| Message ID | 20200916183411.64756-5-david@redhat.com (mailing list archive) |
|---|---|
| State | RFC, archived |
| Headers | show |
| Series | mm: place pages to the freelist tail when onling and undoing isolation | expand |
On 9/16/20 8:34 PM, David Hildenbrand wrote: > __free_pages_core() is used when exposing fresh memory to the buddy > during system boot and when onlining memory in generic_online_page(). > > generic_online_page() is used in two cases: > > 1. Direct memory onlining in online_pages(). > 2. Deferred memory onlining in memory-ballooning-like mechanisms (HyperV > balloon and virtio-mem), when parts of a section are kept > fake-offline to be fake-onlined later on. > > In 1, we already place pages to the tail of the freelist. Pages will be > freed to MIGRATE_ISOLATE lists first and moved to the tail of the freelists > via undo_isolate_page_range(). > > In 2, we currently don't implement a proper rule. In case of virtio-mem, > where we currently always online MAX_ORDER - 1 pages, the pages will be > placed to the HEAD of the freelist - undesireable. While the hyper-v > balloon calls generic_online_page() with single pages, usually it will > call it on successive single pages in a larger block. > > The pages are fresh, so place them to the tail of the freelists and avoid > the PCP. > > Note: If we detect that the new behavior is undesireable for > __free_pages_core() during boot, we can let the caller specify the > behavior. > > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: Alexander Duyck <alexander.h.duyck@linux.intel.com> > Cc: Mel Gorman <mgorman@techsingularity.net> > Cc: Michal Hocko <mhocko@kernel.org> > Cc: Dave Hansen <dave.hansen@intel.com> > Cc: Vlastimil Babka <vbabka@suse.cz> > Cc: Wei Yang <richard.weiyang@linux.alibaba.com> > Cc: Oscar Salvador <osalvador@suse.de> > Cc: Mike Rapoport <rppt@kernel.org> > Cc: "K. Y. Srinivasan" <kys@microsoft.com> > Cc: Haiyang Zhang <haiyangz@microsoft.com> > Cc: Stephen Hemminger <sthemmin@microsoft.com> > Cc: Wei Liu <wei.liu@kernel.org> > Signed-off-by: David Hildenbrand <david@redhat.com> Reviewed-by: Vlastimil Babka <vbabka@suse.cz> > --- > mm/page_alloc.c | 32 ++++++++++++++++++++------------ > 1 file changed, 20 insertions(+), 12 deletions(-) > > diff --git a/mm/page_alloc.c b/mm/page_alloc.c > index 75b0f49b4022..50746e6dc21b 100644 > --- a/mm/page_alloc.c > +++ b/mm/page_alloc.c > @@ -264,7 +264,8 @@ bool pm_suspended_storage(void) > unsigned int pageblock_order __read_mostly; > #endif > > -static void __free_pages_ok(struct page *page, unsigned int order); > +static void __free_pages_ok(struct page *page, unsigned int order, > + fop_t fop_flags); > > /* > * results with 256, 32 in the lowmem_reserve sysctl: > @@ -676,7 +677,7 @@ static void bad_page(struct page *page, const char *reason) > void free_compound_page(struct page *page) > { > mem_cgroup_uncharge(page); > - __free_pages_ok(page, compound_order(page)); > + __free_pages_ok(page, compound_order(page), FOP_NONE); > } > > void prep_compound_page(struct page *page, unsigned int order) > @@ -1402,17 +1403,15 @@ static void free_pcppages_bulk(struct zone *zone, int count, > spin_unlock(&zone->lock); > } > > -static void free_one_page(struct zone *zone, > - struct page *page, unsigned long pfn, > - unsigned int order, > - int migratetype) > +static void free_one_page(struct zone *zone, struct page *page, unsigned long pfn, > + unsigned int order, int migratetype, fop_t fop_flags) > { > spin_lock(&zone->lock); > if (unlikely(has_isolate_pageblock(zone) || > is_migrate_isolate(migratetype))) { > migratetype = get_pfnblock_migratetype(page, pfn); > } > - __free_one_page(page, pfn, zone, order, migratetype, FOP_NONE); > + __free_one_page(page, pfn, zone, order, migratetype, fop_flags); > spin_unlock(&zone->lock); > } > > @@ -1490,7 +1489,8 @@ void __meminit reserve_bootmem_region(phys_addr_t start, phys_addr_t end) > } > } > > -static void __free_pages_ok(struct page *page, unsigned int order) > +static void __free_pages_ok(struct page *page, unsigned int order, > + fop_t fop_flags) > { > unsigned long flags; > int migratetype; > @@ -1502,7 +1502,8 @@ static void __free_pages_ok(struct page *page, unsigned int order) > migratetype = get_pfnblock_migratetype(page, pfn); > local_irq_save(flags); > __count_vm_events(PGFREE, 1 << order); > - free_one_page(page_zone(page), page, pfn, order, migratetype); > + free_one_page(page_zone(page), page, pfn, order, migratetype, > + fop_flags); > local_irq_restore(flags); > } > > @@ -1523,7 +1524,13 @@ void __free_pages_core(struct page *page, unsigned int order) > > atomic_long_add(nr_pages, &page_zone(page)->managed_pages); > set_page_refcounted(page); > - __free_pages(page, order); > + > + /* > + * Bypass PCP and place fresh pages right to the tail, primarily > + * relevant for memory onlining. > + */ > + page_ref_dec(page); > + __free_pages_ok(page, order, FOP_TO_TAIL); > } > > #ifdef CONFIG_NEED_MULTIPLE_NODES > @@ -3167,7 +3174,8 @@ static void free_unref_page_commit(struct page *page, unsigned long pfn) > */ > if (migratetype >= MIGRATE_PCPTYPES) { > if (unlikely(is_migrate_isolate(migratetype))) { > - free_one_page(zone, page, pfn, 0, migratetype); > + free_one_page(zone, page, pfn, 0, migratetype, > + FOP_NONE); > return; > } > migratetype = MIGRATE_MOVABLE; > @@ -4984,7 +4992,7 @@ static inline void free_the_page(struct page *page, unsigned int order) > if (order == 0) /* Via pcp? */ > free_unref_page(page); > else > - __free_pages_ok(page, order); > + __free_pages_ok(page, order, FOP_NONE); > } > > void __free_pages(struct page *page, unsigned int order) >
On Wed, Sep 16, 2020 at 08:34:11PM +0200, David Hildenbrand wrote: > @@ -1523,7 +1524,13 @@ void __free_pages_core(struct page *page, unsigned int order) > > atomic_long_add(nr_pages, &page_zone(page)->managed_pages); > set_page_refcounted(page); > - __free_pages(page, order); > + > + /* > + * Bypass PCP and place fresh pages right to the tail, primarily > + * relevant for memory onlining. > + */ > + page_ref_dec(page); > + __free_pages_ok(page, order, FOP_TO_TAIL); Sorry, I must be missing something obvious here, but I am a bit confused here. I get the part of placing them at the tail so rmqueue_bulk() won't find them, but I do not get why we decrement page's refcount. IIUC, its refcount will be 0, but why do we want to do that? Another thing a bit unrelated... we mess three times with page's refcount (two before this patch). Why do we have this dance in place? Thanks
On 28.09.20 09:58, Oscar Salvador wrote: > On Wed, Sep 16, 2020 at 08:34:11PM +0200, David Hildenbrand wrote: >> @@ -1523,7 +1524,13 @@ void __free_pages_core(struct page *page, unsigned int order) >> >> atomic_long_add(nr_pages, &page_zone(page)->managed_pages); >> set_page_refcounted(page); >> - __free_pages(page, order); >> + >> + /* >> + * Bypass PCP and place fresh pages right to the tail, primarily >> + * relevant for memory onlining. >> + */ >> + page_ref_dec(page); >> + __free_pages_ok(page, order, FOP_TO_TAIL); > > Sorry, I must be missing something obvious here, but I am a bit confused here. > I get the part of placing them at the tail so rmqueue_bulk() won't > find them, but I do not get why we decrement page's refcount. > IIUC, its refcount will be 0, but why do we want to do that? > > Another thing a bit unrelated... we mess three times with page's refcount > (two before this patch). > Why do we have this dance in place? Hi Oscar! Old code: set_page_refcounted(): sets the refcount to 1. __free_pages() -> put_page_testzero(): sets it to 0 -> free_the_page()->__free_pages_ok() New code: set_page_refcounted(): sets the refcount to 1. page_ref_dec(page): sets it to 0 __free_pages_ok(): We could skip the set_page_refcounted() + page_ref_dec(page) and lose a couple of sanity checks but we could simply use a VM_BUG_ON_PAGE(page_ref_count(page), page), which is what we really care about when onlining memory.
On Mon, Sep 28, 2020 at 10:36:00AM +0200, David Hildenbrand wrote: > Hi Oscar! Hi David :-) > > Old code: > > set_page_refcounted(): sets the refcount to 1. > __free_pages() > -> put_page_testzero(): sets it to 0 > -> free_the_page()->__free_pages_ok() > > New code: > > set_page_refcounted(): sets the refcount to 1. > page_ref_dec(page): sets it to 0 > __free_pages_ok(): bleh, I misread the patch, somehow I managed to not see that you replaced __free_pages with __free_pages_ok. To be honest, now that we do not need the page's refcount to be 1 for the put_page_testzero to trigger (and since you are decrementing it anyways), I think it would be much clear for those two to be gone. But not strong, so: Reviewed-by: Oscar Salvador <osalvador@suse.de>
diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 75b0f49b4022..50746e6dc21b 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -264,7 +264,8 @@ bool pm_suspended_storage(void) unsigned int pageblock_order __read_mostly; #endif -static void __free_pages_ok(struct page *page, unsigned int order); +static void __free_pages_ok(struct page *page, unsigned int order, + fop_t fop_flags); /* * results with 256, 32 in the lowmem_reserve sysctl: @@ -676,7 +677,7 @@ static void bad_page(struct page *page, const char *reason) void free_compound_page(struct page *page) { mem_cgroup_uncharge(page); - __free_pages_ok(page, compound_order(page)); + __free_pages_ok(page, compound_order(page), FOP_NONE); } void prep_compound_page(struct page *page, unsigned int order) @@ -1402,17 +1403,15 @@ static void free_pcppages_bulk(struct zone *zone, int count, spin_unlock(&zone->lock); } -static void free_one_page(struct zone *zone, - struct page *page, unsigned long pfn, - unsigned int order, - int migratetype) +static void free_one_page(struct zone *zone, struct page *page, unsigned long pfn, + unsigned int order, int migratetype, fop_t fop_flags) { spin_lock(&zone->lock); if (unlikely(has_isolate_pageblock(zone) || is_migrate_isolate(migratetype))) { migratetype = get_pfnblock_migratetype(page, pfn); } - __free_one_page(page, pfn, zone, order, migratetype, FOP_NONE); + __free_one_page(page, pfn, zone, order, migratetype, fop_flags); spin_unlock(&zone->lock); } @@ -1490,7 +1489,8 @@ void __meminit reserve_bootmem_region(phys_addr_t start, phys_addr_t end) } } -static void __free_pages_ok(struct page *page, unsigned int order) +static void __free_pages_ok(struct page *page, unsigned int order, + fop_t fop_flags) { unsigned long flags; int migratetype; @@ -1502,7 +1502,8 @@ static void __free_pages_ok(struct page *page, unsigned int order) migratetype = get_pfnblock_migratetype(page, pfn); local_irq_save(flags); __count_vm_events(PGFREE, 1 << order); - free_one_page(page_zone(page), page, pfn, order, migratetype); + free_one_page(page_zone(page), page, pfn, order, migratetype, + fop_flags); local_irq_restore(flags); } @@ -1523,7 +1524,13 @@ void __free_pages_core(struct page *page, unsigned int order) atomic_long_add(nr_pages, &page_zone(page)->managed_pages); set_page_refcounted(page); - __free_pages(page, order); + + /* + * Bypass PCP and place fresh pages right to the tail, primarily + * relevant for memory onlining. + */ + page_ref_dec(page); + __free_pages_ok(page, order, FOP_TO_TAIL); } #ifdef CONFIG_NEED_MULTIPLE_NODES @@ -3167,7 +3174,8 @@ static void free_unref_page_commit(struct page *page, unsigned long pfn) */ if (migratetype >= MIGRATE_PCPTYPES) { if (unlikely(is_migrate_isolate(migratetype))) { - free_one_page(zone, page, pfn, 0, migratetype); + free_one_page(zone, page, pfn, 0, migratetype, + FOP_NONE); return; } migratetype = MIGRATE_MOVABLE; @@ -4984,7 +4992,7 @@ static inline void free_the_page(struct page *page, unsigned int order) if (order == 0) /* Via pcp? */ free_unref_page(page); else - __free_pages_ok(page, order); + __free_pages_ok(page, order, FOP_NONE); } void __free_pages(struct page *page, unsigned int order)
__free_pages_core() is used when exposing fresh memory to the buddy during system boot and when onlining memory in generic_online_page(). generic_online_page() is used in two cases: 1. Direct memory onlining in online_pages(). 2. Deferred memory onlining in memory-ballooning-like mechanisms (HyperV balloon and virtio-mem), when parts of a section are kept fake-offline to be fake-onlined later on. In 1, we already place pages to the tail of the freelist. Pages will be freed to MIGRATE_ISOLATE lists first and moved to the tail of the freelists via undo_isolate_page_range(). In 2, we currently don't implement a proper rule. In case of virtio-mem, where we currently always online MAX_ORDER - 1 pages, the pages will be placed to the HEAD of the freelist - undesireable. While the hyper-v balloon calls generic_online_page() with single pages, usually it will call it on successive single pages in a larger block. The pages are fresh, so place them to the tail of the freelists and avoid the PCP. Note: If we detect that the new behavior is undesireable for __free_pages_core() during boot, we can let the caller specify the behavior. Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Alexander Duyck <alexander.h.duyck@linux.intel.com> Cc: Mel Gorman <mgorman@techsingularity.net> Cc: Michal Hocko <mhocko@kernel.org> Cc: Dave Hansen <dave.hansen@intel.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Wei Yang <richard.weiyang@linux.alibaba.com> Cc: Oscar Salvador <osalvador@suse.de> Cc: Mike Rapoport <rppt@kernel.org> Cc: "K. Y. Srinivasan" <kys@microsoft.com> Cc: Haiyang Zhang <haiyangz@microsoft.com> Cc: Stephen Hemminger <sthemmin@microsoft.com> Cc: Wei Liu <wei.liu@kernel.org> Signed-off-by: David Hildenbrand <david@redhat.com> --- mm/page_alloc.c | 32 ++++++++++++++++++++------------ 1 file changed, 20 insertions(+), 12 deletions(-)