Re: careadlinkatcwd's signature

2011-06-05 Thread Paul Eggert
> 2011-06-05 Bruno Haible > > careadlinkat: Avoid mismatch between ssize_t and int. > * lib/careadlinkat.h (careadlinkatcwd): Declare as a function always. > * lib/careadlinkat.c (careadlinkatcwd): Define always. Thanks, that patch looks good. Could you please install it?

ansi-c++-opt: interoperability with libtool

2011-06-05 Thread Bruno Haible
GNU gettext uses the gl_PROG_ANSI_CXX and also uses LT_LANG([C++]) On a machine with a broken C++ compiler and no working C++ preprocessor, then despite of configure option --disable-c++, configure bails out like this: checking whether we are using the GNU C++ compiler... no checking whether

Re: [PATCH] getopt: for ambiguous options, enumerate the possibilities.

2011-06-05 Thread James Youngman
On Sat, Jun 4, 2011 at 2:34 PM, Bruno Haible wrote: > Another nit is that "gcc -Wall" gives a warning (seen on OpenBSD 4.9): > >  getopt.c: In function '_getopt_internal_r': >  getopt.c:491: warning: unused variable 'ambig' > > I'm fixing it like this: Thanks for spotting (and fixing) that. Jame

Re: Syntax check prohibit_doubled_word: false positives on non-English text

2011-06-05 Thread James Youngman
Thanks for your comments. I had fixed the Texinfo problems substantially as you suggested, but I opted to entirely exempt the .po files from the check. More generally, the sntax-check rules discovered a number of potential improvements for findutils, including unused headers, opportunities to us

acl: test failure on AIX

2011-06-05 Thread Bruno Haible
On AIX 7.1, an ACL test fails: ./test-copy-acl.sh[5]: 12189730 Segmentation fault(coredump) FAIL: test-copy-acl.sh It's due to a bug in the 'test-sameacls' program. This fixes it. Thanks to Jørn Amundsen, who reported it already in

pipe-filter-ii: fix test failure

2011-06-05 Thread Bruno Haible
On AIX 5.1, 7.1 and IRIX 6.5, the pipe-filter-ii test fails like this: ./test-pipe-filter-ii1: write to tr subprocess failed: Resource temporarily unavailable FAIL: test-pipe-filter-ii1.sh This fixes it. 2011-06-05 Bruno Haible pipe-filter-ii: Fix test failure on AIX and IRIX.

realpath on AIX 7.1

2011-06-05 Thread Bruno Haible
Hi Eric, On AIX 7.1 I'm seeing this test failure: test-canonicalize-lgpl.c:69: assertion failed FAIL: test-canonicalize-lgpl The reason is apparently a bug in realpath. Here's a test case: #include #include #

localename on AIX 7

2011-06-05 Thread Bruno Haible
On AIX 7.1, I get a link error: xlc -D_ALL_SOURCE -O -g -o test-localename test-localename.o ../gllib/libgnu.a ld: 0711-317 ERROR: Undefined symbol: .pthread_mutex_lock ld: 0711-317 ERROR: Undefined symbol: .pthread_mutex_unlock ld: 0711-345 Use the -bloadmap or -bnoquiet option to o

error: avoid gcc warning

2011-06-05 Thread Bruno Haible
On OSF/1 5.1 with gcc, I'm seeing this warning: error.c:170: warning: implicit declaration of function 'strerror_r' It's because /usr/include/string.h on this system does not always declare strerror_r (only if _REENTRANT or _THREAD_SAFE is defined). This fixes it (without requiring the 'strerro

setenv, unsetenv warnings

2011-06-05 Thread Bruno Haible
On OSF/1 5.1 with gcc, I'm seeing these warnings: setenv.c:366: warning: implicit declaration of function 'setenv' unsetenv.c:116: warning: implicit declaration of function 'unsetenv' This fixes it. 2011-06-05 Bruno Haible setenv: Avoid gcc warning. * lib/setenv.c (seten

sys_select: tweak for AIX

2011-06-05 Thread Bruno Haible
On AIX 7.1 with gcc, I'm seeing this warning: test-sys_select.c:32: warning: implicit declaration of function 'memset' test-sys_select.c:32: warning: incompatible implicit declaration of built-in function 'memset' This should fix it. 2011-06-05 Bruno Haible sys_select: Ensure m

careadlinkatcwd's signature

2011-06-05 Thread Bruno Haible
Hi Paul, On AIX 7.1 with gcc, I get this warning: areadlink.c:38: warning: passing argument 6 of 'careadlinkat' from incompatible pointer type The reason is obviously that 'readlinkat' has a return type of 'int', but careadlinkat() expects a function pointer that has a return type of ssize_t.

Re: Syntax check prohibit_doubled_word: false positives on non-English text

2011-06-05 Thread Jim Meyering
James Youngman wrote: > I had an interesting failure from the prohibit_doubled_word syntax check: > > prohibit_doubled_word_RE_ ?= \ > /\b(then?|[iao]n|i[fst]|but|f?or|at|and|[dt]o)\s+\1\b/gims > prohibit_doubled_word_ = \ > -e 'while ($(prohibit_d