Re: blocking socket is nonblocking after calling gnulib select() in windows

On Thu, Apr 21, 2011 at 22:17, Bastien ROUCARIES wrote: > What give FILE_MODE_INFORMATION query using NtQueryInformationFile on > (Handle) Socket ? http://msdn.microsoft.com/en-us/library/cc232119%28v=prot.10%29.aspx Nothing related to blocking/nonblocking state. That's done in the guts of win

Re: [6/7] Proposed patches to remove several inttypes-related dependencies

On 04/22/11 03:31, Eric Blake wrote: > An alternative might be to have a new module, inttypes-h, which provides > the shell replacement header, and keep inttypes as the module that pulls > in everything C99 related Yes, thanks, that would be more backward-compatible. Here's a further patch that do

Re: [1/7] Proposed patches to remove several inttypes-related dependencies

On 04/22/11 03:07, Eric Blake wrote: > How can the egrep check ever succeed if the declaration check (which > detects macros) failed? You can further trim this .m4 file. On hosts that do this: #define strtoimax(a,b,c) __strtoll(a,b,c) the declaration check will fail, because it uses strtoimax

Re: Proposed patches to remove several inttypes-related dependencies

On 04/22/2011 03:54 AM, Paul Eggert wrote: > Subject: [PATCH 7/7] Fix typo in modules/strtoumax: it needs strtoul.c too. > > --- > ChangeLog |2 +- > modules/strtoumax |1 + > 2 files changed, 2 insertions(+), 1 deletions(-) > > diff --git a/ChangeLog b/ChangeLog > index babb259.

Re: [6/7] Proposed patches to remove several inttypes-related dependencies

On 04/22/2011 04:31 AM, Eric Blake wrote: >> +2011-04-22 inttypesThis module no longer arranges for >> +to declare imaxabs; that has been moved to the >> +imaxabs module. Similarly for imaxdiv, >> strtoimax, >> +

Re: [6/7] Proposed patches to remove several inttypes-related dependencies

On 04/22/2011 03:54 AM, Paul Eggert wrote: > diff --git a/ChangeLog b/ChangeLog > index 1ed5995..babb259 100644 > --- a/ChangeLog > +++ b/ChangeLog > @@ -1,5 +1,29 @@ > 2011-04-22 Paul Eggert > > + inttypes: migrate 'configure' checks to modules that need it > + This module was quite

Re: [3/7] Proposed patches to remove several inttypes-related dependencies

On 04/22/2011 03:54 AM, Paul Eggert wrote: > + strtoll: remove dependency on strtol > + This is so that 'configure' need not check for strtol merely because > + the application needs strtoll. > + * modules/strtoll (Files): Add lib/strtol.c. > + (Depends-on): Remove strtol. > + >

Re: [2/7] Proposed patches to remove several inttypes-related dependencies

On 04/22/2011 03:54 AM, Paul Eggert wrote: > + AC_CHECK_DECLS_ONCE([strtoumax]) > + if test "$ac_cv_have_decl_strtoumax" != yes; then > +HAVE_DECL_STRTOUMAX=0 > + > +AC_CACHE_CHECK([whether defines strtoumax as a macro], > + gl_cv_func_strtoumax_macro, > + [AC_EGREP_CPP([inttyp

Re: [1/7] Proposed patches to remove several inttypes-related dependencies

On 04/22/2011 03:54 AM, Paul Eggert wrote: > Here's a series of proposed patches designed to remove unnecessary > dependencies from gnulib's implementation of inttypes.h and related > functions. I've become accustomed to the habit of patch series with one patch per email; it's easier to reply (whe

Re: propose removing strnlen dependency on memchr

On 04/22/11 02:21, Jim Meyering wrote: > Gnulib-fixed memchr bugs affect only glibc 2.10 and 2.11, > and glibc has had strnlen forever. So I agree. > Nice. Thanks! I pushed that.

Re: Proposed patches to remove several inttypes-related dependencies

On 04/22/2011 03:54 AM, Paul Eggert wrote: > Sorry about the long "Subject:" in the first patch: I don't > know what went wrong. > >>From 402d05161d2fcfdd2308790969d4b44dae0f9c50 Mon Sep 17 00:00:00 2001 > From: Paul Eggert > Date: Fri, 22 Apr 2011 01:11:48 -0700 > Subject: [PATCH 1/7] strtoimax:

Proposed patches to remove several inttypes-related dependencies

Here's a series of proposed patches designed to remove unnecessary dependencies from gnulib's implementation of inttypes.h and related functions. This patch (and the previous strtol patch) effectively removed 11 files from a test version of Emacs that uses strtoumax (in lib: inttypes.h, strtoll,c

Re: propose removing strnlen dependency on memchr

Paul Eggert wrote: > I considered adding the strnlen module to Emacs, but this had > the problem of bringing in the memchr module, which is pretty > heavyweight. I propose that we remove this dependency, as > the bugs fixed by the memchr module are not tickled by > the gnulib strnlen implementatio

propose removing strnlen dependency on memchr

I considered adding the strnlen module to Emacs, but this had the problem of bringing in the memchr module, which is pretty heavyweight. I propose that we remove this dependency, as the bugs fixed by the memchr module are not tickled by the gnulib strnlen implementation. Possibly other dependenci

[PATCH] strtol: remove dependency on wchar

I added the strtoumax module to a test version of Emacs, and it pulled in a lot of gnulib stuff that it didn't need. The diff is about 200 kB One simple fix can remove the need for the wchar module, s

Re: blocking socket is nonblocking after calling gnulib select() in windows

> Really nice to see that, ntdeviceiocontrol thread could be put in > alertable state and thus safely stopable by an APC. APC to stop? Do you mean call WSAIoctl and then use an APC to terminate the thread? I don't think that would be safe and I assume there would remain a pending operation if