| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-Id: <1333670495-7016-4-git-send-email-suresh.b.siddha@intel.com>
Date: Thu, 5 Apr 2012 17:01:35 -0700
From: Suresh Siddha <suresh.b.siddha@...el.com>
To: Konstantin Khlebnikov <khlebnikov@...nvz.org>,
Konstantin Khlebnikov <koct9i@...il.com>, linux-mm@...ck.org,
Andrew Morton <akpm@...ux-foundation.org>,
linux-kernel@...r.kernel.org
Cc: Andi Kleen <andi@...stfloor.org>,
Suresh Siddha <suresh.b.siddha@...el.com>,
Pallipadi Venkatesh <venki@...gle.com>,
Ingo Molnar <mingo@...hat.com>,
"H. Peter Anvin" <hpa@...or.com>,
Linus Torvalds <torvalds@...ux-foundation.org>,
Nick Piggin <npiggin@...nel.dk>
Subject: [v3 VM_PAT PATCH 3/3] mm, x86, PAT: rework linear pfn-mmap tracking
From: Konstantin Khlebnikov <khlebnikov@...nvz.org>
This patch replaces generic vma-flag VM_PFN_AT_MMAP with x86-only VM_PAT.
We can toss mapping address from remap_pfn_range() into track_pfn_vma_new(),
and collect all PAT-related logic together in arch/x86/.
This patch also restores orignal frustration-free is_cow_mapping() check in
remap_pfn_range(), as it was before commit v2.6.28-rc8-88-g3c8bb73
("x86: PAT: store vm_pgoff for all linear_over_vma_region mappings - v3")
is_linear_pfn_mapping() checks can be removed from mm/huge_memory.c,
because it already handled by VM_PFNMAP in VM_NO_THP bit-mask.
-v2: Reset the VM_PAT flag as part of untrack_pfn_vma()
-v3: Adapt to the track_pfn_insert/track_pfn_remap API
Signed-off-by: Konstantin Khlebnikov <khlebnikov@...nvz.org>
Signed-off-by: Suresh Siddha <suresh.b.siddha@...el.com>
Cc: Venkatesh Pallipadi <venki@...gle.com>
Cc: H. Peter Anvin <hpa@...or.com>
Cc: Nick Piggin <npiggin@...nel.dk>
Cc: Ingo Molnar <mingo@...hat.com>
---
arch/x86/mm/pat.c | 17 ++++++++++++-----
include/asm-generic/pgtable.h | 6 ++++--
include/linux/mm.h | 15 +--------------
mm/huge_memory.c | 7 +++----
mm/memory.c | 12 ++++++------
5 files changed, 26 insertions(+), 31 deletions(-)
diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c
index d0553bf..bef33df 100644
--- a/arch/x86/mm/pat.c
+++ b/arch/x86/mm/pat.c
@@ -665,7 +665,7 @@ int track_pfn_copy(struct vm_area_struct *vma)
unsigned long vma_size = vma->vm_end - vma->vm_start;
pgprot_t pgprot;
- if (is_linear_pfn_mapping(vma)) {
+ if (vma->vm_flags & VM_PAT) {
/*
* reserve the whole chunk covered by vma. We need the
* starting address and protection from pte.
@@ -687,14 +687,20 @@ int track_pfn_copy(struct vm_area_struct *vma)
* single reserve_pfn_range call.
*/
int track_pfn_remap(struct vm_area_struct *vma, pgprot_t *prot,
- unsigned long pfn, unsigned long size)
+ unsigned long pfn, unsigned long addr, unsigned long size)
{
unsigned long flags;
int i;
/* reserve the whole chunk starting from pfn */
- if (is_linear_pfn_mapping(vma))
- return reserve_pfn_range(pfn << PAGE_SHIFT, size, prot, 0);
+ if (addr == vma->vm_start && size == (vma->vm_end - vma->vm_start)) {
+ int ret;
+
+ ret = reserve_pfn_range(pfn << PAGE_SHIFT, size, prot, 0);
+ if (!ret)
+ vma->vm_flags |= VM_PAT;
+ return ret;
+ }
if (!pat_enabled)
return 0;
@@ -741,7 +747,7 @@ void untrack_pfn(struct vm_area_struct *vma, unsigned long pfn,
resource_size_t paddr;
unsigned long prot;
- if (!is_linear_pfn_mapping(vma))
+ if (!(vma->vm_flags & VM_PAT))
return;
/* free the chunk starting from pfn or the whole chunk */
@@ -755,6 +761,7 @@ void untrack_pfn(struct vm_area_struct *vma, unsigned long pfn,
size = vma->vm_end - vma->vm_start;
}
free_pfn_range(paddr, size);
+ vma->vm_flags &= ~VM_PAT;
}
pgprot_t pgprot_writecombine(pgprot_t prot)
diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
index a877649..ddd613e 100644
--- a/include/asm-generic/pgtable.h
+++ b/include/asm-generic/pgtable.h
@@ -392,7 +392,8 @@ static inline void ptep_modify_prot_commit(struct mm_struct *mm,
* by remap_pfn_range() for physical range indicated by pfn and size.
