| Message ID | 1422635852-21569-1-git-send-email-andre.przywara@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Fri, Jan 30, 2015 at 04:37:32PM +0000, Andre Przywara wrote: > Commit ff61d185f4e7 ("mm: convert p[te|md]_mknonnuma and remaining > page table manipulations") removed a check in > mm/pgtable-generic.c:pmdp_invalidate(), which leaves the > pmd_mknotpresent macro the only user of the entry variable. > For ARM/LPAE we use a constant 0 without referencing the argument to > mark this condition, so the entry variable is no longer used here: > > mm/pgtable-generic.c: In function 'pmdp_invalidate': > mm/pgtable-generic.c:195:8: warning: unused variable 'entry' [-Wunused-variable] > pmd_t entry = *pmdp; > ^ > > Replace the ARM macro implementation with a static inline function to > get rid of this warning. > > Signed-off-by: Andre Przywara <andre.przywara@arm.com> > --- > Hi Andrew, > > Mel mentioned that I should send that simple fix below to you to > merge it with his original mmotm patch: > mm-convert-p_mknonnuma-and-remaining-page-table-manipulations.patch > The commit message above contains the mmotm-reference to this patch, > which will break when it hits mainline. > > Can you consider this? Given that's only a warning fix I am also > happy with resending it after -rc1, if you reckon that this arch/arm > change should not go via mmotm. This isn't correct though. > /* represent a notpresent pmd by zero, this is used by pmdp_invalidate */ > -#define pmd_mknotpresent(pmd) (__pmd(0)) > +static inline pmd_t pmd_mknotpresent(pmd_t pmd) > +{ > + return 0; > +} typedef struct { pmdval_t pmd; } pmd_t; typedef u32 pmdval_t; "0" is not compatible with pmd_t when STRICT_MM_TYPECHECKS is enabled.
diff --git a/arch/arm/include/asm/pgtable-3level.h b/arch/arm/include/asm/pgtable-3level.h index 370684d..0e51fbe 100644 --- a/arch/arm/include/asm/pgtable-3level.h +++ b/arch/arm/include/asm/pgtable-3level.h @@ -259,7 +259,10 @@ PMD_BIT_FUNC(mkyoung, |= PMD_SECT_AF); #define mk_pmd(page,prot) pfn_pmd(page_to_pfn(page),prot) /* represent a notpresent pmd by zero, this is used by pmdp_invalidate */ -#define pmd_mknotpresent(pmd) (__pmd(0)) +static inline pmd_t pmd_mknotpresent(pmd_t pmd) +{ + return 0; +} static inline pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot) {
Commit ff61d185f4e7 ("mm: convert p[te|md]_mknonnuma and remaining page table manipulations") removed a check in mm/pgtable-generic.c:pmdp_invalidate(), which leaves the pmd_mknotpresent macro the only user of the entry variable. For ARM/LPAE we use a constant 0 without referencing the argument to mark this condition, so the entry variable is no longer used here: mm/pgtable-generic.c: In function 'pmdp_invalidate': mm/pgtable-generic.c:195:8: warning: unused variable 'entry' [-Wunused-variable] pmd_t entry = *pmdp; ^ Replace the ARM macro implementation with a static inline function to get rid of this warning. Signed-off-by: Andre Przywara <andre.przywara@arm.com> --- Hi Andrew, Mel mentioned that I should send that simple fix below to you to merge it with his original mmotm patch: mm-convert-p_mknonnuma-and-remaining-page-table-manipulations.patch The commit message above contains the mmotm-reference to this patch, which will break when it hits mainline. Can you consider this? Given that's only a warning fix I am also happy with resending it after -rc1, if you reckon that this arch/arm change should not go via mmotm. On 30/01/15 10:06, Mel Gorman wrote: > Patch looks fine to me but this is against a mmotm patch that is > included in mmotm. The commit ID is not going to last. This should be > sent to Andrew Morton with a note saying it's a fix to the mmotm patch > mm-convert-p_mknonnuma-and-remaining-page-table-manipulations.patch . > He'll then fold your fix into the original patch so that bisection works > correctly when merged to mainline. Do you mind doing that please? Cheers, Andre arch/arm/include/asm/pgtable-3level.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)