On Friday 09 October 2015 03:38 PM, Kirill A. Shutemov wrote:
> On Tue, Sep 22, 2015 at 04:04:54PM +0530, Vineet Gupta wrote:
>
> Commit message: -ENOENT.
>
> Otherwise, looks good:
>
> Acked-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
With updated change log and some reworking in the source code comment !
---------------->
>From 96537a576f99be29f65c5682d6e0e7b31028d5ba Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@xxxxxxxxxxxx>
Date: Fri, 20 Feb 2015 10:36:28 +0530
Subject: [PATCH v3] mm,thp: introduce flush_pmd_tlb_range
ARCHes with special requirements for evicting THP backing TLB entries
can implement this.
Otherwise also, it can help optimize TLB flush in THP regime.
stock flush_tlb_range() typically has optimization to nuke the entire
TLB if flush span is greater than a certain threshhold, which will
likely be true for a single huge page. Thus a single thp flush will
invalidate the entrire TLB which is not desirable.
e.g. see arch/arc: flush_pmd_tlb_range
Signed-off-by: Vineet Gupta <vgupta@xxxxxxxxxxxx>
---
mm/huge_memory.c | 2 +-
mm/pgtable-generic.c | 26 ++++++++++++++++++++------
2 files changed, 21 insertions(+), 7 deletions(-)
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 4b06b8db9df2..e25eb3d2081a 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1880,7 +1880,7 @@ static int __split_huge_page_map(struct page *page,
* here). But it is generally safer to never allow
* small and huge TLB entries for the same virtual
* address to be loaded simultaneously. So instead of
- * doing "pmd_populate(); flush_tlb_range();" we first
+ * doing "pmd_populate(); flush_pmd_tlb_range();" we first
* mark the current pmd notpresent (atomically because
* here the pmd_trans_huge and pmd_trans_splitting
* must remain set at all times on the pmd until the
diff --git a/mm/pgtable-generic.c b/mm/pgtable-generic.c
index c9c59bb75a17..7d3db0247983 100644
--- a/mm/pgtable-generic.c
+++ b/mm/pgtable-generic.c
@@ -84,6 +84,20 @@ pte_t ptep_clear_flush(struct vm_area_struct *vma, unsigned
long address,
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+#ifndef __HAVE_ARCH_FLUSH_PMD_TLB_RANGE
+
+/*
+ * ARCHes with special requirements for evicting THP backing TLB entries can
+ * implement this. Otherwise also, it can help optimize normal TLB flush in
+ * THP regime. stock flush_tlb_range() typically has optimization to nuke the
+ * entire TLB TLB if flush span is greater than a threshhold, which will
+ * likely be true for a single huge page. Thus a single thp flush will
+ * invalidate the entire TLB which is not desitable.
+ * e.g. see arch/arc: flush_pmd_tlb_range
+ */
+#define flush_pmd_tlb_range(vma, addr, end) flush_tlb_range(vma, addr, end)
+#endif
+
#ifndef __HAVE_ARCH_PMDP_SET_ACCESS_FLAGS
int pmdp_set_access_flags(struct vm_area_struct *vma,
unsigned long address, pmd_t *pmdp,
@@ -93,7 +107,7 @@ int pmdp_set_access_flags(struct vm_area_struct *vma,
VM_BUG_ON(address & ~HPAGE_PMD_MASK);
if (changed) {
set_pmd_at(vma->vm_mm, address, pmdp, entry);
- flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
+ flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
}
return changed;
}
@@ -107,7 +121,7 @@ int pmdp_clear_flush_young(struct vm_area_struct *vma,
VM_BUG_ON(address & ~HPAGE_PMD_MASK);
young = pmdp_test_and_clear_young(vma, address, pmdp);
if (young)
- flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
+ flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
return young;
}
#endif
@@ -120,7 +134,7 @@ pmd_t pmdp_huge_clear_flush(struct vm_area_struct *vma,
unsigned long address,
VM_BUG_ON(address & ~HPAGE_PMD_MASK);
VM_BUG_ON(!pmd_trans_huge(*pmdp));
pmd = pmdp_huge_get_and_clear(vma->vm_mm, address, pmdp);
- flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
+ flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
return pmd;
}
#endif
@@ -133,7 +147,7 @@ void pmdp_splitting_flush(struct vm_area_struct *vma, unsigned
long address,
VM_BUG_ON(address & ~HPAGE_PMD_MASK);
set_pmd_at(vma->vm_mm, address, pmdp, pmd);
/* tlb flush only to serialize against gup-fast */
- flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
+ flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
}
#endif
@@ -179,7 +193,7 @@ void pmdp_invalidate(struct vm_area_struct *vma, unsigned long
address,
{
pmd_t entry = *pmdp;
set_pmd_at(vma->vm_mm, address, pmdp, pmd_mknotpresent(entry));
- flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
+ flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
}
#endif
@@ -196,7 +210,7 @@ pmd_t pmdp_collapse_flush(struct vm_area_struct *vma, unsigned
long address,
VM_BUG_ON(address & ~HPAGE_PMD_MASK);
VM_BUG_ON(pmd_trans_huge(*pmdp));
pmd = pmdp_huge_get_and_clear(vma->vm_mm, address, pmdp);
- flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
+ flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE);
return pmd;
}
#endif
--
1.9.1
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>