On 06/03/2013 06:44 PM, Albert Chin wrote:
> What's the best way to fix this?
Can you find out why _OSF_SOURCE isn't defined for you?
http://h30097.www3.hp.com/docs/base_doc/DOCUMENTATION/V51_HTML/MAN/MAN5/0001.HTM
says that _OSF_SOURCE should be on by default.
If we can figure out why it's
With latest gnulib:
$ uname -a
OSF1 gen V5.1 732 alpha
$ ./gnulib-tool --test --with-tests fwriting
...
source='../../gllib/freading.c' object='freading.o' libtool=no DEPDIR=.deps
depmode=tru64 /bin/ksh ../../build-aux/depcomp cc -DHAVE_CONFIG_H -I.
-I../../gllib -I.. -DGNULIB_STRICT_
> 2012-06-24 Bruno Haible
>
> ptsname_r: Make it consistent with ptsname on OSF/1.
> * lib/ptsname_r.c (__ptsname_r): Add a different implementation for
> OSF/1.
Nearly the same code works also on AIX. Unfortunately, it does not fix
the problem that clo
On OSF/1, gnulib's ptsname_r function returns a file name like
"/dev/ttyp5". But ptsname() returns something like "/dev/pts/5".
How does it do this? 'truss' reveals it:
ioctl(3, 0x20007447, 0x)= 7340037 = 0x75
stat("/dev/pts/
On OSF/1, I'm seeing this test failure in an *interactive* "make check" of
a ptsname_r testdir:
test-ptsname_r.c:193: assertion failed
FAIL: test-ptsname_r
It turns out the cause is a bug in gnulib's ttyname_r: It fails to add a
trailing NUL byte to the output string. T
[adding bug-gnulib]
How about having compare "know" about /dev/null.
Then it can perform the test -s and warn if the file is nonempty.
With that, all existing (and there are many) /dev/null-using
compare uses will benefit.
>>
>> Nice idea.
> Yes. In fact, c-set was against gn
on failed
FAIL: test-ptsname_r
Cause: ptsname_r fails with errno = ENOENT.
2011-11-10 Bruno Haible
ptsname_r: Avoid compilation error on OSF/1 5.1.
* lib/stdlib.in.h (ptsname_r): Override if REPLACE_PTSNAME_R is 1.
* m4/stdlib_h.m4 (gl_STDLIB_H_DEFAULTS): Initialize REPLACE_PTSNAME_
grep pthread_join
__pthread_join | 0004395905075472 | T | 0008
This fixes it: Ensure that S["LIB_PTHREAD"]="-lpthread" on this platform.
2011-09-07 Bruno Haible
pthread: Determine $(LIB_PTHREAD) correctly on OSF/1 5.1.
* m4/pthread.m4 (gl_PTHREAD_CHECK): Use
Paul Eggert wrote:
> On 09/07/11 02:49, Bruno Haible wrote:
> > -#include
> > +/* Write "fcntl.h" here, not , otherwise OSF/1 5.1 DTK cc
> > miscompiles
> > + openat.c because that file has a preliminary #include . */
> > +#include "fcntl.h"
>
> Thanks. Is a similar change needed to lib/open
On 09/07/11 02:49, Bruno Haible wrote:
> -#include
> +/* Write "fcntl.h" here, not , otherwise OSF/1 5.1 DTK cc
> miscompiles
> + openat.c because that file has a preliminary #include . */
> +#include "fcntl.h"
Thanks. Is a similar change needed to lib/open.c,
which plays the same trick with
Bruno Haible wrote:
> Hi Jim, Paul,
>
>> I hope to be able to make the coreutils-8.13 release on Wednesday.
>
> There is a regression: A new compilation error on OSF/1 5.1.
>
> CC openat.o
> cc: Error: openat.h, line 59: In this statement, "AT_SYML
Hi Jim, Paul,
> I hope to be able to make the coreutils-8.13 release on Wednesday.
There is a regression: A new compilation error on OSF/1 5.1.
CC openat.o
cc: Error: openat.h, line 59: In this statement, "AT_SYMLINK_NOFOLLOW" is not
declared. (undeclared)
return fchow
> > On OSF/1 5.1:
> > Build error.
The next build error is:
$ gmake V=1
...
cc -O -o df df.o find-mount-point.o libver.a ../lib/libcoreutils.a
../lib/libcoreutils.a /home/haible/prefix-osf51-cc/lib/libiconv.so -rpath
/home/haible/prefix-osf51-cc/lib
ld:
Unresolved:
strto
Tom G. Christensen wrote:
> With this change there is still a link failure on CentOS 5.6:
> gcc -std=gnu99 -g -O2 -Wall -o test-spawn-pipe-child
> test-spawn-pipe-child.o -lm -lm -lm -lm -lm -lm -lm -lm -lm -lm -lm
> test-spawn-pipe-child.o: In function `main':
> /home/tgc/projects/gnulib/buil
On 06/17/2011 01:45 PM, Bruno Haible wrote:
On 2011-06-06 I wrote:
spawn-pipe tests: Like the child program only against libc.
2011-06-17 Bruno Haible
spawn-pipe tests: Fix link error.
* tests/test-spawn-pipe-child.c: Undefine fflush.
With this change there is still
On 2011-06-06 I wrote:
> spawn-pipe tests: Like the child program only against libc.
One link failure is still present:
gcc -g -O2 -o test-spawn-pipe-child test-spawn-pipe-child.o
test-spawn-pipe-child.o: In function `main':
/tmp/testdir/gltests/test-spawn-pipe-child.c:86: undefined re
I wrote:
> This failure is probably due to the many libraries against which the child
> program is linked:
>
> $ ldd gettext-0.18.2/gettext-tools/gnulib-tests/.libs/*test-spawn-pipe*
>
> Main =>
> gettext-0.18.2/gettext-tools/gnulib-tests/.libs/lt-test-spawn-pipe
> libiconv.so
I Bruno. Hope you doon't mind if I chime in just to signal ...
On Monday 06 June 2011, Bruno Haible wrote:
>
> 2011-06-06 Bruno Haible
>
> spawn-pipe tests: Like the child program only against libc.
>
... this probable typo (should s/Like/Link/ I think).
Regards,
Stefano
On OSF/1 5.1 with cc (but not with gcc), and only in the gettext build, not in
testdirs, I see a test failure like this:
test-spawn-pipe.c:80: assertion failed
(null):
/home/haible/gettext-0.18.2/gettext-tools/gnulib-tests/.libs/lt-test-spawn-pipe
subprocess got fatal signal 6
test-spawn
Bruno Haible wrote:
> Hi Eric,
>
>> > -ASSERT (errno == EINVAL || errno == EBUSY);
>> > +ASSERT (errno == EINVAL || errno == EBUSY || errno == EEXIST);
>
>> The patch to relax the testsuite looks okay to me; we're already
>> allowing other non-standard errno values here, so no one h
Hi Eric,
> > -ASSERT (errno == EINVAL || errno == EBUSY);
> > +ASSERT (errno == EINVAL || errno == EBUSY || errno == EEXIST);
> The patch to relax the testsuite looks okay to me; we're already
> allowing other non-standard errno values here, so no one has complained
> about it cau
On OSF/1 5.1 I'm seeing this test failure:
test-getlogin_r.c:76: assertion failed
sh: 363055 Abort - core dumped
FAIL: test-getlogin_r
The reason is that the OSF/1 getlogin_r() function produces a truncated
result when the buffer is not large enough. As if with strncpy. This
is not exactly
On OSF/1 5.1, the ptsname test fails:
test-ptsname.c:78: assertion failed
sh: 366922 Abort - core dumped
FAIL: test-ptsname
The reason is that for e.g. a master of "/dev/ptype" the ptsname function
returns the slave "/dev/pts/14" but our code expects "/dev/ttype".
On OSF/1 5.1, the ttyname_r test fails:
test-ttyname_r.c:45: assertion failed
sh: 367584 Abort - core dumped
FAIL: test-ttyname_r
The reason is that ttyname_r stores its result in the buffer, ignoring
the buffer's size. This patch provides a workaround:
2010-12-22 Bruno H
On OSF/1 5.1, I get these test failures:
test-rename.h:342: assertion failed
sh: 366977 Abort - core dumped
FAIL: test-rename
test-rename.h:342: assertion failed
sh: 366978 Abort - core dumped
FAIL: test-renameat
The reason are different errno values. The following patch fixes it for me.
OK to
On 12/22/2010 08:31 AM, Bruno Haible wrote:
> On OSF/1 5.1, I get these test failures:
>
> test-rename.h:342: assertion failed
> sh: 366977 Abort - core dumped
> FAIL: test-rename
>
> test-rename.h:342: assertion failed
> sh: 366978 Abort - core dumped
> FAIL: test-r
Bruno Haible wrote:
> On OSF/1 5.1 I got these test build failures:
>
> depbase=`echo test-stdio-c++.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`; g++
> -DHAVE_CONFIG_H -I. -DGNULIB_STRICT_CHECKING=1 -I. -I. -I.. -I./..
> -I../gllib -I./../gllib -mieee -MT
Hi Simon,
> > * test-gethostname.c:24: warning: initialization from incompatible pointer
> > type
> > Similarly:
> > - POSIX: int gethostname(char *, size_t);
> > - OSF/1: int gethostname(char *, int);
> >
> > Likewise.
> >
> > * test-inet_ntop.c:24: warning: initialization from incompatible
Bruno Haible writes:
> * test-gethostname.c:24: warning: initialization from incompatible pointer
> type
> Similarly:
> - POSIX: int gethostname(char *, size_t);
> - OSF/1: int gethostname(char *, int);
>
> Likewise.
>
> * test-inet_ntop.c:24: warning: initialization from incompatible pointe
On OSF/1 5.1, I'm seeing this gcc warning:
wcsrtombs.c: In function 'rpl_wcsrtombs':
wcsrtombs.c:44: warning: passing argument 4 of 'wcsrtombs' from incompatible
pointer type
The reason is that wcsrtombs exists but mbstate_t and mbsinit() are
replaced by gnu
Still on OSF/1 5.1, I see this error:
depbase=`echo test-wctype-c++.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`; g++
-DHAVE_CONFIG_H -I. -DGNULIB_STRICT_CHECKING=1 -I. -I. -I.. -I./..
-I../gllib -I./../gllib -Wall -mieee -MT test-wctype-c++.o -MD -MP -MF
$depbase.Tpo -c -o tes
On OSF/1 5.1, I'm seeing this warning:
strsignal.c: In function 'strsignal':
strsignal.c:111: warning: format '%d' expects type 'int', but argument 4 has
type 'long int'
The reason is that while POSIX says that SIGRTMIN and SIGRTMAX expand to
On OSF/1, the functions recv(), recvfrom(), send(), sendto() are declared
with an old-style declaration (return value 'int' instead of 'ssize_t',
and no use of 'socklen_t') by default. The fix is easy, this time: just
setting _POSIX_PII_SOCKET does
On OSF/1 5.1, the function tcgetsid() exists but is not declared by default.
(It would be with _XOPEN_SOURCE=500, but this is not a choice we can make.)
This leads to an error:
In file included from test-termios-c++.cc:22:
../gllib/termios.h:355: error: '::tcgetsid' has not bee
On OSF/1 5.1 I got these test build failures:
depbase=`echo test-stdio-c++.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`; g++
-DHAVE_CONFIG_H -I. -DGNULIB_STRICT_CHECKING=1 -I. -I. -I.. -I./..
-I../gllib -I./../gllib -mieee -MT test-stdio-c++.o -MD -MP -MF $depbase.Tpo
-c -o te
On OSF/1 5.1 I'm also seeing these gcc warnings:
* test-getaddrinfo.c:26: warning: initialization from incompatible pointer type
This one is because the gai_strerror declaration is different from POSIX:
- POSIX: const char *gai_strerror(int);
- OSF/1: char *gai_strerror(int);
I
On OSF/1, unsetenv() is also not declared by default. As a consequence,
there is an error in the C++ tests:
../gllib/stdlib.h:1022: error: 'unsetenv' was not declared in this scope
../gllib/stdlib.h:1022: error: invalid type in declaration before ';' token
and also the conf
r: invalid type in declaration before ';' token
This fixes it.
2010-12-19 Bruno Haible
setenv: Add missing declaration on OSF/1 5.1.
* lib/stdlib.in.h (setenv): Test HAVE_DECL_SETENV, not HAVE_SETENV.
* m4/setenv.m4 (gl_FUNC_SETENV_SEPARATE): Test whether s
On OSF/1 with g++, I'm seeing this error in the test suite:
depbase=`echo test-glob-c++.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`; g++
-DHAVE_CONFIG_H -I. -DGNULIB_STRICT_CHECKING=1 -I. -I. -I.. -I./..
-I../gllib -I./../gllib -mieee -MT test-glob-c++.o -MD -MP -MF $depb
Hi,
On OSF/1 5.1, the declarations of fseeko() and ftello() are missing by
default. Setting _XOPEN_SOURCE=500 would provide these declarations but
is not the right fix (see previous mail about fchdir()). I'll add this
fix for fseeko and similarly for ftello:
2010-12-19 Bruno H
fchdir: Add missing declaration on OSF/1 5.1.
* lib/unistd.in.h (fchdir): Provide declaration if systems lacks it.
* m4/fchdir.m4 (gl_FUNC_FCHDIR): Check whether fchdir is declared.
* m4/unistd_h.m4 (gl_UNISTD_H_DEFAULTS): Initialize HAVE_DECL_FCHDIR.
* modul
On OSF/1 5.1, the gettext build fails like this:
cc: Error: ../gnulib-lib/stdlib.h, line 50: In this declaration, the struct
"random_data" is redefined. (redefstruct)
struct random_data
^
*** Exit 1
The reason is that this platform has 'struct random_data' declared in
, n
Paolo Bonzini wrote:
> > The patch is obviously correct. You can apply it, together with this
> > enhancement of the unit test. I'll test it in the next few days.
>
> Done.
I'm updating the doc:
2008-10-05 Bruno Haible <[EMAIL PROTECTED]>
* doc/posix-headers/sys_select.texi: Mention '
Bruno Haible wrote:
> Hi Paolo,
>
>> What about this patch? Bruno, can you give it a try on OSF/1?
>
> The patch is obviously correct. You can apply it, together with this
> enhancement of the unit test. I'll test it in the next few days.
Done.
Paolo
Hi Paolo,
> What about this patch? Bruno, can you give it a try on OSF/1?
The patch is obviously correct. You can apply it, together with this
enhancement of the unit test. I'll test it in the next few days.
Bruno
*** tests/test-sys_select.c.orig2008-09-29 12:42:28.000
Simon Josefsson wrote:
> Bruno Haible <[EMAIL PROTECTED]> writes:
>
>> On OSF/1 4.0, exists but provides only a forward declaration
>> of 'struct timeval', no definition of this type.
>
> Sigh. So it seems we need a sys_select header module and a separ
Bruno Haible <[EMAIL PROTECTED]> writes:
> On OSF/1 4.0, exists but provides only a forward declaration
> of 'struct timeval', no definition of this type.
Sigh. So it seems we need a sys_select header module and a separate
select function after all: someone might w
On OSF/1 4.0, exists but provides only a forward declaration
of 'struct timeval', no definition of this type. This leads to this error:
cc -nodtk -O -std -std -DHAVE_CONFIG_H -I. -I.. -I. -I. -I.. -I./..
-I../gnulib-lib -I./../gnulib-lib
-I/nfs/visu/telecom/users/haible/gnu/a
On OSF/1 4.0d (but not on OSF/1 5.1) the 'truncl' test fails:
/bin/ksh: 3488 Memory fault(coredump)
FAIL: test-truncl
The reason is that the system truncl function exists and is declared, but is
completely broken:
$ cat foo.c
#include
int main ()
{
truncl (0.0l);
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to Gary E. Barnes on 11/6/2007 12:42 PM:
[adding the gnulib list]
> There is no snprintf on the system. If I modify src/system.c to call
> sprintf instead of snprintf I can link tar. But then "make check" fails
> when trying to compile ar
On OSF/1 5.1, the printf() and fprintf() replacements no longer produced
output of more than 1999 bytes. Wh...
2007-11-03 Bruno Haible <[EMAIL PROTECTED]>
* lib/vasnprintf.c (VASNPRINTF): Don't assume that snprintf's return
value is C99 compliant.
Hi Simon,
This getaddrinfo.c error on OSF/1 is now the only remaining compilation error
of all of gnulib on OSF/1, HP-UX, AIX, Solaris, IRIX.
> > int getnameinfo(
> > const struct sockaddr *sa,
> > socklen_t salen,
> > char *node,
>
Bruno Haible <[EMAIL PROTECTED]> writes:
> Hi Simon,
>
> On OSF/1 5.1, with cc as compiler, the getaddrinfo module fails to compile:
>
> cc: Warning: getaddrinfo.c, line 334: In this declaration, parameter 2 has a
> different type than specified in an earlier decl
Hi Simon,
On OSF/1 5.1, with cc as compiler, the getaddrinfo module fails to compile:
cc: Warning: getaddrinfo.c, line 334: In this declaration, parameter 2 has a
different type than specified in an earlier declaration of this function.
(mismatparam)
int getnameinfo(const struct sockaddr
Eric Blake-1 <[EMAIL PROTECTED]> writes:
> Why don't we instead add a new module, wchar, which provides a replacement
> using the same techniques as we do with stdint.m4.
Sounds reasonable to me.
> Confirmed. Thanks, Albert! I wasn't aware that this was the same bug
> that we work around in so many other places.
>
> I'm adding the fix below to gnulib. No need any more for -I/usr/include
> or option -nodtk.
Why don't we instead add a new module, wchar, which provides a replacement
using
Albert Chin wrote:
> > > cc: Error: /usr/include.dtk/wchar.h, line 87: Ill-formed parameter type
> > > list. (parmtyplist)
> > > __VA_LIST__ __arg);
> > > ^
> > > configure:14834: $? = 1
>
> The fix is to #include before .
Confirmed. Thanks, Albert! I wasn't aware
On Mon, Aug 07, 2006 at 08:47:38PM +0200, Bruno Haible wrote:
> Gary V. Vaughan wrote:
> > Here is the culprit:
> >
> > configure:14791: checking whether wcwidth is declared
> > configure:14828: cc -c -g conftest.c >&5
> > cc: Error: /usr/include.dtk/wchar.h, line 78: Ill-formed parameter type
>
Hello Bruno,
* Bruno Haible wrote on Tue, Aug 08, 2006 at 03:32:33PM CEST:
>
> + If GNU CC is not installed, it is therefore recommended to try
> +
> + env CC="cc" ./configure
> +
> + and if that doesn't work, try
> +
> + env CC="cc -nodtk" ./configure
What does it need to persuad
. yes
> alphaev67-dec-osf5.1$ /opt/TWWfsw/gmake38/bin/make all check
>[[...]]
> All checks successful
Interesting! Thanks for this solution. I'm going to document it in
gettext's INSTALL file.
Bruno
*******
*** 171,176
--- 171,186
env CC="cc -Ae" .
All:
Bruno Haible wrote:
> Gary V. Vaughan wrote:
>> Here is the culprit:
>>
>> configure:14791: checking whether wcwidth is declared
>> configure:14828: cc -c -g conftest.c >&5
>> cc: Error: /usr/include.dtk/wchar.h, line 78: Ill-formed parameter type
>> list. (parmtyplist)
>> __
Gary V. Vaughan wrote:
> Here is the culprit:
>
> configure:14791: checking whether wcwidth is declared
> configure:14828: cc -c -g conftest.c >&5
> cc: Error: /usr/include.dtk/wchar.h, line 78: Ill-formed parameter type
> list. (parmtyplist)
> __VA_LIST__ __arg);
> --
Hi Bruno,
Gary V. Vaughan wrote:
> alphaev67-dec-osf5.1$ ../configure --enable-changeword CC=gcc |
>> fgrep wcwidth
> checking for wcwidth... yes
> checking whether wcwidth is declared... yes
> alphaev67-dec-osf5.1$ fgrep WCWIDTH config.h
> #define HAVE_DECL_WCWIDTH 1
> #define HAVE_WCWIDTH 1
Hi Bruno,
Bruno Haible wrote:
> Gary V. Vaughan wrote:
>
>> A gnulib bug on osf (4.0d and 5.1):
>> alphaev67-dec-osf5.1$ make all check
>> cc -DHAVE_CONFIG_H -I. -I../../lib -I.. -g -c ../../lib/mbchar.c
>> cc: Warning: ../../lib/wcwidth.h, line 70: In this declaration,
>> parameter 1
Gary V. Vaughan wrote:
> A gnulib bug on osf (4.0d and 5.1):
> alphaev67-dec-osf5.1$ make all check
> cc -DHAVE_CONFIG_H -I. -I../../lib -I.. -g -c ../../lib/mbchar.c
> cc: Warning: ../../lib/wcwidth.h, line 70: In this declaration,
> parameter 1 has a different type than specified in a
65 matches
Mail list logo
