time to start thinking about GDB 7.3 (as well!)

time to start thinking about GDB 7.3 (as well!)

[Joel Brobecker]

According to the tentative calendar, it's also time to start thinking
about the GDB 7.3 release process.  Normally, we were planning on
branching next week.

Judging from the entries in the NEWS file, I think we have enough
material to make a new release.  However, I also believe that the
same issues that we identified as a blocker for 7.2.1 are also
relevant for 7.3.

There is a Wiki page for the 7.3 release:
    http://sourceware.org/gdb/wiki/GDB_7.3_Release

Can we start putting entries for things we know need to be fixed
before branching/release?

We should also discuss features that we think would be worthwhile
waiting for. Some Python work being finished, for instance, which
might nicely complements some of the new enhancements we got?
Tracepoints? gdbserver work? Etc.

What I propose is to produce 7.2.1 as soon as the blocking issues
are fixed.  We might also be creating the 7.3 branch pretty much
at the same time, which means that 7.3 might be coming shortly
after 7.2.1, but I think that's OK. I'd rather not have a 7.2.1
if that means artificially delaying 7.3 - but I don't a problem
with producing both.

-- 
Joel

Re: time to start thinking about GDB 7.3 (as well!)

[Tom Tromey]

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

Joel> According to the tentative calendar, it's also time to start
Joel> thinking about the GDB 7.3 release process.  Normally, we were
Joel> planning on branching next week.

Joel> We should also discuss features that we think would be worthwhile
Joel> waiting for. Some Python work being finished, for instance, which
Joel> might nicely complements some of the new enhancements we got?
Joel> Tracepoints? gdbserver work? Etc.

I think Phil is planning to resubmit his Python breakpoint changes this
week.

Jan said on irc that the ifunc branch might be close to merge-ready.
This is important because, without it, `strcmp' is not usable (on
systems with a glibc that uses ifunc) in a breakpoint condition --
pretty basic stuff.

Sergio and I have been working on exposing SystemTap static probe points
to gdb.  We are reasonably close to finishing this; the basic
functionality is fine, and this week Sergio is working on the probe
arguments while I work on tests.  So, assuming others are amenable to
this idea, we would like this to be in 7.3.  I am not sure whether all
the details will be acceptable in the right timeframe; I could post the
"bad" parts if people want to take a look now.


For 7.2.1, I think Jan and Keith have been discussing the C++ regressions.
I'm not sure what they have decided, I'm a little behind on that again.

Tom

Re: time to start thinking about GDB 7.3 (as well!)

[Mike Frysinger]

[-- Attachment #1: Type: text/plain, Size: 523 bytes --]

On Monday, February 14, 2011 09:50:10 Joel Brobecker wrote:
> We should also discuss features that we think would be worthwhile
> waiting for. Some Python work being finished, for instance, which
> might nicely complements some of the new enhancements we got?
> Tracepoints? gdbserver work? Etc.

i'd like to the Blackfin sim port merged before the 7.3 branch.  i'm not 
expecting a comprehensive review or anything since the port is so large 
(==comprehensive).  the only thing it needs is an approval ...
-mike

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

Re: time to start thinking about GDB 7.3 (as well!)

[Mike Frysinger]

On Mon, Feb 14, 2011 at 3:14 PM, Mike Frysinger wrote:
> On Monday, February 14, 2011 09:50:10 Joel Brobecker wrote:
>> We should also discuss features that we think would be worthwhile
>> waiting for. Some Python work being finished, for instance, which
>> might nicely complements some of the new enhancements we got?
>> Tracepoints? gdbserver work? Etc.
>
> i'd like to the Blackfin sim port merged before the 7.3 branch.  i'm not
> expecting a comprehensive review or anything since the port is so large
> (==comprehensive).  the only thing it needs is an approval ...

well my stuff is in now :), and the wiki page in question is blank ...
-mike

Re: time to start thinking about GDB 7.3 (as well!)

