On mingw, getcwd(NULL,1) succeeds, even though glibc documents that
with a non-zero size, the allocation will not exceed that many bytes.
On mingw, getcwd has the wrong signature. However, we don't have
to check for this if anything else triggers the replacement.
Also, fix a type bug that crept
In order to eventually add socket nonblocking status into
fd shadowing, it is useful to move the existing fchdir name
shadowing into a common file.
* modules/fd-hook (Depends-on): Add dosname, dup2,
filenamecat-lgpl, getcwd-lgpl, malloc-posix, realloc-posix,
stdbool, strdup-posix.
* modules/fchdir
On 04/27/2011 11:58 AM, Eric Blake wrote:
> On 04/26/2011 05:35 PM, Eric Blake wrote:
>> * tests/test-getcwd-lgpl.c: New file, taken from...
>> * tests/test-getcwd.c: ...old contents. Rewrite this file to
>> repeat long path stress tests from m4 probe.
>> * modules/getcwd-lgpl-tests: New module.
>
Eric Blake wrote:
> > + dies from a fatal signal such as SIGQUIT, SIGKILL, or SIGABRT, SIGSEGV,
> > + SIGBUS, SIGILL, SIGFPE, */
>
> The grammar is a bit off, though. Should the "or" be moved to the end,
SIGQUIT, SIGKILL are signals forceful kill signals from outside the program,
whereas SI
Rely on the new xalloc-oversized module to avoid requiring
xalloc-die for functions documented as returning NULL on
potential allocation overflow.
* lib/hash.c (includes): Adjust includes.
* lib/mgetgroups.c (includes): Likewise.
(xgetgroups): Move...
* lib/xgetgroups.c: ...to new file.
* modules/
Ever since commit 2392b18099, openat and fdopendir no longer drag in
xalloc-die through indirect dependencies, so these tests no longer
need $(LIBINTL).
* modules/areadlinkat-tests (Makefile.am): Drop stale LDADD.
* modules/dirent-safer-tests (Makefile.am): Likewise.
* modules/fdopendir-tests (Mak
Due to inline functions, mere inclusion of xalloc.h can result
in a link dependency on xalloc_die() on some platforms. However,
there are several modules that want to use just xalloc_oversized
in order to short-circuit the potential to call xalloc_die.
Splitting the macro into a new header and mod
utimecmp's only use of malloc was for a cache; we can always bypass
the cache in low memory, and thus avoid dragging in xalloc-die.
* lib/utimecmp.c (utimecmp): Work even if hash table cache fails
due to memory pressure.
* modules/utimecmp (Depends-on): Drop xalloc.
Signed-off-by: Eric Blake
---
On 04/27/2011 04:37 PM, Bruno Haible wrote:
> Eric Blake wrote:
>>> +# On some hosts, mkstemp creates files with mode 0666, which is a security
>>> +# problem.
>>
>> Also, it was valid in POSIX 2001, but violates POSIX 2008.
>
> What about mkstemps? Does it have the same problem as mkstemp?
In al
Eric Blake wrote:
> > +# On some hosts, mkstemp creates files with mode 0666, which is a security
> > +# problem.
>
> Also, it was valid in POSIX 2001, but violates POSIX 2008.
What about mkstemps? Does it have the same problem as mkstemp?
The answer is fortunately "no". The platforms x symbols
On 04/27/2011 02:08 AM, Bruno Haible wrote:
> Eric Blake wrote:
>> * modules/xgetgroups: New file, split from...
>> * modules/mgetgroups: ...here.
>
> A NEWS entry should explain this backward incompatible change.
> And coreutils/bootstrap.conf will need an update.
Good catch, although the coreut
On 27/04/11 16:47, J.T. Conklin wrote:
> First some background.
>
> The continuous integration build framework I maintain at work uses
> timestamp files for "last_build", "last_success", etc. The contents
> of each file is the timestamp in ISO 8601 format (the inode mtime is
> set to the same v
On MacOS X 10.5, mkstemps is declared in , not in . This
leads to a failure in a testdir created by
$ ./gnulib-tool --create-testdir --dir=... --with-tests --with-c++-tests
mkstemps
g++ -DHAVE_CONFIG_H -I. -DGNULIB_STRICT_CHECKING=1 -I. -I. -I.. -I./..
-I../gllib -I./../gllib -Wall -MT test
On 04/27/2011 03:29 PM, Bruno Haible wrote:
> Just "less likely"?! The code in module 'clean-temp' goes through all kinds of
> hoops with signal handlers and 'volatile' variables to make _sure_ that the
> temporary files are cleaned up. This has also been tested as part of the
> 'msgfmt' program.
>
Jim Meyering writes:
>> I'm interested in comments on both the idea of supporting such time
>> formats, and whether there is a better implementation.
>
> Thank you! This patch looks fine.
>
> As Paul mentioned, a complete change would be welcome,
> if you can fill out paperwork. Considering that
> 2011-04-27 Reuben Thomas
> and Eric Blake
>
> + mkstemp: replace if system version uses wrong permissions
> + * m4/mkstemp.m4 (gl_FUNC_MKSTEMP): Add test for non-owner
> + read/write mode bits set in file created by mkstemp.
> + * doc/posix-functions/mkstemp.texi (mks
Eric Blake wrote:
> I also think doc/posix-functions/mkstemp.texi could
> use a similar comment. I added that, and pushed the modified result.
Thanks, but...
"The gnulib module clean-temp can create temporary files that are less
likely to be left behind on signals such as SIGINT."
Just "le
On 04/27/2011 12:24 AM, Jim Meyering wrote:
>> fdutimensat-tests still pulls in xalloc, via utimecmp (which
>> calls xalloc_die if it can't manage a hash table). So this
>> patch isn't quite perfect yet, but getting closer. Any ideas
>> on whether it would be possible to make utimecmp work withou
On 04/26/2011 05:35 PM, Eric Blake wrote:
> On Linux, with a forced ac_cv_func_fchdir=no, the test failed
> because dup2 was not replaced, all because the shell variable
> ac_cv_func_dup2 had not been set yet.
>
> Meanwhile, computing the canonical pathname of a directory on
> mingw was rather exp
On 04/26/2011 05:35 PM, Eric Blake wrote:
> * modules/linkat (Depends-on): Require getcwd-lgpl.
> * modules/renameat (Depends-on): Likewise.
6 and 7 were both pretty uncontroversial; I've now pushed them.
--
Eric Blake ebl...@redhat.com+1-801-349-2682
Libvirt virtualization library http://
FYI, here's the log:
http://hydra.nixos.org/build/1058543
Here's the relevant snippet from the log:
building test-lock.log
/bin/sh: line 1: 83575 Abort trap EXEEXT='' srcdir='.'
LOCALE_FR='fr_FR.ISO8859-1' LOCALE_FR_UTF8='fr_FR.UTF-8' LIBSIGSEGV=''
LOCALE_FR='fr_FR.ISO8859-1' LO
On 04/27/2011 10:03 AM, Jim Meyering wrote:
> Eric Blake wrote:
>> On 04/27/2011 01:37 AM, Jim Meyering wrote:
>>> Eric Blake wrote:
save-cwd needs either a working fchdir or a working getcwd(NULL).
The only platform where getcwd fixes problems not solved by
getcwd-lgpl is Linux, but
J.T. Conklin wrote:
> First some background.
>
> The continuous integration build framework I maintain at work uses
> timestamp files for "last_build", "last_success", etc. The contents
> of each file is the timestamp in ISO 8601 format (the inode mtime is
> set to the same value, but thats beside
On 04/26/2011 05:35 PM, Eric Blake wrote:
> * tests/test-getcwd-lgpl.c: New file, taken from...
> * tests/test-getcwd.c: ...old contents. Rewrite this file to
> repeat long path stress tests from m4 probe.
> * modules/getcwd-lgpl-tests: New module.
> * modules/getcwd-tests (Depends-on): Depend on
On 04/26/2011 06:54 PM, Bruno Haible wrote:
> Hi Eric,
>
> Now the code is clear enough that it can be understood :-) I have three
> remarks:
The first two were trivial fixes.
>> AC_DEFUN([gl_PREREQ_GETCWD],
>> [
>>AC_REQUIRE([gl_USE_SYSTEM_EXTENSIONS])
>>AC_REQUIRE([gl_CHECK_TYPE_STR
On 04/27/2011 10:24 AM, Reuben Thomas wrote:
> On 27 April 2011 16:33, Eric Blake wrote:
>>
>> Here's the final version I'd like to push. It's modified a bit from
>> your version; are you okay with it still being in your name?
>
> Thanks very much for polishing and testing this patch; I don't mi
On 04/27/11 08:47, J.T. Conklin wrote:
> I'm interested in comments on both the idea of supporting such time
> formats, and whether there is a better implementation.
Thanks. I like the idea of supporting these formats, and know of
no better implementation. But we'd need to change
parse-datetime.
Detected on Fedora 13 with -Wshadow.
* lib/passfd.c (sendfd, recvfd): Avoid shadowing names.
Reported by Laine Stump.
Signed-off-by: Eric Blake
---
ChangeLog|6 ++
lib/passfd.c |8
2 files changed, 10 insertions(+), 4 deletions(-)
diff --git a/ChangeLog b/ChangeLog
ind
On 27 April 2011 16:33, Eric Blake wrote:
>
> Here's the final version I'd like to push. It's modified a bit from
> your version; are you okay with it still being in your name?
Thanks very much for polishing and testing this patch; I don't mind if
my name is still on it.
--
http://rrt.sc3d.org
On 04/27/2011 09:23 AM, J.T. Conklin wrote:
> For backwards compatibility, the 4.4BSD join used by NetBSD supports a
> "-a" option as if it was "-a 1 -a 2". Unfortunately, this was done in
> such a way that "-a 1" and "-a 2" need to be passed as "-a1" or "-a2".
Thanks. That's a violation of POSI
Eric Blake wrote:
> On 04/27/2011 01:37 AM, Jim Meyering wrote:
>> Eric Blake wrote:
>>> save-cwd needs either a working fchdir or a working getcwd(NULL).
>>> The only platform where getcwd fixes problems not solved by
>>> getcwd-lgpl is Linux, but fchdir works there.
>>>
>> Hi Eric,
>>
>> Unfortun
First some background.
The continuous integration build framework I maintain at work uses
timestamp files for "last_build", "last_success", etc. The contents
of each file is the timestamp in ISO 8601 format (the inode mtime is
set to the same value, but thats besides the point).
I used this fo
For backwards compatibility, the 4.4BSD join used by NetBSD supports a
"-a" option as if it was "-a 1 -a 2". Unfortunately, this was done in
such a way that "-a 1" and "-a 2" need to be passed as "-a1" or "-a2".
---
ChangeLog |5 +
gnulib-tool |2 +-
2 files changed, 6 insertions(
From: Reuben Thomas
* m4/mkstemp.m4 (gl_FUNC_MKSTEMP): Add test for non-owner
read/write mode bits set in file created by mkstemp.
* doc/posix-functions/mkstemp.texi (mkstemp): Document the fix.
Signed-off-by: Eric Blake
---
Here's the final version I'd like to push. It's modified a bit from
On 27 April 2011 15:17, Eric Blake wrote:
>
> Can you convince your mailer to send patches inline, or at least with
> text/plain MIME type?
I think my only options are to rename the patch file (which will
change the mailer's diagnosed MIME type) or to manually paste the
patch into the body of the
On 04/27/2011 07:52 AM, Reuben Thomas wrote:
> On 27 April 2011 14:00, Eric Blake wrote:
>> On 04/27/2011 04:05 AM, Reuben Thomas wrote:
>>> On 27 April 2011 03:24, Bruno Haible wrote:
> Does setting a 0600 umask (as glibc does) sound like a good thing to
> add to the mkstemp-safer functi
On 04/27/2011 07:32 AM, Reuben Thomas wrote:
> On 27 April 2011 14:00, Eric Blake wrote:
>> On 04/27/2011 04:05 AM, Reuben Thomas wrote:
>>
>>> However, this suggests that the gnulib documentation for mkstemp
>>> should direct the programmer to clean-temp, something like:
>>>
>>> "If you are creat
On 27 April 2011 14:00, Eric Blake wrote:
> On 04/27/2011 04:05 AM, Reuben Thomas wrote:
>> On 27 April 2011 03:24, Bruno Haible wrote:
Does setting a 0600 umask (as glibc does) sound like a good thing to
add to the mkstemp-safer functions?
>
> Setting umask() is bad for other reasons -
On 27 April 2011 14:00, Eric Blake wrote:
> On 04/27/2011 04:05 AM, Reuben Thomas wrote:
>
>> However, this suggests that the gnulib documentation for mkstemp
>> should direct the programmer to clean-temp, something like:
>>
>> "If you are creating temporary files which will later be removed, cons
On 04/27/2011 04:05 AM, Reuben Thomas wrote:
> On 27 April 2011 03:24, Bruno Haible wrote:
>>> Does setting a 0600 umask (as glibc does) sound like a good thing to
>>> add to the mkstemp-safer functions?
Setting umask() is bad for other reasons - it is global state, and
should not be done in mult
On 04/27/2011 01:37 AM, Jim Meyering wrote:
> Eric Blake wrote:
>> save-cwd needs either a working fchdir or a working getcwd(NULL).
>> The only platform where getcwd fixes problems not solved by
>> getcwd-lgpl is Linux, but fchdir works there.
>>
> Hi Eric,
>
> Unfortunately, that change would in
On 27 April 2011 03:24, Bruno Haible wrote:
>> Does setting a 0600 umask (as glibc does) sound like a good thing to
>> add to the mkstemp-safer functions?
>
> The application can call chmod() on its own. Why push this
> into mkstemp-safer?
I'm not suggesting chmod (that would surely create a race
Eric Blake wrote:
> * modules/xgetgroups: New file, split from...
> * modules/mgetgroups: ...here.
A NEWS entry should explain this backward incompatible change.
And coreutils/bootstrap.conf will need an update.
Bruno
Eric Blake wrote:
> save-cwd needs either a working fchdir or a working getcwd(NULL).
> The only platform where getcwd fixes problems not solved by
> getcwd-lgpl is Linux, but fchdir works there.
>
> * modules/save-cwd (Depends-on): Use getcwd-lgpl.
>
> Signed-off-by: Eric Blake
> ---
> ChangeLog
44 matches
Mail list logo
