This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
RE: Startup problem
- From: Rycerz Roman-G11565 <roman dot rycerz at motorola dot com>
- To: ecos-discuss at sources dot redhat dot com
- Date: Thu, 10 Jun 2004 13:28:02 -0500
- Subject: RE: [ECOS] Startup problem
I have tried running and re-running the same binary..... Same results ran a memcheck program and it checks out ok.
I have increased my stack to 32k, and already removed the -O2 option
Roman
-----Original Message-----
From: Gary Thomas [mailto:gary@mlbassoc.com]
Sent: Thursday, June 10, 2004 11:01 AM
To: Rycerz Roman-G11565
Cc: ecos-discuss@sources.redhat.com
Subject: RE: [ECOS] Startup problem
On Thu, 2004-06-10 at 09:49, Rycerz Roman-G11565 wrote:
> Sorry I keep forgetting to include the environments
>
> * What target platform [processor architecture] -> Power PC on a MPC5554
> * What tools are you using -> gcc 2.95.3-10, Insight 5.2, ecos config tool 2.11
> * What eCos configuration -> I/O subsystem, Serial Device Drivers, eCos Kernel
> * Is this repeatable or random -> random in the sense that from 1 compile to the next it comes and goes
>
What about rerunning the same binary? If it fails, does it always fail, or only sometimes? If it's the latter, I'd suspect a RAM problem. If it always fails, then it may be a compiler issue.
>
> Roman
>
>
>
> -----Original Message-----
> From: Gary Thomas [mailto:gary@mlbassoc.com]
> Sent: Thursday, June 10, 2004 10:33 AM
> To: Sven Rehfuss
> Cc: Rycerz Roman-G11565; ecos-discuss@sources.redhat.com
> Subject: Re: [ECOS] Startup problem
>
>
> On Thu, 2004-06-10 at 09:23, Sven Rehfuss wrote:
> > Hi,
> > use a debugger and check the code. If the routines modify stack
> > variables before decrementing the stack pointer, it maybe is the same
> > problem I had. If it is the same problem, I can give you the solution.
>
> Why not share this solution with us all? After all, that's what this
> forum is for!
>
> >
> > Regards,
> > Sven Rehfu?
> >
> >
> > Am Do, den 10.06.2004 schrieb Rycerz Roman-G11565 um 17:08:
> > > Hi
> > >
> > > I have an intermittent problem. When starting up the eCos kernel
> > > the call to "cyg_hal_invoke_constructors" sometimes fails with a
> > > segmentation violation. It seems to happening when the destructors
> > > are being called. in thread.cxx
> > >
> > > its in the routine "~cyg_thread_remove_from_list", it seem to be
> > > working it way back thru a linked list assigning
> > > prev = prev->list_next when it blows up prev is set to the contents of memory "0", location a null ptr problem.
> > >
> > > I know this is a quite vague description, and I may all wrong..
> > > But
> > > does anyone have any ideas why this is happening ???
>
> You've not told us much about your environment.
> * What target platform [processor architecture]?
> * What tools are you using
> * What eCos configuration
> * Is this repeatable or random
>
> --
> Gary Thomas <gary@mlbassoc.com>
> MLB Associates
--
Gary Thomas <gary@mlbassoc.com>
MLB Associates
--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss