8000 [BE] Move cuda12.6 builds to gcc11 (#148740) · pytorch/pytorch@7b79e17 · GitHub
[go: up one dir, main page]
Skip to content

Commit 7b79e17

Browse files
malfetpytorchmergebot
malfet
authored and
pytorchmergebot
committed
[BE] Move cuda12.6 builds to gcc11 (#148740)
I.e. `s/pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9/pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11/` Which accidentally fixes undefined symbol references errors namely ``` /usr/bin/ld: /var/lib/jenkins/cpp-build/caffe2/build/lib/libtorch_cuda.so: undefined reference to `std::__throw_bad_array_new_length()' ``` Which happens because `libmagma.a` that were build with gcc-11 (after #148135 ) contains symbols which are defined in `/opt/rh/gcc-toolset-11/root/usr/lib/gcc/x86_64-redhat-linux/11/libstdc++_nonshared.a` but missing from the corresponding library bundled with `g++-9`) Though I could not figure out what flags one must use to trigger generation of those symbols, see https://godbolt.org/z/E9KfdhzzY or ``` $ echo "int* foo(int x) { return new int[x];}"|g++ -std=c++17 -S -O3 -x c++ -o - - .file "" .text .section .text.unlikely,"ax",@progbits .LCOLDB0: .text .LHOTB0: .p2align 4 .globl _Z3fooi .type _Z3fooi, @function _Z3fooi: .LFB0: .cfi_startproc endbr64 movslq %edi, %rdi subq $8, %rsp .cfi_def_cfa_offset 16 movabsq $2305843009213693950, %rax cmpq %rax, %rdi ja .L2 salq $2, %rdi addq $8, %rsp .cfi_def_cfa_offset 8 jmp _Znam@PLT .cfi_endproc .section .text.unlikely .cfi_startproc .type _Z3fooi.cold, @function _Z3fooi.cold: .LFSB0: .L2: .cfi_def_cfa_offset 16 call __cxa_throw_bad_array_new_length@PLT .cfi_endproc ``` Fixes #148728 and #148495 Pull Request resolved: #148740 Approved by: https://github.com/wdvr, https://github.com/atalman, https://github.com/Skylion007, https://github.com/ZainRizvi
1 parent 08baaa7 commit 7b79e17

File tree

6 files changed

+74
-74
lines changed

6 files changed

+74
-74
lines changed

‎.ci/docker/build.sh

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -99,11 +99,11 @@ fi
9999
# configuration, so we hardcode everything here rather than do it
100100
# from scratch
101101
case "$image" in
102-
pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9)
102+
pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11)
103103
CUDA_VERSION=12.6.3
104104
CUDNN_VERSION=9
105105
ANACONDA_PYTHON_VERSION=3.10
106-
GCC_VERSION=9
106+
GCC_VERSION=11
107107
PROTOBUF=yes
108108
DB=yes
109109
VISION=yes

‎.github/workflows/docker-builds.yml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -52,7 +52,7 @@ jobs:
5252
pytorch-linux-focal-cuda12.4-cudnn9-py3-gcc9-inductor-benchmarks,
5353
pytorch-linux-focal-cuda12.4-cudnn9-py3.12-gcc9-inductor-benchmarks,
5454
pytorch-linux-focal-cuda12.4-cudnn9-py3.13-gcc9-inductor-benchmarks,
55-
pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9,
55+
pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11,
5656
pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9-inductor-benchmarks,
5757
pytorch-linux-focal-cuda12.6-cudnn9-py3.12-gcc9-inductor-benchmarks,
5858
pytorch-linux-focal-cuda12.6-cudnn9-py3.13-gcc9-inductor-benchmarks,

‎.github/workflows/periodic.yml

Lines changed: 24 additions & 24 deletions
Original file line numberDiff line numberDiff line change
@@ -49,14 +49,14 @@ jobs:
4949
curr_branch: ${{ github.head_ref || github.ref_name }}
5050
curr_ref_type: ${{ github.ref_type }}
5151

52-
linux-focal-cuda12_6-py3_10-gcc9-build:
53-
name: linux-focal-cuda12.6-py3.10-gcc9
52+
linux-focal-cuda12_6-py3_10-gcc11-build:
53+
name: linux-focal-cuda12.6-py3.10-gcc11
5454
uses: ./.github/workflows/_linux-build.yml
5555
needs: get-label-type
5656
with:
5757
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
58-
build-environment: linux-focal-cuda12.6-py3.10-gcc9
59-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
58+
build-environment: linux-focal-cuda12.6-py3.10-gcc11
59+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
6060
test-matrix: |
6161
{ include: [
6262
{ config: "nogpu_AVX512", shard: 1, num_shards: 2, runner: "${{ needs.get-label-type.outputs.label-type }}linux.4xlarge" },
@@ -67,16 +67,16 @@ jobs:
6767
]}
6868
secrets: inherit
6969

70-
linux-focal-cuda12_6-py3_10-gcc9-test:
71-
name: linux-focal-cuda12.6-py3.10-gcc9
70+
linux-focal-cuda12_6-py3_10-gcc11-test:
71+
name: linux-focal-cuda12.6-py3.10-gcc11
7272
uses: ./.github/workflows/_linux-test.yml
7373
needs:
74-
- linux-focal-cuda12_6-py3_10-gcc9-build
74+
- linux-focal-cuda12_6-py3_10-gcc11-build
7575
- target-determination
7676
with:
77-
build-environment: linux-focal-cuda12.6-py3.10-gcc9
78-
docker-image: ${{ needs.linux-focal-cuda12_6-py3_10-gcc9-build.outputs.docker-image }}
79-
test-matrix: ${{ needs.linux-focal-cuda12_6-py3_10-gcc9-build.outputs.test-matrix }}
77+
build-environment: linux-focal-cuda12.6-py3.10-gcc11
78+
docker-image: ${{ needs.linux-focal-cuda12_6-py3_10-gcc11-build.outputs.docker-image }}
79+
test-matrix: ${{ needs.linux-focal-cuda12_6-py3_10-gcc11-build.outputs.test-matrix }}
8080
secrets: inherit
8181

8282
linux-focal-cuda11_8-py3_9-gcc9-build:
@@ -170,14 +170,14 @@ jobs:
170170
test-matrix: ${{ needs.linux-focal-rocm6_3-py3_10-build.outputs.test-matrix }}
171171
secrets: inherit
172172

173-
linux-focal-cuda12_6-py3-gcc9-slow-gradcheck-build:
174-
name: linux-focal-cuda12.6-py3-gcc9-slow-gradcheck
173+
linux-focal-cuda12_6-py3-gcc11-slow-gradcheck-build:
174+
name: linux-focal-cuda12.6-py3-gcc11-slow-gradcheck
175175
uses: ./.github/workflows/_linux-build.yml
176176
needs: get-label-type
177177
with:
178178
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
179-
build-environment: linux-focal-cuda12.6-py3-gcc9-slow-gradcheck
180-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
179+
build-environment: linux-focal-cuda12.6-py3-gcc11-slow-gradcheck
180+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
181181
cuda-arch-list: 8.6
182182
test-matrix: |
183183
{ include: [
@@ -192,27 +192,27 @@ jobs:
192192
]}
193193
secrets: inherit
194194

195-
linux-focal-cuda12_6-py3-gcc9-slow-gradcheck-test:
196-
name: linux-focal-cuda12.6-py3-gcc9-slow-gradcheck
195+
linux-focal-cuda12_6-py3-gcc11-slow-gradcheck-test:
196+
name: linux-focal-cuda12.6-py3-gcc11-slow-gradcheck
197197
uses: ./.github/workflows/_linux-test.yml
198198
needs:
199-
- linux-focal-cuda12_6-py3-gcc9-slow-gradcheck-build
199+
- linux-focal-cuda12_6-py3-gcc11-slow-gradcheck-build
200200
- target-determination
201201
with:
202-
build-environment: linux-focal-cuda12.6-py3-gcc9-slow-gradcheck
203-
docker-image: ${{ needs.linux-focal-cuda12_6-py3-gcc9-slow-gradcheck-build.outputs.docker-image }}
204-
test-matrix: ${{ needs.linux-focal-cuda12_6-py3-gcc9-slow-gradcheck-build.outputs.test-matrix }}
202+
build-environment: linux-focal-cuda12.6-py3-gcc11-slow-gradcheck
203+
docker-image: ${{ needs.linux-focal-cuda12_6-py3-gcc11-slow-gradcheck-build.outputs.docker-image }}
204+
test-matrix: ${{ needs.linux-focal-cuda12_6-py3-gcc11-slow-gradcheck-build.outputs.test-matrix }}
205205
timeout-minutes: 300
206206
secrets: inherit
207207

