ALT Linux Team development discussions
 help / color / mirror / Atom feed
* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  @ 2008-11-08 20:31 ` Alexey I. Froloff
  2008-11-08 20:41   ` Dmitry V. Levin
  0 siblings, 1 reply; 11+ messages in thread
From: Alexey I. Froloff @ 2008-11-08 20:31 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 287 bytes --]

* QA Team Robot <qa@> [081103 20:33]:
> Package: mutt1.5-3:1.5.18-alt3
...
> hasher-priv: master: idle time limit (3600 seconds) exceeded
> hsh-rebuild: rebuild of `mutt1.5-1.5.18-alt3.src.rpm' failed.
Это что-то со сборочечницей, или пакет поломался?

-- 
Regards,
Sir Raorn.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-08 20:31 ` [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1] Alexey I. Froloff
@ 2008-11-08 20:41   ` Dmitry V. Levin
  2008-11-08 22:26     ` Alexey I. Froloff
  0 siblings, 1 reply; 11+ messages in thread
From: Dmitry V. Levin @ 2008-11-08 20:41 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 440 bytes --]

On Sat, Nov 08, 2008 at 11:31:56PM +0300, Alexey I. Froloff wrote:
> * QA Team Robot <qa@> [081103 20:33]:
> > Package: mutt1.5-3:1.5.18-alt3
> ...
> > hasher-priv: master: idle time limit (3600 seconds) exceeded
> > hsh-rebuild: rebuild of `mutt1.5-1.5.18-alt3.src.rpm' failed.
> Это что-то со сборочечницей, или пакет поломался?

Это воспроизводимая ошибка, неделю назад сборка закончилась точно таким
же образом.


-- 
ldv

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-08 20:41   ` Dmitry V. Levin
@ 2008-11-08 22:26     ` Alexey I. Froloff
  2008-11-08 22:33       ` Dmitry V. Levin
  0 siblings, 1 reply; 11+ messages in thread
From: Alexey I. Froloff @ 2008-11-08 22:26 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 1292 bytes --]

* Dmitry V. Levin <ldv@> [081108 23:45]:
> > > Package: mutt1.5-3:1.5.18-alt3
> > ...
> > > hasher-priv: master: idle time limit (3600 seconds) exceeded
> > > hsh-rebuild: rebuild of `mutt1.5-1.5.18-alt3.src.rpm' failed.
> > Это что-то со сборочечницей, или пакет поломался?
> Это воспроизводимая ошибка, неделю назад сборка закончилась точно таким
> же образом.
Очень забавная ошибка...

i586-alt-linux-gcc -DPKGDATADIR=\"/usr/share/mutt\" -DSYSCONFDIR=\"/etc\" -DBINDIR=\"/usr/bin\" -DMUTTLOCALEDIR=\"/usr/share/locale\" -DHAVE_CONFIG_H=1 -I.  -I. -I. -I./imap  -Iintl -I/usr/include/krb5 -I/usr/include/et -I/include -I./intl -DMD5UTIL -Wall -pedantic -pipe -Wall -O2 -march=i586 -mtune=i686 -std=c99 -c -o mutt_md5-md5.o `test -f 'md5.c' || echo './'`md5.c
i586-alt-linux-gcc -DMD5UTIL -Wall -pedantic -pipe -Wall -O2 -march=i586 -mtune=i686 -std=c99  -L/lib -o mutt_md5 mutt_md5-md5.o  -lidn 

И в результате получаем нерабочий mutt_md5.  Ессно сборка с
--enable debug проходит нормально.  Работать начинает при
отключении опций от -fno-tree-ccp до -fno-tree-vrp включительно
(список опций получил так:
i586-alt-linux-gcc -Wall -pedantic -Wall -O2 -march=i586 -mtune=i686 -std=c99 -c -Q --help=optimizers | grep enabled
).

Куда копать?

-- 
Regards,
Sir Raorn.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-08 22:26     ` Alexey I. Froloff
@ 2008-11-08 22:33       ` Dmitry V. Levin
  2008-11-09 11:31         ` Alexey I. Froloff
  0 siblings, 1 reply; 11+ messages in thread
From: Dmitry V. Levin @ 2008-11-08 22:33 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 1080 bytes --]

On Sun, Nov 09, 2008 at 01:26:34AM +0300, Alexey I. Froloff wrote:
[...]
> Очень забавная ошибка...
> 
> i586-alt-linux-gcc -DPKGDATADIR=\"/usr/share/mutt\" -DSYSCONFDIR=\"/etc\" -DBINDIR=\"/usr/bin\" -DMUTTLOCALEDIR=\"/usr/share/locale\" -DHAVE_CONFIG_H=1 -I.  -I. -I. -I./imap  -Iintl -I/usr/include/krb5 -I/usr/include/et -I/include -I./intl -DMD5UTIL -Wall -pedantic -pipe -Wall -O2 -march=i586 -mtune=i686 -std=c99 -c -o mutt_md5-md5.o `test -f 'md5.c' || echo './'`md5.c
> i586-alt-linux-gcc -DMD5UTIL -Wall -pedantic -pipe -Wall -O2 -march=i586 -mtune=i686 -std=c99  -L/lib -o mutt_md5 mutt_md5-md5.o  -lidn 
> 
> И в результате получаем нерабочий mutt_md5.  Ессно сборка с
> --enable debug проходит нормально.  Работать начинает при
> отключении опций от -fno-tree-ccp до -fno-tree-vrp включительно
> (список опций получил так:
> i586-alt-linux-gcc -Wall -pedantic -Wall -O2 -march=i586 -mtune=i686 -std=c99 -c -Q --help=optimizers | grep enabled
> ).
> 
> Куда копать?

Если есть подозрение на miscompile, то надо минимизировать test case.


-- 
ldv

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-08 22:33       ` Dmitry V. Levin
@ 2008-11-09 11:31         ` Alexey I. Froloff
  2008-11-09 12:09           ` Dmitry V. Levin
  0 siblings, 1 reply; 11+ messages in thread
From: Alexey I. Froloff @ 2008-11-09 11:31 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 644 bytes --]

* Dmitry V. Levin <ldv@> [081109 01:39]:
> > И в результате получаем нерабочий mutt_md5.
> Если есть подозрение на miscompile, то надо минимизировать test case.
Смиялсо.

$ git log --pretty=oneline --abbrev-commit md5.c
7482ae1... Build local md5 tool for hcachever.sh. Closes #3025.
f9b5f95... Replace RFC md5 implementation with GPL version from coreutils

И там действительно копия lib/md5.c из gnulib, версия от Author:
Simon Josefsson <simon@josefsson.org>  2005-10-17 16:55:00,
commit id 8523c7381469c7279a26fd88f4f3be20d43c4e1d .  Обновление
до последней версии (ldv/coreutils.git) не помогает.

-- 
Regards,
Sir Raorn.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-09 11:31         ` Alexey I. Froloff
@ 2008-11-09 12:09           ` Dmitry V. Levin
  2008-11-09 13:04             ` Alexey I. Froloff
  0 siblings, 1 reply; 11+ messages in thread
From: Dmitry V. Levin @ 2008-11-09 12:09 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 896 bytes --]

On Sun, Nov 09, 2008 at 02:31:31PM +0300, Alexey I. Froloff wrote:
> * Dmitry V. Levin <ldv@> [081109 01:39]:
> > > И в результате получаем нерабочий mutt_md5.
> > Если есть подозрение на miscompile, то надо минимизировать test case.
> Смиялсо.
> 
> $ git log --pretty=oneline --abbrev-commit md5.c
> 7482ae1... Build local md5 tool for hcachever.sh. Closes #3025.
> f9b5f95... Replace RFC md5 implementation with GPL version from coreutils
> 
> И там действительно копия lib/md5.c из gnulib, версия от Author:
> Simon Josefsson <simon@josefsson.org>  2005-10-17 16:55:00,
> commit id 8523c7381469c7279a26fd88f4f3be20d43c4e1d .  Обновление
> до последней версии (ldv/coreutils.git) не помогает.

