From e16faf26780fc0c8dd693ea9ee8420a7706cb2f5 Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Tue, 22 Mar 2022 14:43:17 -0700 Subject: cma: factor out minimum alignment requirement Patch series "mm: enforce pageblock_order < MAX_ORDER". Having pageblock_order >= MAX_ORDER seems to be able to happen in corner cases and some parts of the kernel are not prepared for it. For example, Aneesh has shown [1] that such kernels can be compiled on ppc64 with 64k base pages by setting FORCE_MAX_ZONEORDER=8, which will run into a WARN_ON_ONCE(order >= MAX_ORDER) in comapction code right during boot. We can get pageblock_order >= MAX_ORDER when the default hugetlb size is bigger than the maximum allocation granularity of the buddy, in which case we are no longer talking about huge pages but instead gigantic pages. Having pageblock_order >= MAX_ORDER can only make alloc_contig_range() of such gigantic pages more likely to succeed. Reliable use of gigantic pages either requires boot time allcoation or CMA, no need to overcomplicate some places in the kernel to optimize for corner cases that are broken in other areas of the kernel. This patch (of 2): Let's enforce pageblock_order < MAX_ORDER and simplify. Especially patch #1 can be regarded a cleanup before: [PATCH v5 0/6] Use pageblock_order for cma and alloc_contig_range alignment. [2] [1] https://lkml.kernel.org/r/87r189a2ks.fsf@linux.ibm.com [2] https://lkml.kernel.org/r/20220211164135.1803616-1-zi.yan@sent.com Link: https://lkml.kernel.org/r/20220214174132.219303-2-david@redhat.com Signed-off-by: David Hildenbrand Reviewed-by: Zi Yan Acked-by: Rob Herring Cc: Aneesh Kumar K.V Cc: Michael Ellerman Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Frank Rowand Cc: Michael S. Tsirkin Cc: Christoph Hellwig Cc: Marek Szyprowski Cc: Robin Murphy Cc: Minchan Kim Cc: Vlastimil Babka Cc: John Garry via iommu Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/cma.h | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'include/linux/cma.h') diff --git a/include/linux/cma.h b/include/linux/cma.h index bd801023504b..75fe188ec4a1 100644 --- a/include/linux/cma.h +++ b/include/linux/cma.h @@ -20,6 +20,15 @@ #define CMA_MAX_NAME 64 +/* + * TODO: once the buddy -- especially pageblock merging and alloc_contig_range() + * -- can deal with only some pageblocks of a higher-order page being + * MIGRATE_CMA, we can use pageblock_nr_pages. + */ +#define CMA_MIN_ALIGNMENT_PAGES max_t(phys_addr_t, MAX_ORDER_NR_PAGES, \ + pageblock_nr_pages) +#define CMA_MIN_ALIGNMENT_BYTES (PAGE_SIZE * CMA_MIN_ALIGNMENT_PAGES) + struct cma; extern unsigned long totalcma_pages; -- cgit v1.2.3 From b3d40a2b6d10c9d0424d2b398bf962fb6adad87e Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Tue, 22 Mar 2022 14:43:20 -0700 Subject: mm: enforce pageblock_order < MAX_ORDER Some places in the kernel don't really expect pageblock_order >= MAX_ORDER, and it looks like this is only possible in corner cases: 1) CONFIG_DEFERRED_STRUCT_PAGE_INIT we'll end up freeing pageblock_order pages via __free_pages_core(), which cannot possibly work. 2) find_zone_movable_pfns_for_nodes() will roundup the ZONE_MOVABLE start PFN to MAX_ORDER_NR_PAGES. Consequently with a bigger pageblock_order, we could have a single pageblock partially managed by two zones. 3) compaction code runs into __fragmentation_index() with order >= MAX_ORDER, when checking WARN_ON_ONCE(order >= MAX_ORDER). [1] 4) mm/page_reporting.c won't be reporting any pages with default page_reporting_order == pageblock_order, as we'll be skipping the reporting loop inside page_reporting_process_zone(). 