29 16:58:51 +08:00

llama-kv-cache-msa.cpp

model: M3: Move MSA into a new memory implementation (#26338)

2026-08-03 16:30:08 +03:00

llama-kv-cache-msa.h

model: M3: Move MSA into a new memory implementation (#26338)

2026-08-03 16:30:08 +03:00

llama-kv-cache.cpp

kv-cache: check the mirrored slot layout on a whole-context restore too

2026-08-27 06:23:08 +00:00

llama-kv-cache.h

qwen4exp: keep the indexer cache in step across server slots

2026-08-27 06:23:08 +00:00

llama-kv-cells.h

kv-cache : avoid kv cells copies (#24277)

2026-06-07 21:42:54 +03:00

llama-memory-hybrid-idx.cpp

llama: give the qwen4exp full memory context its indexer cache

2026-08-27 10:01:47 +00:00

llama-memory-hybrid-idx.h

llama: give the qwen4exp full memory context its indexer cache

2026-08-27 10:01:47 +00:00

llama-memory-hybrid-iswa.cpp

llama-batch: add n_keep_tail in split_equal for recurrent models (#25278)

2026-07-08 15:55:19 +08:00

llama-memory-hybrid-iswa.h

llama + spec: MTP Support (#22673)

2026-05-16 20:06:23 +08:00

llama-memory-hybrid.cpp

llama: hold the qwen4exp indexer cache in a new llama_memory_hybrid_idx

2026-08-26 15:09:53 +00:00

llama-memory-hybrid.h

llama: hold the qwen4exp indexer cache in a new llama_memory_hybrid_idx

2026-08-26 15:09:53 +00:00

llama-memory-recurrent.cpp

qwen4exp: give the PLE conv history its own mirrored recurrent row

2026-08-27 06:23:08 +00:00

llama-memory-recurrent.h

qwen4exp: give the PLE conv history its own mirrored recurrent row

2026-08-27 06:23:08 +00:00

llama-memory.cpp

memory : correctly handle failure in apply() (#14438)

2025-06-30 18:03:03 +03:00

llama-memory.h

llama : add Gemma4 MTP (#23398)

2026-06-07 20:50:54 +08:00

llama-mmap.cpp

llama: fold the random-access prefetch into its own feature flag

2026-08-27 10:01:47 +00:00

llama-mmap.h

llama: fold the random-access prefetch into its own feature flag

2026-08-27 10:01:47 +00:00

llama-model-loader.cpp

llama: narrow the random-access mmap advice to the gather table

2026-08-27 10:01:47 +00:00

llama-model-loader.h

llama: narrow the random-access mmap advice to the gather table

2026-08-27 10:01:47 +00:00

llama-model-saver.cpp

tests: cover qwen4exp in test-llama-archs

2026-08-26 13:41:50 +00:00

llama-model-saver.h

tests: cover qwen4exp in test-llama-archs

2026-08-26 13:41:50 +00:00

llama-model.cpp

llama: narrow the random-access mmap advice to the gather table

2026-08-27 10:01:47 +00:00

llama-model.h

llama: narrow the random-access mmap advice to the gather table

2026-08-27 10:01:47 +00:00

llama-quant.cpp

quantize: dequantize and quantize large tensors in row bands

2026-08-27 00:45:01 +00:00

llama-quant.h

llama : refactor src/llama.cpp (#10902)

2025-01-03 10:18:53 +02:00

llama-sampler.cpp

llama : support multi-output backend sampling (#25532)

2026-08-10 16:58:56 +03:00

llama-sampler.h

llama : support multi-output backend sampling (#25532)

2026-08-10 16:58:56 +03:00

llama-vocab.cpp

vocab : support integer tokenizer scores (#27260)

2026-08-17 20:13:15 +02:00

llama-vocab.h

Add support for Laguna XS.2 & M.1 (#25165)

2026-07-22 09:54:08 +08:00

llama.cpp

llama : allow virtual igpu devices (#26953)

2026-08-14 15:14:19 +03:00

unicode-data.cpp

server : better security control for public deployments (#9776)

2024-10-08 13:27:04 +02:00

unicode-data.h

llama : reduce compile time and binary size (#9712)

2024-10-02 15:49:55 +02:00

unicode.cpp

unicode : include '~' in collapsed symbol class (#26972)

2026-08-18 15:15:22 +02:00

unicode.h

vocab: fix Gemma4 tokenizer (#21343)

2026-04-03 10:33:03 +02:00