[racket-dev] racket-5.3.[34] compilation failure SIGSEGV MAPERR si_code 1 fault

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 14-06-13 11:39, Marijn wrote:
> Hi list,
> 
> racket-5.3.3 and 5.3.4 are today both failing to compile for me
> with:
> 
> raco setup:  in r5rs/private raco setup:  in
> typed-racket/optimizer SIGSEGV MAPERR si_code 1 fault on addr
> 0x130
> 
> I get this error using gcc versions 4.5.3, 4.6.3 and 4.7.3. My
> glibc is 2.17, but I have a bug report of a very similar failure[1]
> using glibc-2.15:
> 
> raco setup: 1 making: math/private/vector raco setup: 1 making:
> math/scribblings raco setup: --- updating info-domain tables --- 
> SIGSEGV MAPERR si_code 1 fault on addr 0x40
> 
> which also contains this backtrace:
> 
> [New LWP 9579] [New LWP 9514] [New LWP 9528] [New LWP 9581] [Thread
> debugging using libthread_db enabled] Using host libthread_db
> library "/lib/libthread_db.so.1". Core was generated by 
> `/var/tmp/portage/dev-scheme/racket-5.3.3/work/racket-5.3.3/src/racket/.libs/rac'.
>
> 
Program terminated with signal 6, Aborted.
> #0  0xb76e58d2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2 
> (gdb) bt #0  0xb76e58d2 in _dl_sysinfo_int80 () from
> /lib/ld-linux.so.2 #1  0xb71bf2b1 in *__GI_raise (sig=6) at 
> ../nptl/sysdeps/unix/sysv/linux/raise.c:64 #2  0xb71c0a8e in
> *__GI_abort () at abort.c:91 #3  0xb761f427 in fault_handler ()
> from 
> /var/tmp/portage/dev-scheme/racket-5.3.3/work/racket-5.3.3/src/racket/.libs/libracket3m-5.3.3.so
>
> 
#4  <signal handler called>
> #5  0xa2e9d7a9 in ?? () from /usr/lib/libGL.so.1 #6  0xb7320d3f in
> __nptl_deallocate_tsd () at pthread_create.c:155 #7  0xb7320f15 in
> start_thread (arg=0xb200bb40) at pthread_create.c:312 #8
> 0xb727e64e in clone () at
> ../sysdeps/unix/sysv/linux/i386/clone.S:133 (gdb) q
> 
> 
> I was able to compile 5.3.3 fine in February, although drracket
> won't start (anymore?), but gracket will:
> 
> 
> set-loaded-mask method in bitmap%: contract violation expected:
> (is-a?/c bitmap%) given: #f in: the 1st argument of the
> set-loaded-mask method in bitmap%/c contract from:
> <collects>/racket/draw.rkt blaming: 
> <collects>/drracket/private/frame-icon.rkt at:
> <collects>/racket/draw.rkt:58.19 context...: 
> /usr/lib64/racket/collects/racket/contract/private/blame.rkt:89:0: 
> raise-blame-error 
> /usr/lib64/racket/collects/drracket/private/frame-icon.rkt: 
> [running body] 
> /usr/lib64/racket/collects/drracket/private/drracket-normal.rkt: 
> [traversing imports] 
> /usr/lib64/racket/collects/drracket/drracket.rkt: [running body]
> 
> 
> My cairo is version 1.12.14.
> 
> I tried googling the SIGSEGV MAPERR error and got some other
> mentions of the problem dating back a long time, but no-one seemed
> to have any clue what was going on.
> 
> Marijn
> 
> [1]:https://bugs.gentoo.org/show_bug.cgi?id=471978

Version 5.3.6 seems to have cleared up this problem.

Marijn
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.20 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlITMZIACgkQp/VmCx0OL2xQkwCeO40LLhfM2P43JpGRxA2DKOlh
rbwAnRBYawszidy2b6KWgiForE2g1Kfs
=LHfi
-----END PGP SIGNATURE-----