RE : isatty() and errno

isatty should be pached under windows. Try isatty(/dev/null) Bastien Le 8 nov. 2011 00:53, "Eric Blake" a écrit : Our assumption in ptsname.c is that isatty() will set errno on invalid fds. But this is not required by POSIX, and indeed not enforced by Solaris 10; that is, our ptsname.c impleme

[PATCH 1/4] ptsname_r: new module

For now, this replacement focuses solely on compilation compatibility, and assumes that isatty() and ttyname_r() work on a master side pty; if this assumption fails, or if thread-safety is also required, then a later patch can follow the lead of strerror_r.c in wrapping the system ptsname() with a

[PATCH 3/4] ptsname_r: port to Solaris

Blindly implementing ptsname_r from scratch using the glibc algorithm fails on Solaris 10, where isatty() doesn't set errno when returning 0. While not thread-safe, wrapping ptsname() instead of implementing things ourselves has better results. * m4/ptsname_r.m4 (gl_FUNC_PTSNAME_R): Also check fo

[PATCH 4/4] ptsname_r: work around FreeBSD issue

POSIX does not require ptsname() to set errno on failure, and FreeBSD 8.2 leaves it unchanged. * lib/ptsname_r.c (ptsname_r): Guarantee errno on ptsname failure. * doc/posix-functions/ptsname.texi (ptsname): Document the issue. Signed-off-by: Eric Blake --- ChangeLog|

[PATCH 2/4] ptsname_r-tests: new test module

For now, this tests glibc behavior of both returning an errno value as well as setting errno; we may need to relax that. * modules/ptsname_r-tests: New module. * tests/test-ptsname_r.c: New file. Signed-off-by: Eric Blake --- ChangeLog |4 + modules/ptsname_r-tests | 14 +++

[PATCH 0/4] add ptsname_r

I propose adding ptsname_r; I'll push this in 72 hours if I don't get any feedback. [Among other things, I would expect feedback to include whether ptsname_r should drag in locking, like our strerror_r module, or just assume that no one else uses ptsname, like our gmtime_r replacement in our time_

isatty() and errno

Our assumption in ptsname.c is that isatty() will set errno on invalid fds. But this is not required by POSIX, and indeed not enforced by Solaris 10; that is, our ptsname.c implementation was copied from glibc, and was making an assumption only guaranteed by glibc's implementation choice for i

Re: isatty() and errno

On 11/07/2011 04:18 PM, Eric Blake wrote: Our assumption in ptsname.c is that isatty() will set errno on invalid fds. But this is not required by POSIX, and indeed not enforced by Solaris 10; that is, our ptsname.c implementation was copied from glibc, and was making an assumption only guaranteed

Re: gnutls missing largefile support

Nikos Mavrogiannopoulos wrote: > We only use file > support to open configuration files, or files containing certificates, > files that if they exceed 2GB then something is wrong. Indeed, after taking a look at gnutls-3.0.6/lib/opencdk/keydb.[hc] I can even tell you what will be wrong: If someone

[PATCH] grantpt: fix typo

* lib/stdlib.in.h (grantpt): Check correct function. Signed-off-by: Eric Blake --- ChangeLog |3 +++ lib/stdlib.in.h |2 +- 2 files changed, 4 insertions(+), 1 deletions(-) diff --git a/ChangeLog b/ChangeLog index 820decf..1761f56 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,

Re: math error reporting

Bastien ROUCARIES wrote: > msvc is compliant but a the microsoft taste > https://www.securecoding.cert.org/confluence/display/seccode/FLP03-C.+Detect+and+handle+floating+point+errors Thanks, this was helpful. I wasn't aware of this _controlfp_s function. Bruno -- In memoriam Hannah Szenes

Re: gnulib largefile support

On 11/07/2011 03:02 PM, Nikos Mavrogiannopoulos wrote: On 11/07/2011 10:52 PM, Bruno Haible wrote: [CCing Ludovic. The thread starts at ] Nikos Mavrogiannopoulos wrote: Wouldn't it be better to add an option to enable it only

Re: gnulib largefile support

Hi Bruno, Bruno Haible skribis: > You are confusing trigger and cause. While a gnulib update *triggered* > a malfunction [1], the *cause* was a guile bug [2], and it was fixed in > guile 2.0 already two years ago [3]. I agree about the cause being a Guile bug. That said, if the change to _FILE

Re: gnulib largefile support

On 11/07/2011 10:52 PM, Bruno Haible wrote: > [CCing Ludovic. The thread starts at > ] > Nikos Mavrogiannopoulos wrote: >> Wouldn't it be better to add an option to enable >> it only in systems where it is desirable? > Like Eric s

Re: gnulib largefile support

[CCing Ludovic. The thread starts at ] Nikos Mavrogiannopoulos wrote: > Wouldn't it be better to add an option to enable > it only in systems where it is desirable? Like Eric said, large file support *is* desirable on all platfo

Re: gnulib largefile support

Paul Eggert wrote: > Well, I can imagine a small embedded system where files and > streams cannot possibly be larger than 2 GiB. Yes, such as my router here: prompt> df Filesystem 1k-blocks Used Available Use% Mounted on /dev/ram0 4475 3487 988 78% / /de

Re: gnulib largefile support

On 11/07/11 12:38, Eric Blake wrote: > Large file support IS desirable on all platforms. Well, I can imagine a small embedded system where files and streams cannot possibly be larger than 2 GiB. On such a system it might be reasonable to use gnulib modules in small-file mode. If this is a real i

Re: gnulib largefile support

On 11/07/2011 01:30 PM, Nikos Mavrogiannopoulos wrote: Hello, After adding few modules to gnulib the gnutls-guile library had few failures in testsuite in some arm and mips systems. After some investigation[0] it seems that the issue was that the new gnulib modules were depending on largefile m

gnulib largefile support

Hello, After adding few modules to gnulib the gnutls-guile library had few failures in testsuite in some arm and mips systems. After some investigation[0] it seems that the issue was that the new gnulib modules were depending on largefile module which defines a 64-bit off_t even on 32-bit systems.

[PATCH] maint: silence new syntax check

* top/maint.mk (sc_prohibit_dirent_without_use): Add missing @. Signed-off-by: Eric Blake --- I noticed that the latest gnulib was rather noisy at syntax check. ChangeLog|5 + top/maint.mk |2 +- 2 files changed, 6 insertions(+), 1 deletions(-) diff --git a/ChangeLog b/ChangeL

Re: math error reporting

On Sun, Nov 6, 2011 at 8:36 PM, Bruno Haible wrote: > Hi, > > Considering how gnulib-implemented functions should do their > error reporting, I ran a test program to see how the various systems > implement math errors. > > ==

Re: math error reporting

>OSF/1 5.1:0, errno ( incomplete) With the module 'fpieee', this becomes OSF/1 5.1:NaN, errno ( incomplete) So, we are able to support IEEE NaN on all platforms. Bruno -- In memoriam Hannah Szenes