5) __rmqueue_fallback() will never be able to steal with ALLOC_NOFRAGMENT. pageblock_order >= MAX_ORDER is weird either way: it's a pure optimization for making alloc_contig_range(), as used for allcoation of gigantic pages, a little more reliable to succeed. However, if there is demand for somewhat reliable allocation of gigantic pages, affected setups should be using CMA or boottime allocations instead. So let's make sure that pageblock_order < MAX_ORDER and simplify. [1] https://lkml.kernel.org/r/87r189a2ks.fsf@linux.ibm.com Link: https://lkml.kernel.org/r/20220214174132.219303-3-david@redhat.com Signed-off-by: David Hildenbrand Reviewed-by: Zi Yan Cc: Aneesh Kumar K.V Cc: Benjamin Herrenschmidt Cc: Christoph Hellwig Cc: Frank Rowand Cc: John Garry via iommu Cc: Marek Szyprowski Cc: Michael Ellerman Cc: Michael S. Tsirkin Cc: Minchan Kim Cc: Paul Mackerras Cc: Rob Herring Cc: Robin Murphy Cc: Vlastimil Babka Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/virtio/virtio_mem.c | 9 +++------ include/linux/cma.h | 3 +-- include/linux/pageblock-flags.h | 7 +++++-- mm/Kconfig | 3 +++ mm/page_alloc.c | 32 ++++++++------------------------ 5 files changed, 20 insertions(+), 34 deletions(-) (limited to 'include/linux/cma.h') diff --git a/drivers/virtio/virtio_mem.c b/drivers/virtio/virtio_mem.c index 38becd8d578c..e7d6b679596d 100644 --- a/drivers/virtio/virtio_mem.c +++ b/drivers/virtio/virtio_mem.c @@ -2476,13 +2476,10 @@ static int virtio_mem_init_hotplug(struct virtio_mem *vm) VIRTIO_MEM_DEFAULT_OFFLINE_THRESHOLD); /* - * We want subblocks to span at least MAX_ORDER_NR_PAGES and - * pageblock_nr_pages pages. This: - * - Is required for now for alloc_contig_range() to work reliably - - * it doesn't properly handle smaller granularity on ZONE_NORMAL. + * TODO: once alloc_contig_range() works reliably with pageblock + * granularity on ZONE_NORMAL, use pageblock_nr_pages instead. */ - sb_size = max_t(uint64_t, MAX_ORDER_NR_PAGES, - pageblock_nr_pages) * PAGE_SIZE; + sb_size = PAGE_SIZE * MAX_ORDER_NR_PAGES; sb_size = max_t(uint64_t, vm->device_block_size, sb_size); if (sb_size < memory_block_size_bytes() && !force_bbm) { diff --git a/include/linux/cma.h b/include/linux/cma.h index 75fe188ec4a1..b1ba94f1cc9c 100644 --- a/include/linux/cma.h +++ b/include/linux/cma.h @@ -25,8 +25,7 @@ * -- can deal with only some pageblocks of a higher-order page being * MIGRATE_CMA, we can use pageblock_nr_pages. */ -#define CMA_MIN_ALIGNMENT_PAGES max_t(phys_addr_t, MAX_ORDER_NR_PAGES, \ - pageblock_nr_pages) +#define CMA_MIN_ALIGNMENT_PAGES MAX_ORDER_NR_PAGES #define CMA_MIN_ALIGNMENT_BYTES (PAGE_SIZE * CMA_MIN_ALIGNMENT_PAGES) struct cma; diff --git a/include/linux/pageblock-flags.h b/include/linux/pageblock-flags.h index 973fd731a520..83c7248053a1 100644 --- a/include/linux/pageblock-flags.h +++ b/include/linux/pageblock-flags.h @@ -37,8 +37,11 @@ extern unsigned int pageblock_order; #else /* CONFIG_HUGETLB_PAGE_SIZE_VARIABLE */ -/* Huge pages are a constant size */ -#define pageblock_order HUGETLB_PAGE_ORDER +/* + * Huge pages are a constant size, but don't exceed the maximum allocation + * granularity. + */ +#define pageblock_order min_t(unsigned int, HUGETLB_PAGE_ORDER, MAX_ORDER - 1) #endif /* CONFIG_HUGETLB_PAGE_SIZE_VARIABLE */ diff --git a/mm/Kconfig b/mm/Kconfig index 3326ee3903f3..4c91b92e7537 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -262,6 +262,9 @@ config HUGETLB_PAGE_SIZE_VARIABLE HUGETLB_PAGE_ORDER when there are multiple HugeTLB page sizes available on a platform. + Note that the pageblock_order cannot exceed MAX_ORDER - 1 and will be + clamped down