[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [microblaze-uclinux] ptrace and gdbserver (with MMU)
Hi Steve,
The ptrace / gdbserver combo has not been tested or validated for MMU,
so I'm not surprised if there are issues. There are also quite likely
to be some changed requirements in mb-linux-gdb. To be honest, until
this week no-one has ever asked about gdb/gdbserver for the MMU
variant.
user/gdb/gdbserver has modifications in it to support noMMU - I don't
recall what sort of #ifdef these are guarded by for MicroBlaze, but
when this work was done for MicroBlaze, the MMU was not on the
horizon. So, there's a starting point to go looking for issues.
The ptrace kernel support should be mostly there, again it all works
for noMMU but has not been tested for MMU. It should be pretty
generic - actually I seem to recall a patch on LKLM recently that
moved microblaze to an almost completely generic ptrace
implementation. That might help you in trying to get it going in
2.6.20.
As mb-*-gdb, again I put the mommu/uclinux/gdbserver hooks into the
nommu microblaze GDB, but we haven't looked at the MMU variant.
All of this is on the agenda for the work being done currently at
kernel.org, so if you are in a hurry you may awnt to start looking
yourself, otherwise it's likely that as we cover these issues in
mainline we can backport the fixes to petalinux.
Regards,
John
On Thu, May 21, 2009 at 2:09 AM, Steven Hein <ssh@xxxxxxx> wrote:
> Hi all--
>
> Is gdbserver (and the underlying ptrace functionality) expected to
> be working properly in the Petalogix SVN tree? I've been trying
> to get gdbserver working on my platform, the app being debugged
> would segv immediately after the first "continue" from GDB.
> While trying to debug this, I found some simple ptrace
> examples (i.e. to simply trace system calls in a child process),
> and found that this didn't appear to work (the PTRACE_SYSCALL
> op didn't trace system calls, and the PC register value that I got
> didn't seem to be valid either).
>
> I saw these failures on my own custom kernel, which is very
> close to the current SVN kernel (with MMU enabled). Then
> I tried running my test app on the pre-built platform for my
> board (S3ADSP1800 w/ MMU), and I saw the same problems.
>
> Can anyone tell me if they have gdbserver working on an
> MMU-enabled Microblaze platform? If it's not working right now
> in the SVN, could anyone give me any hints as to where the
> problem might be?
>
> Thanks!
> Steve Hein
> ___________________________
> microblaze-uclinux mailing list
> microblaze-uclinux@xxxxxxxxxxxxxx
> Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
> Mailing List Archive :
> http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/
>
>
--
John Williams, PhD, B.Eng, B.IT
PetaLogix - Linux Solutions for a Reconfigurable World
w: www.petalogix.com p: +61-7-30090663 f: +61-7-30090663
___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/