Известно, что lib/md5.c из gnulib в сборке coreutils вполне рабочее, по
крайней мере coreutils testsuite проходит.

Я пока не улавливаю суть происходящего с mutt_md5.


-- 
ldv

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-09 12:09           ` Dmitry V. Levin
@ 2008-11-09 13:04             ` Alexey I. Froloff
  2008-11-09 13:38               ` Dmitry V. Levin
  0 siblings, 1 reply; 11+ messages in thread
From: Alexey I. Froloff @ 2008-11-09 13:04 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 5581 bytes --]

* Dmitry V. Levin <ldv@> [081109 15:10]:
> Я пока не улавливаю суть происходящего с mutt_md5.
Код (md5_process_bytes):

  /* Move remaining bytes in internal buffer.  */
  if (len > 0)
    {
      size_t left_over = ctx->buflen;

      memcpy (&((char *) ctx->buffer)[left_over], buffer, len);
      left_over += len;
      if (left_over >= 64)
	{
	  md5_process_block (ctx->buffer, 64, ctx);
	  left_over -= 64;
	  memcpy (ctx->buffer, &ctx->buffer[16], left_over);
	}
      ctx->buflen = left_over;
    }

Соответственно затык происходит на первом memcpy.

(gdb) run
Starting program: /home/raorn/src/gitsnaps/packages/mutt1.5/work/mutt-1.5.18/mutt_md5 
foobar

Breakpoint 1, md5_process_bytes (buffer=0xbfd62df4, len=7, ctx=0xbfd63e3c) at md5.c:223
223       if (ctx->buflen != 0)
(gdb) nexti
220     {
(gdb) 
0x080497ac      220     {
(gdb) 
0x080497af      220     {
(gdb) 
0x080497b2      220     {
(gdb) 
223       if (ctx->buflen != 0)
(gdb) 
0x080497b8      223       if (ctx->buflen != 0)
(gdb) 
0x080497ba      223       if (ctx->buflen != 0)
(gdb) 
247       if (len >= 64)
(gdb) 
0x080497bf      247       if (len >= 64)
(gdb) 
269       if (len > 0)
(gdb) 
0x080497c3      269       if (len > 0)
(gdb) 
271           size_t left_over = ctx->buflen;
(gdb) 
52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
(gdb) 
271           size_t left_over = ctx->buflen;
(gdb) 
52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
(gdb) 
0x080497e5      52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
(gdb) 
0x080497e9      52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
(gdb) 
274           left_over += len;
(gdb) 
52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
(gdb) 
0x080497f2      52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
(gdb) 

^C
Program received signal SIGINT, Interrupt.
memcpy (__dest=0xbfd62df4, __src=0x7, __len=3218488892) at /usr/include/bits/string3.h:52
52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));

Здесь __dest == buffer, __src == len, __len == ctx, т.е. то
самое, что передали в md5_process_bytes.

Результат (неполный) disasssemble:

0x080497a0 <md5_process_bytes+0>:       push   %ebp
0x080497a1 <md5_process_bytes+1>:       mov    %esp,%ebp
0x080497a3 <md5_process_bytes+3>:       sub    $0x28,%esp
0x080497a6 <md5_process_bytes+6>:       mov    0x10(%ebp),%eax
0x080497a9 <md5_process_bytes+9>:       mov    %esi,0xfffffff8(%ebp)
0x080497ac <md5_process_bytes+12>:      mov    0xc(%ebp),%esi
0x080497af <md5_process_bytes+15>:      mov    %ebx,0xfffffff4(%ebp)
0x080497b2 <md5_process_bytes+18>:      mov    %edi,0xfffffffc(%ebp)
0x080497b5 <md5_process_bytes+21>:      mov    0x18(%eax),%edi
0x080497b8 <md5_process_bytes+24>:      test   %edi,%edi
0x080497ba <md5_process_bytes+26>:      jne    0x8049830 <md5_process_bytes+144>
0x080497bc <md5_process_bytes+28>:      cmp    $0x3f,%esi
0x080497bf <md5_process_bytes+31>:      ja     0x8049808 <md5_process_bytes+104>
0x080497c1 <md5_process_bytes+33>:      test   %esi,%esi
0x080497c3 <md5_process_bytes+35>:      jne    0x80497d8 <md5_process_bytes+56>
0x080497c5 <md5_process_bytes+37>:      mov    0xfffffff4(%ebp),%ebx
0x080497c8 <md5_process_bytes+40>:      mov    0xfffffff8(%ebp),%esi
0x080497cb <md5_process_bytes+43>:      mov    0xfffffffc(%ebp),%edi
0x080497ce <md5_process_bytes+46>:      mov    %ebp,%esp
0x080497d0 <md5_process_bytes+48>:      pop    %ebp
0x080497d1 <md5_process_bytes+49>:      ret    
0x080497d2 <md5_process_bytes+50>:      lea    0x0(%esi),%esi
0x080497d8 <md5_process_bytes+56>:      mov    0x10(%ebp),%eax
0x080497db <md5_process_bytes+59>:      mov    0x8(%ebp),%edx
0x080497de <md5_process_bytes+62>:      mov    0x18(%eax),%ebx
0x080497e1 <md5_process_bytes+65>:      mov    %esi,0x8(%esp)
0x080497e5 <md5_process_bytes+69>:      mov    %edx,0x4(%esp)
0x080497e9 <md5_process_bytes+73>:      lea    0x1c(%eax,%ebx,1),%eax
0x080497ed <md5_process_bytes+77>:      add    %ebx,%esi
0x080497ef <md5_process_bytes+79>:      mov    %eax,(%esp)
0x080497f2 <md5_process_bytes+82>:      call   0x8049650 <memcpy>
0x080497f7 <md5_process_bytes+87>:      cmp    $0x3f,%esi

А вот во что компилится вышеприведённый кусок кода без -O2:

0x08048a40 <md5_process_bytes+378>:     mov    0x10(%ebp),%eax
0x08048a43 <md5_process_bytes+381>:     mov    0x18(%eax),%eax
0x08048a46 <md5_process_bytes+384>:     mov    %eax,0xfffffff0(%ebp)
0x08048a49 <md5_process_bytes+387>:     mov    0x10(%ebp),%eax
0x08048a4c <md5_process_bytes+390>:     add    $0x1c,%eax
0x08048a4f <md5_process_bytes+393>:     mov    %eax,%edx
0x08048a51 <md5_process_bytes+395>:     add    0xfffffff0(%ebp),%edx
0x08048a54 <md5_process_bytes+398>:     mov    0xc(%ebp),%eax
0x08048a57 <md5_process_bytes+401>:     mov    %eax,0x8(%esp)
0x08048a5b <md5_process_bytes+405>:     mov    0x8(%ebp),%eax
0x08048a5e <md5_process_bytes+408>:     mov    %eax,0x4(%esp)
0x08048a62 <md5_process_bytes+412>:     mov    %edx,(%esp)
0x08048a65 <md5_process_bytes+415>:     call   0x80483fc <memcpy@plt>
0x08048a6a <md5_process_bytes+420>:     mov    0xc(%ebp),%eax
0x08048a6d <md5_process_bytes+423>:     add    %eax,0xfffffff0(%ebp)
0x08048a70 <md5_process_bytes+426>:     cmpl   $0x3f,0xfffffff0(%ebp)

Куда копать? (ц)

-- 
Regards,
Sir Raorn.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-09 13:04             ` Alexey I. Froloff
@ 2008-11-09 13:38               ` Dmitry V. Levin
  2008-11-09 13:54                 ` Alexey I. Froloff
  0 siblings, 1 reply; 11+ messages in thread
From: Dmitry V. Levin @ 2008-11-09 13:38 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 2499 bytes --]

On Sun, Nov 09, 2008 at 04:04:11PM +0300, Alexey I. Froloff wrote:
> * Dmitry V. Levin <ldv@> [081109 15:10]:
> > Я пока не улавливаю суть происходящего с mutt_md5.
> Код (md5_process_bytes):
> 
>   /* Move remaining bytes in internal buffer.  */
>   if (len > 0)
>     {
>       size_t left_over = ctx->buflen;
> 
>       memcpy (&((char *) ctx->buffer)[left_over], buffer, len);
>       left_over += len;
>       if (left_over >= 64)
> 	{
> 	  md5_process_block (ctx->buffer, 64, ctx);
> 	  left_over -= 64;
> 	  memcpy (ctx->buffer, &ctx->buffer[16], left_over);
> 	}
>       ctx->buflen = left_over;
>     }
> 
> Соответственно затык происходит на первом memcpy.
> 
> (gdb) run
> Starting program: /home/raorn/src/gitsnaps/packages/mutt1.5/work/mutt-1.5.18/mutt_md5 
> foobar
> 
> Breakpoint 1, md5_process_bytes (buffer=0xbfd62df4, len=7, ctx=0xbfd63e3c) at md5.c:223
> 223       if (ctx->buflen != 0)
> (gdb) nexti
> 220     {
> (gdb) 
> 0x080497ac      220     {
> (gdb) 
> 0x080497af      220     {
> (gdb) 
> 0x080497b2      220     {
> (gdb) 
> 223       if (ctx->buflen != 0)
> (gdb) 
> 0x080497b8      223       if (ctx->buflen != 0)
> (gdb) 
> 0x080497ba      223       if (ctx->buflen != 0)
> (gdb) 
> 247       if (len >= 64)
> (gdb) 
> 0x080497bf      247       if (len >= 64)
> (gdb) 
> 269       if (len > 0)
> (gdb) 
> 0x080497c3      269       if (len > 0)
> (gdb) 
> 271           size_t left_over = ctx->buflen;
> (gdb) 
> 52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
> (gdb) 
> 271           size_t left_over = ctx->buflen;
> (gdb) 
> 52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
> (gdb) 
> 0x080497e5      52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
> (gdb) 
> 0x080497e9      52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
> (gdb) 
> 274           left_over += len;
> (gdb) 
> 52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
> (gdb) 
> 0x080497f2      52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
> (gdb) 
> 
> ^C
> Program received signal SIGINT, Interrupt.
> memcpy (__dest=0xbfd62df4, __src=0x7, __len=3218488892) at /usr/include/bits/string3.h:52
> 52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));

__len=3218488892?  Не многовато ли?


-- 
ldv

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-09 13:38               ` Dmitry V. Levin
@ 2008-11-09 13:54                 ` Alexey I. Froloff
  2008-11-09 21:15                   ` Dmitry V. Levin
  0 siblings, 1 reply; 11+ messages in thread
From: Alexey I. Froloff @ 2008-11-09 13:54 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 492 bytes --]

* Dmitry V. Levin <ldv@> [081109 16:48]:
> > Program received signal SIGINT, Interrupt.
> > memcpy (__dest=0xbfd62df4, __src=0x7, __len=3218488892) at /usr/include/bits/string3.h:52
> > 52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
> __len=3218488892?  Не многовато ли?
Да и __src=0x7 тоже не должен просто так проскакивать.  Это те
параметры, которые передаются в md5_process_bytes().  Я не знаю
как это отлаживать :-(

