Commit Graph

Select branches

Hide Pull Requests

abstractions_4

add_local_buffer

after_qol

algebraic_upat

all_bv_no_idx

all_sink

amd_isel

amd_refac

amd_smu_reset_patch

amd_sqtt

amd_uop

amd_uop_cleanups

assign_copy

assign_try_3

before2

better_spec

better_syn

bitcast_spec

broadcast_shape_2

bt_ei

buf_heur

bump_nv_2404

cache_reshape

call_inline

call_is_kernels

callify

cast_vec_as_cast

cat_mop

changes_from_rangify

check_clsetarg

check_dims

cifar_30

cifar_30_clean

clean_load

clean_spec

clone_tg

codegen_try_2

contig_on_after

continue_mat

deepsek

default_late_allreduce

delete_kernel_py

delete_ones

delete_slow_rangeify

delta_attention

delta_attention_scan

demote_op

disable_rules

dsp_search

dsp_search_merged

dtype_shape

earlier_gater

emu_cdna

emulate_ctx

end_all_ranges

endrange

er_prioity

execution_unit

failing_assign

fancy_usb

fast_llm

fix_fa_fwd_clang_22

fix_fuse

fix_llvm_cast

fix_range_merging_sd

fix_rangeify_tests

fix_rc

fix_sched_copy

fix_slice_store

fix_src_spec

fixup_fix_kernel

full_spec

fun_w_egraphs

gate-on-load-store

gate_rewrite

gemma4_gpt

gen_pdf_fast

gh-pages

gl_dims

gldims

glm_flash

grad_outerworld

grok_x86_novec

group_div_rules

hcq2_vram_copies

heur_4096

heur_group

hk_gemm

http_threading

hwtest_fixes

id_store

if_range_cleanups

ignore_oob

image_idx_xy

image_in_mc

image_no_vec

image_refactors_stack

index_to_slice

invalid_try_3

jinja_template

kernel4

kernel_is_call

kernelless

kimi_slop

kitten_matmul

late_Locals

late_add_load

late_allreduce

late_codegen

late_decomp_is_late

late_lower_idx

late_view

late_weakint

lds_double_buffering

less_wmma_arg

lil_dg_change

lil_rangeify_td

llama_trainer

llm_262k_minimal

llm_flash_attn

llm_qwen

llm_serve_fixes

load_view_late

local_cache

locals_have_device

loop_end_op

loop_is_rng

lowerer_hacks

mac_pytest

manual_kernel_3

master

mega_lowerer

min_new_lin

min_outer_range

minor_views

mixins7

mlperf_training_v4.1

mlperf_training_v5.0

mlperf_training_v6.0

modern_scan

more_apps

more_early_comps

more_folding_rules

more_kernelcount

more_pcode

more_rangeify_pm_tests

more_ren_clean

more_uop_gc_test

more_uop_mm

more_vmap

move_devec

move_dtype_decomps

move_expander

move_gates_to_load_store

move_mop_cleanup

move_schedule

move_simp_views

moveleftright

mt_typing

multi_range

multioutput

new_ck

new_devec

new_invalid_and_image

new_linearizer

new_shape

new_test

no_assign_2

no_buf_0

no_decimals

no_late_simplify_marg

no_merge_ends

no_merge_views

no_mstack_in_allreduce

no_one_ranges

no_sip_usbgpu

no_tsink_base

no_tuplize

no_user_ck_contig

no_vec_x86_4

no_weakint_variable

noop_index

novecimg

num_batches_tracked_shape

olmo3_support

one_compiler

one_hcopt

only_assign_buffers

only_loop_backends

only_reg_emu

only_reg_emu2

outer_range

outerworld_work

pad_invalid_work

pad_mixin

param_call_is_function

parse_sqtt_more

pci_iommu

port_nir

postrange_hcopts

precompile_backward

prefill_qwen_fast

print_special_ops

ps_clean

pyremu_try

python_prg

python_speed

q4k_support

qcom-openpilot-13ms

qcom_830

qcom_mmapeak

qcom_speed_codex

qol_string

qwen36_27b_amd_900

qwen36_27b_amd_minimal

qwen36_amd_bugfixes

qwen_mergable

r_cleanups

r_new_opt

ranged_call

rangeify_2

rangeify_const_folding

rdna3_vibes

rdna4_asm

rdna4_gemm

real_pyrender

recurse_sub

reg_coalese

relu_pattern

remove_assign

remove_cast_hack

remove_ck_contig

remove_define_var

remove_forced_re

remove_on_stack

remove_programspec

remove_pylint

remove_slow_tests

remove_vec_2

remove_vec_3

renderers_use_shapes_2

replace_if_with_range

reprocess_node_experiments

reshape_trait

resnet-ci

revert-11547-fix_mismatch_red

revert-11715-revert-11711-amd_llvm_tc_cast

revert-11809-no_ki_in_gpudims

revert-11811-better_div_nesting

revert-11876-memplan_vbig

revert-12104-simplify_valid_in_symbolic

revert-12308-bufferize_early

revert-12731-jitlink_init_check

revert-12813-after_cleanups

revert-13046-better_cleanup_arange_buffers

revert-13311-prof_link_sqtt

revert-13583-rocr_desc_2

revert-14734-tload

revert-17141-full_rewrite_to_sink

revert-17408-_broadcasted-weak

revert-9438-simple_matmul_np_rand

revert_wrong_ck_fix

rewrite_rangeify2

rtoposort

scan_assign

scan_op_8

sched_cache

sched_lin

sd_load_simple

self_type

set_testpath

shrink_in_render

simp_spec

simple_gidxs

simple_pool

simple_priority

simple_rangeify

simpler-sd-compile

simpler_end

simpler_fusion

simpler_image

simpler_postrange

simpler_reduce

slice_like_shrink

spec_3

spec_cleanup

spec_ren

speed_one_linear

split_buf_idx

split_decomp_pass

split_sub

stack_0

store_is_void

string_ranges

support_opts_in_contig

swizzle_in_kernel

sym_llm

sym_work

symbolic-chunk-split

szdiff_bot_test

tc_is_heur

tensor_graph_single_sink

test_amd_crash

test_fa

test_from_postopt

test_hgemm

test_unit

tiny_scan

tmr

tool_call_support

topoprop

ttomsa-new_linearizer

tuplegtuple

two_stage_remove

typed_checks

typing_work

unused_permute_arg

uop_prg

upd_test4

update_benchmark

update_benchmark_search

update_mlperf

user_failure

vecless_x86_glm

verify_sink_dims_clean

vf_ish

view_after_const

view_in_codegen

viz_cbroadcast

warp_fun

wgpu-f16-packed

working_postopt

x86_moves

x86_numel

v0.10.0

v0.10.1

v0.10.2

v0.10.3

v0.11.0

v0.12.0

v0.13.0

v0.4.0

v0.5.0

v0.6.0

v0.7.0

v0.8.0

v0.9.0

v0.9.1

v0.9.2

Mono Color

• 391d14438e DEFINE_VAR prereqs for VALID [run_process_replay] (#6637) qazalandGitHub 2024-09-21 13:28:39 +08:00
• c1bbd15bd9 Sharded SDXL Inference (#6328) 3 people 2024-09-21 01:26:43 -04:00
• b91aa1c3d1 validhack cleanup parse valid [run_process_replay] (#6635) chenyuandGitHub 2024-09-20 11:51:56 -04:00
• 683857de5d make match a method on UPat [run_process_replay] (#6634) George HotzandGitHub 2024-09-20 20:00:03 +08:00
• dbe890b358 lowerer don't re-init UOp if we aren't rewriting [run_process_replay] (#6633) qazalandGitHub 2024-09-20 19:26:49 +08:00
• 21f2d79461 qcom match gpu impl for reg a6xx_sp_cs_unknown_a9b1 (#6631) nimlgenandGitHub 2024-09-20 18:14:00 +08:00
• 053c4dee55 qcom test for image pitch (#6621) nimlgenandGitHub 2024-09-20 18:13:48 +08:00
• 37ddd971e6 validhack explicity check valid has CMPLT [run_process_replay] (#6630) chenyuandGitHub 2024-09-20 06:13:05 -04:00
• 581a389a58 limit ctx tracking in TrackedPatternMatcher [run_process_replay] (#6629) qazalandGitHub 2024-09-20 18:06:05 +08:00
• 641586cb87 qcom updated ioctl (#6627) nimlgenandGitHub 2024-09-20 18:02:40 +08:00
• 98644a047b use UOp.define_var in Variable shape [run_process_replay] (#6626) qazalandGitHub 2024-09-20 17:58:29 +08:00
• acef3e67fa add an example that idx is const and valid cannot be removed (#6625) chenyuandGitHub 2024-09-20 05:46:27 -04:00
• 5707503048 x//a<b -> x <a*b for positive a (#6622) chenyuandGitHub 2024-09-20 04:38:47 -04:00
• 72c7087420 viz add kernel code (#6620) qazalandGitHub 2024-09-20 16:31:47 +08:00
• 2dfb1e022c UOp st prereqs for valid [run_process_replay] (#6618) qazalandGitHub 2024-09-20 15:55:35 +08:00
• 74f8f86631 viz kernel tree view (#6614) qazalandGitHub 2024-09-20 15:52:12 +08:00
• b14c1bc417 UOps.RANGE is_increasing (#6615) chenyuandGitHub 2024-09-20 03:14:52 -04:00
• 76aa6416d7 qcom: add disassembler with DEBUG >= 5 Comma Device 2024-09-20 07:04:28 +00:00
• 036c2f5b26 validhack use the new style ge for upper bound valid (#6612) chenyuandGitHub 2024-09-19 23:45:42 -04:00
• c4d5575c61 beat mlx at resnet 18 (#6611) George HotzandGitHub 2024-09-20 11:28:01 +08:00
• 785aaec67c make VIZ more responsive for big graphs (#6610) qazalandGitHub 2024-09-20 08:56:13 +08:00
• 78699d9924 15% more folder speed [run_process_replay] (#6607) George HotzandGitHub 2024-09-19 22:34:42 +08:00
• a37e92081a fix unrolled arange folding (#6606) chenyuandGitHub 2024-09-19 09:03:01 -04:00
• eebd23155c move scheduler rewrites into full_ast_rewrite [run_process_replay] (#6609) qazalandGitHub 2024-09-19 20:03:28 +08:00
• 31748c72c4 refactor viz to parse_qs (#6608) qazalandGitHub 2024-09-19 19:51:41 +08:00
• 944cc46e11 qcom fix image pitch (#6600) nimlgenandGitHub 2024-09-19 18:50:02 +08:00
• a1a882b006 arange folding with new ge (#6604) George HotzandGitHub 2024-09-19 18:01:28 +08:00
• 224151a958 update indexing with UPat.any [run_process_replay] (#6605) George HotzandGitHub 2024-09-19 17:40:17 +08:00
• d148a62f8d more generic simplify_valid_image_load (#6603) chenyuandGitHub 2024-09-19 05:33:37 -04:00
• 718ecad2ee add UPat.any support [run_process_replay] (#6602) George HotzandGitHub 2024-09-19 17:11:24 +08:00
• d06b36e527 viz open UPat links in editor (#6601) qazalandGitHub 2024-09-19 16:48:09 +08:00
• 94effe2a71 simple VIZ=1 and get_location changes (#6599) qazalandGitHub 2024-09-19 15:58:33 +08:00
• eeee032b14 tiny cleanup of test_image_valid (#6597) chenyuandGitHub 2024-09-19 03:09:47 -04:00
• 012a2c449a fix lt_folding VCONST issue [run_process_replay] (#6424) George HotzandGitHub 2024-09-19 14:59:20 +08:00
• 309ea63c03 include cached replaces in VIZ=1 (#6596) qazalandGitHub 2024-09-19 14:48:31 +08:00
• 44c18a39a5 fix upat .location for the type verifier (#6592) qazalandGitHub 2024-09-19 14:13:12 +08:00
• 496806ce75 another example of openpilot conv with valid (#6595) chenyuandGitHub 2024-09-19 01:54:01 -04:00
• 0c9b7c9167 more detailed UPat view in VIZ (#6594) qazalandGitHub 2024-09-19 13:18:11 +08:00
