Skip to content
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

Integrate LLVM at llvm/llvm-project@ef8de6 #18004

Merged
merged 10 commits into from
Jul 25, 2024
Merged

Conversation

ScottTodd
Copy link
Member

@ScottTodd ScottTodd commented Jul 24, 2024

Advance to llvm/llvm-project@ef8de68

…sforms] Dialect conversion: Simplify handling of dropped arguments (#97213) (Matthias Springer on 2024-07-20 10:12:13 +0200) (0 of 29)
… for emit functions in BytecodeWriter.cpp (#99558) (Kevin Gleason on 2024-07-20 12:57:50 -0500) (0 of 28)
… unused bool arguments from createMCObjectStreamer callers (Fangrui Song on 2024-07-20 21:27:36 -0700) (0 of 27)
…] Add comments in a test (nfc) (#99810) (Andrzej Warzyński on 2024-07-21 17:44:12 +0100) (0 of 26)
…ate for new deps (#100166) (Keith Smiley on 2024-07-23 10:09:52 -0700) (15 of 25)
…ee0226b4836e66574f225f724e768 [MLIR][DebugInfo] Enable the use of DILocalVariable DIFlags (#100190) (Walter Erquinigo on 2024-07-23 19:40:22 -0400) (0 of 9)
@ScottTodd ScottTodd changed the title Integrate LLVM at llvm/llvm-project@ef8de68faebee0226b4836e66574f225f724e768 Integrate LLVM at llvm/llvm-project@ef8de6 Jul 24, 2024
@ScottTodd ScottTodd marked this pull request as ready for review July 25, 2024 18:35
@ScottTodd ScottTodd requested a review from aviator19941 July 25, 2024 18:45
Copy link

Abbreviated Benchmark Summary

@ commit 1c5005a6d5dc99bdb75910831f289f10c9c14170 (vs. base 80be52bdc2cdc63f3d45c1413ac62d62a635a8ed)

Data-Tiling Comparison Table

Click to show
Name No-DT (baseline) DT-Only DT-UK
BertLargeTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[30-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 745.328 (1.0X) 276.020 (2.7X) 224.383 (3.3X)
DeepLabV3_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 7.077 (1.0X) 9.373 (0.8X) 8.560 (0.8X)
EfficientNetV2STF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 36.272 (1.0X) 36.604 (1.0X) 34.833 (1.0X)
EfficientNet_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.867 (1.0X) 10.975 (0.5X) 5.061 (1.2X)
GPT2_117M_TF_1X1XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 9.234 (1.0X) 8.550 (1.1X) 8.536 (1.1X)
GPT2_117M_TF_1X4XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 11.029 (1.0X) 9.025 (1.2X) 9.010 (1.2X)
MiniLML12H384Uncased(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 12.071 (1.0X) 15.461 (0.8X) 14.053 (0.9X)
MobileBertSquad_fp16(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 34.794 (1.0X) 66.308 (0.5X) 62.064 (0.6X)
MobileBertSquad_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 35.453 (1.0X) 66.185 (0.5X) 62.293 (0.6X)
MobileBertSquad_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 68.811 (1.0X) 134.524 (0.5X) 64.973 (1.1X)
MobileNetV1_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 4.787 (1.0X) 5.332 (0.9X) 4.618 (1.0X)
MobileNetV2_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 3.813 (1.0X) 5.315 (0.7X) 4.958 (0.8X)
MobileNetV2_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.978 (1.0X) 9.542 (0.6X) 5.456 (1.1X)
MobileNetV3Small_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 2.851 (1.0X) 3.389 (0.8X) 2.819 (1.0X)
MobileSSD_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 8.612 (1.0X) 11.030 (0.8X) 9.964 (0.9X)
PersonDetect_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 0.779 (1.0X) 1.406 (0.6X) 0.611 (1.3X)
PoseNet_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 4.197 (1.0X) 5.832 (0.7X) 5.278 (0.8X)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 7.565 (1.0X) 7.554 (1.0X) 7.577 (1.0X)
matmul_256x256x2048_i8_i8_i32_tile_config_default(linalg) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 6.669 (1.0X) 13.238 (0.5X) 1.808 (3.7X)
BertForMaskedLMTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[30-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 217.293 (1.0X) 135.949 (1.6X) 108.382 (2.0X)
DeepLabV3_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 32.256 (1.0X) 36.673 (0.9X) 29.964 (1.1X)
EfficientNetV2STF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 277.779 (1.0X) 258.898 (1.1X) 230.839 (1.2X)
EfficientNet_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 27.006 (1.0X) 51.428 (0.5X) 13.088 (2.1X)
GPT2_117M_TF_1X1XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 69.765 (1.0X) 38.495 (1.8X) 37.997 (1.8X)
GPT2_117M_TF_1X4XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 87.186 (1.0X) 41.808 (2.1X) 41.052 (2.1X)
MiniLML12H384Uncased(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 80.198 (1.0X) 76.977 (1.0X) 56.282 (1.4X)
MobileBertSquad_fp16(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 180.364 (1.0X) 249.822 (0.7X) 186.391 (1.0X)
MobileBertSquad_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 182.243 (1.0X) 253.276 (0.7X) 191.277 (1.0X)
MobileBertSquad_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 516.276 (1.0X) 1085.368 (0.5X) 241.046 (2.1X)
MobileNetV1_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 25.365 (1.0X) 22.880 (1.1X) 17.898 (1.4X)
MobileNetV2_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 11.922 (1.0X) 14.997 (0.8X) 11.459 (1.0X)
MobileNetV2_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 21.576 (1.0X) 42.280 (0.5X) 11.885 (1.8X)
MobileNetV3Small_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 2.751 (1.0X) 3.264 (0.8X) 2.704 (1.0X)
MobileSSD_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 33.423 (1.0X) 39.377 (0.8X) 31.642 (1.1X)
PersonDetect_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 0.711 (1.0X) 1.314 (0.5X) 0.549 (1.3X)
PoseNet_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 17.846 (1.0X) 23.682 (0.8X) 19.015 (0.9X)
matmul_1x256x2048_i8_i4_i32_tile_config_default(linalg) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 0.054 (1.0X) 0.054 (1.0X) 0.054 (1.0X)
matmul_1x256x2048_i8_i8_i32_tile_config_default(linalg) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 0.042 (1.0X) 0.225 (0.2X) 0.021 (2.0X)

No improved or regressed benchmarks 🏖️

No improved or regressed compilation metrics 🏖️

For more information:

Source Workflow Run

@ScottTodd ScottTodd merged commit 456d80c into main Jul 25, 2024
53 checks passed
@ScottTodd ScottTodd deleted the integrates/llvm-20240724 branch July 25, 2024 22:52
LLITCHEV pushed a commit to LLITCHEV/iree that referenced this pull request Jul 30, 2024
Advance to
llvm/llvm-project@ef8de68

* Still carrying a revert of
llvm/llvm-project@fa06668
* New revert of
llvm/llvm-project@bbd4af5
due to asserts/crashes during f64 to f32 demotion
* A patch like
ScottTodd/llvm-project@f766cd2
is being discussed to fix-forward
* Patch for ambiguous conversion in
llvm/llvm-project@e3b30bc
* Also landed that patch upstream, for our next integrate:
llvm/llvm-project@4f80508
* Local build fixes for
llvm/llvm-project@ef8de68

Signed-off-by: Lubo Litchev <lubol@google.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants