Re: [RFC PATCH 1/2] powerpc/mm: Update prot_none implementation using _PAGE_READ
From: Paul Mackerras <hidden>
Date: 2016-03-02 01:08:03
On Fri, Feb 26, 2016 at 08:50:49AM +0530, Aneesh Kumar K.V wrote:
Now that we have _PAGE_READ use that to implement prot none. With this prot_none is _PAGE_PRESENT with none of the access bits set. While hashing we map that to PP bit 00. With this implementation, we will now take a prot fault for prot none ptes, whereas before, we never inserted such a pte to hash. Hence we always got nohpte fault before. This is in preparation to remove _PAGE_USER from book3s 64
Mostly looks good, but I have a comment:
quoted hunk ↗ jump to hunk
@@ -176,10 +177,12 @@ #define PAGE_SHARED __pgprot(_PAGE_BASE | _PAGE_USER | _PAGE_RW) #define PAGE_SHARED_X __pgprot(_PAGE_BASE | _PAGE_USER | _PAGE_RW | \ _PAGE_EXEC)
Don't we need _PAGE_READ in PAGE_SHARED[_X] now? Paul.