[Bug bootstrap/11745] New: Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[Bug bootstrap/11745] New: Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[tobias dot schlueter at web dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745

           Summary: Bootstrap fails when compiling insn-attrtab.c with
                    stage1/xgcc
           Product: gcc
           Version: tree-ssa
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: bootstrap
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: tobias dot schlueter at web dot de
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-cygwin
  GCC host triplet: i686-pc-cygwin
GCC target triplet: i686-pc-cygwin

I try to build with 

../gcc/configure --prefix=/tmp/gcc-g95 --enable-languages=c,f95 
make bootstrap-lean

The bootstrap then ends with:

echo timestamp > s-attrtab
stage1/xgcc.exe -Bstage1/ -B/tmp/gcc-g95/i686-pc-cygwin/bin/   -g -O2  -DIN_GCC
  -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -W
no-long-long -Werror -fno-common   -DHAVE_CONFIG_H    -I. -I. -I../../gcc/gcc -I
../../gcc/gcc/. -I../../gcc/gcc/config -I../../gcc/gcc/../include -I../../gcc/gc
c/../libbanshee/libcompat -I../../gcc/gcc/../libbanshee -I../../gcc/gcc/../libba
nshee/points-to -c insn-attrtab.c \
  -o insn-attrtab.o
make[2]: *** [insn-attrtab.o] Error 1
make[2]: Leaving directory `/home/tobi/src/gcc/build/gcc'
make[1]: *** [stage2_build] Error 2
make[1]: Leaving directory `/home/tobi/src/gcc/build/gcc'
make: *** [bootstrap-lean] Error 2

Please note that there's no error message in the above. This happens less than
ten seconds after this command is issued. Compiling manually with -v gives the
following output:

tobi@Marktplatz ~/src/gcc/build/gcc
$ stage1/xgcc.exe -Bstage1/ -B/tmp/gcc-g95/i686-pc-cygwin/bin/   -g -O2  -DIN_G
CC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic
-Wno-long-long -Werror -fno-common   -DHAVE_CONFIG_H    -I. -I. -I../../gcc/gcc
 -I../../gcc/gcc/. -I../../gcc/gcc/config -I../../gcc/gcc/../include -I../../gc
c/gcc/../libbanshee/libcompat -I../../gcc/gcc/../libbanshee -I../../gcc/gcc/../
libbanshee/points-to -c insn-attrtab.c  -o insn-attrtab.o -v
Reading specs from stage1/specs
Configured with: ../gcc/configure --prefix=/tmp/gcc-g95 --enable-languages=c,f95

Thread model: single
gcc version 3.5-tree-ssa 20030731 (merged 20030722)
 stage1/cc1.exe -quiet -v -I. -I. -I../../gcc/gcc -I../../gcc/gcc/. -I../../gcc/
gcc/config -I../../gcc/gcc/../include -I../../gcc/gcc/../libbanshee/libcompat -I
../../gcc/gcc/../libbanshee -I../../gcc/gcc/../libbanshee/points-to -iprefix /ho
me/tobi/src/gcc/build/gcc/stage1/../lib/gcc-lib/i686-pc-cygwin/3.5-tree-ssa/ -is
ystem include -isystem stage1/include -isystem /tmp/gcc-g95/i686-pc-cygwin/bin/i
nclude -D__CYGWIN32__ -D__CYGWIN__ -Dunix -D__unix__ -D__unix -idirafter /usr/li
b/../include/w32api -idirafter ../../include/w32api -DIN_GCC -DHAVE_CONFIG_H ins
n-attrtab.c -quiet -dumpbase insn-attrtab.c -mtune=pentiumpro -auxbase-strip ins
n-attrtab.o -g -O2 -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-protot
ypes -pedantic -Wno-long-long -Werror -version -fno-common -o /cygdrive/c/DOCUME
~1/TOBI~1.MAR/LOCALS~1/Temp/ccMDExKy.s
ignoring nonexistent directory "stage1/include"
ignoring nonexistent directory "/tmp/gcc-g95/i686-pc-cygwin/bin/include"
ignoring nonexistent directory "/home/tobi/src/gcc/build/gcc/lib/gcc-lib/i686-pc
$ stage1/xgcc.exe -Bstage1/ -B/tmp/gcc-g95/i686-pc-cygwin/bin/   -g -O2  -DIN_G
CC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic c
-Wno-long-long -Werror -fno-common   -DHAVE_CONFIG_H    -I. -I. -I../../gcc/gcc
 -I../../gcc/gcc/. -I../../gcc/gcc/config -I../../gcc/gcc/../include -I../../gc
c/gcc/../libbanshee/libcompat -I../../gcc/gcc/../libbanshee -I../../gcc/gcc/../e
libbanshee/points-to -c insn-attrtab.c  -o insn-attrtab.o -v
Reading specs from stage1/specs"/tmp/gcc-g95/i686-pc-cygwin/include"
Configured with: ../gcc/configure --prefix=/tmp/gcc-g95 --enable-languages=c,f95
ignoring duplicate directory "."
Thread model: singleirectory "../../gcc/gcc"
gcc version 3.5-tree-ssa 20030731 (merged 20030722)
 stage1/cc1.exe -quiet -v -I. -I. -I../../gcc/gcc -I../../gcc/gcc/. -I../../gcc/
gcc/config -I../../gcc/gcc/../include -I../../gcc/gcc/../libbanshee/libcompat -I
../../gcc/gcc/../libbanshee -I../../gcc/gcc/../libbanshee/points-to -iprefix /ho
me/tobi/src/gcc/build/gcc/stage1/../lib/gcc-lib/i686-pc-cygwin/3.5-tree-ssa/ -is
ystem include -isystem stage1/include -isystem /tmp/gcc-g95/i686-pc-cygwin/bin/i
nclude -D__CYGWIN32__ -D__CYGWIN__ -Dunix -D__unix__ -D__unix -idirafter /usr/li
b/../include/w32api -idirafter ../../include/w32api -DIN_GCC -DHAVE_CONFIG_H ins
n-attrtab.c -quiet -dumpbase insn-attrtab.c -mtune=pentiumpro -auxbase-strip ins
n-attrtab.o -g -O2 -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-protot
ypes -pedantic -Wno-long-long -Werror -version -fno-common -o /cygdrive/c/DOCUME
~1/TOBI~1.MAR/LOCALS~1/Temp/ccmT37xj.s
ignoring nonexistent directory "stage1/include"
ignoring nonexistent directory "/tmp/gcc-g95/i686-pc-cygwin/bin/include"
ignoring nonexistent directory "/home/tobi/src/gcc/build/gcc/lib/gcc-lib/i686-pc
-cygwin/3.5-tree-ssa/include"
ignoring nonexistent directory "/home/tobi/src/gcc/build/gcc/lib/gcc-lib/i686-pc
-cygwin/3.5-tree-ssa/../../../../i686-pc-cygwin/include"
ignoring nonexistent directory "/tmp/gcc-g95/include"
ignoring nonexistent directory "/tmp/gcc-g95/lib/gcc-lib/i686-pc-cygwin/3.5-tree
-ssa/include"
ignoring nonexistent directory "/tmp/gcc-g95/i686-pc-cygwin/include"
ignoring nonexistent directory "../../include/w32api"
ignoring duplicate directory "."
ignoring duplicate directory "../../gcc/gcc"
#include "..." search starts here:
#include <...> search starts here:
 .
 ../../gcc/gcc
 ../../gcc/gcc/config
 ../../gcc/include
 ../../gcc/libbanshee/libcompat
 ../../gcc/libbanshee
 ../../gcc/libbanshee/points-to
 include
 /usr/local/include
 /usr/include
 /usr/include/w32api
End of search list.
GNU C version 3.5-tree-ssa 20030731 (merged 20030722) (i686-pc-cygwin)
        compiled by GNU C version 3.2 20020927 (prerelease).
GGC heuristics: --param ggc-min-expand=47 --param ggc-min-heapsize=32702

tobi@Marktplatz ~/src/gcc/build/gcc
$

If you want me to attach the generated insn-attrtab.c (or the .i-file) I will. I
refrained from doing so because of its size. If you tell me what to enter into

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[tobias dot schlueter at web dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From tobias dot schlueter at web dot de  2003-07-31 14:14 -------
Ok, I just ran it in gdb. It looks like infinite recursion is happening in
walk_tree(). The lower end of the call stack looks like this:

