GDB 11.2 ("re-spin") update (2021-11-21)

GDB 11.2 ("re-spin") update (2021-11-21)

[Joel Brobecker]

Hello everyone,

It's been just over two months since we release GDB 11.1,
and so let's try to get GDB 11.2 out.

So far, we have 3 fixes that got in:

 * PR build/28318 ([gdb/build] std::thread support configure check does not use CXX_DIALECT)
 * PR tui/28483 ([gdb/tui] breakpoint creation not displayed)
 * PR build/28555 (uclibc compile failure since commit 4655f8509fd44e6efabefa373650d9982ff37fd6)

I know of a couple of items (listed below). Are there any other issues
you're aware of that you think we should fix for the .2?

Finally, just a quick reminder that all fixes going in the .2
should have a corresponding PR ;-).

Thank you!

Ongoing Items:
--------------

 * [Andrew?] <PR gdb/28405>
   arm-none-eabi: internal-error: first_resumed_thread != nullptr failed assertion
   https://sourceware.org/bugzilla/show_bug.cgi?id=28405

       Andrew sent a patch on Oct 5th:
       https://sourceware.org/pipermail/gdb-patches/2021-October/182394.html

       Simon's review:
       https://sourceware.org/pipermail/gdb-patches/2021-November/183388.html

  * [Tromey] <*** NEEDS BUGZILLA PR ***>
    Allow DW_ATE_UTF for Rust characters

    Patch available at:
    https://sourceware.org/pipermail/gdb-patches/2021-October/182960.html


Not Critical, but Requested:
----------------------------

  * [SimonM as awesome supporter and reviewer] <PR gdb/24069>
    gdb on OSX Mojave hangs in darwin_decode_message
    https://sourceware.org/bugzilla/show_bug.cgi?id=24069

        No progress since Jul 2021.

        I know that the state of GDB on MacOS is very poor at the moment,
        and so having this PR fixed for 11.1 would be very nice. I don't
        think this PR should block the release, though. This is not new,
        and I'm not sure whether we can expect for us to converge without
        too much delay. There are good signs, though, with regular
        activity on the PR, lately. So let's see how it goes...

  * [BerndE] Improve debugging of optimized code
    https://sourceware.org/pipermail/gdb-patches/2021-January/175617.html

    This is reported to fix <PR c++/25987>
    Wrong line for code containing inlined functions

    Andrew xref-ed to https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94474

    (2021-09-05) v3 posted at:
      [0/3] https://sourceware.org/pipermail/gdb-patches/2021-September/181829.html
      [1/3] https://sourceware.org/pipermail/gdb-patches/2021-September/181830.html
      [2/3] https://sourceware.org/pipermail/gdb-patches/2021-September/181831.html
      [3/3] https://sourceware.org/pipermail/gdb-patches/2021-September/181832.html

    Looks like those patches are a challenge to review
    https://sourceware.org/pipermail/gdb-patches/2021-September/181930.html

  * [Simon] <PR gdb/26868>
    Follow-up on x86-64 stub submission
    https://sourceware.org/bugzilla/show_bug.cgi?id=26868

    This is the submission in question:
    https://sourceware.org/pipermail/gdb-patches/2020-November/173182.html

    For me, although I am not a lawyer, I don't think we can accept
    the contribution without the author signing a copyright assignment
    form.

  * [PhilippeW] Fix/complete option list/description in manual and in gdb --help

    RFA v2 (2021-06-20):
        [RFAV2] Fix/complete option list/description in manual and in gdb --help.
        https://sourceware.org/pipermail/gdb-patches/2021-June/180186.html

-- 
Joel

Re: GDB 11.2 ("re-spin") update (2021-11-21)

[Bruno Larsen]

On 11/21/21 10:47, Joel Brobecker via Gdb-patches wrote:
> Hello everyone,
> 
> It's been just over two months since we release GDB 11.1,
> and so let's try to get GDB 11.2 out.
> 
> So far, we have 3 fixes that got in:
> 
>   * PR build/28318 ([gdb/build] std::thread support configure check does not use CXX_DIALECT)
>   * PR tui/28483 ([gdb/tui] breakpoint creation not displayed)
>   * PR build/28555 (uclibc compile failure since commit 4655f8509fd44e6efabefa373650d9982ff37fd6)
> 
> I know of a couple of items (listed below). Are there any other issues
> you're aware of that you think we should fix for the .2?

I'm wondering if PR 28480 (https://sourceware.org/bugzilla/show_bug.cgi?id=28480) is relevant enough to be on .2; A solution is already on the list (https://sourceware.org/pipermail/gdb-patches/2021-November/183243.html) but hasn't been accepted yet

Basically, with this proposed solutions, things will stop breaking silently and will now loudly warn if ambiguous requests are made.