*/
static inline int track_pfn_remap(struct vm_area_struct *vma, pgprot_t *prot,
- unsigned long pfn, unsigned long size)
+ unsigned long pfn, unsigned long addr,
+ unsigned long size)
{
return 0;
}
@@ -427,7 +428,8 @@ static inline void untrack_pfn(struct vm_area_struct *vma,
}
#else
extern int track_pfn_remap(struct vm_area_struct *vma, pgprot_t *prot,
- unsigned long pfn, unsigned long size);
+ unsigned long pfn, unsigned long addr,
+ unsigned long size);
extern int track_pfn_insert(struct vm_area_struct *vma, pgprot_t *prot,
unsigned long pfn);
extern int track_pfn_copy(struct vm_area_struct *vma);
diff --git a/include/linux/mm.h b/include/linux/mm.h
index d8738a4..b8e5fe5 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -117,7 +117,7 @@ extern unsigned int kobjsize(const void *objp);
#define VM_CAN_NONLINEAR 0x08000000 /* Has ->fault & does nonlinear pages */
#define VM_MIXEDMAP 0x10000000 /* Can contain "struct page" and pure PFN pages */
#define VM_SAO 0x20000000 /* Strong Access Ordering (powerpc) */
-#define VM_PFN_AT_MMAP 0x40000000 /* PFNMAP vma that is fully mapped at mmap time */
+#define VM_PAT 0x40000000 /* PAT reserves whole VMA at once (x86) */
#define VM_MERGEABLE 0x80000000 /* KSM may merge identical pages */
/* Bits set in the VMA until the stack is in its final location */
@@ -158,19 +158,6 @@ extern pgprot_t protection_map[16];
#define FAULT_FLAG_RETRY_NOWAIT 0x10 /* Don't drop mmap_sem and wait when retrying */
#define FAULT_FLAG_KILLABLE 0x20 /* The fault task is in SIGKILL killable region */
-/*
- * This interface is used by x86 PAT code to identify a pfn mapping that is
- * linear over entire vma. This is to optimize PAT code that deals with
- * marking the physical region with a particular prot. This is not for generic
- * mm use. Note also that this check will not work if the pfn mapping is
- * linear for a vma starting at physical address 0. In which case PAT code
- * falls back to slow path of reserving physical range page by page.
- */
-static inline int is_linear_pfn_mapping(struct vm_area_struct *vma)
-{
- return !!(vma->vm_flags & VM_PFN_AT_MMAP);
-}
-
static inline int is_pfn_mapping(struct vm_area_struct *vma)
{
return !!(vma->vm_flags & VM_PFNMAP);
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index f0e5306..cf827da 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1650,7 +1650,7 @@ int khugepaged_enter_vma_merge(struct vm_area_struct *vma)
* If is_pfn_mapping() is true is_learn_pfn_mapping() must be
* true too, verify it here.
*/
- VM_BUG_ON(is_linear_pfn_mapping(vma) || vma->vm_flags & VM_NO_THP);
+ VM_BUG_ON(vma->vm_flags & VM_NO_THP);
hstart = (vma->vm_start + ~HPAGE_PMD_MASK) & HPAGE_PMD_MASK;
hend = vma->vm_end & HPAGE_PMD_MASK;
if (hstart < hend)
@@ -1908,7 +1908,7 @@ static void collapse_huge_page(struct mm_struct *mm,
* If is_pfn_mapping() is true is_learn_pfn_mapping() must be
* true too, verify it here.
*/
- VM_BUG_ON(is_linear_pfn_mapping(vma) || vma->vm_flags & VM_NO_THP);
+ VM_BUG_ON(vma->vm_flags & VM_NO_THP);
pgd = pgd_offset(mm, address);
if (!pgd_present(*pgd))
@@ -2150,8 +2150,7 @@ static unsigned int khugepaged_scan_mm_slot(unsigned int pages,
* If is_pfn_mapping() is true is_learn_pfn_mapping()
* must be true too, verify it here.
*/
- VM_BUG_ON(is_linear_pfn_mapping(vma) ||
- vma->vm_flags & VM_NO_THP);
+ VM_BUG_ON(vma->vm_flags & VM_NO_THP);
hstart = (vma->vm_start + ~HPAGE_PMD_MASK) & HPAGE_PMD_MASK;
hend = vma->vm_end & HPAGE_PMD_MASK;
diff --git a/mm/memory.c b/mm/memory.c
index 4cdcf53..2ade15b 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -2282,23 +2282,23 @@ int remap_pfn_range(struct vm_area_struct *vma, unsigned long addr,
* There's a horrible special case to handle copy-on-write
* behaviour that some programs depend on. We mark the "original"
* un-COW'ed pages by matching them up with "vma->vm_pgoff".
+ * See vm_normal_page() for details.
*/
- if (addr == vma->vm_start && end == vma->vm_end) {
+ if (is_cow_mapping(vma->vm_flags)) {
+ if (addr != vma->vm_start || end != vma->vm_end)
+ return -EINVAL;
vma->vm_pgoff = pfn;
- vma->vm_flags |= VM_PFN_AT_MMAP;
- } else if (is_cow_mapping(vma->vm_flags))
- return -EINVAL;
+ }
vma->vm_flags |= VM_IO | VM_RESERVED | VM_PFNMAP;
- err = track_pfn_remap(vma, &prot, pfn, PAGE_ALIGN(size));
+ err = track_pfn_remap(vma, &prot, pfn, addr, PAGE_ALIGN(size));
if (err) {
/*
* To indicate that track_pfn related cleanup is not
* needed from higher level routine calling unmap_vmas
*/
vma->vm_flags &= ~(VM_IO | VM_RESERVED | VM_PFNMAP);
- vma->vm_flags &= ~VM_PFN_AT_MMAP;
return -EINVAL;
}
--
1.7.6.5
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists