Race condition (although with x86 being globally ordered, it probably can't
actually happen.) The bitmask is probably the way to go.
On April 21, 2014 6:28:12 PM PDT, Andrew Lutomirski <aml...@gmail.com> wrote:
>On Mon, Apr 21, 2014 at 6:14 PM, H. Peter Anvin <h...@zytor.com> wrote:
>> I wanted to avoid the "another cpu made this allocation, now I have
>to free" crap, but I also didn't want to grab the lock if there was no
>work needed.
>
>I guess you also want to avoid bouncing all these cachelines around on
>boot on bit multicore machines.
>
>I'd advocate using the bitmap approach or simplifying the existing
>code. For example:
>
>+ for (n = 0; n < ESPFIX_PUD_CLONES; n++) {
>+ pud = ACCESS_ONCE(pud_p[n]);
>+ if (!pud_present(pud))
>+ return false;
>+ }
>
>I don't see why that needs to be a loop. How can one clone exist but
>not the others?
>
>--Andy
--
Sent from my mobile phone. Please pardon brevity and lack of formatting.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
