| Issue |
73992
|
| Summary |
[BOLT] ananicy-cpp binary broken with BOLT optimizations applied
|
| Labels |
BOLT
|
| Assignees |
|
| Reporter |
ms178
|
I've been experimenting with the BOLT script in ananicy-cpp for some time with clang-git, I've noticed only recently that the binary is actually broken when using BOLT (ddc6ef46418cc419041d0bd85de40d0eb9396848) on my Haswell-EP system.
For details, either take a look at the [PKGBUILD_BOLT](https://github.com/ms178/archpkgbuilds/blob/main/packages/ananicy-cpp/PKGBUILD_BOLT) in my repo, I will attach it here as well. I've tried some extra BOLT optimizations and apply them via the [patch.patch](https://github.com/ms178/archpkgbuilds/blob/main/packages/ananicy-cpp/patch.patch), but in my tests the binary was also broken with the out-of-the box options.
[PKGBUILD_BOLT.txt](https://github.com/llvm/llvm-project/files/13518068/PKGBUILD_BOLT.txt)
The compilation process itself went fine with the first set of flags, but a `sudo systemctl status ananicy-cpp.service` reveals that the service fails to run.
These flags in /etc/makepkg.conf compile fine but should reproduce the broken binary on any Arch-based distro:
```
export CC=clang
export CXX=clang++
export CC_LD=lld
export CXX_LD=lld
export AR=llvm-ar
export NM=llvm-nm
export STRIP=llvm-strip
export OBJCOPY=llvm-objcopy
export OBJDUMP=llvm-objdump
export READELF=llvm-readelf
export RANLIB=llvm-ranlib
export HOSTCC=clang
export HOSTCXX=clang++
export HOSTAR=llvm-ar
export CPPFLAGS="-D_FORTIFY_SOURCE=0"
export CFLAGS="-O3 -march=native -mtune=native -maes -mllvm -inline-threshold=1000 -mllvm -extra-vectorizer-passes -mllvm -enable-cond-stores-vec -mllvm -slp-vectorize-hor-store -mllvm -enable-loopinterchange -mllvm -enable-loop-distribute -mllvm -enable-unroll-and-jam -mllvm -enable-loop-flatten -mllvm -interleave-small-loop-scalar-reduction -mllvm -unroll-runtime-multi-exit -mllvm -aggressive-ext-opt -fno-math-errno -fno-trapping-math -falign-functions=32 -funroll-loops -fno-semantic-interposition -fcf-protection=none -mharden-sls=none -fomit-frame-pointer -mprefer-vector-width=256 -flto -freroll-loops -mllvm -adce-remove-loops -mllvm -enable-ext-tsp-block-placement=1 -mllvm -enable-gvn-hoist -mllvm -vp-counters-per-site=6 -fgnuc-version=6.5.0"
export CXXFLAGS="${CFLAGS}"
export LDFLAGS="-Wl,--lto-O3,-O3,-Bsymbolic-functions,--as-needed -march=native -maes -flto -fuse-ld=lld -Wl,-zmax-page-size=0x200000 -freroll-loops"
CCLDFLAGS="$LDFLAGS"
CXXLDFLAGS="$LDFLAGS"
export ASFLAGS="-D__AVX__=1 -D__AVX2__=1 -msse2avx -D__FMA__=1"
```
When using the following flags, BOLT even crashs:
```
export CC=clang
export CXX=clang++
export CC_LD=lld
export CXX_LD=lld
export AR=llvm-ar
export NM=llvm-nm
export STRIP=llvm-strip
export OBJCOPY=llvm-objcopy
export OBJDUMP=llvm-objdump
export READELF=llvm-readelf
export RANLIB=llvm-ranlib
export HOSTCC=clang
export HOSTCXX=clang++
export HOSTAR=llvm-ar
