[Openib-windows] Some more additions questions about the patch:

[Tzachi Dar]

While working on checking my patch with a kernel mode client I have found
out that there are some more things that should be changed in the IPOIB
code.
 
1)     gp_mem_tracker: It seems that there is a call to __cl_mem_track_start
every time there is an adapter that is being created, and a similar call
when an adapter is being destroyed. It is not clear to me, why this is so
and you are not doing this in the Driver-Entry Driver Unload functions. In
any case if you keep things as they are, please note that on
__cl_mem_track_stop you have to set gp_mem_tracker to null if the reference
has reached zero, otherwise if the reference has reached zero, things won't
work again. (by the way, if you keep things as they are things won't work in
any case, since there is a race between the reference count and the global
variable :-(, it probably must be done in DriverEntry)
2)     I have moved the place in which the adapter is being created. It
should better be created when the number of adapters in the list goes from 0
to 1, This is the case in which there is someone holding the device, the
IPOIB ports are being disabled and re-enabled again.
3)     Please add the case of IRP_MJ_INTERNAL_DEVICE_CONTROL to the function
__ipoib_dispatch. It should be handled just like IRP_MJ_DEVICE_CONTROL. 
 
Thanks
Tzachi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/ofw/attachments/20050918/2da751fa/attachment.html>