This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug bootstrap/59985] stage2/3 compare error on lto-streamer-in.o
- From: "jakub at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Thu, 30 Jan 2014 15:55:16 +0000
- Subject: [Bug bootstrap/59985] stage2/3 compare error on lto-streamer-in.o
- Auto-submitted: auto-generated
- References: <bug-59985-4 at http dot gcc dot gnu dot org/bugzilla/>
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59985
Jakub Jelinek <jakub at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |vmakarov at gcc dot gnu.org
--- Comment #9 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Seems it is LRA where things change. First of all, with -g there is an extra
- 151: r0:SI=r151:SI
+ Split reuse change 151->260:
+ 77: debug b => r260:SI
+ 10150: r0:SI=r151:SI
message in -fdump-rtl-reload-details dump (obviously, insn numbers are
different).
But, even when comparing say -g --param min-nondebug-insn-uid=100 dump with
-g --param min-nondebug-insn-uid=1000 dump I see more important change:
- Choosing alt 4 in insn 336: (0) &r (1) 0 (2) De {*anddi3_insn}
- Creating newreg=262 from oldreg=249, assigning class GENERAL_REGS to
r262
- 336: r262:DI=r262:DI&0x1
+ Choosing alt 5 in insn 10224: (0) &r (1) r (2) De {*anddi3_insn}
+ Creating newreg=262 from oldreg=231, assigning class GENERAL_REGS to
r262
+ 10224: r249:DI=r262:DI&0x1