threads / patch / 59656

patchdeveloper: remove gcc 12 workaround

Subject: [PATCH] developer: remove gcc 12 workaround

## tl;dr

4 messages between Apr 26, 2023 and Apr 26, 2023. Diffs are folded; open one to read it.

replies: 3people: 3as markdown or json

Carlo Marcelo Arenas Belón· Apr 26, 2023, 01:34 UTC · lore

Since 846a29afb0 (config.mak.dev: workaround gcc 12 bug affecting "pedantic" CI job, 2022-04-15), DEVELOPER mode has this workaround for the version of gcc that was released with Fedora 36.

That version of Fedora is about to be EOL and latest versions of the compiler don't have that bug anymore, so remove the workaround.

Tested not to trigger with latest gcc packages from the last 3 Fedora releases, including gcc 13 from Fedora 38.

Signed-off-by: Carlo Marcelo Arenas Belón <carenas@gmail.com>
---
 config.mak.dev | 5 -----
 1 file changed, 5 deletions(-)

Show changes to config.mak.dev +0 −5

diff --git a/config.mak.dev b/config.mak.dev
index 981304727c..6d07ef3692 100644
--- a/config.mak.dev
+++ b/config.mak.dev
@@ -94,9 +94,4 @@ endif
 endif
 endif
 
-# https://bugzilla.redhat.com/show_bug.cgi?id=2075786
-ifneq ($(filter gcc12,$(COMPILER_FEATURES)),)
-DEVELOPER_CFLAGS += -Wno-error=stringop-overread
-endif
-
 GIT_TEST_PERL_FATAL_WARNINGS = YesPlease

-- 
2.39.2 (Apple Git-143)

Eric Sunshine· Apr 26, 2023, 03:45 UTC · re: Carlo Marcelo Arenas Belón · lore

Re: [PATCH] developer: remove gcc 12 workaround

On Tue, Apr 25, 2023 at 9:47 PM Carlo Marcelo Arenas Belón <carenas@gmail.com> wrote:

Show 9 quoted lines

> Since 846a29afb0 (config.mak.dev: workaround gcc 12 bug affecting
> "pedantic" CI job, 2022-04-15), DEVELOPER mode has this workaround
> for the version of gcc that was released with Fedora 36.
>
> That version of Fedora is about to be EOL and latest versions of
> the compiler don't have that bug anymore, so remove the workaround.
>
> Tested not to trigger with latest gcc packages from the last 3
> Fedora releases, including gcc 13 from Fedora 38.

The commit message doesn't explain the benefit of removing this workaround. Is it because it's a maintenance burden? Or is it preventing some future planned improvement in this area? What is the justification for penalizing users who might be stuck on Fedora 36 for some reason or another?

(I ask these questions as a person who regularly is stuck on old platforms for which it is increasingly painful or impossible to build or install modern software.)

Show 16 quoted lines

> Signed-off-by: Carlo Marcelo Arenas Belón <carenas@gmail.com>
> ---
> diff --git a/config.mak.dev b/config.mak.dev
> index 981304727c..6d07ef3692 100644
> --- a/config.mak.dev
> +++ b/config.mak.dev
> @@ -94,9 +94,4 @@ endif
>  endif
>  endif
>
> -# https://bugzilla.redhat.com/show_bug.cgi?id=2075786
> -ifneq ($(filter gcc12,$(COMPILER_FEATURES)),)
> -DEVELOPER_CFLAGS += -Wno-error=stringop-overread
> -endif
> -
>  GIT_TEST_PERL_FATAL_WARNINGS = YesPlease

Carlo Arenas· Apr 26, 2023, 06:41 UTC · re: Eric Sunshine · lore

Re: [PATCH] developer: remove gcc 12 workaround

On Tue, Apr 25, 2023 at 8:45 PM Eric Sunshine <sunshine@sunshineco.com> wrote:

Show 16 quoted lines

>
> On Tue, Apr 25, 2023 at 9:47 PM Carlo Marcelo Arenas Belón
> <carenas@gmail.com> wrote:
> > Since 846a29afb0 (config.mak.dev: workaround gcc 12 bug affecting
> > "pedantic" CI job, 2022-04-15), DEVELOPER mode has this workaround
> > for the version of gcc that was released with Fedora 36.
> >
> > That version of Fedora is about to be EOL and latest versions of
> > the compiler don't have that bug anymore, so remove the workaround.
> >
> > Tested not to trigger with latest gcc packages from the last 3
> > Fedora releases, including gcc 13 from Fedora 38.
>
> The commit message doesn't explain the benefit of removing this
> workaround. Is it because it's a maintenance burden? Or is it
> preventing some future planned improvement in this area?

The workaround was added for the benefit of the CI and to prevent failures because the buggy warning will otherwise make it fail because of -Werror when Fedora 36 was released and our CI started using it.

Our CI doesn't need it anymore to avoid that issue.

> What is the
> justification for penalizing users who might be stuck on Fedora 36 for
> some reason or another?

I certainly didn't intend on penalizing any users, but I frankly suspect there aren't any, as the bug was fixed long ago, and it wouldn't affect them unless no gcc updates were ever done on their systems.

Carlo

Phillip Wood· Apr 26, 2023, 15:00 UTC · re: Carlo Arenas · lore

Re: [PATCH] developer: remove gcc 12 workaround

Hi Carlo

On 26/04/2023 07:41, Carlo Arenas wrote:

Show 23 quoted lines

> On Tue, Apr 25, 2023 at 8:45 PM Eric Sunshine <sunshine@sunshineco.com> wrote:
>>
>> On Tue, Apr 25, 2023 at 9:47 PM Carlo Marcelo Arenas Belón
>> <carenas@gmail.com> wrote:
>>> Since 846a29afb0 (config.mak.dev: workaround gcc 12 bug affecting
>>> "pedantic" CI job, 2022-04-15), DEVELOPER mode has this workaround
>>> for the version of gcc that was released with Fedora 36.
>>>
>>> That version of Fedora is about to be EOL and latest versions of
>>> the compiler don't have that bug anymore, so remove the workaround.
>>>
>>> Tested not to trigger with latest gcc packages from the last 3
>>> Fedora releases, including gcc 13 from Fedora 38.
>>
>> The commit message doesn't explain the benefit of removing this
>> workaround. Is it because it's a maintenance burden? Or is it
>> preventing some future planned improvement in this area?
> 
> The workaround was added for the benefit of the CI and to prevent
> failures because the buggy warning will otherwise make it fail because
> of -Werror when Fedora 36 was released and our CI started using it.
> 
> Our CI doesn't need it anymore to avoid that issue.

Our CI may not need it but what users of other distributions using gcc 12? The link[1] in the commit message for 846a29afb0 suggests that debian was also affected at that time so the bug was not a fedora specific. We could perhaps tighten the check if we know which specific versions of gcc 12 are affected but it is not clear that removing it entirely is a good idea.

Best Wishes

Phillip

[1] https://bugzilla.redhat.com/show_bug.cgi?id=2075786

Show 10 quoted lines

>> What is the
>> justification for penalizing users who might be stuck on Fedora 36 for
>> some reason or another?
> 
> I certainly didn't intend on penalizing any users, but I frankly
> suspect there aren't any, as the bug was fixed long ago, and it
> wouldn't affect them unless no gcc updates were ever done on their
> systems.
> 
> Carlo