Re: Failure to bootstrap for trunk on i686-pc-linux-gnu

[Phil Edwards <phil at codesourcery dot com>]

> > /home/pme/src/trunk/libiberty/regex.c:2321: internal compiler error: in avail_expr_eq, at tree-ssa-dom.c:3648
> > Please submit a full bug report,
> > with preprocessed source if appropriate.
> > See <URL:http://gcc.gnu.org/bugs.html> for instructions.
> > make[2]: *** [regex.o] Error 1
> > make[2]: Leaving directory `/mnt/build/build-2004-07-12-trunk/i686-pc-linux-gnu/libiberty'
> > make[1]: *** [all-target-libiberty] Error 2
> > make[1]: Leaving directory `/mnt/build/build-2004-07-12-trunk'
> > make: *** [bootstrap] Error 2

Am I the only one seeing this for the past couple days?  It's an
ENABLE_CHECKING bit of code, comparing hash values for equality.

(gdb) p ((struct expr_hash_elt *)p1)->hash
$1 = 3232635056
(gdb) p ((struct expr_hash_elt *)p2)->hash
$2 = 3498006046

Nothing obvious jumps out of the changelogs.  RTH, Zdenek, Diego and others
have all been making SSA changes, but nobody's touched the file where the
ICE occurs (we all know how little that means).

The backtrace follows, fwiw.  I assume that "line 140 million" is actually
one o' them thar new-fangled combined-location values.


(gdb) bt
#0  fancy_abort (file=0x84873c8 "/home/pme/src/trunk/gcc/tree-ssa-dom.c",
    line=140774920,
    function=0x84873c8 "/home/pme/src/trunk/gcc/tree-ssa-dom.c")
    at /home/pme/src/trunk/gcc/diagnostic.c:594
#1  0x081125cc in avail_expr_eq (p1=0x8640e08, p2=0x8602ec0)
    at /home/pme/src/trunk/gcc/tree-ssa-dom.c:3648
#2  0x0845264e in htab_find_slot_with_hash (htab=0x864e4f0, element=0x8602ec0,
    hash=13, insert=INSERT) at /home/pme/src/trunk/libiberty/hashtab.c:679
#3  0x08115ea1 in lookup_avail_expr (stmt=0x405f94ec,
    block_avail_exprs_p=0x862ea30, insert=1 '\001')
    at /home/pme/src/trunk/gcc/tree-ssa-dom.c:3292
#4  0x08116d2e in eliminate_redundant_computations (walk_data=0xbffff230,
    stmt=0x405f94ec, ann=0x40187040)
    at /home/pme/src/trunk/gcc/tree-ssa-dom.c:2645
#5  0x08118da8 in optimize_stmt (walk_data=0xbffff230, bb=0x40186414, si=
      {tsi = {ptr = 0x405f62a0, container = 0x405bf708}, bb = 0x40186414})
    at /home/pme/src/trunk/gcc/tree-ssa-dom.c:3106
#6  0x0811b8b1 in walk_dominator_tree (walk_data=0xbffff230, bb=0x40186414)
    at /home/pme/src/trunk/gcc/domwalk.c:189
#7  0x0811b6b7 in walk_dominator_tree (walk_data=0xbffff230, bb=0x4018632c)
    at /home/pme/src/trunk/gcc/domwalk.c:205

...many recursive calls to walk_dominator_tree with varying 'bb' args...

#48 0x0811b6b7 in walk_dominator_tree (walk_data=0xbffff230, bb=0x404c03a0)
    at /home/pme/src/trunk/gcc/domwalk.c:205
#49 0x0811b6b7 in walk_dominator_tree (walk_data=0xbffff230, bb=0x404c01d0)
    at /home/pme/src/trunk/gcc/domwalk.c:205
#50 0x0811405d in tree_ssa_dominator_optimize ()
    at /home/pme/src/trunk/gcc/tree-ssa-dom.c:599
#51 0x080c7922 in execute_pass_list (pass=0x85a8ae8)
    at /home/pme/src/trunk/gcc/tree-optimize.c:410
#52 0x080c7975 in execute_pass_list (pass=0x85053e0)
    at /home/pme/src/trunk/gcc/tree-optimize.c:436
#53 0x080c7ae2 in tree_rest_of_compilation (fndecl=0x40339a6c, nested_p=0 '\0')
    at /home/pme/src/trunk/gcc/tree-optimize.c:511
#54 0x08062dc5 in c_expand_body (fndecl=0x40339a6c)
    at /home/pme/src/trunk/gcc/c-decl.c:6365
#55 0x083d8b5e in cgraph_expand_function (node=0x40591d98)
    at /home/pme/src/trunk/gcc/cgraphunit.c:797
#56 0x083d999f in cgraph_optimize ()
    at /home/pme/src/trunk/gcc/cgraphunit.c:1669
#57 0x08063d81 in c_write_global_declarations ()
    at /home/pme/src/trunk/gcc/c-decl.c:6676
#58 0x08381475 in toplev_main (argc=138965960, argv=0xbffff4f4)
    at /home/pme/src/trunk/gcc/toplev.c:994
#59 0x4003e7f8 in __libc_start_main () from /lib/tls/libc.so.6
#60 0x080499c1 in _start () at ../sysdeps/i386/elf/start.S:102