Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Simon Sobisch]

Joel Brobecker <brobecker@adacore.com> wrote Sat Jan 8 08:27:18 GMT 2022

 > Hi everyone,
 >
 > The last fix went in just before the holidays, and so we should
 > be ready to create the release. Unless there are last-minute
 > issues being discovered, I will create it next weekend.
 >
 > Thanks everyone for helping us get there :)
 >
 > --
 > Joel

I guess that means that the patch-sets for "Source highlight non utf-8 
characters using Python" by Anbdrew Burgess and "Improvements to Python 
parameters" by Tom Tromey won't be in there, then.

Is that (sadly) correct?
[If yes: is there any rough estimation when those are likely available 
as release, fixing issues that seem to be around since GDB 10?]

Thanks for working on GDB to all and this time special thanks to Andrew 
and Tom for those patch sets (and Eli for checking those).

Simon

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Joel Brobecker]

> I guess that means that the patch-sets for "Source highlight non utf-8
> characters using Python" by Anbdrew Burgess and "Improvements to Python
> parameters" by Tom Tromey won't be in there, then.
> 
> Is that (sadly) correct?
> [If yes: is there any rough estimation when those are likely available as
> release, fixing issues that seem to be around since GDB 10?]
> 
> Thanks for working on GDB to all and this time special thanks to Andrew and
> Tom for those patch sets (and Eli for checking those).

I fear it might be too tight, but I will let Andrew and Tom comment
on that.

Are the issues critical enough and the associated fix sufficiently
safe that we might consider delaying the release for this?

  - Source highlight non utf-8 characters using Python:
    Not sure.

  - Improvements to Python parameter
    I don't think this is new, and I am not sure it's worth delaying
    the release for this.

-- 
Joel

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Tom Tromey]

>>>>> "Joel" == Joel Brobecker <brobecker@adacore.com> writes:

Joel>   - Improvements to Python parameter
Joel>     I don't think this is new, and I am not sure it's worth delaying
Joel>     the release for this.

FWIW the bugs fixed there have been around for years.
I wasn't really considering backporting these to 11.2.
Hopefully 12.1 isn't all that far away.

Tom

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Andrew Burgess]

* Joel Brobecker via Gdb-patches <gdb-patches@sourceware.org> [2022-01-12 09:12:20 +0400]:

> > I guess that means that the patch-sets for "Source highlight non utf-8
> > characters using Python" by Anbdrew Burgess and "Improvements to Python
> > parameters" by Tom Tromey won't be in there, then.
> > 
> > Is that (sadly) correct?
> > [If yes: is there any rough estimation when those are likely available as
> > release, fixing issues that seem to be around since GDB 10?]
> > 
> > Thanks for working on GDB to all and this time special thanks to Andrew and
> > Tom for those patch sets (and Eli for checking those).
> 
> I fear it might be too tight, but I will let Andrew and Tom comment
> on that.
> 
> Are the issues critical enough and the associated fix sufficiently
> safe that we might consider delaying the release for this?
> 
>   - Source highlight non utf-8 characters using Python:
>     Not sure.

My current guess is that this patch is going to take a while to hit
master - anything that involves Python and Unicode seems to be
problematic - so I had not imagined this landing in 11.2.

I'd be disappointed if this (or something similar) hadn't landed by
12.1 though.

Thanks,
Andrew

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Joel Brobecker]

> > I fear it might be too tight, but I will let Andrew and Tom comment
> > on that.
> > 
> > Are the issues critical enough and the associated fix sufficiently
> > safe that we might consider delaying the release for this?
> > 
> >   - Source highlight non utf-8 characters using Python:
> >     Not sure.
> 
> My current guess is that this patch is going to take a while to hit
> master - anything that involves Python and Unicode seems to be
> problematic - so I had not imagined this landing in 11.2.

I share the feeling, so this makes sense to me.

