Robert Millan wrote:
- In the average case, you can get away with 1 strftime call instead of 2,
if you preallocate a buffer on the stack:
char buf[256];
len = strftime (buf, sizeof (buf), %c, loctime) + 1;
In the case where 256 bytes are not enough, the function
Jim Meyering wrote:
Rich Jones wanted to use xstrtoull and/or xstrtoll in libguestfs,
but it didn't exist yet. So here is the new module.
I took the opportunity to write the test using the new init.sh framework.
A couple things can be improved:
Currently, every test that uses init.sh must
On 22/01/10 17:09, Jim Meyering wrote:
Eric Blake wrote:
According to Jim Meyering on 1/22/2010 6:17 AM:
However, it'd sure be nice to use something more generic than
lib/config.h. IMHO, autoconf should make configure AC_SUBST its
currently-internal-only CONFIG_HEADERS variable. While we
On Sat, Jan 23, 2010 at 12:13:51PM +0100, Jim Meyering wrote:
diff --git a/lib/xstrtol.h b/lib/xstrtol.h
index 95475f0..3a94a9c 100644
--- a/lib/xstrtol.h
+++ b/lib/xstrtol.h
@@ -46,6 +46,11 @@ _DECLARE_XSTRTOL (xstrtoul, unsigned long int)
_DECLARE_XSTRTOL (xstrtoimax, intmax_t)
Pádraig Brady wrote:
On 22/01/10 17:09, Jim Meyering wrote:
Eric Blake wrote:
According to Jim Meyering on 1/22/2010 6:17 AM:
However, it'd sure be nice to use something more generic than
lib/config.h. IMHO, autoconf should make configure AC_SUBST its
currently-internal-only CONFIG_HEADERS
Thomas Treichl wrote:
a month ago I was still working on an older OS X 10.4 system and tried
to compile recent sources of GNU Octave which now uses the bootstrap
file. I failed with a local copy of gnulib (I used the option
--gnulib-srcdir=path) because I didn't have any of the tools
On Mon, Jan 25, 2010 at 12:39:02PM +0100, Jim Meyering wrote:
Richard W.M. Jones wrote:
On Sat, Jan 23, 2010 at 12:13:51PM +0100, Jim Meyering wrote:
diff --git a/lib/xstrtol.h b/lib/xstrtol.h
index 95475f0..3a94a9c 100644
--- a/lib/xstrtol.h
+++ b/lib/xstrtol.h
@@ -46,6 +46,11 @@
According to Richard W.M. Jones on 1/25/2010 4:34 AM:
This didn't work for me. I had to add '#include config.h' near the
top of 'lib/xstrtol.h' in order to get the symbol HAVE_LONG_LONG_INT
defined.
In general, we've specifically avoided including config.h in .h files;
they should only be
Richard W.M. Jones wrote:
On Sat, Jan 23, 2010 at 12:13:51PM +0100, Jim Meyering wrote:
diff --git a/lib/xstrtol.h b/lib/xstrtol.h
index 95475f0..3a94a9c 100644
--- a/lib/xstrtol.h
+++ b/lib/xstrtol.h
@@ -46,6 +46,11 @@ _DECLARE_XSTRTOL (xstrtoul, unsigned long int)
_DECLARE_XSTRTOL
On 25/01/10 11:51, Jim Meyering wrote:
Pádraig Brady wrote:
Subject: [PATCH] syntax-check: detect incorrect boolean macro values in config.h
* modules/maintainer-makefile (configure.ac): Parameterize the location
of config.h. The logic is from Eric Blake and the location indicated
by Jim
Pádraig Brady wrote:
On 25/01/10 11:51, Jim Meyering wrote:
Pádraig Brady wrote:
Subject: [PATCH] syntax-check: detect incorrect boolean macro values in
config.h
* modules/maintainer-makefile (configure.ac): Parameterize the location
of config.h. The logic is from Eric Blake and the
Hi,
This message concerns both gnulib and grub. As discussed on irc and on
the list [1], ENABLE_NLS is not used correctly, which leads to a build
failure when gettext is not detected (or with configure option
--disable-nls).
ENABLE_NLS is defined in AM_GNU_GETTEXT and the documentation of
According to Grégoire Sutre on 1/25/2010 2:53 AM:
Hi,
This message concerns both gnulib and grub. As discussed on irc and on
the list [1], ENABLE_NLS is not used correctly, which leads to a build
failure when gettext is not detected (or with configure option
--disable-nls).
ENABLE_NLS
Hi Jim,
diff --git a/tests/test-xstrtoll.c b/tests/test-xstrtoll.c
index 47a552e..03dd232 100644
--- a/tests/test-xstrtoll.c
+++ b/tests/test-xstrtoll.c
@@ -1,4 +1,4 @@
#define __xstrtol xstrtoll
#define __strtol_t long long int
-#define __spec lld
+#define __spec PRId64
#include
Bruno Haible wrote:
Hi Jim,
diff --git a/tests/test-xstrtoll.c b/tests/test-xstrtoll.c
index 47a552e..03dd232 100644
--- a/tests/test-xstrtoll.c
+++ b/tests/test-xstrtoll.c
@@ -1,4 +1,4 @@
#define __xstrtol xstrtoll
#define __strtol_t long long int
-#define __spec lld
+#define __spec
Hi,
Grégoire Sutre wrote:
ENABLE_NLS is defined in AM_GNU_GETTEXT and the documentation of this
macro [2] does not require ENABLE_NLS to be defined when gettext is not
available.
Correct: ENABLE_NLS is not meant to be defined to empty. It is meant to
be undefined or defined to 0 (both
Hi Jim,
Have you considered making gnulib-tool diagnose (I'd prefer failure)
such redundant dependencies?
gnulib-tool already gives a warning about this. You insisted on it, years ago.
Both changes are fine. Thanks.
Applied.
Bruno
Bruno Haible wrote:
Hi Jim,
Have you considered making gnulib-tool diagnose (I'd prefer failure)
such redundant dependencies?
gnulib-tool already gives a warning about this. You insisted on it, years ago.
I should have guessed.
Then how about an option to make any warning (or at least those
Jim Meyering wrote:
Then how about an option to make any warning (or at least those warnings)
be treated as an error? Otherwise, it's just too easy to overlook
the warnings in the reams of usual output.
If someone is not looking at the output and wants to see warnings nevertheless,
he can use
I noticed this while bootstrapping m4; ignore-value.h stood out like a sore
thumb when sorted among the list of system headers, when all other gnulib-
specific headers used by m4 used . Any objections to this patch? I did not
touch eealloc, although that is another module that could probably
Then how about an option to make any warning (or at least those warnings)
be treated as an error? Otherwise, it's just too easy to overlook
the warnings in the reams of usual output.
If someone is not looking at the output and wants to see warnings
nevertheless,
he can use the option
21 matches
Mail list logo
