GDB 7.4.1 release on Wed?

GDB 7.4.1 release on Wed?

[Joel Brobecker]

Hello,

I am thinking of creating the 7.4.1 release on Wed. If there is anything
that we should be waiting for, please let me know!

-- 
Joel

Re: GDB 7.4.1 release on Wed?

[Paweł Sikora]

On Monday 23 of April 2012 16:53:01 Joel Brobecker wrote:
> Hello,
> 
> I am thinking of creating the 7.4.1 release on Wed. If there is anything
> that we should be waiting for, please let me know!

just for note, i've recently reported a major user-blocker:

http://sourceware.org/bugzilla/show_bug.cgi?id=14002

current 7.4 and master are unusable for big c++ development :(

Re: GDB 7.4.1 release on Wed?

[Joel Brobecker]

> just for note, i've recently reported a major user-blocker:
> http://sourceware.org/bugzilla/show_bug.cgi?id=14002
> 
> current 7.4 and master are unusable for big c++ development :(

Do you confirm that the current head is still very slow for you?
Jan pointed at a couple of patches that fixed some of the slowness,
one of them being:

    http://sourceware.org/ml/gdb-patches/2012-02/msg00409.html

But if master is still drastically slow, then something else
is going on.

-- 
Joel

Re: GDB 7.4.1 release on Wed?

[Joel Brobecker]

> current master (57a68b7) is faster (~1 minute for the testing step)
> and eats less memory (~500M -> ~1G during step) than 7.4-head but
> the 7.3-head is the fastest one (2..3 seconds! for testing step).

Hmmm, in that case, we're going to need more time to figure out
what's going on, and a lot of us are really quite busy working
on other things. So I am afraid that we'll be out of luck for 7.4.1.
We can port the patch I mentioned in my previous email to make
things a little better, but otherwise we'll have to wait for 7.5
for a proper fix.

-- 
Joel

Re: GDB 7.4.1 release on Wed?

[Paweł Sikora]

On Tuesday 24 of April 2012 09:20:59 Joel Brobecker wrote:
> > current master (57a68b7) is faster (~1 minute for the testing step)
> > and eats less memory (~500M -> ~1G during step) than 7.4-head but
> > the 7.3-head is the fastest one (2..3 seconds! for testing step).
> 
> Hmmm, in that case, we're going to need more time to figure out
> what's going on, and a lot of us are really quite busy working
> on other things.

fyi, i've posted a comment about cpu hog on git/master:
http://sourceware.org/bugzilla/show_bug.cgi?id=14002#c1