public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug rtl-optimization/41646]  New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register
@ 2009-10-09 14:40 jakub at gcc dot gnu dot org
  2009-10-09 14:42 ` [Bug rtl-optimization/41646] " jakub at gcc dot gnu dot org
                   ` (7 more replies)
  0 siblings, 8 replies; 9+ messages in thread
From: jakub at gcc dot gnu dot org @ 2009-10-09 14:40 UTC (permalink / raw)
  To: gcc-bugs

struct A { unsigned long a; };
struct B { unsigned short b, c, d; };
struct B bar (unsigned long);
char *
foo (char *a, struct A *x)
{
  struct B b = bar (x->a);
  unsigned char c;
  unsigned short d;
  a[3] = ((unsigned char) (b.b % 10) + 48);
  d = b.b / 10;
  a[2] = ((unsigned char) (d % 10) + 48);
  d = d / 10;
  a[1] = ((unsigned char) (d % 10) + 48);
  a[0] = ((unsigned char) ((d / 10) % 10) + 48);
  a[4] = 46;
  c = (unsigned char) b.c;
  a[6] = (c % 10 + 48);
  a[5] = ((c / 10) % 10 + 48);
  a[7] = 46;
  c = b.d;
  a[9] = (c % 10 + 48);
  a[8] = ((c / 10) % 10 + 48);
  return a + 10;
}

ICEs with -O2 on x86_64-linux:
/tmp/tt.c: In function 'foo':
/tmp/tt.c:25:1: error: unable to find a register to spill in class 'AREG'
/tmp/tt.c:25:1: error: this is the insn:
(insn 82 81 83 2 /tmp/tt.c:18 (parallel [
            (set (reg:HI 1 dx [170])
                (mult:HI (zero_extend:HI (reg/v:QI 4 si [orig:82 c ] [82]))
                    (zero_extend:HI (reg:QI 2 cx [169]))))
            (clobber (reg:CC 17 flags))
        ]) 337 {*umulqihi3_1} (expr_list:REG_UNUSED (reg:CC 17 flags)
        (expr_list:REG_EQUAL (mult:HI (zero_extend:HI (reg/v:QI 4 si [orig:82 c
] [82]))
                (const_int 205 [0xcd]))
            (nil))))
/tmp/tt.c:25:1: internal compiler error: in spill_failure, at reload1.c:2151
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

Worked in 3.4/4.0, ICEs from 4.1 onwards up to current trunk.

The problem seems to be in the combiner, it extremely lengthens the lifetime of
hard register (%rax):

in *.dce:
...
(insn 13 12 14 2 /tmp/tt.c:7 (parallel [
            (set (reg:DI 113)
                (lshiftrt:DI (reg:DI 0 ax)
                    (const_int 32 [0x20])))
            (clobber (reg:CC 17 flags))
        ]) 549 {*lshrdi3_1_rex64} (expr_list:REG_DEAD (reg:DI 0 ax)
        (expr_list:REG_UNUSED (reg:CC 17 flags)
            (nil))))
(insn 14 13 16 2 /tmp/tt.c:7 (set (reg:HI 114)
        (subreg:HI (reg:DI 113) 0)) 53 {*movhi_1} (expr_list:REG_DEAD (reg:DI
113)
        (nil)))
... 49 insns ...
(insn 82 81 83 2 /tmp/tt.c:18 (parallel [
            (set (reg:HI 168)
                (mult:HI (zero_extend:HI (reg/v:QI 82 [ c ]))
                    (zero_extend:HI (reg:QI 169))))
            (clobber (reg:CC 17 flags))
        ]) 337 {*umulqihi3_1} (expr_list:REG_UNUSED (reg:CC 17 flags)
        (expr_list:REG_EQUAL (mult:HI (zero_extend:HI (reg/v:QI 82 [ c ]))
                (const_int 205 [0xcd]))
            (nil))))
... 18 insns ...
(insn 108 107 110 2 /tmp/tt.c:21 (set (reg/v:QI 93 [ c ])
        (subreg:QI (reg:HI 114) 0)) 62 {*movqi_1} (expr_list:REG_DEAD (reg:HI
114)
        (nil)))
...
in *.combine:
...
(note 13 12 14 2 NOTE_INSN_DELETED)
(note 14 13 16 2 NOTE_INSN_DELETED)
... many insns ...
(insn 82 81 83 2 /tmp/tt.c:18 (parallel [
            (set (reg:HI 168)
                (mult:HI (zero_extend:HI (reg/v:QI 82 [ c ]))
                    (zero_extend:HI (reg:QI 169))))
            (clobber (reg:CC 17 flags))
        ]) 337 {*umulqihi3_1} (expr_list:REG_UNUSED (reg:CC 17 flags)
        (expr_list:REG_EQUAL (mult:HI (zero_extend:HI (reg/v:QI 82 [ c ]))
                (const_int 205 [0xcd]))
            (nil))))
... many insns ...
(insn 108 107 110 2 /tmp/tt.c:21 (parallel [
            (set (subreg:DI (reg/v:QI 93 [ c ]) 0)
                (lshiftrt:DI (reg:DI 0 ax)
                    (const_int 32 [0x20])))
            (clobber (reg:CC 17 flags))
        ]) 549 {*lshrdi3_1_rex64} (expr_list:REG_UNUSED (reg:CC 17 flags)
        (expr_list:REG_DEAD (reg:DI 0 ax)
            (nil))))
...


-- 
           Summary: [4.3/4.4/4.5 Regression] Reload ICE due to combiner
                    extending life time of a hard register
           Product: gcc
           Version: 4.4.1
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: rtl-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: jakub at gcc dot gnu dot org
GCC target triplet: x86_64-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/41646] [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register
  2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
@ 2009-10-09 14:42 ` jakub at gcc dot gnu dot org
  2009-10-09 15:35 ` jakub at gcc dot gnu dot org
                   ` (6 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: jakub at gcc dot gnu dot org @ 2009-10-09 14:42 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #1 from jakub at gcc dot gnu dot org  2009-10-09 14:42 -------
Or perhaps the bug is that we should force the return value into a register to
avoid having hard regs in the arithmetic instructions so early.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/41646] [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register
  2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
  2009-10-09 14:42 ` [Bug rtl-optimization/41646] " jakub at gcc dot gnu dot org
@ 2009-10-09 15:35 ` jakub at gcc dot gnu dot org
  2009-10-09 19:02 ` jakub at gcc dot gnu dot org
                   ` (5 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: jakub at gcc dot gnu dot org @ 2009-10-09 15:35 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #2 from jakub at gcc dot gnu dot org  2009-10-09 15:35 -------
Created an attachment (id=18767)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18767&action=view)
gcc45-pr41646.patch

Patch I'm going to bootstrap/regtest.

cant_combine_insn_p already takes care of not combining in hard registers with
likely spilled classes in the (set (reg) (hardreg)) cases, so it is just when
expand doesn't honor this.


-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |jakub at gcc dot gnu dot org
                   |dot org                     |
             Status|UNCONFIRMED                 |ASSIGNED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/41646] [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register
  2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
  2009-10-09 14:42 ` [Bug rtl-optimization/41646] " jakub at gcc dot gnu dot org
  2009-10-09 15:35 ` jakub at gcc dot gnu dot org
@ 2009-10-09 19:02 ` jakub at gcc dot gnu dot org
  2009-10-12  8:48 ` jakub at gcc dot gnu dot org
                   ` (4 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: jakub at gcc dot gnu dot org @ 2009-10-09 19:02 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #3 from jakub at gcc dot gnu dot org  2009-10-09 19:02 -------
Subject: Bug 41646

Author: jakub
Date: Fri Oct  9 19:01:53 2009
New Revision: 152597

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=152597
Log:
        PR rtl-optimization/41646
        * calls.c (expand_call): For BLKmode types returned in registers
        avoid likely spilled hard regs in copy_blkmode_from_reg generated
        insns.

        * gcc.c-torture/compile/pr41646.c: New test.

Added:
    trunk/gcc/testsuite/gcc.c-torture/compile/pr41646.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/calls.c
    trunk/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/41646] [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register
  2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
                   ` (2 preceding siblings ...)
  2009-10-09 19:02 ` jakub at gcc dot gnu dot org
@ 2009-10-12  8:48 ` jakub at gcc dot gnu dot org
  2009-10-12 13:38 ` jakub at gcc dot gnu dot org
                   ` (3 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: jakub at gcc dot gnu dot org @ 2009-10-12  8:48 UTC (permalink / raw)
  To: gcc-bugs



-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.3.5


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/41646] [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register
  2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
                   ` (3 preceding siblings ...)
  2009-10-12  8:48 ` jakub at gcc dot gnu dot org
@ 2009-10-12 13:38 ` jakub at gcc dot gnu dot org
  2009-10-12 13:54 ` [Bug rtl-optimization/41646] [4.3 " jakub at gcc dot gnu dot org
                   ` (2 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: jakub at gcc dot gnu dot org @ 2009-10-12 13:38 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #4 from jakub at gcc dot gnu dot org  2009-10-12 13:38 -------
Subject: Bug 41646

Author: jakub
Date: Mon Oct 12 13:38:29 2009
New Revision: 152666

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=152666
Log:
        PR rtl-optimization/41646
        * calls.c (expand_call): For BLKmode types returned in registers
        avoid likely spilled hard regs in copy_blkmode_from_reg generated
        insns.

