* Multi process debugging using gdb - references?
@ 2007-03-26 14:39 Fred Fish
2007-03-26 14:51 ` Daniel Jacobowitz
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: Fred Fish @ 2007-03-26 14:39 UTC (permalink / raw)
To: gdb
I'm curious about the status of using gdb to debug multiple processes.
I've read most of the past messages I could find by doing a search on
"multiprocess" and "multi process" so I know about the recent work to
support multiple processes that share the same executable file, such
as following forked processes.
What I'm more interested in is being able to control and debug
multiple processes that don't share the same executable, may not even
be running on the same machine, and may not be running on machines of
the same architecture.
My gut feeling says that using some front end to control multiple
instances of gdb is probably a better solution than trying to enhance
gdb to handle this using a single gdb process.
Also, any pointers/references to published papers on multi process
debugging, other debuggers that implement multi process debugging
(such as Totalview), works in progress, etc, would be appreciated.
-Fred
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Multi process debugging using gdb - references?
2007-03-26 14:39 Multi process debugging using gdb - references? Fred Fish
@ 2007-03-26 14:51 ` Daniel Jacobowitz
2007-03-26 15:54 ` Gary Funck
[not found] ` <1174945438.24110.48.camel@localhost.localdomain>
2 siblings, 0 replies; 4+ messages in thread
From: Daniel Jacobowitz @ 2007-03-26 14:51 UTC (permalink / raw)
To: gdb
On Mon, Mar 26, 2007 at 07:39:13AM -0700, Fred Fish wrote:
> My gut feeling says that using some front end to control multiple
> instances of gdb is probably a better solution than trying to enhance
> gdb to handle this using a single gdb process.
I have not done any work in the multi-process area in years, but
that's my gut feeling too. GDB's internals and user interface are
both very wound up in a single process model.
That's not to say it couldn't be changed. But you'd have to convince
us first that it was worth changing, and then that it was practical.
--
Daniel Jacobowitz
CodeSourcery
^ permalink raw reply [flat|nested] 4+ messages in thread
* RE: Multi process debugging using gdb - references?
2007-03-26 14:39 Multi process debugging using gdb - references? Fred Fish
2007-03-26 14:51 ` Daniel Jacobowitz
@ 2007-03-26 15:54 ` Gary Funck
[not found] ` <1174945438.24110.48.camel@localhost.localdomain>
2 siblings, 0 replies; 4+ messages in thread
From: Gary Funck @ 2007-03-26 15:54 UTC (permalink / raw)
To: gdb
> From: Fred Fish
> Sent: Monday, March 26, 2007 7:39 AM
[...]
> Also, any pointers/references to published papers on multi process
> debugging, other debuggers that implement multi process debugging
> (such as Totalview), works in progress, etc, would be appreciated.
Here are a couple. Since the first reference is an abstract, and
the second is a survey, it might be best to follow up with the
authors, or to follow the references. - Gary
http://www2.imm.dtu.dk/~jw/para04/Abstracts/susanne_balle/susanne_balle.html
Like Ygdrasil, the world tree of Norse mythology, the Grid promises to cover
the entire world, holding it together and feeding its inhabitants with data
and processing power. Serving this potentially huge number of processors
(1,000+) affects many software tools such as programming development tools,
performance enhancement tools and system administration tools. These tools
need to deal with massive amount of data created by the large number of
processes as well as be scalable. The data need to be condensed and
reorganized into useful information before it is presented to the user. We
present a modular aggregating network toolkit, which can be tailored to
almost any purpose involving reduction of data. The latter allows the user
to visualize as well as re-arrange the data to fit his need (integration
with statistical packages, visualization, etc.). The user can write his own
plug-ins that fit his needs and thereby instruct the network about what is
getting aggregated, how to aggregate the data (merge identical data, discard
data, etc.), as well as how to display the data in the user portal.
http://www.ogf.org/documents/GFD.33.pdf
This memo provides information to the Grid community regarding program
development tools currently used by users developing programs on large scale
systems and on the Grid. In addition, this memo provides information
regarding what tools are needed in order to facilitate users' transition to
the Grid.
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Multi process debugging using gdb - references?
[not found] ` <1174945438.24110.48.camel@localhost.localdomain>
@ 2007-03-26 22:13 ` Michael Snyder
0 siblings, 0 replies; 4+ messages in thread
From: Michael Snyder @ 2007-03-26 22:13 UTC (permalink / raw)
To: fnf; +Cc: gdb
On Mon, 2007-03-26 at 14:43 -0700, Michael Snyder wrote:
> On Mon, 2007-03-26 at 07:39 -0700, Fred Fish wrote:
> > I'm curious about the status of using gdb to debug multiple processes.
> >
>
> I used to know about a project that did as you suggest -- use a
> front end to control multiple gdbs. Let me see if I can dredge
> up the memory.
>
Fred, google up "P2D2", by Robert Hood (he was at NASA at the
time, don't know if he still is). Here's one reference:
http://www.dl.ac.uk/TCSC/Subjects/Parallel_Algorithms/tools_survey/node37.html
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2007-03-26 22:13 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-03-26 14:39 Multi process debugging using gdb - references? Fred Fish
2007-03-26 14:51 ` Daniel Jacobowitz
2007-03-26 15:54 ` Gary Funck
[not found] ` <1174945438.24110.48.camel@localhost.localdomain>
2007-03-26 22:13 ` Michael Snyder
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox