Re: mark atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, strcspn, strpbrk as obsolete

Bruno Haible <[EMAIL PROTECTED]> wrote: ... >> A more conservative approach would be to make each of those >> proposed-obsolete module AC_REQUIRE a new macro with code to run at >> configure-time if ever the module is required. That code could make >> configure fail by default with a diagnostic >

Re: test-poll vs ioctl

> a) Rename winsock.c to w32sock.h. > Create a file connect.c containing > >#define GNULIB_CONNECT 1 >#include "w32sock.h" > > and similarly for the other modules that make up winsock.c. > > Remove the gl_MODULE_INDICATOR([$1]) line from the definition of >

Re: test-poll vs ioctl

Simon Josefsson wrote: > The current situation breaks simple test-cases like this: > > rm -rf m;gnulib-tool --create-testdir --with-tests --dir m poll; cd m; > ./configure --host=i586-mingw32msvc --build=i686-pc-linux-gnu; make check; cd > .. > > Regardless of the solution, I think the above co

Re: sys_socket: inconsistent linker warnings?

Hi Simon, > Linking gsasl on mingw fails with: > > gsasl.o: In function `readln': > /home/jas/src/gsasl/gsasl-0.2.29/src/gsasl.c:101: undefined reference to > `_recv_used_without_requesting_gnulib_module_recv' > gsasl.o: In function `main': > /home/jas/src/gsasl/gsasl-0.2.29/src/gsasl.c:480: und

Re: abort() on Windows 2008 Server

Eric Blake wrote: > > int > > sigfillset (sigset_t *set) > > { > > ! *set = ((2U << (NSIG - 1)) - 1) > > !#ifdef SIGABRT_COMPAT > > !& ~(1U << SIGABRT_COMPAT) > > !#endif > > !; > > However, I'm not the biggest fan of #ifdef inside a statement. Maybe a > helper macro would

Re: declare lstat in sys/stat.h

Jim Meyering wrote: > This looks fine. Applied.

Re: poll-tests: Pipe test crashes wineserver

Simon Josefsson wrote: > Running this > > [EMAIL PROTECTED]:~/src/gnulib master$ rm -rf m;gnulib-tool --create-testdir > --with-tests --dir m poll; cd m; ./configure --host=i586-mingw32msvc > --build=i686-pc-linux-gnu; make check; cd .. > > I get this output: > > Unconnected socket test... pas

Re: declare getusershell in unistd.h

Jim Meyering wrote: > At first I removed those declarations from su.c, > but then the ones from the system unistd.h were > not visible on recent Linux. glibc's declares getusershell() properly when _GNU_SOURCE is defined. At least that's what I can see in the glibc sources. Do you have more detai

Re: mark atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, strcspn, strpbrk as obsolete

Jim Meyering wrote: > Since you listed Irix 6.2 and AIX 4.3.2 as being immune, I presumed > that versions earlier than those, yet with the same major number, > may be impacted. Not necessarily. Simply, we don't have data about IRIX 6.1 and AIX 4.2. > If, as you now imply, all of 6.x and 4.x.x are

Re: mark atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, strcspn, strpbrk as obsolete

Tom G. Christensen wrote: > > The systems which don't have atexit() ... raise() are things like > > Solaris 2.3, IRIX 5, SunOS 4, NeXTstep, Ultrix. Museumware. We haven't > > heard any report from anyone using such old systems in 3 years. > > > Perhaps I misunderstand but on my request you recentl

Re: mark atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, strcspn, strpbrk as obsolete

On Sun, Oct 19, 2008 at 11:38:01PM +0200, Bruno Haible wrote: > Hi Jim, > > > Gnulib is more than a little about portability. > > Do we really want to discourage creation of packages > > that will work on IRIX 6.1 and AIX older than 4.3.2? > > The systems which don't have atexit() ... raise() are

Re: getaddrinfo, netdb, canon-host

Simon Josefsson <[EMAIL PROTECTED]> wrote: ... > I see you already did, thanks. > > There seems to be some lag on gmane's archive of this list, I didn't > receive your reply until now. Yes, I've noticed it on other lists, too, recently. Makes me feel justified for preferring email ;-)

