bug-groff
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[bug #55018] [PATCH] src/utils/indxbib/signal.c: define _POSIX_C_SOURCE


From: Ingo Schwarze
Subject: [bug #55018] [PATCH] src/utils/indxbib/signal.c: define _POSIX_C_SOURCE
Date: Mon, 20 Jan 2020 12:03:33 -0500 (EST)
User-agent: Mozilla/5.0 (X11; OpenBSD amd64; rv:71.0) Gecko/20100101 Firefox/71.0

Follow-up Comment #2, bug #55018 (project groff):

To me, this ticket looks rather bogus.

I have no idea what Bjarni is even doing, he doesn't show the compiler command
line he is running, nor does he show his config.h.

Using the standard C kill(2) function certainly does not require defining
arcane feature-test macros.  Just including <signal.h> is enough according to
the POSIX standard, see
https://pubs.opengroup.org/onlinepubs/9699919799/functions/kill.html - and the
file already includes that header.

So far, i failed to reproduce the problem, using clang version 8.0.1 on
OpenBSD-current/amd64.

Unfortunately, the groff build system hides compiler command lines from the
logs, so i cannot easily vary compiler options to try harder to reproduce.  It
only prints "CC src/utils/indxbib/signal.o" which is obviously not what it
actually does.  I strongly feel that obfuscation ought to be removed.  Such
obfuscation really makes debugging harder for no benefit whatsoever.

Either way, even if there is a problem at all, which i seriously doubt,
defining _POSIX_C_SOURCE to 1 sounds like an extremely bad idea.  That is
requesting a long obsolete standard, and i'm not aware that the devellopers
currently working on groff are expected to code according to that obsolete
standard.

If there is a problem, then the root cause needs to be explained and fixed,
not some cargo cult thrown in without even considering whether that may cause
other problems down the line.

By the way, on a Debian jessie box i have access to, the default value of
_POSIX_C_SOURCE when the application defines nothing appears to be 200809L,
which looks quite sane to me.

So it appears Bjarni has misconfigured his compiler in some way he doesn't
specify and maybe isn't even aware of.

Unless somebody can provide a proper explanation, i suggest closing this
ticket as invalid.

    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?55018>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/




reply via email to

[Prev in Thread] Current Thread [Next in Thread]