Thanks Andrew.
-- 
Joel

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Aaron Merey]

On Wed, Jan 12, 2022 at 10:58 PM Joel Brobecker <brobecker@adacore.com> wrote:
> With that said, unless the patch you are proposing has been deployed
> and in real-world use for a while, my experience tells me that it is
> a really risky proposition to be pushing a patch like that at the very
> last minute. On the surface, the patch looks OK, and it's contained
> within the download side of debuginfod, but we could imagine a bug
> in this patch causing debuginfod downloads to crash, thus rendering
> the entire feature no longer usable. When we are in the last stretches
> of the preparation of a release which is going to be the last release
> from this branch, unless the fix is absolutely critical or obviously
> safe, my personal approach has been to remember that the .2 release
> is first and foremost a corrective release. Just my 2 cents.

On Thu, Jan 13, 2022 at 5:49 AM Andrew Burgess <aburgess@redhat.com> wrote:
> I agree with Joel, this isn't a bug fix, so I don't think it's
> appropriate to include in 11.2.

Fair points. To be safe let's not rush my patch then.

Thanks,
Aaron

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Andrew Burgess]

* Aaron Merey via Gdb-patches <gdb-patches@sourceware.org> [2022-01-12 17:29:03 -0500]:

> Hi Joel,
> 
> On Sat, Jan 8, 2022 at 3:27 AM Joel Brobecker via Gdb-patches
> <gdb-patches@sourceware.org> wrote:
> > The last fix went in just before the holidays, and so we should
> > be ready to create the release. Unless there are last-minute
> > issues being discovered, I will create it next weekend.
> 
> I realize it's a bit of a stretch but I'm wondering if there is any way
> the following patch could make it in time:
> 
> https://sourceware.org/pipermail/gdb-patches/2022-January/185034.html
> 
> It's a small, straightforward patch and I've received feedback from
> some who would like to see this improvement in GDB.

I agree with Joel, this isn't a bug fix, so I don't think it's
appropriate to include in 11.2.

Others are free to disagree though.

Thanks,
Andrew

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Joel Brobecker]

Hi Aaron,

> > The last fix went in just before the holidays, and so we should
> > be ready to create the release. Unless there are last-minute
> > issues being discovered, I will create it next weekend.
> 
> I realize it's a bit of a stretch but I'm wondering if there is any way
> the following patch could make it in time:
> 
> https://sourceware.org/pipermail/gdb-patches/2022-January/185034.html
> 
> It's a small, straightforward patch and I've received feedback from
> some who would like to see this improvement in GDB.

If you create a PR in bugzilla for this change, add it to the commit's
revision history, and then you can get the patch reviewed and approved
by Saturday (RH has a couple of Global Maintainer who can help), then
the patch looks isolated-enough that I do not have an objection against
backporting it to the gdb-11-branch.

With that said, unless the patch you are proposing has been deployed
and in real-world use for a while, my experience tells me that it is
a really risky proposition to be pushing a patch like that at the very
last minute. On the surface, the patch looks OK, and it's contained
within the download side of debuginfod, but we could imagine a bug
in this patch causing debuginfod downloads to crash, thus rendering
the entire feature no longer usable. When we are in the last stretches
of the preparation of a release which is going to be the last release
from this branch, unless the fix is absolutely critical or obviously
safe, my personal approach has been to remember that the .2 release
is first and foremost a corrective release. Just my 2 cents.

-- 
Joel

Re: [heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Aaron Merey]

Hi Joel,

On Sat, Jan 8, 2022 at 3:27 AM Joel Brobecker via Gdb-patches
<gdb-patches@sourceware.org> wrote:
> The last fix went in just before the holidays, and so we should
> be ready to create the release. Unless there are last-minute
> issues being discovered, I will create it next weekend.

I realize it's a bit of a stretch but I'm wondering if there is any way
the following patch could make it in time:

https://sourceware.org/pipermail/gdb-patches/2022-January/185034.html

It's a small, straightforward patch and I've received feedback from
some who would like to see this improvement in GDB.

Thanks,
Aaron

[heads up] GDB 11.2 "respin" release next weekend (Jan 15-16)

[Joel Brobecker]

Hi everyone,

The last fix went in just before the holidays, and so we should
be ready to create the release. Unless there are last-minute
issues being discovered, I will create it next weekend.

Thanks everyone for helping us get there :)

-- 
Joel