#8132 0x006a659c in walk_tree (tp=0x10ffe23c, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8133 0x006a659c in walk_tree (tp=0x10ffe07c, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8134 0x006a659c in walk_tree (tp=0x10ffe1dc, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8135 0x006a659c in walk_tree (tp=0x10faec1c, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8136 0x006a659c in walk_tree (tp=0x10faeabc, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8137 0x006a659c in walk_tree (tp=0x10fb301c, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8138 0x006a659c in walk_tree (tp=0x10ffc594, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8139 0x006a659c in walk_tree (tp=0x112d631c, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8140 0x006a659c in walk_tree (tp=0x10ff9fdc, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8141 0x006a659c in walk_tree (tp=0x10fa8fe4, func=0x6d92fe <copy_if_shared_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8142 0x006a659c in walk_tree (tp=0x22eb00, func=0x6d92fe <copy_if_shared_r>, da
ta=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
#8143 0x006d92d9 in unshare_all_trees (t=0x10fa8fc8) at ../../gcc/gcc/gimplify.c
:2622
#8144 0x006cb7b3 in gimplify_body (body_p=0x22eb60, fndecl=0x107da0e0) at ../../
gcc/gcc/gimplify.c:269
#8145 0x006cb70a in gimplify_function_tree (fndecl=0x107da0e0) at ../../gcc/gcc/
gimplify.c:239
#8146 0x004b7b63 in c_genericize (fndecl=0x107da0e0) at ../../gcc/gcc/c-simplify
.c:161
#8147 0x00437230 in finish_function (nested=0, can_defer_p=1) at ../../gcc/gcc/c
-decl.c:6247
#8148 0x0040b206 in yyparse () at c-parse.y:385
#8149 0x00416719 in c_parse_file () at c-parse.y:3022
#8150 0x004864c4 in c_common_parse_file (set_yydebug=0) at ../../gcc/gcc/c-opts.
c:1206
#8151 0x004dd8d1 in compile_file () at ../../gcc/gcc/toplev.c:1793
#8152 0x004e3074 in do_compile () at ../../gcc/gcc/toplev.c:4584
#8153 0x004e3110 in toplev_main (argc=52, argv=0x10021b60) at ../../gcc/gcc/topl
ev.c:4625
#8154 0x004d8f80 in main (argc=52, argv=0x10021b60) at ../../gcc/gcc/main.c:35


Giving up a few times I see gcc's error handlers:
(gdb) up
#1  0x01010101 in _libkernel32_a_iname ()
(gdb) up
#2  0x6101c3e4 in cygwin_internal () from /usr/bin/cygwin1.dll
(gdb) up
#3  0x6101c963 in cygwin_internal () from /usr/bin/cygwin1.dll
(gdb) up
#4  0x610170a1 in cygwin_internal () from /usr/bin/cygwin1.dll
(gdb) up
#5  0x6107e30f in writev () from /usr/bin/cygwin1.dll
(gdb) up
#6  0x6107db0c in write () from /usr/bin/cygwin1.dll
(gdb) up
#7  0x610c6d26 in wmemset () from /usr/bin/cygwin1.dll
(gdb) up
#8  0x610c55da in wmemset () from /usr/bin/cygwin1.dll
(gdb) up
#9  0x610c5978 in wmemset () from /usr/bin/cygwin1.dll
(gdb) up
#10 0x610b3899 in fputs () from /usr/bin/cygwin1.dll
(gdb) up
#11 0x00507775 in output_buffer_to_stream (buffer=0x9d4390) at ../../gcc/gcc/dia
gnostic.c:432
432       fputs (text, output_buffer_attached_stream (buffer));
(gdb) up
#12 0x00508d78 in lhd_print_error_function (context=0x9d4390, file=0x22fdf6 "ins
n-attrtab.c") at ../../gcc/gcc/diagnostic.c:953
953           output_buffer_to_stream (&context->buffer);
(gdb) up
#13 0x00508db5 in diagnostic_report_current_function (context=0x9d4390) at ../..
/gcc/gcc/diagnostic.c:967
967       (*lang_hooks.print_error_function) (context, input_filename);
(gdb) up
#14 0x00508ef3 in default_diagnostic_starter (context=0x9d4390, diagnostic=0x32f
50) at ../../gcc/gcc/diagnostic.c:1002
1002      diagnostic_report_current_function (context);
(gdb) up
#15 0x00508f80 in diagnostic_report_diagnostic (context=0x9d4390, diagnostic=0x3
2f50) at ../../gcc/gcc/diagnostic.c:1028
1028          (*diagnostic_starter (context)) (context, diagnostic);
(gdb) up
#16 0x0050942e in internal_error (msgid=0x4dc7b0 "%s") at ../../gcc/gcc/diagnost
ic.c:1232
1232      report_diagnostic (&diagnostic);
(gdb) up
#17 0x004dc7e8 in crash_signal (signo=11) at ../../gcc/gcc/toplev.c:1371
1371      internal_error ("%s", strsignal (signo));
(gdb) up
#18 0x610149e3 in cygwin_stackdump () from /usr/bin/cygwin1.dll
(gdb) up
#19 0x006a619e in walk_tree (tp=0x10cb2218, func=0x6d939a <mostly_copy_tree_r>,
data=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1463
1463      result = (*func) (tp, &walk_subtrees, data);
(gdb) up
#20 0x006d933a in copy_if_shared_r (tp=0x10cb2218, walk_subtrees=0x331cc, data=0
x0) at ../../gcc/gcc/gimplify.c:2643
2643          walk_tree (tp, mostly_copy_tree_r, NULL, NULL);
(gdb) up
#21 0x006a61b4 in walk_tree (tp=0x10cb2218, func=0x6d92fe <copy_if_shared_r>, da
ta=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1463
1463      result = (*func) (tp, &walk_subtrees, data);
(gdb) up
#22 0x006a659c in walk_tree (tp=0x111aadbc, func=0x6d92fe <copy_if_shared_r>, da
ta=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
1511            WALK_SUBTREE (TREE_OPERAND (*tp, i));
(gdb) up
#23 0x006a659c in walk_tree (tp=0x111aadf8, func=0x6d92fe <copy_if_shared_r>, da
ta=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
1511            WALK_SUBTREE (TREE_OPERAND (*tp, i));
(gdb) up
#24 0x006a659c in walk_tree (tp=0x111aaabc, func=0x6d92fe <copy_if_shared_r>, da
ta=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
1511            WALK_SUBTREE (TREE_OPERAND (*tp, i));
(gdb) up
#25 0x006a659c in walk_tree (tp=0x111aaa9c, func=0x6d92fe <copy_if_shared_r>, da
ta=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
1511            WALK_SUBTREE (TREE_OPERAND (*tp, i));
(gdb) up
#26 0x006a659c in walk_tree (tp=0x111aaa5c, func=0x6d92fe <copy_if_shared_r>, da
ta=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
1511            WALK_SUBTREE (TREE_OPERAND (*tp, i));
(gdb) up
#27 0x006a659c in walk_tree (tp=0x111aaa1c, func=0x6d92fe <copy_if_shared_r>, da
ta=0x0, htab_=0x0) at ../../gcc/gcc/tree-inline.c:1511
1511            WALK_SUBTREE (TREE_OPERAND (*tp, i));
(gdb)

