https://bugzilla.novell.com/show_bug.cgi?id=413842
User schueffler@softgarden.de added comment
https://bugzilla.novell.com/show_bug.cgi?id=413842#c46
--- Comment #46 from Stefan Schueffler 2008-09-04 12:24:17 MDT ---
I now tested this with a different controller from the same vendor:
LSI-20320-RB (mpt-driver-vendor-id: LSI53C1030). This one also is a
ultra-320-scsi-controller card, but it is the pci-X-variant
(the formerly used LSI-20320-IE / LSI53C1020 is the PCI-express-variant
of the same card).
Both use the MPT-driver, but surprisingly the PCI-X-one is working fine
when booting xen as opposite to the PCI-express.
The difference seems to be in the ioremap-call
mem = ioremap(mem_phys, msize) in mtpbase.c, line 1505
emitted in the driver (see comment #17):
Having the not-working (PCI-express) card equipped,
it will be called with
mem_phys = fc6e0000
msize = 131072
with the working (pci-X)-card
mem_phys = fc7e0000
msize = 131072
As the msize is the same, the problem seems to be related with the different
starting offset of the card-memory (mem_phys). I do not know what determines
this offset, and if the offset for both of these cards should be equal. (I just
changed the backplane from pci-express to pci-x, and changed the one card by
the other).
The next two weeks i do have an "idle" server with the 1030 card and the dell
remote-access-controller 5 free for testing purposes. If it could help you to
gather test-results or debugging information, i can give you access to the
hardware through the drac-5-controller and ssh-account. Just email me to give
you the account-infos (drac-access and root).
regards
Stefan Schueffler
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.