Re: Seven more unaligned PCI configuration space access bugs

[Matthias Drochner <M.Drochner%fz-juelich.de@localhost>]

david%l8s.co.uk@localhost said:
> Maybe there ought to be a pci_conf_read_8() etc ? 

I don't think this is worth the effort. It would have
to be implemented for all config space access methods
on all platforms, and it would be used on only a few
places.

Also, I believe there are host bridges which only
implement aligned 32-bit accesses. These would have
to emulate 8/16-bit accesses in software. Since there
are registers with write-one-to-clear semantics, this
cannot be done safely in generic code.

I'd say what we do now is reasonable.

best regards
Matthias




-------------------------------------------------------------------
-------------------------------------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich

Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzende des Aufsichtsrats: MinDir'in Baerbel Brumme-Bothe
Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender),
Dr. Ulrich Krafft (stellv. Vorsitzender), Prof. Dr. Harald Bolt,
Dr. Sebastian M. Schmidt
-------------------------------------------------------------------
-------------------------------------------------------------------