export CPPFLAGS="-D_FORTIFY_SOURCE=0"
export CFLAGS="-O3 -march=native -mtune=native -maes -mbmi2 -mpclmul -mllvm -inline-threshold=1000 -mllvm -polly -mllvm -polly-position=early -mllvm -polly-dependences-computeout=900000000 -mllvm -polly-detect-profitability-min-per-loop-insts=40 -mllvm -polly-tiling=true -mllvm -polly-prevect-width=256 -mllvm -polly-vectorizer=stripmine -mllvm -polly-invariant-load-hoisting -mllvm -polly-loopfusion-greedy -mllvm -polly-run-inliner -mllvm -polly-run-dce -mllvm -polly-enable-delicm=true -mllvm -polly -fmerge-all-constants -mllvm -extra-vectorizer-passes -mllvm -enable-cond-stores-vec -mllvm -slp-vectorize-hor-store -mllvm -enable-loopinterchange -mllvm -enable-loop-distribute -mllvm -enable-unroll-and-jam -mllvm -enable-loop-flatten -mllvm -interleave-small-loop-scalar-reduction -mllvm -unroll-runtime-multi-exit -mllvm -aggressive-ext-opt -fno-math-errno -fno-trapping-math -falign-functions=32 -funroll-loops -fno-semantic-interposition -fcf-protection=none -mharden-sls=none -fomit-frame-pointer -mprefer-vector-width=256 -flto -fwhole-program-vtables -fsplit-lto-unit -mllvm -adce-remove-loops -mllvm -enable-ext-tsp-block-placement=1 -mllvm -enable-gvn-hoist -mllvm -enable-dfa-jump-thread -Wno-error -fgnuc-version=13.2.1 -ffp-contract=fast -fdata-sections -ffunction-sections -fno-unique-section-names -fsplit-machine-functions -freroll-loops -w"
export CXXFLAGS="${CFLAGS} -Wp,-U_GLIBCXX_ASSERTIONS"
export LDFLAGS="-Wl,--lto-CGO3 -Wl,--gc-sections -Wl,--icf=all -Wl,--lto-O3,-O3,-Bsymbolic-functions,--as-needed -fcf-protection=none -mharden-sls=none -Wl,-mllvm,-polly -Wl,-mllvm,-polly-invariant-load-hoisting -Wl,-mllvm -Wl,-polly-loopfusion-greedy -Wl,-mllvm -Wl,-polly-run-inliner -Wl,-mllvm -Wl,-polly-run-dce -Wl,-mllvm -Wl,-polly-enable-delicm=true -Wl,-mllvm -Wl,-extra-vectorizer-passes -Wl,-mllvm -Wl,-enable-cond-stores-vec -Wl,-mllvm -Wl,-slp-vectorize-hor-store -Wl,-mllvm -Wl,-enable-loopinterchange -Wl,-mllvm -Wl,-enable-loop-distribute -Wl,-mllvm -Wl,-enable-unroll-and-jam -Wl,-mllvm -Wl,-enable-loop-flatten -Wl,-mllvm -Wl,-interleave-small-loop-scalar-reduction -Wl,-mllvm -Wl,-unroll-runtime-multi-exit -Wl,-mllvm -Wl,-aggressive-ext-opt -Wl,-mllvm -Wl,-enable-interleaved-mem-accesses -Wl,-mllvm -Wl,-enable-masked-interleaved-mem-accesses -march=native -maes -mbmi2 -mpclmul -flto -fwhole-program-vtables -fuse-ld=lld -Wl,-zmax-page-size=0x200000 -freroll-loops -Wl,-mllvm -Wl,-adce-remove-loops -Wl,-mllvm -Wl,-enable-ext-tsp-block-placement=1 -Wl,-mllvm -Wl,-enable-gvn-hoist -Wl,-mllvm -Wl,-enable-dfa-jump-thread -Wl,--push-state -Wl,-whole-archive -ljemalloc_pic -Wl,--pop-state -lpthread -lstdc++ -lm -ldl -Wl,--undefined-version"
export CCLDFLAGS="$LDFLAGS"
export CXXLDFLAGS="$LDFLAGS"
export ASFLAGS="-D__AVX__=1 -D__AVX2__=1 -msse2avx -D__FMA__=1"
```
![]()
_______________________________________________
llvm-bugs mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