to MAX_ORDER - 1. + config CONTIG_ALLOC def_bool (MEMORY_ISOLATION && COMPACTION) || CMA diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 06c683754334..a4efbdad2f2b 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1072,14 +1072,12 @@ static inline void __free_one_page(struct page *page, int migratetype, fpi_t fpi_flags) { struct capture_control *capc = task_capc(zone); + unsigned int max_order = pageblock_order; unsigned long buddy_pfn; unsigned long combined_pfn; - unsigned int max_order; struct page *buddy; bool to_tail; - max_order = min_t(unsigned int, MAX_ORDER - 1, pageblock_order); - VM_BUG_ON(!zone_is_initialized(zone)); VM_BUG_ON_PAGE(page->flags & PAGE_FLAGS_CHECK_AT_PREP, page); @@ -2259,19 +2257,8 @@ void __init init_cma_reserved_pageblock(struct page *page) } while (++p, --i); set_pageblock_migratetype(page, MIGRATE_CMA); - - if (pageblock_order >= MAX_ORDER) { - i = pageblock_nr_pages; - p = page; - do { - set_page_refcounted(p); - __free_pages(p, MAX_ORDER - 1); - p += MAX_ORDER_NR_PAGES; - } while (i -= MAX_ORDER_NR_PAGES); - } else { - set_page_refcounted(page); - __free_pages(page, pageblock_order); - } + set_page_refcounted(page); + __free_pages(page, pageblock_order); adjust_managed_page_count(page, pageblock_nr_pages); page_zone(page)->cma_pages += pageblock_nr_pages; @@ -7382,16 +7369,15 @@ static inline void setup_usemap(struct zone *zone) {} /* Initialise the number of pages represented by NR_PAGEBLOCK_BITS */ void __init set_pageblock_order(void) { - unsigned int order; + unsigned int order = MAX_ORDER - 1; /* Check that pageblock_nr_pages has not already been setup */ if (pageblock_order) return; - if (HPAGE_SHIFT > PAGE_SHIFT) + /* Don't let pageblocks exceed the maximum allocation granularity. */ + if (HPAGE_SHIFT > PAGE_SHIFT && HUGETLB_PAGE_ORDER < order) order = HUGETLB_PAGE_ORDER; - else - order = MAX_ORDER - 1; /* * Assume the largest contiguous order of interest is a huge page. @@ -8979,14 +8965,12 @@ struct page *has_unmovable_pages(struct zone *zone, struct page *page, #ifdef CONFIG_CONTIG_ALLOC static unsigned long pfn_max_align_down(unsigned long pfn) { - return pfn & ~(max_t(unsigned long, MAX_ORDER_NR_PAGES, - pageblock_nr_pages) - 1); + return ALIGN_DOWN(pfn, MAX_ORDER_NR_PAGES); } static unsigned long pfn_max_align_up(unsigned long pfn) { - return ALIGN(pfn, max_t(unsigned long, MAX_ORDER_NR_PAGES, - pageblock_nr_pages)); + return ALIGN(pfn, MAX_ORDER_NR_PAGES); } #if defined(CONFIG_DYNAMIC_DEBUG) || \ -- cgit v1.2.3 From 27d121d0ec6d604d0147c5b579e4181b688a2d64 Mon Sep 17 00:00:00 2001 From: Hari Bathini Date: Tue, 22 Mar 2022 14:46:14 -0700 Subject: mm/cma: provide option to opt out from exposing pages on activation failure Patch series "powerpc/fadump: handle CMA activation failure appropriately", v3. Commit 072355c1cf2d ("mm/cma: expose all pages to the buddy if activation of an area fails") started exposing all pages to buddy allocator on CMA activation failure. But there can be CMA users that want to handle the reserved memory differently on CMA allocation failure. Provide an option to opt out from exposing pages to buddy for such cases. Link: https://lkml.kernel.org/r/20220117075246.36072-1-hbathini@linux.ibm.com Link: https://lkml.kernel.org/r/20220117075246.36072-2-hbathini@linux.ibm.com Signed-off-by: Hari Bathini Reviewed-by: David Hildenbrand Cc: Oscar Salvador Cc: Mike Kravetz Cc: Mahesh Salgaonkar Cc: Sourabh Jain Cc: Michael Ellerman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/cma.h | 2 ++ mm/cma.c | 11 +++++++++-- mm/cma.h | 1 + 3 files changed, 12 insertions(+), 2 deletions(-) (limited