> 
> Finally, just a quick reminder that all fixes going in the .2
> should have a corresponding PR ;-).
> 
> Thank you!
> 
> Ongoing Items:
> --------------
> 
>   * [Andrew?] <PR gdb/28405>
>     arm-none-eabi: internal-error: first_resumed_thread != nullptr failed assertion
>     https://sourceware.org/bugzilla/show_bug.cgi?id=28405
> 
>         Andrew sent a patch on Oct 5th:
>         https://sourceware.org/pipermail/gdb-patches/2021-October/182394.html
> 
>         Simon's review:
>         https://sourceware.org/pipermail/gdb-patches/2021-November/183388.html
> 
>    * [Tromey] <*** NEEDS BUGZILLA PR ***>
>      Allow DW_ATE_UTF for Rust characters
> 
>      Patch available at:
>      https://sourceware.org/pipermail/gdb-patches/2021-October/182960.html
> 
> 
> Not Critical, but Requested:
> ----------------------------
> 
>    * [SimonM as awesome supporter and reviewer] <PR gdb/24069>
>      gdb on OSX Mojave hangs in darwin_decode_message
>      https://sourceware.org/bugzilla/show_bug.cgi?id=24069
> 
>          No progress since Jul 2021.
> 
>          I know that the state of GDB on MacOS is very poor at the moment,
>          and so having this PR fixed for 11.1 would be very nice. I don't
>          think this PR should block the release, though. This is not new,
>          and I'm not sure whether we can expect for us to converge without
>          too much delay. There are good signs, though, with regular
>          activity on the PR, lately. So let's see how it goes...
> 
>    * [BerndE] Improve debugging of optimized code
>      https://sourceware.org/pipermail/gdb-patches/2021-January/175617.html
> 
>      This is reported to fix <PR c++/25987>
>      Wrong line for code containing inlined functions
> 
>      Andrew xref-ed to https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94474
> 
>      (2021-09-05) v3 posted at:
>        [0/3] https://sourceware.org/pipermail/gdb-patches/2021-September/181829.html
>        [1/3] https://sourceware.org/pipermail/gdb-patches/2021-September/181830.html
>        [2/3] https://sourceware.org/pipermail/gdb-patches/2021-September/181831.html
>        [3/3] https://sourceware.org/pipermail/gdb-patches/2021-September/181832.html
> 
>      Looks like those patches are a challenge to review
>      https://sourceware.org/pipermail/gdb-patches/2021-September/181930.html
> 
>    * [Simon] <PR gdb/26868>
>      Follow-up on x86-64 stub submission
>      https://sourceware.org/bugzilla/show_bug.cgi?id=26868
> 
>      This is the submission in question:
>      https://sourceware.org/pipermail/gdb-patches/2020-November/173182.html
> 
>      For me, although I am not a lawyer, I don't think we can accept
>      the contribution without the author signing a copyright assignment
>      form.
> 
>    * [PhilippeW] Fix/complete option list/description in manual and in gdb --help
> 
>      RFA v2 (2021-06-20):
>          [RFAV2] Fix/complete option list/description in manual and in gdb --help.
>          https://sourceware.org/pipermail/gdb-patches/2021-June/180186.html
> 


-- 
Cheers!
Bruno Larsen

Re: GDB 11.2 ("re-spin") update (2021-11-21)

[Joel Brobecker]

Hi Bruno,

> I'm wondering if PR 28480
> (https://sourceware.org/bugzilla/show_bug.cgi?id=28480) is relevant
> enough to be on .2; A solution is already on the list
> (https://sourceware.org/pipermail/gdb-patches/2021-November/183243.html)
> but hasn't been accepted yet
> 
> Basically, with this proposed solutions, things will stop breaking
> silently and will now loudly warn if ambiguous requests are made.

This seems reasonable to me, provided the patch doesn't drag for
too long, and then a Global Maintainer approves the patch for
backporting to 11.2 (it looks relatively contained, so no objection
a priori myself, but I don't know much about this area of the code).

The PR suggests this is a regression from GDB 10, but it sounds
like it was really luck of the draw that made it work before?

-- 
Joel

Re: GDB 11.2 ("re-spin") update (2021-11-21)

[Bruno Larsen]

On 11/23/21 01:54, Joel Brobecker wrote:
> Hi Bruno,
> 
>> I'm wondering if PR 28480
>> (https://sourceware.org/bugzilla/show_bug.cgi?id=28480) is relevant
>> enough to be on .2; A solution is already on the list
>> (https://sourceware.org/pipermail/gdb-patches/2021-November/183243.html)
>> but hasn't been accepted yet
>>
>> Basically, with this proposed solutions, things will stop breaking
>> silently and will now loudly warn if ambiguous requests are made.
> 
> This seems reasonable to me, provided the patch doesn't drag for
> too long, and then a Global Maintainer approves the patch for
> backporting to 11.2 (it looks relatively contained, so no objection
> a priori myself, but I don't know much about this area of the code).

I wouldn't expect the patch to drag, since the last comments were to do with
formatting, more than about the solution itself.

I would be surprised if it wasn't self contained, but it is possible, I am
not very familiar with the code as a whole.

> 
> The PR suggests this is a regression from GDB 10, but it sounds
> like it was really luck of the draw that made it work before?
> 

Pretty much. It was reported as a regression, but we eventually understood that
it is just the "bad commit" was just not thorough enough and that the code was
relying on undefined behavior.

-- 
Cheers!
Bruno Larsen

Re: GDB 11.2 ("re-spin") update (2021-11-21)

[Joel Brobecker]

> > > I'm wondering if PR 28480
> > > (https://sourceware.org/bugzilla/show_bug.cgi?id=28480) is relevant
> > > enough to be on .2; A solution is already on the list
> > > (https://sourceware.org/pipermail/gdb-patches/2021-November/183243.html)
> > > but hasn't been accepted yet
> > > 
> > > Basically, with this proposed solutions, things will stop breaking
> > > silently and will now loudly warn if ambiguous requests are made.
> > 
> > This seems reasonable to me, provided the patch doesn't drag for
> > too long, and then a Global Maintainer approves the patch for
> > backporting to 11.2 (it looks relatively contained, so no objection
> > a priori myself, but I don't know much about this area of the code).
> 
> I wouldn't expect the patch to drag, since the last comments were to do with
> formatting, more than about the solution itself.

OK, good.

I'll add this PR to the list of candidates for the 11.2.

-- 
Joel