[Bug debug/48203] New: ICE in dwarf2out.c while building eglibc.

[Bug debug/48203] New: ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

           Summary: ICE in dwarf2out.c while building eglibc.
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: debug
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: ramana@gcc.gnu.org


With this reduced testcase (min.i) and the original test.i cleaned from
ld-collate.c,  trunk ICE's building eglibc trunk. 


GCC was configured as --with-cpu=cortex-a9 --with-mode=thumb
--with-float=softfp --with-fpu=vfpv3-d16 

Compiling command line options are -O2 -g :

Here's a debug backtrace. 

(gdb)bt
#0  fancy_abort (file=0x116d200
"/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c", line=13885,
function=0x11741c0 "mem_loc_descriptor") at /home/ramana/cross-build/src/trun\
k/gcc/diagnostic.c:892
#1  0x000000000069c359 in mem_loc_descriptor (rtl=0x7ffff7355c70, mode=SImode,
initialized=VAR_INIT_STATUS_INITIALIZED) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:1\
3885
#2  0x000000000069c4f0 in mem_loc_descriptor (rtl=0x7ffff7246d68, mode=SImode,
initialized=VAR_INIT_STATUS_INITIALIZED) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:1\
3913
#3  0x000000000069dd2f in loc_descriptor (rtl=0x7ffff7246d50, mode=SImode,
initialized=VAR_INIT_STATUS_INITIALIZED) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:14562
#4  0x000000000069de43 in loc_descriptor (rtl=0x7ffff7247280, mode=SImode,
initialized=VAR_INIT_STATUS_INITIALIZED) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:14590
#5  0x000000000069ebb4 in dw_loc_list_1 (loc=0x7ffff746de60,
varloc=0x7ffff7247280, want_address=2, initialized=VAR_INIT_STATUS_INITIALIZED)
at /home/ramana/cross-build/src/tr\
unk/gcc/dwarf2out.c:14861
#6  0x000000000069f43c in dw_loc_list (loc_list=0x7ffff7246f48,
decl=0x7ffff746de60, want_address=2) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:15117
#7  0x00000000006a06b2 in loc_list_from_tree (loc=0x7ffff746de60,
want_address=2) at /home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:15469
#8  0x00000000006a6c26 in add_location_or_const_value_attribute
(die=0x7ffff723b2d0, decl=0x7ffff746de60, attr=DW_AT_location) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2\
out.c:16954
#9  0x00000000006b0973 in gen_variable_die (decl=0x7ffff746de60, origin=0x0,
context_die=0x7ffff723bd20) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:19831
#10 0x00000000006b6e21 in gen_decl_die (decl=0x7ffff746de60, origin=0x0,
context_die=0x7ffff723bd20) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:21317
#11 0x00000000006b4eab in process_scope_var (stmt=0x7ffff74b1000,
decl=0x7ffff746de60, origin=0x0, context_die=0x7ffff723bd20) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2\
out.c:20864
#12 0x00000000006b4f33 in decls_for_scope (stmt=0x7ffff74b1000,
context_die=0x7ffff723bd20, depth=0) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:20886
#13 0x00000000006af0d7 in gen_subprogram_die (decl=0x7ffff74b0000,
context_die=0x7ffff7560320) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:19457
#14 0x00000000006b6aa9 in gen_decl_die (decl=0x7ffff74b0000, origin=0x0,
context_die=0x7ffff7560320) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:21250
#15 0x00000000006b7ad3 in dwarf2out_decl (decl=0x7ffff74b0000) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:21624
#16 0x00000000006b7aea in dwarf2out_function_decl (decl=0x7ffff74b0000) at
/home/ramana/cross-build/src/trunk/gcc/dwarf2out.c:21632
#17 0x0000000000728fc0 in rest_of_handle_final () at
/home/ramana/cross-build/src/trunk/gcc/final.c:4280
#18 0x00000000009041f1 in execute_one_pass (pass=0x16dc860) at
/home/ramana/cross-build/src/trunk/gcc/passes.c:1556
#19 0x00000000009045b8 in execute_pass_list (pass=0x16dc860) at
/home/ramana/cross-build/src/trunk/gcc/passes.c:1611
#20 0x00000000009045d6 in execute_pass_list (pass=0x16dd220) at
/home/ramana/cross-build/src/trunk/gcc/passes.c:1612
#21 0x00000000009045d6 in execute_pass_list (pass=0x16dd1c0) at
/home/ramana/cross-build/src/trunk/gcc/passes.c:1612
#22 0x0000000000a89dcd in tree_rest_of_compilation (fndecl=0x7ffff74b0000) at
/home/ramana/cross-build/src/trunk/gcc/tree-optimize.c:422
#23 0x0000000000d018e8 in cgraph_expand_function (node=0x7ffff74ac580) at
/home/ramana/cross-build/src/trunk/gcc/cgraphunit.c:1576
#24 0x0000000000d01a98 in cgraph_expand_all_functions () at
/home/ramana/cross-build/src/trunk/gcc/cgraphunit.c:1635
#25 0x0000000000d0211c in cgraph_optimize () at
/home/ramana/cross-build/src/trunk/gcc/cgraphunit.c:1899
#26 0x0000000000cff4ee in cgraph_finalize_compilation_unit () at
/home/ramana/cross-build/src/trunk/gcc/cgraphunit.c:1096
#27 0x00000000004beb81 in c_write_global_declarations () at
/home/ramana/cross-build/src/trunk/gcc/c-decl.c:9871
#28 0x00000000009fb08a in compile_file () at
/home/ramana/cross-build/src/trunk/gcc/toplev.c:591
#29 0x00000000009fd388 in do_compile () at
/home/ramana/cross-build/src/trunk/gcc/toplev.c:1900
#30 0x00000000009fd4b3 in toplev_main (argc=4, argv=0x7fffffffe728) at
/home/ramana/cross-build/src/trunk/gcc/toplev.c:1963
#31 0x00000000005958df in main (argc=4, argv=0x7fffffffe728) at
/home/ramana/cross-build/src/trunk/gcc/main.c:36


