[Bug rtl-optimization/26945] New: [3.4/4.0/4.1 regression] Attempt to delete prologue/epilogue insn

[Bug rtl-optimization/26945] New: [3.4/4.0/4.1 regression] Attempt to delete prologue/epilogue insn

[tausq at debian dot org]

Compile attached test case with optimization causes an ICE:

$ gcc -c -O1 bug.c
bug.c: In function 'parsearguments':
bug.c:46: error: Attempt to delete prologue/epilogue insn:
(insn/f 97 96 98 0 (set (mem:SI (plus:SI (reg/f:SI 30 %r30)
                (const_int -124 [0xffffff84])) [0 S4 A32])
        (reg:SI 4 %r4)) -1 (nil)
    (nil))
bug.c:46: internal compiler error: in propagate_one_insn, at flow.c:1690
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
For Debian GNU/Linux specific bug reporting instructions,
see <URL:file:///usr/share/doc/gcc-4.0/README.Bugs>.

This may be related to PR12535 but the code that triggers the problem is
different.


-- 
           Summary: [3.4/4.0/4.1 regression] Attempt to delete
                    prologue/epilogue insn
           Product: gcc
           Version: 4.0.3
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: rtl-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: tausq at debian dot org
 GCC build triplet: hppa-unknown-linux
  GCC host triplet: hppa-unknown-linux
GCC target triplet: hppa-unknown-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26945

[Bug rtl-optimization/26945] [3.4/4.0/4.1 regression] Attempt to delete prologue/epilogue insn

[tausq at debian dot org]

------- Comment #1 from tausq at debian dot org  2006-03-30 16:28 -------
Created an attachment (id=11167)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11167&action=view)
testcase


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26945

[Bug rtl-optimization/26945] [3.4/4.0/4.1 regression] Attempt to delete prologue/epilogue insn

[tausq at debian dot org]

------- Comment #2 from tausq at debian dot org  2006-03-30 16:31 -------
The code is valid syntactically, but there is actually a bug. the longoptions
array only has 3 elements, but we try to fill in 4 entries. if we make the
longoptions array bigger than it doesn't cause the ICE.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26945

[Bug rtl-optimization/26945] [3.4/4.0/4.1 regression] Attempt to delete prologue/epilogue insn

[pinskia at gcc dot gnu dot org]

------- Comment #3 from pinskia at gcc dot gnu dot org  2006-03-30 19:59 -------
I am starting to think we should change a[outofbounds] = 1 to be an
__builtin_trap() so that we will not run into stuff like this again.

Also we should have a keyword for ice-on-undefined-code too :).


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |minor
  GCC build triplet|hppa-unknown-linux          |
   GCC host triplet|hppa-unknown-linux          |
           Keywords|ice-on-valid-code           |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26945

[Bug rtl-optimization/26945] [4.0/4.1 regression] Attempt to delete prologue/epilogue insn

[pinskia at gcc dot gnu dot org]

-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[3.4/4.0/4.1 regression]    |[4.0/4.1 regression] Attempt
                   |Attempt to delete           |to delete prologue/epilogue
                   |prologue/epilogue insn      |insn
   Target Milestone|---                         |4.0.4


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26945

[Bug rtl-optimization/26945] [4.0/4.1 regression] Attempt to delete prologue/epilogue insn

[ted at midg3t dot net]

------- Comment #4 from ted at midg3t dot net  2006-07-06 09:38 -------
I came across a similar error, but because the test case attached to this
report did not trigger on my machine I filed a separate report, bug #28282. It
looks like it's probably related.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26945

[Bug rtl-optimization/26945] [4.0/4.1 regression] Attempt to delete prologue/epilogue insn

[rguenth at gcc dot gnu dot org]

------- Comment #5 from rguenth at gcc dot gnu dot org  2006-07-07 09:45 -------
Dup of 28282 which has a simpler testcase.

*** This bug has been marked as a duplicate of 28282 ***


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |DUPLICATE


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26945