This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug debug/66691] [5/6 Regression] ICE on valid code at -O3 with -g enabled in simplify_subreg, at simplify-rtx.c:5744
- From: "ubizjak at gmail dot com" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Sun, 28 Jun 2015 08:13:14 +0000
- Subject: [Bug debug/66691] [5/6 Regression] ICE on valid code at -O3 with -g enabled in simplify_subreg, at simplify-rtx.c:5744
- Auto-submitted: auto-generated
- References: <bug-66691-4 at http dot gcc dot gnu dot org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66691
UroÅ Bizjak <ubizjak at gmail dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |vmakarov at gcc dot gnu.org
--- Comment #3 from UroÅ Bizjak <ubizjak at gmail dot com> ---
In _.ira dump, we have:
(insn 146 144 147 21 (set (reg:HI 96 [ D.1574 ])
(const_int 0 [0])) pr66691.c:42 88 {*movhi_internal}
(expr_list:REG_EQUAL (const_int 0 [0])
(nil)))
[...]
(debug_insn 148 147 30 21 (var_location:QI r (subreg:QI (reg:HI 96 [ D.1574 ])
0)) pr66691.c:42 -1
(nil))
And in _.reload, we get invalid RTX when (const_int 0) propagates to
(debug_insn 148):
(note 146 144 429 21 NOTE_INSN_DELETED)
[...]
(debug_insn 148 147 30 21 (var_location:QI r (subreg:QI (const_int 0 [0]) 0))
pr66691.c:42 -1
(nil))
Looks like LRA problem, adding CC.