This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug rtl-optimization/61559] FAIL: gcc.dg/builtin-bswap-8.c on i686 with -mmovbe
- From: "ubizjak at gmail dot com" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Thu, 19 Jun 2014 09:22:22 +0000
- Subject: [Bug rtl-optimization/61559] FAIL: gcc.dg/builtin-bswap-8.c on i686 with -mmovbe
- Auto-submitted: auto-generated
- References: <bug-61559-4 at http dot gcc dot gnu dot org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61559
--- Comment #2 from UroÅ Bizjak <ubizjak at gmail dot com> ---
On a related note, -march=i386 generates long sequences of rol insns, e.g.:
foo5:
movl 4(%esp), %eax
rolw $8, %ax
roll $16, %eax
rolw $8, %ax
movl 8(%esp), %edx
rolw $8, %dx
roll $16, %edx
rolw $8, %dx
andl %edx, %eax
rolw $8, %ax
roll $16, %eax
rolw $8, %ax
ret
Please note that this arch doesn't have bswap insn, but this insn isn't
generated for optimized code anyway.