This fixes a miscompile of glibc trunk on i?86 which happens because
the sched-deps treats a debug-insn with volatile ASM_OPERANDS
as a barrier.
Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
Ok for trunk and branches?
Thanks,
Richard.
2014-07-16 Richard Biener <rguent...@suse.de>
PR rtl-optimization/61801
* sched-deps.c (sched_analyze_2): For ASM_OPERANDS and
ASM_INPUT don't set reg_pending_barrier if it appears in a
debug-insn.
Index: gcc/sched-deps.c
===================================================================
--- gcc/sched-deps.c (revision 212580)
+++ gcc/sched-deps.c (working copy)
@@ -2750,7 +2750,8 @@ sched_analyze_2 (struct deps_desc *deps,
Consider for instance a volatile asm that changes the fpu rounding
mode. An insn should not be moved across this even if it only uses
pseudo-regs because it might give an incorrectly rounded result. */
- if (code != ASM_OPERANDS || MEM_VOLATILE_P (x))
+ if ((code != ASM_OPERANDS || MEM_VOLATILE_P (x))
+ && !DEBUG_INSN_P (insn))
reg_pending_barrier = TRUE_BARRIER;
/* For all ASM_OPERANDS, we must traverse the vector of input operands.