• 5e358cf179 qcom set ctx prio (#6593) nimlgenandGitHub 2024-09-19 12:30:00 +08:00
• 7f9fd556b0 _min_max for WHERE (#6564) chenyuandGitHub 2024-09-18 23:47:48 -04:00
• 1b6eee02ad failed test case for openpilot validhack conv (#6590) chenyuandGitHub 2024-09-18 23:12:30 -04:00
• dfcc9c9aa3 remove unused view.expr [run_process_replay] (#6591) George HotzandGitHub 2024-09-19 11:09:42 +08:00
• e015b41ce9 remove e( function just alu( [run_process_replay] (#6589) George HotzandGitHub 2024-09-19 10:24:02 +08:00
• fa0f678d5a use the PatternMatcher to validate UOps type [run_process_replay] (#6583) George HotzandGitHub 2024-09-19 09:59:06 +08:00
• d01e011a8c start multi graph VIZ=1 (#6587) qazalandGitHub 2024-09-19 08:31:56 +08:00
• 5a7cb8d5a5 qcom set power to max (#6578) nimlgenandGitHub 2024-09-18 18:27:06 +08:00
• bd40a26b8b image valid test case that current approach does not work (#6584) chenyuandGitHub 2024-09-18 06:06:03 -04:00
• 1ec6bd5125 restructure simplify_valid_image_load [run_process_replay] (#6581) chenyuandGitHub 2024-09-18 04:46:41 -04:00
• d02bb270b7 add copyin copyout for image on GPU [run_process_replay] (#6580) George HotzandGitHub 2024-09-18 16:06:20 +08:00
• 162ead02a9 remove LOAD where valid is an empty set (#6579) chenyuandGitHub 2024-09-18 03:49:41 -04:00
• d4b662c318 new openpilot compile (#6573) George HotzandGitHub 2024-09-18 14:22:50 +08:00
• c3a70dbf0d 20 jitted steps in openpilot benchmark (#6577) chenyuandGitHub 2024-09-18 02:15:16 -04:00
• a72d51e277 brute force VALIDHACK matching (#6575) chenyuandGitHub 2024-09-18 01:59:50 -04:00
• d8e5d5c663 move VIZ=1 tests to fuzzers (#6574) qazalandGitHub 2024-09-18 12:12:03 +08:00
• ca8bad90a1 Fix typo in ops.py (#6572) ethanreidelandGitHub 2024-09-17 18:25:00 -07:00
• 9894f20684 dsp offset buffer (#6570) nimlgenandGitHub 2024-09-17 23:34:21 +08:00
• 28e565dc0d prune independent kernels for openpilot [run_process_replay] (#6569) George HotzandGitHub 2024-09-17 20:02:38 +08:00
• 9295bc0189 viz more work [run_process_replay] (#6568) qazalandGitHub 2024-09-17 19:27:09 +08:00
• 455a27dd43 start viz unittests (#6550) qazalandGitHub 2024-09-17 18:58:23 +08:00
• 67a03e72bb remove expr_idxs [run_process_replay] (#6567) George HotzandGitHub 2024-09-17 18:34:51 +08:00
• 9ebbedc37f hotfix: remove expr_idxs from graph geohot 2024-09-17 18:01:17 +08:00
• b947db3de1 don't fold mul mod for common factor (#6566) chenyuandGitHub 2024-09-17 06:01:27 -04:00
• a2f446653e add swizzle_st [run_process_replay] (#6561) qazalandGitHub 2024-09-17 15:37:39 +08:00
• f214bb140d test: relax tolerance of test_broadcastdot (#6560) Gaétan LepageandGitHub 2024-09-17 09:26:39 +02:00
• 5fb877c78c generic valid match criteria of #6552 (#6558) chenyuandGitHub 2024-09-17 02:40:36 -04:00
• 0ab06d5840 push geps through wmma (#6559) George HotzandGitHub 2024-09-17 14:38:40 +08:00
• 5a30a32af8 small viz fixups from the swizzle pads branch [run_process_replay] (#6557) qazalandGitHub 2024-09-17 14:37:53 +08:00
• ffce3ed896 add some new rules (#6555) George HotzandGitHub 2024-09-17 13:59:55 +08:00
• c62b6fd8f0 match any statement in valid for simplification (#6554) chenyuandGitHub 2024-09-17 01:39:47 -04:00
