From: Yinghai Lu Date: Sun, 23 Mar 2008 07:16:49 +0000 (-0700) Subject: x86: fix trim mtrr not to setup_memory two times X-Git-Tag: v3.12-rc1~22255^2~1 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=76c324182bbd29dfe4298ca65efb15be18055df1;p=kernel%2Fkernel-generic.git x86: fix trim mtrr not to setup_memory two times we could call find_max_pfn() directly instead of setup_memory() to get max_pfn needed for mtrr trimming. otherwise setup_memory() is called two times... that is duplicated... [ mingo@elte.hu: both Thomas and me simulated a double call to setup_bootmem_allocator() and can confirm that it is a real bug which can hang in certain configs. It's not been reported yet but that is probably due to the relatively scarce nature of MTRR-trimming systems. ] Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c index a1d7071..2b3e5d4 100644 --- a/arch/x86/kernel/setup_32.c +++ b/arch/x86/kernel/setup_32.c @@ -406,8 +406,6 @@ static unsigned long __init setup_memory(void) */ min_low_pfn = PFN_UP(init_pg_tables_end); - find_max_pfn(); - max_low_pfn = find_max_low_pfn(); #ifdef CONFIG_HIGHMEM @@ -764,12 +762,13 @@ void __init setup_arch(char **cmdline_p) if (efi_enabled) efi_init(); - max_low_pfn = setup_memory(); - /* update e820 for memory not covered by WB MTRRs */ + find_max_pfn(); mtrr_bp_init(); if (mtrr_trim_uncached_memory(max_pfn)) - max_low_pfn = setup_memory(); + find_max_pfn(); + + max_low_pfn = setup_memory(); #ifdef CONFIG_VMI /* diff --git a/arch/x86/mm/discontig_32.c b/arch/x86/mm/discontig_32.c index c394ca0..8e25e06 100644 --- a/arch/x86/mm/discontig_32.c +++ b/arch/x86/mm/discontig_32.c @@ -324,7 +324,6 @@ unsigned long __init setup_memory(void) * this space and use it to adjust the boundary between ZONE_NORMAL * and ZONE_HIGHMEM. */ - find_max_pfn(); get_memcfg_numa(); kva_pages = calculate_numa_remap_pages();