[Bug rtl-optimization/104961] [9/10/11/12 Regression] compilation never (?) finishes at -Og

["mpolacek at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104961

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2022-03-16
           Priority|P3                          |P2
                 CC|                            |mpolacek at gcc dot gnu.org
   Target Milestone|---                         |9.5
            Summary|compilation never (?)       |[9/10/11/12 Regression]
                   |finishes at -Og             |compilation never (?)
                   |                            |finishes at -Og
             Status|UNCONFIRMED                 |NEW

--- Comment #1 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Confirmed.  Started with

commit 7293e3f58edb7535d826c85160b9fa6d4132ffea
Author: Vladimir Makarov <vmakarov@redhat.com>
Date:   Fri Mar 16 18:48:26 2018 +0000

    re PR target/84876 (ICE on invalid code in lra_assign at
gcc/lra-assigns.c:1601 since r258504)

    2018-03-16  Vladimir Makarov  <vmakarov@redhat.com>

            PR target/84876
            * lra-assigns.c (lra_split_hard_reg_for): Don't use
            regno_allocno_class_array and sorted_pseudos.
            * lra-constraints.c (spill_hard_reg_in_range): Ignore hard regs in
            insns where regno is used.

    2018-03-16  Vladimir Makarov  <vmakarov@redhat.com>

            PR target/84876
            * gcc.target/i386/pr84876.c: New test.

    From-SVN: r258602

which fixed an ICE:
/home/mpolacek/q.c: In function ‘foo’:
/home/mpolacek/q.c:12:1: error: unable to find a register to spill
 }
 ^
/home/mpolacek/q.c:12:1: error: this is the insn:
(insn 15 32 33 2 (parallel [
            (set (reg/v:TI 105 [orig:91 j ] [91])
                (mult:TI (zero_extend:TI (subreg/j:DI (reg/v:TI 105 [orig:91 j
] [91]) 0))
                    (zero_extend:TI (reg:DI 102 [ i.0_1 ]))))
            (clobber (reg:CC 17 flags))
        ]) "/home/mpolacek/q.c":8 356 {*umulditi3_1}
     (expr_list:REG_DEAD (reg:DI 102 [ i.0_1 ])
        (expr_list:REG_UNUSED (reg:CC 17 flags)
            (nil))))
during RTL pass: reload
/home/mpolacek/q.c:12:1: internal compiler error: in lra_split_hard_reg_for, at
lra-assigns.c:1805
unrecognized DWARF version in .debug_info at 6
0xf1752d _fatal_insn(char const*, rtx_def const*, char const*, int, char
const*)
        ../../gcc/rtl-error.c:108
0xd773ae lra_split_hard_reg_for()
        ../../gcc/lra-assigns.c:1805
0xd6fcb8 lra(_IO_FILE*)
        ../../gcc/lra.c:2507
0xd17143 do_reload
        ../../gcc/ira.c:5465
0xd17636 execute
        ../../gcc/ira.c:5649

Compiled fine with GCC 5 -> regression.