Re: test-poll vs ioctl

Paolo Bonzini <[EMAIL PROTECTED]> writes: >> There is no -I for the $gltests/ directory in $gl/, so this fails. >> >> I think the ioctl replacement function should be moved out from >> winsock.c to a new ioctl.c. >> >> What do you think of the patch below? > > Before applying something like this

poll-tests: Pipe test crashes wineserver

Running this [EMAIL PROTECTED]:~/src/gnulib master$ rm -rf m;gnulib-tool --create-testdir --with-tests --dir m poll; cd m; ./configure --host=i586-mingw32msvc --build=i686-pc-linux-gnu; make check; cd .. I get this output: Unconnected socket test... passed Connected sockets test... passed Gene

Re: getaddrinfo, netdb, canon-host

Jim Meyering <[EMAIL PROTECTED]> writes: > Simon Josefsson <[EMAIL PROTECTED]> wrote: >> Jim, the canon-host module needs this patch to work with the latest >> changes in the getaddrinfo module. > > > Thanks, Simon. > That looks obviously fine. > Please apply. I see you already did, thanks. Ther

Re: test-poll vs ioctl

> There is no -I for the $gltests/ directory in $gl/, so this fails. > > I think the ioctl replacement function should be moved out from > winsock.c to a new ioctl.c. > > What do you think of the patch below? Before applying something like this, it would make sense to understand if the root pro

test-poll vs ioctl

Bruno Haible <[EMAIL PROTECTED]> writes: > This defines a module for the 'ioctl' function, and moves its declaration > from to . The poll-tests module uses sys/ioctl.h and ioctl, even though poll doesn't need them, so I added these modules to the poll-tests module dependencies. However that doe

Re: new module 'fclose'

-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Simon Josefsson on 10/20/2008 7:10 AM: > And indeed gl_REPLACE_FCLOSE is called even if the module is avoided, > see close.m4: > > > Is there a good solution here? You could add m4_pushdef([gl_REPLACE_FCLOSE]) in your configure.ac. - -

Re: new module 'fclose'

Bruno Haible <[EMAIL PROTECTED]> writes: > We're overriding close(). But the system's fclose(), when called on a FILE > stream that was created with fdopen() from a socket descriptor, will not > invoke our overridden close(). So we have to override fclose() as well. > > I'm committing this. It has

sys_socket: inconsistent linker warnings?

Linking gsasl on mingw fails with: gsasl.o: In function `readln': /home/jas/src/gsasl/gsasl-0.2.29/src/gsasl.c:101: undefined reference to `_recv_used_without_requesting_gnulib_module_recv' gsasl.o: In function `main': /home/jas/src/gsasl/gsasl-0.2.29/src/gsasl.c:480: undefined reference to `_rp

Re: getaddrinfo, netdb, canon-host

Jim Meyering <[EMAIL PROTECTED]> wrote: > Simon Josefsson <[EMAIL PROTECTED]> wrote: >> Jim, the canon-host module needs this patch to work with the latest >> changes in the getaddrinfo module. > > Thanks, Simon. > That looks obviously fine. > Please apply. I went ahead and pushed that, since the

Re: mark atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, strcspn, strpbrk as obsolete

Bruno Haible <[EMAIL PROTECTED]> writes: > Hi Simon, > >> If it is only a warning, +1 from me. > > It's a notice, and soon we will also remove dependencies to these modules, > since we don't want these notices to appear for people who have not explicitly > asked for these modules. I like this. >

Re: abort() on Windows 2008 Server

-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Bruno Haible on 10/18/2008 3:37 PM: > > Thanks for this analysis. The sigset_t datatype introduced by gnulib has > to be adjusted. I propose this patch. Eric, what do you think? Generally looks okay to me. > int > sigfillset (sigset

another posix_spawn bug on AIX

The implementation of posix_spawn on AIX 5.3..6.1 has yet another bug: it fails when asked to open a file whose name contains a '*'. Apparently the implementors (from a subcontractor of IBM, not IBM itself, I am being told) just took the primitive sample implementation from

fix posix_spawn test for MacOS X

On MacOS X the posix_spawn currently fails, although the system function behaves perfectly fine - just differently than the glibc implementation. This fixes it. 2008-10-20 Bruno Haible <[EMAIL PROTECTED]> * m4/posix_spawn.m4 (gl_POSIX_SPAWN_WORKS): Make the check on MacOS X. *** m4/pos

Re: getaddrinfo, netdb

Bruno Haible <[EMAIL PROTECTED]> writes: > Simon Josefsson wrote: >> -#ifdef __USE_GNU >> -# ifndef EAI_INPROGRESS >> -# define EAI_INPROGRESS-100/* Processing request in progress. */ >> -# define EAI_CANCELED -101/* Request canceled. */ >> -# define EAI_NOTCANCELED

Re: declare getusershell in unistd.h

Bruno Haible <[EMAIL PROTECTED]> wrote: >> 2008-10-18 Bruno Haible <[EMAIL PROTECTED]> >> >> * lib/unistd.in.h (getusershell, setusershell, endusershell): New >> declarations. >> * lib/getusershell.c: Include unistd.h. >> * m4/getusershell.m4 (gl_FUNC_GETUSERSHELL): Require >>

Re: declare lstat in sys/stat.h

Bruno Haible <[EMAIL PROTECTED]> wrote: > lstat() is declared in according to POSIX. Here's a proposed > patch for gnulib to do the same. > > The only tricky issue here is that on some systems we have a > "#define lstat lstat64", and this gets in the way of redefining the function > while at the s

Re: getaddrinfo, netdb, canon-host

Simon Josefsson <[EMAIL PROTECTED]> wrote: > Jim, the canon-host module needs this patch to work with the latest > changes in the getaddrinfo module. Thanks, Simon. That looks obviously fine. Please apply.

Re: declare gethostname in unistd.h

Bruno Haible <[EMAIL PROTECTED]> writes: > Hi Simon, > > Here's a proposed patch so that on platforms where gethostname() is missing > and substituted by gnulib, it is declared in . OK to commit? Hi. Yes, please do. > Note that this does not resolve the issue with mingw. On mingw, > declares g

Re: mark atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, strcspn, strpbrk as obsolete

Bruno Haible <[EMAIL PROTECTED]> wrote: >> Gnulib is more than a little about portability. >> Do we really want to discourage creation of packages >> that will work on IRIX 6.1 and AIX older than 4.3.2? > > The systems which don't have atexit() ... raise() are things like > Solaris 2.3, IRIX 5, Sun

Re: mark atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, strcspn, strpbrk as obsolete

Hi Simon, > If it is only a warning, +1 from me. It's a notice, and soon we will also remove dependencies to these modules, since we don't want these notices to appear for people who have not explicitly asked for these modules. > What about strdup? Indeed, we can treat it the same way: --- mo

Re: getaddrinfo, netdb

Simon Josefsson wrote: > -#ifdef __USE_GNU > -# ifndef EAI_INPROGRESS > -# define EAI_INPROGRESS -100/* Processing request in progress. */ > -# define EAI_CANCELED -101/* Request canceled. */ > -# define EAI_NOTCANCELED-102/* Request not canceled. */ > -# de

Re: getaddrinfo, netdb, canon-host

Jim, the canon-host module needs this patch to work with the latest changes in the getaddrinfo module. /Simon >From d6843ea1869f2c5762179de1699c0c12d357277a Mon Sep 17 00:00:00 2001 From: Simon Josefsson <[EMAIL PROTECTED]> Date: Mon, 20 Oct 2008 09:59:07 +0200 Subject: [PATCH] canon-host: Includ

Re: getaddrinfo, netdb, canon-host

I have pushed the following change. The canon-host module needs to be fixed, I'll send a patch separately. /Simon >From dbf8ca54bfdeaef894c8d0912df6e05b4c1a7302 Mon Sep 17 00:00:00 2001 From: Simon Josefsson <[EMAIL PROTECTED]> Date: Mon, 20 Oct 2008 09:53:42 +0200 Subject: [PATCH] Move getaddri

Re: mark atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, strcspn, strpbrk as obsolete

Bruno Haible <[EMAIL PROTECTED]> writes: > Hi, > > The functions atexit, memchr, memcmp, memcpy, memmove, memset, raise, rmdir, > strcspn, strpbrk are all present in systems newer than ca. 1998. They are > not missing in systems as old as AIX 4.3.2, HP-UX 10.20, IRIX 6.2, OSF/1 4.0, > Solaris 2.5.