Eric Blake wrote:
> > Since -isystem is some burden on the gnulib user (not a big one, but anyway)
> > I propose to add
> >#ifdef __GNUC__
> ># pragma GCC system_header
> >#endif
> > to all gnulib headers that use @[EMAIL PROTECTED] Fortunately gnulib does
> > not
> > use __STDC__ in a
This removes a one-year-old FIXME from m4/include_next.m4.
2008-08-29 Bruno Haible <[EMAIL PROTECTED]>
* m4/include_next.m4 (gl_INCLUDE_NEXT): Don't define HAVE_INCLUDE_NEXT
any more.
--- m4/include_next.m4.orig Sat Aug 30 02:10:51 2008
+++ m4/include_next.m4 Sat Aug 30 02
Simon Josefsson wrote:
> +
> +#include
> +
The module description needs to list the dependency to the 'stdint' module.
Bruno
Ben Pfaff wrote:
> Since you're using the inline keyword, you should add a
> dependency on the inline module.
He's only using 'static inline'; therefore an AC_REQUIRE([AC_C_INLINE])
is all that he needs.
> > +/* Given an unsigned 32-bit argument X, return the value corresponding
> > + to rotati
Eric Blake <[EMAIL PROTECTED]> writes:
> how about we go one step
> further, and have a deprecation period, where for a year or so, we provide
> mbase_name (returns NULL on failure) and xbase_name (calls xmalloc on
> failure) but no base_name.
There's a better solution that doesn't involve waitin
Simon Josefsson <[EMAIL PROTECTED]> writes:
> I have pushed the patch below, but i still appreciate further review of
> the code.
Since you're using the inline keyword, you should add a
dependency on the inline module.
> +/* Given an unsigned 32-bit argument X, return the value corresponding
> +
> +/* Given an unsigned 16-bit argument X, return the value corresponding
> + to rotating the bits N steps to the left. N must be between 1 to
> + 15 inclusive. */
> +static inline uint16_t
> +rotl16 (uint16_t x, int n)
> +{
> + return ((x << n) | (x >> (16 - n))) & 0x;
> +}
& 0xFFF
Bruno Haible <[EMAIL PROTECTED]> writes:
> Simon Josefsson wrote:
>> I'd prefer to solve it by having m4 code check whether
>> the non-portable code works or not, and then AC_SUBST in the proper
>> replacement function, either getline-fast.c or getline-slow.c. This
>> allows a gnulib module 'getl
Paul Eggert <[EMAIL PROTECTED]> writes:
> Simon Josefsson <[EMAIL PROTECTED]> writes:
>
>> +#define rotl32(x,n) \
>> + x) << ((uint16_t)(n))) | ((x) >> (32 - (uint16_t)(n & 0x)
>
> I suggest using inline functions rather than macros, to avoid hassles
> with double-evaluating argum
Eric Blake <[EMAIL PROTECTED]> wrote:
> According to Jim Meyering on 8/29/2008 6:00 AM:
>>> 1) About the naming of the functions. We have one convention so far to
>>>distinguish a library-safe function and one that calls xalloc_die()
>>>upon memory allocation error: the prefix "x" ("checki
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to Jim Meyering on 8/29/2008 6:00 AM:
>> 1) About the naming of the functions. We have one convention so far to
>>distinguish a library-safe function and one that calls xalloc_die()
>>upon memory allocation error: the prefix "x" ("che
Bruno Haible <[EMAIL PROTECTED]> wrote:
...
> This would also solve a problem that was mentioned some time ago: 'argp'
> is an LGPLed module, but it uses 'dirname' which is not under LGPL.
>
>> I wanted a version of findprog that I could use from a library
>> (libvirt), but the existing one can cal
Hi Jim,
> ended up with
> a new module, mfilenamecat (wrapper around filenamecat), and mfilename
> (contents of basename.c and dirname.c, adjusted not to exit).
This would also solve a problem that was mentioned some time ago: 'argp'
is an LGPLed module, but it uses 'dirname' which is not under L
FYI, I've just pushed this:
>From 2477d329d432a5969750486f2ea7657f5c54abae Mon Sep 17 00:00:00 2001
From: Jim Meyering <[EMAIL PROTECTED]>
Date: Fri, 29 Aug 2008 12:53:50 +0200
Subject: [PATCH] bootstrap: merge changes from coreutils
* build-aux/bootstrap (cp_mark_as_generated): Preserve perms
of
Jim Meyering <[EMAIL PROTECTED]> wrote:
...
> diff --git a/lib/mfilename.c b/lib/mfilename.c
> new file mode 100644
> index 000..10bdfbd
> --- /dev/null
> +++ b/lib/mfilename.c
> @@ -0,0 +1,182 @@
> +/* simple file-name-manipulation functions that never exit
> +
> + Copyright (C) 1996-2008 Fr
Simon Josefsson wrote:
> I'd prefer to solve it by having m4 code check whether
> the non-portable code works or not, and then AC_SUBST in the proper
> replacement function, either getline-fast.c or getline-slow.c. This
> allows a gnulib module 'getline-slow' for when people prefer a small and
> s
> Thanks. It's fine now: the newest MODULES.html fixes the problems with the
> hyperlinks.
I'm removing the --cvs-urls option. Note that the problem was that these URLs
referred to the old cvs repository. The git-mirroring cvs repository works
fine, but AFAIK it is not viewable through http.
Brun
Bruno Haible <[EMAIL PROTECTED]> wrote:
> Hi Jim,
>
>> Will commit this shortly:
>>
>> * modules/getndelim2 (License): Relicense to LGPLv2+.
>
> Do you have RMS's backing for this move? The original version of this code
> was contributed to the FSF by Jan Brittenson, who is not on this mailing
Bruno Haible wrote:
> Hi Jim,
>
>> Will commit this shortly:
>>
>> * modules/getndelim2 (License): Relicense to LGPLv2+.
>
> Do you have RMS's backing for this move? The original version of this code
> was contributed to the FSF by Jan Brittenson, who is not on this mailing list.
> It was my
19 matches
Mail list logo
