From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932298AbWHCHoA (ORCPT ); Thu, 3 Aug 2006 03:44:00 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932325AbWHCHoA (ORCPT ); Thu, 3 Aug 2006 03:44:00 -0400 Received: from ozlabs.tip.net.au ([203.10.76.45]:8902 "EHLO ozlabs.org") by vger.kernel.org with ESMTP id S932298AbWHCHn7 (ORCPT ); Thu, 3 Aug 2006 03:43:59 -0400 Subject: [PATCH 3/4] Trivial move of ptep_set_access_flags From: Rusty Russell To: Linus Torvalds Cc: lkml - Kernel Mailing List , Andi Kleen In-Reply-To: <1154590994.11423.54.camel@localhost.localdomain> References: <1154590902.11423.52.camel@localhost.localdomain> <1154590994.11423.54.camel@localhost.localdomain> Content-Type: text/plain Date: Thu, 03 Aug 2006 17:43:57 +1000 Message-Id: <1154591038.11423.56.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Evolution 2.6.1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Move ptep_set_access_flags to be closer to the other ptep accessors, and make the indentation standard. Signed-off-by: Zachary Amsden Signed-off-by: Rusty Russell diff -urpN --exclude TAGS -X /home/rusty/devel/kernel/kernel-patches/current-dontdiff --minimal working-2.6.18-rc2-mm1-zach-2/include/asm-i386/pgtable.h working-2.6.18-rc2-mm1-zach-3/include/asm-i386/pgtable.h --- working-2.6.18-rc2-mm1-zach-2/include/asm-i386/pgtable.h 2006-08-03 17:28:26.000000000 +1000 +++ working-2.6.18-rc2-mm1-zach-3/include/asm-i386/pgtable.h 2006-08-03 17:29:59.000000000 +1000 @@ -246,6 +246,22 @@ static inline pte_t pte_mkhuge(pte_t pte # include #endif +/* + * We only update the dirty/accessed state if we set + * the dirty bit by hand in the kernel, since the hardware + * will do the accessed bit for us, and we don't want to + * race with other CPU's that might be updating the dirty + * bit at the same time. + */ +#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS +#define ptep_set_access_flags(vma, address, ptep, entry, dirty) \ +do { \ + if (dirty) { \ + (ptep)->pte_low = (entry).pte_low; \ + flush_tlb_page(vma, address); \ + } \ +} while (0) + #define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_DIRTY static inline int ptep_test_and_clear_dirty(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep) { @@ -415,23 +431,8 @@ extern void noexec_setup(const char *str /* * The i386 doesn't have any external MMU info: the kernel page * tables contain all the necessary information. - * - * Also, we only update the dirty/accessed state if we set - * the dirty bit by hand in the kernel, since the hardware - * will do the accessed bit for us, and we don't want to - * race with other CPU's that might be updating the dirty - * bit at the same time. */ #define update_mmu_cache(vma,address,pte) do { } while (0) -#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS -#define ptep_set_access_flags(__vma, __address, __ptep, __entry, __dirty) \ - do { \ - if (__dirty) { \ - (__ptep)->pte_low = (__entry).pte_low; \ - flush_tlb_page(__vma, __address); \ - } \ - } while (0) - #endif /* !__ASSEMBLY__ */ #ifdef CONFIG_FLATMEM -- Help! Save Australia from the worst of the DMCA: http://linux.org.au/law