From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:39586) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fn6yX-0007pO-Is for qemu-devel@nongnu.org; Tue, 07 Aug 2018 14:44:58 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fn6yS-0006II-QX for qemu-devel@nongnu.org; Tue, 07 Aug 2018 14:44:57 -0400 Received: from mx3-rdu2.redhat.com ([66.187.233.73]:38098 helo=mx1.redhat.com) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1fn6yS-0006I0-JS for qemu-devel@nongnu.org; Tue, 07 Aug 2018 14:44:52 -0400 Date: Tue, 7 Aug 2018 19:44:47 +0100 From: "Dr. David Alan Gilbert" Message-ID: <20180807184447.GS2556@work-vm> References: <20180722193350.6028-1-zhangckid@gmail.com> <20180722193350.6028-9-zhangckid@gmail.com> <20180727185131.GD2498@work-vm> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Subject: Re: [Qemu-devel] [PATCH V10 08/20] ram/COLO: Record the dirty pages that SVM received List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Zhang Chen Cc: qemu-devel@nongnu.org, Paolo Bonzini , Juan Quintela , Jason Wang , Eric Blake , Markus Armbruster , zhanghailiang , Li Zhijian * Zhang Chen (zhangckid@gmail.com) wrote: > On Sat, Jul 28, 2018 at 2:51 AM, Dr. David Alan Gilbert > wrote: > > > * Zhang Chen (zhangckid@gmail.com) wrote: > > > We record the address of the dirty pages that received, > > > it will help flushing pages that cached into SVM. > > > > > > Here, it is a trick, we record dirty pages by re-using migration > > > dirty bitmap. In the later patch, we will start the dirty log > > > for SVM, just like migration, in this way, we can record both > > > the dirty pages caused by PVM and SVM, we only flush those dirty > > > pages from RAM cache while do checkpoint. > > > > > > Signed-off-by: zhanghailiang > > > Reviewed-by: Dr. David Alan Gilbert > > > --- > > > migration/ram.c | 33 +++++++++++++++++++++++++++++++++ > > > 1 file changed, 33 insertions(+) > > > > > > diff --git a/migration/ram.c b/migration/ram.c > > > index 33ebd09d70..d1060f1337 100644 > > > --- a/migration/ram.c > > > +++ b/migration/ram.c > > > @@ -3325,6 +3325,15 @@ static inline void *colo_cache_from_block_offset(RAMBlock > > *block, > > > __func__, block->idstr); > > > return NULL; > > > } > > > + > > > + /* > > > + * During colo checkpoint, we need bitmap of these migrated pages. > > > + * It help us to decide which pages in ram cache should be flushed > > > + * into VM's RAM later. > > > + */ > > > + if (!test_and_set_bit(offset >> TARGET_PAGE_BITS, block->bmap)) { > > > + ram_state->migration_dirty_pages++; > > > + } > > > return block->colo_cache + offset; > > > } > > > > > > @@ -3555,6 +3564,24 @@ int colo_init_ram_cache(void) > > > memcpy(block->colo_cache, block->host, block->used_length); > > > } > > > rcu_read_unlock(); > > > + /* > > > + * Record the dirty pages that sent by PVM, we use this dirty bitmap > > together > > > + * with to decide which page in cache should be flushed into SVM's > > RAM. Here > > > + * we use the same name 'ram_bitmap' as for migration. > > > + */ > > > + if (ram_bytes_total()) { > > > + RAMBlock *block; > > > + > > > + QLIST_FOREACH_RCU(block, &ram_list.blocks, next) { > > > > I think those need updating to check for 'qemu_ram_is_migratable' - it > > might be worth moving RAMBLOCK_FOREACH_MIGRATABLE. > > > > > OK, I will change to the " RAMBLOCK_FOREACH_MIGRATABLE " in next version. > > > > > > > + unsigned long pages = block->max_length >> TARGET_PAGE_BITS; > > > + > > > + block->bmap = bitmap_new(pages); > > > + bitmap_set(block->bmap, 0, pages); > > > > Would it make sense to use the 'receivedmap' that was recently added > > for other uses rather than needing your own? > > > > Maybe we can do this job as optimization in the future, currently we want > to make this series merged into upstream first. Sure; I think it's getting there; I don't see anything too bad in the migration bits as is; hopefully Jason can check the networking set again and Markus/Eric check the QMP side of things. Dave > Thanks > Zhang Chen > > > > > > Dave > > > > > + } > > > + } > > > + ram_state = g_new0(RAMState, 1); > > > + ram_state->migration_dirty_pages = 0; > > > + > > > return 0; > > > > > > out_locked: > > > @@ -3574,6 +3601,10 @@ void colo_release_ram_cache(void) > > > { > > > RAMBlock *block; > > > > > > + QLIST_FOREACH_RCU(block, &ram_list.blocks, next) { > > > + g_free(block->bmap); > > > + block->bmap = NULL; > > > + } > > > rcu_read_lock(); > > > QLIST_FOREACH_RCU(block, &ram_list.blocks, next) { > > > if (block->colo_cache) { > > > @@ -3582,6 +3613,8 @@ void colo_release_ram_cache(void) > > > } > > > } > > > rcu_read_unlock(); > > > + g_free(ram_state); > > > + ram_state = NULL; > > > } > > > > > > /** > > > -- > > > 2.17.1 > > > > > -- > > Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK > > -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK