On Mon, Apr 22, 2013 at 03:30:57PM +0530, Aneesh Kumar K.V wrote: > From: "Aneesh Kumar K.V" <aneesh.ku...@linux.vnet.ibm.com> > > Replace find_linux_pte with find_linux_pte_or_hugepte and explicitly > document why we don't need to handle transparent hugepages at callsites.
> diff --git a/arch/powerpc/kvm/book3s_hv_rm_mmu.c > b/arch/powerpc/kvm/book3s_hv_rm_mmu.c > index 19c93ba..aa6a351 100644 > --- a/arch/powerpc/kvm/book3s_hv_rm_mmu.c > +++ b/arch/powerpc/kvm/book3s_hv_rm_mmu.c > @@ -24,13 +24,15 @@ > /* Translate address of a vmalloc'd thing to a linear map address */ > static void *real_vmalloc_addr(void *x) > { > + unsigned shift; > unsigned long addr = (unsigned long) x; > pte_t *p; > > - p = find_linux_pte(swapper_pg_dir, addr); > + p = find_linux_pte_or_hugepte(swapper_pg_dir, addr, &shift); > if (!p || !pte_present(*p)) > return NULL; > /* assume we don't have huge pages in vmalloc space... */ > + BUG_ON(shift); Please don't add BUG_ON in this file. At this point we're basically still in guest context (though in real mode), and BUG_ON would cause a trap which we would handle very badly... Paul. _______________________________________________ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev