On Thu, 14 Jul 2016, Georg-Johann Lay wrote:
> On 12.07.2016 11:38, Richard Biener wrote:
> >
> > The following patch does $subject which is requested in a comment in
> > PR50417 as this restriction defeats the purpose of having memcpy
> > as a portable and efficient way to circumvent strict-alia
On 12.07.2016 11:38, Richard Biener wrote:
The following patch does $subject which is requested in a comment in
PR50417 as this restriction defeats the purpose of having memcpy
as a portable and efficient way to circumvent strict-aliasing violations
(or even as a portable and efficient way to do
On Thu, 14 Jul 2016, Eric Botcazou wrote:
> > Boostrap / regtest on strict-alignment platforms welcome.
>
> Bootstrap/regtest on SPARC/Solaris is OK, modulo:
>
> XPASS: gcc.dg/memmove-4.c scan-tree-dump-not optimized "memmove"
> FAIL: gcc.dg/strlenopt-8.c scan-tree-dump-times strlen "strlen
> Boostrap / regtest on strict-alignment platforms welcome.
Bootstrap/regtest on SPARC/Solaris is OK, modulo:
XPASS: gcc.dg/memmove-4.c scan-tree-dump-not optimized "memmove"
FAIL: gcc.dg/strlenopt-8.c scan-tree-dump-times strlen "strlen (" 0
FAIL: gcc.dg/strlenopt-8.c scan-tree-dump-times st
> I have added a testcase that should exercise most relevant cases so
> we can look for fallout on "interesting" targets.
It passes on SPARC/Solaris 32-bit.
> Boostrap / regtest on strict-alignment platforms welcome.
Preliminary testing looks good, with a few expected regressions I guess:
+XPAS
The following patch does $subject which is requested in a comment in
PR50417 as this restriction defeats the purpose of having memcpy
as a portable and efficient way to circumvent strict-aliasing violations
(or even as a portable and efficient way to do unaligned loads).
Bootstrap / regtest runni
