From 188f87f2648b13f5de17d5e068f18d317e0c1f98 Mon Sep 17 00:00:00 2001 From: Eric Chanudet Date: Wed, 22 May 2024 16:38:01 -0400 Subject: mm/mm_init: use node's number of cpus in deferred_page_init_max_threads x86_64 is already using the node's cpu as maximum threads. Make that the default for all archs setting DEFERRED_STRUCT_PAGE_INIT. This returns to the behavior prior making the function arch-specific with commit ecd096506922 ("mm: make deferred init's max threads arch-specific"). Setting DEFERRED_STRUCT_PAGE_INIT and testing on a few arm64 platforms shows faster deferred_init_memmap completions: | | x13s | SA8775p-ride | Ampere R137-P31 | Ampere HR330 | | | Metal, 32GB | VM, 36GB | VM, 58GB | Metal, 128GB | | | 8cpus | 8cpus | 8cpus | 32cpus | |---------|-------------|--------------|-----------------|--------------| | threads | ms (%) | ms (%) | ms (%) | ms (%) | |---------|-------------|--------------|-----------------|--------------| | 1 | 108 (0%) | 72 (0%) | 224 (0%) | 324 (0%) | | cpus | 24 (-77%) | 36 (-50%) | 40 (-82%) | 56 (-82%) | Michael Ellerman reported: : On a machine here (1TB, 40 cores, 4KB pages) the existing code gives: : : [ 0.500124] node 2 deferred pages initialised in 210ms : [ 0.515790] node 3 deferred pages initialised in 230ms : [ 0.516061] node 0 deferred pages initialised in 230ms : [ 0.516522] node 7 deferred pages initialised in 230ms : [ 0.516672] node 4 deferred pages initialised in 230ms : [ 0.516798] node 6 deferred pages initialised in 230ms : [ 0.517051] node 5 deferred pages initialised in 230ms : [ 0.523887] node 1 deferred pages initialised in 240ms : : vs with the patch: : : [ 0.379613] node 0 deferred pages initialised in 90ms : [ 0.380388] node 1 deferred pages initialised in 90ms : [ 0.380540] node 4 deferred pages initialised in 100ms : [ 0.390239] node 6 deferred pages initialised in 100ms : [ 0.390249] node 2 deferred pages initialised in 100ms : [ 0.390786] node 3 deferred pages initialised in 110ms : [ 0.396721] node 5 deferred pages initialised in 110ms : [ 0.397095] node 7 deferred pages initialised in 110ms : : Which is a nice speedup. [echanude@redhat.com: v3] Link: https://lkml.kernel.org/r/20240528185455.643227-4-echanude@redhat.com Link: https://lkml.kernel.org/r/20240522203758.626932-4-echanude@redhat.com Signed-off-by: Eric Chanudet Tested-by: Michael Ellerman (powerpc) Reviewed-by: Baoquan He Acked-by: Alexander Gordeev Acked-by: Mike Rapoport (IBM) Cc: Andy Lutomirski Cc: Borislav Petkov (AMD) Cc: Dave Hansen Cc: "H. Peter Anvin" Cc: Ingo Molnar Cc: Nicholas Piggin Cc: Peter Zijlstra Cc: Thomas Gleixner Signed-off-by: Andrew Morton --- arch/x86/mm/init_64.c | 12 ------------ 1 file changed, 12 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index 7e177856ee4f..adec42928ec1 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -1354,18 +1354,6 @@ void __init mem_init(void) preallocate_vmalloc_pages(); } -#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT -int __init deferred_page_init_max_threads(const struct cpumask *node_cpumask) -{ - /* - * More CPUs always led to greater speedups on tested systems, up to - * all the nodes' CPUs. Use all since the system is otherwise idle - * now. - */ - return max_t(int, cpumask_weight(node_cpumask), 1); -} -#endif - int kernel_set_to_readonly; void mark_rodata_ro(void) -- cgit v1.2.3 From d39b6af2189095f2c411a0a261d44e0ac4be3ad1 Mon Sep 17 00:00:00 2001 From: Peter Xu Date: Mon, 27 May 2024 11:48:55 -0400 Subject: mm: drop leftover comment references to pxx_huge() pxx_huge() has been removed in recent commit 9636f055dae1 ("mm/treewide: remove pXd_huge()"), however there are still three comments referencing the API that got overlooked. Remove them. Link: https://lkml.kernel.org/r/20240527154855.528816-1-peterx@redhat.com Signed-off-by: Peter Xu Reported-by: Christophe Leroy Reviewed-by: David Hildenbrand Cc: Jason Gunthorpe Signed-off-by: Andrew Morton --- arch/x86/mm/pat/set_memory.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/pat/set_memory.c b/arch/x86/mm/pat/set_memory.c index 19fdfbb171ed..8b2164509b4d 100644 --- a/arch/x86/mm/pat/set_memory.c +++ b/arch/x86/mm/pat/set_memory.c @@ -1120,8 +1120,8 @@ __split_large_page(struct cpa_data *cpa, pte_t *kpte, unsigned long address, lpinc = PMD_SIZE; /* * Clear the PSE flags if the PRESENT flag is not set - * otherwise pmd_present/pmd_huge will return true - * even on a non present pmd. + * otherwise pmd_present() will return true even on a non + * present pmd. */ if (!(pgprot_val(ref_prot) & _PAGE_PRESENT)) pgprot_val(ref_prot) &= ~_PAGE_PSE; -- cgit v1.2.3 From 7bc131c56853cd6c1d51d8cfc1befc0a23278395 Mon Sep 17 00:00:00 2001 From: Oscar Salvador Date: Mon, 27 May 2024 06:45:23 +0200 Subject: arch/x86: do not explicitly clear Reserved flag in free_pagetable In free_pagetable() we use the non-atomic version for clearing the PageReserved bit from the page. free_pagetable() will either call free_reserved_page() or put_page_bootmem(), which will eventually end up calling free_reserved_page(), and in there we already clear the PageReserved flag. Link: https://lkml.kernel.org/r/20240527044523.29207-1-osalvador@suse.de Signed-off-by: Oscar Salvador Acked-by: David Hildenbrand Cc: Dave Hansen Signed-off-by: Andrew Morton --- arch/x86/mm/init_64.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index adec42928ec1..8f6b14003ff2 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -980,8 +980,6 @@ static void __meminit free_pagetable(struct page *page, int order) /* bootmem page has reserved flag */ if (PageReserved(page)) { - __ClearPageReserved(page); - magic = page->index; if (magic == SECTION_INFO || magic == MIX_SECTION_INFO) { while (nr_pages--) -- cgit v1.2.3