* Product branches in the GDB repository?
@ 2005-03-22 0:20 Daniel Jacobowitz
2005-03-22 0:37 ` Stan Shebs
2005-03-25 19:47 ` Daniel Jacobowitz
0 siblings, 2 replies; 10+ messages in thread
From: Daniel Jacobowitz @ 2005-03-22 0:20 UTC (permalink / raw)
To: gdb
We're already pretty flexible about creating branches for development. Does
anyone have an object to branches for non-FSF releases? I think it's a
logical extension of current practice.
CodeSourcery does regular releases of an ARM toolchain, and we're going to
have a bunch of as yet unsubmitted GDB patches in our next release. Nothing
deliberately unsubmitted, of course, but things which need additional
polishing or review or input before they're ready for HEAD. I would prefer
to maintain them in the main repository, simply because it's easier with CVS.
And that way the patches are easily available for any curious GDB
developers.
FWIW, we already do this for both gcc and binutils.
--
Daniel Jacobowitz
CodeSourcery, LLC
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-22 0:20 Product branches in the GDB repository? Daniel Jacobowitz
@ 2005-03-22 0:37 ` Stan Shebs
2005-03-22 1:21 ` Nick Roberts
2005-03-25 19:47 ` Daniel Jacobowitz
1 sibling, 1 reply; 10+ messages in thread
From: Stan Shebs @ 2005-03-22 0:37 UTC (permalink / raw)
To: Daniel Jacobowitz; +Cc: gdb
Daniel Jacobowitz wrote:
>We're already pretty flexible about creating branches for development. Does
>anyone have an object to branches for non-FSF releases? I think it's a
>logical extension of current practice.
>
>CodeSourcery does regular releases of an ARM toolchain, and we're going to
>have a bunch of as yet unsubmitted GDB patches in our next release. Nothing
>deliberately unsubmitted, of course, but things which need additional
>polishing or review or input before they're ready for HEAD. I would prefer
>to maintain them in the main repository, simply because it's easier with CVS.
>And that way the patches are easily available for any curious GDB
>developers.
>
>FWIW, we already do this for both gcc and binutils.
>
>
Yes, Apple has been doing it with GCC on a pretty large scale,
and it doesn't seem as though it's been troublesome for anyone.
So I'm OK with doing the same for GDB; as with the GCC branches,
there should be documentation about the branch and its policies
both on web page and in a branch-only file.
(Does that mean Apple GDB should get a branch here too?
Interesting question...)
Stan
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-22 0:37 ` Stan Shebs
@ 2005-03-22 1:21 ` Nick Roberts
2005-03-22 2:38 ` Daniel Jacobowitz
2005-03-22 19:43 ` Stan Shebs
0 siblings, 2 replies; 10+ messages in thread
From: Nick Roberts @ 2005-03-22 1:21 UTC (permalink / raw)
To: Stan Shebs; +Cc: Daniel Jacobowitz, gdb
> (Does that mean Apple GDB should get a branch here too?
> Interesting question...)
In that case, do Apple intend to merge their changes with HEAD at some stage?
I am particularly interested in their changes to GDB/MI.
Nick
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-22 1:21 ` Nick Roberts
@ 2005-03-22 2:38 ` Daniel Jacobowitz
2005-03-22 19:43 ` Stan Shebs
1 sibling, 0 replies; 10+ messages in thread
From: Daniel Jacobowitz @ 2005-03-22 2:38 UTC (permalink / raw)
To: Nick Roberts; +Cc: Stan Shebs, gdb
On Tue, Mar 22, 2005 at 02:21:39PM +1200, Nick Roberts wrote:
>
> > (Does that mean Apple GDB should get a branch here too?
> > Interesting question...)
In my opinion - absolutely yes. We might want a separate list for
patches for non-FSF branches, though. Just to keep the development
load sane.
> In that case, do Apple intend to merge their changes with HEAD at some stage?
> I am particularly interested in their changes to GDB/MI.
I am under the impression that Apple has always wanted to do this, but
is persistently short of the manpower to do it - just like everyone
else in the world, they don't have enough hours in their day.
I hope that, eventually, progress will be made!
--
Daniel Jacobowitz
CodeSourcery, LLC
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-22 1:21 ` Nick Roberts
2005-03-22 2:38 ` Daniel Jacobowitz
@ 2005-03-22 19:43 ` Stan Shebs
2005-03-29 0:53 ` Bob Rossi
1 sibling, 1 reply; 10+ messages in thread
From: Stan Shebs @ 2005-03-22 19:43 UTC (permalink / raw)
To: Nick Roberts; +Cc: Daniel Jacobowitz, gdb
Nick Roberts wrote:
> > (Does that mean Apple GDB should get a branch here too?
> > Interesting question...)
>
>In that case, do Apple intend to merge their changes with HEAD at some stage?
>I am particularly interested in their changes to GDB/MI.
>
I can't speak for our future plans, but I personally would like to
make it happen, and I recently got reassigned to debugger hacking,
which gives us a 33% increase in our GDB hack power. So there's
some reason for optimism.
I currently have an FSF GDB sandbox that can read mach-o files
and run programs, but haven't quite figured out how to make them
stop. :-)
Stan
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-22 19:43 ` Stan Shebs
@ 2005-03-29 0:53 ` Bob Rossi
2005-03-29 1:25 ` Stan Shebs
0 siblings, 1 reply; 10+ messages in thread
From: Bob Rossi @ 2005-03-29 0:53 UTC (permalink / raw)
To: Stan Shebs; +Cc: Nick Roberts, Daniel Jacobowitz, gdb
On Tue, Mar 22, 2005 at 11:43:20AM -0800, Stan Shebs wrote:
> Nick Roberts wrote:
>
> >> (Does that mean Apple GDB should get a branch here too?
> >> Interesting question...)
> >
> >In that case, do Apple intend to merge their changes with HEAD at some
> >stage?
> >I am particularly interested in their changes to GDB/MI.
> >
> I can't speak for our future plans, but I personally would like to
> make it happen, and I recently got reassigned to debugger hacking,
> which gives us a 33% increase in our GDB hack power. So there's
> some reason for optimism.
It would certainly benefit everyone if you guys could contribute back.
For example, I'm probably going to be reimplementing some MI features
you guys already have. I hate to do that if the work has already been
done. (ie. Asyncronous commands need to return the type of command they
are, so the command can be analsized semantically)
Bob Rossi
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-29 0:53 ` Bob Rossi
@ 2005-03-29 1:25 ` Stan Shebs
0 siblings, 0 replies; 10+ messages in thread
From: Stan Shebs @ 2005-03-29 1:25 UTC (permalink / raw)
To: Bob Rossi; +Cc: Nick Roberts, Daniel Jacobowitz, gdb
Bob Rossi wrote:
>On Tue, Mar 22, 2005 at 11:43:20AM -0800, Stan Shebs wrote:
>
>>Nick Roberts wrote:
>>
>>
>>>>(Does that mean Apple GDB should get a branch here too?
>>>>Interesting question...)
>>>>
>>>In that case, do Apple intend to merge their changes with HEAD at some
>>>stage?
>>>I am particularly interested in their changes to GDB/MI.
>>>
>>>
>>I can't speak for our future plans, but I personally would like to
>>make it happen, and I recently got reassigned to debugger hacking,
>>which gives us a 33% increase in our GDB hack power. So there's
>>some reason for optimism.
>>
>
>It would certainly benefit everyone if you guys could contribute back.
>For example, I'm probably going to be reimplementing some MI features
>you guys already have. I hate to do that if the work has already been
>done. (ie. Asyncronous commands need to return the type of command they
>are, so the command can be analsized semantically)
>
Absolutely we want to work together on this. I'm being careful
not to overpromise because there has been some unfortunate history
(an euphemism for past interactions now described by all sides
with unprintable words :-) ) in connection with Apple and GDB.