(gdb) p rtl
$4 = (rtx) 0x7ffff7246d50
(gdb) pr
(plus:SI (entry_value:SI (mem/f/c/i:SI (plus:SI (reg/f:SI 26 afp)
                (const_int 4 [0x4])) [3 result+0 S4 A32]))
    (const_int 12 [0xc]))

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #1 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-19 14:59:37 UTC ---
Created attachment 23720
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23720
Testcase

Reduced testcase with delta.

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #2 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-19 15:02:15 UTC ---
Created attachment 23722
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23722
Original test.

Unreduced testcase.

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #3 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-19 15:12:55 UTC ---
This also breaks bootstrap in stage3 for arm-linux-gnueabi while building
c-common.c 

Ramana

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

Ramana Radhakrishnan <ramana at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2011.03.19 16:36:13
                 CC|                            |jakub at gcc dot gnu.org
     Ever Confirmed|0                           |1

--- Comment #4 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-19 16:36:13 UTC ---
svn+ssh://gcc.gnu.org/svn/gcc/trunk@171033 => Broken
svn+ssh://gcc.gnu.org/svn/gcc/trunk@171031 => Good

Ramana

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

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

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-19 18:35:38 UTC ---
Created attachment 23725
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23725
gcc47-pr48203.patch

Untested fix.

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #6 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-19 19:13:30 UTC ---
(In reply to comment #5)
> Created attachment 23725 [details]
> gcc47-pr48203.patch
> 
> Untested fix

Bootstrap and cross-testing on in parallel.

