[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread vries at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 Tom de Vries changed: What|Removed |Added Status|NEW |RESOLVED Resolution|---

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread vries at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #12 from Tom de Vries --- Author: vries Date: Thu Oct 26 20:09:24 2017 New Revision: 254120 URL: https://gcc.gnu.org/viewcvs?rev=254120&root=gcc&view=rev Log: Fix unsharing of GIMPLE_OMP_{SINGLE,TARGET,TEAMS} in gimple_copy 2017-10-

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread jakub at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #11 from Jakub Jelinek --- Comment on attachment 42477 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=42477 More minimal patch Ok, thanks. Not really happy this C++ification has been introduced, but now we have to live with th

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread vries at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 Tom de Vries changed: What|Removed |Added Attachment #42476|0 |1 is obsolete|

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread jakub at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #9 from Jakub Jelinek --- Comment on attachment 42476 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=42476 Updated tentative patch Both TEAMS and SINGLE don't have any inlines that actually need gomp_teams or gomp_single, they

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread vries at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 Tom de Vries changed: What|Removed |Added Attachment #42475|0 |1 is obsolete|

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread jakub at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #7 from Jakub Jelinek --- Ah, in that case it is hard to reproduce with single/teams, because usually those don't appear in try/finally cleanup actions. That said, gimple_copy is used in many other spots. But, if you want, just deal

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread vries at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #6 from Tom de Vries --- (In reply to Jakub Jelinek from comment #5) > If the problem is missing unsharing, Minimal example (with c compiler and -fexceptions): ... extern void bar (void); void foo (void) { int e[8]; #pragma acc de

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-26 Thread jakub at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #5 from Jakub Jelinek --- If the problem is missing unsharing, then I think many other constructs have the same problem and need their clauses to be unshared too. So, in addition to TARGET also SINGLE and TEAMS.

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-25 Thread vries at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #4 from Tom de Vries --- Created attachment 42475 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=42475&action=edit Tentative patch

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-25 Thread vries at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 Tom de Vries changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed|

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-25 Thread ro at CeBiTec dot Uni-Bielefeld.DE
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #2 from ro at CeBiTec dot Uni-Bielefeld.DE --- > --- Comment #1 from Tom de Vries --- [...] > I can't reproduce though at r253997. Can somebody post a configure line and > command line? I'm seeing it on i386-pc-solaris2.11 (no speci

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-25 Thread vries at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 --- Comment #1 from Tom de Vries --- Mentioned here ( https://gcc.gnu.org/ml/gcc-testresults/2017-10/msg01907.html ) for r253982, Intel64 -fpic -mcmodel=medium -mcmodel=large. Mentioned here ( https://gcc.gnu.org/ml/gcc-testresults/2017-10/msg01

[Bug tree-optimization/82707] ICE in verify_gimple_in_cfg at tree-cfg.c:5395

2017-10-24 Thread ro at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82707 Rainer Orth changed: What|Removed |Added Target Milestone|--- |8.0