Error trying to compile pspp using up-to-date gnulib

Hi Guys, When I tried to compile pspp using both pspp and gnulib git versions I got this using mingw and windows: Creating directory ./"unilbrk mkdir: cannot create directory `./"unilbrk': No such file or directory /home/Michel/gnulib//gnulib-tool: *** failed /home/Michel/gnulib//gnulib-tool: ***

Re: [PATCH 1/2] selinux-h: new module (from coreutils/gl/)

Two more remarks: Jim Meyering wrote: > +# define security_context_t char* Is it really a 'char *' also in the real SELinux? If not, then this typedef will allow assigments between security_context_t and string values, without producing gcc warnings. If you want to avoid this possibility, somethi

Re: [PATCH 1/2] selinux-h: new module (from coreutils/gl/)

Hi Jim, A few random remarks. Why is the module name 'selinux-h', not 'selinux'? (We have 'stdlib', 'unistd' etc. for include files.) Is there a separate module 'selinux' to come later? > +# define is_selinux_enabled() 0 > +Description: > +SELinux-related headers for systems that lack them. Co

[PATCH 1/2] selinux-h: new module (from coreutils/gl/)

Hi, I've just added the following new module, from coreutils. The second change set (below) cleans up a little. I haven't yet listed the new module in MODULES.html.in, mainly because I haven't spotted a good place yet. Suggestions welcome. >From 8a5ff29ccb28ea665463f0ead178d4b0f801c2d1 Mon Sep 1

Re: winsock.c split

Simon Josefsson <[EMAIL PROTECTED]> writes: >>> In particular, I'm not sure the close module is handled correctly. >> >> It's correct. >> >> I'm proposing this addendum, mostly simplifications. > > Looks fine, please push and I can prepare a new gsasl release that use > all all these changes. I p

[PATCH] prepare to move selinux-h module to gnulib

FYI, Since this module is useful to augeas , I've relaxed the license and will soon move it to gnulib. >From 1a13e78ca14dd8b50c00af18d472dee33825560b Mon Sep 17 00:00:00 2001 From: Jim Meyering <[EMAIL PROTECTED]> Date: Tue, 21 Oct 2008 15:38:48 +0200 Subject: [PATCH] prepare

Re: winsock.c split

Bruno Haible <[EMAIL PROTECTED]> writes: > Hi Simon, > >> I have pushed this, but would appreciate testing and review. > > Everything's fine, except the change of doc/posix-functions/close.texi. > (close() sets errno also under Windows, only closesocket() does not.) Oh, I didn't know some Windows

Re: winsock.c split

Hi Simon, > I have pushed this, but would appreciate testing and review. Everything's fine, except the change of doc/posix-functions/close.texi. (close() sets errno also under Windows, only closesocket() does not.) > In particular, I'm not sure the close module is handled correctly. It's correc

Re: test-poll vs ioctl

Bruno Haible <[EMAIL PROTECTED]> writes: > Sure, that's due to the missing dependency to 'ioctl' in modules/poll-tests > and modules/select-tests. When you add that dependency, as you said in your > previous mail, compilation works fine. Ah, I forgot these... > To get out of this confusion, I'm

Re: test-poll vs ioctl

Hi Simon, > >> 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

Re: test-poll vs ioctl

Bruno Haible <[EMAIL PROTECTED]> writes: > 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 .. >>

Re: poll-tests: Pipe test crashes wineserver

Paolo Bonzini <[EMAIL PROTECTED]> writes: > 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