Ramana(In reply to comment #5)
> Created attachment 23725 [details]
> gcc47-pr48203.patch
> 
> Untested fix.

ramana@gcc10:~/cross-build/arm-none-linux-gnueabi/obj/gcc2/gcc$ ./xgcc -B`pwd`
-S -O2 -g ~/min.i 
/home/ramana/min.i:96:1: warning: useless storage class specifier in empty
declaration [enabled by default]
/home/ramana/min.i:101:49: warning: ‘struct tm’ declared inside parameter list
[enabled by default]
/home/ramana/min.i:101:49: warning: its scope is only this definition or
declaration, which is probably not what you want [enabled by default]
/home/ramana/min.i:131:8: warning: ‘struct random_data’ declared inside
parameter list [enabled by default]
/home/ramana/min.i:133:16: warning: ‘struct drand48_data’ declared inside
parameter list [enabled by default]
/home/ramana/min.i:164:16: warning: ‘struct timezone’ declared inside parameter
list [enabled by default]
/home/ramana/min.i:241:5: warning: no semicolon at end of struct or union
[enabled by default]
/home/ramana/min.i:267:10: warning: ‘struct iovec’ declared inside parameter
list [enabled by default]
/home/ramana/min.i:448:14: warning: conflicting types for built-in function
‘__strcpy_chk’ [enabled by default]
/home/ramana/min.i:450:14: warning: conflicting types for built-in function
‘__strncat_chk’ [enabled by default]
/home/ramana/min.i:548:1: warning: declaration does not declare anything
[enabled by default]
/home/ramana/min.i:548:1: warning: no semicolon at end of struct or union
[enabled by default]
/home/ramana/min.i:562:14: warning: conflicting types for built-in function
‘dcgettext’ [enabled by default]
/home/ramana/min.i:573:1: warning: unnamed struct/union that defines no
instances [enabled by default]
/home/ramana/min.i:575:7: warning: ‘struct __gconv_step’ declared inside
parameter list [enabled by default]
/home/ramana/min.i: In function ‘read_directions’:
/home/ramana/min.i:906:32: warning: incompatible implicit declaration of
built-in function ‘calloc’ [enabled by default]
/home/ramana/min.i: In function ‘insert_weights’:
/home/ramana/min.i:976:4: warning: incompatible implicit declaration of
built-in function ‘memcpy’ [enabled by default]
/home/ramana/min.i: In function ‘handle_ellipsis’:
/home/ramana/min.i:1043:21: warning: incompatible implicit declaration of
built-in function ‘strlen’ [enabled by default]
/home/ramana/min.i:1059:14: warning: assignment makes pointer from integer
without a cast [enabled by default]
/home/ramana/min.i: In function ‘collate_read’:
/home/ramana/min.i:1238:16: warning: assignment makes pointer from integer
without a cast [enabled by default]
/home/ramana/min.i:1260:21: warning: assignment makes pointer from integer
without a cast [enabled by default]
/home/ramana/min.i:1306:6: warning: incompatible implicit declaration of
built-in function ‘sprintf’ [enabled by default]
/home/ramana/min.i:1307:8: warning: passing argument 2 of ‘sprintf’ makes
pointer from integer without a cast [enabled by default]
/home/ramana/min.i:1307:8: note: expected ‘const char *’ but argument is of
type ‘int’
/home/ramana/min.i:1424:120: warning: incompatible implicit declaration of
built-in function ‘strlen’ [enabled by default]
/home/ramana/min.i:1447:82: warning: initialization makes pointer from integer
without a cast [enabled by default]
/home/ramana/min.i:1447:4: warning: passing argument 5 of ‘error_at_line’ makes
pointer from integer without a cast [enabled by default]
/home/ramana/min.i:363:1: note: expected ‘const char *’ but argument is of type
‘int’
/home/ramana/min.i: In function ‘read_directions’:
/home/ramana/min.i:941:1: internal compiler error: in mem_loc_descriptor, at
dwarf2out.c:13885
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.


cheers
Ramana

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #7 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-19 19:16:34 UTC ---
(In reply to comment #5)
> Created attachment 23725 [details]
> gcc47-pr48203.patch
> 
> Untested fix.

Please ignore comment #6 - I was using the wrong compiler accidentally. 


This fixes the ICE with min.i but the ICE with ld-collate.i still remains at a
different place. 


ramana@gcc10:~/cross-build/arm-none-linux-gnueabi/objpr48203/gcc2/gcc$ ./xgcc
-B`pwd` -S -O2 -g /tmp/ld-collate.i 
programs/ld-collate.c: In function ‘collate_output’:
programs/ld-collate.c:2500:1: internal compiler error: in
dwarf2out_var_location, at dwarf2out.c:21983
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

I think there are a couple of problems here.

cheers
Ramana

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #8 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-19 19:21:27 UTC ---
./cc1 -O2 -g ld-collate.i -quiet; echo $?
0

if you are talking about #c2 testcase.  I have cross configured with
../configure --disable-bootstrap --target armv5tel-unknown-linux-gnueabi
--enable-languages=c
Perhaps some -m* cpu selection etc. options are needed to reproduce?

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #9 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-19 19:25:30 UTC ---
(In reply to comment #8)
> ./cc1 -O2 -g ld-collate.i -quiet; echo $?
> 0
> 
> if you are talking about #c2 testcase.  I have cross configured with
> ../configure --disable-bootstrap --target armv5tel-unknown-linux-gnueabi
> --enable-languages=c
> Perhaps some -m* cpu selection etc. options are needed to reproduce?

Can you try with -mcpu=cortex-a9 -mfpu=vfpv3-d16 -mfloat-abi=softfp -mthumb ?
It fails in that case . #c1 has the configuration options that I used. 

cheers
Ramana

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #10 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-19 20:02:04 UTC ---
(In reply to comment #9)
> Can you try with -mcpu=cortex-a9 -mfpu=vfpv3-d16 -mfloat-abi=softfp -mthumb ?
> It fails in that case . #c1 has the configuration options that I used. 

Can't reproduce, neither with the above options, nor with a fresh cross
configured the way you mentioned in #c0.

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #11 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-19 22:14:50 UTC ---
(In reply to comment #10)
> (In reply to comment #9)
> > Can you try with -mcpu=cortex-a9 -mfpu=vfpv3-d16 -mfloat-abi=softfp -mthumb ?
> > It fails in that case . #c1 has the configuration options that I used. 
> 
> Can't reproduce, neither with the above options, nor with a fresh cross
> configured the way you mentioned in #c0.

I see the ICE with -mcpu=cortex-a9 -mfpu=vfpv3-d16 -mfloat-abi=softfp -marm . 

It appears as though when I did a quick rebuild of the tools to verify your
patch I missed the --with-mode=thumb. Sorry about the confusion. 


I've dug into this a bit further and it appears as though the ICE at 


21979    gcc_assert (prev
21980                && (CALL_P (prev)
21981                || (NONJUMP_INSN_P (prev)
21982                && GET_CODE (PATTERN (prev)) == SEQUENCE
21983                && CALL_P (XVECEXP (PATTERN (prev), 0, 0)))));




Looking at the backtrace it appears as though the assertion failure is because
we don't see the call before the note. Looking at the dumps the ARM backend
ends up generating the minipools to interleave the constant pools within the
function because of range issues and thus end up breaking the assertion. The
work around is then to emit the barriers and the minipools after the notes
rather than just after the "call insn".

 2265 call [`obstack_free'] argc:0
      REG_DEAD: r1:SI
      REG_DEAD: r0:SI
 4874 pc=L4876
i4875: barrier
L4877:
 4878 unspec/v{0;}
L4879:
 4880 unspec/v{1.399999999999999911182158029987476766109466552734375e+0;}
 4881 unspec/v{`swap_endianness_p';}
 4882 unspec/v{const(`*.LANCHOR1'+0x7d0);}
 4883 unspec/v{const(`*.LANCHOR1'+0x2c);}
 4884 unspec/v{const(`*.LANCHOR1'+0x7bc);}
 4885 unspec/v{const(`*.LANCHOR1'+0x7a8);}
 4886 unspec/v{const(`*.LANCHOR1'+0x70c);}
 4887 unspec/v{const(`*.LANCHOR1'+0x218);}
 4888 unspec/v{0;}
i4889: barrier
L4876:
 4011 NOTE_INSN_CALL_ARG_LOCATION
 4012 NOTE_INSN_VAR_LOCATION
 4013 NOTE_INSN_VAR_LOCATION

