Re: [PATCH] ARM: reset arm_fp16_format

2021-06-23 Thread Martin Liška
On 6/1/21 5:35 PM, Richard Earnshaw wrote: My initial reaction is 'that can't be right'. How would -mfp16-format=alternative ever work in this case? You are right. I'm going to revert the hunk done in g:ebd5e86c0f41dc1d692f9b2b68a510b1f6835a3e Martin >From 371c1992624c9269e2d5747561a8b27b30e4

RE: [PATCH] ARM: reset arm_fp16_format

2021-06-02 Thread Tamar Christina via Gcc-patches
ška > Sent: Tuesday, June 1, 2021 3:06 PM > To: gcc-patches@gcc.gnu.org > Cc: Christophe Lyon ; Tamar Christina > ; Kyrylo Tkachov > Subject: [PATCH] ARM: reset arm_fp16_format > > Hello. > > The patch fixes https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98636#c20 > wher

Re: [PATCH] ARM: reset arm_fp16_format

2021-06-01 Thread Richard Earnshaw via Gcc-patches
On 01/06/2021 15:05, Martin Liška wrote: Hello. The patch fixes https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98636#c20 where target option restore can be called and arm_fp16_format should be reset to ARM_FP16_FORMAT_NONE. It fixes the ICE in the PR. Can please ARM folks test me the patch

[PATCH] ARM: reset arm_fp16_format

2021-06-01 Thread Martin Liška
Hello. The patch fixes https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98636#c20 where target option restore can be called and arm_fp16_format should be reset to ARM_FP16_FORMAT_NONE. It fixes the ICE in the PR. Can please ARM folks test me the patch on a Arm machine? Thanks, Martin gcc/ChangeLog