        * gcc.c-torture/compile/pr41646.c: New test.

Added:
    branches/gcc-4_4-branch/gcc/testsuite/gcc.c-torture/compile/pr41646.c
Modified:
    branches/gcc-4_4-branch/gcc/ChangeLog
    branches/gcc-4_4-branch/gcc/calls.c
    branches/gcc-4_4-branch/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/41646] [4.3 Regression] Reload ICE due to combiner extending life time of a hard register
  2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
                   ` (4 preceding siblings ...)
  2009-10-12 13:38 ` jakub at gcc dot gnu dot org
@ 2009-10-12 13:54 ` jakub at gcc dot gnu dot org
  2010-01-02 16:29 ` rguenth at gcc dot gnu dot org
  2010-05-22 18:43 ` rguenth at gcc dot gnu dot org
  7 siblings, 0 replies; 9+ messages in thread
From: jakub at gcc dot gnu dot org @ 2009-10-12 13:54 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #5 from jakub at gcc dot gnu dot org  2009-10-12 13:54 -------
Fixed in 4.4/4.5 for now.


-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|jakub at gcc dot gnu dot org|unassigned at gcc dot gnu
                   |                            |dot org
             Status|ASSIGNED                    |UNCONFIRMED
            Summary|[4.3/4.4/4.5 Regression]    |[4.3 Regression] Reload ICE
                   |Reload ICE due to combiner  |due to combiner extending
                   |extending life time of a    |life time of a hard register
                   |hard register               |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/41646] [4.3 Regression] Reload ICE due to combiner extending life time of a hard register
  2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
                   ` (5 preceding siblings ...)
  2009-10-12 13:54 ` [Bug rtl-optimization/41646] [4.3 " jakub at gcc dot gnu dot org
@ 2010-01-02 16:29 ` rguenth at gcc dot gnu dot org
  2010-05-22 18:43 ` rguenth at gcc dot gnu dot org
  7 siblings, 0 replies; 9+ messages in thread
From: rguenth at gcc dot gnu dot org @ 2010-01-02 16:29 UTC (permalink / raw)
  To: gcc-bugs



-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
      Known to fail|                            |4.3.4 4.4.1
      Known to work|                            |4.4.2 4.5.0
           Priority|P3                          |P2
   Last reconfirmed|0000-00-00 00:00:00         |2010-01-02 16:29:44
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/41646] [4.3 Regression] Reload ICE due to combiner extending life time of a hard register
  2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
                   ` (6 preceding siblings ...)
  2010-01-02 16:29 ` rguenth at gcc dot gnu dot org
@ 2010-05-22 18:43 ` rguenth at gcc dot gnu dot org
  7 siblings, 0 replies; 9+ messages in thread
From: rguenth at gcc dot gnu dot org @ 2010-05-22 18:43 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #6 from rguenth at gcc dot gnu dot org  2010-05-22 18:13 -------
GCC 4.3.5 is being released, adjusting target milestone.


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.3.5                       |4.3.6


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41646


^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~2010-05-22 18:33 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-10-09 14:40 [Bug rtl-optimization/41646] New: [4.3/4.4/4.5 Regression] Reload ICE due to combiner extending life time of a hard register jakub at gcc dot gnu dot org
2009-10-09 14:42 ` [Bug rtl-optimization/41646] " jakub at gcc dot gnu dot org
2009-10-09 15:35 ` jakub at gcc dot gnu dot org
2009-10-09 19:02 ` jakub at gcc dot gnu dot org
2009-10-12  8:48 ` jakub at gcc dot gnu dot org
2009-10-12 13:38 ` jakub at gcc dot gnu dot org
2009-10-12 13:54 ` [Bug rtl-optimization/41646] [4.3 " jakub at gcc dot gnu dot org
2010-01-02 16:29 ` rguenth at gcc dot gnu dot org
2010-05-22 18:43 ` rguenth at gcc dot gnu dot org

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).