George Hotz
b384c27314
ln
2026-03-12 08:23:06 +08:00
George Hotz
5ed68aeed5
cleanups
2026-03-12 08:18:31 +08:00
George Hotz
f682af2a31
better uop matmul
2026-03-12 08:13:09 +08:00
George Hotz
62dbf12655
call matmul
2026-03-11 21:50:31 +08:00
George Hotz
b17e15d1aa
support ranges on call
2026-03-11 15:22:46 +08:00
Christopher Milan
2fb8a7f60f
fix test_invalid_tensor when before values are nan ( #15215 )
2026-03-10 23:51:19 -04:00
chenyu
fce87f19a8
better fold_add_divmod_recombine ( #15214 )
2026-03-10 23:24:22 -04:00
chenyu
df8deec949
test for nest_by_factor selection ( #15213 )
2026-03-10 22:41:31 -04:00
chenyu
be6b0bce1f
variations of (x%c)+(x//c)*c ( #15212 )
...
put those into one function
2026-03-10 22:41:14 -04:00
qazal
a408d90f4f
viz: always detect sqtt packet overlaps, add timeline tests ( #15211 )
...
* test
* work
* it's called CALL, better assert
* qol
* row_ends
2026-03-11 05:32:38 +09:00
nimlgen
d9c7290eb0
nv: nvdec as NVDEC:0 device ( #15209 )
2026-03-10 14:44:50 +03:00
Christopher Milan
25d86ec9e1
start using Invalid in image_conv2d ( #15208 )
2026-03-10 07:11:06 -04:00
chenyu
ecbddfcffe
clean up gcd_with_remainder [pr] ( #15207 )
...
this can operate with int gcd directly and not through UOp
2026-03-10 06:13:20 -04:00
chenyu
bb7888b281
cleanup (x%(k*c))//c and (x%(k*c))%c ( #15206 )
...
these two are in the same family
2026-03-10 05:21:32 -04:00
chenyu
8389a8d7c5
remove_nested_mod can work with negative ( #15205 )
2026-03-10 03:10:08 -04:00
Christopher Milan
ffaafd391a
Invalid in Tensor ( #15154 )
2026-03-10 02:49:54 -04:00
chenyu
68c7c3ca84
divmod test_gcd_with_remainder ( #15204 )
...
test cases for gcd_with_remainder
2026-03-09 23:51:47 -04:00
chenyu
a53187eef7
fix TestPartialAssignToSharedBuffer ( #15202 )
...
bufferize_to_store issue with assign
2026-03-09 23:14:23 -04:00
wozeparrot
525a178966
llama: jit more ( #15199 )
2026-03-10 11:04:59 +08:00
George Hotz
315ad50a1a
make late allreduce the default ( #15125 )
...
Co-authored-by: wozeparrot <wozeparrot@gmail.com >
2026-03-09 17:42:57 -07:00
chenyu
6b354b906d
fold_divmod_general cleanups [pr] ( #15196 )
2026-03-09 19:43:16 -04:00
qazal
02ceeab3a7
viz: ui cleanups from the sqtt real time branch ( #15195 )
...
* label location for packets
* work
* OTHER_* packets always get filtered out
* less
2026-03-10 05:33:53 +09:00
qazal
a615ed8ebe
sqtt: update RDNA timestamp marker fields ( #15194 )
...
* rt:realtime field name, correct RDNA4
* share rdna4 and rdna3
2026-03-10 05:18:47 +09:00
nimlgen
8bd6d270c5
rm ops.encdec ( #15193 )
...
* rm ops.encdec
* x
2026-03-09 18:52:48 +03:00
qazal
81ab499b4b
viz: small ui code cleanups ( #15192 )
...
* less
* more work
* tabulate returns node like colored
2026-03-09 21:17:33 +09:00
chenyu
60215deb60
tiebreak in fold_divmod_congruence ( #15190 )
...
need to try both direction
2026-03-09 03:40:39 -04:00
chenyu
a8d8351e5a
match IDIV and MOD in nest_by_factor ( #15188 )
2026-03-09 00:50:38 -04:00
Christopher Milan
7592622562
fix QCOMCLRenderer pickle ( #15189 )
2026-03-09 00:36:16 -04:00
Christopher Milan
2bb0970512
QCOM CL compiler prints LLVMIR when DEBUG>=8 ( #15187 )
2026-03-09 00:15:20 -04:00
chenyu
83b80da8f3
even more divmod recombine ( #15163 )
2026-03-08 23:52:26 -04:00
chenyu
82f7734501
use backward_slice in reduce_mul_chain [pr] ( #15186 )
2026-03-08 21:44:53 -04:00
qazal
25e82a9aca
viz: exclude redundant traceback from SDMA ( #15185 )
...
* viz: exclude redundant traceback from SDMA
* ctx
* cpu_profile
2026-03-09 05:12:14 +09:00
nimlgen
6ac99fd4c9
memplanner opt copy bufs ( #15110 )
...
* mtp
* x
* tests
* ss
* simp
* less slop
* x
* cleaner
* rm
* m
* c
* x
* f
2026-03-08 22:28:01 +03:00
nimlgen
633264feae
am: flush sdma pipeline ( #15184 )
...
* am: flush sdma pipeline
* f
* f
* fix
2026-03-08 20:27:56 +03:00
b1tg
891a73befc
llm: fix chunked prefill ( #15182 )
...
* llm: fix chunked prefill
* less lines
---------
Co-authored-by: b1tg <b1tg@users.noreply.github.com >
2026-03-07 22:08:31 +08:00
chenyu
5d58b1c396
don't use intermediate dict in onnx parse ( #15181 )
...
also don't parse fields that are never used
2026-03-07 00:08:03 -05:00
nimlgen
086081e35b
tbgpu: add stapler to the script ( #15180 )
2026-03-07 00:07:27 +03:00
qazal
a03f512147
viz: clean up old / unused paths in sidebar rendering ( #15179 )
...
* src is unused
* less
2026-03-07 05:36:10 +09:00
chenyu
605b37c03f
use backward_slice in count_divmod [pr] ( #15178 )
2026-03-06 14:03:53 -05:00
Ananta Ranganathan
5bdad8ee41
update mxfp4 tests to use the same patterns as the others ( #15177 )
...
* update mxfp4 tests to use the same patterns as the others
* fix typo in test call not sure how it committed
2026-03-06 13:21:40 -05:00
qazal
d85109f9f7
viz: walk PROGRAM UOp back to source and binary only ( #15174 )
...
* work
* simpler
2026-03-07 01:39:07 +09:00
Ananta Ranganathan
5c50035e0d
avoid using arithmetic for mxfp4 ( #15172 )
...
* avoid using arithmetic for mxfp4
* update tests to use assert equal
* no longer todo
2026-03-06 11:17:56 -05:00
qazal
f064db0ac6
viz: later tooltip rendering ( #15170 )
2026-03-06 23:00:15 +09:00
Roelof van Dijk
4ed8bb7445
tie break for divmod ( #15169 )
2026-03-06 08:05:38 -05:00
qazal
83f1faa142
sqtt: update CDNA wave packet field, start unskipping tests ( #15168 )
...
* correct field names
* packet types
* packet 5 is regc
* test skips
2026-03-06 21:37:44 +09:00
Christopher Milan
7810be8d3c
compile QCOM without opening device ( #15165 )
...
Co-authored-by: Comma Device <device@comma.ai >
2026-03-06 06:24:27 -05:00
George Hotz
6fd18ef875
rename CAT to VCAT ( #15167 )
2026-03-06 18:46:28 +08:00
Roelof van Dijk
059c6326c0
metal uint32 icb offset overflow ( #15156 )
...
* metal uint32 icb offset overflow
fix: diff
supports_exec_item
GraphRunner.supports_exec_item
tests
fix: can't import on non-metal
stricter
* also test the non-metal buffer case
* imports on non-mac
2026-03-06 00:54:39 +03:00
chenyu
da61088ca4
more divmod recombine ( #15162 )
2026-03-05 12:53:22 -05:00
chenyu
167a1d56a6
improve divmod folding ( #15148 )
...
canonicalize to div than mod which enables more simplifcation
2026-03-05 10:07:36 -05:00
