George Hotz
5d66fa479b
linearize API
2025-10-10 16:52:46 +08:00
George Hotz
3d760bba51
store ranges aren't needed anymore
2025-10-10 15:21:37 +08:00
George Hotz
9ec77ff587
ttomsa linearizer
2025-10-10 15:08:28 +08:00
George Hotz
b38f3595bf
Merge branch 'master' into new_linearizer
2025-10-10 14:40:38 +08:00
Sieds Lykles
cbdc13279d
fix openpilot gated reads ( #12570 )
...
* fix gated image counts
* slice correctly
2025-10-10 04:52:57 +02:00
chenyu
c8dfd10257
ShapeTracker.real_strides -> is_expanded [pr] ( #12579 )
...
only keep the used part
2025-10-09 22:52:45 -04:00
qazal
88ce63a49a
remove outdated comment in multi [pr] ( #12580 )
2025-10-10 05:50:49 +03:00
George Hotz
5977df267f
outerworld uses expand ( #12578 )
2025-10-10 10:25:25 +08:00
chenyu
f2c3a72b0c
remove RANGEIFY flag [pr] ( #12577 )
2025-10-09 21:52:54 -04:00
George Hotz
9b66c2b0b7
fix weekly commits table (i didn't know we linted extra)
2025-10-10 09:23:33 +08:00
George Hotz
658b96cbfb
weekly commits table
2025-10-10 09:15:41 +08:00
ttomsa
9e4cd9b27c
bye
2025-10-09 23:00:50 +01:00
ttomsa
7ea44c60f6
fix
2025-10-09 22:08:32 +01:00
ttomsa
8b63413e7b
yes mr linter
2025-10-09 22:03:16 +01:00
ttomsa
2f6c5c1256
don't run the failed test
2025-10-09 21:55:25 +01:00
ttomsa
d4eb1fb45a
linter
2025-10-09 21:53:44 +01:00
ttomsa
91916dfee7
fix these
2025-10-09 21:19:48 +01:00
ttomsa
621a3e1598
fix test
2025-10-09 20:54:41 +01:00
ttomsa
26b89acc0f
ok, now!
2025-10-09 20:34:02 +01:00
ttomsa
6ea7fbe257
more correct
2025-10-09 20:00:11 +01:00
ttomsa
86782182a3
gpu work now?
2025-10-09 19:22:50 +01:00
ttomsa
2aeca4fae1
Merge remote-tracking branch 'upstream/master' into new_linearizer
2025-10-09 18:46:21 +01:00
qazal
b86ad6053a
test_schedule independent of RANGEIFY flag ( #12568 )
...
* test_schedule independent of RANGEIFY flag
* comment for expectedFailure + test_cast_padded_view
* test_cast_padded_const works
* don't use full_shape it's fine
* add todos for the rest
2025-10-09 20:00:50 +03:00
nimlgen
502e613c9c
amd: clean up uppercased vars ( #12571 )
2025-10-09 19:39:27 +08:00
George Hotz
840d2bf1ea
fix div rules ( #12567 )
...
* group div rules
* merge those pattern matchers
* revert
2025-10-09 19:28:21 +08:00
nimlgen
8a1c3dc1bf
amd: use soc headers from rocm ( #12566 )
2025-10-09 19:10:46 +08:00
qazal
e0694fdb8e
remove UPat.__repr__ [pr] ( #12565 )
2025-10-09 12:35:34 +03:00
chenyu
678f83e41b
delete ShapeTracker to_valid_uop and substitute [pr] ( #12563 )
2025-10-09 05:06:10 -04:00
nimlgen
a11b686c71
amd: sqtt for all gfx11 ( #12546 )
...
* amd: general sqtt for gfx11
* target
* ops
* no gfx12 here
2025-10-09 17:04:06 +08:00
chenyu
a0cbbc35ad
remove LLAMA_LAYERS in ci ( #12562 )
2025-10-09 04:46:41 -04:00
chenyu
fe94453d52
delete CONTIGUOUS with RANGE in st [pr] ( #12561 )
2025-10-09 04:32:31 -04:00
chenyu
f793cdeb87
clean up shape changing logic to not use st [pr] ( #12560 )
2025-10-09 04:13:02 -04:00
chenyu
1bcea19846
remove ShapeTracker.reduce [pr] ( #12559 )
2025-10-09 03:54:11 -04:00
chenyu
c1cc277fc3
don't call src[0].shape multiple times in MULTI st [pr] ( #12558 )
2025-10-09 03:40:17 -04:00
qazal
2551a60d97
viz: split out shape links ( #12557 )
2025-10-09 10:34:55 +03:00
George Hotz
e7aa26ed29
make remove bufferize fast ( #12555 )
...
* add more uop gc test
* make remove bufferize fast
* substitute is fast too
* fix tests
2025-10-09 15:20:02 +08:00
chenyu
cf8232ec6a
clean up more RANGEIFY flag ( #12556 )
2025-10-09 03:06:48 -04:00
nimlgen
658c566e22
vars in gated_read_image_count ( #12486 )
...
* vars in gated_read_image_count
* nc
2025-10-09 14:54:15 +08:00
George Hotz
a8a9ac0e95
add more uop gc test ( #12553 )
2025-10-09 14:49:32 +08:00
chenyu
250f05a776
run some hashing test only on METAL ( #12554 )
...
quite slow on CPU
2025-10-09 02:39:49 -04:00
qazal
da9425c1a7
viz: sum all buffers in zoomed out memory graph ( #11898 )
...
* viz: switch to transformation matrix
* simpler axes domains
* less domain
* split loops
* flatten
* tiny rects
* solid proxy but still too big
* cache FileNotFound
* gridlines instead of padding
* not this
* like METAL -> METAL memory -> graph
* less colors
* better
* more grid work
* glitch
* clamp
* add range index
* pixel grids
* set min width
* y cords
* pruning
* test: clip in world units
* keep linear scan
* switch to interval tree
* fps counter
* work
* visible is the easiest
* shapes api
* math
* test bitgrid
* checkout
* work
* simpler
* work
* draw
* it's just a polygon
* merge polygons
* cleanup old stuff
* switch to hashmap there too
* add tooltips
* fix that
* better color
* better
2025-10-09 09:30:37 +03:00
chenyu
ae51bdd06a
remove trivial use of RANGEIFY flag ( #12550 )
...
some tests need update still
2025-10-09 02:29:38 -04:00
George Hotz
80d99d52a5
reduce_unparented only checks ranges ( #12548 )
2025-10-09 14:14:03 +08:00
nimlgen
375ee2c576
faster backward_slice ( #12515 )
...
* not cached backward_slice
* mypy
* just speed
* faster
2025-10-09 14:12:20 +08:00
George Hotz
1dc500426e
remove restrictions on range ending in indexing ( #12543 )
...
* remove restrictions on range ending in indexing
* early simplify
* Revert "early simplify"
This reverts commit 657d9972c2
2025-10-09 13:53:08 +08:00
chenyu
585bd95b50
fix ruff 0.14.0 [pr] ( #12547 )
2025-10-09 01:52:30 -04:00
qazal
6af29b913b
viz: format rewrite time as a comment ( #12545 )
...
* viz: format rewrite time as a comment
* put above
2025-10-09 07:14:27 +03:00
qazal
baab7e334d
put match times in viz ( #12544 )
...
* put match times in viz
* float
2025-10-09 06:56:10 +03:00
George Hotz
51420d1f99
rangeify profiling ( #12540 )
...
* clean up stable diffusion weight loading
* add profiling to run_rangeify
* fix tests
2025-10-09 11:32:34 +08:00
chenyu
43bce1f39f
delete View minify [pr] ( #12538 )
2025-10-08 23:25:53 -04:00
