headers
Eric Blake | 3 Oct 2009 02:03
Gravatar

Re: make check failure on Debian GNU/Hurd


According to Hiroyuki Yamamoto on 9/23/2009 8:26 AM:
>> the c-stack module needs to be fixed
>> for GNU/Hurd.  Meanwhile, could you help debug this further?  I don't have
>> access to GNU/Hurd to reproduce this.
> 
> I do anything if possible, though there is not confidence.

I now have access to GNU/Hurd (thanks to the people at
http://www.gnu.org/software/hurd/public_hurd_boxen.html), so I hope to
look into it.

> 
> Then I tried making with libsigsegv library (Version 2.5-3 of Debian GNU/Hurd package).

There have been some bug fixes since, which might impact the Hurd.
Perhaps using libsigsegv 2.7 might help?

--
Don't work too hard, make some time for fun as well!

Eric Blake             ebb9 <at> byu.net
Permalink | Reply | 0 comments |
headers
Tim Wood | 6 Oct 2009 19:40

Possible m4-1.4.13 issue

TWIMC:

Platform info:
Linux axolotl.home.tim.wood.name 2.4.20-30.7.legacy #1 Fri Feb 20 14:31:41 PST 2004 i586 unknown
gcc --version: 2.96
make --version: GNU Make version 3.79.1, by Richard Stallman and Roland McGrath.
Built for i386-redhat-linux-gnu

I ran into this test failure building m4-1.4.13:
---
Skipping test: ungetc cannot handle arbitrary bytes
SKIP: test-fseeko2.sh
test-ftello.c:119: assertion failed
./test-ftello.sh: line 3: 16223 Aborted                 ./test-ftello${EXEEXT} 1 <"$srcdir/test-ftello.sh"
FAIL: test-ftello.sh
Skipping test: ungetc cannot handle arbitrary bytes
SKIP: test-ftello2.sh
---
The offending lines in test-ftello.c:
     114 #if !defined __MINT__ /* FreeMiNT has problems seeking past end of file         */
     115   /* Test ftell beyond end of file.  */
     116   ASSERT (fseek (stdin, 0, SEEK_END) == 0);
     117   ch = ftello (stdin);
     118   ASSERT (fseek (stdin, 10, SEEK_END) == 0);
     119   ASSERT (ftell (stdin) == ch + 10);
     120   ASSERT (ftello (stdin) == ch + 10);
     121 #endif
---
test-ftello executable invoked by test-ftello.sh:
       1 #!/bin/sh
(Continue reading)

Permalink | Reply | 1 comment |
headers
Eric Blake | 7 Oct 2009 02:45
Gravatar

Re: Possible m4-1.4.13 issue


[adding bug-gnulib, since the gnulib test-ftello test failed]

According to Tim Wood on 10/6/2009 11:40 AM:
> TWIMC:
> 
> Platform info:
> Linux axolotl.home.tim.wood.name 2.4.20-30.7.legacy #1 Fri Feb 20

> test-ftello.c:119: assertion failed
> ./test-ftello.sh: line 3: 16223 Aborted                
> ./test-ftello${EXEEXT} 1 <"$srcdir/test-ftello.sh"
> FAIL: test-ftello.sh

Thanks for the report.  This has been previously reported; it is a known
bug in ungetc() in a VERY old version of glibc.  No one has bothered to
update the testsuite to ignore the bug, because most machines are using a
fixed glibc by now.  It does not impact m4's behavior, since m4 does not
use ungetc in a manner that tickles the bug.

--
Don't work too hard, make some time for fun as well!

Eric Blake             ebb9 <at> byu.net
Permalink | Reply | 0 comments |
headers
Jorma Karvonen | 16 Oct 2009 20:07
Picon

Present-But-Cannot-Be-Compiled bug

Hello,

I think that since version 1.4.9 there has been so called
Present-But-Cannot-Be-Compiled bug in m4 source code:

http://www.gnu.org/software/autoconf/manual/html_node/Present-But-Cannot-Be-Compiled.html

In my development environment I have autoconf 2.64 and error messages
are complaining that autoconf 2.61 is needed.

Best regards,

Jorma Karvonen
Permalink | Reply | 4 comments |
headers
Eric Blake | 16 Oct 2009 20:25
Gravatar

Re: Present-But-Cannot-Be-Compiled bug


According to Jorma Karvonen on 10/16/2009 12:07 PM:
> Hello,
> 
> I think that since version 1.4.9 there has been so called
> Present-But-Cannot-Be-Compiled bug in m4 source code:
> 
> http://www.gnu.org/software/autoconf/manual/html_node/Present-But-Cannot-Be-Compiled.html
> 
> In my development environment I have autoconf 2.64 and error messages
> are complaining that autoconf 2.61 is needed.

Thanks for the report.  However, without any details, such as the command
you typed, what system you are on, and the actual error message displayed
on the console as well as config.log detailing the failure, there is
nothing I can do about it.

--
Don't work too hard, make some time for fun as well!

Eric Blake             ebb9 <at> byu.net
Permalink | Reply | 3 comments |
headers
Jorma Karvonen | 16 Oct 2009 20:59
Picon

Re: Present-But-Cannot-Be-Compiled bug

Hello,

here are my commands:

cd m4-1.4.13/
mkdir target
cd target/
../configure --prefix=/usr/local
...
checking whether gettimeofday clobbers localtime buffer... yes
checking sys/timeb.h usability... no
checking sys/timeb.h presence... yes
configure: WARNING: sys/timeb.h: present but cannot be compiled
configure: WARNING: sys/timeb.h:     check for missing prerequisite headers?
configure: WARNING: sys/timeb.h: see the Autoconf documentation
configure: WARNING: sys/timeb.h:     section "Present But Cannot Be Compiled"
configure: WARNING: sys/timeb.h: proceeding with the compiler's result
configure: WARNING:     ## ----------------------------- ##
configure: WARNING:     ## Report this to bug-m4 <at> gnu.org ##
configure: WARNING:     ## ----------------------------- ##
checking for sys/timeb.h... no
checking for _ftime... no
checking whether the compiler generally respects inline... yes
checking whether isnan(double) can be used without linking with libm... no
checking where to find the exponent in a 'double'... unknown
checking whether isnan(float) can be used without linking with libm... no
checking where to find the exponent in a 'float'... (cached) word 0 bit 23
checking whether isnan(long double) can be used without linking with libm... no
checking where to find the exponent in a 'long double'... unknown
checking for libsigsegv... (cached) no, consider installing GNU libsigsegv
(Continue reading)

Permalink | Reply | 2 comments |
headers
Eric Blake | 17 Oct 2009 00:45
Gravatar

Re: m4-1.4.13 config.log


Please keep the list in the loop.

According to Jorma Karvonen on 10/16/2009 1:02 PM:
> Hello,
> 
> here is my config.log.

You failed to mention your platform, but config.log states:

> hostname = 2(SERVFAIL)
> uname -m = i686
> uname -r = 2.6.18.8-0.13-default
> uname -s = Linux
> uname -v = #1 SMP Thu Nov 6 13:35:13 UTC 2008

> configure:3792: gcc --version >&5
> gcc (GCC) 4.3.4

It would also be nice to see the condensed output of running ./configure;
config.log has everything, but it doesn't make it easy to pinpoint where
you are first seeing unexpected failures.

Meanwhile, this is the first suspicious message in the log:

> /tmp/ccJTzTZJ.s: Assembler messages:
> /tmp/ccJTzTZJ.s:2287: Error: symbol `fstatat64' is already defined
> /tmp/ccJTzTZJ.s:2316: Error: symbol `fstat64' is already defined
> /tmp/ccJTzTZJ.s:2341: Error: symbol `lstat64' is already defined
> /tmp/ccJTzTZJ.s:2366: Error: symbol `stat64' is already defined
(Continue reading)

Permalink | Reply | 1 comment |
headers
Eric Blake | 17 Oct 2009 01:03
Gravatar

Re: Present-But-Cannot-Be-Compiled bug


Please don't top-post on technical lists.

