[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 Segher Boessenkool changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED --- Comment #9 from Segher Boessenkool --- Fixed.
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 --- Comment #8 from Segher Boessenkool --- Author: segher Date: Mon Oct 29 07:36:45 2018 New Revision: 265582 URL: https://gcc.gnu.org/viewcvs?rev=265582=gcc=rev Log: combine: Fix various shortcomings in make_more_copies (PR87701, PR87780) This rewrites most of make_more_copies, in the process fixing a few PRs and some other bugs, and working around a few target problems. Certain notes turn out to actually change the meaning of the RTL, so we cannot drop them; and i386 takes subregs of hard regs. PR rtl-optimization/87701 PR rtl-optimization/87780 * combine.c (make_more_copies): Rewrite. Modified: trunk/gcc/ChangeLog trunk/gcc/combine.c
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701
--- Comment #7 from Segher Boessenkool ---
I have a patch.
rtl.texi has this to say about subregs of hard regs:
===
@item hard registers
It is seldom necessary to wrap hard registers in @code{subreg}s; such
registers would normally reduce to a single @code{reg} rtx. This use of
@code{subreg}s is discouraged and may not be supported in the future.
===
So why is the i386 backend doing this? It seems wrong.
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 Segher Boessenkool changed: What|Removed |Added Status|NEW |ASSIGNED Assignee|unassigned at gcc dot gnu.org |segher at gcc dot gnu.org --- Comment #6 from Segher Boessenkool --- We start off with (insn 36 35 9 2 (set (reg:V4SI 104) (subreg:V4SI (reg:TI 1 dx [ b ]) 0)) "87701.cc":13:31 -1 (expr_list:REG_DEAD (reg:DI 2 cx) (expr_list:REG_DEAD (reg:DI 1 dx) (nil and replace that with (insn 41 35 42 2 (set (reg:V4SI 107) (reg:TI 1 dx [ b ])) "87701.cc":13:31 -1 (expr_list:REG_DEAD (reg:TI 1 dx [ b ]) (nil))) 42: r104:V4SI=r107:V4SI REG_DEAD r107:V4SI (no long form dump for that last one, sorry). Note the subreg has gone missing. So which is correct here: with or without subreg?
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 --- Comment #5 from Segher Boessenkool --- I see the problem. Mine.
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 Segher Boessenkool changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed|2018-10-23 00:00:00 |2018-10-27 Ever confirmed|0 |1 --- Comment #4 from Segher Boessenkool --- Confirmed.
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 David Binderman changed: What|Removed |Added CC||dcb314 at hotmail dot com --- Comment #3 from David Binderman --- Code compiles ok at -O2, so -O3 required. Bug seems to first appear somewhere between revision 265367 and 265411.
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 --- Comment #2 from Martin Liška --- (In reply to Richard Biener from comment #1) > x86? Yes! I forgot to mention that.
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 Richard Biener changed: What|Removed |Added Target||x86_64-*-* --- Comment #1 from Richard Biener --- x86?
[Bug rtl-optimization/87701] [9 Regression] ICE in elimination_costs_in_insn, at reload1.c:3640 since r265398
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87701 Martin Liška changed: What|Removed |Added Last reconfirmed||2018-10-23 Known to work||8.2.0 Target Milestone|--- |9.0 Known to fail||9.0
