[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-20 Thread ubizjak at gmail dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 --- Comment #23 from Uroš Bizjak --- (In reply to Jakub Jelinek from comment #22) > IMHO that is way too expensive. Agreed. Anyway, it looks you got this problem under control ;)

[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-20 Thread jakub at gcc dot gnu.org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 --- Comment #22 from Jakub Jelinek --- (In reply to Uroš Bizjak from comment #21) > --- config/i386/i386.c (revision 207935) > +++ config/i386/i386.c (working copy) > @@ -40666,9 +40666,9 @@ ix86_vectorize_vec_perm_const_ok (enum machine_mod >

[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-20 Thread ubizjak at gmail dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 --- Comment #21 from Uroš Bizjak --- --cut here-- Index: config/i386/i386.c === --- config/i386/i386.c (revision 207935) +++ config/i386/i386.c (working copy) @@ -40666,9 +40666,9 @

[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-20 Thread jakub at gcc dot gnu.org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 --- Comment #20 from Jakub Jelinek --- So, from quick skim of relevant routines that do gen_reg_rtx, I see: gen_reg_rtx called only after if (d->testing_p) return true;, thus ok: expand_vec_perm_blend expand_vec_perm_pshufb expand_vec_perm_palign

[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-20 Thread jakub at gcc dot gnu.org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 Jakub Jelinek changed: What|Removed |Added Component|middle-end |target --- Comment #19 from Jakub Jelinek

[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-19 Thread ubizjak at gmail dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 --- Comment #11 from Uroš Bizjak --- (In reply to Uroš Bizjak from comment #10) > (In reply to Vittorio Zecca from comment #5) > > > Adding option -m32 I get ICE in ix86_expand_prologue, at > > config/i386/i386.c:10559 > > I can confirm this wit

[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-19 Thread ubizjak at gmail dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 --- Comment #10 from Uroš Bizjak --- (In reply to Vittorio Zecca from comment #5) > Adding option -m32 I get ICE in ix86_expand_prologue, at > config/i386/i386.c:10559 I can confirm this with: gcc version 4.8.3 20140219 (prerelease) [gcc-4_8-br

[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-19 Thread ubizjak at gmail dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 --- Comment #9 from Uroš Bizjak --- (In reply to Vittorio Zecca from comment #6) > As an aside, in gcc 4.8.1 source code, before line 6995 of gcc/expr.c I put > > printf("\nexpr.c:6995 value->code=%d NUM_RTX_CODE=%d\n",(int) > value->code,NUM_R

[Bug target/57896] [4.8 Regression] ICE in expand_expr_real_2

2014-02-19 Thread mpolacek at gcc dot gnu.org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57896 Marek Polacek changed: What|Removed |Added Component|c |target --- Comment #8 from Marek Polacek