According to Jorma Karvonen on 10/16/2009 12:59 PM:
> checking whether gettimeofday clobbers localtime buffer... yes
> checking sys/timeb.h usability... no
> checking sys/timeb.h presence... yes
> configure: WARNING: sys/timeb.h: present but cannot be compiled
> configure: WARNING: sys/timeb.h:     check for missing prerequisite headers?
> configure: WARNING: sys/timeb.h: see the Autoconf documentation
> configure: WARNING: sys/timeb.h:     section "Present But Cannot Be Compiled"
> configure: WARNING: sys/timeb.h: proceeding with the compiler's result

Looking more at your config.log, I see:

> configure:14781: checking sys/timeb.h usability
> configure:14781: gcc -std=gnu99 -c -g -O2  conftest.c >&5
> In file included from /usr/include/stdio.h:34,
>                  from conftest.c:136:
> /usr/local/lib/gcc/i686-pc-linux-gnu/4.3.4/include/stddef.h:214: error: duplicate 'unsigned'
> /usr/local/lib/gcc/i686-pc-linux-gnu/4.3.4/include/stddef.h:214: error: two or more data types
in declaration specifiers
> In file included from /usr/include/inttypes.h:28,
>                  from conftest.c:161:
> /usr/include/stdint.h:139: error: 'long long long' is too long for GCC
> /usr/include/stdint.h:139: error: 'long long long' is too long for GCC

What is at line 214 of that stddef.h file, and line 139 of stdint.h?  It
might be a macro being previously (but incorrectly) redefined by
configure?  This may also be a case where your system headers came from a
(Continue reading)

Permalink | Reply | 1 comment |
headers
Jorma Karvonen | 17 Oct 2009 01:11
Picon

Re: m4-1.4.13 config.log

2009/10/17 Eric Blake <ebb9 <at> byu.net>

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Please keep the list in the loop.
>
> According to Jorma Karvonen on 10/16/2009 1:02 PM:
> > Hello,
> >
> > here is my config.log.
>
> You failed to mention your platform, but config.log states:
>
> > hostname = 2(SERVFAIL)
> > uname -m = i686
> > uname -r = 2.6.18.8-0.13-default
> > uname -s = Linux
> > uname -v = #1 SMP Thu Nov 6 13:35:13 UTC 2008
>
> > configure:3792: gcc --version >&5
> > gcc (GCC) 4.3.4
>

Actually, I told those things at the end of previous message:

/usr/include/stdint.h:139: error: ”long long long” is too long for GCC
 /usr/include/stdint.h:139: error: ”long long long” is too long for GCC
 In file included from ../../lib/gl_avltree_oset.c:23:
 ./stdlib.h:57: error: redefinition of ”struct random_data”
(Continue reading)

Permalink | Reply | 0 comments |
headers
Jorma Karvonen | 17 Oct 2009 01:20
Picon

Re: Present-But-Cannot-Be-Compiled bug

2009/10/17, Eric Blake <ebb9 <at> byu.net>:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Please don't top-post on technical lists.
>
> According to Jorma Karvonen on 10/16/2009 12:59 PM:
>> checking whether gettimeofday clobbers localtime buffer... yes
>> checking sys/timeb.h usability... no
>> checking sys/timeb.h presence... yes
>> configure: WARNING: sys/timeb.h: present but cannot be compiled
>> configure: WARNING: sys/timeb.h:     check for missing prerequisite
>> headers?
>> configure: WARNING: sys/timeb.h: see the Autoconf documentation
>> configure: WARNING: sys/timeb.h:     section "Present But Cannot Be
>> Compiled"
>> configure: WARNING: sys/timeb.h: proceeding with the compiler's result
>
> Looking more at your config.log, I see:
>
>> configure:14781: checking sys/timeb.h usability
>> configure:14781: gcc -std=gnu99 -c -g -O2  conftest.c >&5
>> In file included from /usr/include/stdio.h:34,
>>                  from conftest.c:136:
>> /usr/local/lib/gcc/i686-pc-linux-gnu/4.3.4/include/stddef.h:214: error:
>> duplicate 'unsigned'
>> /usr/local/lib/gcc/i686-pc-linux-gnu/4.3.4/include/stddef.h:214: error:
>> two or more data types in declaration specifiers
>> In file included from /usr/include/inttypes.h:28,
>>                  from conftest.c:161:
(Continue reading)

Permalink | Reply | 0 comments |

Gmane