Re: [PATCH, GCC/LRA] Teach LRA to not use same register value for multiple output operands of an insn

2016-09-20 Thread Thomas Preudhomme
Hi Vladimir, I'm sorry I failed to give a proper amount of context in that email. You already approved that patch for trunk one or two months ago and I was just replying to your approval to ask if you are now ok to commit it to gcc-6-branch. I've committed it now then since it has been on tru

Re: [PATCH, GCC/LRA] Teach LRA to not use same register value for multiple output operands of an insn

2016-09-19 Thread Vladimir N Makarov
On 07/08/2016 11:07 AM, Thomas Preudhomme wrote: Hi, While investigating the root cause a testsuite regression for the ARM/embedded-5-branch GCC in gcc.dg/vect/slp-perm-5.c, we found that the bug seems to also affect trunk. The bug manifests itself as an ICE in cselib due to a parallel insn wi

Re: [PATCH, GCC/LRA] Teach LRA to not use same register value for multiple output operands of an insn

2016-09-15 Thread Thomas Preudhomme
Hi Vladimir & release managers, The patch applies cleanly to gcc-6-branch. Ok to backport? Best regards, Thomas On 14/07/16 17:25, Vladimir Makarov wrote: On 07/08/2016 11:07 AM, Thomas Preudhomme wrote: Hi, While investigating the root cause a testsuite regression for the ARM/embedded-5-br

Re: [PATCH, GCC/LRA] Teach LRA to not use same register value for multiple output operands of an insn

2016-07-14 Thread Vladimir Makarov
On 07/08/2016 11:07 AM, Thomas Preudhomme wrote: Hi, While investigating the root cause a testsuite regression for the ARM/embedded-5-branch GCC in gcc.dg/vect/slp-perm-5.c, we found that the bug seems to also affect trunk. The bug manifests itself as an ICE in cselib due to a parallel insn with

Re: [PATCH, GCC/LRA] Teach LRA to not use same register value for multiple output operands of an insn

2016-07-14 Thread Thomas Preudhomme
On Thursday 14 July 2016 10:32:48 Thomas Preudhomme wrote: > On Friday 08 July 2016 09:05:55 Mike Stump wrote: > > On Jul 8, 2016, at 8:07 AM, Thomas Preudhomme > > wrote: > > > While investigating the root cause a testsuite regression for the > > > ARM/embedded-5-branch GCC in gcc.dg/vect/slp-pe

Re: [PATCH, GCC/LRA] Teach LRA to not use same register value for multiple output operands of an insn

2016-07-14 Thread Thomas Preudhomme
On Friday 08 July 2016 09:05:55 Mike Stump wrote: > On Jul 8, 2016, at 8:07 AM, Thomas Preudhomme wrote: > > While investigating the root cause a testsuite regression for the > > ARM/embedded-5-branch GCC in gcc.dg/vect/slp-perm-5.c, we found that the > > bug seems to also affect trunk. > > Hum.

Re: [PATCH, GCC/LRA] Teach LRA to not use same register value for multiple output operands of an insn

2016-07-08 Thread Mike Stump
On Jul 8, 2016, at 8:07 AM, Thomas Preudhomme wrote: > While investigating the root cause a testsuite regression for the > ARM/embedded-5-branch GCC in gcc.dg/vect/slp-perm-5.c, we found that the bug > seems to also affect trunk. Hum... If in 6.x, and safe to back port to 6, a back port would