From 6ca497a83e592d64e050c4d04b6dedb8c915f39a Mon Sep 17 00:00:00 2001 From: wengang wang Date: Fri, 6 Mar 2009 21:29:10 +0800 Subject: ocfs2: fix rare stale inode errors when exporting via nfs For nfs exporting, ocfs2_get_dentry() returns the dentry for fh. ocfs2_get_dentry() may read from disk when the inode is not in memory, without any cross cluster lock. this leads to the file system loading a stale inode. This patch fixes above problem. Solution is that in case of inode is not in memory, we get the cluster lock(PR) of alloc inode where the inode in question is allocated from (this causes node on which deletion is done sync the alloc inode) before reading out the inode itsself. then we check the bitmap in the group (the inode in question allcated from) to see if the bit is clear. if it's clear then it's stale. if the bit is set, we then check generation as the existing code does. We have to read out the inode in question from disk first to know its alloc slot and allot bit. And if its not stale we read it out using ocfs2_iget(). The second read should then be from cache. And also we have to add a per superblock nfs_sync_lock to cover the lock for alloc inode and that for inode in question. this is because ocfs2_get_dentry() and ocfs2_delete_inode() lock on them in reverse order. nfs_sync_lock is locked in EX mode in ocfs2_get_dentry() and in PR mode in ocfs2_delete_inode(). so that mutliple ocfs2_delete_inode() can run concurrently in normal case. [mfasheh@suse.com: build warning fixes and comment cleanups] Signed-off-by: Wengang Wang Acked-by: Joel Becker Signed-off-by: Mark Fasheh --- fs/ocfs2/export.c | 84 ++++++++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 77 insertions(+), 7 deletions(-) (limited to 'fs/ocfs2/export.c') diff --git a/fs/ocfs2/export.c b/fs/ocfs2/export.c index 2f27b332d8b3..de3da8eb558c 100644 --- a/fs/ocfs2/export.c +++ b/fs/ocfs2/export.c @@ -31,6 +31,7 @@ #include "ocfs2.h" +#include "alloc.h" #include "dir.h" #include "dlmglue.h" #include "dcache.h" @@ -38,6 +39,7 @@ #include "inode.h" #include "buffer_head_io.h" +#include "suballoc.h" struct ocfs2_inode_handle { @@ -49,29 +51,97 @@ static struct dentry *ocfs2_get_dentry(struct super_block *sb, struct ocfs2_inode_handle *handle) { struct inode *inode; + struct ocfs2_super *osb = OCFS2_SB(sb); + u64 blkno = handle->ih_blkno; + int status, set; struct dentry *result; mlog_entry("(0x%p, 0x%p)\n", sb, handle); - if (handle->ih_blkno == 0) { - mlog_errno(-ESTALE); - return ERR_PTR(-ESTALE); + if (blkno == 0) { + mlog(0, "nfs wants inode with blkno: 0\n"); + result = ERR_PTR(-ESTALE); + goto bail; + } + + inode = ocfs2_ilookup(sb, blkno); + /* + * If the inode exists in memory, we only need to check it's + * generation number + */ + if (inode) + goto check_gen; + + /* + * This will synchronize us against ocfs2_delete_inode() on + * all nodes + */ + status = ocfs2_nfs_sync_lock(osb, 1); + if (status < 0) { + mlog(ML_ERROR, "getting nfs sync lock(EX) failed %d\n", status); + goto check_err; + } + + status = ocfs2_test_inode_bit(osb, blkno, &set); + if (status < 0) { + if (status == -EINVAL) { + /* + * The blkno NFS gave us doesn't even show up + * as an inode, we return -ESTALE to be + * nice + */ + mlog(0, "test inode bit failed %d\n", status); + status = -ESTALE; + } else { + mlog(ML_ERROR, "test inode bit failed %d\n", status); + } + goto unlock_nfs_sync; + } + + /* If the inode allocator bit is clear, this inode must be stale */ + if (!set) { + mlog(0, "inode %llu suballoc bit is clear\n", blkno); + status = -ESTALE; + goto unlock_nfs_sync; } - inode = ocfs2_iget(OCFS2_SB(sb), handle->ih_blkno, 0, 0); + inode = ocfs2_iget(osb, blkno, 0, 0); - if (IS_ERR(inode)) - return (void *)inode; +unlock_nfs_sync: + ocfs2_nfs_sync_unlock(osb, 1); +check_err: + if (status < 0) { + if (status == -ESTALE) { + mlog(0, "stale inode ino: %llu generation: %u\n", + blkno, handle->ih_generation); + } + result = ERR_PTR(status); + goto bail; + } + + if (IS_ERR(inode)) { + mlog_errno(PTR_ERR(inode)); + result = (void *)inode; + goto bail; + } + +check_gen: if (handle->ih_generation != inode->i_generation) { iput(inode); - return ERR_PTR(-ESTALE); + mlog(0, "stale inode ino: %llu generation: %u\n", blkno, + handle->ih_generation); + result = ERR_PTR(-ESTALE); + goto bail; } result = d_obtain_alias(inode); if (!IS_ERR(result)) result->d_op = &ocfs2_dentry_ops; + else + mlog_errno(PTR_ERR(result)); +bail: mlog_exit_ptr(result); return result; } -- cgit v1.2.2