From patchwork Tue Dec 2 12:39:27 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Miao Xie X-Patchwork-Id: 5419691 Return-Path: X-Original-To: patchwork-linux-btrfs@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork1.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.19.201]) by patchwork1.web.kernel.org (Postfix) with ESMTP id BBEEE9F1C5 for ; Tue, 2 Dec 2014 12:38:37 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id F24F6201ED for ; Tue, 2 Dec 2014 12:38:34 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 27C0A202AE for ; Tue, 2 Dec 2014 12:38:32 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933049AbaLBMi2 (ORCPT ); Tue, 2 Dec 2014 07:38:28 -0500 Received: from cn.fujitsu.com ([59.151.112.132]:15348 "EHLO heian.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S932916AbaLBMiM (ORCPT ); Tue, 2 Dec 2014 07:38:12 -0500 X-IronPort-AV: E=Sophos;i="5.04,848,1406563200"; d="scan'208";a="44350253" Received: from unknown (HELO edo.cn.fujitsu.com) ([10.167.33.5]) by heian.cn.fujitsu.com with ESMTP; 02 Dec 2014 20:34:48 +0800 Received: from G08CNEXCHPEKD03.g08.fujitsu.local (localhost.localdomain [127.0.0.1]) by edo.cn.fujitsu.com (8.14.3/8.13.1) with ESMTP id sB2Cbj20001473 for ; Tue, 2 Dec 2014 20:37:45 +0800 Received: from miao.g08.fujitsu.local (10.167.226.169) by G08CNEXCHPEKD03.g08.fujitsu.local (10.167.33.89) with Microsoft SMTP Server (TLS) id 14.3.181.6; Tue, 2 Dec 2014 20:38:06 +0800 From: Miao Xie To: Subject: [PATCH v4 06/10] Btrfs, raid56: support parity scrub on raid56 Date: Tue, 2 Dec 2014 20:39:27 +0800 Message-ID: <1417523971-15553-7-git-send-email-miaox@cn.fujitsu.com> X-Mailer: git-send-email 1.9.3 In-Reply-To: <1417523971-15553-1-git-send-email-miaox@cn.fujitsu.com> References: <1417523971-15553-1-git-send-email-miaox@cn.fujitsu.com> MIME-Version: 1.0 X-Originating-IP: [10.167.226.169] Sender: linux-btrfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-btrfs@vger.kernel.org X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, T_RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=ham version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP The implementation is: - Read and check all the data with checksum in the same stripe. All the data which has checksum is COW data, and we are sure that it is not changed though we don't lock the stripe. because the space of that data just can be reclaimed after the current transction is committed, and then the fs can use it to store the other data, but when doing scrub, we hold the current transaction, that is that data can not be recovered, it is safe that read and check it out of the stripe lock. - Lock the stripe - Read out all the data without checksum and parity The data without checksum and the parity may be changed if we don't lock the stripe, so we need read it in the stripe lock context. - Check the parity - Re-calculate the new parity and write back it if the old parity is not right - Unlock the stripe If we can not read out the data or the data we read is corrupted, we will try to repair it. If the repair fails. we will mark the horizontal sub-stripe(pages on the same horizontal) as corrupted sub-stripe, and we will skip the parity check and repair of that horizontal sub-stripe. And in order to skip the horizontal sub-stripe that has no data, we introduce a bitmap. If there is some data on the horizontal sub-stripe, we will the relative bit to 1, and when we check and repair the parity, we will skip those horizontal sub-stripes that the relative bits is 0. Signed-off-by: Miao Xie --- Changelog v3 -> v4: - None. Changelog v2 -> v3: - Fix wrong stripe start logical address calculation which was reported by Chris. - Fix unhandled raid bios for parity scrub, which are added into the plug list of the head raid bio. - Modify the code that is used to avoid the rbio merge. Changelog v1 -> v2: - None. --- fs/btrfs/raid56.c | 514 ++++++++++++++++++++++++++++++++++++++++++++- fs/btrfs/raid56.h | 12 ++ fs/btrfs/scrub.c | 609 ++++++++++++++++++++++++++++++++++++++++++++++++++++-- 3 files changed, 1115 insertions(+), 20 deletions(-) diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c index 4924388..58a8408 100644 --- a/fs/btrfs/raid56.c +++ b/fs/btrfs/raid56.c @@ -72,6 +72,7 @@ enum btrfs_rbio_ops { BTRFS_RBIO_WRITE = 0, BTRFS_RBIO_READ_REBUILD = 1, + BTRFS_RBIO_PARITY_SCRUB = 2, }; struct btrfs_raid_bio { @@ -130,6 +131,7 @@ struct btrfs_raid_bio { /* number of data stripes (no p/q) */ int nr_data; + int stripe_npages; /* * set if we're doing a parity rebuild * for a read from higher up, which is handled @@ -144,6 +146,7 @@ struct btrfs_raid_bio { /* second bad stripe (for raid6 use) */ int failb; + int scrubp; /* * number of pages needed to represent the full * stripe @@ -178,6 +181,11 @@ struct btrfs_raid_bio { * here for faster lookup */ struct page **bio_pages; + + /* + * bitmap to record which horizontal stripe has data + */ + unsigned long *dbitmap; }; static int __raid56_parity_recover(struct btrfs_raid_bio *rbio); @@ -192,6 +200,10 @@ static void __free_raid_bio(struct btrfs_raid_bio *rbio); static void index_rbio_pages(struct btrfs_raid_bio *rbio); static int alloc_rbio_pages(struct btrfs_raid_bio *rbio); +static noinline void finish_parity_scrub(struct btrfs_raid_bio *rbio, + int need_check); +static void async_scrub_parity(struct btrfs_raid_bio *rbio); + /* * the stripe hash table is used for locking, and to collect * bios in hopes of making a full stripe @@ -593,10 +605,20 @@ static int rbio_can_merge(struct btrfs_raid_bio *last, cur->raid_map[0]) return 0; - /* reads can't merge with writes */ - if (last->operation != cur->operation) { + /* we can't merge with different operations */ + if (last->operation != cur->operation) + return 0; + /* + * We've need read the full stripe from the drive. + * check and repair the parity and write the new results. + * + * We're not allowed to add any new bios to the + * bio list here, anyone else that wants to + * change this stripe needs to do their own rmw. + */ + if (last->operation == BTRFS_RBIO_PARITY_SCRUB || + cur->operation == BTRFS_RBIO_PARITY_SCRUB) return 0; - } return 1; } @@ -789,9 +811,12 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio) if (next->operation == BTRFS_RBIO_READ_REBUILD) async_read_rebuild(next); - else if (next->operation == BTRFS_RBIO_WRITE){ + else if (next->operation == BTRFS_RBIO_WRITE) { steal_rbio(rbio, next); async_rmw_stripe(next); + } else if (next->operation == BTRFS_RBIO_PARITY_SCRUB) { + steal_rbio(rbio, next); + async_scrub_parity(next); } goto done_nolock; @@ -957,9 +982,11 @@ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_root *root, struct btrfs_raid_bio *rbio; int nr_data = 0; int num_pages = rbio_nr_pages(stripe_len, bbio->num_stripes); + int stripe_npages = DIV_ROUND_UP(stripe_len, PAGE_SIZE); void *p; - rbio = kzalloc(sizeof(*rbio) + num_pages * sizeof(struct page *) * 2, + rbio = kzalloc(sizeof(*rbio) + num_pages * sizeof(struct page *) * 2 + + DIV_ROUND_UP(stripe_npages, BITS_PER_LONG / 8), GFP_NOFS); if (!rbio) return ERR_PTR(-ENOMEM); @@ -974,6 +1001,7 @@ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_root *root, rbio->fs_info = root->fs_info; rbio->stripe_len = stripe_len; rbio->nr_pages = num_pages; + rbio->stripe_npages = stripe_npages; rbio->faila = -1; rbio->failb = -1; atomic_set(&rbio->refs, 1); @@ -987,6 +1015,7 @@ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_root *root, p = rbio + 1; rbio->stripe_pages = p; rbio->bio_pages = p + sizeof(struct page *) * num_pages; + rbio->dbitmap = p + sizeof(struct page *) * num_pages * 2; if (raid_map[bbio->num_stripes - 1] == RAID6_Q_STRIPE) nr_data = bbio->num_stripes - 2; @@ -1781,6 +1810,14 @@ static void __raid_recover_end_io(struct btrfs_raid_bio *rbio) index_rbio_pages(rbio); for (pagenr = 0; pagenr < nr_pages; pagenr++) { + /* + * Now we just use bitmap to mark the horizontal stripes in + * which we have data when doing parity scrub. + */ + if (rbio->operation == BTRFS_RBIO_PARITY_SCRUB && + !test_bit(pagenr, rbio->dbitmap)) + continue; + /* setup our array of pointers with pages * from each stripe */ @@ -1925,7 +1962,13 @@ cleanup_io: } else if (err == 0) { rbio->faila = -1; rbio->failb = -1; - finish_rmw(rbio); + + if (rbio->operation == BTRFS_RBIO_WRITE) + finish_rmw(rbio); + else if (rbio->operation == BTRFS_RBIO_PARITY_SCRUB) + finish_parity_scrub(rbio, 0); + else + BUG(); } else { rbio_orig_end_io(rbio, err, 0); } @@ -2133,3 +2176,462 @@ static void read_rebuild_work(struct btrfs_work *work) rbio = container_of(work, struct btrfs_raid_bio, work); __raid56_parity_recover(rbio); } + +/* + * The following code is used to scrub/replace the parity stripe + * + * Note: We need make sure all the pages that add into the scrub/replace + * raid bio are correct and not be changed during the scrub/replace. That + * is those pages just hold metadata or file data with checksum. + */ + +struct btrfs_raid_bio * +raid56_parity_alloc_scrub_rbio(struct btrfs_root *root, struct bio *bio, + struct btrfs_bio *bbio, u64 *raid_map, + u64 stripe_len, struct btrfs_device *scrub_dev, + unsigned long *dbitmap, int stripe_nsectors) +{ + struct btrfs_raid_bio *rbio; + int i; + + rbio = alloc_rbio(root, bbio, raid_map, stripe_len); + if (IS_ERR(rbio)) + return NULL; + bio_list_add(&rbio->bio_list, bio); + /* + * This is a special bio which is used to hold the completion handler + * and make the scrub rbio is similar to the other types + */ + ASSERT(!bio->bi_iter.bi_size); + rbio->operation = BTRFS_RBIO_PARITY_SCRUB; + + for (i = 0; i < bbio->num_stripes; i++) { + if (bbio->stripes[i].dev == scrub_dev) { + rbio->scrubp = i; + break; + } + } + + /* Now we just support the sectorsize equals to page size */ + ASSERT(root->sectorsize == PAGE_SIZE); + ASSERT(rbio->stripe_npages == stripe_nsectors); + bitmap_copy(rbio->dbitmap, dbitmap, stripe_nsectors); + + return rbio; +} + +void raid56_parity_add_scrub_pages(struct btrfs_raid_bio *rbio, + struct page *page, u64 logical) +{ + int stripe_offset; + int index; + + ASSERT(logical >= rbio->raid_map[0]); + ASSERT(logical + PAGE_SIZE <= rbio->raid_map[0] + + rbio->stripe_len * rbio->nr_data); + stripe_offset = (int)(logical - rbio->raid_map[0]); + index = stripe_offset >> PAGE_CACHE_SHIFT; + rbio->bio_pages[index] = page; +} + +/* + * We just scrub the parity that we have correct data on the same horizontal, + * so we needn't allocate all pages for all the stripes. + */ +static int alloc_rbio_essential_pages(struct btrfs_raid_bio *rbio) +{ + int i; + int bit; + int index; + struct page *page; + + for_each_set_bit(bit, rbio->dbitmap, rbio->stripe_npages) { + for (i = 0; i < rbio->bbio->num_stripes; i++) { + index = i * rbio->stripe_npages + bit; + if (rbio->stripe_pages[index]) + continue; + + page = alloc_page(GFP_NOFS | __GFP_HIGHMEM); + if (!page) + return -ENOMEM; + rbio->stripe_pages[index] = page; + ClearPageUptodate(page); + } + } + return 0; +} + +/* + * end io function used by finish_rmw. When we finally + * get here, we've written a full stripe + */ +static void raid_write_parity_end_io(struct bio *bio, int err) +{ + struct btrfs_raid_bio *rbio = bio->bi_private; + + if (err) + fail_bio_stripe(rbio, bio); + + bio_put(bio); + + if (!atomic_dec_and_test(&rbio->stripes_pending)) + return; + + err = 0; + + if (atomic_read(&rbio->error)) + err = -EIO; + + rbio_orig_end_io(rbio, err, 0); +} + +static noinline void finish_parity_scrub(struct btrfs_raid_bio *rbio, + int need_check) +{ + struct btrfs_bio *bbio = rbio->bbio; + void *pointers[bbio->num_stripes]; + int nr_data = rbio->nr_data; + int stripe; + int pagenr; + int p_stripe = -1; + int q_stripe = -1; + struct page *p_page = NULL; + struct page *q_page = NULL; + struct bio_list bio_list; + struct bio *bio; + int ret; + + bio_list_init(&bio_list); + + if (bbio->num_stripes - rbio->nr_data == 1) { + p_stripe = bbio->num_stripes - 1; + } else if (bbio->num_stripes - rbio->nr_data == 2) { + p_stripe = bbio->num_stripes - 2; + q_stripe = bbio->num_stripes - 1; + } else { + BUG(); + } + + /* + * Because the higher layers(scrubber) are unlikely to + * use this area of the disk again soon, so don't cache + * it. + */ + clear_bit(RBIO_CACHE_READY_BIT, &rbio->flags); + + if (!need_check) + goto writeback; + + p_page = alloc_page(GFP_NOFS | __GFP_HIGHMEM); + if (!p_page) + goto cleanup; + SetPageUptodate(p_page); + + if (q_stripe != -1) { + q_page = alloc_page(GFP_NOFS | __GFP_HIGHMEM); + if (!q_page) { + __free_page(p_page); + goto cleanup; + } + SetPageUptodate(q_page); + } + + atomic_set(&rbio->error, 0); + + for_each_set_bit(pagenr, rbio->dbitmap, rbio->stripe_npages) { + struct page *p; + void *parity; + /* first collect one page from each data stripe */ + for (stripe = 0; stripe < nr_data; stripe++) { + p = page_in_rbio(rbio, stripe, pagenr, 0); + pointers[stripe] = kmap(p); + } + + /* then add the parity stripe */ + pointers[stripe++] = kmap(p_page); + + if (q_stripe != -1) { + + /* + * raid6, add the qstripe and call the + * library function to fill in our p/q + */ + pointers[stripe++] = kmap(q_page); + + raid6_call.gen_syndrome(bbio->num_stripes, PAGE_SIZE, + pointers); + } else { + /* raid5 */ + memcpy(pointers[nr_data], pointers[0], PAGE_SIZE); + run_xor(pointers + 1, nr_data - 1, PAGE_CACHE_SIZE); + } + + /* Check scrubbing pairty and repair it */ + p = rbio_stripe_page(rbio, rbio->scrubp, pagenr); + parity = kmap(p); + if (memcmp(parity, pointers[rbio->scrubp], PAGE_CACHE_SIZE)) + memcpy(parity, pointers[rbio->scrubp], PAGE_CACHE_SIZE); + else + /* Parity is right, needn't writeback */ + bitmap_clear(rbio->dbitmap, pagenr, 1); + kunmap(p); + + for (stripe = 0; stripe < bbio->num_stripes; stripe++) + kunmap(page_in_rbio(rbio, stripe, pagenr, 0)); + } + + __free_page(p_page); + if (q_page) + __free_page(q_page); + +writeback: + /* + * time to start writing. Make bios for everything from the + * higher layers (the bio_list in our rbio) and our p/q. Ignore + * everything else. + */ + for_each_set_bit(pagenr, rbio->dbitmap, rbio->stripe_npages) { + struct page *page; + + page = rbio_stripe_page(rbio, rbio->scrubp, pagenr); + ret = rbio_add_io_page(rbio, &bio_list, + page, rbio->scrubp, pagenr, rbio->stripe_len); + if (ret) + goto cleanup; + } + + nr_data = bio_list_size(&bio_list); + if (!nr_data) { + /* Every parity is right */ + rbio_orig_end_io(rbio, 0, 0); + return; + } + + atomic_set(&rbio->stripes_pending, nr_data); + + while (1) { + bio = bio_list_pop(&bio_list); + if (!bio) + break; + + bio->bi_private = rbio; + bio->bi_end_io = raid_write_parity_end_io; + BUG_ON(!test_bit(BIO_UPTODATE, &bio->bi_flags)); + submit_bio(WRITE, bio); + } + return; + +cleanup: + rbio_orig_end_io(rbio, -EIO, 0); +} + +static inline int is_data_stripe(struct btrfs_raid_bio *rbio, int stripe) +{ + if (stripe >= 0 && stripe < rbio->nr_data) + return 1; + return 0; +} + +/* + * While we're doing the parity check and repair, we could have errors + * in reading pages off the disk. This checks for errors and if we're + * not able to read the page it'll trigger parity reconstruction. The + * parity scrub will be finished after we've reconstructed the failed + * stripes + */ +static void validate_rbio_for_parity_scrub(struct btrfs_raid_bio *rbio) +{ + if (atomic_read(&rbio->error) > rbio->bbio->max_errors) + goto cleanup; + + if (rbio->faila >= 0 || rbio->failb >= 0) { + int dfail = 0, failp = -1; + + if (is_data_stripe(rbio, rbio->faila)) + dfail++; + else if (is_parity_stripe(rbio->faila)) + failp = rbio->faila; + + if (is_data_stripe(rbio, rbio->failb)) + dfail++; + else if (is_parity_stripe(rbio->failb)) + failp = rbio->failb; + + /* + * Because we can not use a scrubbing parity to repair + * the data, so the capability of the repair is declined. + * (In the case of RAID5, we can not repair anything) + */ + if (dfail > rbio->bbio->max_errors - 1) + goto cleanup; + + /* + * If all data is good, only parity is correctly, just + * repair the parity. + */ + if (dfail == 0) { + finish_parity_scrub(rbio, 0); + return; + } + + /* + * Here means we got one corrupted data stripe and one + * corrupted parity on RAID6, if the corrupted parity + * is scrubbing parity, luckly, use the other one to repair + * the data, or we can not repair the data stripe. + */ + if (failp != rbio->scrubp) + goto cleanup; + + __raid_recover_end_io(rbio); + } else { + finish_parity_scrub(rbio, 1); + } + return; + +cleanup: + rbio_orig_end_io(rbio, -EIO, 0); +} + +/* + * end io for the read phase of the rmw cycle. All the bios here are physical + * stripe bios we've read from the disk so we can recalculate the parity of the + * stripe. + * + * This will usually kick off finish_rmw once all the bios are read in, but it + * may trigger parity reconstruction if we had any errors along the way + */ +static void raid56_parity_scrub_end_io(struct bio *bio, int err) +{ + struct btrfs_raid_bio *rbio = bio->bi_private; + + if (err) + fail_bio_stripe(rbio, bio); + else + set_bio_pages_uptodate(bio); + + bio_put(bio); + + if (!atomic_dec_and_test(&rbio->stripes_pending)) + return; + + /* + * this will normally call finish_rmw to start our write + * but if there are any failed stripes we'll reconstruct + * from parity first + */ + validate_rbio_for_parity_scrub(rbio); +} + +static void raid56_parity_scrub_stripe(struct btrfs_raid_bio *rbio) +{ + int bios_to_read = 0; + struct btrfs_bio *bbio = rbio->bbio; + struct bio_list bio_list; + int ret; + int pagenr; + int stripe; + struct bio *bio; + + ret = alloc_rbio_essential_pages(rbio); + if (ret) + goto cleanup; + + bio_list_init(&bio_list); + + atomic_set(&rbio->error, 0); + /* + * build a list of bios to read all the missing parts of this + * stripe + */ + for (stripe = 0; stripe < bbio->num_stripes; stripe++) { + for_each_set_bit(pagenr, rbio->dbitmap, rbio->stripe_npages) { + struct page *page; + /* + * we want to find all the pages missing from + * the rbio and read them from the disk. If + * page_in_rbio finds a page in the bio list + * we don't need to read it off the stripe. + */ + page = page_in_rbio(rbio, stripe, pagenr, 1); + if (page) + continue; + + page = rbio_stripe_page(rbio, stripe, pagenr); + /* + * the bio cache may have handed us an uptodate + * page. If so, be happy and use it + */ + if (PageUptodate(page)) + continue; + + ret = rbio_add_io_page(rbio, &bio_list, page, + stripe, pagenr, rbio->stripe_len); + if (ret) + goto cleanup; + } + } + + bios_to_read = bio_list_size(&bio_list); + if (!bios_to_read) { + /* + * this can happen if others have merged with + * us, it means there is nothing left to read. + * But if there are missing devices it may not be + * safe to do the full stripe write yet. + */ + goto finish; + } + + /* + * the bbio may be freed once we submit the last bio. Make sure + * not to touch it after that + */ + atomic_set(&rbio->stripes_pending, bios_to_read); + while (1) { + bio = bio_list_pop(&bio_list); + if (!bio) + break; + + bio->bi_private = rbio; + bio->bi_end_io = raid56_parity_scrub_end_io; + + btrfs_bio_wq_end_io(rbio->fs_info, bio, + BTRFS_WQ_ENDIO_RAID56); + + BUG_ON(!test_bit(BIO_UPTODATE, &bio->bi_flags)); + submit_bio(READ, bio); + } + /* the actual write will happen once the reads are done */ + return; + +cleanup: + rbio_orig_end_io(rbio, -EIO, 0); + return; + +finish: + validate_rbio_for_parity_scrub(rbio); +} + +static void scrub_parity_work(struct btrfs_work *work) +{ + struct btrfs_raid_bio *rbio; + + rbio = container_of(work, struct btrfs_raid_bio, work); + raid56_parity_scrub_stripe(rbio); +} + +static void async_scrub_parity(struct btrfs_raid_bio *rbio) +{ + btrfs_init_work(&rbio->work, btrfs_rmw_helper, + scrub_parity_work, NULL, NULL); + + btrfs_queue_work(rbio->fs_info->rmw_workers, + &rbio->work); +} + +void raid56_parity_submit_scrub_rbio(struct btrfs_raid_bio *rbio) +{ + if (!lock_stripe_add(rbio)) + async_scrub_parity(rbio); +} diff --git a/fs/btrfs/raid56.h b/fs/btrfs/raid56.h index b310e8c..3d4ddb3 100644 --- a/fs/btrfs/raid56.h +++ b/fs/btrfs/raid56.h @@ -39,6 +39,9 @@ static inline int nr_data_stripes(struct map_lookup *map) #define is_parity_stripe(x) (((x) == RAID5_P_STRIPE) || \ ((x) == RAID6_Q_STRIPE)) +struct btrfs_raid_bio; +struct