[Pedro Alves]

On Monday 07 March 2011 11:18:03, Mike Frysinger wrote:
> On Mon, Feb 14, 2011 at 3:14 PM, Mike Frysinger wrote:
> > On Monday, February 14, 2011 09:50:10 Joel Brobecker wrote:
> >> We should also discuss features that we think would be worthwhile
> >> waiting for. Some Python work being finished, for instance, which
> >> might nicely complements some of the new enhancements we got?
> >> Tracepoints? gdbserver work? Etc.
> >
> > i'd like to the Blackfin sim port merged before the 7.3 branch.  i'm not
> > expecting a comprehensive review or anything since the port is so large
> > (==comprehensive).  the only thing it needs is an approval ...
> 
> well my stuff is in now :), and the wiki page in question is blank ...

Let me put some stuff there then.  :-)  

Joel, do I still have time to get my second and third parts
of the <unavailable> values support (posted the other week) into 7.3?

 <http://sourceware.org/ml/gdb-patches/2011-02/msg00581.html>
 <http://sourceware.org/ml/gdb-patches/2011-02/msg00611.html>

WDYT?  This makes the tracing experience much better compared to 7.2.

-- 
Pedro Alves

Re: time to start thinking about GDB 7.3 (as well!)

[Joel Brobecker]

> Let me put some stuff there then.  :-)  
> 
> Joel, do I still have time to get my second and third parts
> of the <unavailable> values support (posted the other week) into 7.3?
> 
>  <http://sourceware.org/ml/gdb-patches/2011-02/msg00581.html>
>  <http://sourceware.org/ml/gdb-patches/2011-02/msg00611.html>
> 
> WDYT?  This makes the tracing experience much better compared to 7.2.

I think you do.  The progress I'm seeing on the real blocker
seems to be vewwwy vewwwy slow :-(.

-- 
Joel

Re: time to start thinking about GDB 7.3 (as well!)

[Pedro Alves]

On Monday 07 March 2011 12:13:31, Joel Brobecker wrote:
> > Let me put some stuff there then.  :-)  
> > 
> > Joel, do I still have time to get my second and third parts
> > of the <unavailable> values support (posted the other week) into 7.3?
> > 
> >  <http://sourceware.org/ml/gdb-patches/2011-02/msg00581.html>
> >  <http://sourceware.org/ml/gdb-patches/2011-02/msg00611.html>
> > 
> > WDYT?  This makes the tracing experience much better compared to 7.2.
> 
> I think you do.

Thanks!  I've added it to the wiki.

-- 
Pedro Alves

Re: time to start thinking about GDB 7.3 (as well!)

[Hui Zhu]

7.2.1

On Mon, Feb 14, 2011 at 22:50, Joel Brobecker <brobecker@adacore.com> wrote:
> According to the tentative calendar, it's also time to start thinking
> about the GDB 7.3 release process.  Normally, we were planning on
> branching next week.
>
> Judging from the entries in the NEWS file, I think we have enough
> material to make a new release.  However, I also believe that the
> same issues that we identified as a blocker for 7.2.1 are also
> relevant for 7.3.
>
> There is a Wiki page for the 7.3 release:
>    http://sourceware.org/gdb/wiki/GDB_7.3_Release
>
> Can we start putting entries for things we know need to be fixed
> before branching/release?
>
> We should also discuss features that we think would be worthwhile
> waiting for. Some Python work being finished, for instance, which
> might nicely complements some of the new enhancements we got?
> Tracepoints? gdbserver work? Etc.


I am woking on the tracepoint printf support.  Wish it can be in 7.3.


>
> What I propose is to produce 7.2.1 as soon as the blocking issues
> are fixed.  We might also be creating the 7.3 branch pretty much
> at the same time, which means that 7.3 might be coming shortly
> after 7.2.1, but I think that's OK. I'd rather not have a 7.2.1
> if that means artificially delaying 7.3 - but I don't a problem
> with producing both.
>

Wish we can get 7.2.1 early.  :)

Thanks,
Hui