Re: [TCWG CI] Regression caused by gcc: rs6000: Harden mma_init_builtins
> On 31 May 2022, at 17:43, Peter Bergner wrote: > > On 5/31/22 8:09 AM, ci_not...@linaro.org wrote: >> [TCWG CI] Regression caused by gcc: rs6000: Harden mma_init_builtins: >> commit 6278065af07634278ba30029d92a82b089969baa >> Author: Peter Bergner >> >>rs6000: Harden mma_init_builtins >> >> Results regressed to >> # reset_artifacts: >> -10 >> # build_abe binutils: >> -9 >> # build_abe stage1: >> -5 >> # build_abe qemu: >> -2 >> # linux_n_obj: >> 33 >> # First few build errors in logs: >> # 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such >> file or directory >> # 00:03:13 make[2]: *** [scripts/gcc-plugins/latent_entropy_plugin.so] Error >> 1 >> # 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such >> file or directory >> # 00:03:13 make[2]: *** [scripts/gcc-plugins/stackleak_plugin.so] Error 1 >> # 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such >> file or directory >> # 00:03:13 make[2]: *** [scripts/gcc-plugins/randomize_layout_plugin.so] >> Error 1 >> # 00:03:13 make[1]: *** [scripts/gcc-plugins] Error 2 >> # 00:03:14 make: *** [scripts] Error 2 > > It seems your CI tester really doesn't like me! ;-) > Given my patch above could not have affected the existence of that > header file, I'll ignore this one too. Hi Peter, I'm suspecting a makefile bug in Linux kernel that makes build process unreliable. It seems there's a missing dependency between latent_entropy_plugin.so and generated/utsrelease.h header. I'm continuing to investigate this, but, meanwhile, [1] should fix spamming of upstream developers. [1] https://review.linaro.org/c/toolchain/jenkins-scripts/+/41433 Regards, -- Maxim Kuvyrkov https://www.linaro.org ___ linaro-toolchain mailing list -- linaro-toolchain@lists.linaro.org To unsubscribe send an email to linaro-toolchain-le...@lists.linaro.org
Re: [TCWG CI] Regression caused by gcc: rs6000: Harden mma_init_builtins
On 5/31/22 8:09 AM, ci_not...@linaro.org wrote: > [TCWG CI] Regression caused by gcc: rs6000: Harden mma_init_builtins: > commit 6278065af07634278ba30029d92a82b089969baa > Author: Peter Bergner > > rs6000: Harden mma_init_builtins > > Results regressed to > # reset_artifacts: > -10 > # build_abe binutils: > -9 > # build_abe stage1: > -5 > # build_abe qemu: > -2 > # linux_n_obj: > 33 > # First few build errors in logs: > # 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such > file or directory > # 00:03:13 make[2]: *** [scripts/gcc-plugins/latent_entropy_plugin.so] Error 1 > # 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such > file or directory > # 00:03:13 make[2]: *** [scripts/gcc-plugins/stackleak_plugin.so] Error 1 > # 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such > file or directory > # 00:03:13 make[2]: *** [scripts/gcc-plugins/randomize_layout_plugin.so] > Error 1 > # 00:03:13 make[1]: *** [scripts/gcc-plugins] Error 2 > # 00:03:14 make: *** [scripts] Error 2 It seems your CI tester really doesn't like me! ;-) Given my patch above could not have affected the existence of that header file, I'll ignore this one too. Peter ___ linaro-toolchain mailing list -- linaro-toolchain@lists.linaro.org To unsubscribe send an email to linaro-toolchain-le...@lists.linaro.org
[TCWG CI] Regression caused by gcc: rs6000: Harden mma_init_builtins
[TCWG CI] Regression caused by gcc: rs6000: Harden mma_init_builtins:
commit 6278065af07634278ba30029d92a82b089969baa
Author: Peter Bergner
rs6000: Harden mma_init_builtins
Results regressed to
# reset_artifacts:
-10
# build_abe binutils:
-9
# build_abe stage1:
-5
# build_abe qemu:
-2
# linux_n_obj:
33
# First few build errors in logs:
# 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such file
or directory
# 00:03:13 make[2]: *** [scripts/gcc-plugins/latent_entropy_plugin.so] Error 1
# 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such file
or directory
# 00:03:13 make[2]: *** [scripts/gcc-plugins/stackleak_plugin.so] Error 1
# 00:03:13 cc1plus: fatal error: ./include/generated/utsrelease.h: No such file
or directory
# 00:03:13 make[2]: *** [scripts/gcc-plugins/randomize_layout_plugin.so] Error 1
# 00:03:13 make[1]: *** [scripts/gcc-plugins] Error 2
# 00:03:14 make: *** [scripts] Error 2
from
# reset_artifacts:
-10
# build_abe binutils:
-9
# build_abe stage1:
-5
# build_abe qemu:
-2
# linux_n_obj:
31574
# linux build successful:
all
# linux boot successful:
boot
THIS IS THE END OF INTERESTING STUFF. BELOW ARE LINKS TO BUILDS, REPRODUCTION
INSTRUCTIONS, AND THE RAW COMMIT.
This commit has regressed these CI configurations:
- tcwg_kernel/gnu-release-aarch64-next-allmodconfig
First_bad build:
https://ci.linaro.org/job/tcwg_kernel-gnu-bisect-gnu-release-aarch64-next-allmodconfig/129/artifact/artifacts/build-6278065af07634278ba30029d92a82b089969baa/
Last_good build:
https://ci.linaro.org/job/tcwg_kernel-gnu-bisect-gnu-release-aarch64-next-allmodconfig/129/artifact/artifacts/build-a5b5c7c2840ebb0aaa0d0f1e090fb2fc6942196d/
Baseline build:
https://ci.linaro.org/job/tcwg_kernel-gnu-bisect-gnu-release-aarch64-next-allmodconfig/129/artifact/artifacts/build-baseline/
Even more details:
https://ci.linaro.org/job/tcwg_kernel-gnu-bisect-gnu-release-aarch64-next-allmodconfig/129/artifact/artifacts/
Reproduce builds:
mkdir investigate-gcc-6278065af07634278ba30029d92a82b089969baa
cd investigate-gcc-6278065af07634278ba30029d92a82b089969baa
# Fetch scripts
git clone https://git.linaro.org/toolchain/jenkins-scripts
# Fetch manifests and test.sh script
mkdir -p artifacts/manifests
curl -o artifacts/manifests/build-baseline.sh
https://ci.linaro.org/job/tcwg_kernel-gnu-bisect-gnu-release-aarch64-next-allmodconfig/129/artifact/artifacts/manifests/build-baseline.sh
--fail
curl -o artifacts/manifests/build-parameters.sh
https://ci.linaro.org/job/tcwg_kernel-gnu-bisect-gnu-release-aarch64-next-allmodconfig/129/artifact/artifacts/manifests/build-parameters.sh
--fail
curl -o artifacts/test.sh
https://ci.linaro.org/job/tcwg_kernel-gnu-bisect-gnu-release-aarch64-next-allmodconfig/129/artifact/artifacts/test.sh
--fail
chmod +x artifacts/test.sh
# Reproduce the baseline build (build all pre-requisites)
./jenkins-scripts/tcwg_kernel-build.sh @@ artifacts/manifests/build-baseline.sh
# Save baseline build state (which is then restored in artifacts/test.sh)
mkdir -p ./bisect
rsync -a --del --delete-excluded --exclude /bisect/ --exclude /artifacts/
--exclude /gcc/ ./ ./bisect/baseline/
cd gcc
# Reproduce first_bad build
git checkout --detach 6278065af07634278ba30029d92a82b089969baa
../artifacts/test.sh
# Reproduce last_good build
git checkout --detach a5b5c7c2840ebb0aaa0d0f1e090fb2fc6942196d
../artifacts/test.sh
cd ..
Full commit (up to 1000 lines):
commit 6278065af07634278ba30029d92a82b089969baa
Author: Peter Bergner
Date: Wed Jul 7 11:39:34 2021 -0500
rs6000: Harden mma_init_builtins
The previous MMA patch added some fragile code to initialize its new
built-ins. This patch hardens the initialization.
2021-07-07 Peter Bergner
gcc/
* config/rs6000/rs6000-call.c (mma_init_builtins): Use
VSX_BUILTIN_LXVP
and VSX_BUILTIN_STXVP.
---
gcc/config/rs6000/rs6000-call.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/gcc/config/rs6000/rs6000-call.c b/gcc/config/rs6000/rs6000-call.c
index 6115e3b34d9..904e104c058 100644
--- a/gcc/config/rs6000/rs6000-call.c
+++ b/gcc/config/rs6000/rs6000-call.c
@@ -14293,7 +14293,8 @@ mma_init_builtins (void)
/* Some MMA built-ins that are expanded into gimple are converted
into internal MMA built-ins that are expanded into rtl.
The internal built-in follows immediately after this built-in. */
- if (d[1].icode != CODE_FOR_nothing)
+ if (d->code != VSX_BUILTIN_LXVP
+ && d->code != VSX_BUILTIN_STXVP)
{
op[nopnds++] = void_type_node;
icode = d[1].icode;
___
linaro-toolchain mailing list -- linaro-toolchain@lists.linaro.org
To unsubscribe send an email to linaro-toolchain-le...@lists.linaro.org