• 006c7c5747 remove unused rules in new expand [run_process_replay] (#6553) George HotzandGitHub 2024-09-17 13:18:18 +08:00
• a2239c812e minimum new style expand (#6534) George HotzandGitHub 2024-09-17 13:02:41 +08:00
• f5dd25d376 enable whisper batch for long sequences (#6458) 3 people 2024-09-17 06:42:10 +02:00
• 7c942418a1 other side of simple out of bound valid case (#6552) chenyuandGitHub 2024-09-16 23:57:15 -04:00
• aeaf7894a7 more generic version of #6548 (#6549) chenyuandGitHub 2024-09-16 23:17:16 -04:00
• 596f41eb46 simple drop image valid case (#6548) chenyuandGitHub 2024-09-16 22:54:07 -04:00
• 798be6bb74 add gated read_image count in openpilot compile2 (#6546) chenyuandGitHub 2024-09-16 21:17:00 -04:00
• 665b4203f8 dsp power managment (#6544) nimlgenandGitHub 2024-09-16 23:34:01 +08:00
• 25d8f3046a dsp do not flush libs to ds (#6531) nimlgenandGitHub 2024-09-16 16:42:15 +08:00
• dae3615008 replace viz graph when it's sink (#6541) qazalandGitHub 2024-09-16 16:00:27 +08:00
• 2a5a53c3db remove extra scheduler graph call, VIZ does this [run_process_replay] (#6540) qazalandGitHub 2024-09-16 14:52:50 +08:00
• c1b2472dea reorder alu/vectorize (#6538) George HotzandGitHub 2024-09-16 14:28:14 +08:00
• 42ba887daa remove logic to vectorize reduces (#6536) George HotzandGitHub 2024-09-16 14:04:48 +08:00
• 607113fcdf fix vectorized dtype repr [run_process_replay] (#6535) qazalandGitHub 2024-09-16 13:42:55 +08:00
• 9b9b83b8b0 viz tests (#6532) qazalandGitHub 2024-09-16 13:08:42 +08:00
• 07bd6e070d add more uops tests for vmin/vmax/const_factor/divides (#6533) George HotzandGitHub 2024-09-16 13:06:31 +08:00
• c447ec2190 Fix amx shape [run_process_replay] (#6524) ignaciosicaandGitHub 2024-09-15 22:49:55 -03:00
• 1683b274b6 main example we want the valid removed (#6527) chenyuandGitHub 2024-09-15 21:49:10 -04:00
• e1b21879a7 minor changes from new expand [run_process_replay] (#6528) George HotzandGitHub 2024-09-16 09:48:37 +08:00
• 3450382a77 Don't re-check patterns when uop.arg is None (#6525) Tim BeckerandGitHub 2024-09-15 18:46:59 -07:00
• a104ecf79b refactor for SWIZZLE with different st dims [run_process_replay] (#6526) qazalandGitHub 2024-09-16 09:41:03 +08:00
• 21835fc08c more graph rewrite tests (#6521) George HotzandGitHub 2024-09-16 09:20:54 +08:00
• 6be0cc387c _get_add_chain(x) -> _get_chain(x, BinaryOps.ADD) (#6523) chenyuandGitHub 2024-09-15 10:54:13 -04:00
• b2c286f567 fix typing for test_ops (#6520) chenyuandGitHub 2024-09-15 06:18:36 -04:00
• cd90092f14 graph rewrite tests (#6519) George HotzandGitHub 2024-09-15 17:29:16 +08:00
• 89b950c6b3 viz more work (#6517) qazalandGitHub 2024-09-15 16:42:17 +08:00
• f69251c6b4 assert pyint in linearize_uop [run_process_replay] (#6518) qazalandGitHub 2024-09-15 16:29:05 +08:00
• 5132bab48d hotfix: add TYPED=1 support geohot 2024-09-15 14:44:26 +08:00
• 2d53e47b14 refactor viz saved context (prereq for tree view) (#6516) qazalandGitHub 2024-09-15 14:08:55 +08:00
• 893a24f60f viz minor stuff (#6515) qazalandGitHub 2024-09-15 12:05:31 +08:00
• d0262ac6ab make ScheduleItem hashable [run_process_replay] (#6512) qazalandGitHub 2024-09-14 18:31:33 +08:00