Re: [PATCH 4/4] x86/MSI-X: cleanup

[Andrew Cooper <andrew.cooper3@citrix.com>]

On 10/03/15 16:29, Jan Beulich wrote:
> - __pci_enable_msix() now checks that an MSI-X capability was actually
>   found
> - pass "pos" to msix_capability_init() as both callers already know it
>   (and hence there's no need to re-obtain it)
> - call __pci_disable_msi{,x}() directly instead of via
>   pci_disable_msi() from __pci_enable_msi{x,}() state validation paths

"__pci_enable_msi{x,}()'s state validation paths" ?

> - use msix_control_reg() instead of open coding it
> - log message adjustments
> - coding style corrections
>
> Signed-off-by: Jan Beulich <jbeulich@suse.com>

Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

It occurs to me that for further cleanup, it might be quite useful to
have a concrete sbdf type.

All this code does huge quantities of moving sbdf representations in and
out of a u32, and using 4 parameters instead of 1 to functions will
cause excessive register scheduling issues.  In addition, a custom %p
format identifier would help us move in the direction of consistent
representation (which we are currently a long way from).

~Andrew