From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750977AbdIITTO (ORCPT ); Sat, 9 Sep 2017 15:19:14 -0400 Received: from mail.kernel.org ([198.145.29.99]:53050 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750805AbdIITTN (ORCPT ); Sat, 9 Sep 2017 15:19:13 -0400 DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org DFCD021B67 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=kernel.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=jaegeuk@kernel.org Date: Sat, 9 Sep 2017 12:19:12 -0700 From: Jaegeuk Kim To: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-f2fs-devel@lists.sourceforge.net Cc: Jaegeuk Kim Subject: Re: [PATCH] f2fs: speed up gc_urgent mode with SSR and more discard commands Message-ID: <20170909191912.GA81339@jaegeuk-macbookpro.roam.corp.google.com> References: <20170909181730.78225-1-jaegeuk@kernel.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20170909181730.78225-1-jaegeuk@kernel.org> User-Agent: Mutt/1.8.2 (2017-04-18) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org It turns out it doesn't need to change discarding policy below. I submitted another patch to modify SSR mode only. Thanks, On 09/09, Jaegeuk Kim wrote: > From: Jaegeuk Kim > > This patch activates SSR and issuing more discard commands in gc_urgent mode. > > Signed-off-by: Jaegeuk Kim > --- > fs/f2fs/f2fs.h | 1 + > fs/f2fs/segment.c | 19 ++++++++++++++++++- > fs/f2fs/segment.h | 13 ------------- > 3 files changed, 19 insertions(+), 14 deletions(-) > > diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h > index 7a467fcd4bcd..3fbf55e41694 100644 > --- a/fs/f2fs/f2fs.h > +++ b/fs/f2fs/f2fs.h > @@ -2517,6 +2517,7 @@ void destroy_node_manager_caches(void); > /* > * segment.c > */ > +bool need_SSR(struct f2fs_sb_info *sbi); > void register_inmem_page(struct inode *inode, struct page *page); > void drop_inmem_pages(struct inode *inode); > void drop_inmem_page(struct inode *inode, struct page *page); > diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c > index 9e708e525ba8..d6e0c160a268 100644 > --- a/fs/f2fs/segment.c > +++ b/fs/f2fs/segment.c > @@ -168,6 +168,21 @@ static unsigned long __find_rev_next_zero_bit(const unsigned long *addr, > return result - size + __reverse_ffz(tmp); > } > > +bool need_SSR(struct f2fs_sb_info *sbi) > +{ > + int node_secs = get_blocktype_secs(sbi, F2FS_DIRTY_NODES); > + int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS); > + int imeta_secs = get_blocktype_secs(sbi, F2FS_DIRTY_IMETA); > + > + if (test_opt(sbi, LFS)) > + return false; > + if (sbi->gc_thread && sbi->gc_thread->gc_urgent) > + return true; > + > + return free_sections(sbi) <= (node_secs + 2 * dent_secs + imeta_secs + > + 2 * reserved_sections(sbi)); > +} > + > void register_inmem_page(struct inode *inode, struct page *page) > { > struct f2fs_inode_info *fi = F2FS_I(inode); > @@ -1203,6 +1218,7 @@ static int issue_discard_thread(void *data) > wait_queue_head_t *q = &dcc->discard_wait_queue; > unsigned int wait_ms = DEF_MIN_DISCARD_ISSUE_TIME; > int issued; > + bool urgent = false; > > set_freezable(); > > @@ -1218,13 +1234,14 @@ static int issue_discard_thread(void *data) > > if (dcc->discard_wake) { > dcc->discard_wake = 0; > + urgent = sbi->gc_thread && sbi->gc_thread->gc_urgent; > if (sbi->gc_thread && sbi->gc_thread->gc_urgent) > mark_discard_range_all(sbi); > } > > sb_start_intwrite(sbi->sb); > > - issued = __issue_discard_cmd(sbi, true); > + issued = __issue_discard_cmd(sbi, !urgent); > if (issued) { > __wait_discard_cmd(sbi, true); > wait_ms = DEF_MIN_DISCARD_ISSUE_TIME; > diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h > index 613b2fa7b1c1..e0a6cc23ace3 100644 > --- a/fs/f2fs/segment.h > +++ b/fs/f2fs/segment.h > @@ -497,19 +497,6 @@ static inline int reserved_sections(struct f2fs_sb_info *sbi) > return GET_SEC_FROM_SEG(sbi, (unsigned int)reserved_segments(sbi)); > } > > -static inline bool need_SSR(struct f2fs_sb_info *sbi) > -{ > - int node_secs = get_blocktype_secs(sbi, F2FS_DIRTY_NODES); > - int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS); > - int imeta_secs = get_blocktype_secs(sbi, F2FS_DIRTY_IMETA); > - > - if (test_opt(sbi, LFS)) > - return false; > - > - return free_sections(sbi) <= (node_secs + 2 * dent_secs + imeta_secs + > - 2 * reserved_sections(sbi)); > -} > - > static inline bool has_not_enough_free_secs(struct f2fs_sb_info *sbi, > int freed, int needed) > { > -- > 2.14.0.rc1.383.gd1ce394fe2-goog