Unfortunately the code wasn't that ready for holes in the processor id
range that I thought, so here is an updated patch that probably
address all sites.
Bert
On Fri, Mar 21, 2008 at 12:50 PM, Bert Wesarg
<bert.wesarg_at_[hidden]> wrote:
> To support holes in the logical processor IDs one can either try to
> access() all PLPA_BITMASK_CPU_MAX processors and continue on failure,
> or use readdir() to scan all entries of format "cpu%d". This patch
> does the latter.
>
> Bert
>
> ---
>
|