... and so on.

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[tobias dot schlueter at web dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From tobias dot schlueter at web dot de  2003-07-31 14:16 -------
Sorry for being so obnoxious as to post several times but I forgot to mention
that I've been seeing this same error already for some time. 

This has already been happening around the time I posted to gcc@gcc.gnu.org
http://gcc.gnu.org/ml/gcc/2003-07/msg00602.html
at that time I couldn't reliably reproduce it, though.

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[pinskia at physics dot uc dot edu]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745


pinskia at physics dot uc dot edu changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |build, ice-on-valid-code
   Target Milestone|3.4                         |tree-ssa

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[jv244 at cam dot ac dot uk]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From jv244 at cam dot ac dot uk  2003-08-13 19:39 -------
notice that today's version of ssa reports an error and location

stage1/xgcc.exe -Bstage1/ -B/home/vondele/gcc-g95/tmp/i686-pc-cygwin/bin/
-g -O2  -DIN_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes
-Wmissing-prototypes -pedantic -Wno-long-long -Werror -fno-common
-DHAVE_CONFIG_H    -I. -I. -I../../gcc/gcc -I../../gcc/gcc/.
-I../../gcc/gcc/config -I../../gcc/gcc/../include
-I../../gcc/gcc/../libbanshee/libcompat -I../../gcc/gcc/../libbanshee
-I../../gcc/gcc/../libbanshee/points-to -I/home/vondele/download/gmp-4.1.2
-c insn-attrtab.c \
  -o insn-attrtab.o
insn-attrtab.c: In function `insn_default_latency':
insn-attrtab.c:4189: internal compiler error: Segmentation fault

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[tobias dot schlueter at web dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From tobias dot schlueter at web dot de  2003-08-13 21:50 -------
Subject: Re:  Bootstrap fails when compiling insn-attrtab.c
 with stage1/xgcc



jv244 at cam dot ac dot uk wrote:

> notice that today's version of ssa reports an error and location
> 

not for me, I still get the same error I described above. And I've been 
constantly trying since even before I reported this bug.

I use the same bootstrap compiler as Jost (see: 
http://gcc.gnu.org/ml/gcc/2003-08/msg00855.html), but I use 
bootstrap-lean instead. I don't see how this could make a difference though.

- Tobi

> stage1/xgcc.exe -Bstage1/ -B/home/vondele/gcc-g95/tmp/i686-pc-cygwin/bin/
> -g -O2  -DIN_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes
> -Wmissing-prototypes -pedantic -Wno-long-long -Werror -fno-common
> -DHAVE_CONFIG_H    -I. -I. -I../../gcc/gcc -I../../gcc/gcc/.
> -I../../gcc/gcc/config -I../../gcc/gcc/../include
> -I../../gcc/gcc/../libbanshee/libcompat -I../../gcc/gcc/../libbanshee
> -I../../gcc/gcc/../libbanshee/points-to -I/home/vondele/download/gmp-4.1.2
> -c insn-attrtab.c \
>   -o insn-attrtab.o
> insn-attrtab.c: In function `insn_default_latency':
> insn-attrtab.c:4189: internal compiler error: Segmentation fault
> 
> 
> 
> ------- You are receiving this mail because: -------
> You reported the bug, or are watching the reporter.
>

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[tobias dot schlueter at web dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From tobias dot schlueter at web dot de  2003-08-13 21:58 -------
Subject: Re:  Bootstrap fails when compiling insn-attrtab.c
 with stage1/xgcc

tobias dot schlueter at web dot de wrote:
> not for me, I still get the same error I described above. And I've been 
> constantly trying since even before I reported this bug.
                    ^ and always getting this error

Sorry, I left out the relevant part.
- Tobi

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[pinskia at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745


pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jv244 at cam dot ac dot uk
             Status|UNCONFIRMED                 |WAITING


------- Additional Comments From pinskia at gcc dot gnu dot org  2003-08-15 03:30 -------
Do you know what the default stacksize on cygwin is?
I think if you increase the stacksize to 10M or higher you might get passed the first stage.
Can one of you two try it?
I know on powerpc-apple-darwin6.6 I have to increase the stacksize to 8M from 1/2M (the 
too low default).

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[jv244 at cam dot ac dot uk]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From jv244 at cam dot ac dot uk  2003-08-15 07:21 -------
sounds like the stack size could be the problem.
I get the following info in cygwin :
ulimit -s
2047
ulimit -s 8000
bash: ulimit: stack size: cannot modify limit: invalid argument

from the following mail on the cygwin list:
http://sources.redhat.com/ml/cygwin/2003-03/msg01394.html

This is the real answer. The system call to which the ulimit shell built-in 
delegates its work is not implemented in Cygwin.

I believe link-time is the only way to override the default stack allotment is 
during compilation (or linking, as the case may be): E.g. (from a message by 
Gerrit P. Haase, subject "Re: increase stacksize" sent Wed, 17 Jul 2002 
15:54:42 +0200):

gcc -Wl,--stack,8388608 -o your.exe your.source.c

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[tobias dot schlueter at web dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From tobias dot schlueter at web dot de  2003-08-15 11:17 -------
Subject: Re:  Bootstrap fails when compiling insn-attrtab.c
 with stage1/xgcc



jv244 at cam dot ac dot uk wrote:
> 
> This is the real answer. The system call to which the ulimit shell built-in 
> delegates its work is not implemented in Cygwin.
> 
> I believe link-time is the only way to override the default stack allotment is 
> during compilation (or linking, as the case may be): E.g. (from a message by 
> Gerrit P. Haase, subject "Re: increase stacksize" sent Wed, 17 Jul 2002 
> 15:54:42 +0200):
> 
> gcc -Wl,--stack,8388608 -o your.exe your.source.c
> 

I tried this. Running
	CC="gcc -Wl,--stack=0x2000000" make bootstrap-lean
didn't change anything.

Configuring with
	CC="gcc -Wl,--stack=0x2000000" ../gcc/configure etc.
confuses configure, I get a lot of lines like below:

checking for sys/file.h... grep: conftest.out: No such file or directory
yes
checking for sys/param.h... grep: conftest.out: No such file or directory
yes

The build then bails out with an error relatively fast.

Is there any place where I can set linker options, so I don't confuse 
the compiler with the '-Wl'?

- Tobi

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[jv244 at cam dot ac dot uk]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From jv244 at cam dot ac dot uk  2003-08-15 11:51 -------
I've tried with
setenv FC_FLAGS '-g -Wl,--stack,8388608'
../configure ...
These options go into the objdir/Makefile, but not in e.g. 
objdir/gcc/Makefile. As a result xgcc.exe is still linked without these 
options.

