Re: spin_unlock optimization(i386)

Alan Cox (alan@lxorguk.ukuu.org.uk)
Mon, 29 Nov 1999 20:59:05 +0000 (GMT)

> i believe this still doesnt turn off store forwarding. If two PCI commands
> are written to the same register then they might get 'merged', resulting
> in the first written command to be lost. This usually doesnt happen, but
> can happen, and does happen in some cases.

For the PC side read the PCI 2.1 specification. It say precisely what is
allowed and what isnt allowed on the PCI side and what a bridge may merge and
the ordering rules. There are no "might"s involved in the PCI bridge side of
things.

For CPU side merging Im not sure of the rules.

> the only MTRR stuff we do right now is mmap()-ing the frame buffer via
> /dev/mem, right? In that case we are doing things outside of ioremap()'s
> scope anyway.

We use it for I2O already. I'm playing at using it with other
things. On the I2O message posting its a "must have"

Alan

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/