Commit Graph

Select branches

Hide Pull Requests

giga

main

Mono Color

• 359dd67287 regex ir opt code rewrite gen2, window scanning Youmu Liu 2026-06-26 19:07:16 +08:00
• 13fcfe4a4c regex ir opt code rewrite gen2, window scanning Youmu Liu 2026-06-26 18:41:50 +08:00
• 5cd2b8b4ba regex ir opt code on rewrite Youmu Liu 2026-06-25 19:52:11 +08:00
• 45f9b04b6d add more ir opt code, clearified what to opt, so current opt code may need rewrite Youmu Liu 2026-06-24 18:09:05 +08:00
• 2b05d626c7 begin regex ir opt Youmu Liu 2026-06-23 20:08:33 +08:00
• 7e5e4d1f7a discuss regex ir opt Youmu Liu 2026-06-22 18:18:16 +08:00
• 062ffacf88 regex ir emitter is working Youmu Liu 2026-06-18 19:20:14 +08:00
• 8114ef4bf3 add essential typst file Youmu Liu 2026-06-18 19:12:38 +08:00
• eeca5721fb regex ir emitter is working Youmu Liu 2026-06-18 19:10:30 +08:00
• 5570c6f338 add COUNT and CONCAT ir emitter Youmu Liu 2026-06-17 18:32:55 +08:00
• a7bb3209dd add ir emitter Youmu Liu 2026-06-16 21:04:51 +08:00
• 3cfd29f793 Merge branch 'main' of https://github.com/Nitload-NSI/CCD_Github into local merge missing commits Youmu Liu 2026-06-16 14:15:15 +08:00
• 2c2997ef2e before merge Youmu Liu 2026-06-16 14:15:10 +08:00
• 61b4c6012d add a merge tail alt operation Nitload 2026-06-16 00:23:43 +08:00
• 861b50575a ngra string regex use ir as jit engine as well Youmu Liu 2026-06-15 18:08:14 +08:00
• 2741b34583 left and right use diff pack kernel Youmu Liu 2026-06-12 18:25:04 +08:00
• 06a119cc6f make a trial on packing GEMM kernel Youmu Liu 2026-06-11 20:03:02 +08:00
• 6b449a341e all mat kernel func use new task pack convension Youmu Liu 2026-06-11 17:27:44 +08:00
• 09c1cace48 remove mat panel, mat_task goes on span like Youmu Liu 2026-06-10 18:10:52 +08:00
• 0356f5aff8 mat task adapts L4 Youmu Liu 2026-06-09 18:13:36 +08:00
• 413e41db45 other gemm kernel use half-vector-machine style Youmu Liu 2026-06-08 18:09:15 +08:00
• 9d5c0fd8b4 other gemm kernel use half-vector-machine style Youmu Liu 2026-06-08 16:18:03 +08:00
• 1c515ea2ae FMA-avx512-f64 kernel use half-vector-machine design Nitload 2026-06-08 01:02:34 +08:00
• 4336b0ca5a Move some marcros from x86 ro base head Youmu Liu 2026-06-05 13:32:59 +08:00
• 94b2d759ae Gemm/Fma kernel on avx2 use new solution, with less permute and shuff instruction Youmu Liu 2026-06-04 21:05:38 +08:00
• 2690cf9341 Gemm/Fma kernel on avx512 use new solution, with less permute and shuff instruction Youmu Liu 2026-06-04 16:43:17 +08:00
• 6c38bcfade Find solution to fma, change some type define macros Youmu Liu 2026-06-03 18:06:37 +08:00
• 9f3fc3b345 Compress L4 metadata into 32 bytes Youmu Liu 2026-06-02 18:27:58 +08:00
• 07f8178431 Remove id from L4 head, and other some useless types. Nitload 2026-06-02 00:35:08 +08:00
• 9a09d32960 TLSF in l4 use a shorter head, but may be useless Nitload 2026-06-01 00:36:46 +08:00
• ddb59c38ce Little change in C code, remove isa specific code in std header Youmu Liu 2026-05-28 18:44:53 +08:00
• 15f90aed29 L4 is now a burden Youmu Liu 2026-05-21 19:40:00 +08:00
• 0681237c3b L4 is now a burden Youmu Liu 2026-05-20 18:07:54 +08:00
• 4694fd9433 L4 is now a burden Youmu Liu 2026-05-20 18:07:21 +08:00
• 53ed4de530 L4 is now a burden Youmu Liu 2026-05-19 18:09:24 +08:00
• f3cfa37b1c working on L4's tlsf Youmu Liu 2026-05-15 18:53:58 +08:00
• b87943495d working on L4's tlsf Youmu Liu 2026-05-14 20:15:07 +08:00
• c9fb11b47b working on L4's tlsf Youmu Liu 2026-05-13 20:13:37 +08:00
• ae57de4040 tlsf is ok, and fast Youmu Liu 2026-05-13 15:08:30 +08:00
• bf9879f59f tlsf briefly ok, debugging Youmu Liu 2026-05-12 18:25:48 +08:00
• a99b19fa3f add libs' cmakelists file Youmu Liu 2026-05-11 18:11:28 +08:00
• b59d414e20 Remove hlsf, TLSF.alloc is ok Youmu Liu 2026-05-11 18:08:35 +08:00
• 5706c89563 Remove ignored files Youmu Liu 2026-05-11 09:30:04 +08:00
• 6baae849f8 Improve .gitignore Youmu Liu 2026-05-11 09:29:18 +08:00
• 8105af6ca3 Update on 20260511-001810 by Nitload Youmu. Additional Message : Abort HLSF design, use a non-standard TLSF as instead Nitload 2026-05-11 00:18:10 +08:00
• 5e73469c2e after merge Nitload 2026-05-10 06:39:52 +08:00
• 9234dbe99b Abort hlsf, goes back to tlsf, may use 'Arena-ed tlsf' Youmu Liu 2026-05-09 18:27:34 +08:00
• 942933dcac Abort hlsf, goes back to tlsf, may use 'Arena-ed tlsf' Youmu Liu 2026-05-09 18:18:57 +08:00
• d1aca8f745 hlsf goes bad Youmu Liu 2026-05-08 18:09:12 +08:00
• 156f6afe85 after merge Nitload 2026-05-04 17:25:01 +08:00
• 8aa319a120 going on reconstruct of HLSF Youmu Liu 2026-05-07 18:03:33 +08:00
• c343e763ed before rewrite of HLSF Youmu Liu 2026-05-07 11:16:41 +08:00
• e636410ba6 mat queue and mat predictor confuses. i need time to design and organize Youmu Liu 2026-05-06 18:07:31 +08:00
• e1167bb071 before merge Nitload 2026-05-04 17:22:29 +08:00
• 1338bf2801 add design of .ngra file Youmu Liu 2026-04-30 20:38:44 +08:00
• 37a96ba4fc matrix parallel start to use two different policy Youmu Liu 2026-04-29 20:01:07 +08:00
• ae664cb541 matrix parallel start to use two different policy Youmu Liu 2026-04-27 20:08:10 +08:00
• ecb1f04770 matrix thread pool use ParaWaveFront and heavy sync Youmu Liu 2026-04-26 17:38:10 +08:00
• 251762791e L4 will not main stream in matrix schedualing Youmu Liu 2026-04-24 17:31:49 +08:00
• 41350b0487 add a pre-kP evaluation function in c Youmu Liu 2026-04-24 15:46:58 +08:00
• 2f6617447c L4 is now on new algo, using tricky way to init entryies Youmu Liu 2026-04-23 18:07:14 +08:00
• 1a6ce24a27 L4 is now on new algo, using new eviction arrays, using union of hlsf and cache line head. Youmu Liu 2026-04-23 17:39:38 +08:00
• 5252be9436 L4 removes RRPV array, will not support cross-predictor reuse, will use out-line predictcount and map count array, will not use inline predict count. Youmu Liu 2026-04-22 18:11:00 +08:00
• 4f3e206dcb HLSf use nuint rather than long, L4 now uses hlsf rather than array to alloc cache line. Youmu Liu 2026-04-21 20:32:28 +08:00
• 2e2d64a2ce HLSf use nuint rather than long, L4 now uses hlsf rather than array to alloc cache line. Youmu Liu 2026-04-20 18:10:14 +08:00
• 1b7c929be4 Merge branch 'main' of https://github.com/Nitload-NSI/CCD_Github into local Youmu Liu 2026-04-20 13:31:40 +08:00
• 501bd583e2 before merge Youmu Liu 2026-04-20 13:31:36 +08:00
• 5fc8e62f28 solution to length unfixed cacheline in l4 Nitload 2026-04-19 02:30:35 +08:00
• f2560f3d5b need a new design of L4 Youmu Liu 2026-04-17 18:13:45 +08:00
• f999e88629 can select most spared numa node when schedualing Youmu Liu 2026-04-16 18:20:53 +08:00
• 2203a46eb7 can create and retire matrix thread by numa infomation Youmu Liu 2026-04-15 18:08:08 +08:00
• fdf11c729f find a way to create Thread binded to LogicalCore properly Youmu Liu 2026-04-14 18:08:22 +08:00
• b07013a003 find a way to create Thread binded to LogicalCore properly Youmu Liu 2026-04-13 19:01:28 +08:00
• 61a5b49c22 solution to unsafe operation in l4 Nitload 2026-04-13 00:12:29 +08:00
• 556008ad8c L4 is now very unsafe Nitload 2026-04-12 23:55:18 +08:00
• 9aa096dc63 L4 completely add support for multi predictor, working on ver 5 mat para Youmu Liu 2026-04-10 18:06:08 +08:00
• 8a66c1697a L4 completely add support for multi predictor, working on ver 5 mat para Youmu Liu 2026-04-10 18:05:40 +08:00
• b5bee5c442 fix M512 member index bug Youmu Liu 2026-04-09 18:10:11 +08:00
• f7187ca7fd add multi predictor support to l4, working for ver 5 of mat parallel Youmu Liu 2026-04-09 18:07:49 +08:00
• 0bad6bb480 few changed Youmu Liu 2026-04-08 18:21:00 +08:00
• 9fb08b672d finally find pos to inject prediction to predictor Nitload 2026-04-08 00:25:21 +08:00
• 94a3127c0d small changed Youmu Liu 2026-04-07 18:08:34 +08:00
• 7bf2104edb remove folder Youmu Liu 2026-04-07 08:59:41 +08:00
• 12a216d404 new compute kernel design, less mem read delay Nitload 2026-04-07 01:04:05 +08:00
• d702b3224c adjust for clang-gcc Youmu Liu 2026-04-03 18:40:11 +08:00
• 2f74e5e8a9 add startup numa sense Youmu Liu 2026-04-02 18:13:32 +08:00
• aa7394503e Initial plan copilot-swe-agent[bot] 2026-04-02 02:12:07 +00:00
• ba30c02cb3 advanced userinterface design Nitload 2026-04-01 22:22:22 +08:00
• aa7ece4ed0 advanced userinterface design Nitload 2026-04-01 22:20:49 +08:00
• 8f515a2a50 few change Youmu Liu 2026-04-01 18:03:04 +08:00
• 24498c2717 matrix pool init ok, working on task publish Nitload 2026-03-30 01:42:09 +08:00
• 61303635ad mat queue find solution, native supports clang Youmu Liu 2026-03-27 18:14:54 +08:00
• 06e3aa119b mat Youmu Liu 2026-03-26 18:02:13 +08:00
• d960d02972 back to matrix, simplified thread init Youmu Liu 2026-03-24 13:45:14 +08:00
• eee88e1098 back to matrix Youmu Liu 2026-03-23 20:11:08 +08:00
• 0ac7b2b71f back to matrix, but trouble in combining L4 predict with matrixparallelqueue Youmu Liu 2026-03-20 20:04:45 +08:00
• 48c308ae3c find solution to combinition of predictor and queue Nitload 2026-03-20 00:15:58 +08:00
• dcc2bb699a merge to main stream Nitload 2026-03-19 22:03:30 +08:00
• e521bed222 back to matrix, but trouble in combining L4 predict with matrixparallelqueue Youmu Liu 2026-03-19 19:53:22 +08:00
• fd82829788 L4 briefly pass unit test Youmu Liu 2026-03-18 18:00:34 +08:00