On May. 24, 2010, 21:28 +0300, andros@xxxxxxxxxx wrote:
> From: Andy Adamson <andros@xxxxxxxxxx>
>
> Taking a reference on the nfs_open_context to signal that a layoutcommit is
> needed is problematic because the last reference to the context triggers a
> close (nfs_release). But, if the layout holds a reference on the
> nfs_open_context, then close will not be triggered.
>
> Since we only use the rpc credential from the layoutcommit_ctx, replace the
> layoutcommit_ctx with the rpc_cred.
>
> Hold a reference on the rpc_cred until the layoutcommit rpc returns.
>
> If the layoutdriver fails to setup the layoutcommit, clear the layoutcommit
> properties and put the credential.
>
> Signed-off-by: Andy Adamson <andros@xxxxxxxxxx>
> ---
> fs/nfs/inode.c | 4 ++--
> fs/nfs/nfs4state.c | 2 +-
> fs/nfs/pnfs.c | 42 +++++++++++++++++-------------------------
> fs/nfs/write.c | 2 +-
> include/linux/nfs4_pnfs.h | 6 ++++++
> include/linux/nfs_fs.h | 3 +--
> include/linux/pnfs_xdr.h | 1 -
> 7 files changed, 28 insertions(+), 32 deletions(-)
>
> diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
> index 7fd33d9..2b8e8e6 100644
> --- a/fs/nfs/inode.c
> +++ b/fs/nfs/inode.c
> @@ -1110,7 +1110,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
> /*
> * file needs layout commit, server attributes may be stale
> */
> - if (nfsi->layoutcommit_ctx && nfsi->change_attr >= fattr->change_attr) {
> + if (do_layoutcommit(nfsi) && nfsi->change_attr >= fattr->change_attr) {
Andy, the patch looks good overall, thanks!
One nit though: mind if I rename "do_layoutcommit" which reads as if it actually performs
the layoutcommit call to "layoutcommit_needed"?
Benny
> dprintk("NFS: %s: layoutcommit is needed for file %s/%ld\n",
> __func__, inode->i_sb->s_id, inode->i_ino);
> return 0;
> @@ -1331,7 +1331,7 @@ static void pnfs_alloc_init_inode(struct nfs_inode *nfsi)
> nfsi->pnfs_layout_state = 0;
> memset(&nfsi->layout.stateid, 0, NFS4_STATEID_SIZE);
> nfsi->layout.roc_iomode = 0;
> - nfsi->layoutcommit_ctx = NULL;
> + nfsi->lo_cred = NULL;
> nfsi->pnfs_write_begin_pos = 0;
> nfsi->pnfs_write_end_pos = 0;
> #endif /* CONFIG_NFS_V4_1 */
> diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
> index d145de1..bf03fde 100644
> --- a/fs/nfs/nfs4state.c
> +++ b/fs/nfs/nfs4state.c
> @@ -589,7 +589,7 @@ static void __nfs4_close(struct path *path, struct nfs4_state *state, fmode_t fm
> #ifdef CONFIG_NFS_V4_1
> struct nfs_inode *nfsi = NFS_I(state->inode);
>
> - if (nfsi->layoutcommit_ctx)
> + if (do_layoutcommit(nfsi))
> pnfs_layoutcommit_inode(state->inode, wait);
> if (has_layout(nfsi) && nfsi->layout.roc_iomode) {
> struct nfs4_pnfs_layout_segment range;
> diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
> index 20285bc..ee530c8 100644
> --- a/fs/nfs/pnfs.c
> +++ b/fs/nfs/pnfs.c
> @@ -145,21 +145,19 @@ find_pnfs(u32 id, struct pnfs_module **module) {
> return 0;
> }
>
> -/* Set context to indicate we require a layoutcommit
> +/* Set lo_cred to indicate we require a layoutcommit
> * If we don't even have a layout, we don't need to commit it.
> */
> void
> pnfs_need_layoutcommit(struct nfs_inode *nfsi, struct nfs_open_context *ctx)
> {
> - dprintk("%s: has_layout=%d layoutcommit_ctx=%p ctx=%p\n", __func__,
> - has_layout(nfsi), nfsi->layoutcommit_ctx, ctx);
> + dprintk("%s: has_layout=%d ctx=%p\n", __func__, has_layout(nfsi), ctx);
> spin_lock(&nfsi->lo_lock);
> - if (has_layout(nfsi) && !nfsi->layoutcommit_ctx) {
> - nfsi->layoutcommit_ctx = get_nfs_open_context(ctx);
> + if (has_layout(nfsi) && !do_layoutcommit(nfsi)) {
> + nfsi->lo_cred = get_rpccred(ctx->state->owner->so_cred);
> nfsi->change_attr++;
> spin_unlock(&nfsi->lo_lock);
> - dprintk("%s: Set layoutcommit_ctx=%p\n", __func__,
> - nfsi->layoutcommit_ctx);
> + dprintk("%s: Set layoutcommit\n", __func__);
> return;
> }
> spin_unlock(&nfsi->lo_lock);
> @@ -755,7 +753,7 @@ _pnfs_return_layout(struct inode *ino, struct nfs4_pnfs_layout_segment *range,
> !pnfs_return_layout_barrier(nfsi, &arg));
> }
>
> - if (nfsi->layoutcommit_ctx) {
> + if (do_layoutcommit(nfsi)) {
> status = pnfs_layoutcommit_inode(ino, wait);
> if (status) {
> dprintk("%s: layoutcommit failed, status=%d. "
> @@ -1910,16 +1908,9 @@ pnfs_layoutcommit_done(struct pnfs_layoutcommit_data *data)
>
> dprintk("%s: (status %d)\n", __func__, data->status);
>
> - /* TODO: For now, set an error in the open context (just like
> - * if a commit failed) We may want to do more, much more, like
> - * replay all writes through the NFSv4
> - * server, or something.
> - */
> - if (data->status < 0) {
> + if (data->status < 0)
> printk(KERN_ERR "%s, Layoutcommit Failed! = %d\n",
> __func__, data->status);
> - data->ctx->error = data->status;
> - }
>
> /* TODO: Maybe we should avoid this by allowing the layout driver
> * to directly xdr its layout on the wire.
> @@ -1929,9 +1920,7 @@ pnfs_layoutcommit_done(struct pnfs_layoutcommit_data *data)
> &nfsi->layout,
> &data->args,
> data->status);
> -
> - /* release the open_context acquired in pnfs_writeback_done */
> - put_nfs_open_context(data->ctx);
> + put_rpccred(data->cred);
> }
>
> /*
> @@ -1995,24 +1984,27 @@ pnfs_layoutcommit_inode(struct inode *inode, int sync)
> return -ENOMEM;
>
> spin_lock(&nfsi->lo_lock);
> - if (!nfsi->layoutcommit_ctx)
> + if (!do_layoutcommit(nfsi))
> goto out_unlock;
>
> data->args.inode = inode;
> - data->cred = nfsi->layoutcommit_ctx->cred;
> - data->ctx = nfsi->layoutcommit_ctx;
> + data->cred = nfsi->lo_cred;
>
> /* Set up layout commit args*/
> status = pnfs_layoutcommit_setup(data, sync);
> - if (status)
> - goto out_unlock;
>
> /* Clear layoutcommit properties in the inode so
> * new lc info can be generated
> */
> nfsi->pnfs_write_begin_pos = 0;
> nfsi->pnfs_write_end_pos = 0;
> - nfsi->layoutcommit_ctx = NULL;
> + nfsi->lo_cred = NULL;
> +
> + if (status) {
> + /* The layout driver failed to setup the layoutcommit */
> + put_rpccred(data->cred);
> + goto out_unlock;
> + }
>
> /* release lock on pnfs layoutcommit attrs */
> spin_unlock(&nfsi->lo_lock);
> diff --git a/fs/nfs/write.c b/fs/nfs/write.c
> index a4c95a0..2c1918e 100644
> --- a/fs/nfs/write.c
> +++ b/fs/nfs/write.c
> @@ -1490,7 +1490,7 @@ static int nfs_commit_inode(struct inode *inode, int how)
> nfs_wait_bit_killable,
> TASK_KILLABLE);
> #ifdef CONFIG_NFS_V4_1
> - if (may_wait && NFS_I(inode)->layoutcommit_ctx) {
> + if (may_wait && do_layoutcommit(NFS_I(inode))) {
> error = pnfs_layoutcommit_inode(inode, 1);
> if (error < 0)
> return error;
> diff --git a/include/linux/nfs4_pnfs.h b/include/linux/nfs4_pnfs.h
> index 4d47b48..c9ef43e 100644
> --- a/include/linux/nfs4_pnfs.h
> +++ b/include/linux/nfs4_pnfs.h
> @@ -80,6 +80,12 @@ has_layout(struct nfs_inode *nfsi)
> return nfsi->layout.ld_data != NULL;
> }
>
> +static inline bool
> +do_layoutcommit(struct nfs_inode *nfsi)
> +{
> + return nfsi->lo_cred != NULL;
> +}
> +
> #endif /* CONFIG_NFS_V4_1 */
>
> struct pnfs_layout_segment {
> diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
> index 98a8dc0..478b00c 100644
> --- a/include/linux/nfs_fs.h
> +++ b/include/linux/nfs_fs.h
> @@ -203,11 +203,10 @@ struct nfs_inode {
> #define NFS_INO_RW_LAYOUT_FAILED 1 /* get rw layout failed stop trying */
> #define NFS_INO_LAYOUT_ALLOC 2 /* bit lock for layout allocation */
> time_t pnfs_layout_suspend;
> + struct rpc_cred *lo_cred; /* layoutcommit credential */
> wait_queue_head_t lo_waitq;
> spinlock_t lo_lock;
> struct pnfs_layout_type layout;
> - /* use rpc_creds in this open_context to send LAYOUTCOMMIT to MDS */
> - struct nfs_open_context *layoutcommit_ctx;
> /* DH: These vars keep track of the maximum write range
> * so the values can be used for layoutcommit.
> */
> diff --git a/include/linux/pnfs_xdr.h b/include/linux/pnfs_xdr.h
> index a0bf341..154b04e 100644
> --- a/include/linux/pnfs_xdr.h
> +++ b/include/linux/pnfs_xdr.h
> @@ -86,7 +86,6 @@ struct pnfs_layoutcommit_data {
> bool is_sync;
> struct rpc_cred *cred;
> struct nfs_fattr fattr;
> - struct nfs_open_context *ctx;
> struct pnfs_layoutcommit_arg args;
> struct pnfs_layoutcommit_res res;
> int status;
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
