Skip to content

Develop upstream sync 250427 #2939

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1,344 commits into
base: develop-upstream
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1344 commits
Select commit Hold shift + click to select a range
86d0d43
Update GraphDef version to 2205.
tensorflower-gardener Apr 22, 2025
fe66657
compat: Update forward compatibility horizon to 2025-04-22
tensorflower-gardener Apr 22, 2025
b0f1414
[XLA:CPU] Port scatter to kernel API
WillFroom Apr 22, 2025
1df34ae
Mark Tensorflow Python compatible with Protobuf v30+.
tensorflower-gardener Apr 22, 2025
01024eb
Automated Code Change
tensorflower-gardener Apr 22, 2025
8467d15
[XLA:GPU] Use correct type for intermediate tensor in the (new) Split…
mooskagh Apr 22, 2025
491f212
PR #25166: Add slow operation alarm for argument initialization in th…
sergachev Apr 22, 2025
a49d238
Add more device attributes.
changhuilin Apr 22, 2025
9ff5249
Automated Code Change
tensorflower-gardener Apr 22, 2025
5a092a5
PR #25275: Propagate loop metadata in double_buffer_loop_unrolling.
jreiffers Apr 22, 2025
c26c5ef
PR #25388: [ROCM] Fix gpu_kernel_test
alekstheod Apr 22, 2025
1a46717
Avoid using corrupt memory on early termination.
Apr 22, 2025
bfa66b1
Rewrite GpuSolverRewriterTest to not require a GPU.
akuegel Apr 22, 2025
892d0b3
Reland NVSHMEM allocator usage PR.
tensorflower-gardener Apr 22, 2025
aff55b4
Move CUB sort kernels behind FFI
beckerhe Apr 22, 2025
aa39bea
#sdy Optionally try and get the tensor sharding off the frontend attr…
bartchr808 Apr 22, 2025
6e75e93
[xla:gpu] Allow S4 in generic triton emitter.
chsigg Apr 22, 2025
3cb06ab
Reverts 9d4f5c15ff88144612d9cd328aa8e5bb66dc2c01
tensorflower-gardener Apr 22, 2025
2d24785
PR #25489: Fix typos in gpu_dot_fusion_cost_model.h
dimvar Apr 22, 2025
6edb207
Internal visibility cleanup
junwhanahn Apr 22, 2025
158450d
[xla:cpu] Implement more pthreadpool APIs in XLA adaptor
ezhulenev Apr 22, 2025
cd38ebd
[XLA:GPU] update ncu-rep parser to work with multiple kernels
metaflow Apr 22, 2025
cda6306
Run `nvidia-smi` on TF/JAX GPU builds to make it easier to get driver…
ddunl Apr 22, 2025
264c48f
[XLA:GPU] Enable normalization and reduction tests in `fusion_emitter…
bchetioui Apr 22, 2025
f31b47a
[XLA] Add BenchmarkConfig proto and configs to set up microbenchmarki…
juliagmt-google Apr 22, 2025
b4727b1
Implementation of a new heuristic, called RandomPathGreedy, which sam…
tensorflower-gardener Apr 22, 2025
5152e64
[xla:cpu] ParallelLoopRunner: adjust tile dimensions for dynamic para…
ezhulenev Apr 22, 2025
45bdc5b
Move hlo_op_code_test near its base files.
toli-y Apr 22, 2025
7fdaa84
Plumbing for has-a between LoadedExec and Exec
jparkerh Apr 22, 2025
e654528
Simplify the `Layout` API: remove `SetProto()`.
tensorflower-gardener Apr 22, 2025
399932e
quant_add_quantization_unit_loc pass migrates to TF quant dialect fro…
ecalubaquib Apr 22, 2025
b9dacf6
[xla:cpu] Fix work_queue clang warnings
ezhulenev Apr 22, 2025
1a4529a
Use the utility function `sdy::getFirstFreeAxisIter` in `shard_map_ex…
ZixuanJiang Apr 22, 2025
17eaaf2
Move analysis of dynamic-update-slice to utility module.
ZixuanJiang Apr 22, 2025
4acddbb
More StableHLO Ops : Direct StableHLO -> HLO Translation
abhigunj Apr 22, 2025
0d4f494
Remove unsafe API `Shape::mutable_dimensions()`.
tensorflower-gardener Apr 22, 2025
2d6a14a
Reverts 6edb20704f31377487224365fe13eea791e955a5
tensorflower-gardener Apr 22, 2025
2241334
[XLA] Fix deprecated headers in generate_math_impl.py.
GleasonK Apr 22, 2025
3a90b9c
Migrate OpenXLA to `absl::string_view` per the Google C++ style guide.
tensorflower-gardener Apr 22, 2025
0784c8b
[XLA:MSA] Move coloring requirement properties to the `AllocationRequ…
subhankarshah Apr 22, 2025
13ce747
Apply split shapes to defining position as well as derivative instruc…
tensorflower-gardener Apr 22, 2025
9dc0ebf
Reverts 2d6a14ac610e06e9798ffd5a30064cd78c302305
tensorflower-gardener Apr 22, 2025
6d325b8
Add `darwin_arm64` config setting and use it to depend on `:hh_neon` …
ddunl Apr 22, 2025
723e472
Add `GetCompiledMemoryStats()` method for TfrtGpuExecutable.
changhuilin Apr 22, 2025
0c66814
Add no-op scheduling annotation
tensorflower-gardener Apr 22, 2025
87cabb3
Remove profiler client from the public package namespace
tensorflower-gardener Apr 22, 2025
c4ccbef
Roll forward of non-blocking NCCL communicators change.
mwhittaker Apr 22, 2025
632eb08
Convert HloTestBase tests to HloHardwareIndependentTestBase tests whe…
nvgrw Apr 22, 2025
8eb1e90
Automated Code Change
tensorflower-gardener Apr 22, 2025
c1ab077
Make the `xla::Layout` API less error-prone: make dimensions in a lay…
tensorflower-gardener Apr 22, 2025
d9062cf
Set TFCI_MACOS_UPGRADE_PYENV_ENABLE=1 for python 3.13
ecalubaquib Apr 22, 2025
964250c
IFRT Proxy: Disable executable_impl_test_tfrt_cpu since it is flaky i…
tensorflower-gardener Apr 23, 2025
a9b3625
Manage host_dma_ptr with unique_ptr in tests
apivovarov Apr 23, 2025
e5e7f51
Change unique_ptr to shared_ptr for latency_estimator and async_track…
tensorflower-gardener Apr 23, 2025
c0d4187
Upgrade to XCode 16
majiddadashi Apr 23, 2025
2b8fd7d
Reverts aff55b46f83131f4873d542a26e9f1bab7e78631
hawkinsp Apr 23, 2025
5c7cbdd
Support movable callback types inside AsyncValue.AndThen(...);
pschuh Apr 23, 2025
8e0b80a
[tf2xla] Migrate TF2XLA patterns to use StableHLO.
GleasonK Apr 23, 2025
b6c895c
Automated Code Change
tensorflower-gardener Apr 23, 2025
49d7f2c
Remove the unused op name
tensorflower-gardener Apr 23, 2025
a39720e
Automated Code Change
tensorflower-gardener Apr 23, 2025
49f2c56
Merge pull request #91854 from NicholasWilsonDEV:patch-1
tensorflower-gardener Apr 23, 2025
0a1c43d
Automated Code Change
