From 2b4a08150e0ce2f6eb5d0987fdfe3524ec799313 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Mon, 12 Sep 2005 18:49:24 +0200 Subject: [PATCH] x86-64: Increase TLB flush array size The generic TLB flush functions kept upto 506 pages per CPU to avoid too frequent IPIs. This value was done for the L1 cache of older x86 CPUs, but with modern CPUs it does not make much sense anymore. TLB flushing is slow enough that using the L2 cache is fine. This patch increases the flush array on x86-64 to cache 5350 pages. That is roughly 20MB with 4K pages. It speeds up large munmaps in multithreaded processes on SMP considerably. The cost is roughly 42k of memory per CPU, which is reasonable. I only increased it on x86-64 for now, but it would probably make sense to increase it everywhere. Embedded architectures with SMP may keep it smaller to save some memory per CPU. Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds --- include/asm-x86_64/tlbflush.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'include/asm-x86_64') diff --git a/include/asm-x86_64/tlbflush.h b/include/asm-x86_64/tlbflush.h index 505b0cf906d..4a9c20ea9b1 100644 --- a/include/asm-x86_64/tlbflush.h +++ b/include/asm-x86_64/tlbflush.h @@ -109,6 +109,10 @@ static inline void flush_tlb_range(struct vm_area_struct * vma, unsigned long st #define TLBSTATE_OK 1 #define TLBSTATE_LAZY 2 +/* Roughly an IPI every 20MB with 4k pages for freeing page table + ranges. Cost is about 42k of memory for each CPU. */ +#define ARCH_FREE_PTE_NR 5350 + #endif #define flush_tlb_kernel_range(start, end) flush_tlb_all() -- cgit v1.2.3