[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 Jakub Jelinek changed: What|Removed |Added Status|UNCONFIRMED |ASSIGNED Last reconfirmed|

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 --- Comment #2 from Uros Bizjak 2012-02-29 20:32:06 UTC --- This patch adds missing alternative, and also disparages alternatives that end with excess register->mem moves. Index: sse.md ===

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 --- Comment #3 from Jakub Jelinek 2012-02-29 20:37:49 UTC --- I've tried: @@ -3899,7 +3899,7 @@ (vec_merge:VI4F_128 (vec_duplicate:VI4F_128 (match_operand: 2 "general_operand" - " x,m,*r,m,x,x,*rm,*rm,x,fF,*r")) +

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 Jakub Jelinek changed: What|Removed |Added CC||uros at gcc dot gnu.org --- Comment #4 fr

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 --- Comment #5 from Uros Bizjak 2012-02-29 21:00:26 UTC --- (In reply to comment #4) > movsi uses re constraint, not rn, so I'd think we should use re. re also includes symbols and labels and whatnot (please see x86_64_immediate_operand predicat

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 --- Comment #6 from Uros Bizjak 2012-02-29 21:14:34 UTC --- (In reply to comment #5) > (In reply to comment #4) > > movsi uses re constraint, not rn, so I'd think we should use re. On a second thought, re won't hurt and it will also handle symbo

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 --- Comment #7 from Jakub Jelinek 2012-02-29 21:17:05 UTC --- Created attachment 26791 --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=26791 gcc47-pr52437.patch So does this look ok?

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 --- Comment #8 from Uros Bizjak 2012-02-29 21:19:19 UTC --- (In reply to comment #7) > Created attachment 26791 [details] > gcc47-pr52437.patch > > So does this look ok? Yes, perfect!

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 --- Comment #9 from Jakub Jelinek 2012-02-29 23:45:14 UTC --- Author: jakub Date: Wed Feb 29 23:45:08 2012 New Revision: 184676 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=184676 Log: PR target/52437 * config/i386/sse.md (vec_s

[Bug target/52437] [4.7 Regression] internal compiler error: in spill_failure, at reload1.c:2120

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52437 Jakub Jelinek changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|