From 691b45ddbd182a4ce0bc91953c70c845cf0935f1 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 19 Aug 2019 18:36:19 -0400 Subject: SUNRPC: Remove rpc_wake_up_queued_task_on_wq() Clean up: commit c544577daddb ("SUNRPC: Clean up transport write space handling") appears to have removed the last caller of rpc_wake_up_queued_task_on_wq(). Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/sched.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h index baa3ecdb882f..d1283bddd218 100644 --- a/include/linux/sunrpc/sched.h +++ b/include/linux/sunrpc/sched.h @@ -243,9 +243,6 @@ void rpc_sleep_on_priority_timeout(struct rpc_wait_queue *queue, void rpc_sleep_on_priority(struct rpc_wait_queue *, struct rpc_task *, int priority); -void rpc_wake_up_queued_task_on_wq(struct workqueue_struct *wq, - struct rpc_wait_queue *queue, - struct rpc_task *task); void rpc_wake_up_queued_task(struct rpc_wait_queue *, struct rpc_task *); void rpc_wake_up_queued_task_set_status(struct rpc_wait_queue *, -- cgit v1.2.3 From aeaed4848234c97fb720b0e51a0c56dc8de0eeed Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 19 Aug 2019 18:40:58 -0400 Subject: xprtrdma: Boost client's max slot table size to match Linux server I've heard rumors of an NFS/RDMA server implementation that has a default credit limit of 1024. The client's default setting remains at 128. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xprtrdma.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprtrdma.h b/include/linux/sunrpc/xprtrdma.h index 86fc38ff0355..16c239e0d6dd 100644 --- a/include/linux/sunrpc/xprtrdma.h +++ b/include/linux/sunrpc/xprtrdma.h @@ -49,9 +49,9 @@ * fully-chunked NFS message (read chunks are the largest). Note only * a single chunk type per message is supported currently. */ -#define RPCRDMA_MIN_SLOT_TABLE (2U) +#define RPCRDMA_MIN_SLOT_TABLE (4U) #define RPCRDMA_DEF_SLOT_TABLE (128U) -#define RPCRDMA_MAX_SLOT_TABLE (256U) +#define RPCRDMA_MAX_SLOT_TABLE (16384U) #define RPCRDMA_MIN_INLINE (1024) /* min inline thresh */ #define RPCRDMA_DEF_INLINE (4096) /* default inline thresh */ -- cgit v1.2.3 From cc204d01262a69218b2d0db5cdea371de85871d9 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 10 Sep 2019 13:01:35 -0400 Subject: SUNRPC: Dequeue the request from the receive queue while we're re-encoding Ensure that we dequeue the request from the transport receive queue while we're re-encoding to prevent issues like use-after-free when we release the bvec. Fixes: 7536908982047 ("SUNRPC: Ensure the bvecs are reset when we re-encode...") Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org # v4.20+ Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xprt.h | 1 + net/sunrpc/clnt.c | 6 ++--- net/sunrpc/xprt.c | 54 ++++++++++++++++++++++++++------------------- 3 files changed, 35 insertions(+), 26 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 13e108bcc9eb..d783e15ba898 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -352,6 +352,7 @@ bool xprt_prepare_transmit(struct rpc_task *task); void xprt_request_enqueue_transmit(struct rpc_task *task); void xprt_request_enqueue_receive(struct rpc_task *task); void xprt_request_wait_receive(struct rpc_task *task); +void xprt_request_dequeue_xprt(struct rpc_task *task); bool xprt_request_need_retransmit(struct rpc_task *task); void xprt_transmit(struct rpc_task *task); void xprt_end_transmit(struct rpc_task *task); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index d8679b6027e9..0359466947e2 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1862,6 +1862,7 @@ rpc_xdr_encode(struct rpc_task *task) req->rq_rbuffer, req->rq_rcvsize); + req->rq_reply_bytes_recvd = 0; req->rq_snd_buf.head[0].iov_len = 0; xdr_init_encode(&xdr, &req->rq_snd_buf, req->rq_snd_buf.head[0].iov_base, req); @@ -1881,6 +1882,8 @@ call_encode(struct rpc_task *task) if (!rpc_task_need_encode(task)) goto out; dprint_status(task); + /* Dequeue task from the receive queue while we're encoding */ + xprt_request_dequeue_xprt(task); /* Encode here so that rpcsec_gss can use correct sequence number. */ rpc_xdr_encode(task); /* Did the encode result in an error condition? */ @@ -2501,9 +2504,6 @@ call_decode(struct rpc_task *task) return; case -EAGAIN: task->tk_status = 0; - xdr_free_bvec(&req->rq_rcv_buf); - req->rq_reply_bytes_recvd = 0; - req->rq_rcv_buf.len = 0; if (task->tk_client->cl_discrtry) xprt_conditional_disconnect(req->rq_xprt, req->rq_connect_cookie); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 783748dc5e6f..02d5b2125c07 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -1323,6 +1323,36 @@ xprt_request_dequeue_transmit(struct rpc_task *task) spin_unlock(&xprt->queue_lock); } +/** + * xprt_request_dequeue_xprt - remove a task from the transmit+receive queue + * @task: pointer to rpc_task + * + * Remove a task from the transmit and receive queues, and ensure that + * it is not pinned by the receive work item. + */ +void +xprt_request_dequeue_xprt(struct rpc_task *task) +{ + struct rpc_rqst *req = task->tk_rqstp; + struct rpc_xprt *xprt = req->rq_xprt; + + if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate) || + test_bit(RPC_TASK_NEED_RECV, &task->tk_runstate) || + xprt_is_pinned_rqst(req)) { + spin_lock(&xprt->queue_lock); + xprt_request_dequeue_transmit_locked(task); + xprt_request_dequeue_receive_locked(task); + while (xprt_is_pinned_rqst(req)) { + set_bit(RPC_TASK_MSG_PIN_WAIT, &task->tk_runstate); + spin_unlock(&xprt->queue_lock); + xprt_wait_on_pinned_rqst(req); + spin_lock(&xprt->queue_lock); + clear_bit(RPC_TASK_MSG_PIN_WAIT, &task->tk_runstate); + } + spin_unlock(&xprt->queue_lock); + } +} + /** * xprt_request_prepare - prepare an encoded request for transport * @req: pointer to rpc_rqst @@ -1754,28 +1784,6 @@ void xprt_retry_reserve(struct rpc_task *task) xprt_do_reserve(xprt, task); } -static void -xprt_request_dequeue_all(struct rpc_task *task, struct rpc_rqst *req) -{ - struct rpc_xprt *xprt = req->rq_xprt; - - if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate) || - test_bit(RPC_TASK_NEED_RECV, &task->tk_runstate) || - xprt_is_pinned_rqst(req)) { - spin_lock(&xprt->queue_lock); - xprt_request_dequeue_transmit_locked(task); - xprt_request_dequeue_receive_locked(task); - while (xprt_is_pinned_rqst(req)) { - set_bit(RPC_TASK_MSG_PIN_WAIT, &task->tk_runstate); - spin_unlock(&xprt->queue_lock); - xprt_wait_on_pinned_rqst(req); - spin_lock(&xprt->queue_lock); - clear_bit(RPC_TASK_MSG_PIN_WAIT, &task->tk_runstate); - } - spin_unlock(&xprt->queue_lock); - } -} - /** * xprt_release - release an RPC request slot * @task: task which is finished with the slot @@ -1795,7 +1803,7 @@ void xprt_release(struct rpc_task *task) } xprt = req->rq_xprt; - xprt_request_dequeue_all(task, req); + xprt_request_dequeue_xprt(task); spin_lock(&xprt->transport_lock); xprt->ops->release_xprt(xprt, task); if (xprt->ops->release_request) -- cgit v1.2.3 From f925ab926d1a9c2112d34ecb59fbb050bb58646c Mon Sep 17 00:00:00 2001 From: Benjamin Coddington Date: Mon, 16 Sep 2019 07:59:38 -0400 Subject: SUNRPC: Rename xdr_buf_read_netobj to xdr_buf_read_mic Let the name reflect the single use. The function now assumes the GSS MIC is the last object in the buffer. Signed-off-by: Benjamin Coddington Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xdr.h | 2 +- net/sunrpc/auth_gss/auth_gss.c | 2 +- net/sunrpc/xdr.c | 52 ++++++++++++++++++++++++------------------ 3 files changed, 32 insertions(+), 24 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index 8a87d8bcb197..f33e5013bdfb 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -186,7 +186,7 @@ xdr_adjust_iovec(struct kvec *iov, __be32 *p) extern void xdr_shift_buf(struct xdr_buf *, size_t); extern void xdr_buf_from_iov(struct kvec *, struct xdr_buf *); extern int xdr_buf_subsegment(struct xdr_buf *, struct xdr_buf *, unsigned int, unsigned int); -extern int xdr_buf_read_netobj(struct xdr_buf *, struct xdr_netobj *, unsigned int); +extern int xdr_buf_read_mic(struct xdr_buf *, struct xdr_netobj *, unsigned int); extern int read_bytes_from_xdr_buf(struct xdr_buf *, unsigned int, void *, unsigned int); extern int write_bytes_to_xdr_buf(struct xdr_buf *, unsigned int, void *, unsigned int); diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 4ce42c62458e..d75fddca44c9 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1960,7 +1960,7 @@ gss_unwrap_resp_integ(struct rpc_task *task, struct rpc_cred *cred, if (xdr_buf_subsegment(rcv_buf, &integ_buf, data_offset, integ_len)) goto unwrap_failed; - if (xdr_buf_read_netobj(rcv_buf, &mic, mic_offset)) + if (xdr_buf_read_mic(rcv_buf, &mic, mic_offset)) goto unwrap_failed; maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &integ_buf, &mic); if (maj_stat == GSS_S_CONTEXT_EXPIRED) diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 24ec2ab5bfa5..14ba9e72a204 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -1236,52 +1236,60 @@ xdr_encode_word(struct xdr_buf *buf, unsigned int base, u32 obj) } EXPORT_SYMBOL_GPL(xdr_encode_word); -/* If the netobj starting offset bytes from the start of xdr_buf is contained - * entirely in the head, pages, or tail, set object to point to it; otherwise - * shift the buffer until it is contained entirely within the pages or tail. +/** + * xdr_buf_read_mic() - obtain the address of the GSS mic from xdr buf + * @buf: pointer to buffer containing a mic + * @mic: on success, returns the address of the mic + * @offset: the offset in buf where mic may be found + * + * This function may modify the xdr buf if the mic is found to be straddling + * a boundary between head, pages, and tail. On success the mic can be read + * from the address returned. There is no need to free the mic. + * + * Return: Success returns 0, otherwise an integer error. */ -int xdr_buf_read_netobj(struct xdr_buf *buf, struct xdr_netobj *obj, unsigned int offset) +int xdr_buf_read_mic(struct xdr_buf *buf, struct xdr_netobj *mic, unsigned int offset) { struct xdr_buf subbuf; unsigned int boundary; - if (xdr_decode_word(buf, offset, &obj->len)) + if (xdr_decode_word(buf, offset, &mic->len)) return -EFAULT; offset += 4; - /* Is the obj partially in the head? */ + /* Is the mic partially in the head? */ boundary = buf->head[0].iov_len; - if (offset < boundary && (offset + obj->len) > boundary) + if (offset < boundary && (offset + mic->len) > boundary) xdr_shift_buf(buf, boundary - offset); - /* Is the obj partially in the pages? */ + /* Is the mic partially in the pages? */ boundary += buf->page_len; - if (offset < boundary && (offset + obj->len) > boundary) + if (offset < boundary && (offset + mic->len) > boundary) xdr_shrink_pagelen(buf, boundary - offset); - if (xdr_buf_subsegment(buf, &subbuf, offset, obj->len)) + if (xdr_buf_subsegment(buf, &subbuf, offset, mic->len)) return -EFAULT; - /* Is the obj contained entirely in the head? */ - obj->data = subbuf.head[0].iov_base; - if (subbuf.head[0].iov_len == obj->len) + /* Is the mic contained entirely in the head? */ + mic->data = subbuf.head[0].iov_base; + if (subbuf.head[0].iov_len == mic->len) return 0; - /* ..or is the obj contained entirely in the tail? */ - obj->data = subbuf.tail[0].iov_base; - if (subbuf.tail[0].iov_len == obj->len) + /* ..or is the mic contained entirely in the tail? */ + mic->data = subbuf.tail[0].iov_base; + if (subbuf.tail[0].iov_len == mic->len) return 0; - /* Find a contiguous area in @buf to hold all of @obj */ - if (obj->len > buf->buflen - buf->len) + /* Find a contiguous area in @buf to hold all of @mic */ + if (mic->len > buf->buflen - buf->len) return -ENOMEM; if (buf->tail[0].iov_len != 0) - obj->data = buf->tail[0].iov_base + buf->tail[0].iov_len; + mic->data = buf->tail[0].iov_base + buf->tail[0].iov_len; else - obj->data = buf->head[0].iov_base + buf->head[0].iov_len; - __read_bytes_from_xdr_buf(&subbuf, obj->data, obj->len); + mic->data = buf->head[0].iov_base + buf->head[0].iov_len; + __read_bytes_from_xdr_buf(&subbuf, mic->data, mic->len); return 0; } -EXPORT_SYMBOL_GPL(xdr_buf_read_netobj); +EXPORT_SYMBOL_GPL(xdr_buf_read_mic); /* Returns 0 on success, or else a negative error code. */ static int -- cgit v1.2.3 From 406cd91533dcc5e82ef2373c39e6a531d944131e Mon Sep 17 00:00:00 2001 From: Benjamin Coddington Date: Fri, 13 Sep 2019 08:29:02 -0400 Subject: NFS: Refactor nfs_instantiate() for dentry referencing callers Since commit b0c6108ecf64 ("nfs_instantiate(): prevent multiple aliases for directory inode"), nfs_instantiate() may succeed without actually instantiating the dentry that was passed in. That can be problematic for some callers in NFSv3, so this patch breaks things up so we can get the actual dentry obtained. Signed-off-by: Benjamin Coddington Signed-off-by: Anna Schumaker --- fs/nfs/dir.c | 41 +++++++++++++++++++++++++++-------------- include/linux/nfs_fs.h | 3 +++ 2 files changed, 30 insertions(+), 14 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 8d501093660f..d1bbf2fb6ac7 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1669,24 +1669,23 @@ static int nfs4_lookup_revalidate(struct dentry *dentry, unsigned int flags) #endif /* CONFIG_NFSV4 */ -/* - * Code common to create, mkdir, and mknod. - */ -int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle, +struct dentry * +nfs_add_or_obtain(struct dentry *dentry, struct nfs_fh *fhandle, struct nfs_fattr *fattr, struct nfs4_label *label) { struct dentry *parent = dget_parent(dentry); struct inode *dir = d_inode(parent); struct inode *inode; - struct dentry *d; - int error = -EACCES; + struct dentry *d = NULL; + int error; d_drop(dentry); /* We may have been initialized further down */ if (d_really_is_positive(dentry)) goto out; + if (fhandle->size == 0) { error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, fhandle, fattr, NULL); if (error) @@ -1702,18 +1701,32 @@ int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle, } inode = nfs_fhget(dentry->d_sb, fhandle, fattr, label); d = d_splice_alias(inode, dentry); - if (IS_ERR(d)) { - error = PTR_ERR(d); - goto out_error; - } - dput(d); out: dput(parent); - return 0; + return d; out_error: nfs_mark_for_revalidate(dir); - dput(parent); - return error; + d = ERR_PTR(error); + goto out; +} +EXPORT_SYMBOL_GPL(nfs_add_or_obtain); + +/* + * Code common to create, mkdir, and mknod. + */ +int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle, + struct nfs_fattr *fattr, + struct nfs4_label *label) +{ + struct dentry *d; + + d = nfs_add_or_obtain(dentry, fhandle, fattr, label); + if (IS_ERR(d)) + return PTR_ERR(d); + + /* Callers don't care */ + dput(d); + return 0; } EXPORT_SYMBOL_GPL(nfs_instantiate); diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 0a11712a80e3..570a60c2f4f4 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -490,6 +490,9 @@ extern const struct file_operations nfs_dir_operations; extern const struct dentry_operations nfs_dentry_operations; extern void nfs_force_lookup_revalidate(struct inode *dir); +extern struct dentry *nfs_add_or_obtain(struct dentry *dentry, + struct nfs_fh *fh, struct nfs_fattr *fattr, + struct nfs4_label *label); extern int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fh, struct nfs_fattr *fattr, struct nfs4_label *label); extern int nfs_may_open(struct inode *inode, const struct cred *cred, int openflags); -- cgit v1.2.3