Messages in this thread | | | Date | Wed, 16 Jul 2003 16:25:32 +0100 | From | Dave Jones <> | Subject | Re: Question about free_one_pgd() changes in these 3.5G patches |
| |
On Wed, Jul 16, 2003 at 01:31:39PM +0100, Hugh Dickins wrote:
Both this..
> > for (j = 0; j < PTRS_PER_PMD ; j++) { > > prefetchw(pmd+j+(PREFETCH_STRIDE/16)); > > free_one_pmd(pmd+j); > > }
and this..
> > pmd_t * pmd, * md, * emd; > > for (md = pmd, emd = pmd + PTRS_PER_PMD; md < emd; md++) { > > prefetchw(md+(PREFETCH_STRIDE/16)); > > free_one_pmd(md); > > }
both use the prefetch that was removed in 2.5 for 'being bogus'. It can prefetch past the end iirc, which is fatal on some CPUs.
Dave
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |