[Bug target/70048] [6 Regression][AArch64] Inefficient local array addressing

wdijkstr at arm dot com gcc-bugzilla@gcc.gnu.org
Fri Mar 11 13:57:00 GMT 2016


--- Comment #17 from Wilco <wdijkstr at arm dot com> ---
(In reply to Jiong Wang from comment #16)

>   * for the second patch at #c10, if we always do the following no matter
>     op0 is virtual & eliminable or not
>      "op1 = force_operand (op1, NULL_RTX);"
>     then there is no performance regression as well. Even more, there is
>     significant perf improvement beyond current trunk on one benchmark.

I ran this modified patch through a few benchmarks and there are no
regressions. The codesize of SPEC2006 reduces significantly in several cases
(-0.25%), while there are only 2 minor increases of +0.02%, so the patch
reduces instruction counts quite effectively.

So we should go with the modified patch for GCC6.

More information about the Gcc-bugs mailing list