tensorflower-gardener Apr 23, 2025
0108de3
Automated Code Change
tensorflower-gardener Apr 23, 2025
9600cb1
Automated Code Change
tensorflower-gardener Apr 23, 2025
2f35f24
Automated Code Change
tensorflower-gardener Apr 23, 2025
17946cc
Automated Code Change
tensorflower-gardener Apr 23, 2025
038b891
Automated Code Change
tensorflower-gardener Apr 23, 2025
71f296d
Automated Code Change
tensorflower-gardener Apr 23, 2025
33b30ad
Automated Code Change
tensorflower-gardener Apr 23, 2025
51e8e19
Automated Code Change
tensorflower-gardener Apr 23, 2025
25f5028
Automated Code Change
tensorflower-gardener Apr 23, 2025
7b93b75
Automated Code Change
tensorflower-gardener Apr 23, 2025
fd46e70
Automated Code Change
tensorflower-gardener Apr 23, 2025
5f10c1e
Automated Code Change
tensorflower-gardener Apr 23, 2025
7d54743
Merge pull request #91897 from wangw-1991:master
tensorflower-gardener Apr 23, 2025
6dc6ba9
Automated Code Change
tensorflower-gardener Apr 23, 2025
b15acf9
Automated Code Change
tensorflower-gardener Apr 23, 2025
2a6bbde
Automated Code Change
tensorflower-gardener Apr 23, 2025
7f187bc
Automated Code Change
tensorflower-gardener Apr 23, 2025
67629d4
Automated Code Change
tensorflower-gardener Apr 23, 2025
18e991b
Automated Code Change
tensorflower-gardener Apr 23, 2025
9ad15a8
Automated Code Change
tensorflower-gardener Apr 23, 2025
d95f9e6
Automated Code Change
tensorflower-gardener Apr 23, 2025
f3f7363
[XLA:GPU] Clean up and enable parametrized comparison tests for the g…
bchetioui Apr 23, 2025
4c8262e
Automated Code Change
tensorflower-gardener Apr 23, 2025
c11f551
Automated Code Change
tensorflower-gardener Apr 23, 2025
85adc0e
Automated Code Change
tensorflower-gardener Apr 23, 2025
30bf49b
Automated Code Change
tensorflower-gardener Apr 23, 2025
da23590
Automated Code Change
tensorflower-gardener Apr 23, 2025
fd06b8b
Automated Code Change
tensorflower-gardener Apr 23, 2025
08a8e8e
[XLA:GPU] Enable the parametrized `MixedTypeTest`s with the new emitter.
bchetioui Apr 23, 2025
4e270d6
Automated Code Change
tensorflower-gardener Apr 23, 2025
9074028
Automated Code Change
tensorflower-gardener Apr 23, 2025
a3fd616
Update GraphDef version to 2206.
tensorflower-gardener Apr 23, 2025
7e014ea
compat: Update forward compatibility horizon to 2025-04-23
tensorflower-gardener Apr 23, 2025
2a241b5
PR #25492: Fix build issue
shraiysh Apr 23, 2025
96a4586
use move ctor for mutable_entry_computation_layout
tomnatan30 Apr 23, 2025
e3b3dec
Clean up autotuner tests to prepare enabling dynamic search space
gflegar Apr 23, 2025
fe5115b
Automated Code Change
tensorflower-gardener Apr 23, 2025
2c52c02
Automated Code Change
tensorflower-gardener Apr 23, 2025
915bb9a
Automated Code Change
tensorflower-gardener Apr 23, 2025
3ef99b4
Automated Code Change
tensorflower-gardener Apr 23, 2025
2470dd4
Automated Code Change
tensorflower-gardener Apr 23, 2025
7250609
Automated Code Change
tensorflower-gardener Apr 23, 2025
a07ccbe
Automated Code Change
tensorflower-gardener Apr 23, 2025
2c2f141
Automated Code Change
tensorflower-gardener Apr 23, 2025
6ba7387
Automated Code Change
tensorflower-gardener Apr 23, 2025
1c43bc8
Automated Code Change
tensorflower-gardener Apr 23, 2025
2c92857
Automated Code Change
tensorflower-gardener Apr 23, 2025
91aacfa
[XLA:GPU] Update emitters documentation with the new flag command
tensorflower-gardener Apr 23, 2025
ae2837a
Enforce use-tuple-args when Shardy is enabled but sharding propagatio…
tomnatan30 Apr 23, 2025
016aaa5
[XLA:GPU] Return std::optional when detecting if a topology is multi …
golechwierowicz Apr 23, 2025
e02f83e
Log "unknown" if memory overflows an int
tf-marissaw Apr 23, 2025
db97430
[xla:cpu:benchmarks] Add `RunHloBenchmark` variants that take `HloMod…
penpornk Apr 23, 2025
8728276
[XLA:GPU] When determining whether a dot has a packed operand, also g…
mooskagh Apr 23, 2025
f6918ea
Fix incorrect migration tag + base class combination in CPU test.
nvgrw Apr 23, 2025
6ce1d9a
Use patched definitions of absl nullability macros to allow their use.
tensorflower-gardener Apr 23, 2025
835bf3a
[xla:cpu:benchmarks] Add Gemma3 1B dot sizes to dot_benchmark_test.
penpornk Apr 23, 2025
2133648
[xla:cpu] Enable pthreadpool adaptor for XLA on Linux platform
ezhulenev Apr 23, 2025
ef3a0ae
Allow the number of quantization parameters for `BATCH_MATMUL` to be …
gonnet Apr 23, 2025
bdf136b
Explicitly set a timeout to moderate for swift ios_unit_test
majiddadashi Apr 23, 2025
7b91561
Fix bug where `dimensions()` is called on a non-array `Shape`.
tensorflower-gardener Apr 23, 2025
3e71611
Generalize the "permutation not needed" check for all dimensions.
nafi3000 Apr 23, 2025
5062ec7
[xla:cpu] Return communicator event as collective thunk execution event
ezhulenev Apr 23, 2025
83d3f06
Integrate StableHLO at openxla/stablehlo@a54938f0
sdasgup3 Apr 23, 2025
a6cc083
Introduce PjRtDeviceEvent which can be returned from (for example)
pschuh Apr 23, 2025
4d0b9e9
Migrate AddLegalizeTFToStablehloPasses to tensorflow/compiler/mlir/st…
ecalubaquib Apr 23, 2025
763ff8a
Reverts 18482601f1ab17f4b0bfd2e5588f620c4f6b7a54
reedwm Apr 23, 2025
f0f71d9
[XLA:GPU][Emitters] Fix the case when there are N transposes with dif…
pifon2a Apr 23, 2025
8a66c6f
Fix xla util.cc for android 27 or older platforms
apivovarov Apr 23, 2025
31331a9
No need to really create device buffers for an error PjRt Buffer.
tensorflower-gardener Apr 23, 2025
793b45e
#sdy Support multiple src/tgt pairs in sdy_all_to_all
tensorflower-gardener Apr 24, 2025
6f696f0
Add debugging logs in tfrt_gpu_client.cc
tensorflower-gardener Apr 24, 2025
b5cd557
Fix bug where `DeleteDimension()` doesn't adjust dimension attributes…
tensorflower-gardener Apr 24, 2025
42d6877
create a default GpuExecutableRunOptions for TfrtGpuClient
tensorflower-gardener Apr 24, 2025
0985782
Implement SubByteTransfer in TfrtGpuClient
hhb Apr 24, 2025
20ba38f
Migrate :legalize_tf used by AddLegalizeTFToStablehloPasses to tensor…
ecalubaquib Apr 24, 2025
4348435
Automated Code Change
tensorflower-gardener Apr 24, 2025
eaa9f80
[XLA:GPU] Move collectives registration from collective_thunk to a se…
tensorflower-gardener Apr 24, 2025
41f820d
Automated Code Change
tensorflower-gardener Apr 24, 2025
8e83acf
Integrate LLVM at llvm/llvm-project@f9c01b59e3d5
tensorflower-gardener Apr 24, 2025
14fec2f
Automated Code Change
tensorflower-gardener Apr 24, 2025
ceff49f
compat: Update forward compatibility horizon to 2025-04-24
tensorflower-gardener Apr 24, 2025
d003359
Update GraphDef version to 2207.
tensorflower-gardener Apr 24, 2025
b2347b8
Update platform_version() method to be consistent with the current PJ…
changhuilin Apr 24, 2025
2dd9dab
temporarily remove references to `sdy.all_to_all` in lit tests becaus…
tomnatan30 Apr 24, 2025
287055c
[XLA:GPU] Remove the `FP8ToFP8EndToEnd` from fusion_emitter_device_test.
Moerafaat Apr 24, 2025
935449b
[XLA:GPU] Remove TODOs related to "support more types" through Triton…
Moerafaat Apr 24, 2025
75c30e4
Allow disabling nvshmem for build targets.
tensorflower-gardener Apr 24, 2025
34c203b
Make PrecisionTestsForTriton less implementation-dependent and more r…
gflegar Apr 24, 2025
6bf4ec6
[XLA:GPU] Expose `FindOutputTileSizesForEpilogue` via nest_gemm_fusio…
golechwierowicz Apr 24, 2025
0de76d6
Reverts 763ff8abced374c8effd14c156ad2a1086e50093
beckerhe Apr 24, 2025
d07d588
[XLA:GPU][Emitters] Enable s4 reductions.
pifon2a Apr 24, 2025
5e7bf8c
Clean up test HLOs.
chsigg Apr 24, 2025
db98ef3
Remove the special handling for Quant OPs in TAC
tensorflower-gardener Apr 24, 2025
53f36d4
Adds and populates additional fields in the SolverParams structure.
tensorflower-gardener Apr 24, 2025
39e595a
Added `WorkerThread` abstraction.
mwhittaker Apr 24, 2025
e015e8c
Allow `ResourceVariable::AssignFrom` to not initialize the newly allo…
tensorflower-gardener Apr 24, 2025
d8ae94c
[xla:ffi] Use std::string_view in FFI API headers
ezhulenev Apr 24, 2025
ab4b941
[xla:cpu] Resolve custom call target at construction time
ezhulenev Apr 24, 2025
f1d17d4
Traverse computations DFS (using post-ordered/scheduled instructions)…
tensorflower-gardener Apr 24, 2025
43ec7fd
Add support on checking duplicated donation and error inputs.
tensorflower-gardener Apr 24, 2025
cd6fe0c
[XLA:GPU] run transpose dimension folding before nesting gemms
metaflow Apr 24, 2025
10d3c55
Remove upper bound for Protobuf compatibility.
tensorflower-gardener Apr 24, 2025
5f77d50
Fixing in lowering to LLIR:
Moerafaat Apr 24, 2025
86a96b1
Add --platforms flags to .bazelrc for Android, iOS, Macos and elinux.
tensorflower-gardener Apr 24, 2025
36c59bd
[xla:ffi] Add benchmark for enum attribute decoding
ezhulenev Apr 24, 2025
0b769c3
[XLA:GPU] copy shape with layout when hoisting bitcasts in nested_gem…
metaflow Apr 24, 2025
3a8ff3c
Re-enable CompiledProgramsCount Test
tensorflower-gardener Apr 24, 2025
7621d1b
Add PjRtDeviceEventOrPromise which represents a promise of a future D…
pschuh Apr 24, 2025
c0993c8
[PJRT] Add link to the PJRT plugin tutorial slides and recording in t…
GleasonK Apr 24, 2025
bbd1aaf
Add more descriptive errors in hlo-translate.
GleasonK Apr 24, 2025
fa95b3a
Internal changes for cleanup
junwhanahn Apr 24, 2025
c3c511e
Internal cleanup of unused libraries.
tensorflower-gardener Apr 24, 2025
e240a6d
#mu add ArrayAttr and DictAttr
chunnienc Apr 24, 2025
3dd62ae
[XLA:Python] Add nanobind_pywrap_extension (from JAX) and add _ops an…
hawkinsp Apr 24, 2025
de0b9d0
[XLA:benchmark] Add TargetMetric to BenchmarkConfig and update regist…
juliagmt-google Apr 24, 2025
130e63c
fork post_quantize.cc to tf_post_quantize.cc without lite dependencies
ecalubaquib Apr 24, 2025
7ac25fc
Update pyenv before install
ecalubaquib Apr 24, 2025
cdd12c9
Removed unused `CoordinationServiceAgent` indirection.
mwhittaker Apr 24, 2025
cb392b4
Use `numpy 1.26.4` for Python 3.13
ecalubaquib Apr 24, 2025
dc348e7
Fix CompiledProgramsCount for real
tensorflower-gardener Apr 25, 2025
90eab35
Update XNNPACK version
tensorflower-gardener Apr 25, 2025
f22e270
[XLA:CPU][bugfix] Collectives should CHECK errors only if they are set.
basioli-k Apr 25, 2025
d5b782d
[XLA][Numerics][HLO Value Tracking] Renames a function to avoid confu…
jcai19 Apr 25, 2025
aa3f76a
Change AsyncWorkRunner from accepting absl::AnyInvocable<void()> -> a…
pschuh Apr 25, 2025
36d328c
Allow passing in preserved weights to custom combiner BWD ops.
tensorflower-gardener Apr 25, 2025
df68b32
Update ops-related pbtxt files.
tensorflower-gardener Apr 25, 2025
6ff4663
Optimize IsDmaMapped by using a ordered map.
hhb Apr 25, 2025
0e249b5
Support apply_model_transformations flag to tflite::gpu::BuildFromFla…
tensorflower-gardener Apr 25, 2025
b438e6d
PR #25628: Fix dumping non-default debug options
shraiysh Apr 25, 2025
3132183
Automated Code Change
tensorflower-gardener Apr 25, 2025
fabccbf
Automated Code Change
tensorflower-gardener Apr 25, 2025
16bc194
Automated Code Change
tensorflower-gardener Apr 25, 2025
a0ec356
Add a field to PresetAssignments that specifies the largets post-HloM…
sparc1998 Apr 25, 2025
d8cad48
[XLA:GPU] Move cuda_collectives static methods to the cuda_executor.
tensorflower-gardener Apr 25, 2025
274d072
[xla:cpu] NFC: fusion_compiler: fix typo
cota Apr 25, 2025
893c85d
Update GraphDef version to 2208.
tensorflower-gardener Apr 25, 2025
9546d4f
compat: Update forward compatibility horizon to 2025-04-25
tensorflower-gardener Apr 25, 2025
61929c6
Automated Code Change
tensorflower-gardener Apr 25, 2025
cd78399
[XLA:GPU] Also handle hoisting downwards past convert ops correctly.
akuegel Apr 25, 2025
9867f5b
[XLA:CPU] Use UWTableKind from LLVM dialect
WillFroom Apr 25, 2025
8e94390
PR #25661: Report an error message when user specify a non-existed pg…
shawnwang18 Apr 25, 2025
f7b85b2
[XLA:GPU/TMA] Add a flag to enable/disable TMA through Triton (disabl…
Moerafaat Apr 25, 2025
ae7123b
Automated Code Change
tensorflower-gardener Apr 25, 2025
dcbb682
PR #25680: [ROCm] disable fp8 case in gpu_compiler_test
ScXfjiang Apr 25, 2025
637a0a8
Integrate Triton up to [f60465e](f60465e16fa801ffc12e2c905a29f2722101…
loislo Apr 25, 2025
91995a2
[mlir][tosa] Add legalization for tfl.bitwise_xor (#92199)
IanTaylerLessa-arm Apr 25, 2025
34c9eb8
Remove duplication of broadcasts as it is no longer required.
SandSnip3r Apr 25, 2025
7495552
[xla] Make rendezvous TraceMe a little more helpful
ezhulenev Apr 25, 2025
692df3c
Document the format of the layout notation, which is highly non-obvious.
tensorflower-gardener Apr 25, 2025
3998724
[XLA:Python] [JAX] Change JAX to use the _profiler module defined in …
hawkinsp Apr 25, 2025
97e3ace
Remove implicit conversion to `absl::AlphaNum`.
allanrenucci Apr 25, 2025
9d7834b
[XLA:LatencyHidingScheduler] Record Memory Pressure in ScheduleProto
kuterd Apr 25, 2025
955b2d0
[XLA][collectives] Change Communicator::AllToAll interface to pass se…
basioli-k Apr 25, 2025
0740425
Change `SerDes::Serialize` to take a const reference
junwhanahn Apr 25, 2025
1ab9f47
[xla:ffi] Annotate decoding functions as always inline and tag unlike…
ezhulenev Apr 25, 2025
62574c7
Fix a bug where AllocationBlock::GetColocationsCount could check fail…
sparc1998 Apr 25, 2025
f03d460
Add pass timing tests for GPU backend.
tensorflower-gardener Apr 25, 2025
11f9ffd
[xla:ffi] Add an Ffi::Wrapper to convert runtime function pointers to…
ezhulenev Apr 25, 2025
e0c883d
Fixed dynamic size when creating dynamic shaped fake arguments.
tensorflower-gardener Apr 25, 2025
7cca416
Deprecate `tsl::strings::Hex` in favor of `absl::Hex`.
allanrenucci Apr 25, 2025
6864106
Remove out-dated comments.
tensorflower-gardener Apr 25, 2025
0612410
[XLA:Python] [JAX] Move XlaBuilder bindings out of JAX and into XLA:P…
hawkinsp Apr 25, 2025
2c0131a
Fix the bug where the dimension attributes aren't normalized when pri…
tensorflower-gardener Apr 25, 2025
965bb35
[xla:cpu] Fix gcc build error
ezhulenev Apr 25, 2025
4bb91ac
add device_ordinal in execute_fn xprof event
tensorflower-gardener Apr 25, 2025
15aad5e
Fix vis issue for tflite dep in litert oss
LukeBoyer Apr 25, 2025
84ee83d
[XLA:HeapSimulator] Add a method to clear the BufferIntervalTree and …
subhankarshah Apr 25, 2025
ecd71c0
[JAX] [XLA:Python] Move ShapeIndex bindings out of JAX and into XLA.
hawkinsp Apr 25, 2025
3b2b4cf
reset dynamic size to 0 when creating dynamic shaped fake arguments.
tensorflower-gardener Apr 25, 2025
68bee5b
Remove before/after LayoutAssignment distinction and configurable "ho…
SandSnip3r Apr 25, 2025
9df8267
[xla:cpu] Use object pool to keep call frames in a custom call thunk
ezhulenev Apr 26, 2025
5571a93
[xla] Move ObjectPool to top level xla/runtime
ezhulenev Apr 26, 2025
70dab8b
[xla:gpu] Use object pool of FFI CallFrames in custom call thunk and …
ezhulenev Apr 26, 2025
85959f4
Reverts 5f8f455c10d7da20944ca1dba377f751c2633790
bartchr808 Apr 26, 2025
8f69040
Change RunWhenReady to accept absl::AnyInvocable<void()&&> which is more
pschuh Apr 26, 2025
7dcf0b5
Add a CommonPjRtClient which is supposed to implement common APIs bas…
pschuh Apr 26, 2025
0f29598
Integrate LLVM at llvm/llvm-project@c60f24dca96d
slackito Apr 26, 2025
1f35e71
Automated Code Change
tensorflower-gardener Apr 26, 2025
10efb89
Update GraphDef version to 2209.
tensorflower-gardener Apr 26, 2025
a0cca90
compat: Update forward compatibility horizon to 2025-04-26
tensorflower-gardener Apr 26, 2025
f072b03
Remove argument `getStringAttr` in `convertToHloShardingAttr` since i…
ZixuanJiang Apr 26, 2025
6309d14
Remove `SetProto()` from `Layout` and `SplitConfig`.
tensorflower-gardener Apr 26, 2025
662716e
Enable tosa support under conditional compilation (#92055)
mtrofin Apr 26, 2025
16e81ea
Add `@local_config_nccl//:nccl` cuda_dep_filter for `tensorflow_frame…
tensorflower-gardener Apr 26, 2025
b0d7512
Remove EIGEN_NO_THROW decorator on leaky_relu_op. The EIGEN_NO_THROW …
tensorflower-gardener Apr 27, 2025
245d82a
stage all merge conflicts, commit all files as-is
linchen1-robot Apr 27, 2025
0bb76bc
Resolved merge conflicts
linchen1-robot Apr 27, 2025
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
The diff you're trying to view is too large. We only load the first 3000 changed files.
29 changes: 24 additions & 5 deletions .bazelrc
Original file line number Diff line number Diff line change
Expand Up @@ -164,15 +164,19 @@ build:android --host_crosstool_top=@bazel_tools//tools/cpp:toolchain
build:android_arm --config=android
build:android_arm --cpu=armeabi-v7a
build:android_arm --fat_apk_cpu=armeabi-v7a
build:android_arm --platforms=@org_tensorflow//tensorflow/tools/toolchains/android:armeabi-v7a
build:android_arm64 --config=android
build:android_arm64 --cpu=arm64-v8a
build:android_arm64 --fat_apk_cpu=arm64-v8a
build:android_arm64 --platforms=@org_tensorflow//tensorflow/tools/toolchains/android:arm64-v8a
build:android_x86 --config=android
build:android_x86 --cpu=x86
build:android_x86 --fat_apk_cpu=x86
build:android_x86 --platforms=@org_tensorflow//tensorflow/tools/toolchains/android:x86
build:android_x86_64 --config=android
build:android_x86_64 --cpu=x86_64
build:android_x86_64 --fat_apk_cpu=x86_64
build:android_x86_64 --platforms=@org_tensorflow//tensorflow/tools/toolchains/android:x86_64

# Build everything statically for Android since all static libs are later
# bundled together into a single .so for deployment.
Expand Down Expand Up @@ -205,6 +209,7 @@ build:apple-toolchain --host_crosstool_top=@local_config_apple_cc//:toolchain
# Settings for MacOS on ARM CPUs.
build:macos_arm64 --cpu=darwin_arm64
build:macos_arm64 --macos_minimum_os=11.0
build:macos_arm64 --platforms=@build_bazel_apple_support//configs/platforms:darwin_arm64

# iOS configs for each architecture and the fat binary builds.
build:ios --apple_platform_type=ios
Expand All @@ -213,14 +218,19 @@ build:ios --copt=-Wno-c++11-narrowing
build:ios --config=apple-toolchain
build:ios_armv7 --config=ios
build:ios_armv7 --cpu=ios_armv7
build:ios_armv7 --platforms=@org_tensorflow//tensorflow/tools/toolchains/ios:ios_armv7
build:ios_arm64 --config=ios
build:ios_arm64 --cpu=ios_arm64
build:ios_arm64 --platforms=@build_bazel_apple_support//configs/platforms:ios_arm64
build:ios_arm64e --config=ios
build:ios_arm64e --cpu=ios_arm64e
build:ios_arm64e --platforms=@build_bazel_apple_support//configs/platforms:ios_arm64e
build:ios_sim_arm64 --config=ios
build:ios_sim_arm64 --cpu=ios_sim_arm64
build:ios_sim_arm64 --platforms=@build_bazel_apple_support//configs/platforms:ios_sim_arm64
build:ios_x86_64 --config=ios
build:ios_x86_64 --cpu=ios_x86_64
build:ios_x86_64 --platforms=@build_bazel_apple_support//configs/platforms:ios_x86_64
build:ios_fat --config=ios
build:ios_fat --ios_multi_cpus=armv7,arm64,i386,x86_64

Expand Down Expand Up @@ -257,13 +267,15 @@ build:mkl_aarch64 -c opt
build:mkl_aarch64_threadpool --define=build_with_mkl_aarch64=true
build:mkl_aarch64_threadpool -c opt

# Default CUDA and CUDNN versions.
build:cuda_version --repo_env=HERMETIC_CUDA_VERSION="12.5.1"
build:cuda_version --repo_env=HERMETIC_CUDNN_VERSION="9.3.0"

# CUDA: This config refers to building CUDA op kernels with nvcc.
build:cuda --repo_env TF_NEED_CUDA=1
build:cuda --crosstool_top=@local_config_cuda//crosstool:toolchain
build:cuda --@local_config_cuda//:enable_cuda
# Default CUDA and CUDNN versions.
build:cuda --repo_env=HERMETIC_CUDA_VERSION="12.5.1"
build:cuda --repo_env=HERMETIC_CUDNN_VERSION="9.3.0"
build:cuda --config=cuda_version
# This flag is needed to include CUDA libraries.
build:cuda --@local_config_cuda//cuda:include_cuda_libs=true

Expand Down Expand Up @@ -293,8 +305,7 @@ build:cuda_clang --linkopt="-lm"

# Set up compilation CUDA version and paths and use the CUDA Clang toolchain.
build:cuda_clang_official --config=cuda_clang
build:cuda_clang_official --repo_env=HERMETIC_CUDA_VERSION="12.5.1"
build:cuda_clang_official --repo_env=HERMETIC_CUDNN_VERSION="9.3.0"
build:cuda_clang_official --config=cuda_version
build:cuda_clang_official --action_env=CLANG_CUDA_COMPILER_PATH="/usr/lib/llvm-18/bin/clang"
build:cuda_clang_official --crosstool_top="@local_config_cuda//crosstool:toolchain"

Expand Down Expand Up @@ -623,6 +634,12 @@ build:rbe_linux_cpu --python_path="/usr/bin/python3"
# These you may need to change for your own GCP project.
common:rbe_linux_cpu --remote_instance_name=projects/tensorflow-testing/instances/default_instance

# Download CUDA/CUDNN redistributions to preserve the repositories cache between
# CPU and GPU builds.
# TODO(ybaturina): Uncomment when RBE is ready to support this.
# build:rbe_linux_cpu --repo_env USE_CUDA_REDISTRIBUTIONS=1
# build:rbe_linux_cpu --config=cuda_version

# TODO(kanglan): Remove it after toolchain update is complete.
build:rbe_linux_cpu_old --config=rbe_linux
build:rbe_linux_cpu_old --host_crosstool_top="@ubuntu20.04-gcc9_manylinux2014-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain"
Expand Down Expand Up @@ -676,8 +693,10 @@ build:elinux --crosstool_top=@local_config_embedded_arm//:toolchain
build:elinux --host_crosstool_top=@bazel_tools//tools/cpp:toolchain
build:elinux_aarch64 --config=elinux
build:elinux_aarch64 --cpu=aarch64
build:elinux_aarch64 --platforms=@org_tensorflow//tensorflow/tools/toolchains/linux:linux_aarch64
build:elinux_armhf --config=elinux
build:elinux_armhf --cpu=armhf
build:elinux_armhf --platforms=@org_tensorflow//tensorflow/tools/toolchains/linux:linux_armhf
build:elinux_armhf --copt -mfp16-format=ieee

# Config-specific options should come above this line.
Expand Down
2 changes: 1 addition & 1 deletion .github/workflows/osv-scanner-scheduled.yml
Original file line number Diff line number Diff line change
Expand Up @@ -28,7 +28,7 @@ permissions:
jobs:
scan-scheduled:
if: github.repository == 'tensorflow/tensorflow'
uses: "google/osv-scanner-action/.github/workflows/osv-scanner-reusable.yml@v1.9.2"
uses: "google/osv-scanner-action/.github/workflows/osv-scanner-reusable.yml@v2.0.0"
with:
scan-args: |-
--lockfile=requirements.txt:./requirements_lock_3_9.txt
Expand Down
2 changes: 1 addition & 1 deletion .github/workflows/pylint-presubmit.yml
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ jobs:
run: |
echo Changed files: ${{ steps.get_file_changes.outputs.files }}
- name: Set up Python 3.9
uses: actions/setup-python@42375524e23c412d93fb67b49958b491fce71c38 # v5.4.0
uses: actions/setup-python@8d9ed9ac5c53483de85588cdf95a591a75ab9f55 # v5.5.0
with:
python-version: "3.9"
- name: Install Python dependencies
Expand Down
2 changes: 1 addition & 1 deletion .github/workflows/release-branch-cherrypick.yml
Original file line number Diff line number Diff line change
Expand Up @@ -58,7 +58,7 @@ jobs:
echo "SHORTSHA=$(git log -1 ${{ github.event.inputs.git_commit }} --format="%h")" >> "$GITHUB_OUTPUT"
echo "TITLE=$(git log -1 ${{ github.event.inputs.git_commit }} --format="%s")" >> "$GITHUB_OUTPUT"
- name: Create Pull Request with changes
uses: peter-evans/create-pull-request@dd2324fc52d5d43c699a5636bcf19fceaa70c284 # v7.0.7
uses: peter-evans/create-pull-request@271a8d0340265f705b14b6d32b9829c1cb33d45e # v7.0.8
with:
title: '${{ github.event.inputs.release_branch }} cherry-pick: ${{ steps.cherrypick.outputs.SHORTSHA }} "${{ steps.cherrypick.outputs.TITLE }}"'
committer: TensorFlow Release Automation <[email protected]>
Expand Down
4 changes: 2 additions & 2 deletions .github/workflows/scorecards-analysis.yml
Original file line number Diff line number Diff line change
Expand Up @@ -55,7 +55,7 @@ jobs:
# Upload the results as artifacts (optional). Commenting out will disable uploads of run results in SARIF
# format to the repository Actions tab.
- name: "Upload artifact"
uses: actions/upload-artifact@4cec3d8aa04e39d1a68397de0c4cd6fb9dce8ec1 # v4.6.1
uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
with:
name: SARIF file
path: results.sarif
Expand All @@ -64,6 +64,6 @@ jobs:
# Upload the results to GitHub's code scanning dashboard (optional).
# Commenting out will disable upload of results to your repo's Code Scanning dashboard
- name: "Upload to code-scanning"
uses: github/codeql-action/upload-sarif@b56ba49b26e50535fa1e7f7db0f4f7b4bf65d80d # v3.28.10
uses: github/codeql-action/upload-sarif@1b549b9259bda1cb5ddde3b41741a82a2d15a841 # v3.28.13
with:
sarif_file: results.sarif
2 changes: 1 addition & 1 deletion .github/workflows/update-rbe.yml
Original file line number Diff line number Diff line change
Expand Up @@ -130,7 +130,7 @@ jobs:
map sigbuild-r2.17-clang-python3.11 2.17-python3.11
map sigbuild-r2.17-clang-python3.12 2.17-python3.12
- name: Create Pull Request with changes
uses: peter-evans/create-pull-request@dd2324fc52d5d43c699a5636bcf19fceaa70c284 # v7.0.7
uses: peter-evans/create-pull-request@271a8d0340265f705b14b6d32b9829c1cb33d45e # v7.0.8
with:
title: Update the RBE images to the latest container versions
committer: TensorFlow Release Automation <[email protected]>
Expand Down
4 changes: 2 additions & 2 deletions RELEASE.md
Original file line number Diff line number Diff line change
Expand Up @@ -3234,7 +3234,7 @@ This release introduces several vulnerability fixes:

* Keras been split into a separate PIP package (`keras`), and its code has
been moved to the GitHub
repository[keras-team/keras](http://github.com/keras-team/keras). The
repository[keras-team/keras](https://github.com/keras-team/keras). The
API endpoints for `tf.keras` stay unchanged, but are now backed by the
`keras` PIP package. The existing code in tensorflow/python/keras is a
staled copy and will be removed in future release (2.7). Please remove
Expand Down Expand Up @@ -10309,7 +10309,7 @@ answered questions, and were part of inspiring discussions.
## Major Features And Improvements

* `tf.keras` is now part of the core TensorFlow API.
* [`tf.data`](http://tensorflow.org/guide/data) is now part of the core
* [`tf.data`](https://tensorflow.org/guide/data) is now part of the core
TensorFlow API.
* The API is now subject to backwards compatibility guarantees.
* For a guide to migrating from the `tf.contrib.data` API, see the
Expand Down
1 change: 1 addition & 0 deletions WORKSPACE
Original file line number Diff line number Diff line change
Expand Up @@ -43,6 +43,7 @@ python_init_repositories(
"3.10": "//:requirements_lock_3_10.txt",
"3.11": "//:requirements_lock_3_11.txt",
"3.12": "//:requirements_lock_3_12.txt",
"3.13": "//:requirements_lock_3_13.txt",
},
)

