Ports failing with -fno-common with clang 9/gcc 9

Dewayne Geraghty dewayne at heuristicsystems.com.au
Wed Jun 24 00:36:45 UTC 2020


On 24/06/2020 7:00 am, Kyle Evans wrote:
> On Mon, Jun 22, 2020 at 1:35 AM Tobias Kortkamp <tobik at freebsd.org> wrote:
>>
>> On Thu, Apr 30, 2020, at 14:56, Kyle Evans wrote:
>>> In any event, I would urge folks to be proactive and identify this
>>> stuff, reporting issues upstream and spreading awareness of the
>>> impending default change for those projects that may not already be
>>> actively aware.
>>>
>>> On a closing note, I'm just going to kinda drop these patches here for
>>> anyone that's willing/able to help make this a collective effort to
>>> identify/fix/report problems here; they backport the default
>>> -fno-common patch from future-LLVM11 to the base system compiler on a
>>> system near you:
>>>
>>> HEAD: https://people.freebsd.org/~kevans/llvm-fnocommon-head.diff
>>
>> Can you ask for an exp-run with this patch applied?  There needs
>> to be a comprehensive list of failing ports for people to be able
>> to work on this.  We see some failing ports because of default
>> -fno-common in GCC 10 [1] already.
>>
>> [1] https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=246700
> 
> I can, but my only problem in doing so is that I cannot take
> responsibility for following up on the issues discovered. I have some
> outstanding exp-run that I've dropped the ball on, I'm a bit wobbly on
> personally requesting one for this unless some group of people
> can/will offer to make sure the issues are fixed so that the exp-run
> can actually be completed in a reasonable timeframe.
> 
> Thanks,
> 
> Kyle Evans
> _______________________________________________
> freebsd-ports at freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-ports
> To unsubscribe, send any mail to "freebsd-ports-unsubscribe at freebsd.org"
> 

Kyle, at best the maintainers should be advised to either:
1. add CFLAGS+=-fcommon to their problematic port's Makefile
2. pursue their respective upstream to address the issue as its going to
"endemic" once clang10+ and gcc10+ become widespread.

I don't see this as a problem for you alone to carry.

I built all my ports using -fno-common in make.conf, and added
CFLAGS+=-fcommon to the problem ports.  It can be a simple fix, or you
could apply the same kind of patching (llvm-fnocommon-head.diff) to
gcc10+ and llvm10+ in ports to delay "the problem" being experienced for
FreeBSD folk, but... (in my view, -fno-common is a good thing)

PS if I had a send-pr functionality, I would've happily raised the 47
PR's (of the 1400+ ports built/used) ;)


More information about the freebsd-ports mailing list