Hi Ludovic,

thank you for coming back on my call.

On Mi, 08 Feb 2006, Ludovic Rousseau wrote:

> On 08/02/06, Georg Lohrer <[EMAIL PROTECTED]> wrote:

[snipped]

> > IFDHGetCapabilities: Lun 0, Tag 4014, Length -1210531548
>
> This length is very suspect. The log line comes from the manufacturer
> driver. I don't know the application you are using by it looks like
> the application is calling SCardGetAttrib() with a strange length
> parameter.
>
> That may be the cause of the problem if the memory is then corrupted
> by the manufacturer driver.

Ah, that will gabe me the first anchor point to dive into the code more
deeply.
There is not deeper application behind the pcsc-lite but only the
ctapi-driver for the ReinerSCT cyberJack as provided by the manufacturer.
I
will do a gdb session to collect some more details.

>
> > What's ongoing there? Why will the system freeze?
> > As far as I can see in EHDestroyEventHandler() the
> > EHStatusHandlerThread should
> > be cancelled by setting dwLockId to 0xFFFF. The pthread_detach()
> > (SYS_ThreadDetach()) will be called successfully, but calling the
> > pthread_exit() (SYS_ThreadExit()) will result in holding the whole
> > system.
>
> A user application should _never_ block the kernel.
> Can you also try with a newer kernel?

Yes, absolutely.
I only know these situations having a race condition in a kernel driver. But
here it is all user-space, so it must happen in a syscall or as you suspected
by ruining the memory. Because I let pcscd run with root-privileges it might
run into lots of problems.

I thing the gdb will bring up some more details.

Thank your for your help.

Ciao, Georg
_______________________________________________
Muscle mailing list
[email protected]
http://lists.drizzle.com/mailman/listinfo/muscle

Reply via email to