From: Jiaju Zhang Date: Thu, 22 Jul 2010 00:05:42 +0800 Subject: [Ocfs2-devel] [PATCH V2] Fix the nested PR lock calling issue In-Reply-To: <20100721151701.GA3579@laptop.jp.oracle.com> References: <20100721144805.GA12822@linux-jjzhang> <20100721151701.GA3579@laptop.jp.oracle.com> Message-ID: <20100721160542.GA19036@linux-jjzhang> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ocfs2-devel@oss.oracle.com On Wed, Jul 21, 2010 at 11:17:01PM +0800, Wengang Wang wrote: > On 10-07-21 22:48, Jiaju Zhang wrote: > > Hi, > > > > This is an improved patch for the bug > > https://bugzilla.novell.com/show_bug.cgi?id=614332 > > (also http://oss.oracle.com/bugzilla/show_bug.cgi?id=1278) > > > > It is a nested PR lock calling issue, the referenced stack trace is > > as: > > smbd D ffff8800013d0600 0 9522 5608 0x00000000 > > ffff88002ca7fb18 0000000000000282 ffff88002f964500 ffff88002ca7fa98 > > ffff8800013d0600 ffff88002ca7fae0 ffff88002f964340 ffff88002f964340 > > ffff88002ca7ffd8 ffff88002ca7ffd8 ffff88002f964340 ffff88002f964340 > > Call Trace: > > [] schedule_timeout+0x175/0x210 > > [] wait_for_common+0xf0/0x210 > > [] __ocfs2_cluster_lock+0x3b9/0xa90 [ocfs2] > > [] ocfs2_inode_lock_full_nested+0x255/0xdb0 [ocfs2] > > [] ocfs2_get_acl+0x69/0x120 [ocfs2] > > [] ocfs2_check_acl+0x28/0x80 [ocfs2] > > [] acl_permission_check+0x57/0xb0 > > [] generic_permission+0x1d/0xc0 > > [] ocfs2_permission+0x10a/0x1d0 [ocfs2] > > [] inode_permission+0x45/0x100 > > [] sys_chdir+0x53/0x90 > > [] system_call_fastpath+0x16/0x1b > > [<00007f34a4ef6927>] 0x7f34a4ef6927 > > > > Many thanks for your review and comments;) > > > > Thanks, > > Jiaju > > > > Signed-off-by: > > Cc: Joel Becker > > Cc: Mark Fasheh > > Cc: Sunil Mushran > > Cc: Tao Ma > > Cc: Tiger Yang > > Cc: Wengang Wang > > --- > > fs/ocfs2/acl.c | 21 +++++++++++++++++---- > > 1 files changed, 17 insertions(+), 4 deletions(-) > > > > diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c > > index da70229..ecf73f4 100644 > > --- a/fs/ocfs2/acl.c > > +++ b/fs/ocfs2/acl.c > > @@ -290,17 +290,30 @@ static int ocfs2_set_acl(handle_t *handle, > > > > int ocfs2_check_acl(struct inode *inode, int mask) > > { > > - struct posix_acl *acl = ocfs2_get_acl(inode, ACL_TYPE_ACCESS); > > + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); > > + struct posix_acl *acl; > > + struct buffer_head *di_bh = NULL; > > + int ret = -EAGAIN; > > + > > + if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL)) > > + return ret; > > + > > + ret = ocfs2_read_inode_block(inode, &di_bh); > > + if (ret < 0) { > > + mlog_errno(ret); > > + return ret; > > + } > > + > > + acl = ocfs2_get_acl_nolock(inode, ACL_TYPE_ACCESS, di_bh); > > >From here on, you need to release di_bh. Oh yes, sorry, I should check the code more closely before sending it out;) Here is the new patch: Signed-off-by: Jiaju Zhang --- fs/ocfs2/acl.c | 23 +++++++++++++++++++---- 1 files changed, 19 insertions(+), 4 deletions(-) diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c index da70229..a1d25ca 100644 --- a/fs/ocfs2/acl.c +++ b/fs/ocfs2/acl.c @@ -290,17 +290,32 @@ static int ocfs2_set_acl(handle_t *handle, int ocfs2_check_acl(struct inode *inode, int mask) { - struct posix_acl *acl = ocfs2_get_acl(inode, ACL_TYPE_ACCESS); + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct posix_acl *acl; + struct buffer_head *di_bh = NULL; + int ret = -EAGAIN; + + if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL)) + return ret; + + ret = ocfs2_read_inode_block(inode, &di_bh); + if (ret < 0) { + mlog_errno(ret); + return ret; + } + + acl = ocfs2_get_acl_nolock(inode, ACL_TYPE_ACCESS, di_bh); + + brelse(di_bh); if (IS_ERR(acl)) return PTR_ERR(acl); if (acl) { - int ret = posix_acl_permission(inode, acl, mask); + ret = posix_acl_permission(inode, acl, mask); posix_acl_release(acl); - return ret; } - return -EAGAIN; + return ret; } int ocfs2_acl_chmod(struct inode *inode)