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[pinskia at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From pinskia at gcc dot gnu dot org  2003-08-15 18:07 -------
I think this will work:
CFLAGS="-g -Wl,--stack,8388608"
export CFLAGS
pathToGCCsrc/configure configureoptions
make bootstrap
can one of you two try this?

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[jv244 at cam dot ac dot uk]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From jv244 at cam dot ac dot uk  2003-08-15 18:45 -------
Unhappily this is not working.

The flag is added for some of the compiles e.g.

gcc -c -DHAVE_CONFIG_H -g -Wl,--stack,8388608 -I. -I/home/vondele/gcc-
g95/gcc/libiberty/../include  -W -Wall -Wtraditional -
pedantic  /home/vondele/gcc-g95/gcc/libiberty/stpncpy.c -o pic/stpncpy.o

but not for linking xgcc.exe (where I suppose it is mostly needed)

gcc   -g  -DIN_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-
prototypes -pedantic -Wno-long-long -fno-common   -DHAVE_CONFIG_H  -o xgcc.exe 
gcc.o gccspec.o intl.o \
  prefix.o version.o cygwin1.o ../libiberty/libiberty.a -lintl -
liconv ../libbanshee/points-
to/libandersen.a ../libbanshee/engine/libbansheeengine.a ../libbanshee/libcompa
t/libbansheecompat.a

the final error remains the same.
(BTW using these CFLAGS also produces a lot of warnings :
gcc: --stack: linker input file unused because linking not done
gcc: 8388608: linker input file unused because linking not done
)

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[pinskia at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From pinskia at gcc dot gnu dot org  2003-08-15 19:10 -------
What about setting LDFLAGS to "-Wl,--stack,8388608"?

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc

[jv244 at cam dot ac dot uk]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From jv244 at cam dot ac dot uk  2003-08-15 20:03 -------
using LDFLAGS to "-Wl,--stack,8388608" doesn't help either, the option seems 
unused at all (grep 838 log) ...

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc (stacksize too low)

[pinskia at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745


pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2003-08-15 20:06:15
               date|                            |
            Summary|Bootstrap fails when        |Bootstrap fails when
                   |compiling insn-attrtab.c    |compiling insn-attrtab.c
                   |with stage1/xgcc            |with stage1/xgcc (stacksize
                   |                            |too low)


------- Additional Comments From pinskia at gcc dot gnu dot org  2003-08-15 20:06 -------
Thanks for trying I am just going to put this back to new but now we know where the problem is, 
maybe we can work around it in the sources so it can be built without setting the stacksize.

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc (stacksize too low)

[david dot billinghurst at riotinto dot com]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From david dot billinghurst at riotinto dot com  2003-08-20 05:54 -------
Subject: RE:  Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc (stacksize too low)

The problem goes away if I link cc1.exe with -Wl,--stack,8388608, 
which confirms that it is a stack overflow.

I did this by editing gcc/Makefile.

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc (stacksize too low)

[tobias dot schlueter at web dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From tobias dot schlueter at web dot de  2003-09-22 11:32 -------
Subject: Re:  Bootstrap fails when compiling insn-attrtab.c
 with stage1/xgcc (stacksize too low)


FYI, building with --disable-checking is possible. I tried this with 
yesterday's sources.

Running the testsuite right now & will post testresults once available.

- Tobi

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc (stacksize too low)

[pinskia at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From pinskia at gcc dot gnu dot org  2003-09-22 15:28 -------
Some of the speedup that are going on must of fixed this.

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc (stacksize too low)

[Tobias dot Schlueter at physik dot uni-muenchen dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745



------- Additional Comments From Tobias dot Schlueter at physik dot uni-muenchen dot de  2003-11-01 20:57 -------
Subject: Re:  Bootstrap fails when compiling insn-attrtab.c
 with stage1/xgcc (stacksize too low)


fixed by this patch (which is not yet in the tree) 
http://gcc.gnu.org/ml/gcc-patches/2003-11/msg00044.html

- Tobi

[Bug bootstrap/11745] Bootstrap fails when compiling insn-attrtab.c with stage1/xgcc (stacksize too low)

[tobias dot schlueter at web dot de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11745


tobias dot schlueter at web dot de changed:

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


------- Additional Comments From tobias dot schlueter at web dot de  2003-11-02 18:21 -------
Zdenek has comitted his patch. Hence I'm closing this bug.