https://gcc.gnu.org/bugzilla/show_bug.cgi?id=96827
Bug ID: 96827
Summary: __m128i from _mm_set_epi32 is backwards with -O3
Product: gcc
Version: 10.2.1
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: gcc at froghat dot ca
Target Milestone: ---
Created attachment 49142
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=49142&action=edit
minimalish preprocessed sources that compiles to differently behaved programs
based on optimization level using gcc 10
Hi. Attached reproduces an issue where an __m128i value appears backwards on
gcc versions 10 and later with -O3.
I don't know what's going on but the compiler output does a different thing
than compared to that of different compilers, older version of gcc, or gcc 10
with lower optimization levels.
The output I expect is "6 4 2 0", with gcc 10 -O3 I get "0 2 4 6".
There are a couple modifications to the source that make the issue go away,
like adding...
> if (dude_[0] == 1234) dude_[0]--;
...after the for loop (but not before). Or using a loop conditional of `i < 2`
instead of `i < 3`.
And again this happens with -O3 but not with -O2.
clang 10 and apparently gcc 5 and 8 both give the expected output.
But this is not the case with gcc 10 from my system as well as when compiled
from a recent checkout of gcc (82030d51017323c5706d58d8c8626324ece007e4)
My system gcc:
gcc version 10.2.1 20200723 (Red Hat 10.2.1-1) (GCC)
Target: x86_64-redhat-linux
Configured with: ../configure --enable-bootstrap
--enable-languages=c,c++,fortran,objc,obj-c++,ada,go,d,lto --prefix=/usr
--mandir=/usr/share/man --infodir=/usr/share/info
--with-bugurl=http://bugzilla.redhat.com/bugzilla --enable-shared
--enable-threads=posix --enable-checking=release --enable-multilib
--with-system-zlib --enable-__cxa_atexit --disable-libunwind-exceptions
--enable-gnu-unique-object --enable-linker-build-id
--with-gcc-major-version-only --with-linker-hash-style=gnu --enable-plugin
--enable-initfini-array --with-isl --enable-offload-targets=nvptx-none
--without-cuda-driver --enable-gnu-indirect-function --enable-cet
--with-tune=generic --with-arch_32=i686 --build=x86_64-redhat-linux
A gcc from a recent git checkout:
Target: x86_64-pc-linux-gnu
Configured with: /home/sqwishy/src/gcc/configure --enable-languages=c
--disable-multilib
gcc version 11.0.0 20200827 (experimental) (GCC)
This is the command and output of compiling the attached file (min.i) with the
version of gcc I built from a recent checkout.
> ./usr/local/bin/gcc -v -O3 -Wall -Wextra -fno-strict-aliasing -fwrapv
> -fno-aggressive-loop-optimizations min.i
Using built-in specs.
COLLECT_GCC=./usr/local/bin/gcc
COLLECT_LTO_WRAPPER=/opt/gcc-inst/usr/local/bin/../libexec/gcc/x86_64-pc-linux-gnu/11.0.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /home/sqwishy/src/gcc/configure --enable-languages=c
--disable-multilib
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 11.0.0 20200827 (experimental) (GCC)
COLLECT_GCC_OPTIONS='-v' '-O3' '-Wall' '-Wextra' '-fno-strict-aliasing'
'-fwrapv' '-fno-aggressive-loop-optimizations' '-mtune=generic' '-march=x86-64'
'-dumpdir' 'a-'
/opt/gcc-inst/usr/local/bin/../libexec/gcc/x86_64-pc-linux-gnu/11.0.0/cc1
-fpreprocessed min.i -quiet -dumpdir a- -dumpbase min.i -dumpbase-ext .i
-mtune=generic -march=x86-64 -O3 -Wall -Wextra -version -fno-strict-aliasing
-fwrapv -fno-aggressive-loop-optimizations -o /tmp/ccUHO9zb.s
GNU C17 (GCC) version 11.0.0 20200827 (experimental) (x86_64-pc-linux-gnu)
compiled by GNU C version 11.0.0 20200827 (experimental), GMP version
6.1.2, MPFR version 4.0.2-p9, MPC version 1.1.0, isl version none
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096
GNU C17 (GCC) version 11.0.0 20200827 (experimental) (x86_64-pc-linux-gnu)
compiled by GNU C version 11.0.0 20200827 (experimental), GMP version
6.1.2, MPFR version 4.0.2-p9, MPC version 1.1.0, isl version none
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096
Compiler executable checksum: 04849eadbd492e04db2b98f5ed9cc34b
COLLECT_GCC_OPTIONS='-v' '-O3' '-Wall' '-Wextra' '-fno-strict-aliasing'
'-fwrapv' '-fno-aggressive-loop-optimizations' '-mtune=generic' '-march=x86-64'
'-dumpdir' 'a-'
as -v --64 -o /tmp/ccMwZ4Yb.o /tmp/ccUHO9zb.s
GNU assembler version 2.34 (x86_64-redhat-linux) using BFD version version
2.34-4.fc32
COMPILER_PATH=/opt/gcc-inst/usr/local/bin/../libexec/gcc/x86_64-pc-linux-gnu/11.0.0/:/opt/gcc-inst/usr/local/bin/../libexec/gcc/
LIBRARY_PATH=/opt/gcc-inst/usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/11.0.0/:/opt/gcc-inst/usr/local/bin/../lib/gcc/:/opt/gcc-inst/usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/11.0.0/../../../../lib64/:/lib/../lib64/:/usr/lib/../lib64/:/opt/gcc-inst/usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/11.0.0/../../../:/lib/:/usr/lib/
COLLECT_GCC_OPTIONS='-v' '-O3' '-Wall' '-Wextra' '-fno-strict-aliasing'
'-fwrapv' '-fno-aggressive-loop-optimizations' '-mtune=generic' '-march=x86-64'
'-dumpdir' 'a.'
/opt/gcc-inst/usr/local/bin/../libexec/gcc/x86_64-pc-linux-gnu/11.0.0/collect2
-plugin
/opt/gcc-inst/usr/local/bin/../libexec/gcc/x86_64-pc-linux-gnu/11.0.0/liblto_plugin.so
-plugin-opt=/opt/gcc-inst/usr/local/bin/../libexec/gcc/x86_64-pc-linux-gnu/11.0.0/lto-wrapper
-plugin-opt=-fresolution=/tmp/ccF6LwQe.res -plugin-opt=-pass-through=-lgcc
-plugin-opt=-pass-through=-lgcc_s -plugin-opt=-pass-through=-lc
-plugin-opt=-pass-through=-lgcc -plugin-opt=-pass-through=-lgcc_s
--eh-frame-hdr -m elf_x86_64 -dynamic-linker /lib64/ld-linux-x86-64.so.2
/lib/../lib64/crt1.o /lib/../lib64/crti.o
/opt/gcc-inst/usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/11.0.0/crtbegin.o
-L/opt/gcc-inst/usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/11.0.0
-L/opt/gcc-inst/usr/local/bin/../lib/gcc
-L/opt/gcc-inst/usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/11.0.0/../../../../lib64
-L/lib/../lib64 -L/usr/lib/../lib64
-L/opt/gcc-inst/usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/11.0.0/../../..
/tmp/ccMwZ4Yb.o -lgcc --push-state --as-needed -lgcc_s --pop-state -lc -lgcc
--push-state --as-needed -lgcc_s --pop-state
/opt/gcc-inst/usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/11.0.0/crtend.o
/lib/../lib64/crtn.o
COLLECT_GCC_OPTIONS='-v' '-O3' '-Wall' '-Wextra' '-fno-strict-aliasing'
'-fwrapv' '-fno-aggressive-loop-optimizations' '-mtune=generic' '-march=x86-64'
'-dumpdir' 'a.'
