| Message ID | 1370462344-7128-1-git-send-email-gmbnomis@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Jun 05, 2013 at 08:59:04PM +0100, Simon Baatz wrote: > Commit f8b63c1 made flush_kernel_dcache_page a no-op assuming that the pages > it needs to handle are kernel mapped only. However, for example when doing > direct I/O, pages with user space mappings may occur. > > Thus, continue to do lazy flushing if there are no user space mappings. > Otherwise, flush the kernel cache lines directly. > > Signed-off-by: Simon Baatz <gmbnomis@gmail.com> > Cc: <stable@vger.kernel.org> # 3.2+ > CC: Catalin Marinas <catalin.marinas@arm.com> > Cc: Russell King <linux@arm.linux.org.uk> Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Hi Catalin, On Mon, Jun 10, 2013 at 11:23:07AM +0100, Catalin Marinas wrote: > On Wed, Jun 05, 2013 at 08:59:04PM +0100, Simon Baatz wrote: > > Commit f8b63c1 made flush_kernel_dcache_page a no-op assuming that the pages > > it needs to handle are kernel mapped only. However, for example when doing > > direct I/O, pages with user space mappings may occur. > > > > Thus, continue to do lazy flushing if there are no user space mappings. > > Otherwise, flush the kernel cache lines directly. > > > > Signed-off-by: Simon Baatz <gmbnomis@gmail.com> > > Cc: <stable@vger.kernel.org> # 3.2+ > > CC: Catalin Marinas <catalin.marinas@arm.com> > > Cc: Russell King <linux@arm.linux.org.uk> > > Reviewed-by: Catalin Marinas <catalin.marinas@arm.com> Thanks for the review. This is submitted now as patch #7755/1 in Russell's patch system. - Simon
diff --git a/arch/arm/include/asm/cacheflush.h b/arch/arm/include/asm/cacheflush.h index bff7138..17d0ae8 100644 --- a/arch/arm/include/asm/cacheflush.h +++ b/arch/arm/include/asm/cacheflush.h @@ -320,9 +320,7 @@ static inline void flush_anon_page(struct vm_area_struct *vma, } #define ARCH_HAS_FLUSH_KERNEL_DCACHE_PAGE -static inline void flush_kernel_dcache_page(struct page *page) -{ -} +extern void flush_kernel_dcache_page(struct page *); #define flush_dcache_mmap_lock(mapping) \ spin_lock_irq(&(mapping)->tree_lock) diff --git a/arch/arm/mm/flush.c b/arch/arm/mm/flush.c index 0d473cc..32aa586 100644 --- a/arch/arm/mm/flush.c +++ b/arch/arm/mm/flush.c @@ -301,6 +301,39 @@ void flush_dcache_page(struct page *page) EXPORT_SYMBOL(flush_dcache_page); /* + * Ensure cache coherency for the kernel mapping of this page. We can + * assume that the page is pinned via kmap. + * + * If the page only exists in the page cache and there are no user + * space mappings, this is a no-op since the page was already marked + * dirty at creation. Otherwise, we need to flush the dirty kernel + * cache lines directly. + */ +void flush_kernel_dcache_page(struct page *page) +{ + if (cache_is_vivt() || cache_is_vipt_aliasing()) { + struct address_space *mapping; + + mapping = page_mapping(page); + + if (!mapping || mapping_mapped(mapping)) { + void *addr; + + addr = page_address(page); + /* + * kmap_atomic() doesn't set the page virtual + * address for highmem pages, and + * kunmap_atomic() takes care of cache + * flushing already. + */ + if (!IS_ENABLED(CONFIG_HIGHMEM) || addr) + __cpuc_flush_dcache_area(addr, PAGE_SIZE); + } + } +} +EXPORT_SYMBOL(flush_kernel_dcache_page); + +/* * Flush an anonymous page so that users of get_user_pages() * can safely access the data. The expected sequence is: *
Commit f8b63c1 made flush_kernel_dcache_page a no-op assuming that the pages it needs to handle are kernel mapped only. However, for example when doing direct I/O, pages with user space mappings may occur. Thus, continue to do lazy flushing if there are no user space mappings. Otherwise, flush the kernel cache lines directly. Signed-off-by: Simon Baatz <gmbnomis@gmail.com> Cc: <stable@vger.kernel.org> # 3.2+ CC: Catalin Marinas <catalin.marinas@arm.com> Cc: Russell King <linux@arm.linux.org.uk> --- Changes: in V5: - simplified highmem case; we can assume kmapped pages in V4: - get back to simpler version of flush_kernel_dcache_page() and use the logic from __flush_dcache_page() to flush the kernel mapping (which also takes care of highmem pages) in V3: - Followed Catalin's suggestion to reverse the order of the patches in V2: - flush_kernel_dcache_page() follows flush_dcache_page() now, except that it does not flush the user mappings arch/arm/include/asm/cacheflush.h | 4 +--- arch/arm/mm/flush.c | 33 +++++++++++++++++++++++++++++++++ 2 files changed, 34 insertions(+), 3 deletions(-)