gmp revision to sync to Gnulib?

Karl, you and I seem to have differing "latest" versions of the GMP library[1][2][3]. I'm using the "gmp" branch from gmplib.org, which is the one first-listed in . Are you using some other branch? I have been using the "gmp" branch since it contains bugfixe

Re: [PATCH 3/5] Import filename.h from gnulib

On 12/31/20 3:13 PM, Joseph Myers wrote: Anything used in glibc libraries should be LGPLv2.1+. Thanks for catching that. I installed the attached patch into Gnulib so that glibc and Gnulib copies of filename.h can be identical. >From ec51ca6c9d59037df66afe0804394f66a2a8ff54 Mon Sep 17 00:00:0

Re: [PATCH 2/5] Import idx.h from gnulib

On 12/31/20 3:12 PM, Joseph Myers wrote: Anything used in glibc libraries should be LGPLv2.1+, not GPL. That was fixed in the most recently-published version of the patch: https://sourceware.org/pipermail/libc-alpha/2020-December/121178.html

Re: posixtm failure on Solaris

On 12/31/20 3:43 PM, Bruno Haible wrote: Since, as you said, it's not worth worrying about, I would guess that there is no need to update the documentation? Yes, the posixtm bug's not worth worrying about. However the underlying localtime issue might be worth a mention. I was feeling energetic

ptsname_r: Fix test failure on Solaris

The ptsname_r test fails on Solaris 11 OpenIndiana and Solaris 11 OmniOS: ../../gltests/test-ptsname_r.c:167: assertion 'result == 0' failed FAIL test-ptsname_r (exit status: 262) This patch fixes it, by removing the test on the major number of the device. This major number is (according to '

posixtm failure on Solaris

Hi Paul, On Solaris 11.3, Solaris OpenIndiana, and Solaris OmniOS, I'm seeing this test failure: 0101.00 mismatch (-: actual; +:expected) --62167132800 +-62167219200 12312359.59 return value mismatch: got 0, expected 1 FAIL test-posixtm (exit status: 1) It was already reported in 201

Re: [PATCH 3/5] Import filename.h from gnulib

On Thu, 24 Dec 2020, Adhemerval Zanella via Libc-alpha wrote: > diff --git a/include/filename.h b/include/filename.h > new file mode 100644 > index 00..4598fb1d63 > --- /dev/null > +++ b/include/filename.h > @@ -0,0 +1,110 @@ > +/* Basic filename support macros. > + Copyright (C) 2001-20

Re: [PATCH 2/5] Import idx.h from gnulib

On Thu, 24 Dec 2020, Adhemerval Zanella via Libc-alpha wrote: > + Copyright (C) 2020 Free Software Foundation, Inc. > + > + This program is free software; you can redistribute it and/or modify > + it under the terms of the GNU General Public License as published by > + the Free Software Fo

poll tests: Avoid test failure on AIX

Another POLLHUP test is failing on AIX, but here not for the sockets, but for the pipes. This is an AIX bug w.r.t. POSIX. Since we cannot easily work around it in Gnulib, this patch - documents it, - avoids a test failure (since I don't want to see it over and over again, now that it's docu

poll tests: Avoid test failure on BSD and Solaris systems

The poll test fails on *BSD and Solaris systems, already for years: Unconnected socket test... passed Connected sockets test... failed (expecting POLLHUP after shutdown) General socket test with fork... failed (expecting POLLHUP after shutdown) Pipe test... passed FAIL test-poll (exit st

Re: [PATCH 5/5] posix: Sync fnmatch with gnulib

On 12/30/20 12:15 PM, Adhemerval Zanella wrote: - ssize_t level; + size_t level; 'level' should be ptrdiff_t not ssize_t, for portability to (now-ancient, but still allowed by POSIX) hosts where ssize_t is 32 bits and size_t is 64 bits. -CHAR str[]; +CHAR str[FLEXIBLE_ARRAY_MEMB

Re: [PATCH 4/5] Sync flexmember.h with gnulib

Thanks, this patch looks fine.

Re: [PATCH 3/5] Sync intprops.h with gnulib

Thanks, this patch looks fine.

Re: [PATCH 2/5] posix: Sync glob code with gnulib

On 12/30/20 12:15 PM, Adhemerval Zanella wrote: It sync with gnulib commit 43ee1a6bf and fixes and use-after-free bug (gnulib commit 717766da8926e36). The main change over gnulib is: --- posix/glob.c +++ lib/glob.c @@ -59,6 +59,12 @@ # define readdir(str) __readdir64 (str) # define getpwnam

Re: [PATCH 1/5] posix: Sync regex code with gnulib

This patch looks good to me, thanks.

Re: memalign: Work around Solaris bug

On Thu, Dec 31, 2020 at 4:21 PM Bruno Haible wrote: > > On Solaris 11.4, Solaris OpenIndiana, Solaris OmniOS, I see this test failure: > > ../../gltests/test-memalign.c:51: assertion 'aligned2_blocks[i] != NULL' > failed > FAIL test-memalign (exit status: 262) > > Apparently memalign() does not s

memalign: Work around Solaris bug

On Solaris 11.4, Solaris OpenIndiana, Solaris OmniOS, I see this test failure: ../../gltests/test-memalign.c:51: assertion 'aligned2_blocks[i] != NULL' failed FAIL test-memalign (exit status: 262) Apparently memalign() does not support alignments 1 and 2 on this platform. This patch fixes it, by

unistd: Fix portability warnings

This patch fixes some missing update from when the modules 'fchownat' and 'unlinkat' were split off from module 'openat' in 2011. 2020-12-31 Bruno Haible unistd: Fix portability warnings. * lib/unistd.in.h (fchownat, unlinkat): Fix module name in warning. diff --git a/lib/uni

Re: [PATCH 1/6] canonicalize-lgpl: fix EOVERFLOW bug

Tom G. Christensen wrote: > FWIW lib/utime.c now fails to build on Solaris < 10 because of > 0003-stat-failing-with-EOVERFLOW-implies-existence.patch > (340dcbd879d050aebcd7081ec5f2208f0e2418b6) > > gcc -std=gnu11 -DHAVE_CONFIG_H -DEXEEXT=\"\" -DEXEEXT=\"\" -DNO_XMALLOC > -DEXEEXT=\"\" -I. -I..

Re: [PATCH 1/6] canonicalize-lgpl: fix EOVERFLOW bug

On 24/12/2020 21:10, Paul Eggert wrote: On 12/18/20 6:13 AM, Adhemerval Zanella wrote: The same tests I pointed out on BZ#24970 comment #2 still fails with gnulib version 0aa8ef424. I finally got some time free to look at this, and came up with some patches that should fix that problem, alon