Commit Graph

6388 Commits

Author SHA1 Message Date
Comma Device ad119af6a5 compile fixes 2024-11-11 15:58:27 -05:00
nimlgen 408c0a5e7f qcom match texture/sampler descriptors to OpenCL (#7622)
* qcom ioctl compare more regs

* bug fix
2024-11-11 15:57:42 -05:00
nimlgen 45db7d9045
fuzz qcom vs opencl (#7130)
* fuzz qcom vs opencl

* fix nv

* bettre?

* typo

* open both devs
2024-10-17 18:49:08 +03:00
qazal 188eef959d
early rewrite UOps.CONTIGUOUS (#7132)
* early rewrite UOps.CONTIGUOUS

* add metaops too

* just the contig diff
2024-10-17 18:35:19 +03:00
chenyu 287a198c4f
increase test_strongly_connected_DAG threshold (#7131)
flaky
2024-10-17 11:08:50 -04:00
George Hotz c23ef7e2f8
real_remove_const (#7128) 2024-10-17 21:58:41 +08:00
qazal 2087abc999
get membufs with dedup [pr] (#7127) 2024-10-17 16:06:06 +03:00
George Hotz be9a433a60
fix a bug in flops counting + touchups [pr] (#7126) 2024-10-17 21:02:11 +08:00
qazal a2eefa6f97
move assign st override to upat (#7122)
* move assign st override to upat

* merge view
2024-10-17 13:33:37 +03:00
George Hotz ded1b38b84
minor dtype cleanup [pr] (#7124)
* minor dtype cleanup [pr]

* use ptr() function
2024-10-17 17:41:23 +08:00
George Hotz 0b2621f63f
improve render_dtype [pr] (#7117)
* improve render_dtype [pr]

* don't deref in index
2024-10-17 14:50:40 +08:00
George Hotz ca0dca35f7
move ptx renderer [pr] (#7118) 2024-10-17 14:50:32 +08:00
George Hotz d990a16326
fix tests to use render (#7116) 2024-10-17 14:35:22 +08:00
George Hotz 9f4ca88218 hotfix: relax target pct for beautiful_mnist 2024-10-17 12:36:07 +08:00
chenyu 51cd0e7c0d
idx_given_valid -> uop_given_valid [pr] (#7110)
will reuse this to simplify valid independent of idx
2024-10-16 18:16:36 -04:00
chenyu 842fe444df
test case for valid only simplification (#7108) 2024-10-16 16:40:46 -04:00
chenyu 9d109c5382
remove outdated symbolic comments (#7105) 2024-10-16 14:51:59 -04:00
Francis Lata 90eff347e2
tinytqdm write support (#6359)
* add write support

* add test

* update test case to compare write outputs

* assert final write output

* flush when using write

* update write logic

* Revert "update write logic"

This reverts commit 5e0e611b46cde7a22e41aa5770bc4ccad20de073.

---------

Co-authored-by: chenyu <chenyu@fastmail.com>
2024-10-16 14:51:41 -04:00
nimlgen d1094fce5e
amd reports on hang (#7101) 2024-10-16 21:32:44 +03:00
nimlgen 39ab67e9ef
beam capture and replay in fuzz (#7099)
* beam capture and reply in fuzz

* clean a bit
2024-10-16 20:26:58 +03:00
George Hotz eac58eaaba
no SIGALRM on windows [pr] (#7104) 2024-10-17 00:21:04 +08:00
qazal 9d5342976e
skip TestSetitem.test_setitem_inplace_operator (#7103) 2024-10-16 18:59:08 +03:00
nimlgen 83e7dbd89e
nv fix reallocation local memory when oom (#7098) 2024-10-16 18:17:50 +03:00
George Hotz cd61e81f55
beautiful mnist works on windows (#7100)
* beautiful mnist works on windows [pr]

* add comment for that (no pr)
2024-10-16 23:00:05 +08:00
qazal fc306ba89d
simple to_uop [pr] (#7096)
* simple to_uop [pr]

* fix metadata
2024-10-16 15:52:11 +03:00
qazal 568a4b5d92
viz touchups [pr] (#7095)
* viz touchups [pr]