btrfs_device; + int raid56_parity_recover(struct btrfs_root *root, struct bio *bio, struct btrfs_bio *bbio, u64 *raid_map, u64 stripe_len, int mirror_num, int hold_bbio); @@ -46,6 +49,15 @@ int raid56_parity_write(struct btrfs_root *root, struct bio *bio, struct btrfs_bio *bbio, u64 *raid_map, u64 stripe_len); +struct btrfs_raid_bio * +raid56_parity_alloc_scrub_rbio(struct btrfs_root *root, struct bio *bio, + struct btrfs_bio *bbio, u64 *raid_map, + u64 stripe_len, struct btrfs_device *scrub_dev, + unsigned long *dbitmap, int stripe_nsectors); +void raid56_parity_add_scrub_pages(struct btrfs_raid_bio *rbio, + struct page *page, u64 logical); +void raid56_parity_submit_scrub_rbio(struct btrfs_raid_bio *rbio); + int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info); void btrfs_free_stripe_hash_table(struct btrfs_fs_info *info); #endif diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index 738ccff..e3f0b0f 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -74,6 +74,7 @@ struct scrub_page { struct scrub_block *sblock; struct page *page; struct btrfs_device *dev; + struct list_head list; u64 flags; /* extent flags */ u64 generation; u64 logical; @@ -114,14 +115,52 @@ struct scrub_block { atomic_t outstanding_pages; atomic_t ref_count; /* free mem on transition to zero */ struct scrub_ctx *sctx; + struct scrub_parity *sparity; struct { unsigned int header_error:1; unsigned int checksum_error:1; unsigned int no_io_error_seen:1; unsigned int generation_error:1; /* also sets header_error */ + + /* The following is for the data used to check parity */ + /* It is for the data with checksum */ + unsigned int data_corrected:1; }; }; +/* Used for the chunks with parity stripe such RAID5/6 */ +struct scrub_parity { + struct scrub_ctx *sctx; + + struct btrfs_device *scrub_dev; + + u64 logic_start; + + u64 logic_end; + + int nsectors; + + int stripe_len; + + atomic_t ref_count; + + struct list_head spages; + + /* Work of parity check and repair */ + struct btrfs_work work; + + /* Mark the parity blocks which have data */ + unsigned long *dbitmap; + + /* + * Mark the parity blocks which have data, but errors happen when + * read data or check data + */ + unsigned long *ebitmap; + + unsigned long bitmap[0]; +}; + struct scrub_wr_ctx { struct scrub_bio *wr_curr_bio; struct btrfs_device *tgtdev; @@ -227,6 +266,8 @@ static void scrub_block_get(struct scrub_block *sblock); static void scrub_block_put(struct scrub_block *sblock); static void scrub_page_get(struct scrub_page *spage); static void scrub_page_put(struct scrub_page *spage); +static void scrub_parity_get(struct scrub_parity *sparity); +static void scrub_parity_put(struct scrub_parity *sparity); static int scrub_add_page_to_rd_bio(struct scrub_ctx *sctx, struct scrub_page *spage); static int scrub_pages(struct scrub_ctx *sctx, u64 logical, u64 len, @@ -944,6 +985,7 @@ static int scrub_handle_errored_block(struct scrub_block *sblock_to_check) */ spin_lock(&sctx->stat_lock); sctx->stat.unverified_errors++; + sblock_to_check->data_corrected = 1; spin_unlock(&sctx->stat_lock); if (sctx->is_dev_replace) @@ -1204,6 +1246,7 @@ nodatasum_case: corrected_error: spin_lock(&sctx->stat_lock); sctx->stat.corrected_errors++; + sblock_to_check->data_corrected = 1; spin_unlock(&sctx->stat_lock); printk_ratelimited_in_rcu(KERN_ERR "BTRFS: fixed up error at logical %llu on dev %s\n", @@ -1645,6 +1688,13 @@ static void scrub_write_block_to_dev_replace(struct scrub_block *sblock) { int page_num; + /* + * This block is used for the check of the parity on the source device, + * so the data needn't be written into the destination device. + */ + if (sblock->sparity) + return; + for (page_num = 0; page_num < sblock->page_count; page_num++) { int ret; @@ -2026,6 +2076,9 @@ static void scrub_block_put(struct scrub_block *sblock) if (atomic_dec_and_test(&sblock->ref_count)) { int i; + if (sblock->sparity) + scrub_parity_put(sblock->sparity); + for (i = 0; i < sblock->page_count; i++) scrub_page_put(sblock->pagev[i]); kfree(sblock); @@ -2283,9 +2336,51 @@ static void scrub_bio_end_io_worker(struct btrfs_work *work) scrub_pending_bio_dec(sctx); } +static inline void __scrub_mark_bitmap(struct scrub_parity *sparity, + unsigned long *bitmap, + u64 start, u64 len) +{ + int offset; + int nsectors; + int sectorsize = sparity->sctx->dev_root->sectorsize; + + if (len >= sparity->stripe_len) { + bitmap_set(bitmap, 0, sparity->nsectors); + return; + } + + start -= sparity->logic_start; + offset = (int)do_div(start, sparity->stripe_len); + offset /= sectorsize; + nsectors = (int)len / sectorsize; + + if (offset + nsectors <= sparity->nsectors) { + bitmap_set(bitmap, offset, nsectors); + return; + } + + bitmap_set(bitmap, offset, sparity->nsectors - offset); + bitmap_set(bitmap, 0, nsectors - (sparity->nsectors - offset)); +} + +static inline void scrub_parity_mark_sectors_error(struct scrub_parity *sparity, + u64 start, u64 len) +{ + __scrub_mark_bitmap(sparity, sparity->ebitmap, start, len); +} + +static inline void scrub_parity_mark_sectors_data(struct scrub_parity *sparity, + u64 start, u64 len) +{ + __scrub_mark_bitmap(sparity, sparity->dbitmap, start, len); +} + static void scrub_block_complete(struct scrub_block *sblock) { + int corrupted = 0; + if (!sblock->no_io_error_seen) { + corrupted = 1; scrub_handle_errored_block(sblock); } else { /* @@ -2293,9 +2388,19 @@ static void scrub_block_complete(struct scrub_block *sblock) * dev replace case, otherwise write here in dev replace * case. */ - if (!scrub_checksum(sblock) && sblock->sctx->is_dev_replace) + corrupted = scrub_checksum(sblock); + if (!corrupted && sblock->sctx->is_dev_replace) scrub_write_block_to_dev_replace(sblock); } + + if (sblock->sparity && corrupted && !sblock->data_corrected) { + u64 start = sblock->pagev[0]->logical; + u64 end = sblock->pagev[sblock->page_count - 1]->logical + + PAGE_SIZE; + + scrub_parity_mark_sectors_error(sblock->sparity, + start, end - start); + } } static int scrub_find_csum(struct scrub_ctx *sctx, u64 logical, u64 len, @@ -2387,6 +2492,132 @@ behind_scrub_pages: return 0; } +static int scrub_pages_for_parity(struct scrub_parity *sparity, + u64 logical, u64 len, + u64 physical, struct btrfs_device *dev, + u64 flags, u64 gen, int mirror_num, u8 *csum) +{ + struct scrub_ctx *sctx = sparity->sctx; + struct scrub_block *sblock; + int index; + + sblock = kzalloc(sizeof(*sblock), GFP_NOFS); + if (!sblock) { + spin_lock(&sctx->stat_lock); + sctx->stat.malloc_errors++; + spin_unlock(&sctx->stat_lock); + return -ENOMEM; + } + + /* one ref inside this function, plus one for each page added to + * a bio later on */ + atomic_set(&sblock->ref_count, 1); + sblock->sctx = sctx; + sblock->no_io_error_seen = 1; + sblock->sparity = sparity; + scrub_parity_get(sparity); + + for (index = 0; len > 0; index++) { + struct scrub_page *spage; + u64 l = min_t(u64, len, PAGE_SIZE); + + spage = kzalloc(sizeof(*spage), GFP_NOFS); + if (!spage) { +leave_nomem: + spin_lock(&sctx->stat_lock); + sctx->stat.malloc_errors++; + spin_unlock(&sctx->stat_lock); + scrub_block_put(sblock); + return -ENOMEM; + } + BUG_ON(index >= SCRUB_MAX_PAGES_PER_BLOCK); + /* For scrub block */ + scrub_page_get(spage); + sblock->pagev[index] = spage; + /* For scrub parity */ + scrub_page_get(spage); + list_add_tail(&spage->list, &sparity->spages); + spage->sblock = sblock; + spage->dev = dev; + spage->flags = flags; + spage->generation = gen; + spage->logical = logical; + spage->physical = physical; + spage->mirror_num = mirror_num; + if (csum) { + spage->have_csum = 1; + memcpy(spage->csum, csum, sctx->csum_size); + } else { + spage->have_csum = 0; + } + sblock->page_count++; + spage->page = alloc_page(GFP_NOFS); + if (!spage->page) + goto leave_nomem; + len -= l; + logical += l; + physical += l; + } + + WARN_ON(sblock->page_count == 0); + for (index = 0; index < sblock->page_count; index++) { + struct scrub_page *spage = sblock->pagev[index]; + int ret; + + ret = scrub_add_page_to_rd_bio(sctx, spage); + if (ret) { + scrub_block_put(sblock); + return ret; + } + } + + /* last one frees, either here or in bio completion for last page */ + scrub_block_put(sblock); + return 0; +} + +static int scrub_extent_for_parity(struct scrub_parity *sparity, + u64 logical, u64 len, + u64 physical, struct btrfs_device *dev, + u64 flags, u64 gen, int mirror_num) +{ + struct scrub_ctx *sctx = sparity->sctx; + int ret; + u8 csum[BTRFS_CSUM_SIZE]; + u32 blocksize; + + if (flags & BTRFS_EXTENT_FLAG_DATA) { + blocksize = sctx->sectorsize; + } else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { + blocksize = sctx->nodesize; + } else { + blocksize = sctx->sectorsize; + WARN_ON(1); + } + + while (len) { + u64 l = min_t(u64, len, blocksize); + int have_csum = 0; + + if (flags & BTRFS_EXTENT_FLAG_DATA) { + /* push csums to sbio */ + have_csum = scrub_find_csum(sctx, logical, l, csum); + if (have_csum == 0) + goto skip; + } + ret = scrub_pages_for_parity(sparity, logical, l, physical, dev, + flags, gen, mirror_num, + have_csum ? csum : NULL); +skip: + if (ret) + return ret; + len -= l; + logical += l; + physical += l; + } + return 0; +} + /* * Given a physical address, this will calculate it's * logical offset. if this is a parity stripe, it will return @@ -2395,7 +2626,8 @@ behind_scrub_pages: * return 0 if it is a data stripe, 1 means parity stripe. */ static int get_raid56_logic_offset(u64 physical, int num, - struct map_lookup *map, u64 *offset) + struct map_lookup *map, u64 *offset, + u64 *stripe_start) { int i; int j = 0; @@ -2406,6 +2638,9 @@ static int get_raid56_logic_offset(u64 physical, int num, last_offset = (physical - map->stripes[num].physical) * nr_data_stripes(map); + if (stripe_start) + *stripe_start = last_offset; + *offset = last_offset; for (i = 0; i < nr_data_stripes(map); i++) { *offset = last_offset + i * map->stripe_len; @@ -2428,13 +2663,330 @@ static int get_raid56_logic_offset(u64 physical, int num, return 1; } +static void scrub_free_parity(struct scrub_parity *sparity) +{ + struct scrub_ctx *sctx = sparity->sctx; + struct scrub_page *curr, *next; + int nbits; + + nbits = bitmap_weight(sparity->ebitmap, sparity->nsectors); + if (nbits) { + spin_lock(&sctx->stat_lock); + sctx->stat.read_errors += nbits; + sctx->stat.uncorrectable_errors += nbits; + spin_unlock(&sctx->stat_lock); + } + + list_for_each_entry_safe(curr, next, &sparity->spages, list) { + list_del_init(&curr->list); + scrub_page_put(curr); + } + + kfree(sparity); +} + +static void scrub_parity_bio_endio(struct bio *bio, int error) +{ + struct scrub_parity *sparity = (struct scrub_parity *)bio->bi_private; + struct scrub_ctx *sctx = sparity->sctx; + + if (error) + bitmap_or(sparity->ebitmap, sparity->ebitmap, sparity->dbitmap, + sparity->nsectors); + + scrub_free_parity(sparity); + scrub_pending_bio_dec(sctx); + bio_put(bio); +} + +static void scrub_parity_check_and_repair(struct scrub_parity *sparity) +{ + struct scrub_ctx *sctx = sparity->sctx; + struct bio *bio; + struct btrfs_raid_bio *rbio; + struct scrub_page *spage; + struct btrfs_bio *bbio = NULL; + u64 *raid_map = NULL; + u64 length; + int ret; + + if (!bitmap_andnot(sparity->dbitmap, sparity->dbitmap, sparity->ebitmap, + sparity->nsectors)) + goto out; + + length = sparity->logic_end - sparity->logic_start + 1; + ret = btrfs_map_sblock(sctx->dev_root->fs_info, REQ_GET_READ_MIRRORS, + sparity->logic_start, + &length, &bbio, 0, &raid_map); + if (ret || !bbio || !raid_map) + goto bbio_out; + + bio = btrfs_io_bio_alloc(GFP_NOFS, 0); + if (!bio) + goto bbio_out; + + bio->bi_iter.bi_sector = sparity->logic_start >> 9; + bio->bi_private = sparity; + bio->bi_end_io = scrub_parity_bio_endio; + + rbio = raid56_parity_alloc_scrub_rbio(sctx->dev_root, bio, bbio, + raid_map, length, + sparity->scrub_dev, + sparity->dbitmap, + sparity->nsectors); + if (!rbio) + goto rbio_out; + + list_for_each_entry(spage, &sparity->spages, list) + raid56_parity_add_scrub_pages(rbio, spage->page, + spage->logical); + + scrub_pending_bio_inc(sctx); + raid56_parity_submit_scrub_rbio(rbio); + return; + +rbio_out: + bio_put(bio); +bbio_out: + kfree(bbio); + kfree(raid_map); + bitmap_or(sparity->ebitmap, sparity->ebitmap, sparity->dbitmap, + sparity->nsectors); + spin_lock(&sctx->stat_lock); + sctx->stat.malloc_errors++; + spin_unlock(&sctx->stat_lock); +out: + scrub_free_parity(sparity); +} + +static inline int scrub_calc_parity_bitmap_len(int nsectors) +{ + return DIV_ROUND_UP(nsectors, BITS_PER_LONG) * (BITS_PER_LONG / 8); +} + +static void scrub_parity_get(struct scrub_parity *sparity) +{ + atomic_inc(&sparity->ref_count); +} + +static void scrub_parity_put(struct scrub_parity *sparity) +{ + if (!atomic_dec_and_test(&sparity->ref_count)) + return; + + scrub_parity_check_and_repair(sparity); +} + +static noinline_for_stack int scrub_raid56_parity(struct scrub_ctx *sctx, + struct map_lookup *map, + struct btrfs_device *sdev, + struct btrfs_path *path, + u64 logic_start, + u64 logic_end) +{ + struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; + struct btrfs_root *root = fs_info->extent_root; + struct btrfs_root *csum_root = fs_info->csum_root; + struct btrfs_extent_item *extent; + u64 flags; + int ret; + int slot; + struct extent_buffer *l; + struct btrfs_key key; + u64 generation; + u64 extent_logical; + u64 extent_physical; + u64 extent_len; + struct btrfs_device *extent_dev; + struct scrub_parity *sparity; + int nsectors; + int bitmap_len; + int extent_mirror_num; + int stop_loop = 0; + + nsectors = map->stripe_len / root->sectorsize; + bitmap_len = scrub_calc_parity_bitmap_len(nsectors); + sparity = kzalloc(sizeof(struct scrub_parity) + 2 * bitmap_len, + GFP_NOFS); + if (!sparity) { + spin_lock(&sctx->stat_lock); + sctx->stat.malloc_errors++; + spin_unlock(&sctx->stat_lock); + return -ENOMEM; + } + + sparity->stripe_len = map->stripe_len; + sparity->nsectors = nsectors; + sparity->sctx = sctx; + sparity->scrub_dev = sdev; + sparity->logic_start = logic_start; + sparity->logic_end = logic_end; + atomic_set(&sparity->ref_count, 1); + INIT_LIST_HEAD(&sparity->spages); + sparity->dbitmap = sparity->bitmap; + sparity->ebitmap = (void *)sparity->bitmap + bitmap_len; + + ret = 0; + while (logic_start < logic_end) { + if (btrfs_fs_incompat(fs_info, SKINNY_METADATA)) + key.type = BTRFS_METADATA_ITEM_KEY; + else + key.type = BTRFS_EXTENT_ITEM_KEY; + key.objectid = logic_start; + key.offset = (u64)-1; + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) + goto out; + + if (ret > 0) { + ret = btrfs_previous_extent_item(root, path, 0); + if (ret < 0) + goto out; + if (ret > 0) { + btrfs_release_path(path); + ret = btrfs_search_slot(NULL, root, &key, + path, 0, 0); + if (ret < 0) + goto out; + } + } + + stop_loop = 0; + while (1) { + u64 bytes; + + l = path->nodes[0]; + slot = path->slots[0]; + if (slot >= btrfs_header_nritems(l)) { + ret = btrfs_next_leaf(root, path); + if (ret == 0) + continue; + if (ret < 0) + goto out; + + stop_loop = 1; + break; + } + btrfs_item_key_to_cpu(l, &key, slot); + + if (key.type == BTRFS_METADATA_ITEM_KEY) + bytes = root->nodesize; + else + bytes = key.offset; + + if (key.objectid + bytes <= logic_start) + goto next; + + if (key.type != BTRFS_EXTENT_ITEM_KEY && + key.type != BTRFS_METADATA_ITEM_KEY) + goto next; + + if (key.objectid > logic_end) { + stop_loop = 1; + break; + } + + while (key.objectid >= logic_start + map->stripe_len) + logic_start += map->stripe_len; + + extent = btrfs_item_ptr(l, slot, + struct btrfs_extent_item); + flags = btrfs_extent_flags(l, extent); + generation = btrfs_extent_generation(l, extent); + + if (key.objectid < logic_start && + (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)) { + btrfs_err(fs_info, + "scrub: tree block %llu spanning stripes, ignored. logical=%llu", + key.objectid, logic_start); + goto next; + } +again: + extent_logical = key.objectid; + extent_len = bytes; + + if (extent_logical < logic_start) { + extent_len -= logic_start - extent_logical; + extent_logical = logic_start; + } + + if (extent_logical + extent_len > + logic_start + map->stripe_len) + extent_len = logic_start + map->stripe_len - + extent_logical; + + scrub_parity_mark_sectors_data(sparity, extent_logical, + extent_len); + + scrub_remap_extent(fs_info, extent_logical, + extent_len, &extent_physical, + &extent_dev, + &extent_mirror_num); + + ret = btrfs_lookup_csums_range(csum_root, + extent_logical, + extent_logical + extent_len - 1, + &sctx->csum_list, 1); + if (ret) + goto out; + + ret = scrub_extent_for_parity(sparity, extent_logical, + extent_len, + extent_physical, + extent_dev, flags, + generation, + extent_mirror_num); + if (ret) + goto out; + + scrub_free_csums(sctx); + if (extent_logical + extent_len < + key.objectid + bytes) { + logic_start += map->stripe_len; + + if (logic_start >= logic_end) { + stop_loop = 1; + break; + } + + if (logic_start < key.objectid + bytes) { + cond_resched(); + goto again; + } + } +next: + path->slots[0]++; + } + + btrfs_release_path(path); + + if (stop_loop) + break; + + logic_start += map->stripe_len; + } +out: + if (ret < 0) + scrub_parity_mark_sectors_error(sparity, logic_start, + logic_end - logic_start + 1); + scrub_parity_put(sparity); + scrub_submit(sctx); + mutex_lock(&sctx->wr_ctx.wr_lock); + scrub_wr_submit(sctx); + mutex_unlock(&sctx->wr_ctx.wr_lock); + + btrfs_release_path(path); + return ret < 0 ? ret : 0; +} + static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, struct map_lookup *map, struct btrfs_device *scrub_dev, int num, u64 base, u64 length, int is_dev_replace) { - struct btrfs_path *path; + struct btrfs_path *path, *ppath; struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; struct btrfs_root *root = fs_info->extent_root; struct btrfs_root *csum_root = fs_info->csum_root; @@ -2461,6 +3013,8 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, u64 extent_logical; u64 extent_physical; u64 extent_len; + u64 stripe_logical; + u64 stripe_end; struct btrfs_device *extent_dev; int extent_mirror_num; int stop_loop = 0; @@ -2486,7 +3040,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, mirror_num = num % map->num_stripes + 1; } else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { - get_raid56_logic_offset(physical, num, map, &offset); + get_raid56_logic_offset(physical, num, map, &offset, NULL); increment = map->stripe_len * nr_data_stripes(map); mirror_num = 1; } else { @@ -2498,6 +3052,12 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, if (!path) return -ENOMEM; + ppath = btrfs_alloc_path(); + if (!ppath) { + btrfs_free_path(ppath); + return -ENOMEM; + } + /* * work on commit root. The related disk blocks are static as * long as COW is applied. This means, it is save to rewrite @@ -2516,7 +3076,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { get_raid56_logic_offset(physical_end, num, - map, &logic_end); + map, &logic_end, NULL); logic_end += base; } else { logic_end = logical + increment * nstripes; @@ -2563,10 +3123,18 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { ret = get_raid56_logic_offset(physical, num, - map, &logical); + map, &logical, &stripe_logical); logical += base; - if (ret) + if (ret) { + stripe_logical += base; + stripe_end = stripe_logical + increment - 1; + ret = scrub_raid56_parity(sctx, map, scrub_dev, + ppath, stripe_logical, + stripe_end); + if (ret) + goto out; goto skip; + } } /* * canceled? @@ -2717,13 +3285,25 @@ again: * loop until we find next data stripe * or we have finished all stripes. */ - do { - physical += map->stripe_len; - ret = get_raid56_logic_offset( - physical, num, - map, &logical); - logical += base; - } while (physical < physical_end && ret); +loop: + physical += map->stripe_len; + ret = get_raid56_logic_offset(physical, + num, map, &logical, + &stripe_logical); + logical += base; + + if (ret && physical < physical_end) { + stripe_logical += base; + stripe_end = stripe_logical + + increment - 1; + ret = scrub_raid56_parity(sctx, + map, scrub_dev, ppath, + stripe_logical, + stripe_end); + if (ret) + goto out; + goto loop; + } } else { physical += map->stripe_len; logical += increment; @@ -2764,6 +3344,7 @@ out: blk_finish_plug(&plug); btrfs_free_path(path); + btrfs_free_path(ppath); return ret < 0 ? ret : 0; }