Expand Down
2 changes: 1 addition & 1 deletion ci/official/envs/linux_arm64
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
TFCI_BAZEL_COMMON_ARGS="--repo_env=HERMETIC_PYTHON_VERSION=$TFCI_PYTHON_VERSION --config release_arm64_linux"
TFCI_BAZEL_COMMON_ARGS="--repo_env=HERMETIC_PYTHON_VERSION=$TFCI_PYTHON_VERSION --repo_env=USE_PYWRAP_RULES=True --config release_arm64_linux"
TFCI_BAZEL_TARGET_SELECTING_CONFIG_PREFIX=linux_arm64
# Note: this is not set to "--cpu", because that changes the package name
# to tensorflow_cpu. These ARM builds are supposed to have the name "tensorflow"
Expand Down
2 changes: 1 addition & 1 deletion ci/official/envs/linux_arm64_cross_compile
Original file line number Diff line number Diff line change
Expand Up @@ -13,5 +13,5 @@
# limitations under the License.
# ==============================================================================
source ci/official/envs/linux_arm64
TFCI_BAZEL_COMMON_ARGS="--repo_env=HERMETIC_PYTHON_VERSION=$TFCI_PYTHON_VERSION --config cross_compile_linux_arm64"
TFCI_BAZEL_COMMON_ARGS="--repo_env=HERMETIC_PYTHON_VERSION=$TFCI_PYTHON_VERSION --config cross_compile_linux_arm64 --repo_env=USE_PYWRAP_RULES=True"
TFCI_BAZEL_TARGET_SELECTING_CONFIG_PREFIX=cross_compile_linux_arm64
7 changes: 6 additions & 1 deletion ci/official/envs/macos_arm64
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
TFCI_BAZEL_COMMON_ARGS="--repo_env=HERMETIC_PYTHON_VERSION=$TFCI_PYTHON_VERSION --config release_macos_arm64"
TFCI_BAZEL_COMMON_ARGS="--repo_env=HERMETIC_PYTHON_VERSION=$TFCI_PYTHON_VERSION --repo_env=USE_PYWRAP_RULES=True --config release_macos_arm64"
TFCI_BAZEL_TARGET_SELECTING_CONFIG_PREFIX=macos_arm64
TFCI_BUILD_PIP_PACKAGE_WHEEL_NAME_ARG="--repo_env=WHEEL_NAME=tensorflow"
TFCI_INDEX_HTML_ENABLE=1
Expand All @@ -29,7 +29,12 @@ case $TFCI_PYTHON_VERSION in
3.11)
TFCI_MACOS_PYENV_INSTALL_ENABLE=0
;;
3.13)
TFCI_MACOS_UPGRADE_PYENV_ENABLE=1
TFCI_MACOS_PYENV_INSTALL_ENABLE=1
;;
*)
TFCI_MACOS_PYENV_INSTALL_ENABLE=1
;;
esac

15 changes: 15 additions & 0 deletions ci/official/envs/py313
Original file line number Diff line number Diff line change
@@ -0,0 +1,15 @@
# Copyright 2023 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
TFCI_PYTHON_VERSION=3.13
4 changes: 2 additions & 2 deletions ci/official/requirements_updater/requirements.in
Original file line number Diff line number Diff line change
Expand Up @@ -28,7 +28,7 @@ requests >= 2.31.0
packaging==23.2
setuptools==70.0.0
jax==0.4.7
zstandard=0.23.0
zstandard==0.23.0
# NVIDIA CUDA dependencies
# Note that the wheels are downloaded only when the targets in bazel command
# contain dependencies on these wheels.
Expand All @@ -44,7 +44,7 @@ nvidia-cusparse-cu12 == 12.5.1.3
nvidia-nccl-cu12 == 2.25.1
nvidia-nvjitlink-cu12 == 12.5.82
# The dependencies below are needed for TF wheel testing.
tensorflow-io-gcs-filesystem==0.37.1
tensorflow-io-gcs-filesystem==0.37.1 ; python_version <= "3.12"
libclang >= 13.0.0
google_pasta ~= 0.2
flatbuffers ~= 24.3.25
6 changes: 5 additions & 1 deletion ci/official/utilities/rename_and_verify_wheels.sh
Original file line number Diff line number Diff line change
Expand Up @@ -69,7 +69,11 @@ fi
# TODO(b/366266944) Remove the check after tf docker image upgrade for NumPy 2
# and numpy 1 support is dropped b/361369076.
if [[ "$TFCI_WHL_NUMPY_VERSION" == 1 ]]; then
"$python" -m pip install numpy==1.26.0
if [[ "$TFCI_PYTHON_VERSION" == "3.13" ]]; then
"$python" -m pip install numpy==1.26.4
else
"$python" -m pip install numpy==1.26.0
fi
fi
"$python" -m pip install *.whl $TFCI_PYTHON_VERIFY_PIP_INSTALL_ARGS
if [[ "$TFCI_WHL_IMPORT_TEST_ENABLE" == "1" ]]; then
Expand Down
21 changes: 17 additions & 4 deletions ci/official/utilities/setup_macos.sh
Original file line number Diff line number Diff line change
Expand Up @@ -61,10 +61,23 @@ fi
# those VMs does not support installing Python 3.12 and above which we need
# for running smoke tests in nightly/release wheel builds.
if [[ "${TFCI_MACOS_UPGRADE_PYENV_ENABLE}" == 1 ]]; then
# The TFCI Mac VM image seems to have uncommitted local changes to the Pyenv
# repository so we have to discard them and reset the working directory before
# we can pull in the latest changes.
cd /Users/kbuilder/.pyenv/ && git reset --hard HEAD && git pull && cd -
echo "Upgrading pyenv..."
echo "Current pyevn version: $(pyenv --version)"

# Check if pyenv is managed by homebrew. If so, update and upgrade pyenv.
# Otherwise, install the latest pyenv from github.
if command -v brew &> /dev/null && brew list pyenv &> /dev/null; then
# On "ventura-slcn" VMs, pyenv is managed via Homebrew.
echo "pyenv is installed and managed by homebrew."
brew update && brew upgrade pyenv
else
echo "pyenv is not managed by homebrew. Installing it via github..."
# On "ventura" VMs, pyenv is not managed by Homebrew. Install the latest
# pyenv from github.
rm -rf "$PYENV_ROOT"
git clone https://github.com/pyenv/pyenv.git "$PYENV_ROOT"
fi
echo "Upgraded pyenv version: $(pyenv --version)"
fi

# "TFCI_MACOS_PYENV_INSTALL_ENABLE" controls whether to use Pyenv to install
Expand Down
Loading