lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

[ARM] 5450/1: Flush only the needed range when unmapping a VMA

Browse Source 
When unmapping N pages (e.g. shared memory) the amount of TLB flushes
done can be (N*PAGE_SIZE/ZAP_BLOCK_SIZE)*N although it should be N at
maximum. With PREEMPT kernel ZAP_BLOCK_SIZE is 8 pages, so there is a
noticeable performance penalty when unmapping a large VMA and the system
is spending its time in flush_tlb_range().

The problem is that tlb_end_vma() is always flushing the full VMA
range. The subrange that needs to be flushed can be calculated by
tlb_remove_tlb_entry(). This approach was suggested by Hugh Dickins,
and is also used by other arches.

The speed increase is roughly 3x for 8M mappings and for larger mappings
even more.

Signed-off-by: Aaro Koskinen <Aaro.Koskinen@nokia.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This commit is contained in:
Aaro Koskinen
2009-04-14 13:07:35 +01:00
committed by Russell King
parent 41609ff430
commit 7fccfc00c0
1 changed files with 21 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
arch/arm/include/asm/tlb.h
View File

@@ -36,6 +36,8 @@
struct mmu_gather { struct mmu_gather {
struct mm_struct *mm; struct mm_struct *mm;
unsigned int fullmm; unsigned int fullmm;
unsigned long range_start;
unsigned long range_end;
}; };
DECLARE_PER_CPU(struct mmu_gather, mmu_gathers); DECLARE_PER_CPU(struct mmu_gather, mmu_gathers);
@@ -63,7 +65,19 @@ tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end)
put_cpu_var(mmu_gathers); put_cpu_var(mmu_gathers);
} }
#define tlb_remove_tlb_entry(tlb,ptep,address) do { } while (0) /*
* Memorize the range for the TLB flush.
*/
static inline void
tlb_remove_tlb_entry(struct mmu_gather *tlb, pte_t *ptep, unsigned long addr)
{
if (!tlb->fullmm) {
if (addr < tlb->range_start)
tlb->range_start = addr;
if (addr + PAGE_SIZE > tlb->range_end)
tlb->range_end = addr + PAGE_SIZE;
}
}
/* /*
* In the case of tlb vma handling, we can optimise these away in the * In the case of tlb vma handling, we can optimise these away in the
@@ -73,15 +87,18 @@ tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end)
static inline void static inline void
tlb_start_vma(struct mmu_gather *tlb, struct vm_area_struct *vma) tlb_start_vma(struct mmu_gather *tlb, struct vm_area_struct *vma)
{ {
if (!tlb->fullmm) if (!tlb->fullmm) {
flush_cache_range(vma, vma->vm_start, vma->vm_end); flush_cache_range(vma, vma->vm_start, vma->vm_end);
tlb->range_start = TASK_SIZE;
tlb->range_end = 0;
}
} }
static inline void static inline void
tlb_end_vma(struct mmu_gather *tlb, struct vm_area_struct *vma) tlb_end_vma(struct mmu_gather *tlb, struct vm_area_struct *vma)
{ {
if (!tlb->fullmm) if (!tlb->fullmm && tlb->range_end > 0)
flush_tlb_range(vma, vma->vm_start, vma->vm_end); flush_tlb_range(vma, tlb->range_start, tlb->range_end);
} }
#define tlb_remove_page(tlb,page) free_page_and_swap_cache(page) #define tlb_remove_page(tlb,page) free_page_and_swap_cache(page)