208-
linux-focal-cuda12_6-py3_10-gcc9-bazel-test:
209-
name: linux-focal-cuda12.6-py3.10-gcc9-bazel-test
208+
linux-focal-cuda12_6-py3_10-gcc11-bazel-test:
209+
name: linux-focal-cuda12.6-py3.10-gcc11-bazel-test
210210
uses: ./.github/workflows/_bazel-build-test.yml
211211
needs: get-label-type
212212
with:
213213
runner: "${{ needs.get-label-type.outputs.label-type }}linux.large"
214-
build-environment: linux-focal-cuda12.6-py3.10-gcc9-bazel-test
215-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
214+
build-environment: linux-focal-cuda12.6-py3.10-gcc11-bazel-test
215+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
216216
cuda-version: "12.6"
217217
test-matrix: |
218218
{ include: [

‎.github/workflows/pull.yml

Lines changed: 29 additions & 29 deletions
Original file line numberDiff line numberDiff line change
@@ -278,14 +278,14 @@ jobs:
278278
test-matrix: ${{ needs.linux-focal-cuda11_8-py3_10-gcc9-build.outputs.test-matrix }}
279279
secrets: inherit
280280

281-
linux-focal-cuda12_6-py3_10-gcc9-build:
282-
name: linux-focal-cuda12.6-py3.10-gcc9
281+
linux-focal-cuda12_6-py3_10-gcc11-build:
282+
name: linux-focal-cuda12.6-py3.10-gcc11
283283
uses: ./.github/workflows/_linux-build.yml
284284
needs: get-label-type
285285
with:
286286
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
287-
build-environment: linux-focal-cuda12.6-py3.10-gcc9
288-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
287+
build-environment: linux-focal-cuda12.6-py3.10-gcc11
288+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
289289
test-matrix: |
290290
{ include: [
291291
{ config: "default", shard: 1, num_shards: 5, runner: "${{ needs.get-label-type.outputs.label-type }}linux.4xlarge.nvidia.gpu" },
@@ -296,17 +296,17 @@ jobs:
296296
]}
297297
secrets: inherit
298298

299-
linux-focal-cuda12_6-py3_10-gcc9-test:
300-
name: linux-focal-cuda12.6-py3.10-gcc9
299+
linux-focal-cuda12_6-py3_10-gcc11-test:
300+
name: linux-focal-cuda12.6-py3.10-gcc11
301301
uses: ./.github/workflows/_linux-test.yml
302302
needs:
303-
- linux-focal-cuda12_6-py3_10-gcc9-build
303+
- linux-focal-cuda12_6-py3_10-gcc11-build
304304
- target-determination
305305
with:
306306
timeout-minutes: 360
307-
build-environment: linux-focal-cuda12.6-py3.10-gcc9
308-
docker-image: ${{ needs.linux-focal-cuda12_6-py3_10-gcc9-build.outputs.docker-image }}
309-
test-matrix: ${{ needs.linux-focal-cuda12_6-py3_10-gcc9-build.outputs.test-matrix }}
307+
build-environment: linux-focal-cuda12.6-py3.10-gcc11
308+
docker-image: ${{ needs.linux-focal-cuda12_6-py3_10-gcc11-build.outputs.docker-image }}
309+
test-matrix: ${{ needs.linux-focal-cuda12_6-py3_10-gcc11-build.outputs.test-matrix }}
310310
secrets: inherit
311311

312312
linux-jammy-py3-clang12-mobile-build:
@@ -381,14 +381,14 @@ jobs:
381381
]}
382382
secrets: inherit
383383

384-
linux-focal-cpu-py3_10-gcc9-bazel-test:
385-
name: linux-focal-cpu-py3.10-gcc9-bazel-test
384+
linux-focal-cpu-py3_10-gcc11-bazel-test:
385+
name: linux-focal-cpu-py3.10-gcc11-bazel-test
386386
uses: ./.github/workflows/_bazel-build-test.yml
387387
needs: get-label-type
388388
with:
389389
runner: "${{ needs.get-label-type.outputs.label-type }}linux.large"
390-
build-environment: linux-focal-cuda12.6-py3.10-gcc9-bazel-test
391-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
390+
build-environment: linux-focal-cuda12.6-py3.10-gcc11-bazel-test
391+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
392392
cuda-version: cpu
393393
test-matrix: |
394394
{ include: [
@@ -402,7 +402,7 @@ jobs:
402402
needs: get-label-type
403403
with:
404404
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
405-
build-environment: linux-jammy-py3.9-gcc111-mobile-lightweight-dispatch-build
405+
build-environment: linux-jammy-py3.9-gcc11-mobile-lightweight-dispatch-build
 10000
406406
docker-image-name: pytorch-linux-jammy-py3.9-gcc11
407407
build-generates-artifacts: false
408408
test-matrix: |
@@ -430,14 +430,14 @@ jobs:
430430
]}
431431
secrets: inherit
432432

433-
linux-focal-cuda12_6-py3_10-gcc9-sm89-build:
434-
name: linux-focal-cuda12.6-py3.10-gcc9-sm89
433+
linux-focal-cuda12_6-py3_10-gcc11-sm89-build:
434+
name: linux-focal-cuda12.6-py3.10-gcc11-sm89
435435
uses: ./.github/workflows/_linux-build.yml
436436
needs: get-label-type
437437
with:
438438
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
439-
build-environment: linux-focal-cuda12.6-py3.10-gcc9-sm89
440-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
439+
build-environment: linux-focal-cuda12.6-py3.10-gcc11-sm89
440+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
441441
cuda-arch-list: 8.9
442442
test-matrix: |
443443
{ include: [
@@ -449,16 +449,16 @@ jobs:
449449
]}
450450
secrets: inherit
451451

452-
unstable-linux-focal-cuda12_6-py3_10-gcc9-sm89-build-xfail:
452+
unstable-linux-focal-cuda12_6-py3_10-gcc11-sm89-build-xfail:
453453
# A version of the build that sets a larger number of jobs for a build. May
454454
# OOM
455-
name: unstable-linux-focal-cuda12.6-py3.10-gcc9-sm89-xfail
455+
name: unstable-linux-focal-cuda12.6-py3.10-gcc11-sm89-xfail
456456
uses: ./.github/workflows/_linux-build.yml
457457
needs: get-label-type
458458
with:
459459
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
460-
build-environment: linux-focal-cuda12.6-py3.10-gcc9-sm89
461-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
460+
build-environment: linux-focal-cuda12.6-py3.10-gcc11-sm89
461+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
462462
cuda-arch-list: 8.9
463463
max-jobs: 4
464464
# Doesn't actually run tests, but need this in order to prevent the build
@@ -469,16 +469,16 @@ jobs:
469469
]}
470 10000 470
secrets: inherit
471471

472-
linux-focal-cuda12_6-py3_10-gcc9-sm89-test:
473-
name: linux-focal-cuda12.6-py3.10-gcc9-sm89
472+
linux-focal-cuda12_6-py3_10-gcc11-sm89-test:
473+
name: linux-focal-cuda12.6-py3.10-gcc11-sm89
474474
uses: ./.github/workflows/_linux-test.yml
475475
needs:
476-
- linux-focal-cuda12_6-py3_10-gcc9-sm89-build
476+
- linux-focal-cuda12_6-py3_10-gcc11-sm89-build
477477
- target-determination
478478
with:
479-
build-environment: linux-focal-cuda12.6-py3.10-gcc9-sm89
480-
docker-image: ${{ needs.linux-focal-cuda12_6-py3_10-gcc9-sm89-build.outputs.docker-image }}
481-
test-matrix: ${{ needs.linux-focal-cuda12_6-py3_10-gcc9-sm89-build.outputs.test-matrix }}
479+
build-environment: linux-focal-cuda12.6-py3.10-gcc11-sm89
480+
docker-image: ${{ needs.linux-focal-cuda12_6-py3_10-gcc11-sm89-build.outputs.docker-image }}
481+
test-matrix: ${{ needs.linux-focal-cuda12_6-py3_10-gcc11-sm89-build.outputs.test-matrix }}
482482
secrets: inherit
483483

484484
linux-jammy-py3-clang12-executorch-build:

‎.github/workflows/slow.yml

Lines changed: 10 additions & 10 deletions
Original file line numberDiff line numberDiff line change
@@ -47,14 +47,14 @@ jobs:
4747
curr_branch: ${{ github.head_ref || github.ref_name }}
4848
curr_ref_type: ${{ github.ref_type }}
4949

50-
linux-focal-cuda12_6-py3_10-gcc9-sm86-build:
51-
name: linux-focal-cuda12.6-py3.10-gcc9-sm86
50+
linux-focal-cuda12_6-py3_10-gcc11-sm86-build:
51+
name: linux-focal-cuda12.6-py3.10-gcc11-sm86
5252
uses: ./.github/workflows/_linux-build.yml
5353
needs: get-label-type
5454
with:
5555
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
56-
build-environment: linux-focal-cuda12.6-py3.10-gcc9-sm86
57-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
56+
build-environment: linux-focal-cuda12.6-py3.10-gcc11-sm86
57+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
5858
cuda-arch-list: 8.6
5959
test-matrix: |
6060
{ include: [
@@ -64,16 +64,16 @@ jobs:
6464
]}
6565
secrets: inherit
6666

67-
linux-focal-cuda12_6-py3_10-gcc9-sm86-test:
68-
name: linux-focal-cuda12.6-py3.10-gcc9-sm86
67+
linux-focal-cuda12_6-py3_10-gcc11-sm86-test:
68+
name: linux-focal-cuda12.6-py3.10-gcc11-sm86
6969
uses: ./.github/workflows/_linux-test.yml
7070
needs:
71-
- linux-focal-cuda12_6-py3_10-gcc9-sm86-build
71+
- linux-focal-cuda12_6-py3_10-gcc11-sm86-build
7272
- target-determination
7373
with:
74-
build-environment: linux-focal-cuda12.6-py3.10-gcc9-sm86
75-
docker-image: ${{ needs.linux-focal-cuda12_6-py3_10-gcc9-sm86-build.outputs.docker-image }}
76-
test-matrix: ${{ needs.linux-focal-cuda12_6-py3_10-gcc9-sm86-build.outputs.test-matrix }}
74+
build-environment: linux-focal-cuda12.6-py3.10-gcc11-sm86
75+
docker-image: ${{ needs.linux-focal-cuda12_6-py3_10-gcc11-sm86-build.outputs.docker-image }}
76+
test-matrix: ${{ needs.linux-focal-cuda12_6-py3_10-gcc11-sm86-build.outputs.test-matrix }}
7777
secrets: inherit
7878

7979
linux-focal-py3_9-clang10-build:

‎.github/workflows/trunk.yml

Lines changed: 8 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -45,13 +45,13 @@ jobs:
4545
curr_branch: ${{ github.head_ref || github.ref_name }}
4646
curr_ref_type: ${{ github.ref_type }}
4747

48-
libtorch-linux-focal-cuda12_6-py3_10-gcc9-debug-build:
49-
name: libtorch-linux-focal-cuda12.6-py3.10-gcc9-debug
48+
libtorch-linux-focal-cuda12_6-py3_10-gcc11-debug-build:
49+
name: libtorch-linux-focal-cuda12.6-py3.10-gcc11-debug
5050
uses: ./.github/workflows/_linux-build.yml
5151
needs: get-label-type
5252
with:
53-
build-environment: libtorch-linux-focal-cuda12.6-py3.10-gcc9
54-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
53+
build-environment: libtorch-linux-focal-cuda12.6-py3.10-gcc11
54+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
5555
build-generates-artifacts: false
5656
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
5757
runner: "linux.4xlarge"
@@ -62,14 +62,14 @@ jobs:
6262
secrets: inherit
6363

6464
# no-ops builds test USE_PER_OPERATOR_HEADERS=0 where ATen/ops is not generated
65-
linux-focal-cuda12_6-py3_10-gcc9-no-ops-build:
66-
name: linux-focal-cuda12.6-py3.10-gcc9-no-ops
65+
linux-focal-cuda12_6-py3_10-gcc11-no-ops-build:
66+
name: linux-focal-cuda12.6-py3.10-gcc11-no-ops
6767
uses: ./.github/workflows/_linux-build.yml
6868
needs: get-label-type
6969
with:
7070
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
71-
build-environment: linux-focal-cuda12.6-py3.10-gcc9-no-ops
72-
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc9
71+
build-environment: linux-focal-cuda12.6-py3.10-gcc11-no-ops
72+
docker-image-name: pytorch-linux-focal-cuda12.6-cudnn9-py3-gcc11
7373
test-matrix: |
7474
{ include: [
7575
{ config: "default", shard: 1, num_shards: 1 },

0 commit comments

Comments
 (0)
0