[tz] [PROPOSED] Fix C23-related conformance bug
Steffen Nurpmeso
steffen at sdaoden.eu
Thu Dec 1 21:04:46 UTC 2022
Paul Eggert wrote in
<20221201183026.262113-1-eggert at cs.ucla.edu>:
|Problem reported by Houge Langley for ‘gcc -std=gnu99’ in:
|https://bugs.gentoo.org/show_bug.cgi?id=883719
|* NEWS: Mention this.
|* date.c, localtime.c, private.h, zdump.c, zic.c:
|Use ATTRIBUTE_* at the start of function declarations,
|not later (such as after the keyword ‘static’).
|This is required for strict conformance to C23.
What i wanted to note all the time since the obsoletion of ISO
C89 (that i regret as i am an -ansi user for C as well as C++
(in development mode the MUA i maintain randomly orders the
tried --std='s "__s1=c89 __s2=c99 __s3=c11 __s4=c18 __s5=c2x"))
is that the worst thing in porting were and are the bugs
(including miscompilations) of specific compiler( version)s, not
only, but also optimization related. And all the peculiarities
to get over overzealous warnings or errors in between different
compiler/ versions, and the ways you can work around them. (For
example _Pragma only lived as #pragma around Y2K in what i could
get my hands on, and could not simply be injected here and there
at will. You all know this, of course.) (My MUA (still)
chooses c99 for non-development code because at one day
i stopped doing so grazy things as disabling specific warnings
for specific compilers in specific modes, and for example 64-bit
integers i always used.)
It may be beneficial, i am thinking about this now that i --
despite total political opposition to the war against North
Korea, Cuba, Iran, China and Russia, and whoever i have
forgotten -- am (back) at github again (to report issues and try
to upstream patches where no other way remains aka was asked for
explicitly) to make use of both the Coverity integration in
Github (it is a >1 GB download otherwise), as well as CI
integration.
This would compile on many different boxes, and some of the tests
make sense when spreaded (could be tailored, too).
This after seeing an OpenSSH commit flew by, their
.github/workflows/selfhosted.yml in
https://github.com/openssh/openssh-portable.git lists an
impressive number of VMs that by far numbers out my own VM list
(especially after i shrunk it a lot to get at BTRFS dup .. a bit
restricted here).
P.S.: i also have had contributions disabled from the start.
I concur very much.
--steffen
|
|Der Kragenbaer, The moon bear,
|der holt sich munter he cheerfully and one by one
|einen nach dem anderen runter wa.ks himself off
|(By Robert Gernhardt)
More information about the tz
mailing list