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #12 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-03-20 07:59:59 UTC ---
Author: ramana
Date: Sun Mar 20 07:59:54 2011
New Revision: 171193

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=171193
Log:

Fix PR debug/48203

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/arm/arm.c

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #13 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-28 23:53:49 UTC ---
Author: jakub
Date: Mon Mar 28 23:53:46 2011
New Revision: 171640

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=171640
Log:
    PR debug/48203
    * cfgexpand.c (expand_debug_expr) <case SSA_NAME>: Only
    create ENTRY_VALUE if incoming or address of incoming's MEM
    is a hard REG.
    * dwarf2out.c (mem_loc_descriptor): Don't emit
    DW_OP_GNU_entry_value of DW_OP_fbreg.
    * var-tracking.c (vt_add_function_parameter): Ensure cselib_lookup
    on ENTRY_VALUE is able to find the canonical parameter VALUE.
    * cselib.c (rtx_equal_for_cselib_1) <case ENTRY_VALUE>: Use
    rtx_equal_p instead of rtx_equal_for_cselib_1 to compare
    ENTRY_VALUE_EXPs.
    (cselib_hash_rtx) <case ENTRY_VALUE>: If ENTRY_VALUE_EXP
    is a REG_P or MEM_P with REG_P address, compute hash directly
    instead of calling cselib_hash_rtx on ENTRY_VALUE_EXP.
    (preserve_only_constants): Don't clear VALUES forwaring
    ENTRY_VALUE to some other VALUE.

    * gcc.dg/pr48203.c: New test.

Added:
    trunk/gcc/testsuite/gcc.dg/pr48203.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/cfgexpand.c
    trunk/gcc/cselib.c
    trunk/gcc/dwarf2out.c
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/var-tracking.c

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[ramana at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

--- Comment #14 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> 2011-05-31 22:30:42 UTC ---
(In reply to comment #13)
> Author: jakub
> Date: Mon Mar 28 23:53:46 2011
> New Revision: 171640
> 
> URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=171640
> Log:
>     PR debug/48203
>     * cfgexpand.c (expand_debug_expr) <case SSA_NAME>: Only
>     create ENTRY_VALUE if incoming or address of incoming's MEM
>     is a hard REG.
>     * dwarf2out.c (mem_loc_descriptor): Don't emit
>     DW_OP_GNU_entry_value of DW_OP_fbreg.
>     * var-tracking.c (vt_add_function_parameter): Ensure cselib_lookup
>     on ENTRY_VALUE is able to find the canonical parameter VALUE.
>     * cselib.c (rtx_equal_for_cselib_1) <case ENTRY_VALUE>: Use
>     rtx_equal_p instead of rtx_equal_for_cselib_1 to compare
>     ENTRY_VALUE_EXPs.
>     (cselib_hash_rtx) <case ENTRY_VALUE>: If ENTRY_VALUE_EXP
>     is a REG_P or MEM_P with REG_P address, compute hash directly
>     instead of calling cselib_hash_rtx on ENTRY_VALUE_EXP.
>     (preserve_only_constants): Don't clear VALUES forwaring
>     ENTRY_VALUE to some other VALUE.
> 
>     * gcc.dg/pr48203.c: New test.
> 
> Added:
>     trunk/gcc/testsuite/gcc.dg/pr48203.c
> Modified:
>     trunk/gcc/ChangeLog
>     trunk/gcc/cfgexpand.c
>     trunk/gcc/cselib.c
>     trunk/gcc/dwarf2out.c
>     trunk/gcc/testsuite/ChangeLog
>     trunk/gcc/var-tracking.c


Jakub, Is this now fixed ?

Ramana

[Bug debug/48203] ICE in dwarf2out.c while building eglibc.

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48203

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED

--- Comment #15 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-06-01 06:32:11 UTC ---
Hopefully.