-- 
Regards,
Sir Raorn.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-09 13:54                 ` Alexey I. Froloff
@ 2008-11-09 21:15                   ` Dmitry V. Levin
  2008-11-09 22:27                     ` Alexey I. Froloff
  0 siblings, 1 reply; 11+ messages in thread
From: Dmitry V. Levin @ 2008-11-09 21:15 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 1206 bytes --]

On Sun, Nov 09, 2008 at 04:54:32PM +0300, Alexey I. Froloff wrote:
> * Dmitry V. Levin <ldv@> [081109 16:48]:
> > > Program received signal SIGINT, Interrupt.
> > > memcpy (__dest=0xbfd62df4, __src=0x7, __len=3218488892) at /usr/include/bits/string3.h:52
> > > 52        return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
> > __len=3218488892?  Не многовато ли?
> Да и __src=0x7 тоже не должен просто так проскакивать.  Это те
> параметры, которые передаются в md5_process_bytes().  Я не знаю
> как это отлаживать :-(

В данном случае проще всего сравнить вывод gcc -E у обычного string.h и
этого md5.c в окрестности определения memcpy.

Нижеприведённое изменение исправляет mutt_md5.  Очевидно, в файле md5.h
сидит гадость (отключение __attribute__), которая искажает string.h;
стоит пожаловаться upstream.
В gnulib'овском md5.h отключения __attribute__ уже нет.
Имеет смысл попробовать обновить md5.c _и_ md5.h из gnulib.

--- md5.c
+++ md5.c
@@ -24,11 +24,11 @@
 # include <config.h>
 #endif
 
-#include "md5.h"
-
 #include <stddef.h>
 #include <string.h>
 
+#include "md5.h"
+
 #if USE_UNLOCKED_IO
 # include "unlocked-io.h"
 #endif


-- 
ldv

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1]
  2008-11-09 21:15                   ` Dmitry V. Levin
@ 2008-11-09 22:27                     ` Alexey I. Froloff
  0 siblings, 0 replies; 11+ messages in thread
From: Alexey I. Froloff @ 2008-11-09 22:27 UTC (permalink / raw)
  To: ALT Devel discussion list

[-- Attachment #1: Type: text/plain, Size: 195 bytes --]

* Dmitry V. Levin <ldv@> [081110 00:25]:
> Имеет смысл попробовать обновить md5.c _и_ md5.h из gnulib.
Ну в общем http://dev.mutt.org/trac/ticket/3133 .  Спасибо.

-- 
Regards,
Sir Raorn.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

^ permalink raw reply	[flat|nested] 11+ messages in thread

end of thread, other threads:[~2008-11-09 22:27 UTC | newest]

Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-11-08 20:31 ` [devel] mutt1.5-3:1.5.18-alt3: i586 rebuild failed [1] Alexey I. Froloff
2008-11-08 20:41   ` Dmitry V. Levin
2008-11-08 22:26     ` Alexey I. Froloff
2008-11-08 22:33       ` Dmitry V. Levin
2008-11-09 11:31         ` Alexey I. Froloff
2008-11-09 12:09           ` Dmitry V. Levin
2008-11-09 13:04             ` Alexey I. Froloff
2008-11-09 13:38               ` Dmitry V. Levin
2008-11-09 13:54                 ` Alexey I. Froloff
2008-11-09 21:15                   ` Dmitry V. Levin
2008-11-09 22:27                     ` Alexey I. Froloff

ALT Linux Team development discussions

This inbox may be cloned and mirrored by anyone:

	git clone --mirror http://lore.altlinux.org/devel/0 devel/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 devel devel/ http://lore.altlinux.org/devel \
		devel@altlinux.org devel@altlinux.ru devel@lists.altlinux.org devel@lists.altlinux.ru devel@linux.iplabs.ru mandrake-russian@linuxteam.iplabs.ru sisyphus@linuxteam.iplabs.ru
	public-inbox-index devel

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://lore.altlinux.org/org.altlinux.lists.devel


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git