[openib-general] Failure in reset HCA with backport-svn4507-to-2.6.9

[Ranjit Pandit]

Has anybody seen this problem before?

modprobe ib_mthca

ib_mthca: Mellanox InfiniBand HCA driver v0.06 (June 23, 2005)
ib_mthca: Initializing 0000:03:00.0
ACPI: PCI interrupt 0000:03:00.0[A] -> GSI 24 (level, low) -> IRQ 233
ib_mthca 0000:03:00.0: PCI device did not come back after reset, aborting.
ib_mthca 0000:03:00.0: Failed to reset HCA, aborting.

I'm using 'infiniband-backport-svn4507-to-2.6.9' on RHEL 4 U2 kernel
(2.6.9-22.EL).

To confirm that the HCA wasn't bad, I tried to bring up SilverStorm
stack and that came up fine - port goes Active etc...

Here is what lspci shows before and after "modprobe ib_mthca"

Before:
02:06.0 PCI bridge: Mellanox Technologies MT23108 PCI Bridge (rev a1)
03:00.0 InfiniBand: Mellanox Technologies MT23108 InfiniHost (rev a1)

After:
02:06.0 PCI bridge: Mellanox Technologies MT23108 PCI Bridge (rev ff)
                                                                      
                        ^^^^^
03:00.0 InfiniBand: Mellanox Technologies MT23108 InfiniHost (rev ff)


I presume that is because of the problem during reset.

Any ideas why it's having problems resetting the card?

thanks,

Ranjit