From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrea Righi Subject: Re: [PATCH 3/9] bio-cgroup controller Date: Thu, 16 Apr 2009 12:42:36 +0200 Message-ID: <20090416104234.GA6656__6936.69005199826$1239878745$gmane$org@linux> References: <1239740480-28125-4-git-send-email-righi.andrea@gmail.com> <20090415111528.b796519a.kamezawa.hiroyu@jp.fujitsu.com> <20090415093716.GA5968@linux> <20090415.213850.226770691.ryov@valinux.co.jp> <20090415132356.GA19270@linux> <20090416085814.8b6d077f.kamezawa.hiroyu@jp.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20090416085814.8b6d077f.kamezawa.hiroyu-+CUm20s59erQFUHtdCDX3A@public.gmane.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org Errors-To: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org To: KAMEZAWA Hiroyuki Cc: randy.dunlap-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org, menage-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org, chlunde-om2ZC0WAoZIXWF+eFR7m5Q@public.gmane.org, eric.rannaud-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org, balbir-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org, fernando-gVGce1chcLdL9jVzuh4AOg@public.gmane.org, dradford-cT2on/YLNlBWk0Htik3J/w@public.gmane.org, agk-9JcytcrH/bA+uJoB2kUjGw@public.gmane.org, subrata-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org, axboe-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org, akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org, containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, dave-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org, matt-cT2on/YLNlBWk0Htik3J/w@public.gmane.org, roberto-5KDOxZqKugI@public.gmane.org, ngupta-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org List-Id: containers.vger.kernel.org On Thu, Apr 16, 2009 at 08:58:14AM +0900, KAMEZAWA Hiroyuki wrote: > On Wed, 15 Apr 2009 15:23:57 +0200 > Andrea Righi wrote: > > > On Wed, Apr 15, 2009 at 09:38:50PM +0900, Ryo Tsuruta wrote: > > > Hi Andrea and Kamezawa-san, > > > > > > > Ryo, it would be great if you can look at this and fix/integrate into > > > > the mainstream bio-cgroup. Otherwise I can try to to schedule this in my > > > > work. > > > > > > O.K. I'll apply those fixes and post patches as soon as I can. > > > > > > > Very good! I've just tested the bio_cgroup_id inclusion in > > page_cgroup->flags. I'm posting the patch on-top-of my patchset. > > > > If you're interested, it should apply cleanly to the original > > bio-cgroup, except for the get/put_cgroup_from_page() part. > > > > Thanks, > > -Andrea > > --- > > bio-cgroup: encode bio_cgroup_id in page_cgroup->flags > > > > Encode the bio_cgroup_id into the flags argument of page_cgroup as > > suggested by Kamezawa. > > > > Lower 16-bits of the flags attribute are used for the actual page_cgroup > > flags. The rest is reserved to store the bio-cgroup id. > > > > This allows to save 4 bytes (in 32-bit architectures) or 8 bytes (in > > 64-bit) for each page_cgroup element. > > > > Signed-off-by: Andrea Righi > > --- > > include/linux/biotrack.h | 2 +- > > include/linux/page_cgroup.h | 24 +++++++++++++++++++++--- > > mm/biotrack.c | 26 ++++++++++++-------------- > > 3 files changed, 34 insertions(+), 18 deletions(-) > > > > diff --git a/include/linux/biotrack.h b/include/linux/biotrack.h > > index 25b8810..4bd0242 100644 > > --- a/include/linux/biotrack.h > > +++ b/include/linux/biotrack.h > > @@ -28,7 +28,7 @@ struct bio_cgroup { > > > > static inline void __init_bio_page_cgroup(struct page_cgroup *pc) > > { > > - pc->bio_cgroup_id = 0; > > + page_cgroup_set_bio_id(pc, 0); > > } > > > > extern struct cgroup *get_cgroup_from_page(struct page *page); > > diff --git a/include/linux/page_cgroup.h b/include/linux/page_cgroup.h > > index 00a49c5..af780a4 100644 > > --- a/include/linux/page_cgroup.h > > +++ b/include/linux/page_cgroup.h > > @@ -16,12 +16,30 @@ struct page_cgroup { > > #ifdef CONFIG_CGROUP_MEM_RES_CTLR > > struct mem_cgroup *mem_cgroup; > > #endif > > -#ifdef CONFIG_CGROUP_BIO > > - int bio_cgroup_id; > > -#endif > > struct list_head lru; /* per cgroup LRU list */ > > }; > > > > +#ifdef CONFIG_CGROUP_BIO > > +/* > > + * use lower 16 bits for flags and reserve the rest for the bio-cgroup id > > + */ > > +#define BIO_CGROUP_ID_SHIFT (16) > > +#define BIO_CGROUP_ID_BITS (8 * sizeof(unsigned long) - BIO_CGROUP_ID_SHIFT) > > + > > +static inline unsigned long page_cgroup_get_bio_id(struct page_cgroup *pc) > > +{ > > + return pc->flags >> BIO_CGROUP_ID_SHIFT; > > +} > > + > > +static inline void page_cgroup_set_bio_id(struct page_cgroup *pc, > > + unsigned long id) > > +{ > > + WARN_ON(id >= (1UL << BIO_CGROUP_ID_BITS)); > > + pc->flags &= (1UL << BIO_CGROUP_ID_SHIFT) - 1; > > + pc->flags |= (unsigned long)(id << BIO_CGROUP_ID_SHIFT); > > +} > > +#endif > > + > Ah, there is "Lock" bit in pc->flags and above "set" code does read-modify-write > without lock_page_cgroup(). > > Could you use lock_page_cgroup() or cmpxchg ? (or using something magical technique ?) If I'm not wrong this should guarantee atomicity without using lock_page_cgroup(). Thanks, -Andrea --- bio-cgroup: encode bio_cgroup_id in page_cgroup->flags Encode the bio_cgroup_id into the flags argument of page_cgroup as suggested by Kamezawa. Lower 16 bits (in 32-bit archs) or lower 32 bits (in 64-bit archs) of the flags attribute are used for the actual page_cgroup flags. The upper bits are reserved to store the bio-cgroup id. This allows to save 4 bytes (in 32-bit architectures) or 8 bytes (in 64-bit) for each page_cgroup element. Signed-off-by: Andrea Righi --- include/linux/page_cgroup.h | 42 +++++++++++++++++++++++++++++++++++++----- mm/biotrack.c | 24 +++++++++++------------- 2 files changed, 48 insertions(+), 18 deletions(-) diff --git a/include/linux/page_cgroup.h b/include/linux/page_cgroup.h index a7249bb..864ad6f 100644 --- a/include/linux/page_cgroup.h +++ b/include/linux/page_cgroup.h @@ -16,14 +16,46 @@ struct page_cgroup { #ifdef CONFIG_CGROUP_MEM_RES_CTLR struct mem_cgroup *mem_cgroup; #endif -#ifdef CONFIG_CGROUP_BIO - int bio_cgroup_id; -#endif -#if defined(CONFIG_CGROUP_MEM_RES_CTLR) || defined(CONFIG_CGROUP_BIO) struct list_head lru; /* per cgroup LRU list */ -#endif }; +#ifdef CONFIG_CGROUP_BIO +/* + * Use the lower 16 bits (in 32-bit archs) or lower 32 bits (in 64-bit archs) + * of page_cgroup->flags for the actual flags and reserve the rest for the + * bio-cgroup id. + * + * This allows to atomically read and write the bio-cgroup id without using + * lock/unlock_page_cgroup(). + */ +#if defined(CONFIG_64BIT) +typedef uint32_t bio_cgroup_id_t; +#elif defined(CONFIG_32BIT) +typedef uint16_t bio_cgroup_id_t; +#else +#error "unsupported architecture" +#endif + +#define BIO_CGROUP_ID_SHIFT (sizeof(bio_cgroup_id_t) * 8) +#define BIO_CGROUP_ID_BITS (sizeof(bio_cgroup_id_t) * 8) +#define BIO_CGROUP_ID_MASK ((1UL << BIO_CGROUP_ID_BITS) - 1) + +static inline unsigned long page_cgroup_get_bio_id(struct page_cgroup *pc) +{ + return (pc->flags >> BIO_CGROUP_ID_SHIFT) & BIO_CGROUP_ID_MASK; +} + +static inline void page_cgroup_set_bio_id(struct page_cgroup *pc, + unsigned long id) +{ + bio_cgroup_id_t *ptr = (bio_cgroup_id_t *)((unsigned char *)&pc->flags + + (BIO_CGROUP_ID_SHIFT / 8)); + + WARN_ON(id >= (1UL << BIO_CGROUP_ID_BITS)); + *ptr = (bio_cgroup_id_t)id; +} +#endif /* CONFIG_CGROUP_BIO */ + void __meminit pgdat_page_cgroup_init(struct pglist_data *pgdat); void __init page_cgroup_init(void); struct page_cgroup *lookup_page_cgroup(struct page *page); diff --git a/mm/biotrack.c b/mm/biotrack.c index d3a35f1..01f83ba 100644 --- a/mm/biotrack.c +++ b/mm/biotrack.c @@ -80,7 +80,7 @@ void bio_cgroup_set_owner(struct page *page, struct mm_struct *mm) if (unlikely(!pc)) return; - pc->bio_cgroup_id = 0; /* 0: default bio_cgroup id */ + __init_bio_page_cgroup(pc); if (!mm) return; /* @@ -93,11 +93,11 @@ void bio_cgroup_set_owner(struct page *page, struct mm_struct *mm) goto out; /* * css_get(&bio->css) isn't called to increment the reference - * count of this bio_cgroup "biog" so pc->bio_cgroup_id might turn - * invalid even if this page is still active. - * This approach is chosen to minimize the overhead. + * count of this bio_cgroup "biog" so the bio-cgroup id might turn + * invalid even if this page is still active. This approach is chosen + * to minimize the overhead. */ - pc->bio_cgroup_id = biog->id; + page_cgroup_set_bio_id(pc, biog->id); out: rcu_read_unlock(); } @@ -110,7 +110,7 @@ void bio_cgroup_reset_owner(struct page *page, struct mm_struct *mm) /* * A little trick: * Just call bio_cgroup_set_owner() for pages which are already - * active since the bio_cgroup_id member of page_cgroup can be + * active since the bio-cgroup id member of page_cgroup can be * updated without any locks. This is because an integer type of * variable can be set a new value at once on modern cpus. */ @@ -146,12 +146,10 @@ void bio_cgroup_copy_owner(struct page *npage, struct page *opage) opc = lookup_page_cgroup(opage); if (unlikely(!opc)) return; - /* - * Do this without any locks. The reason is the same as - * bio_cgroup_reset_owner(). + * XXX: is it safe to do this without locking? */ - npc->bio_cgroup_id = opc->bio_cgroup_id; + page_cgroup_set_bio_id(npc, page_cgroup_get_bio_id(opc)); } /* Create a new bio-cgroup. */ @@ -248,7 +246,7 @@ struct cgroup *get_cgroup_from_page(struct page *page) if (!pc) return NULL; lock_page_cgroup(pc); - biog = find_bio_cgroup(pc->bio_cgroup_id); + biog = find_bio_cgroup(page_cgroup_get_bio_id(pc)); if (biog) { css_get(&biog->css); cgrp = biog->css.cgroup; @@ -266,7 +264,7 @@ void put_cgroup_from_page(struct page *page) if (!pc) return; lock_page_cgroup(pc); - biog = find_bio_cgroup(pc->bio_cgroup_id); + biog = find_bio_cgroup(page_cgroup_get_bio_id(pc)); if (biog) css_put(&biog->css); unlock_page_cgroup(pc); @@ -281,7 +279,7 @@ int get_bio_cgroup_id(struct bio *bio) pc = lookup_page_cgroup(page); if (pc) - id = pc->bio_cgroup_id; + id = page_cgroup_get_bio_id(pc); return id; } EXPORT_SYMBOL(get_bio_cgroup_id);