From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id DFD57EB64DA for ; Mon, 19 Jun 2023 11:00:17 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232164AbjFSLAQ (ORCPT ); Mon, 19 Jun 2023 07:00:16 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58272 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232502AbjFSK7Q (ORCPT ); Mon, 19 Jun 2023 06:59:16 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [IPv6:2604:1380:4641:c500::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7C31710F8 for ; Mon, 19 Jun 2023 03:58:21 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 0F1C960B5F for ; Mon, 19 Jun 2023 10:58:21 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 2663DC433C8; Mon, 19 Jun 2023 10:58:19 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1687172300; bh=MJ9SqAkxj+Uz/O9NmUpcMuzKZlxZQhNVDtS3Pv/kHuI=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=Z03BmcI8r3A8DsIvxz1RQaI6L9y/603PAGYc/S97ySL7eEKVXTXtj4IBk8O5kWwjT 1ILPM9maUntRMa48S6GIyXIcvmVl7cPEs04VzpHC7MMb6I7qOsEnUNxh6sm3JrvDE+ Eg6FTd5jh88vlOhWIzHsi8S9o1THtNVm6c4tBM1Y= From: Greg Kroah-Hartman To: stable@vger.kernel.org Cc: Greg Kroah-Hartman , patches@lists.linux.dev, Qu Wenruo , David Sterba , Sasha Levin Subject: [PATCH 5.15 019/107] btrfs: scrub: try harder to mark RAID56 block groups read-only Date: Mon, 19 Jun 2023 12:30:03 +0200 Message-ID: <20230619102142.445479856@linuxfoundation.org> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230619102141.541044823@linuxfoundation.org> References: <20230619102141.541044823@linuxfoundation.org> User-Agent: quilt/0.67 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: stable@vger.kernel.org From: Qu Wenruo [ Upstream commit 7561551e7ba870b9659083b95feb520fb2dacce3 ] Currently we allow a block group not to be marked read-only for scrub. But for RAID56 block groups if we require the block group to be read-only, then we're allowed to use cached content from scrub stripe to reduce unnecessary RAID56 reads. So this patch would: - Make btrfs_inc_block_group_ro() try harder During my tests, for cases like btrfs/061 and btrfs/064, we can hit ENOSPC from btrfs_inc_block_group_ro() calls during scrub. The reason is if we only have one single data chunk, and trying to scrub it, we won't have any space left for any newer data writes. But this check should be done by the caller, especially for scrub cases we only temporarily mark the chunk read-only. And newer data writes would always try to allocate a new data chunk when needed. - Return error for scrub if we failed to mark a RAID56 chunk read-only Signed-off-by: Qu Wenruo Signed-off-by: David Sterba Signed-off-by: Sasha Levin --- fs/btrfs/block-group.c | 14 ++++++++++++-- fs/btrfs/scrub.c | 9 ++++++++- 2 files changed, 20 insertions(+), 3 deletions(-) diff --git a/fs/btrfs/block-group.c b/fs/btrfs/block-group.c index 19f71c305b988..a76796f153d5f 100644 --- a/fs/btrfs/block-group.c +++ b/fs/btrfs/block-group.c @@ -2576,10 +2576,20 @@ int btrfs_inc_block_group_ro(struct btrfs_block_group *cache, } ret = inc_block_group_ro(cache, 0); - if (!do_chunk_alloc || ret == -ETXTBSY) - goto unlock_out; if (!ret) goto out; + if (ret == -ETXTBSY) + goto unlock_out; + + /* + * Skip chunk alloction if the bg is SYSTEM, this is to avoid system + * chunk allocation storm to exhaust the system chunk array. Otherwise + * we still want to try our best to mark the block group read-only. + */ + if (!do_chunk_alloc && ret == -ENOSPC && + (cache->flags & BTRFS_BLOCK_GROUP_SYSTEM)) + goto unlock_out; + alloc_flags = btrfs_get_alloc_profile(fs_info, cache->space_info->flags); ret = btrfs_chunk_alloc(trans, alloc_flags, CHUNK_ALLOC_FORCE); if (ret < 0) diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index ca8d6979c7887..0d1715ebdef9c 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -3812,13 +3812,20 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx, if (ret == 0) { ro_set = 1; - } else if (ret == -ENOSPC && !sctx->is_dev_replace) { + } else if (ret == -ENOSPC && !sctx->is_dev_replace && + !(cache->flags & BTRFS_BLOCK_GROUP_RAID56_MASK)) { /* * btrfs_inc_block_group_ro return -ENOSPC when it * failed in creating new chunk for metadata. * It is not a problem for scrub, because * metadata are always cowed, and our scrub paused * commit_transactions. + * + * For RAID56 chunks, we have to mark them read-only + * for scrub, as later we would use our own cache + * out of RAID56 realm. + * Thus we want the RAID56 bg to be marked RO to + * prevent RMW from screwing up out cache. */ ro_set = 0; } else if (ret == -ETXTBSY) { -- 2.39.2