to 'include/linux/cma.h') diff --git a/include/linux/cma.h b/include/linux/cma.h index b1ba94f1cc9c..90fd742fd1ef 100644 --- a/include/linux/cma.h +++ b/include/linux/cma.h @@ -58,4 +58,6 @@ extern bool cma_pages_valid(struct cma *cma, const struct page *pages, unsigned extern bool cma_release(struct cma *cma, const struct page *pages, unsigned long count); extern int cma_for_each_area(int (*it)(struct cma *cma, void *data), void *data); + +extern void cma_reserve_pages_on_error(struct cma *cma); #endif diff --git a/mm/cma.c b/mm/cma.c index 5a2cd5851658..eaa4b5c920a2 100644 --- a/mm/cma.c +++ b/mm/cma.c @@ -131,8 +131,10 @@ not_in_zone: bitmap_free(cma->bitmap); out_error: /* Expose all pages to the buddy, they are useless for CMA. */ - for (pfn = base_pfn; pfn < base_pfn + cma->count; pfn++) - free_reserved_page(pfn_to_page(pfn)); + if (!cma->reserve_pages_on_error) { + for (pfn = base_pfn; pfn < base_pfn + cma->count; pfn++) + free_reserved_page(pfn_to_page(pfn)); + } totalcma_pages -= cma->count; cma->count = 0; pr_err("CMA area %s could not be activated\n", cma->name); @@ -150,6 +152,11 @@ static int __init cma_init_reserved_areas(void) } core_initcall(cma_init_reserved_areas); +void __init cma_reserve_pages_on_error(struct cma *cma) +{ + cma->reserve_pages_on_error = true; +} + /** * cma_init_reserved_mem() - create custom contiguous area from reserved memory * @base: Base address of the reserved area diff --git a/mm/cma.h b/mm/cma.h index 2c775877eae2..88a0595670b7 100644 --- a/mm/cma.h +++ b/mm/cma.h @@ -30,6 +30,7 @@ struct cma { /* kobject requires dynamic object */ struct cma_kobject *cma_kobj; #endif + bool reserve_pages_on_error; }; extern struct cma cma_areas[MAX_CMA_AREAS]; -- cgit v1.2.3 From 229a08a4f4e4f9949801cc39b6480ddc9c487183 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 9 Mar 2022 09:37:31 -0800 Subject: ARM/dma-mapping: Remove CMA code when not built with CMA The MAX_CMA_AREAS could be set to 0, which would result in code that would attempt to operate beyond the end of a zero-sized array. If CONFIG_CMA is disabled, just remove this code entirely. Found when building arm on GCC 10.x for several defconfigs (e.g. axm55xx_defconfig) under -Warray-bounds: arch/arm/mm/dma-mapping.c:396:22: warning: array subscript is outside array bounds of 'struct dma_contig_early_reserve[0]' [-Warray-bounds] 396 | dma_mmu_remap[dma_mmu_remap_num].size = size; | ~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~ arch/arm/mm/dma-mapping.c:389:40: note: while referencing 'dma_mmu_remap' 389 | static struct dma_contig_early_reserve dma_mmu_remap[MAX_CMA_AREAS] __initdata; | ^~~~~~~~~~~~~ Cc: Russell King Cc: Logan Gunthorpe Cc: Martin Oliveira Cc: David Hildenbrand Cc: Andrew Morton Cc: Stephen Rothwell Cc: Zi Yan Cc: Hari Bathini Cc: Minchan Kim Cc: Mike Kravetz Cc: linux-arm-kernel@lists.infradead.org Link: https://lore.kernel.org/all/6243ee60.1c69fb81.16de6.7dbf@mx.google.com/ Signed-off-by: Kees Cook Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/lkml/20220310070041.GA24874@lst.de Reviewed-by: David Hildenbrand Link: https://lore.kernel.org/lkml/9059fa71-330f-f04f-b155-2850abb72a71@redhat.com --- arch/arm/mm/dma-mapping.c | 2 ++ arch/arm/mm/mm.h | 4 ++++ include/linux/cma.h | 4 ---- 3 files changed, 6 insertions(+), 4 deletions(-) (limited to 'include/linux/cma.h') diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 4b61541853ea..82ffac621854 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -381,6 +381,7 @@ out: */ postcore_initcall(atomic_pool_init); +#ifdef CONFIG_CMA_AREAS struct dma_contig_early_reserve { phys_addr_t base; unsigned long size; @@ -435,6 +436,7 @@ void __init dma_contiguous_remap(void) iotable_init(&map, 1); } } +#endif static int __dma_update_pte(pte_t *pte, unsigned long addr, void *data) { diff --git a/arch/arm/mm/mm.h b/arch/arm/mm/mm.h index 9ff683612f2a..d7ffccb7fea7 100644 --- a/arch/arm/mm/mm.h +++ b/arch/arm/mm/mm.h @@ -88,6 +88,10 @@ extern phys_addr_t arm_lowmem_limit; void __init bootmem_init(void); void arm_mm_memblock_reserve(void); +#ifdef CONFIG_CMA_AREAS void dma_contiguous_remap(void); +#else +static inline void dma_contiguous_remap(void) { } +#endif unsigned long __clear_cr(unsigned long mask); diff --git a/include/linux/cma.h b/include/linux/cma.h index bd801023504b..2c2ede7f0724 100644 --- a/include/linux/cma.h +++ b/include/linux/cma.h @@ -12,10 +12,6 @@ */ #ifdef CONFIG_CMA_AREAS #define MAX_CMA_AREAS (1 + CONFIG_CMA_AREAS) - -#else -#define MAX_CMA_AREAS (0) - #endif #define CMA_MAX_NAME 64 -- cgit v1.2.3 From 11ac3e87ce09c27f4587a8c4fe0829d814021a82 Mon Sep 17 00:00:00 2001 From: Zi Yan Date: Thu, 12 May 2022 20:22:58 -0700 Subject: mm: cma: use pageblock_order as the single alignment Now alloc_contig_range() works at pageblock granularity. Change CMA allocation, which uses alloc_contig_range(), to use pageblock_nr_pages alignment. Link: https://lkml.kernel.org/r/20220425143118.2850746-6-zi.yan@sent.com Signed-off-by: Zi Yan Cc: Christophe Leroy Cc: David Hildenbrand Cc: Eric Ren Cc: kernel test robot Cc: Mel Gorman Cc: Mike Rapoport Cc: Minchan Kim Cc: Oscar Salvador Cc: Vlastimil Babka Signed-off-by: Andrew Morton --- include/linux/cma.h | 4 ++-- include/linux/mmzone.h | 5 +---- mm/page_alloc.c | 4 ++-- 3 files changed, 5 insertions(+), 8 deletions(-) (limited to 'include/linux/cma.h') diff --git a/include/linux/cma.h b/include/linux/cma.h index a6f637342740..63873b93deaa 100644 --- a/include/linux/cma.h +++ b/include/linux/cma.h @@ -17,11 +17,11 @@ #define CMA_MAX_NAME 64 /* - * TODO: once the buddy -- especially pageblock merging and alloc_contig_range() + * the buddy -- especially pageblock merging and alloc_contig_range() * -- can deal with only some pageblocks of a higher-order page being * MIGRATE_CMA, we can use pageblock_nr_pages. */ -#define CMA_MIN_ALIGNMENT_PAGES MAX_ORDER_NR_PAGES +#define CMA_MIN_ALIGNMENT_PAGES pageblock_nr_pages #define CMA_MIN_ALIGNMENT_BYTES (PAGE_SIZE * CMA_MIN_ALIGNMENT_PAGES) struct cma; diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index 46ffab808f03..aab70355d64f 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -54,10 +54,7 @@ enum migratetype { * * The way to use it is to change migratetype of a range of * pageblocks to MIGRATE_CMA which can be done by - * __free_pageblock_cma() function. What is important though - * is that a range of pageblocks must be aligned to - * MAX_ORDER_NR_PAGES should biggest page be bigger than - * a single pageblock. + * __free_pageblock_cma() function. */ MIGRATE_CMA, #endif diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 3d7de8a96143..732e01e9d0e0 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -9046,8 +9046,8 @@ int __alloc_contig_migrate_range(struct compact_control *cc, * be either of the two. * @gfp_mask: GFP mask to use during compaction * - * The PFN range does not have to be pageblock or MAX_ORDER_NR_PAGES - * aligned. The PFN range must belong to a single zone. + * The PFN range does not have to be pageblock aligned. The PFN range must + * belong to a single zone. * * The first thing this routine does is attempt to MIGRATE_ISOLATE all * pageblocks in the range. Once isolated, the pageblocks should not -- cgit v1.2.3