RE: Did the "Fedora-latest" version on GitHub just get bumped?

[Zimoch Dirk via Core-talk <core-talk at aps.anl.gov>]

I tried to fix it too in https://github.com/epics-base/epics-base/pull/638 but maybe we need to install gcc15 locally and see what else broke.

Dirk

 

From: Core-talk <core-talk-bounces at aps.anl.gov> On Behalf Of Johnson, Andrew N. via Core-talk
Sent: Donnerstag, 17. April 2025 17:14
To: Michael Davidsaver <mdavidsaver at gmail.com>
Cc: EPICS core-talk <core-talk at aps.anl.gov>
Subject: Re: Did the "Fedora-latest" version on GitHub just get bumped?

 

On 4/17/25, 9:27 AM, "Michael Davidsaver" <mdavidsaver at gmail.com> wrote:

On 4/16/25 09:34, Johnson, Andrew N. via Core-talk wrote:

> I saw that a new Fedora was release very recently.

Yes.  Now comes GCC 15, which defaults to C23.

https://urldefense.us/v3/__https://github.com/epics-base/epics-base/issues/632__;!!G_uCfscf7eWS!YHkjFuvcJ3O3DLlrIx18bZ_np7W9OljZ-htT9LK3F3Dz-HpgmTBbEFU6MZqA-7ka1WLkntrENxJ-_pt2$

I saw that commit, how come the build didn’t fail with the same error that other PRs are? Ah that wasn’t a PR, and your Fedora CI job did fail, but apparently we don’t get notifications when non-PR jobs fail. Can/should we change that in the CI?

Looking at the specific code that’s failing, in this case I think the fix is just adding a cast to DEVSUPFUN* since we only print the pointer. I just pushed a commit to try that, this job shows that seemed to work, but there are other similar failures later in the build, which I will look at.

 

- Andrew

 

-- 

Complexity comes for free, Simplicity you have to work for.