http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53844
Bug #: 53844
Summary: GCC generates suboptimal code for unused members of
classes in some cases on multiple targets.
Classification: Unclassified
Product: gcc
Version: unknown
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: middle-end
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: e...@edrosten.com
Created attachment 27737
--> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27737
Source file for which suboptimal code is generated (self contained)
The attached source file contains a very minimal cut out part of a numerics
library with expression templates which was exhibiting poor performance. The
poor performance is due to the compiler emitting instructions to push variables
onto the stack which are ultimately never used.
The attached file seems to be the smallest which replicates the behaviour.
Compiled with:
g++-4.7 -S minimal.cc -O3
The output generated is:
_Z4testRK6VectorI5VBaseERS1_i:
.LFB8:
.cfi_startproc
subq $160, %rsp
.cfi_def_cfa_offset 168
movq (%rsi), %rcx
movq (%rdi), %rdx
leaq -120(%rsp), %rax
movl $1, 8(%rsp)
movl $1, -8(%rsp)
movl $1, -24(%rsp)
movl $1, -40(%rsp)
movq %rax, 24(%rsp)
leaq -104(%rsp), %rax
movl $1, -56(%rsp)
movl $1, -72(%rsp)
movl $1, -88(%rsp)
movq %rax, 40(%rsp)
leaq 24(%rsp), %rax
movl $1, -104(%rsp)
movl $1, -120(%rsp)
movq %rdi, 32(%rsp)
movq %rax, 48(%rsp)
leaq -88(%rsp), %rax
movq %rax, 56(%rsp)
leaq 40(%rsp), %rax
movq %rax, 64(%rsp)
leaq -72(%rsp), %rax
movq %rax, 72(%rsp)
leaq 56(%rsp), %rax
movq %rax, 80(%rsp)
leaq -56(%rsp), %rax
movq %rax, 88(%rsp)
leaq 72(%rsp), %rax
movq %rax, 96(%rsp)
leaq -40(%rsp), %rax
movq %rax, 104(%rsp)
leaq 88(%rsp), %rax
movq %rax, 112(%rsp)
leaq -24(%rsp), %rax
movq %rax, 120(%rsp)
leaq 104(%rsp), %rax
movq %rax, 128(%rsp)
leaq -8(%rsp), %rax
movq %rax, 136(%rsp)
leaq 120(%rsp), %rax
movq %rax, 144(%rsp)
xorl %eax, %eax
.p2align 4,,10
.p2align 3
.L2:
movsd (%rdx,%rax), %xmm0
movsd %xmm0, (%rcx,%rax)
addq $8, %rax
cmpq $800, %rax
jne .L2
addq $160, %rsp
.cfi_def_cfa_offset 8
ret
.cfi_endproc
This majority of the instructions correspond to pushing the numbers (line 68 of
the source file) and corresponding reference (line 39 of the source file) onto
the stack.
The behaviour is quite changable. For example replacing the assignment with
const auto&& v=in*1*1*1*1*1*1*1*1*1*1*1;
out=v;
and enabling C++11 produces much worse results.
The main compiler tested was:
COLLECT_GCC=g++-4.7
COLLECT_LTO_WRAPPER=/usr/local/libexec/gcc/x86_64-unknown-linux-gnu/4.7.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../gcc-4.7.0/configure --program-suffix=-4.7
--enable-languages=c,c++
Thread model: posix
gcc version 4.7.0 (GCC)
The source has been tried with a number of other compilers in 32 and 64 bit
mode where applicable with similar results. The compilers are:
COLLECT_GCC=arm-linux-gnueabi-g++-4.6
COLLECT_LTO_WRAPPER=/usr/lib/gcc/arm-linux-gnueabi/4.6.1/lto-wrapper
Target: arm-linux-gnueabi
Configured with: ../src/configure -v --with-pkgversion='Ubuntu/Linaro
4.6.1-9ubuntu3' --with-bugurl=file:///usr/share/doc/gcc-4.6/README.Bugs
--enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr
--program-suffix=-4.6 --enable-shared --enable-linker-build-id
--with-system-zlib --libexecdir=/usr/lib --without-included-gettext
--enable-threads=posix
--with-gxx-include-dir=/usr/arm-linux-gnueabi/include/c++/4.6.1
--libdir=/usr/lib --enable-nls --enable-clocale=gnu --enable-libstdcxx-debug
--enable-libstdcxx-time=yes --enable-plugin --enable-objc-gc --enable-multilib
--disable-sjlj-exceptions --with-arch=armv7-a --with-float=softfp
--with-fpu=vfpv3-d16 --with-mode=thumb --disable-werror
--enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu
--target=arm-linux-gnueabi --program-prefix=arm-linux-gnueabi-
--includedir=/usr/arm-linux-gnueabi/include
--with-headers=/usr/arm-linux-gnueabi/include
--with-libs=/usr/arm-linux-gnueabi/lib
Thread model: posix
gcc version 4.6.1 (Ubuntu/Linaro 4.6.1-9ubuntu3)
Using built-in specs.
COLLECT_GCC=/usr/bin/g++-4.6.real
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-linux-gnu/4.6.1/lto-wrapper
Target: x86_64-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Ubuntu/Linaro
4.6.1-9ubuntu3' --with-bugurl=file:///usr/share/doc/gcc-4.6/README.Bugs
--enable-languages=c,c++,fortran,objc,obj-c++,go --prefix=/usr
--program-suffix=-4.6 --enable-shared --enable-linker-build-id
--with-system-zlib --libexecdir=/usr/lib --without-included-gettext
--enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.6
--libdir=/usr/lib --enable-nls --with-sysroot=/ --enable-clocale=gnu
--enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-plugin
--enable-objc-gc --disable-werror --with-arch-32=i686 --with-tune=generic
--enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu
--target=x86_64-linux-gnu
Thread model: posix
gcc version 4.6.1 (Ubuntu/Linaro 4.6.1-9ubuntu3)
Using built-in specs.
COLLECT_GCC=/usr/bin/g++-4.5.real
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-linux-gnu/4.5.4/lto-wrapper
Target: x86_64-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Ubuntu/Linaro
4.5.3-9ubuntu1' --with-bugurl=file:///usr/share/doc/gcc-4.5/README.Bugs
--enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr
--program-suffix=-4.5 --enable-shared --enable-linker-build-id
--with-system-zlib --libexecdir=/usr/lib --without-included-gettext
--enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.5
--libdir=/usr/lib --enable-nls --with-sysroot=/ --enable-clocale=gnu
--enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-plugin
--enable-gold --enable-ld=default --with-plugin-ld=ld.gold --enable-objc-gc
--disable-werror --with-arch-32=i686 --with-tune=generic
--enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu
--target=x86_64-linux-gnu
Thread model: posix
gcc version 4.5.4 (Ubuntu/Linaro 4.5.3-9ubuntu1)
Using built-in specs.
Target: x86_64-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Ubuntu/Linaro
4.4.6-11ubuntu2' --with-bugurl=file:///usr/share/doc/gcc-4.4/README.Bugs
--enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr
--program-suffix=-4.4 --enable-shared --enable-linker-build-id
--with-system-zlib --libexecdir=/usr/lib --without-included-gettext
--enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.4
--libdir=/usr/lib --enable-nls --with-sysroot=/ --enable-clocale=gnu
--enable-libstdcxx-debug --enable-objc-gc --disable-werror --with-arch-32=i686
--with-tune=generic --enable-checking=release --build=x86_64-linux-gnu
--host=x86_64-linux-gnu --target=x86_64-linux-gnu
Thread model: posix
gcc version 4.4.6 (Ubuntu/Linaro 4.4.6-11ubuntu2)
