RE: GDB solib interface

["Smith, Stephen \(SWCOE\)" <Stephen dot Smith at honeywell dot com>]

Ok, I have our gdbserver using the updated solib-remote gdb patch using
the dll stop packet.  You said that you've tested the load/unload packet
so as far as I know the patch has been successfully updated to compile
and link with 6.5 (which was the last release when I started work on
it).  I will make sure that the patch compiles/links against head.
Should I propose the code patch before the documentation patch is
created - so that the code can be critiqued?

I remember that I need to make sure that the files that are to be
changed are formatted properly (Kevin Buettner said to use "indent -fca 
-lc80 -l80").  When I do that the patch grows by about a factor of 10
due to unrelated items being reformatted to match the GNU coding
standard.  Do you want me to run indent before or after the patch is
reviewed?

Also how do we get the published changes to the gdb/gdbserver protocol
approved?

The interface documentation does not need changing unless someone here
doesn't like it.  Then again I've never updated the gdb documentation
and so don't know how to go about it.  Could someone point me in the
right direction?

BTW - I have a copyright assignment on file with FSF and nothing has
changed on this end to negate it.

-----Original Message-----
From: Daniel Jacobowitz [mailto:drow@false.org] 
Sent: Thursday, January 04, 2007 9:01 PM
To: Smith, Stephen (SWCOE)
Cc: Stephen & Linda Smith; gdb@sourceware.org
Subject: Re: GDB solib interface

On Thu, Jan 04, 2007 at 08:54:54PM -0700, Smith, Stephen (SWCOE) wrote:
> Ok,  I have the string 'T05dll:dll'.  Single stepping trough the code,
I
> get p1 -> ll:dll and p -> dll:dll.  Therefore I never enter the block.

> 
> What did I do wrong with the packet?

I said I was certain load and unload worked - I hadn't tested dll.
Looks like the leading 'd' means it needs to be handled just like
awatch above.

-- 
Daniel Jacobowitz
CodeSourcery