Hello all, hello Mr. Roberts!
After clearing some general misunderstandings, maybe you could help me
once more with the second part of my former post. I would be very
grateful for your assistance.
I described it as follows:
>Depending on the pipe type I leave the dispatch routine forwarding the
>current Irp and device object to a validation routine for isochronous
>transfers (because they are more complicated) or bulk transfer
>completion routine. Isochronous completion is done as it can be seen
>the original sample. Microsoft did in their samples the validation for
>isochronous and both validation and transaction in the case of bulk
>transfers in the dispatch routine. Although I copied and pasted the
>whole algorithm from their (fine-working [standalone]) drivers, the system
>crashes when I try to perform a bulk/isochronous read or write. What
>rule did I violate?
Moreover I recognized the same crash behaviour when I unplug one or
both devices. I tried to get some debugging messages via DbgView from
Sysinternals, but the system crashes, without storing the information
into the given log file.
I checked the in- and decrements on the stack without result. They seem
okay to me. Since I solely introduced the validation routine for
isochronous transfer and sourced the transfer algorithm of the bulk
transmission out, I assume it must have something to do with the
parameters I forward to these functions. I appreciate your help in
advance, special thanks to Mr. Roberts who helped me this far.