* check if port busy

* url
2024-10-16 15:14:47 +03:00
qazal 6172b42140
all views are UOps.VIEW [pr] (#7090)
* all views are UOps.VIEW

* is it you

* don't recreate st uop [pr]

* first rewrite all elementwise
2024-10-16 13:46:33 +03:00
qazal 6acda43a2c
test a rewrite of permuted reduce [pr] (#7093)
* test a rewrite of permuted reduce [pr]

* addd rewrite tracker

* expected

* passes
2024-10-16 12:49:54 +03:00
qazal 56fbd408a1
viz print the sink tree as it's rewritten [pr] (#7094) 2024-10-16 11:38:27 +03:00
qazal aadc95c2b0
don't recreate st uop [pr] (#7092) 2024-10-16 08:43:19 +03:00
qazal 9c9c241e58
track et in viz [pr] (#7088) 2024-10-16 07:53:08 +03:00
qazal 40f33c110b
big graph var_vals as rewrite context (#7007)
* var_vals as rewrite context

* no default arg

* add st var_vals

* delete some stuff

* add the rewrite rule again

* extra

* this whole part is preschedule

* test with a second context

* redo

* i always forget tensor variable
2024-10-16 07:31:44 +03:00
qazal 390171d686
delete SAVE_SCHEDULE=1 [pr] (#7087) 2024-10-16 07:13:20 +03:00
George Hotz 3169cb386d
remove graph [pr] (#7085) 2024-10-16 11:40:07 +08:00
qazal 53586eac56
late assert post permuted assign [pr] (#7084)
* late assert post permuted assign [pr]

* a lil earlier
2024-10-16 06:26:04 +03:00
George Hotz 023b77cc6e
move MultiGraphRunner logic to GraphRunner [pr] (#7083)
* move MultiGraphRunner logic to GraphRunner [pr]

* _access_resources
2024-10-16 11:04:30 +08:00
qazal 207fbc4bc7
cleanup view on reduce [pr] (#7081) 2024-10-16 05:22:52 +03:00
qazal 067b35e915
add UOp.r [pr] (#7080) 2024-10-16 05:06:02 +03:00
George Hotz 26df50cf43
move memory_planner to memory.py [pr] (#7079) 2024-10-16 10:04:35 +08:00
qazal bddba5897a
generic elementwise view rewrite rule + merge_views (#7078)
* generic elementwise view rewrite rule + merge_views [pr]

* no pr, views merge
2024-10-16 04:36:21 +03:00
qazal fb29de6cc3
split schedule to view_left and view_right [pr] (#7077)
* split schedule to view_left and view_right [pr]

* move valid
2024-10-16 03:39:38 +03:00
chenyu 8601115976
_get_chain -> split_uop [pr] (#7075) 2024-10-15 17:31:25 -04:00
chenyu e136cea027
cleanups around idx_given_valid [pr] (#7074) 2024-10-15 16:59:01 -04:00
qazal 545e79969f
always record matches in viz (#7073)
* always record matches in viz

* simpler
2024-10-15 23:03:12 +03:00
nimlgen b025495e5c
fuzz nv vs cuda (#7066)
* fuzz nv vs cuda

* fixes

* smth

* um

* cmp the same

* dnrt

* correct gpfifo scan

* fix
2024-10-15 22:22:40 +03:00
qazal 8ff6514ba3
delete extra/ops.py [pr] (#7072) 2024-10-15 22:14:21 +03:00
qazal 09de958855
move print_diff to test/helpers (#7071) 2024-10-15 22:00:39 +03:00
qazal 1a45e94f5d
viz late to_json [pr] (#7070) 2024-10-15 21:36:45 +03:00
qazal 52d8afde2b
new viz unittests, isolate the ctx bug (#7069)
* start new test_viz

* test_rewrite_twice

* test_rewrite_with_ctx

* add back some of the old tests

* lints
2024-10-15 18:53:56 +03:00
nimlgen 9f00eacde5
nv tagged memory + resnet failed kernel (#7061)
* nv tagged memory

* linter

* metal fix?
2024-10-15 18:19:58 +03:00