Re: [PATCH] free: preserve errno

2020-12-18 Thread Bruno Haible
I wrote: > * lib/stdlib.in.h (free): New declaration. Oops, I forgot to check this declaration in the test suite. 2020-12-18 Bruno Haible free-posix: Add C++ declaration test. * tests/test-stdlib-c++.cc (free): New declaration. diff --git a/tests/test-stdlib-c++.cc b/t

free-posix: New module, renamed from 'free'

2020-12-18 Thread Bruno Haible
Now that the 'free' module deals only with the errno issue, that is, with a difference between the ISO C and the (future) POSIX standards, a better name is 'free-posix'. Just like we have modules fprintf-posix system-posix (without having modules 'fprintf' and 'system'). 2020-12-18 Bruno Hai

free: Remove support for obsolete platforms

2020-12-18 Thread Bruno Haible
Paul Eggert wrote: > Also, define CANNOT_FREE_NULL if free cannot free NULL. We can simplify this. SunOS 4 stopped being one of our portability targets around 2005. The module 'free' was accordingly marked obsolete in 2008. This old cruft makes the module 'free' now more complex than needed. 202

Re: FYI: s390x: Require GCC 7.1 or later to build glibc due to __builtin_add_overflow

2020-12-18 Thread Paul Eggert
On 12/18/20 10:32 AM, Bruno Haible wrote: Even if I'm wrong, given that GCC 5 and 6 are in decreasing use, I feel it's better to err on the safe side. Thanks, I've followed up on this in the glibc side, here: https://sourceware.org/pipermail/libc-alpha/2020-December/120876.html

Re: FYI: s390x: Require GCC 7.1 or later to build glibc due to __builtin_add_overflow

2020-12-18 Thread Bruno Haible
Paul Eggert wrote: > > Thanks for the heads-up. Indeed, what I understand from this bug report is > > that > > the condition code handling (required by __builtin_add_overflow on many > > platforms) was unreliable before 2017-01-27, i.e. in GCC versions < 7. > > So, gnulib shouldn't make use of it

[PATCH] intprops: port to ICC 2021.1 Beta 20201112

2020-12-18 Thread Paul Eggert
* lib/intprops.h (_GL_HAS_BUILTIN_OVERFLOW_P): Port to ICC 2021.1, which has a non-working __builtin_mul_overflow_p. Add a clause for other compilers that claim to support __builtin_mul_overflow_p, since we might as well find out about slackers other than Clang and ICC. (INT_MULTIPLY_WRAPV): ICC 20

Re: backupfile.c missing feature check for fpathconf

2020-12-18 Thread Adrian Ebeling
On 18.12.20 17:48, Bruno Haible wrote: lib/backupfile.c contains a feature check for pathconf() (HAVE_PATHCONF), but later on fpathconf() is used without a check. Therefore, the file can't be compiled on systems that don't have pathconf. On which system is this? In our experience, the systems

Re: FYI: s390x: Require GCC 7.1 or later to build glibc due to __builtin_add_overflow

2020-12-18 Thread Paul Eggert
On 12/18/20 4:14 AM, Bruno Haible wrote: Thanks for the heads-up. Indeed, what I understand from this bug report is that the condition code handling (required by __builtin_add_overflow on many platforms) was unreliable before 2017-01-27, i.e. in GCC versions < 7. So, gnulib shouldn't make use of

Re: backupfile.c missing feature check for fpathconf

2020-12-18 Thread Bruno Haible
Hi Adrian, Adrian Ebeling wrote: > lib/backupfile.c contains a feature check for pathconf() > (HAVE_PATHCONF), but later on fpathconf() is used without a check. > Therefore, the file can't be compiled on systems that don't have pathconf. On which system is this? In our experience, the systems tha

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

2020-12-18 Thread Paul Eggert
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 am not sure if it would be better to adapt my original patchset to use scratch_buffer (which seems originally a better idea) or if we can work towards f

backupfile.c missing feature check for fpathconf

2020-12-18 Thread Adrian Ebeling
Hi, lib/backupfile.c contains a feature check for pathconf() (HAVE_PATHCONF), but later on fpathconf() is used without a check. Therefore, the file can't be compiled on systems that don't have pathconf. Please add a check for fpathconf(), similar to the pathconf() check in lines 44-46. Regards,

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

2020-12-18 Thread Adhemerval Zanella
On 18/12/2020 09:30, Adhemerval Zanella wrote: > > > On 17/12/2020 07:55, Paul Eggert wrote: >> On 12/11/20 5:03 AM, Adhemerval Zanella wrote: >> >>> I have sent a similar fix to reviews for this very issue for glibc >>> (which is based on the canonicalize-lgpl) along with other fixes that >>>

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

2020-12-18 Thread Adhemerval Zanella
On 17/12/2020 07:55, Paul Eggert wrote: > On 12/11/20 5:03 AM, Adhemerval Zanella wrote: > >> I have sent a similar fix to reviews for this very issue for glibc >> (which is based on the canonicalize-lgpl) along with other fixes that >> you might take a look at [1]. > > Thanks, I looked at tha

Re: FYI: s390x: Require GCC 7.1 or later to build glibc due to __builtin_add_overflow

2020-12-18 Thread Bruno Haible
Stefan Liebler wrote: > just as information, I've committed the glibc patch > > "s390x: Require GCC 7.1 or later to build glibc." > http://sourceware.org/git/?p=glibc.git;a=commit;h=844b4d8b4b937fe6943d2c0c80ce7d871cdb1eb5 > > as if build with gcc 6.5.0, __builtin_add_overflow incorrectly detects

FYI: s390x: Require GCC 7.1 or later to build glibc due to __builtin_add_overflow

2020-12-18 Thread Stefan Liebler
Hi gnulib-developers, just as information, I've committed the glibc patch "s390x: Require GCC 7.1 or later to build glibc." http://sourceware.org/git/?p=glibc.git;a=commit;h=844b4d8b4b937fe6943d2c0c80ce7d871cdb1eb5 as if build with gcc 6.5.0, __builtin_add_overflow incorrectly detects overflow o