Much of my work now is just in reading all the code and figuring
out where things are even at - although it would have been truly
studly to keep up on all the GDB doings while hacking on GCC,
in actuality I've run into my brain limits and GCC bits have
displaced much of my old GDB knowledge. :-(
Stan
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-22 0:20 Product branches in the GDB repository? Daniel Jacobowitz
2005-03-22 0:37 ` Stan Shebs
@ 2005-03-25 19:47 ` Daniel Jacobowitz
2005-03-25 19:56 ` Mark Kettenis
1 sibling, 1 reply; 10+ messages in thread
From: Daniel Jacobowitz @ 2005-03-25 19:47 UTC (permalink / raw)
To: gdb
On Mon, Mar 21, 2005 at 07:20:34PM -0500, Daniel Jacobowitz wrote:
> We're already pretty flexible about creating branches for development. Does
> anyone have an object to branches for non-FSF releases? I think it's a
> logical extension of current practice.
>
> CodeSourcery does regular releases of an ARM toolchain, and we're going to
> have a bunch of as yet unsubmitted GDB patches in our next release. Nothing
> deliberately unsubmitted, of course, but things which need additional
> polishing or review or input before they're ready for HEAD. I would prefer
> to maintain them in the main repository, simply because it's easier with CVS.
> And that way the patches are easily available for any curious GDB
> developers.
>
> FWIW, we already do this for both gcc and binutils.
No one objected, and Stan supported the idea, so I am currently tagging
csl-arm-20050325-branch. This is a release branch for our 2005-Q1 ARM
toolchains; I intend for everything on the branch to be submitted to
mainline eventually, but some bits will take time to review, and others
may need interface changes before they are acceptable. There will be
some native Windows patches and some remote protocol improvements in
addition to ARM-only changes.
Anyone have preferences on where to document branches? We don't
currently do that in GDB. Easiest would be a file gdb/BRANCHES,
parallel to binutils/BRANCHES, or on the "current CVS" page of the
web site.
--
Daniel Jacobowitz
CodeSourcery, LLC
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-25 19:47 ` Daniel Jacobowitz
@ 2005-03-25 19:56 ` Mark Kettenis
2005-03-25 20:54 ` Daniel Jacobowitz
0 siblings, 1 reply; 10+ messages in thread
From: Mark Kettenis @ 2005-03-25 19:56 UTC (permalink / raw)
To: drow; +Cc: gdb
Date: Fri, 25 Mar 2005 14:48:19 -0500
From: Daniel Jacobowitz <drow@false.org>
Anyone have preferences on where to document branches? We don't
currently do that in GDB. Easiest would be a file gdb/BRANCHES,
parallel to binutils/BRANCHES, or on the "current CVS" page of the
web site.
I think gdb/BRANCHES is an excellent idea, especially since that's
what the binutils folks already do.
Hmm, that raises the question where to list the tag if people want to
branch gdb+binutils...
Mark
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Product branches in the GDB repository?
2005-03-25 19:56 ` Mark Kettenis
@ 2005-03-25 20:54 ` Daniel Jacobowitz
0 siblings, 0 replies; 10+ messages in thread
From: Daniel Jacobowitz @ 2005-03-25 20:54 UTC (permalink / raw)
To: Mark Kettenis; +Cc: gdb
On Fri, Mar 25, 2005 at 08:56:16PM +0100, Mark Kettenis wrote:
> Date: Fri, 25 Mar 2005 14:48:19 -0500
> From: Daniel Jacobowitz <drow@false.org>
>
> Anyone have preferences on where to document branches? We don't
> currently do that in GDB. Easiest would be a file gdb/BRANCHES,
> parallel to binutils/BRANCHES, or on the "current CVS" page of the
> web site.
>
> I think gdb/BRANCHES is an excellent idea, especially since that's
> what the binutils folks already do.
>
> Hmm, that raises the question where to list the tag if people want to
> branch gdb+binutils...
Both? :-)
--
Daniel Jacobowitz
CodeSourcery, LLC
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2005-03-29 1:25 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2005-03-22 0:20 Product branches in the GDB repository? Daniel Jacobowitz
2005-03-22 0:37 ` Stan Shebs
2005-03-22 1:21 ` Nick Roberts
2005-03-22 2:38 ` Daniel Jacobowitz
2005-03-22 19:43 ` Stan Shebs
2005-03-29 0:53 ` Bob Rossi
2005-03-29 1:25 ` Stan Shebs
2005-03-25 19:47 ` Daniel Jacobowitz
2005-03-25 19:56 ` Mark Kettenis
2005-03-25 20:54 ` Daniel Jacobowitz
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox