| Message ID | 1372784452-2403-1-git-send-email-jlayton@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Tue, Jul 02, 2013 at 01:00:52PM -0400, Jeff Layton wrote: > Currently the way these get set is a little convoluted. If the dentry is > allocated via lookup from userland, then it gets set by simple_lookup. > If it gets allocated when the kernel is populating the directory, then > it gets set via __rpc_lookup_create_exclusive, which has to check > whether they might already be set. Between both of these, this ensures > that all dentries have their d_op pointer set. > > Instead of doing that, just have them set at d_alloc time by pointing > sb->s_d_op at them. With that change, we no longer want the lookup op > to set them, so we must move to using our own lookup routine. There's a better solution - just make simple_lookup() skip d_set_d_op() if superblock already has ->s_d_op (and thus d_alloc() has already set the damn thing). Voila - we can just set ->s_d_op and leave inode_operations as is. -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Sun, 14 Jul 2013 15:00:11 +0100 Al Viro <viro@ZenIV.linux.org.uk> wrote: > On Tue, Jul 02, 2013 at 01:00:52PM -0400, Jeff Layton wrote: > > Currently the way these get set is a little convoluted. If the dentry is > > allocated via lookup from userland, then it gets set by simple_lookup. > > If it gets allocated when the kernel is populating the directory, then > > it gets set via __rpc_lookup_create_exclusive, which has to check > > whether they might already be set. Between both of these, this ensures > > that all dentries have their d_op pointer set. > > > > Instead of doing that, just have them set at d_alloc time by pointing > > sb->s_d_op at them. With that change, we no longer want the lookup op > > to set them, so we must move to using our own lookup routine. > > There's a better solution - just make simple_lookup() skip d_set_d_op() > if superblock already has ->s_d_op (and thus d_alloc() has already > set the damn thing). Voila - we can just set ->s_d_op and leave > inode_operations as is. Yeah, that is a better solution and the code now in mainline looks reasonable. Trond already merged this patch however, so I'll spin up another patch on top of mainline to convert this back to simple_dir_inode_operations. Thanks,
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index a816b3a..ca5ad70d 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -480,6 +480,23 @@ static const struct dentry_operations rpc_dentry_operations = { .d_delete = rpc_delete_dentry, }; +/* + * Lookup the data. This is trivial - if the dentry didn't already + * exist, we know it is negative. + */ +static struct dentry * +rpc_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) +{ + if (dentry->d_name.len > NAME_MAX) + return ERR_PTR(-ENAMETOOLONG); + d_add(dentry, NULL); + return NULL; +} + +const struct inode_operations rpc_dir_inode_operations = { + .lookup = rpc_lookup, +}; + static struct inode * rpc_get_inode(struct super_block *sb, umode_t mode) { @@ -492,7 +509,7 @@ rpc_get_inode(struct super_block *sb, umode_t mode) switch (mode & S_IFMT) { case S_IFDIR: inode->i_fop = &simple_dir_operations; - inode->i_op = &simple_dir_inode_operations; + inode->i_op = &rpc_dir_inode_operations; inc_nlink(inode); default: break; @@ -666,11 +683,8 @@ static struct dentry *__rpc_lookup_create_exclusive(struct dentry *parent, if (!dentry) return ERR_PTR(-ENOMEM); } - if (dentry->d_inode == NULL) { - if (!dentry->d_op) - d_set_d_op(dentry, &rpc_dentry_operations); + if (dentry->d_inode == NULL) return dentry; - } dput(dentry); return ERR_PTR(-EEXIST); } @@ -1117,6 +1131,7 @@ rpc_fill_super(struct super_block *sb, void *data, int silent) sb->s_blocksize_bits = PAGE_CACHE_SHIFT; sb->s_magic = RPCAUTH_GSSMAGIC; sb->s_op = &s_ops; + sb->s_d_op = &rpc_dentry_operations; sb->s_time_gran = 1; inode = rpc_get_inode(sb, S_IFDIR | S_IRUGO | S_IXUGO);
Currently the way these get set is a little convoluted. If the dentry is allocated via lookup from userland, then it gets set by simple_lookup. If it gets allocated when the kernel is populating the directory, then it gets set via __rpc_lookup_create_exclusive, which has to check whether they might already be set. Between both of these, this ensures that all dentries have their d_op pointer set. Instead of doing that, just have them set at d_alloc time by pointing sb->s_d_op at them. With that change, we no longer want the lookup op to set them, so we must move to using our own lookup routine. Signed-off-by: Jeff Layton <jlayton@redhat.com> --- net/sunrpc/rpc_pipe.c | 25 ++++++++++++++++++++----- 1 file changed, 20 insertions(+), 5 deletions(-)