diff options
Diffstat (limited to 'net/sunrpc')
25 files changed, 955 insertions, 875 deletions
diff --git a/net/sunrpc/Kconfig b/net/sunrpc/Kconfig index 9068e72aa..04ce2c0b6 100644 --- a/net/sunrpc/Kconfig +++ b/net/sunrpc/Kconfig @@ -48,28 +48,16 @@ config SUNRPC_DEBUG If unsure, say Y. -config SUNRPC_XPRT_RDMA_CLIENT - tristate "RPC over RDMA Client Support" +config SUNRPC_XPRT_RDMA + tristate "RPC-over-RDMA transport" depends on SUNRPC && INFINIBAND && INFINIBAND_ADDR_TRANS default SUNRPC && INFINIBAND help - This option allows the NFS client to support an RDMA-enabled - transport. + This option allows the NFS client and server to use RDMA + transports (InfiniBand, iWARP, or RoCE). - To compile RPC client RDMA transport support as a module, - choose M here: the module will be called xprtrdma. + To compile this support as a module, choose M. The module + will be called rpcrdma.ko. - If unsure, say N. - -config SUNRPC_XPRT_RDMA_SERVER - tristate "RPC over RDMA Server Support" - depends on SUNRPC && INFINIBAND && INFINIBAND_ADDR_TRANS - default SUNRPC && INFINIBAND - help - This option allows the NFS server to support an RDMA-enabled - transport. - - To compile RPC server RDMA transport support as a module, - choose M here: the module will be called svcrdma. - - If unsure, say N. + If unsure, or you know there is no RDMA capability on your + hardware platform, say N. diff --git a/net/sunrpc/Makefile b/net/sunrpc/Makefile index 15e6f6c23..b512fbd9d 100644 --- a/net/sunrpc/Makefile +++ b/net/sunrpc/Makefile @@ -5,8 +5,7 @@ obj-$(CONFIG_SUNRPC) += sunrpc.o obj-$(CONFIG_SUNRPC_GSS) += auth_gss/ - -obj-y += xprtrdma/ +obj-$(CONFIG_SUNRPC_XPRT_RDMA) += xprtrdma/ sunrpc-y := clnt.o xprt.o socklib.o xprtsock.o sched.o \ auth.o auth_null.o auth_unix.o auth_generic.o \ @@ -15,6 +14,6 @@ sunrpc-y := clnt.o xprt.o socklib.o xprtsock.o sched.o \ sunrpc_syms.o cache.o rpc_pipe.o \ svc_xprt.o sunrpc-$(CONFIG_SUNRPC_DEBUG) += debugfs.o -sunrpc-$(CONFIG_SUNRPC_BACKCHANNEL) += backchannel_rqst.o bc_svc.o +sunrpc-$(CONFIG_SUNRPC_BACKCHANNEL) += backchannel_rqst.o sunrpc-$(CONFIG_PROC_FS) += stats.o sunrpc-$(CONFIG_SYSCTL) += sysctl.o diff --git a/net/sunrpc/auth.c b/net/sunrpc/auth.c index 47f38be41..02f53674d 100644 --- a/net/sunrpc/auth.c +++ b/net/sunrpc/auth.c @@ -72,7 +72,7 @@ static int param_get_hashtbl_sz(char *buffer, const struct kernel_param *kp) #define param_check_hashtbl_sz(name, p) __param_check(name, p, unsigned int); -static struct kernel_param_ops param_ops_hashtbl_sz = { +static const struct kernel_param_ops param_ops_hashtbl_sz = { .set = param_set_hashtbl_sz, .get = param_get_hashtbl_sz, }; diff --git a/net/sunrpc/auth_gss/gss_krb5_crypto.c b/net/sunrpc/auth_gss/gss_krb5_crypto.c index b5408e8a3..fee3c15a4 100644 --- a/net/sunrpc/auth_gss/gss_krb5_crypto.c +++ b/net/sunrpc/auth_gss/gss_krb5_crypto.c @@ -881,9 +881,7 @@ krb5_rc4_setup_seq_key(struct krb5_ctx *kctx, struct crypto_blkcipher *cipher, if (err) goto out_err; - sg_init_table(sg, 1); - sg_set_buf(sg, &zeroconstant, 4); - + sg_init_one(sg, &zeroconstant, 4); err = crypto_hash_digest(&desc, sg, 4, Kseq); if (err) goto out_err; @@ -951,9 +949,7 @@ krb5_rc4_setup_enc_key(struct krb5_ctx *kctx, struct crypto_blkcipher *cipher, if (err) goto out_err; - sg_init_table(sg, 1); - sg_set_buf(sg, zeroconstant, 4); - + sg_init_one(sg, zeroconstant, 4); err = crypto_hash_digest(&desc, sg, 4, Kcrypt); if (err) goto out_err; diff --git a/net/sunrpc/backchannel_rqst.c b/net/sunrpc/backchannel_rqst.c index 28504dfd3..6255d1411 100644 --- a/net/sunrpc/backchannel_rqst.c +++ b/net/sunrpc/backchannel_rqst.c @@ -37,16 +37,18 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ static inline int xprt_need_to_requeue(struct rpc_xprt *xprt) { - return xprt->bc_alloc_count > 0; + return xprt->bc_alloc_count < atomic_read(&xprt->bc_free_slots); } static inline void xprt_inc_alloc_count(struct rpc_xprt *xprt, unsigned int n) { + atomic_add(n, &xprt->bc_free_slots); xprt->bc_alloc_count += n; } static inline int xprt_dec_alloc_count(struct rpc_xprt *xprt, unsigned int n) { + atomic_sub(n, &xprt->bc_free_slots); return xprt->bc_alloc_count -= n; } @@ -67,6 +69,55 @@ static void xprt_free_allocation(struct rpc_rqst *req) kfree(req); } +static int xprt_alloc_xdr_buf(struct xdr_buf *buf, gfp_t gfp_flags) +{ + struct page *page; + /* Preallocate one XDR receive buffer */ + page = alloc_page(gfp_flags); + if (page == NULL) + return -ENOMEM; + buf->head[0].iov_base = page_address(page); + buf->head[0].iov_len = PAGE_SIZE; + buf->tail[0].iov_base = NULL; + buf->tail[0].iov_len = 0; + buf->page_len = 0; + buf->len = 0; + buf->buflen = PAGE_SIZE; + return 0; +} + +static +struct rpc_rqst *xprt_alloc_bc_req(struct rpc_xprt *xprt, gfp_t gfp_flags) +{ + struct rpc_rqst *req; + + /* Pre-allocate one backchannel rpc_rqst */ + req = kzalloc(sizeof(*req), gfp_flags); + if (req == NULL) + return NULL; + + req->rq_xprt = xprt; + INIT_LIST_HEAD(&req->rq_list); + INIT_LIST_HEAD(&req->rq_bc_list); + + /* Preallocate one XDR receive buffer */ + if (xprt_alloc_xdr_buf(&req->rq_rcv_buf, gfp_flags) < 0) { + printk(KERN_ERR "Failed to create bc receive xbuf\n"); + goto out_free; + } + req->rq_rcv_buf.len = PAGE_SIZE; + + /* Preallocate one XDR send buffer */ + if (xprt_alloc_xdr_buf(&req->rq_snd_buf, gfp_flags) < 0) { + printk(KERN_ERR "Failed to create bc snd xbuf\n"); + goto out_free; + } + return req; +out_free: + xprt_free_allocation(req); + return NULL; +} + /* * Preallocate up to min_reqs structures and related buffers for use * by the backchannel. This function can be called multiple times @@ -87,9 +138,7 @@ static void xprt_free_allocation(struct rpc_rqst *req) */ int xprt_setup_backchannel(struct rpc_xprt *xprt, unsigned int min_reqs) { - struct page *page_rcv = NULL, *page_snd = NULL; - struct xdr_buf *xbufp = NULL; - struct rpc_rqst *req, *tmp; + struct rpc_rqst *req; struct list_head tmp_list; int i; @@ -106,7 +155,7 @@ int xprt_setup_backchannel(struct rpc_xprt *xprt, unsigned int min_reqs) INIT_LIST_HEAD(&tmp_list); for (i = 0; i < min_reqs; i++) { /* Pre-allocate one backchannel rpc_rqst */ - req = kzalloc(sizeof(struct rpc_rqst), GFP_KERNEL); + req = xprt_alloc_bc_req(xprt, GFP_KERNEL); if (req == NULL) { printk(KERN_ERR "Failed to create bc rpc_rqst\n"); goto out_free; @@ -115,41 +164,6 @@ int xprt_setup_backchannel(struct rpc_xprt *xprt, unsigned int min_reqs) /* Add the allocated buffer to the tmp list */ dprintk("RPC: adding req= %p\n", req); list_add(&req->rq_bc_pa_list, &tmp_list); - - req->rq_xprt = xprt; - INIT_LIST_HEAD(&req->rq_list); - INIT_LIST_HEAD(&req->rq_bc_list); - - /* Preallocate one XDR receive buffer */ - page_rcv = alloc_page(GFP_KERNEL); - if (page_rcv == NULL) { - printk(KERN_ERR "Failed to create bc receive xbuf\n"); - goto out_free; - } - xbufp = &req->rq_rcv_buf; - xbufp->head[0].iov_base = page_address(page_rcv); - xbufp->head[0].iov_len = PAGE_SIZE; - xbufp->tail[0].iov_base = NULL; - xbufp->tail[0].iov_len = 0; - xbufp->page_len = 0; - xbufp->len = PAGE_SIZE; - xbufp->buflen = PAGE_SIZE; - - /* Preallocate one XDR send buffer */ - page_snd = alloc_page(GFP_KERNEL); - if (page_snd == NULL) { - printk(KERN_ERR "Failed to create bc snd xbuf\n"); - goto out_free; - } - - xbufp = &req->rq_snd_buf; - xbufp->head[0].iov_base = page_address(page_snd); - xbufp->head[0].iov_len = 0; - xbufp->tail[0].iov_base = NULL; - xbufp->tail[0].iov_len = 0; - xbufp->page_len = 0; - xbufp->len = 0; - xbufp->buflen = PAGE_SIZE; } /* @@ -167,7 +181,10 @@ out_free: /* * Memory allocation failed, free the temporary list */ - list_for_each_entry_safe(req, tmp, &tmp_list, rq_bc_pa_list) { + while (!list_empty(&tmp_list)) { + req = list_first_entry(&tmp_list, + struct rpc_rqst, + rq_bc_pa_list); list_del(&req->rq_bc_pa_list); xprt_free_allocation(req); } @@ -217,9 +234,15 @@ static struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt, __be32 xid) struct rpc_rqst *req = NULL; dprintk("RPC: allocate a backchannel request\n"); - if (list_empty(&xprt->bc_pa_list)) + if (atomic_read(&xprt->bc_free_slots) <= 0) goto not_found; - + if (list_empty(&xprt->bc_pa_list)) { + req = xprt_alloc_bc_req(xprt, GFP_ATOMIC); + if (!req) + goto not_found; + list_add_tail(&req->rq_bc_pa_list, &xprt->bc_pa_list); + xprt->bc_alloc_count++; + } req = list_first_entry(&xprt->bc_pa_list, struct rpc_rqst, rq_bc_pa_list); req->rq_reply_bytes_recvd = 0; @@ -245,11 +268,21 @@ void xprt_free_bc_request(struct rpc_rqst *req) req->rq_connect_cookie = xprt->connect_cookie - 1; smp_mb__before_atomic(); - WARN_ON_ONCE(!test_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state)); clear_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state); smp_mb__after_atomic(); - if (!xprt_need_to_requeue(xprt)) { + /* + * Return it to the list of preallocations so that it + * may be reused by a new callback request. + */ + spin_lock_bh(&xprt->bc_pa_lock); + if (xprt_need_to_requeue(xprt)) { + list_add_tail(&req->rq_bc_pa_list, &xprt->bc_pa_list); + xprt->bc_alloc_count++; + req = NULL; + } + spin_unlock_bh(&xprt->bc_pa_lock); + if (req != NULL) { /* * The last remaining session was destroyed while this * entry was in use. Free the entry and don't attempt @@ -260,14 +293,6 @@ void xprt_free_bc_request(struct rpc_rqst *req) xprt_free_allocation(req); return; } - - /* - * Return it to the list of preallocations so that it - * may be reused by a new callback request. - */ - spin_lock_bh(&xprt->bc_pa_lock); - list_add_tail(&req->rq_bc_pa_list, &xprt->bc_pa_list); - spin_unlock_bh(&xprt->bc_pa_lock); } /* @@ -311,6 +336,7 @@ void xprt_complete_bc_request(struct rpc_rqst *req, uint32_t copied) spin_lock(&xprt->bc_pa_lock); list_del(&req->rq_bc_pa_list); + xprt_dec_alloc_count(xprt, 1); spin_unlock(&xprt->bc_pa_lock); req->rq_private_buf.len = copied; diff --git a/net/sunrpc/bc_svc.c b/net/sunrpc/bc_svc.c deleted file mode 100644 index 15c7a8a1c..000000000 --- a/net/sunrpc/bc_svc.c +++ /dev/null @@ -1,63 +0,0 @@ -/****************************************************************************** - -(c) 2007 Network Appliance, Inc. All Rights Reserved. -(c) 2009 NetApp. All Rights Reserved. - -NetApp provides this source code under the GPL v2 License. -The GPL v2 license is available at -http://opensource.org/licenses/gpl-license.php. - -THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS -"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT -LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR -A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR -CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, -EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, -PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR -PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF -LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING -NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS -SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - -******************************************************************************/ - -/* - * The NFSv4.1 callback service helper routines. - * They implement the transport level processing required to send the - * reply over an existing open connection previously established by the client. - */ - -#include <linux/module.h> - -#include <linux/sunrpc/xprt.h> -#include <linux/sunrpc/sched.h> -#include <linux/sunrpc/bc_xprt.h> - -#define RPCDBG_FACILITY RPCDBG_SVCDSP - -/* Empty callback ops */ -static const struct rpc_call_ops nfs41_callback_ops = { -}; - - -/* - * Send the callback reply - */ -int bc_send(struct rpc_rqst *req) -{ - struct rpc_task *task; - int ret; - - dprintk("RPC: bc_send req= %p\n", req); - task = rpc_run_bc_task(req, &nfs41_callback_ops); - if (IS_ERR(task)) - ret = PTR_ERR(task); - else { - WARN_ON_ONCE(atomic_read(&task->tk_count) != 1); - ret = task->tk_status; - rpc_put_task(task); - } - dprintk("RPC: bc_send ret= %d\n", ret); - return ret; -} - diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index e6ce15173..23608eb0d 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -891,15 +891,8 @@ void rpc_task_set_client(struct rpc_task *task, struct rpc_clnt *clnt) task->tk_flags |= RPC_TASK_SOFT; if (clnt->cl_noretranstimeo) task->tk_flags |= RPC_TASK_NO_RETRANS_TIMEOUT; - if (sk_memalloc_socks()) { - struct rpc_xprt *xprt; - - rcu_read_lock(); - xprt = rcu_dereference(clnt->cl_xprt); - if (xprt->swapper) - task->tk_flags |= RPC_TASK_SWAPPER; - rcu_read_unlock(); - } + if (atomic_read(&clnt->cl_swapper)) + task->tk_flags |= RPC_TASK_SWAPPER; /* Add to the client's list of all tasks */ spin_lock(&clnt->cl_lock); list_add_tail(&task->tk_task, &clnt->cl_tasks); @@ -1031,15 +1024,14 @@ EXPORT_SYMBOL_GPL(rpc_call_async); * rpc_run_bc_task - Allocate a new RPC task for backchannel use, then run * rpc_execute against it * @req: RPC request - * @tk_ops: RPC call ops */ -struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req, - const struct rpc_call_ops *tk_ops) +struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req) { struct rpc_task *task; struct xdr_buf *xbufp = &req->rq_snd_buf; struct rpc_task_setup task_setup_data = { - .callback_ops = tk_ops, + .callback_ops = &rpc_default_ops, + .flags = RPC_TASK_SOFTCONN, }; dprintk("RPC: rpc_run_bc_task req= %p\n", req); @@ -1614,6 +1606,7 @@ call_allocate(struct rpc_task *task) req->rq_callsize + req->rq_rcvsize); if (req->rq_buffer != NULL) return; + xprt_inject_disconnect(xprt); dprintk("RPC: %5u rpc_buffer allocation failed\n", task->tk_pid); @@ -1909,6 +1902,7 @@ call_transmit_status(struct rpc_task *task) switch (task->tk_status) { case -EAGAIN: + case -ENOBUFS: break; default: dprint_status(task); @@ -1935,7 +1929,6 @@ call_transmit_status(struct rpc_task *task) case -ECONNABORTED: case -EADDRINUSE: case -ENOTCONN: - case -ENOBUFS: case -EPIPE: rpc_task_force_reencode(task); } @@ -1951,33 +1944,36 @@ call_bc_transmit(struct rpc_task *task) { struct rpc_rqst *req = task->tk_rqstp; - if (!xprt_prepare_transmit(task)) { - /* - * Could not reserve the transport. Try again after the - * transport is released. - */ - task->tk_status = 0; - task->tk_action = call_bc_transmit; - return; - } + if (!xprt_prepare_transmit(task)) + goto out_retry; - task->tk_action = rpc_exit_task; if (task->tk_status < 0) { printk(KERN_NOTICE "RPC: Could not send backchannel reply " "error: %d\n", task->tk_status); - return; + goto out_done; } + if (req->rq_connect_cookie != req->rq_xprt->connect_cookie) + req->rq_bytes_sent = 0; xprt_transmit(task); + + if (task->tk_status == -EAGAIN) + goto out_nospace; + xprt_end_transmit(task); dprint_status(task); switch (task->tk_status) { case 0: /* Success */ - break; case -EHOSTDOWN: case -EHOSTUNREACH: case -ENETUNREACH: + case -ECONNRESET: + case -ECONNREFUSED: + case -EADDRINUSE: + case -ENOTCONN: + case -EPIPE: + break; case -ETIMEDOUT: /* * Problem reaching the server. Disconnect and let the @@ -2002,6 +1998,13 @@ call_bc_transmit(struct rpc_task *task) break; } rpc_wake_up_queued_task(&req->rq_xprt->pending, task); +out_done: + task->tk_action = rpc_exit_task; + return; +out_nospace: + req->rq_connect_cookie = req->rq_xprt->connect_cookie; +out_retry: + task->tk_status = 0; } #endif /* CONFIG_SUNRPC_BACKCHANNEL */ @@ -2054,12 +2057,13 @@ call_status(struct rpc_task *task) case -ECONNABORTED: rpc_force_rebind(clnt); case -EADDRINUSE: - case -ENOBUFS: rpc_delay(task, 3*HZ); case -EPIPE: case -ENOTCONN: task->tk_action = call_bind; break; + case -ENOBUFS: + rpc_delay(task, HZ>>2); case -EAGAIN: task->tk_action = call_transmit; break; @@ -2476,3 +2480,59 @@ void rpc_show_tasks(struct net *net) spin_unlock(&sn->rpc_client_lock); } #endif + +#if IS_ENABLED(CONFIG_SUNRPC_SWAP) +int +rpc_clnt_swap_activate(struct rpc_clnt *clnt) +{ + int ret = 0; + struct rpc_xprt *xprt; + + if (atomic_inc_return(&clnt->cl_swapper) == 1) { +retry: + rcu_read_lock(); + xprt = xprt_get(rcu_dereference(clnt->cl_xprt)); + rcu_read_unlock(); + if (!xprt) { + /* + * If we didn't get a reference, then we likely are + * racing with a migration event. Wait for a grace + * period and try again. + */ + synchronize_rcu(); + goto retry; + } + + ret = xprt_enable_swap(xprt); + xprt_put(xprt); + } + return ret; +} +EXPORT_SYMBOL_GPL(rpc_clnt_swap_activate); + +void +rpc_clnt_swap_deactivate(struct rpc_clnt *clnt) +{ + struct rpc_xprt *xprt; + + if (atomic_dec_if_positive(&clnt->cl_swapper) == 0) { +retry: + rcu_read_lock(); + xprt = xprt_get(rcu_dereference(clnt->cl_xprt)); + rcu_read_unlock(); + if (!xprt) { + /* + * If we didn't get a reference, then we likely are + * racing with a migration event. Wait for a grace + * period and try again. + */ + synchronize_rcu(); + goto retry; + } + + xprt_disable_swap(xprt); + xprt_put(xprt); + } +} +EXPORT_SYMBOL_GPL(rpc_clnt_swap_deactivate); +#endif /* CONFIG_SUNRPC_SWAP */ diff --git a/net/sunrpc/debugfs.c b/net/sunrpc/debugfs.c index 82962f7e6..e7b4d9356 100644 --- a/net/sunrpc/debugfs.c +++ b/net/sunrpc/debugfs.c @@ -10,9 +10,12 @@ #include "netns.h" static struct dentry *topdir; +static struct dentry *rpc_fault_dir; static struct dentry *rpc_clnt_dir; static struct dentry *rpc_xprt_dir; +unsigned int rpc_inject_disconnect; + struct rpc_clnt_iter { struct rpc_clnt *clnt; loff_t pos; @@ -257,6 +260,8 @@ rpc_xprt_debugfs_register(struct rpc_xprt *xprt) debugfs_remove_recursive(xprt->debugfs); xprt->debugfs = NULL; } + + atomic_set(&xprt->inject_disconnect, rpc_inject_disconnect); } void @@ -266,11 +271,79 @@ rpc_xprt_debugfs_unregister(struct rpc_xprt *xprt) xprt->debugfs = NULL; } +static int +fault_open(struct inode *inode, struct file *filp) +{ + filp->private_data = kmalloc(128, GFP_KERNEL); + if (!filp->private_data) + return -ENOMEM; + return 0; +} + +static int +fault_release(struct inode *inode, struct file *filp) +{ + kfree(filp->private_data); + return 0; +} + +static ssize_t +fault_disconnect_read(struct file *filp, char __user *user_buf, + size_t len, loff_t *offset) +{ + char *buffer = (char *)filp->private_data; + size_t size; + + size = sprintf(buffer, "%u\n", rpc_inject_disconnect); + return simple_read_from_buffer(user_buf, len, offset, buffer, size); +} + +static ssize_t +fault_disconnect_write(struct file *filp, const char __user *user_buf, + size_t len, loff_t *offset) +{ + char buffer[16]; + + if (len >= sizeof(buffer)) + len = sizeof(buffer) - 1; + if (copy_from_user(buffer, user_buf, len)) + return -EFAULT; + buffer[len] = '\0'; + if (kstrtouint(buffer, 10, &rpc_inject_disconnect)) + return -EINVAL; + return len; +} + +static const struct file_operations fault_disconnect_fops = { + .owner = THIS_MODULE, + .open = fault_open, + .read = fault_disconnect_read, + .write = fault_disconnect_write, + .release = fault_release, +}; + +static struct dentry * +inject_fault_dir(struct dentry *topdir) +{ + struct dentry *faultdir; + + faultdir = debugfs_create_dir("inject_fault", topdir); + if (!faultdir) + return NULL; + + if (!debugfs_create_file("disconnect", S_IFREG | S_IRUSR, faultdir, + NULL, &fault_disconnect_fops)) + return NULL; + + return faultdir; +} + void __exit sunrpc_debugfs_exit(void) { debugfs_remove_recursive(topdir); topdir = NULL; + rpc_fault_dir = NULL; rpc_clnt_dir = NULL; rpc_xprt_dir = NULL; } @@ -282,6 +355,10 @@ sunrpc_debugfs_init(void) if (!topdir) return; + rpc_fault_dir = inject_fault_dir(topdir); + if (!rpc_fault_dir) + goto out_remove; + rpc_clnt_dir = debugfs_create_dir("rpc_clnt", topdir); if (!rpc_clnt_dir) goto out_remove; @@ -294,5 +371,6 @@ sunrpc_debugfs_init(void) out_remove: debugfs_remove_recursive(topdir); topdir = NULL; + rpc_fault_dir = NULL; rpc_clnt_dir = NULL; } diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 78974e4d9..5a16d8d8c 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -1290,7 +1290,6 @@ err_bad: svc_putnl(resv, ntohl(rpc_stat)); goto sendit; } -EXPORT_SYMBOL_GPL(svc_process); /* * Process the RPC request. @@ -1338,6 +1337,7 @@ out_drop: svc_drop(rqstp); return 0; } +EXPORT_SYMBOL_GPL(svc_process); #if defined(CONFIG_SUNRPC_BACKCHANNEL) /* @@ -1350,6 +1350,11 @@ bc_svc_process(struct svc_serv *serv, struct rpc_rqst *req, { struct kvec *argv = &rqstp->rq_arg.head[0]; struct kvec *resv = &rqstp->rq_res.head[0]; + struct rpc_task *task; + int proc_error; + int error; + + dprintk("svc: %s(%p)\n", __func__, req); /* Build the svc_rqst used by the common processing routine */ rqstp->rq_xprt = serv->sv_bc_xprt; @@ -1372,21 +1377,36 @@ bc_svc_process(struct svc_serv *serv, struct rpc_rqst *req, /* * Skip the next two words because they've already been - * processed in the trasport + * processed in the transport */ svc_getu32(argv); /* XID */ svc_getnl(argv); /* CALLDIR */ - /* Returns 1 for send, 0 for drop */ - if (svc_process_common(rqstp, argv, resv)) { - memcpy(&req->rq_snd_buf, &rqstp->rq_res, - sizeof(req->rq_snd_buf)); - return bc_send(req); - } else { - /* drop request */ + /* Parse and execute the bc call */ + proc_error = svc_process_common(rqstp, argv, resv); + + atomic_inc(&req->rq_xprt->bc_free_slots); + if (!proc_error) { + /* Processing error: drop the request */ xprt_free_bc_request(req); return 0; } + + /* Finally, send the reply synchronously */ + memcpy(&req->rq_snd_buf, &rqstp->rq_res, sizeof(req->rq_snd_buf)); + task = rpc_run_bc_task(req); + if (IS_ERR(task)) { + error = PTR_ERR(task); + goto out; + } + + WARN_ON_ONCE(atomic_read(&task->tk_count) != 1); + error = task->tk_status; + rpc_put_task(task); + +out: + dprintk("svc: %s(), error=%d\n", __func__, error); + return error; } EXPORT_SYMBOL_GPL(bc_svc_process); #endif /* CONFIG_SUNRPC_BACKCHANNEL */ diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 1d4fe24af..ab5dd621a 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -68,6 +68,7 @@ static void xprt_init(struct rpc_xprt *xprt, struct net *net); static void xprt_request_init(struct rpc_task *, struct rpc_xprt *); static void xprt_connect_status(struct rpc_task *task); static int __xprt_get_cong(struct rpc_xprt *, struct rpc_task *); +static void __xprt_put_cong(struct rpc_xprt *, struct rpc_rqst *); static void xprt_destroy(struct rpc_xprt *xprt); static DEFINE_SPINLOCK(xprt_list_lock); @@ -250,6 +251,8 @@ int xprt_reserve_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task) } xprt_clear_locked(xprt); out_sleep: + if (req) + __xprt_put_cong(xprt, req); dprintk("RPC: %5u failed to lock transport %p\n", task->tk_pid, xprt); task->tk_timeout = 0; task->tk_status = -EAGAIN; @@ -608,8 +611,8 @@ static void xprt_autoclose(struct work_struct *work) struct rpc_xprt *xprt = container_of(work, struct rpc_xprt, task_cleanup); - xprt->ops->close(xprt); clear_bit(XPRT_CLOSE_WAIT, &xprt->state); + xprt->ops->close(xprt); xprt_release_write(xprt, NULL); } @@ -967,6 +970,7 @@ void xprt_transmit(struct rpc_task *task) task->tk_status = status; return; } + xprt_inject_disconnect(xprt); dprintk("RPC: %5u xmit complete\n", task->tk_pid); task->tk_flags |= RPC_TASK_SENT; @@ -1285,6 +1289,7 @@ void xprt_release(struct rpc_task *task) spin_unlock_bh(&xprt->transport_lock); if (req->rq_buffer) xprt->ops->buf_free(req->rq_buffer); + xprt_inject_disconnect(xprt); if (req->rq_cred != NULL) put_rpccred(req->rq_cred); task->tk_rqstp = NULL; diff --git a/net/sunrpc/xprtrdma/Makefile b/net/sunrpc/xprtrdma/Makefile index 579f72bbc..48913de24 100644 --- a/net/sunrpc/xprtrdma/Makefile +++ b/net/sunrpc/xprtrdma/Makefile @@ -1,9 +1,7 @@ -obj-$(CONFIG_SUNRPC_XPRT_RDMA_CLIENT) += xprtrdma.o +obj-$(CONFIG_SUNRPC_XPRT_RDMA) += rpcrdma.o -xprtrdma-y := transport.o rpc_rdma.o verbs.o \ - fmr_ops.o frwr_ops.o physical_ops.o - -obj-$(CONFIG_SUNRPC_XPRT_RDMA_SERVER) += svcrdma.o - -svcrdma-y := svc_rdma.o svc_rdma_transport.o \ - svc_rdma_marshal.o svc_rdma_sendto.o svc_rdma_recvfrom.o +rpcrdma-y := transport.o rpc_rdma.o verbs.o \ + fmr_ops.o frwr_ops.o physical_ops.o \ + svc_rdma.o svc_rdma_transport.o \ + svc_rdma_marshal.o svc_rdma_sendto.o svc_rdma_recvfrom.o \ + module.o diff --git a/net/sunrpc/xprtrdma/fmr_ops.c b/net/sunrpc/xprtrdma/fmr_ops.c index 302d4ebf6..f1e8dafbd 100644 --- a/net/sunrpc/xprtrdma/fmr_ops.c +++ b/net/sunrpc/xprtrdma/fmr_ops.c @@ -11,6 +11,21 @@ * can take tens of usecs to complete. */ +/* Normal operation + * + * A Memory Region is prepared for RDMA READ or WRITE using the + * ib_map_phys_fmr verb (fmr_op_map). When the RDMA operation is + * finished, the Memory Region is unmapped using the ib_unmap_fmr + * verb (fmr_op_unmap). + */ + +/* Transport recovery + * + * After a transport reconnect, fmr_op_map re-uses the MR already + * allocated for the RPC, but generates a fresh rkey then maps the + * MR again. This process is synchronous. + */ + #include "xprt_rdma.h" #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) @@ -50,19 +65,28 @@ fmr_op_init(struct rpcrdma_xprt *r_xprt) struct rpcrdma_mw *r; int i, rc; + spin_lock_init(&buf->rb_mwlock); INIT_LIST_HEAD(&buf->rb_mws); INIT_LIST_HEAD(&buf->rb_all); - i = (buf->rb_max_requests + 1) * RPCRDMA_MAX_SEGS; - dprintk("RPC: %s: initializing %d FMRs\n", __func__, i); + i = max_t(int, RPCRDMA_MAX_DATA_SEGS / RPCRDMA_MAX_FMR_SGES, 1); + i += 2; /* head + tail */ + i *= buf->rb_max_requests; /* one set for each RPC slot */ + dprintk("RPC: %s: initalizing %d FMRs\n", __func__, i); + rc = -ENOMEM; while (i--) { r = kzalloc(sizeof(*r), GFP_KERNEL); if (!r) - return -ENOMEM; + goto out; - r->r.fmr = ib_alloc_fmr(pd, mr_access_flags, &fmr_attr); - if (IS_ERR(r->r.fmr)) + r->r.fmr.physaddrs = kmalloc(RPCRDMA_MAX_FMR_SGES * + sizeof(u64), GFP_KERNEL); + if (!r->r.fmr.physaddrs) + goto out_free; + + r->r.fmr.fmr = ib_alloc_fmr(pd, mr_access_flags, &fmr_attr); + if (IS_ERR(r->r.fmr.fmr)) goto out_fmr_err; list_add(&r->mw_list, &buf->rb_mws); @@ -71,12 +95,24 @@ fmr_op_init(struct rpcrdma_xprt *r_xprt) return 0; out_fmr_err: - rc = PTR_ERR(r->r.fmr); + rc = PTR_ERR(r->r.fmr.fmr); dprintk("RPC: %s: ib_alloc_fmr status %i\n", __func__, rc); + kfree(r->r.fmr.physaddrs); +out_free: kfree(r); +out: return rc; } +static int +__fmr_unmap(struct rpcrdma_mw *r) +{ + LIST_HEAD(l); + + list_add(&r->r.fmr.fmr->list, &l); + return ib_unmap_fmr(&l); +} + /* Use the ib_map_phys_fmr() verb to register a memory region * for remote access via RDMA READ or RDMA WRITE. */ @@ -85,12 +121,24 @@ fmr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, int nsegs, bool writing) { struct rpcrdma_ia *ia = &r_xprt->rx_ia; - struct ib_device *device = ia->ri_id->device; + struct ib_device *device = ia->ri_device; enum dma_data_direction direction = rpcrdma_data_dir(writing); struct rpcrdma_mr_seg *seg1 = seg; - struct rpcrdma_mw *mw = seg1->rl_mw; - u64 physaddrs[RPCRDMA_MAX_DATA_SEGS]; int len, pageoff, i, rc; + struct rpcrdma_mw *mw; + + mw = seg1->rl_mw; + seg1->rl_mw = NULL; + if (!mw) { + mw = rpcrdma_get_mw(r_xprt); + if (!mw) + return -ENOMEM; + } else { + /* this is a retransmit; generate a fresh rkey */ + rc = __fmr_unmap(mw); + if (rc) + return rc; + } pageoff = offset_in_page(seg1->mr_offset); seg1->mr_offset -= pageoff; /* start of page */ @@ -100,7 +148,7 @@ fmr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, nsegs = RPCRDMA_MAX_FMR_SGES; for (i = 0; i < nsegs;) { rpcrdma_map_one(device, seg, direction); - physaddrs[i] = seg->mr_dma; + mw->r.fmr.physaddrs[i] = seg->mr_dma; len += seg->mr_len; ++seg; ++i; @@ -110,11 +158,13 @@ fmr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, break; } - rc = ib_map_phys_fmr(mw->r.fmr, physaddrs, i, seg1->mr_dma); + rc = ib_map_phys_fmr(mw->r.fmr.fmr, mw->r.fmr.physaddrs, + i, seg1->mr_dma); if (rc) goto out_maperr; - seg1->mr_rkey = mw->r.fmr->rkey; + seg1->rl_mw = mw; + seg1->mr_rkey = mw->r.fmr.fmr->rkey; seg1->mr_base = seg1->mr_dma + pageoff; seg1->mr_nsegs = i; seg1->mr_len = len; @@ -137,48 +187,28 @@ fmr_op_unmap(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg) { struct rpcrdma_ia *ia = &r_xprt->rx_ia; struct rpcrdma_mr_seg *seg1 = seg; - struct ib_device *device; + struct rpcrdma_mw *mw = seg1->rl_mw; int rc, nsegs = seg->mr_nsegs; - LIST_HEAD(l); - list_add(&seg1->rl_mw->r.fmr->list, &l); - rc = ib_unmap_fmr(&l); - read_lock(&ia->ri_qplock); - device = ia->ri_id->device; + dprintk("RPC: %s: FMR %p\n", __func__, mw); + + seg1->rl_mw = NULL; while (seg1->mr_nsegs--) - rpcrdma_unmap_one(device, seg++); - read_unlock(&ia->ri_qplock); + rpcrdma_unmap_one(ia->ri_device, seg++); + rc = __fmr_unmap(mw); if (rc) goto out_err; + rpcrdma_put_mw(r_xprt, mw); return nsegs; out_err: + /* The FMR is abandoned, but remains in rb_all. fmr_op_destroy + * will attempt to release it when the transport is destroyed. + */ dprintk("RPC: %s: ib_unmap_fmr status %i\n", __func__, rc); return nsegs; } -/* After a disconnect, unmap all FMRs. - * - * This is invoked only in the transport connect worker in order - * to serialize with rpcrdma_register_fmr_external(). - */ -static void -fmr_op_reset(struct rpcrdma_xprt *r_xprt) -{ - struct rpcrdma_buffer *buf = &r_xprt->rx_buf; - struct rpcrdma_mw *r; - LIST_HEAD(list); - int rc; - - list_for_each_entry(r, &buf->rb_all, mw_all) - list_add(&r->r.fmr->list, &list); - - rc = ib_unmap_fmr(&list); - if (rc) - dprintk("RPC: %s: ib_unmap_fmr failed %i\n", - __func__, rc); -} - static void fmr_op_destroy(struct rpcrdma_buffer *buf) { @@ -188,10 +218,13 @@ fmr_op_destroy(struct rpcrdma_buffer *buf) while (!list_empty(&buf->rb_all)) { r = list_entry(buf->rb_all.next, struct rpcrdma_mw, mw_all); list_del(&r->mw_all); - rc = ib_dealloc_fmr(r->r.fmr); + kfree(r->r.fmr.physaddrs); + + rc = ib_dealloc_fmr(r->r.fmr.fmr); if (rc) dprintk("RPC: %s: ib_dealloc_fmr failed %i\n", __func__, rc); + kfree(r); } } @@ -202,7 +235,6 @@ const struct rpcrdma_memreg_ops rpcrdma_fmr_memreg_ops = { .ro_open = fmr_op_open, .ro_maxpages = fmr_op_maxpages, .ro_init = fmr_op_init, - .ro_reset = fmr_op_reset, .ro_destroy = fmr_op_destroy, .ro_displayname = "fmr", }; diff --git a/net/sunrpc/xprtrdma/frwr_ops.c b/net/sunrpc/xprtrdma/frwr_ops.c index dff0481db..04ea91420 100644 --- a/net/sunrpc/xprtrdma/frwr_ops.c +++ b/net/sunrpc/xprtrdma/frwr_ops.c @@ -11,12 +11,136 @@ * but most complex memory registration mode. */ +/* Normal operation + * + * A Memory Region is prepared for RDMA READ or WRITE using a FAST_REG + * Work Request (frmr_op_map). When the RDMA operation is finished, this + * Memory Region is invalidated using a LOCAL_INV Work Request + * (frmr_op_unmap). + * + * Typically these Work Requests are not signaled, and neither are RDMA + * SEND Work Requests (with the exception of signaling occasionally to + * prevent provider work queue overflows). This greatly reduces HCA + * interrupt workload. + * + * As an optimization, frwr_op_unmap marks MRs INVALID before the + * LOCAL_INV WR is posted. If posting succeeds, the MR is placed on + * rb_mws immediately so that no work (like managing a linked list + * under a spinlock) is needed in the completion upcall. + * + * But this means that frwr_op_map() can occasionally encounter an MR + * that is INVALID but the LOCAL_INV WR has not completed. Work Queue + * ordering prevents a subsequent FAST_REG WR from executing against + * that MR while it is still being invalidated. + */ + +/* Transport recovery + * + * ->op_map and the transport connect worker cannot run at the same + * time, but ->op_unmap can fire while the transport connect worker + * is running. Thus MR recovery is handled in ->op_map, to guarantee + * that recovered MRs are owned by a sending RPC, and not one where + * ->op_unmap could fire at the same time transport reconnect is + * being done. + * + * When the underlying transport disconnects, MRs are left in one of + * three states: + * + * INVALID: The MR was not in use before the QP entered ERROR state. + * (Or, the LOCAL_INV WR has not completed or flushed yet). + * + * STALE: The MR was being registered or unregistered when the QP + * entered ERROR state, and the pending WR was flushed. + * + * VALID: The MR was registered before the QP entered ERROR state. + * + * When frwr_op_map encounters STALE and VALID MRs, they are recovered + * with ib_dereg_mr and then are re-initialized. Beause MR recovery + * allocates fresh resources, it is deferred to a workqueue, and the + * recovered MRs are placed back on the rb_mws list when recovery is + * complete. frwr_op_map allocates another MR for the current RPC while + * the broken MR is reset. + * + * To ensure that frwr_op_map doesn't encounter an MR that is marked + * INVALID but that is about to be flushed due to a previous transport + * disconnect, the transport connect worker attempts to drain all + * pending send queue WRs before the transport is reconnected. + */ + #include "xprt_rdma.h" #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) # define RPCDBG_FACILITY RPCDBG_TRANS #endif +static struct workqueue_struct *frwr_recovery_wq; + +#define FRWR_RECOVERY_WQ_FLAGS (WQ_UNBOUND | WQ_MEM_RECLAIM) + +int +frwr_alloc_recovery_wq(void) +{ + frwr_recovery_wq = alloc_workqueue("frwr_recovery", + FRWR_RECOVERY_WQ_FLAGS, 0); + return !frwr_recovery_wq ? -ENOMEM : 0; +} + +void +frwr_destroy_recovery_wq(void) +{ + struct workqueue_struct *wq; + + if (!frwr_recovery_wq) + return; + + wq = frwr_recovery_wq; + frwr_recovery_wq = NULL; + destroy_workqueue(wq); +} + +/* Deferred reset of a single FRMR. Generate a fresh rkey by + * replacing the MR. + * + * There's no recovery if this fails. The FRMR is abandoned, but + * remains in rb_all. It will be cleaned up when the transport is + * destroyed. + */ +static void +__frwr_recovery_worker(struct work_struct *work) +{ + struct rpcrdma_mw *r = container_of(work, struct rpcrdma_mw, + r.frmr.fr_work); + struct rpcrdma_xprt *r_xprt = r->r.frmr.fr_xprt; + unsigned int depth = r_xprt->rx_ia.ri_max_frmr_depth; + struct ib_pd *pd = r_xprt->rx_ia.ri_pd; + + if (ib_dereg_mr(r->r.frmr.fr_mr)) + goto out_fail; + + r->r.frmr.fr_mr = ib_alloc_fast_reg_mr(pd, depth); + if (IS_ERR(r->r.frmr.fr_mr)) + goto out_fail; + + dprintk("RPC: %s: recovered FRMR %p\n", __func__, r); + r->r.frmr.fr_state = FRMR_IS_INVALID; + rpcrdma_put_mw(r_xprt, r); + return; + +out_fail: + pr_warn("RPC: %s: FRMR %p unrecovered\n", + __func__, r); +} + +/* A broken MR was discovered in a context that can't sleep. + * Defer recovery to the recovery worker. + */ +static void +__frwr_queue_recovery(struct rpcrdma_mw *r) +{ + INIT_WORK(&r->r.frmr.fr_work, __frwr_recovery_worker); + queue_work(frwr_recovery_wq, &r->r.frmr.fr_work); +} + static int __frwr_init(struct rpcrdma_mw *r, struct ib_pd *pd, struct ib_device *device, unsigned int depth) @@ -128,8 +252,8 @@ frwr_sendcompletion(struct ib_wc *wc) /* WARNING: Only wr_id and status are reliable at this point */ r = (struct rpcrdma_mw *)(unsigned long)wc->wr_id; - dprintk("RPC: %s: frmr %p (stale), status %d\n", - __func__, r, wc->status); + pr_warn("RPC: %s: frmr %p flushed, status %s (%d)\n", + __func__, r, ib_wc_status_msg(wc->status), wc->status); r->r.frmr.fr_state = FRMR_IS_STALE; } @@ -137,16 +261,19 @@ static int frwr_op_init(struct rpcrdma_xprt *r_xprt) { struct rpcrdma_buffer *buf = &r_xprt->rx_buf; - struct ib_device *device = r_xprt->rx_ia.ri_id->device; + struct ib_device *device = r_xprt->rx_ia.ri_device; unsigned int depth = r_xprt->rx_ia.ri_max_frmr_depth; struct ib_pd *pd = r_xprt->rx_ia.ri_pd; int i; + spin_lock_init(&buf->rb_mwlock); INIT_LIST_HEAD(&buf->rb_mws); INIT_LIST_HEAD(&buf->rb_all); - i = (buf->rb_max_requests + 1) * RPCRDMA_MAX_SEGS; - dprintk("RPC: %s: initializing %d FRMRs\n", __func__, i); + i = max_t(int, RPCRDMA_MAX_DATA_SEGS / depth, 1); + i += 2; /* head + tail */ + i *= buf->rb_max_requests; /* one set for each RPC slot */ + dprintk("RPC: %s: initalizing %d FRMRs\n", __func__, i); while (i--) { struct rpcrdma_mw *r; @@ -165,6 +292,7 @@ frwr_op_init(struct rpcrdma_xprt *r_xprt) list_add(&r->mw_list, &buf->rb_mws); list_add(&r->mw_all, &buf->rb_all); r->mw_sendcompletion = frwr_sendcompletion; + r->r.frmr.fr_xprt = r_xprt; } return 0; @@ -178,12 +306,12 @@ frwr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, int nsegs, bool writing) { struct rpcrdma_ia *ia = &r_xprt->rx_ia; - struct ib_device *device = ia->ri_id->device; + struct ib_device *device = ia->ri_device; enum dma_data_direction direction = rpcrdma_data_dir(writing); struct rpcrdma_mr_seg *seg1 = seg; - struct rpcrdma_mw *mw = seg1->rl_mw; - struct rpcrdma_frmr *frmr = &mw->r.frmr; - struct ib_mr *mr = frmr->fr_mr; + struct rpcrdma_mw *mw; + struct rpcrdma_frmr *frmr; + struct ib_mr *mr; struct ib_send_wr fastreg_wr, *bad_wr; u8 key; int len, pageoff; @@ -192,12 +320,25 @@ frwr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, u64 pa; int page_no; + mw = seg1->rl_mw; + seg1->rl_mw = NULL; + do { + if (mw) + __frwr_queue_recovery(mw); + mw = rpcrdma_get_mw(r_xprt); + if (!mw) + return -ENOMEM; + } while (mw->r.frmr.fr_state != FRMR_IS_INVALID); + frmr = &mw->r.frmr; + frmr->fr_state = FRMR_IS_VALID; + pageoff = offset_in_page(seg1->mr_offset); seg1->mr_offset -= pageoff; /* start of page */ seg1->mr_len += pageoff; len = -pageoff; if (nsegs > ia->ri_max_frmr_depth) nsegs = ia->ri_max_frmr_depth; + for (page_no = i = 0; i < nsegs;) { rpcrdma_map_one(device, seg, direction); pa = seg->mr_dma; @@ -216,8 +357,6 @@ frwr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, dprintk("RPC: %s: Using frmr %p to map %d segments (%d bytes)\n", __func__, mw, i, len); - frmr->fr_state = FRMR_IS_VALID; - memset(&fastreg_wr, 0, sizeof(fastreg_wr)); fastreg_wr.wr_id = (unsigned long)(void *)mw; fastreg_wr.opcode = IB_WR_FAST_REG_MR; @@ -229,6 +368,7 @@ frwr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, fastreg_wr.wr.fast_reg.access_flags = writing ? IB_ACCESS_REMOTE_WRITE | IB_ACCESS_LOCAL_WRITE : IB_ACCESS_REMOTE_READ; + mr = frmr->fr_mr; key = (u8)(mr->rkey & 0x000000FF); ib_update_fast_reg_key(mr, ++key); fastreg_wr.wr.fast_reg.rkey = mr->rkey; @@ -238,6 +378,7 @@ frwr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, if (rc) goto out_senderr; + seg1->rl_mw = mw; seg1->mr_rkey = mr->rkey; seg1->mr_base = seg1->mr_dma + pageoff; seg1->mr_nsegs = i; @@ -246,10 +387,9 @@ frwr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, out_senderr: dprintk("RPC: %s: ib_post_send status %i\n", __func__, rc); - ib_update_fast_reg_key(mr, --key); - frmr->fr_state = FRMR_IS_INVALID; while (i--) rpcrdma_unmap_one(device, --seg); + __frwr_queue_recovery(mw); return rc; } @@ -261,78 +401,46 @@ frwr_op_unmap(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg) { struct rpcrdma_mr_seg *seg1 = seg; struct rpcrdma_ia *ia = &r_xprt->rx_ia; + struct rpcrdma_mw *mw = seg1->rl_mw; struct ib_send_wr invalidate_wr, *bad_wr; int rc, nsegs = seg->mr_nsegs; - struct ib_device *device; - seg1->rl_mw->r.frmr.fr_state = FRMR_IS_INVALID; + dprintk("RPC: %s: FRMR %p\n", __func__, mw); + + seg1->rl_mw = NULL; + mw->r.frmr.fr_state = FRMR_IS_INVALID; memset(&invalidate_wr, 0, sizeof(invalidate_wr)); - invalidate_wr.wr_id = (unsigned long)(void *)seg1->rl_mw; + invalidate_wr.wr_id = (unsigned long)(void *)mw; invalidate_wr.opcode = IB_WR_LOCAL_INV; - invalidate_wr.ex.invalidate_rkey = seg1->rl_mw->r.frmr.fr_mr->rkey; + invalidate_wr.ex.invalidate_rkey = mw->r.frmr.fr_mr->rkey; DECR_CQCOUNT(&r_xprt->rx_ep); - read_lock(&ia->ri_qplock); - device = ia->ri_id->device; while (seg1->mr_nsegs--) - rpcrdma_unmap_one(device, seg++); + rpcrdma_unmap_one(ia->ri_device, seg++); + read_lock(&ia->ri_qplock); rc = ib_post_send(ia->ri_id->qp, &invalidate_wr, &bad_wr); read_unlock(&ia->ri_qplock); if (rc) goto out_err; + + rpcrdma_put_mw(r_xprt, mw); return nsegs; out_err: - /* Force rpcrdma_buffer_get() to retry */ - seg1->rl_mw->r.frmr.fr_state = FRMR_IS_STALE; dprintk("RPC: %s: ib_post_send status %i\n", __func__, rc); + __frwr_queue_recovery(mw); return nsegs; } -/* After a disconnect, a flushed FAST_REG_MR can leave an FRMR in - * an unusable state. Find FRMRs in this state and dereg / reg - * each. FRMRs that are VALID and attached to an rpcrdma_req are - * also torn down. - * - * This gives all in-use FRMRs a fresh rkey and leaves them INVALID. - * - * This is invoked only in the transport connect worker in order - * to serialize with rpcrdma_register_frmr_external(). - */ -static void -frwr_op_reset(struct rpcrdma_xprt *r_xprt) -{ - struct rpcrdma_buffer *buf = &r_xprt->rx_buf; - struct ib_device *device = r_xprt->rx_ia.ri_id->device; - unsigned int depth = r_xprt->rx_ia.ri_max_frmr_depth; - struct ib_pd *pd = r_xprt->rx_ia.ri_pd; - struct rpcrdma_mw *r; - int rc; - - list_for_each_entry(r, &buf->rb_all, mw_all) { - if (r->r.frmr.fr_state == FRMR_IS_INVALID) - continue; - - __frwr_release(r); - rc = __frwr_init(r, pd, device, depth); - if (rc) { - dprintk("RPC: %s: mw %p left %s\n", - __func__, r, - (r->r.frmr.fr_state == FRMR_IS_STALE ? - "stale" : "valid")); - continue; - } - - r->r.frmr.fr_state = FRMR_IS_INVALID; - } -} - static void frwr_op_destroy(struct rpcrdma_buffer *buf) { struct rpcrdma_mw *r; + /* Ensure stale MWs for "buf" are no longer in flight */ + flush_workqueue(frwr_recovery_wq); + while (!list_empty(&buf->rb_all)) { r = list_entry(buf->rb_all.next, struct rpcrdma_mw, mw_all); list_del(&r->mw_all); @@ -347,7 +455,6 @@ const struct rpcrdma_memreg_ops rpcrdma_frwr_memreg_ops = { .ro_open = frwr_op_open, .ro_maxpages = frwr_op_maxpages, .ro_init = frwr_op_init, - .ro_reset = frwr_op_reset, .ro_destroy = frwr_op_destroy, .ro_displayname = "frwr", }; diff --git a/net/sunrpc/xprtrdma/module.c b/net/sunrpc/xprtrdma/module.c new file mode 100644 index 000000000..560712bd9 --- /dev/null +++ b/net/sunrpc/xprtrdma/module.c @@ -0,0 +1,46 @@ +/* + * Copyright (c) 2015 Oracle. All rights reserved. + */ + +/* rpcrdma.ko module initialization + */ + +#include <linux/module.h> +#include <linux/init.h> +#include <linux/sunrpc/svc_rdma.h> +#include "xprt_rdma.h" + +#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) +# define RPCDBG_FACILITY RPCDBG_TRANS +#endif + +MODULE_AUTHOR("Open Grid Computing and Network Appliance, Inc."); +MODULE_DESCRIPTION("RPC/RDMA Transport"); +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_ALIAS("svcrdma"); +MODULE_ALIAS("xprtrdma"); + +static void __exit rpc_rdma_cleanup(void) +{ + xprt_rdma_cleanup(); + svc_rdma_cleanup(); +} + +static int __init rpc_rdma_init(void) +{ + int rc; + + rc = svc_rdma_init(); + if (rc) + goto out; + + rc = xprt_rdma_init(); + if (rc) + svc_rdma_cleanup(); + +out: + return rc; +} + +module_init(rpc_rdma_init); +module_exit(rpc_rdma_cleanup); diff --git a/net/sunrpc/xprtrdma/physical_ops.c b/net/sunrpc/xprtrdma/physical_ops.c index ba518af16..41985d07f 100644 --- a/net/sunrpc/xprtrdma/physical_ops.c +++ b/net/sunrpc/xprtrdma/physical_ops.c @@ -50,8 +50,7 @@ physical_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, { struct rpcrdma_ia *ia = &r_xprt->rx_ia; - rpcrdma_map_one(ia->ri_id->device, seg, - rpcrdma_data_dir(writing)); + rpcrdma_map_one(ia->ri_device, seg, rpcrdma_data_dir(writing)); seg->mr_rkey = ia->ri_bind_mem->rkey; seg->mr_base = seg->mr_dma; seg->mr_nsegs = 1; @@ -65,19 +64,11 @@ physical_op_unmap(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg) { struct rpcrdma_ia *ia = &r_xprt->rx_ia; - read_lock(&ia->ri_qplock); - rpcrdma_unmap_one(ia->ri_id->device, seg); - read_unlock(&ia->ri_qplock); - + rpcrdma_unmap_one(ia->ri_device, seg); return 1; } static void -physical_op_reset(struct rpcrdma_xprt *r_xprt) -{ -} - -static void physical_op_destroy(struct rpcrdma_buffer *buf) { } @@ -88,7 +79,6 @@ const struct rpcrdma_memreg_ops rpcrdma_physical_memreg_ops = { .ro_open = physical_op_open, .ro_maxpages = physical_op_maxpages, .ro_init = physical_op_init, - .ro_reset = physical_op_reset, .ro_destroy = physical_op_destroy, .ro_displayname = "physical", }; diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c index 2c53ea9e1..84ea37dae 100644 --- a/net/sunrpc/xprtrdma/rpc_rdma.c +++ b/net/sunrpc/xprtrdma/rpc_rdma.c @@ -284,9 +284,6 @@ rpcrdma_create_chunks(struct rpc_rqst *rqst, struct xdr_buf *target, return (unsigned char *)iptr - (unsigned char *)headerp; out: - if (r_xprt->rx_ia.ri_memreg_strategy == RPCRDMA_FRMR) - return n; - for (pos = 0; nchunks--;) pos += r_xprt->rx_ia.ri_ops->ro_unmap(r_xprt, &req->rl_segments[pos]); @@ -732,8 +729,8 @@ rpcrdma_reply_handler(struct rpcrdma_rep *rep) struct rpcrdma_msg *headerp; struct rpcrdma_req *req; struct rpc_rqst *rqst; - struct rpc_xprt *xprt = rep->rr_xprt; - struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt); + struct rpcrdma_xprt *r_xprt = rep->rr_rxprt; + struct rpc_xprt *xprt = &r_xprt->rx_xprt; __be32 *iptr; int rdmalen, status; unsigned long cwnd; @@ -770,7 +767,6 @@ rpcrdma_reply_handler(struct rpcrdma_rep *rep) rep->rr_len); repost: r_xprt->rx_stats.bad_reply_count++; - rep->rr_func = rpcrdma_reply_handler; if (rpcrdma_ep_post_recv(&r_xprt->rx_ia, &r_xprt->rx_ep, rep)) rpcrdma_recv_buffer_put(rep); diff --git a/net/sunrpc/xprtrdma/svc_rdma.c b/net/sunrpc/xprtrdma/svc_rdma.c index c1b627026..2cd252f02 100644 --- a/net/sunrpc/xprtrdma/svc_rdma.c +++ b/net/sunrpc/xprtrdma/svc_rdma.c @@ -38,8 +38,7 @@ * * Author: Tom Tucker <tom@opengridcomputing.com> */ -#include <linux/module.h> -#include <linux/init.h> + #include <linux/slab.h> #include <linux/fs.h> #include <linux/sysctl.h> @@ -295,8 +294,3 @@ int svc_rdma_init(void) destroy_workqueue(svc_rdma_wq); return -ENOMEM; } -MODULE_AUTHOR("Tom Tucker <tom@opengridcomputing.com>"); -MODULE_DESCRIPTION("SVC RDMA Transport"); -MODULE_LICENSE("Dual BSD/GPL"); -module_init(svc_rdma_init); -module_exit(svc_rdma_cleanup); diff --git a/net/sunrpc/xprtrdma/svc_rdma_marshal.c b/net/sunrpc/xprtrdma/svc_rdma_marshal.c index b681855cf..e2fca7617 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_marshal.c +++ b/net/sunrpc/xprtrdma/svc_rdma_marshal.c @@ -50,12 +50,12 @@ /* * Decodes a read chunk list. The expected format is as follows: * descrim : xdr_one - * position : u32 offset into XDR stream - * handle : u32 RKEY + * position : __be32 offset into XDR stream + * handle : __be32 RKEY * . . . * end-of-list: xdr_zero */ -static u32 *decode_read_list(u32 *va, u32 *vaend) +static __be32 *decode_read_list(__be32 *va, __be32 *vaend) { struct rpcrdma_read_chunk *ch = (struct rpcrdma_read_chunk *)va; @@ -67,20 +67,20 @@ static u32 *decode_read_list(u32 *va, u32 *vaend) } ch++; } - return (u32 *)&ch->rc_position; + return &ch->rc_position; } /* * Decodes a write chunk list. The expected format is as follows: * descrim : xdr_one * nchunks : <count> - * handle : u32 RKEY ---+ - * length : u32 <len of segment> | + * handle : __be32 RKEY ---+ + * length : __be32 <len of segment> | * offset : remove va + <count> * . . . | * ---+ */ -static u32 *decode_write_list(u32 *va, u32 *vaend) +static __be32 *decode_write_list(__be32 *va, __be32 *vaend) { unsigned long start, end; int nchunks; @@ -90,14 +90,14 @@ static u32 *decode_write_list(u32 *va, u32 *vaend) /* Check for not write-array */ if (ary->wc_discrim == xdr_zero) - return (u32 *)&ary->wc_nchunks; + return &ary->wc_nchunks; if ((unsigned long)ary + sizeof(struct rpcrdma_write_array) > (unsigned long)vaend) { dprintk("svcrdma: ary=%p, vaend=%p\n", ary, vaend); return NULL; } - nchunks = ntohl(ary->wc_nchunks); + nchunks = be32_to_cpu(ary->wc_nchunks); start = (unsigned long)&ary->wc_array[0]; end = (unsigned long)vaend; @@ -112,10 +112,10 @@ static u32 *decode_write_list(u32 *va, u32 *vaend) * rs_length is the 2nd 4B field in wc_target and taking its * address skips the list terminator */ - return (u32 *)&ary->wc_array[nchunks].wc_target.rs_length; + return &ary->wc_array[nchunks].wc_target.rs_length; } -static u32 *decode_reply_array(u32 *va, u32 *vaend) +static __be32 *decode_reply_array(__be32 *va, __be32 *vaend) { unsigned long start, end; int nchunks; @@ -124,14 +124,14 @@ static u32 *decode_reply_array(u32 *va, u32 *vaend) /* Check for no reply-array */ if (ary->wc_discrim == xdr_zero) - return (u32 *)&ary->wc_nchunks; + return &ary->wc_nchunks; if ((unsigned long)ary + sizeof(struct rpcrdma_write_array) > (unsigned long)vaend) { dprintk("svcrdma: ary=%p, vaend=%p\n", ary, vaend); return NULL; } - nchunks = ntohl(ary->wc_nchunks); + nchunks = be32_to_cpu(ary->wc_nchunks); start = (unsigned long)&ary->wc_array[0]; end = (unsigned long)vaend; @@ -142,15 +142,14 @@ static u32 *decode_reply_array(u32 *va, u32 *vaend) ary, nchunks, vaend); return NULL; } - return (u32 *)&ary->wc_array[nchunks]; + return (__be32 *)&ary->wc_array[nchunks]; } int svc_rdma_xdr_decode_req(struct rpcrdma_msg **rdma_req, struct svc_rqst *rqstp) { struct rpcrdma_msg *rmsgp = NULL; - u32 *va; - u32 *vaend; + __be32 *va, *vaend; u32 hdr_len; rmsgp = (struct rpcrdma_msg *)rqstp->rq_arg.head[0].iov_base; @@ -162,22 +161,17 @@ int svc_rdma_xdr_decode_req(struct rpcrdma_msg **rdma_req, return -EINVAL; } - /* Decode the header */ - rmsgp->rm_xid = ntohl(rmsgp->rm_xid); - rmsgp->rm_vers = ntohl(rmsgp->rm_vers); - rmsgp->rm_credit = ntohl(rmsgp->rm_credit); - rmsgp->rm_type = ntohl(rmsgp->rm_type); - - if (rmsgp->rm_vers != RPCRDMA_VERSION) + if (rmsgp->rm_vers != rpcrdma_version) return -ENOSYS; /* Pull in the extra for the padded case and bump our pointer */ - if (rmsgp->rm_type == RDMA_MSGP) { + if (rmsgp->rm_type == rdma_msgp) { int hdrlen; + rmsgp->rm_body.rm_padded.rm_align = - ntohl(rmsgp->rm_body.rm_padded.rm_align); + be32_to_cpu(rmsgp->rm_body.rm_padded.rm_align); rmsgp->rm_body.rm_padded.rm_thresh = - ntohl(rmsgp->rm_body.rm_padded.rm_thresh); + be32_to_cpu(rmsgp->rm_body.rm_padded.rm_thresh); va = &rmsgp->rm_body.rm_padded.rm_pempty[4]; rqstp->rq_arg.head[0].iov_base = va; @@ -192,7 +186,7 @@ int svc_rdma_xdr_decode_req(struct rpcrdma_msg **rdma_req, * chunk list and a reply chunk list. */ va = &rmsgp->rm_body.rm_chunks[0]; - vaend = (u32 *)((unsigned long)rmsgp + rqstp->rq_arg.len); + vaend = (__be32 *)((unsigned long)rmsgp + rqstp->rq_arg.len); va = decode_read_list(va, vaend); if (!va) return -EINVAL; @@ -211,76 +205,20 @@ int svc_rdma_xdr_decode_req(struct rpcrdma_msg **rdma_req, return hdr_len; } -int svc_rdma_xdr_decode_deferred_req(struct svc_rqst *rqstp) -{ - struct rpcrdma_msg *rmsgp = NULL; - struct rpcrdma_read_chunk *ch; - struct rpcrdma_write_array *ary; - u32 *va; - u32 hdrlen; - - dprintk("svcrdma: processing deferred RDMA header on rqstp=%p\n", - rqstp); - rmsgp = (struct rpcrdma_msg *)rqstp->rq_arg.head[0].iov_base; - - /* Pull in the extra for the padded case and bump our pointer */ - if (rmsgp->rm_type == RDMA_MSGP) { - va = &rmsgp->rm_body.rm_padded.rm_pempty[4]; - rqstp->rq_arg.head[0].iov_base = va; - hdrlen = (u32)((unsigned long)va - (unsigned long)rmsgp); - rqstp->rq_arg.head[0].iov_len -= hdrlen; - return hdrlen; - } - - /* - * Skip all chunks to find RPC msg. These were previously processed - */ - va = &rmsgp->rm_body.rm_chunks[0]; - - /* Skip read-list */ - for (ch = (struct rpcrdma_read_chunk *)va; - ch->rc_discrim != xdr_zero; ch++); - va = (u32 *)&ch->rc_position; - - /* Skip write-list */ - ary = (struct rpcrdma_write_array *)va; - if (ary->wc_discrim == xdr_zero) - va = (u32 *)&ary->wc_nchunks; - else - /* - * rs_length is the 2nd 4B field in wc_target and taking its - * address skips the list terminator - */ - va = (u32 *)&ary->wc_array[ary->wc_nchunks].wc_target.rs_length; - - /* Skip reply-array */ - ary = (struct rpcrdma_write_array *)va; - if (ary->wc_discrim == xdr_zero) - va = (u32 *)&ary->wc_nchunks; - else - va = (u32 *)&ary->wc_array[ary->wc_nchunks]; - - rqstp->rq_arg.head[0].iov_base = va; - hdrlen = (unsigned long)va - (unsigned long)rmsgp; - rqstp->rq_arg.head[0].iov_len -= hdrlen; - - return hdrlen; -} - int svc_rdma_xdr_encode_error(struct svcxprt_rdma *xprt, struct rpcrdma_msg *rmsgp, - enum rpcrdma_errcode err, u32 *va) + enum rpcrdma_errcode err, __be32 *va) { - u32 *startp = va; + __be32 *startp = va; - *va++ = htonl(rmsgp->rm_xid); - *va++ = htonl(rmsgp->rm_vers); - *va++ = htonl(xprt->sc_max_requests); - *va++ = htonl(RDMA_ERROR); - *va++ = htonl(err); + *va++ = rmsgp->rm_xid; + *va++ = rmsgp->rm_vers; + *va++ = cpu_to_be32(xprt->sc_max_requests); + *va++ = rdma_error; + *va++ = cpu_to_be32(err); if (err == ERR_VERS) { - *va++ = htonl(RPCRDMA_VERSION); - *va++ = htonl(RPCRDMA_VERSION); + *va++ = rpcrdma_version; + *va++ = rpcrdma_version; } return (int)((unsigned long)va - (unsigned long)startp); @@ -297,7 +235,7 @@ int svc_rdma_xdr_get_reply_hdr_len(struct rpcrdma_msg *rmsgp) &rmsgp->rm_body.rm_chunks[1]; if (wr_ary->wc_discrim) wr_ary = (struct rpcrdma_write_array *) - &wr_ary->wc_array[ntohl(wr_ary->wc_nchunks)]. + &wr_ary->wc_array[be32_to_cpu(wr_ary->wc_nchunks)]. wc_target.rs_length; else wr_ary = (struct rpcrdma_write_array *) @@ -306,7 +244,7 @@ int svc_rdma_xdr_get_reply_hdr_len(struct rpcrdma_msg *rmsgp) /* skip reply array */ if (wr_ary->wc_discrim) wr_ary = (struct rpcrdma_write_array *) - &wr_ary->wc_array[ntohl(wr_ary->wc_nchunks)]; + &wr_ary->wc_array[be32_to_cpu(wr_ary->wc_nchunks)]; else wr_ary = (struct rpcrdma_write_array *) &wr_ary->wc_nchunks; @@ -325,7 +263,7 @@ void svc_rdma_xdr_encode_write_list(struct rpcrdma_msg *rmsgp, int chunks) ary = (struct rpcrdma_write_array *) &rmsgp->rm_body.rm_chunks[1]; ary->wc_discrim = xdr_one; - ary->wc_nchunks = htonl(chunks); + ary->wc_nchunks = cpu_to_be32(chunks); /* write-list terminator */ ary->wc_array[chunks].wc_target.rs_handle = xdr_zero; @@ -338,7 +276,7 @@ void svc_rdma_xdr_encode_reply_array(struct rpcrdma_write_array *ary, int chunks) { ary->wc_discrim = xdr_one; - ary->wc_nchunks = htonl(chunks); + ary->wc_nchunks = cpu_to_be32(chunks); } void svc_rdma_xdr_encode_array_chunk(struct rpcrdma_write_array *ary, @@ -350,7 +288,7 @@ void svc_rdma_xdr_encode_array_chunk(struct rpcrdma_write_array *ary, struct rpcrdma_segment *seg = &ary->wc_array[chunk_no].wc_target; seg->rs_handle = rs_handle; seg->rs_offset = rs_offset; - seg->rs_length = htonl(write_len); + seg->rs_length = cpu_to_be32(write_len); } void svc_rdma_xdr_encode_reply_header(struct svcxprt_rdma *xprt, @@ -358,10 +296,10 @@ void svc_rdma_xdr_encode_reply_header(struct svcxprt_rdma *xprt, struct rpcrdma_msg *rdma_resp, enum rpcrdma_proc rdma_type) { - rdma_resp->rm_xid = htonl(rdma_argp->rm_xid); - rdma_resp->rm_vers = htonl(rdma_argp->rm_vers); - rdma_resp->rm_credit = htonl(xprt->sc_max_requests); - rdma_resp->rm_type = htonl(rdma_type); + rdma_resp->rm_xid = rdma_argp->rm_xid; + rdma_resp->rm_vers = rdma_argp->rm_vers; + rdma_resp->rm_credit = cpu_to_be32(xprt->sc_max_requests); + rdma_resp->rm_type = cpu_to_be32(rdma_type); /* Encode <nul> chunks lists */ rdma_resp->rm_body.rm_chunks[0] = xdr_zero; diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c index f9f13a32d..2e1348bde 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c +++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c @@ -85,7 +85,7 @@ static void rdma_build_arg_xdr(struct svc_rqst *rqstp, /* RDMA_NOMSG: RDMA READ data should land just after RDMA RECV data */ rmsgp = (struct rpcrdma_msg *)rqstp->rq_arg.head[0].iov_base; - if (be32_to_cpu(rmsgp->rm_type) == RDMA_NOMSG) + if (rmsgp->rm_type == rdma_nomsg) rqstp->rq_arg.pages = &rqstp->rq_pages[0]; else rqstp->rq_arg.pages = &rqstp->rq_pages[1]; @@ -117,8 +117,8 @@ static void rdma_build_arg_xdr(struct svc_rqst *rqstp, static int rdma_read_max_sge(struct svcxprt_rdma *xprt, int sge_count) { - if (rdma_node_get_transport(xprt->sc_cm_id->device->node_type) == - RDMA_TRANSPORT_IWARP) + if (!rdma_cap_read_multi_sge(xprt->sc_cm_id->device, + xprt->sc_cm_id->port_num)) return 1; else return min_t(int, sge_count, xprt->sc_max_sge); diff --git a/net/sunrpc/xprtrdma/svc_rdma_sendto.c b/net/sunrpc/xprtrdma/svc_rdma_sendto.c index 7de33d1af..d25cd430f 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_sendto.c +++ b/net/sunrpc/xprtrdma/svc_rdma_sendto.c @@ -240,6 +240,7 @@ static int send_write_chunks(struct svcxprt_rdma *xprt, u32 xdr_off; int chunk_off; int chunk_no; + int nchunks; struct rpcrdma_write_array *arg_ary; struct rpcrdma_write_array *res_ary; int ret; @@ -251,14 +252,15 @@ static int send_write_chunks(struct svcxprt_rdma *xprt, &rdma_resp->rm_body.rm_chunks[1]; /* Write chunks start at the pagelist */ + nchunks = be32_to_cpu(arg_ary->wc_nchunks); for (xdr_off = rqstp->rq_res.head[0].iov_len, chunk_no = 0; - xfer_len && chunk_no < arg_ary->wc_nchunks; + xfer_len && chunk_no < nchunks; chunk_no++) { struct rpcrdma_segment *arg_ch; u64 rs_offset; arg_ch = &arg_ary->wc_array[chunk_no].wc_target; - write_len = min(xfer_len, ntohl(arg_ch->rs_length)); + write_len = min(xfer_len, be32_to_cpu(arg_ch->rs_length)); /* Prepare the response chunk given the length actually * written */ @@ -270,7 +272,7 @@ static int send_write_chunks(struct svcxprt_rdma *xprt, chunk_off = 0; while (write_len) { ret = send_write(xprt, rqstp, - ntohl(arg_ch->rs_handle), + be32_to_cpu(arg_ch->rs_handle), rs_offset + chunk_off, xdr_off, write_len, @@ -318,13 +320,13 @@ static int send_reply_chunks(struct svcxprt_rdma *xprt, &rdma_resp->rm_body.rm_chunks[2]; /* xdr offset starts at RPC message */ - nchunks = ntohl(arg_ary->wc_nchunks); + nchunks = be32_to_cpu(arg_ary->wc_nchunks); for (xdr_off = 0, chunk_no = 0; xfer_len && chunk_no < nchunks; chunk_no++) { u64 rs_offset; ch = &arg_ary->wc_array[chunk_no].wc_target; - write_len = min(xfer_len, htonl(ch->rs_length)); + write_len = min(xfer_len, be32_to_cpu(ch->rs_length)); /* Prepare the reply chunk given the length actually * written */ @@ -335,7 +337,7 @@ static int send_reply_chunks(struct svcxprt_rdma *xprt, chunk_off = 0; while (write_len) { ret = send_write(xprt, rqstp, - ntohl(ch->rs_handle), + be32_to_cpu(ch->rs_handle), rs_offset + chunk_off, xdr_off, write_len, @@ -515,7 +517,7 @@ int svc_rdma_sendto(struct svc_rqst *rqstp) inline_bytes = rqstp->rq_res.len; /* Create the RDMA response header */ - res_page = svc_rdma_get_page(); + res_page = alloc_page(GFP_KERNEL | __GFP_NOFAIL); rdma_resp = page_address(res_page); reply_ary = svc_rdma_get_reply_array(rdma_argp); if (reply_ary) diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c index f609c1c2d..6b36279e4 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_transport.c +++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c @@ -91,7 +91,7 @@ struct svc_xprt_class svc_rdma_class = { .xcl_name = "rdma", .xcl_owner = THIS_MODULE, .xcl_ops = &svc_rdma_ops, - .xcl_max_payload = RPCSVC_MAXPAYLOAD_RDMA, + .xcl_max_payload = RPCRDMA_MAXPAYLOAD, .xcl_ident = XPRT_TRANSPORT_RDMA, }; @@ -99,12 +99,8 @@ struct svc_rdma_op_ctxt *svc_rdma_get_context(struct svcxprt_rdma *xprt) { struct svc_rdma_op_ctxt *ctxt; - while (1) { - ctxt = kmem_cache_alloc(svc_rdma_ctxt_cachep, GFP_KERNEL); - if (ctxt) - break; - schedule_timeout_uninterruptible(msecs_to_jiffies(500)); - } + ctxt = kmem_cache_alloc(svc_rdma_ctxt_cachep, + GFP_KERNEL | __GFP_NOFAIL); ctxt->xprt = xprt; INIT_LIST_HEAD(&ctxt->dto_q); ctxt->count = 0; @@ -156,12 +152,8 @@ void svc_rdma_put_context(struct svc_rdma_op_ctxt *ctxt, int free_pages) struct svc_rdma_req_map *svc_rdma_get_req_map(void) { struct svc_rdma_req_map *map; - while (1) { - map = kmem_cache_alloc(svc_rdma_map_cachep, GFP_KERNEL); - if (map) - break; - schedule_timeout_uninterruptible(msecs_to_jiffies(500)); - } + map = kmem_cache_alloc(svc_rdma_map_cachep, + GFP_KERNEL | __GFP_NOFAIL); map->count = 0; return map; } @@ -175,8 +167,8 @@ void svc_rdma_put_req_map(struct svc_rdma_req_map *map) static void cq_event_handler(struct ib_event *event, void *context) { struct svc_xprt *xprt = context; - dprintk("svcrdma: received CQ event id=%d, context=%p\n", - event->event, context); + dprintk("svcrdma: received CQ event %s (%d), context=%p\n", + ib_event_msg(event->event), event->event, context); set_bit(XPT_CLOSE, &xprt->xpt_flags); } @@ -191,8 +183,9 @@ static void qp_event_handler(struct ib_event *event, void *context) case IB_EVENT_COMM_EST: case IB_EVENT_SQ_DRAINED: case IB_EVENT_QP_LAST_WQE_REACHED: - dprintk("svcrdma: QP event %d received for QP=%p\n", - event->event, event->element.qp); + dprintk("svcrdma: QP event %s (%d) received for QP=%p\n", + ib_event_msg(event->event), event->event, + event->element.qp); break; /* These are considered fatal events */ case IB_EVENT_PATH_MIG_ERR: @@ -201,9 +194,10 @@ static void qp_event_handler(struct ib_event *event, void *context) case IB_EVENT_QP_ACCESS_ERR: case IB_EVENT_DEVICE_FATAL: default: - dprintk("svcrdma: QP ERROR event %d received for QP=%p, " + dprintk("svcrdma: QP ERROR event %s (%d) received for QP=%p, " "closing transport\n", - event->event, event->element.qp); + ib_event_msg(event->event), event->event, + event->element.qp); set_bit(XPT_CLOSE, &xprt->xpt_flags); break; } @@ -402,7 +396,8 @@ static void sq_cq_reap(struct svcxprt_rdma *xprt) for (i = 0; i < ret; i++) { wc = &wc_a[i]; if (wc->status != IB_WC_SUCCESS) { - dprintk("svcrdma: sq wc err status %d\n", + dprintk("svcrdma: sq wc err status %s (%d)\n", + ib_wc_status_msg(wc->status), wc->status); /* Close the transport */ @@ -490,18 +485,6 @@ static struct svcxprt_rdma *rdma_create_xprt(struct svc_serv *serv, return cma_xprt; } -struct page *svc_rdma_get_page(void) -{ - struct page *page; - - while ((page = alloc_page(GFP_KERNEL)) == NULL) { - /* If we can't get memory, wait a bit and try again */ - printk(KERN_INFO "svcrdma: out of memory...retrying in 1s\n"); - schedule_timeout_uninterruptible(msecs_to_jiffies(1000)); - } - return page; -} - int svc_rdma_post_recv(struct svcxprt_rdma *xprt) { struct ib_recv_wr recv_wr, *bad_recv_wr; @@ -520,7 +503,7 @@ int svc_rdma_post_recv(struct svcxprt_rdma *xprt) pr_err("svcrdma: Too many sges (%d)\n", sge_no); goto err_put_ctxt; } - page = svc_rdma_get_page(); + page = alloc_page(GFP_KERNEL | __GFP_NOFAIL); ctxt->pages[sge_no] = page; pa = ib_dma_map_page(xprt->sc_cm_id->device, page, 0, PAGE_SIZE, @@ -616,7 +599,8 @@ static int rdma_listen_handler(struct rdma_cm_id *cma_id, switch (event->event) { case RDMA_CM_EVENT_CONNECT_REQUEST: dprintk("svcrdma: Connect request on cma_id=%p, xprt = %p, " - "event=%d\n", cma_id, cma_id->context, event->event); + "event = %s (%d)\n", cma_id, cma_id->context, + rdma_event_msg(event->event), event->event); handle_connect_req(cma_id, event->param.conn.initiator_depth); break; @@ -636,7 +620,8 @@ static int rdma_listen_handler(struct rdma_cm_id *cma_id, default: dprintk("svcrdma: Unexpected event on listening endpoint %p, " - "event=%d\n", cma_id, event->event); + "event = %s (%d)\n", cma_id, + rdma_event_msg(event->event), event->event); break; } @@ -669,7 +654,8 @@ static int rdma_cma_handler(struct rdma_cm_id *cma_id, break; case RDMA_CM_EVENT_DEVICE_REMOVAL: dprintk("svcrdma: Device removal cma_id=%p, xprt = %p, " - "event=%d\n", cma_id, xprt, event->event); + "event = %s (%d)\n", cma_id, xprt, + rdma_event_msg(event->event), event->event); if (xprt) { set_bit(XPT_CLOSE, &xprt->xpt_flags); svc_xprt_enqueue(xprt); @@ -677,7 +663,8 @@ static int rdma_cma_handler(struct rdma_cm_id *cma_id, break; default: dprintk("svcrdma: Unexpected event on DTO endpoint %p, " - "event=%d\n", cma_id, event->event); + "event = %s (%d)\n", cma_id, + rdma_event_msg(event->event), event->event); break; } return 0; @@ -848,10 +835,11 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) struct svcxprt_rdma *listen_rdma; struct svcxprt_rdma *newxprt = NULL; struct rdma_conn_param conn_param; + struct ib_cq_init_attr cq_attr = {}; struct ib_qp_init_attr qp_attr; struct ib_device_attr devattr; int uninitialized_var(dma_mr_acc); - int need_dma_mr; + int need_dma_mr = 0; int ret; int i; @@ -900,22 +888,22 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) dprintk("svcrdma: error creating PD for connect request\n"); goto errout; } + cq_attr.cqe = newxprt->sc_sq_depth; newxprt->sc_sq_cq = ib_create_cq(newxprt->sc_cm_id->device, sq_comp_handler, cq_event_handler, newxprt, - newxprt->sc_sq_depth, - 0); + &cq_attr); if (IS_ERR(newxprt->sc_sq_cq)) { dprintk("svcrdma: error creating SQ CQ for connect request\n"); goto errout; } + cq_attr.cqe = newxprt->sc_max_requests; newxprt->sc_rq_cq = ib_create_cq(newxprt->sc_cm_id->device, rq_comp_handler, cq_event_handler, newxprt, - newxprt->sc_max_requests, - 0); + &cq_attr); if (IS_ERR(newxprt->sc_rq_cq)) { dprintk("svcrdma: error creating RQ CQ for connect request\n"); goto errout; @@ -985,35 +973,26 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) /* * Determine if a DMA MR is required and if so, what privs are required */ - switch (rdma_node_get_transport(newxprt->sc_cm_id->device->node_type)) { - case RDMA_TRANSPORT_IWARP: - newxprt->sc_dev_caps |= SVCRDMA_DEVCAP_READ_W_INV; - if (!(newxprt->sc_dev_caps & SVCRDMA_DEVCAP_FAST_REG)) { - need_dma_mr = 1; - dma_mr_acc = - (IB_ACCESS_LOCAL_WRITE | - IB_ACCESS_REMOTE_WRITE); - } else if (!(devattr.device_cap_flags & IB_DEVICE_LOCAL_DMA_LKEY)) { - need_dma_mr = 1; - dma_mr_acc = IB_ACCESS_LOCAL_WRITE; - } else - need_dma_mr = 0; - break; - case RDMA_TRANSPORT_IB: - if (!(newxprt->sc_dev_caps & SVCRDMA_DEVCAP_FAST_REG)) { - need_dma_mr = 1; - dma_mr_acc = IB_ACCESS_LOCAL_WRITE; - } else if (!(devattr.device_cap_flags & - IB_DEVICE_LOCAL_DMA_LKEY)) { - need_dma_mr = 1; - dma_mr_acc = IB_ACCESS_LOCAL_WRITE; - } else - need_dma_mr = 0; - break; - default: + if (!rdma_protocol_iwarp(newxprt->sc_cm_id->device, + newxprt->sc_cm_id->port_num) && + !rdma_ib_or_roce(newxprt->sc_cm_id->device, + newxprt->sc_cm_id->port_num)) goto errout; + + if (!(newxprt->sc_dev_caps & SVCRDMA_DEVCAP_FAST_REG) || + !(devattr.device_cap_flags & IB_DEVICE_LOCAL_DMA_LKEY)) { + need_dma_mr = 1; + dma_mr_acc = IB_ACCESS_LOCAL_WRITE; + if (rdma_protocol_iwarp(newxprt->sc_cm_id->device, + newxprt->sc_cm_id->port_num) && + !(newxprt->sc_dev_caps & SVCRDMA_DEVCAP_FAST_REG)) + dma_mr_acc |= IB_ACCESS_REMOTE_WRITE; } + if (rdma_protocol_iwarp(newxprt->sc_cm_id->device, + newxprt->sc_cm_id->port_num)) + newxprt->sc_dev_caps |= SVCRDMA_DEVCAP_READ_W_INV; + /* Create the DMA MR if needed, otherwise, use the DMA LKEY */ if (need_dma_mr) { /* Register all of physical memory */ @@ -1319,11 +1298,11 @@ void svc_rdma_send_error(struct svcxprt_rdma *xprt, struct rpcrdma_msg *rmsgp, struct ib_send_wr err_wr; struct page *p; struct svc_rdma_op_ctxt *ctxt; - u32 *va; + __be32 *va; int length; int ret; - p = svc_rdma_get_page(); + p = alloc_page(GFP_KERNEL | __GFP_NOFAIL); va = page_address(p); /* XDR encode error */ diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c index 54f23b1be..680f888a9 100644 --- a/net/sunrpc/xprtrdma/transport.c +++ b/net/sunrpc/xprtrdma/transport.c @@ -48,7 +48,6 @@ */ #include <linux/module.h> -#include <linux/init.h> #include <linux/slab.h> #include <linux/seq_file.h> #include <linux/sunrpc/addr.h> @@ -59,11 +58,6 @@ # define RPCDBG_FACILITY RPCDBG_TRANS #endif -MODULE_LICENSE("Dual BSD/GPL"); - -MODULE_DESCRIPTION("RPC/RDMA Transport for Linux kernel NFS"); -MODULE_AUTHOR("Network Appliance, Inc."); - /* * tunables */ @@ -246,6 +240,16 @@ xprt_rdma_connect_worker(struct work_struct *work) xprt_clear_connecting(xprt); } +static void +xprt_rdma_inject_disconnect(struct rpc_xprt *xprt) +{ + struct rpcrdma_xprt *r_xprt = container_of(xprt, struct rpcrdma_xprt, + rx_xprt); + + pr_info("rpcrdma: injecting transport disconnect on xprt=%p\n", xprt); + rdma_disconnect(r_xprt->rx_ia.ri_id); +} + /* * xprt_rdma_destroy * @@ -618,12 +622,6 @@ xprt_rdma_send_request(struct rpc_task *task) if (req->rl_reply == NULL) /* e.g. reconnection */ rpcrdma_recv_buffer_get(req); - if (req->rl_reply) { - req->rl_reply->rr_func = rpcrdma_reply_handler; - /* this need only be done once, but... */ - req->rl_reply->rr_xprt = xprt; - } - /* Must suppress retransmit to maintain credits */ if (req->rl_connect_cookie == xprt->connect_cookie) goto drop_connection; @@ -682,6 +680,17 @@ static void xprt_rdma_print_stats(struct rpc_xprt *xprt, struct seq_file *seq) r_xprt->rx_stats.bad_reply_count); } +static int +xprt_rdma_enable_swap(struct rpc_xprt *xprt) +{ + return -EINVAL; +} + +static void +xprt_rdma_disable_swap(struct rpc_xprt *xprt) +{ +} + /* * Plumbing for rpc transport switch and kernel module */ @@ -700,7 +709,10 @@ static struct rpc_xprt_ops xprt_rdma_procs = { .send_request = xprt_rdma_send_request, .close = xprt_rdma_close, .destroy = xprt_rdma_destroy, - .print_stats = xprt_rdma_print_stats + .print_stats = xprt_rdma_print_stats, + .enable_swap = xprt_rdma_enable_swap, + .disable_swap = xprt_rdma_disable_swap, + .inject_disconnect = xprt_rdma_inject_disconnect }; static struct xprt_class xprt_rdma = { @@ -711,7 +723,7 @@ static struct xprt_class xprt_rdma = { .setup = xprt_setup_rdma, }; -static void __exit xprt_rdma_cleanup(void) +void xprt_rdma_cleanup(void) { int rc; @@ -726,17 +738,24 @@ static void __exit xprt_rdma_cleanup(void) if (rc) dprintk("RPC: %s: xprt_unregister returned %i\n", __func__, rc); + + frwr_destroy_recovery_wq(); } -static int __init xprt_rdma_init(void) +int xprt_rdma_init(void) { int rc; - rc = xprt_register_transport(&xprt_rdma); - + rc = frwr_alloc_recovery_wq(); if (rc) return rc; + rc = xprt_register_transport(&xprt_rdma); + if (rc) { + frwr_destroy_recovery_wq(); + return rc; + } + dprintk("RPCRDMA Module Init, register RPC RDMA transport\n"); dprintk("Defaults:\n"); @@ -753,6 +772,3 @@ static int __init xprt_rdma_init(void) #endif return 0; } - -module_init(xprt_rdma_init); -module_exit(xprt_rdma_cleanup); diff --git a/net/sunrpc/xprtrdma/verbs.c b/net/sunrpc/xprtrdma/verbs.c index 4870d272e..891c4ede2 100644 --- a/net/sunrpc/xprtrdma/verbs.c +++ b/net/sunrpc/xprtrdma/verbs.c @@ -80,7 +80,6 @@ static void rpcrdma_run_tasklet(unsigned long data) { struct rpcrdma_rep *rep; - void (*func)(struct rpcrdma_rep *); unsigned long flags; data = data; @@ -89,14 +88,9 @@ rpcrdma_run_tasklet(unsigned long data) rep = list_entry(rpcrdma_tasklets_g.next, struct rpcrdma_rep, rr_list); list_del(&rep->rr_list); - func = rep->rr_func; - rep->rr_func = NULL; spin_unlock_irqrestore(&rpcrdma_tk_lock_g, flags); - if (func) - func(rep); - else - rpcrdma_recv_buffer_put(rep); + rpcrdma_reply_handler(rep); spin_lock_irqsave(&rpcrdma_tk_lock_g, flags); } @@ -105,32 +99,6 @@ rpcrdma_run_tasklet(unsigned long data) static DECLARE_TASKLET(rpcrdma_tasklet_g, rpcrdma_run_tasklet, 0UL); -static const char * const async_event[] = { - "CQ error", - "QP fatal error", - "QP request error", - "QP access error", - "communication established", - "send queue drained", - "path migration successful", - "path mig error", - "device fatal error", - "port active", - "port error", - "LID change", - "P_key change", - "SM change", - "SRQ error", - "SRQ limit reached", - "last WQE reached", - "client reregister", - "GID change", -}; - -#define ASYNC_MSG(status) \ - ((status) < ARRAY_SIZE(async_event) ? \ - async_event[(status)] : "unknown async error") - static void rpcrdma_schedule_tasklet(struct list_head *sched_list) { @@ -148,7 +116,7 @@ rpcrdma_qp_async_error_upcall(struct ib_event *event, void *context) struct rpcrdma_ep *ep = context; pr_err("RPC: %s: %s on device %s ep %p\n", - __func__, ASYNC_MSG(event->event), + __func__, ib_event_msg(event->event), event->device->name, context); if (ep->rep_connected == 1) { ep->rep_connected = -EIO; @@ -163,7 +131,7 @@ rpcrdma_cq_async_error_upcall(struct ib_event *event, void *context) struct rpcrdma_ep *ep = context; pr_err("RPC: %s: %s on device %s ep %p\n", - __func__, ASYNC_MSG(event->event), + __func__, ib_event_msg(event->event), event->device->name, context); if (ep->rep_connected == 1) { ep->rep_connected = -EIO; @@ -172,35 +140,6 @@ rpcrdma_cq_async_error_upcall(struct ib_event *event, void *context) } } -static const char * const wc_status[] = { - "success", - "local length error", - "local QP operation error", - "local EE context operation error", - "local protection error", - "WR flushed", - "memory management operation error", - "bad response error", - "local access error", - "remote invalid request error", - "remote access error", - "remote operation error", - "transport retry counter exceeded", - "RNR retry counter exceeded", - "local RDD violation error", - "remove invalid RD request", - "operation aborted", - "invalid EE context number", - "invalid EE context state", - "fatal error", - "response timeout error", - "general error", -}; - -#define COMPLETION_MSG(status) \ - ((status) < ARRAY_SIZE(wc_status) ? \ - wc_status[(status)] : "unexpected completion error") - static void rpcrdma_sendcq_process_wc(struct ib_wc *wc) { @@ -209,7 +148,7 @@ rpcrdma_sendcq_process_wc(struct ib_wc *wc) if (wc->status != IB_WC_SUCCESS && wc->status != IB_WC_WR_FLUSH_ERR) pr_err("RPC: %s: SEND: %s\n", - __func__, COMPLETION_MSG(wc->status)); + __func__, ib_wc_status_msg(wc->status)); } else { struct rpcrdma_mw *r; @@ -291,7 +230,7 @@ rpcrdma_recvcq_process_wc(struct ib_wc *wc, struct list_head *sched_list) __func__, rep, wc->byte_len); rep->rr_len = wc->byte_len; - ib_dma_sync_single_for_cpu(rdmab_to_ia(rep->rr_buffer)->ri_id->device, + ib_dma_sync_single_for_cpu(rep->rr_device, rdmab_addr(rep->rr_rdmabuf), rep->rr_len, DMA_FROM_DEVICE); prefetch(rdmab_to_msg(rep->rr_rdmabuf)); @@ -302,7 +241,7 @@ out_schedule: out_fail: if (wc->status != IB_WC_WR_FLUSH_ERR) pr_err("RPC: %s: rep %p: %s\n", - __func__, rep, COMPLETION_MSG(wc->status)); + __func__, rep, ib_wc_status_msg(wc->status)); rep->rr_len = ~0U; goto out_schedule; } @@ -386,31 +325,6 @@ rpcrdma_flush_cqs(struct rpcrdma_ep *ep) rpcrdma_sendcq_process_wc(&wc); } -#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) -static const char * const conn[] = { - "address resolved", - "address error", - "route resolved", - "route error", - "connect request", - "connect response", - "connect error", - "unreachable", - "rejected", - "established", - "disconnected", - "device removal", - "multicast join", - "multicast error", - "address change", - "timewait exit", -}; - -#define CONNECTION_MSG(status) \ - ((status) < ARRAY_SIZE(conn) ? \ - conn[(status)] : "unrecognized connection error") -#endif - static int rpcrdma_conn_upcall(struct rdma_cm_id *id, struct rdma_cm_event *event) { @@ -476,7 +390,7 @@ connected: default: dprintk("RPC: %s: %pIS:%u (ep 0x%p): %s\n", __func__, sap, rpc_get_port(sap), ep, - CONNECTION_MSG(event->event)); + rdma_event_msg(event->event)); break; } @@ -487,7 +401,7 @@ connected: pr_info("rpcrdma: connection to %pIS:%u on %s, memreg '%s', %d credits, %d responders%s\n", sap, rpc_get_port(sap), - ia->ri_id->device->name, + ia->ri_device->name, ia->ri_ops->ro_displayname, xprt->rx_buf.rb_max_requests, ird, ird < 4 && ird < tird / 2 ? " (low!)" : ""); @@ -588,8 +502,9 @@ rpcrdma_ia_open(struct rpcrdma_xprt *xprt, struct sockaddr *addr, int memreg) rc = PTR_ERR(ia->ri_id); goto out1; } + ia->ri_device = ia->ri_id->device; - ia->ri_pd = ib_alloc_pd(ia->ri_id->device); + ia->ri_pd = ib_alloc_pd(ia->ri_device); if (IS_ERR(ia->ri_pd)) { rc = PTR_ERR(ia->ri_pd); dprintk("RPC: %s: ib_alloc_pd() failed %i\n", @@ -597,7 +512,7 @@ rpcrdma_ia_open(struct rpcrdma_xprt *xprt, struct sockaddr *addr, int memreg) goto out2; } - rc = ib_query_device(ia->ri_id->device, devattr); + rc = ib_query_device(ia->ri_device, devattr); if (rc) { dprintk("RPC: %s: ib_query_device failed %d\n", __func__, rc); @@ -606,7 +521,7 @@ rpcrdma_ia_open(struct rpcrdma_xprt *xprt, struct sockaddr *addr, int memreg) if (devattr->device_cap_flags & IB_DEVICE_LOCAL_DMA_LKEY) { ia->ri_have_dma_lkey = 1; - ia->ri_dma_lkey = ia->ri_id->device->local_dma_lkey; + ia->ri_dma_lkey = ia->ri_device->local_dma_lkey; } if (memreg == RPCRDMA_FRMR) { @@ -621,7 +536,7 @@ rpcrdma_ia_open(struct rpcrdma_xprt *xprt, struct sockaddr *addr, int memreg) } } if (memreg == RPCRDMA_MTHCAFMR) { - if (!ia->ri_id->device->alloc_fmr) { + if (!ia->ri_device->alloc_fmr) { dprintk("RPC: %s: MTHCAFMR registration " "not supported by HCA\n", __func__); memreg = RPCRDMA_ALLPHYSICAL; @@ -670,9 +585,6 @@ rpcrdma_ia_open(struct rpcrdma_xprt *xprt, struct sockaddr *addr, int memreg) dprintk("RPC: %s: memory registration strategy is '%s'\n", __func__, ia->ri_ops->ro_displayname); - /* Else will do memory reg/dereg for each chunk */ - ia->ri_memreg_strategy = memreg; - rwlock_init(&ia->ri_qplock); return 0; @@ -702,17 +614,17 @@ rpcrdma_ia_close(struct rpcrdma_ia *ia) dprintk("RPC: %s: ib_dereg_mr returned %i\n", __func__, rc); } + if (ia->ri_id != NULL && !IS_ERR(ia->ri_id)) { if (ia->ri_id->qp) rdma_destroy_qp(ia->ri_id); rdma_destroy_id(ia->ri_id); ia->ri_id = NULL; } - if (ia->ri_pd != NULL && !IS_ERR(ia->ri_pd)) { - rc = ib_dealloc_pd(ia->ri_pd); - dprintk("RPC: %s: ib_dealloc_pd returned %i\n", - __func__, rc); - } + + /* If the pd is still busy, xprtrdma missed freeing a resource */ + if (ia->ri_pd && !IS_ERR(ia->ri_pd)) + WARN_ON(ib_dealloc_pd(ia->ri_pd)); } /* @@ -724,6 +636,7 @@ rpcrdma_ep_create(struct rpcrdma_ep *ep, struct rpcrdma_ia *ia, { struct ib_device_attr *devattr = &ia->ri_devattr; struct ib_cq *sendcq, *recvcq; + struct ib_cq_init_attr cq_attr = {}; int rc, err; /* check provider's send/recv wr limits */ @@ -771,9 +684,9 @@ rpcrdma_ep_create(struct rpcrdma_ep *ep, struct rpcrdma_ia *ia, init_waitqueue_head(&ep->rep_connect_wait); INIT_DELAYED_WORK(&ep->rep_connect_worker, rpcrdma_connect_worker); - sendcq = ib_create_cq(ia->ri_id->device, rpcrdma_sendcq_upcall, - rpcrdma_cq_async_error_upcall, ep, - ep->rep_attr.cap.max_send_wr + 1, 0); + cq_attr.cqe = ep->rep_attr.cap.max_send_wr + 1; + sendcq = ib_create_cq(ia->ri_device, rpcrdma_sendcq_upcall, + rpcrdma_cq_async_error_upcall, ep, &cq_attr); if (IS_ERR(sendcq)) { rc = PTR_ERR(sendcq); dprintk("RPC: %s: failed to create send CQ: %i\n", @@ -788,9 +701,9 @@ rpcrdma_ep_create(struct rpcrdma_ep *ep, struct rpcrdma_ia *ia, goto out2; } - recvcq = ib_create_cq(ia->ri_id->device, rpcrdma_recvcq_upcall, - rpcrdma_cq_async_error_upcall, ep, - ep->rep_attr.cap.max_recv_wr + 1, 0); + cq_attr.cqe = ep->rep_attr.cap.max_recv_wr + 1; + recvcq = ib_create_cq(ia->ri_device, rpcrdma_recvcq_upcall, + rpcrdma_cq_async_error_upcall, ep, &cq_attr); if (IS_ERR(recvcq)) { rc = PTR_ERR(recvcq); dprintk("RPC: %s: failed to create recv CQ: %i\n", @@ -896,8 +809,6 @@ retry: rpcrdma_flush_cqs(ep); xprt = container_of(ia, struct rpcrdma_xprt, rx_ia); - ia->ri_ops->ro_reset(xprt); - id = rpcrdma_create_id(xprt, ia, (struct sockaddr *)&xprt->rx_data.addr); if (IS_ERR(id)) { @@ -911,7 +822,7 @@ retry: * More stuff I haven't thought of! * Rrrgh! */ - if (ia->ri_id->device != id->device) { + if (ia->ri_device != id->device) { printk("RPC: %s: can't reconnect on " "different device!\n", __func__); rdma_destroy_id(id); @@ -1053,7 +964,8 @@ rpcrdma_create_rep(struct rpcrdma_xprt *r_xprt) goto out_free; } - rep->rr_buffer = &r_xprt->rx_buf; + rep->rr_device = ia->ri_device; + rep->rr_rxprt = r_xprt; return rep; out_free: @@ -1177,31 +1089,33 @@ rpcrdma_buffer_destroy(struct rpcrdma_buffer *buf) kfree(buf->rb_pool); } -/* "*mw" can be NULL when rpcrdma_buffer_get_mrs() fails, leaving - * some req segments uninitialized. - */ -static void -rpcrdma_buffer_put_mr(struct rpcrdma_mw **mw, struct rpcrdma_buffer *buf) +struct rpcrdma_mw * +rpcrdma_get_mw(struct rpcrdma_xprt *r_xprt) { - if (*mw) { - list_add_tail(&(*mw)->mw_list, &buf->rb_mws); - *mw = NULL; + struct rpcrdma_buffer *buf = &r_xprt->rx_buf; + struct rpcrdma_mw *mw = NULL; + + spin_lock(&buf->rb_mwlock); + if (!list_empty(&buf->rb_mws)) { + mw = list_first_entry(&buf->rb_mws, + struct rpcrdma_mw, mw_list); + list_del_init(&mw->mw_list); } + spin_unlock(&buf->rb_mwlock); + + if (!mw) + pr_err("RPC: %s: no MWs available\n", __func__); + return mw; } -/* Cycle mw's back in reverse order, and "spin" them. - * This delays and scrambles reuse as much as possible. - */ -static void -rpcrdma_buffer_put_mrs(struct rpcrdma_req *req, struct rpcrdma_buffer *buf) +void +rpcrdma_put_mw(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mw *mw) { - struct rpcrdma_mr_seg *seg = req->rl_segments; - struct rpcrdma_mr_seg *seg1 = seg; - int i; + struct rpcrdma_buffer *buf = &r_xprt->rx_buf; - for (i = 1, seg++; i < RPCRDMA_MAX_SEGS; seg++, i++) - rpcrdma_buffer_put_mr(&seg->rl_mw, buf); - rpcrdma_buffer_put_mr(&seg1->rl_mw, buf); + spin_lock(&buf->rb_mwlock); + list_add_tail(&mw->mw_list, &buf->rb_mws); + spin_unlock(&buf->rb_mwlock); } static void @@ -1211,115 +1125,10 @@ rpcrdma_buffer_put_sendbuf(struct rpcrdma_req *req, struct rpcrdma_buffer *buf) req->rl_niovs = 0; if (req->rl_reply) { buf->rb_recv_bufs[--buf->rb_recv_index] = req->rl_reply; - req->rl_reply->rr_func = NULL; req->rl_reply = NULL; } } -/* rpcrdma_unmap_one() was already done during deregistration. - * Redo only the ib_post_send(). - */ -static void -rpcrdma_retry_local_inv(struct rpcrdma_mw *r, struct rpcrdma_ia *ia) -{ - struct rpcrdma_xprt *r_xprt = - container_of(ia, struct rpcrdma_xprt, rx_ia); - struct ib_send_wr invalidate_wr, *bad_wr; - int rc; - - dprintk("RPC: %s: FRMR %p is stale\n", __func__, r); - - /* When this FRMR is re-inserted into rb_mws, it is no longer stale */ - r->r.frmr.fr_state = FRMR_IS_INVALID; - - memset(&invalidate_wr, 0, sizeof(invalidate_wr)); - invalidate_wr.wr_id = (unsigned long)(void *)r; - invalidate_wr.opcode = IB_WR_LOCAL_INV; - invalidate_wr.ex.invalidate_rkey = r->r.frmr.fr_mr->rkey; - DECR_CQCOUNT(&r_xprt->rx_ep); - - dprintk("RPC: %s: frmr %p invalidating rkey %08x\n", - __func__, r, r->r.frmr.fr_mr->rkey); - - read_lock(&ia->ri_qplock); - rc = ib_post_send(ia->ri_id->qp, &invalidate_wr, &bad_wr); - read_unlock(&ia->ri_qplock); - if (rc) { - /* Force rpcrdma_buffer_get() to retry */ - r->r.frmr.fr_state = FRMR_IS_STALE; - dprintk("RPC: %s: ib_post_send failed, %i\n", - __func__, rc); - } -} - -static void -rpcrdma_retry_flushed_linv(struct list_head *stale, - struct rpcrdma_buffer *buf) -{ - struct rpcrdma_ia *ia = rdmab_to_ia(buf); - struct list_head *pos; - struct rpcrdma_mw *r; - unsigned long flags; - - list_for_each(pos, stale) { - r = list_entry(pos, struct rpcrdma_mw, mw_list); - rpcrdma_retry_local_inv(r, ia); - } - - spin_lock_irqsave(&buf->rb_lock, flags); - list_splice_tail(stale, &buf->rb_mws); - spin_unlock_irqrestore(&buf->rb_lock, flags); -} - -static struct rpcrdma_req * -rpcrdma_buffer_get_frmrs(struct rpcrdma_req *req, struct rpcrdma_buffer *buf, - struct list_head *stale) -{ - struct rpcrdma_mw *r; - int i; - - i = RPCRDMA_MAX_SEGS - 1; - while (!list_empty(&buf->rb_mws)) { - r = list_entry(buf->rb_mws.next, - struct rpcrdma_mw, mw_list); - list_del(&r->mw_list); - if (r->r.frmr.fr_state == FRMR_IS_STALE) { - list_add(&r->mw_list, stale); - continue; - } - req->rl_segments[i].rl_mw = r; - if (unlikely(i-- == 0)) - return req; /* Success */ - } - - /* Not enough entries on rb_mws for this req */ - rpcrdma_buffer_put_sendbuf(req, buf); - rpcrdma_buffer_put_mrs(req, buf); - return NULL; -} - -static struct rpcrdma_req * -rpcrdma_buffer_get_fmrs(struct rpcrdma_req *req, struct rpcrdma_buffer *buf) -{ - struct rpcrdma_mw *r; - int i; - - i = RPCRDMA_MAX_SEGS - 1; - while (!list_empty(&buf->rb_mws)) { - r = list_entry(buf->rb_mws.next, - struct rpcrdma_mw, mw_list); - list_del(&r->mw_list); - req->rl_segments[i].rl_mw = r; - if (unlikely(i-- == 0)) - return req; /* Success */ - } - - /* Not enough entries on rb_mws for this req */ - rpcrdma_buffer_put_sendbuf(req, buf); - rpcrdma_buffer_put_mrs(req, buf); - return NULL; -} - /* * Get a set of request/reply buffers. * @@ -1332,12 +1141,11 @@ rpcrdma_buffer_get_fmrs(struct rpcrdma_req *req, struct rpcrdma_buffer *buf) struct rpcrdma_req * rpcrdma_buffer_get(struct rpcrdma_buffer *buffers) { - struct rpcrdma_ia *ia = rdmab_to_ia(buffers); - struct list_head stale; struct rpcrdma_req *req; unsigned long flags; spin_lock_irqsave(&buffers->rb_lock, flags); + if (buffers->rb_send_index == buffers->rb_max_requests) { spin_unlock_irqrestore(&buffers->rb_lock, flags); dprintk("RPC: %s: out of request buffers\n", __func__); @@ -1356,20 +1164,7 @@ rpcrdma_buffer_get(struct rpcrdma_buffer *buffers) } buffers->rb_send_bufs[buffers->rb_send_index++] = NULL; - INIT_LIST_HEAD(&stale); - switch (ia->ri_memreg_strategy) { - case RPCRDMA_FRMR: - req = rpcrdma_buffer_get_frmrs(req, buffers, &stale); - break; - case RPCRDMA_MTHCAFMR: - req = rpcrdma_buffer_get_fmrs(req, buffers); - break; - default: - break; - } spin_unlock_irqrestore(&buffers->rb_lock, flags); - if (!list_empty(&stale)) - rpcrdma_retry_flushed_linv(&stale, buffers); return req; } @@ -1381,19 +1176,10 @@ void rpcrdma_buffer_put(struct rpcrdma_req *req) { struct rpcrdma_buffer *buffers = req->rl_buffer; - struct rpcrdma_ia *ia = rdmab_to_ia(buffers); unsigned long flags; spin_lock_irqsave(&buffers->rb_lock, flags); rpcrdma_buffer_put_sendbuf(req, buffers); - switch (ia->ri_memreg_strategy) { - case RPCRDMA_FRMR: - case RPCRDMA_MTHCAFMR: - rpcrdma_buffer_put_mrs(req, buffers); - break; - default: - break; - } spin_unlock_irqrestore(&buffers->rb_lock, flags); } @@ -1423,10 +1209,9 @@ rpcrdma_recv_buffer_get(struct rpcrdma_req *req) void rpcrdma_recv_buffer_put(struct rpcrdma_rep *rep) { - struct rpcrdma_buffer *buffers = rep->rr_buffer; + struct rpcrdma_buffer *buffers = &rep->rr_rxprt->rx_buf; unsigned long flags; - rep->rr_func = NULL; spin_lock_irqsave(&buffers->rb_lock, flags); buffers->rb_recv_bufs[--buffers->rb_recv_index] = rep; spin_unlock_irqrestore(&buffers->rb_lock, flags); @@ -1455,9 +1240,9 @@ rpcrdma_register_internal(struct rpcrdma_ia *ia, void *va, int len, /* * All memory passed here was kmalloc'ed, therefore phys-contiguous. */ - iov->addr = ib_dma_map_single(ia->ri_id->device, + iov->addr = ib_dma_map_single(ia->ri_device, va, len, DMA_BIDIRECTIONAL); - if (ib_dma_mapping_error(ia->ri_id->device, iov->addr)) + if (ib_dma_mapping_error(ia->ri_device, iov->addr)) return -ENOMEM; iov->length = len; @@ -1501,8 +1286,8 @@ rpcrdma_deregister_internal(struct rpcrdma_ia *ia, { int rc; - ib_dma_unmap_single(ia->ri_id->device, - iov->addr, iov->length, DMA_BIDIRECTIONAL); + ib_dma_unmap_single(ia->ri_device, + iov->addr, iov->length, DMA_BIDIRECTIONAL); if (NULL == mr) return 0; @@ -1595,15 +1380,18 @@ rpcrdma_ep_post(struct rpcrdma_ia *ia, send_wr.num_sge = req->rl_niovs; send_wr.opcode = IB_WR_SEND; if (send_wr.num_sge == 4) /* no need to sync any pad (constant) */ - ib_dma_sync_single_for_device(ia->ri_id->device, - req->rl_send_iov[3].addr, req->rl_send_iov[3].length, - DMA_TO_DEVICE); - ib_dma_sync_single_for_device(ia->ri_id->device, - req->rl_send_iov[1].addr, req->rl_send_iov[1].length, - DMA_TO_DEVICE); - ib_dma_sync_single_for_device(ia->ri_id->device, - req->rl_send_iov[0].addr, req->rl_send_iov[0].length, - DMA_TO_DEVICE); + ib_dma_sync_single_for_device(ia->ri_device, + req->rl_send_iov[3].addr, + req->rl_send_iov[3].length, + DMA_TO_DEVICE); + ib_dma_sync_single_for_device(ia->ri_device, + req->rl_send_iov[1].addr, + req->rl_send_iov[1].length, + DMA_TO_DEVICE); + ib_dma_sync_single_for_device(ia->ri_device, + req->rl_send_iov[0].addr, + req->rl_send_iov[0].length, + DMA_TO_DEVICE); if (DECR_CQCOUNT(ep) > 0) send_wr.send_flags = 0; @@ -1636,7 +1424,7 @@ rpcrdma_ep_post_recv(struct rpcrdma_ia *ia, recv_wr.sg_list = &rep->rr_rdmabuf->rg_iov; recv_wr.num_sge = 1; - ib_dma_sync_single_for_cpu(ia->ri_id->device, + ib_dma_sync_single_for_cpu(ia->ri_device, rdmab_addr(rep->rr_rdmabuf), rdmab_length(rep->rr_rdmabuf), DMA_BIDIRECTIONAL); diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h index 78e0b8bea..f49dd8b38 100644 --- a/net/sunrpc/xprtrdma/xprt_rdma.h +++ b/net/sunrpc/xprtrdma/xprt_rdma.h @@ -62,6 +62,7 @@ struct rpcrdma_ia { const struct rpcrdma_memreg_ops *ri_ops; rwlock_t ri_qplock; + struct ib_device *ri_device; struct rdma_cm_id *ri_id; struct ib_pd *ri_pd; struct ib_mr *ri_bind_mem; @@ -69,7 +70,6 @@ struct rpcrdma_ia { int ri_have_dma_lkey; struct completion ri_done; int ri_async_rc; - enum rpcrdma_memreg ri_memreg_strategy; unsigned int ri_max_frmr_depth; struct ib_device_attr ri_devattr; struct ib_qp_attr ri_qp_attr; @@ -173,9 +173,8 @@ struct rpcrdma_buffer; struct rpcrdma_rep { unsigned int rr_len; - struct rpcrdma_buffer *rr_buffer; - struct rpc_xprt *rr_xprt; - void (*rr_func)(struct rpcrdma_rep *); + struct ib_device *rr_device; + struct rpcrdma_xprt *rr_rxprt; struct list_head rr_list; struct rpcrdma_regbuf *rr_rdmabuf; }; @@ -203,11 +202,18 @@ struct rpcrdma_frmr { struct ib_fast_reg_page_list *fr_pgl; struct ib_mr *fr_mr; enum rpcrdma_frmr_state fr_state; + struct work_struct fr_work; + struct rpcrdma_xprt *fr_xprt; +}; + +struct rpcrdma_fmr { + struct ib_fmr *fmr; + u64 *physaddrs; }; struct rpcrdma_mw { union { - struct ib_fmr *fmr; + struct rpcrdma_fmr fmr; struct rpcrdma_frmr frmr; } r; void (*mw_sendcompletion)(struct ib_wc *); @@ -281,15 +287,17 @@ rpcr_to_rdmar(struct rpc_rqst *rqst) * One of these is associated with a transport instance */ struct rpcrdma_buffer { - spinlock_t rb_lock; /* protects indexes */ - u32 rb_max_requests;/* client max requests */ - struct list_head rb_mws; /* optional memory windows/fmrs/frmrs */ - struct list_head rb_all; - int rb_send_index; + spinlock_t rb_mwlock; /* protect rb_mws list */ + struct list_head rb_mws; + struct list_head rb_all; + char *rb_pool; + + spinlock_t rb_lock; /* protect buf arrays */ + u32 rb_max_requests; + int rb_send_index; + int rb_recv_index; struct rpcrdma_req **rb_send_bufs; - int rb_recv_index; struct rpcrdma_rep **rb_recv_bufs; - char *rb_pool; }; #define rdmab_to_ia(b) (&container_of((b), struct rpcrdma_xprt, rx_buf)->rx_ia) @@ -350,7 +358,6 @@ struct rpcrdma_memreg_ops { struct rpcrdma_create_data_internal *); size_t (*ro_maxpages)(struct rpcrdma_xprt *); int (*ro_init)(struct rpcrdma_xprt *); - void (*ro_reset)(struct rpcrdma_xprt *); void (*ro_destroy)(struct rpcrdma_buffer *); const char *ro_displayname; }; @@ -413,6 +420,8 @@ int rpcrdma_ep_post_recv(struct rpcrdma_ia *, struct rpcrdma_ep *, int rpcrdma_buffer_create(struct rpcrdma_xprt *); void rpcrdma_buffer_destroy(struct rpcrdma_buffer *); +struct rpcrdma_mw *rpcrdma_get_mw(struct rpcrdma_xprt *); +void rpcrdma_put_mw(struct rpcrdma_xprt *, struct rpcrdma_mw *); struct rpcrdma_req *rpcrdma_buffer_get(struct rpcrdma_buffer *); void rpcrdma_buffer_put(struct rpcrdma_req *); void rpcrdma_recv_buffer_get(struct rpcrdma_req *); @@ -425,6 +434,9 @@ void rpcrdma_free_regbuf(struct rpcrdma_ia *, unsigned int rpcrdma_max_segments(struct rpcrdma_xprt *); +int frwr_alloc_recovery_wq(void); +void frwr_destroy_recovery_wq(void); + /* * Wrappers for chunk registration, shared by read/write chunk code. */ @@ -480,6 +492,11 @@ void rpcrdma_reply_handler(struct rpcrdma_rep *); */ int rpcrdma_marshal_req(struct rpc_rqst *); +/* RPC/RDMA module init - xprtrdma/transport.c + */ +int xprt_rdma_init(void); +void xprt_rdma_cleanup(void); + /* Temporary NFS request map cache. Created in svc_rdma.c */ extern struct kmem_cache *svc_rdma_map_cachep; /* WR context cache. Created in svc_rdma.c */ @@ -487,10 +504,4 @@ extern struct kmem_cache *svc_rdma_ctxt_cachep; /* Workqueue created in svc_rdma.c */ extern struct workqueue_struct *svc_rdma_wq; -#if RPCSVC_MAXPAYLOAD < (RPCRDMA_MAX_DATA_SEGS << PAGE_SHIFT) -#define RPCSVC_MAXPAYLOAD_RDMA RPCSVC_MAXPAYLOAD -#else -#define RPCSVC_MAXPAYLOAD_RDMA (RPCRDMA_MAX_DATA_SEGS << PAGE_SHIFT) -#endif - #endif /* _LINUX_SUNRPC_XPRT_RDMA_H */ diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 66891e32c..003037632 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -527,6 +527,10 @@ static int xs_local_send_request(struct rpc_task *task) true, &sent); dprintk("RPC: %s(%u) = %d\n", __func__, xdr->len - req->rq_bytes_sent, status); + + if (status == -EAGAIN && sock_writeable(transport->inet)) + status = -ENOBUFS; + if (likely(sent > 0) || status == 0) { req->rq_bytes_sent += sent; req->rq_xmit_bytes_sent += sent; @@ -539,6 +543,7 @@ static int xs_local_send_request(struct rpc_task *task) switch (status) { case -ENOBUFS: + break; case -EAGAIN: status = xs_nospace(task); break; @@ -589,6 +594,9 @@ static int xs_udp_send_request(struct rpc_task *task) if (status == -EPERM) goto process_status; + if (status == -EAGAIN && sock_writeable(transport->inet)) + status = -ENOBUFS; + if (sent > 0 || status == 0) { req->rq_xmit_bytes_sent += sent; if (sent >= req->rq_slen) @@ -623,24 +631,6 @@ process_status: } /** - * xs_tcp_shutdown - gracefully shut down a TCP socket - * @xprt: transport - * - * Initiates a graceful shutdown of the TCP socket by calling the - * equivalent of shutdown(SHUT_RDWR); - */ -static void xs_tcp_shutdown(struct rpc_xprt *xprt) -{ - struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); - struct socket *sock = transport->sock; - - if (sock != NULL) { - kernel_sock_shutdown(sock, SHUT_RDWR); - trace_rpc_socket_shutdown(xprt, sock); - } -} - -/** * xs_tcp_send_request - write an RPC request to a TCP socket * @task: address of RPC task that manages the state of an RPC request * @@ -687,9 +677,6 @@ static int xs_tcp_send_request(struct rpc_task *task) dprintk("RPC: xs_tcp_send_request(%u) = %d\n", xdr->len - req->rq_bytes_sent, status); - if (unlikely(sent == 0 && status < 0)) - break; - /* If we've sent the entire packet, immediately * reset the count of bytes sent. */ req->rq_bytes_sent += sent; @@ -699,18 +686,21 @@ static int xs_tcp_send_request(struct rpc_task *task) return 0; } - if (sent != 0) - continue; - status = -EAGAIN; - break; + if (status < 0) + break; + if (sent == 0) { + status = -EAGAIN; + break; + } } + if (status == -EAGAIN && sk_stream_is_writeable(transport->inet)) + status = -ENOBUFS; switch (status) { case -ENOTSOCK: status = -ENOTCONN; /* Should we call xs_close() here? */ break; - case -ENOBUFS: case -EAGAIN: status = xs_nospace(task); break; @@ -721,6 +711,7 @@ static int xs_tcp_send_request(struct rpc_task *task) case -ECONNREFUSED: case -ENOTCONN: case -EADDRINUSE: + case -ENOBUFS: case -EPIPE: clear_bit(SOCK_ASYNC_NOSPACE, &transport->sock->flags); } @@ -786,6 +777,7 @@ static void xs_sock_mark_closed(struct rpc_xprt *xprt) xs_sock_reset_connection_flags(xprt); /* Mark transport as closed and wake up all pending tasks */ xprt_disconnect_done(xprt); + xprt_force_disconnect(xprt); } /** @@ -827,6 +819,9 @@ static void xs_reset_transport(struct sock_xprt *transport) if (sk == NULL) return; + if (atomic_read(&transport->xprt.swapper)) + sk_clear_memalloc(sk); + write_lock_bh(&sk->sk_callback_lock); transport->inet = NULL; transport->sock = NULL; @@ -863,6 +858,13 @@ static void xs_close(struct rpc_xprt *xprt) xprt_disconnect_done(xprt); } +static void xs_inject_disconnect(struct rpc_xprt *xprt) +{ + dprintk("RPC: injecting transport disconnect on xprt=%p\n", + xprt); + xprt_disconnect_done(xprt); +} + static void xs_xprt_free(struct rpc_xprt *xprt) { xs_free_peer_addresses(xprt); @@ -901,7 +903,6 @@ static int xs_local_copy_to_xdr(struct xdr_buf *xdr, struct sk_buff *skb) /** * xs_local_data_ready - "data ready" callback for AF_LOCAL sockets * @sk: socket with data to read - * @len: how much data to read * * Currently this assumes we can read the whole reply in a single gulp. */ @@ -965,7 +966,6 @@ static void xs_local_data_ready(struct sock *sk) /** * xs_udp_data_ready - "data ready" callback for UDP sockets * @sk: socket with data to read - * @len: how much data to read * */ static void xs_udp_data_ready(struct sock *sk) @@ -1389,7 +1389,6 @@ static int xs_tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, uns /** * xs_tcp_data_ready - "data ready" callback for TCP sockets * @sk: socket with data to read - * @bytes: how much data to read * */ static void xs_tcp_data_ready(struct sock *sk) @@ -1886,9 +1885,7 @@ static int xs_local_finish_connecting(struct rpc_xprt *xprt, /** * xs_local_setup_socket - create AF_LOCAL socket, connect to a local endpoint - * @xprt: RPC transport to connect * @transport: socket transport to connect - * @create_sock: function to create a socket of the correct type */ static int xs_local_setup_socket(struct sock_xprt *transport) { @@ -1960,43 +1957,84 @@ static void xs_local_connect(struct rpc_xprt *xprt, struct rpc_task *task) msleep_interruptible(15000); } -#ifdef CONFIG_SUNRPC_SWAP +#if IS_ENABLED(CONFIG_SUNRPC_SWAP) +/* + * Note that this should be called with XPRT_LOCKED held (or when we otherwise + * know that we have exclusive access to the socket), to guard against + * races with xs_reset_transport. + */ static void xs_set_memalloc(struct rpc_xprt *xprt) { struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); - if (xprt->swapper) + /* + * If there's no sock, then we have nothing to set. The + * reconnecting process will get it for us. + */ + if (!transport->inet) + return; + if (atomic_read(&xprt->swapper)) sk_set_memalloc(transport->inet); } /** - * xs_swapper - Tag this transport as being used for swap. + * xs_enable_swap - Tag this transport as being used for swap. * @xprt: transport to tag - * @enable: enable/disable * + * Take a reference to this transport on behalf of the rpc_clnt, and + * optionally mark it for swapping if it wasn't already. */ -int xs_swapper(struct rpc_xprt *xprt, int enable) +static int +xs_enable_swap(struct rpc_xprt *xprt) { - struct sock_xprt *transport = container_of(xprt, struct sock_xprt, - xprt); - int err = 0; + struct sock_xprt *xs = container_of(xprt, struct sock_xprt, xprt); - if (enable) { - xprt->swapper++; - xs_set_memalloc(xprt); - } else if (xprt->swapper) { - xprt->swapper--; - sk_clear_memalloc(transport->inet); - } + if (atomic_inc_return(&xprt->swapper) != 1) + return 0; + if (wait_on_bit_lock(&xprt->state, XPRT_LOCKED, TASK_KILLABLE)) + return -ERESTARTSYS; + if (xs->inet) + sk_set_memalloc(xs->inet); + xprt_release_xprt(xprt, NULL); + return 0; +} - return err; +/** + * xs_disable_swap - Untag this transport as being used for swap. + * @xprt: transport to tag + * + * Drop a "swapper" reference to this xprt on behalf of the rpc_clnt. If the + * swapper refcount goes to 0, untag the socket as a memalloc socket. + */ +static void +xs_disable_swap(struct rpc_xprt *xprt) +{ + struct sock_xprt *xs = container_of(xprt, struct sock_xprt, xprt); + + if (!atomic_dec_and_test(&xprt->swapper)) + return; + if (wait_on_bit_lock(&xprt->state, XPRT_LOCKED, TASK_KILLABLE)) + return; + if (xs->inet) + sk_clear_memalloc(xs->inet); + xprt_release_xprt(xprt, NULL); } -EXPORT_SYMBOL_GPL(xs_swapper); #else static void xs_set_memalloc(struct rpc_xprt *xprt) { } + +static int +xs_enable_swap(struct rpc_xprt *xprt) +{ + return -EINVAL; +} + +static void +xs_disable_swap(struct rpc_xprt *xprt) +{ +} #endif static void xs_udp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) @@ -2057,6 +2095,27 @@ out: xprt_wake_pending_tasks(xprt, status); } +/** + * xs_tcp_shutdown - gracefully shut down a TCP socket + * @xprt: transport + * + * Initiates a graceful shutdown of the TCP socket by calling the + * equivalent of shutdown(SHUT_RDWR); + */ +static void xs_tcp_shutdown(struct rpc_xprt *xprt) +{ + struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); + struct socket *sock = transport->sock; + + if (sock == NULL) + return; + if (xprt_connected(xprt)) { + kernel_sock_shutdown(sock, SHUT_RDWR); + trace_rpc_socket_shutdown(xprt, sock); + } else + xs_reset_transport(transport); +} + static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) { struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); @@ -2067,6 +2126,7 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) unsigned int keepidle = xprt->timeout->to_initval / HZ; unsigned int keepcnt = xprt->timeout->to_retries + 1; unsigned int opt_on = 1; + unsigned int timeo; /* TCP Keepalive options */ kernel_setsockopt(sock, SOL_SOCKET, SO_KEEPALIVE, @@ -2078,6 +2138,12 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) kernel_setsockopt(sock, SOL_TCP, TCP_KEEPCNT, (char *)&keepcnt, sizeof(keepcnt)); + /* TCP user timeout (see RFC5482) */ + timeo = jiffies_to_msecs(xprt->timeout->to_initval) * + (xprt->timeout->to_retries + 1); + kernel_setsockopt(sock, SOL_TCP, TCP_USER_TIMEOUT, + (char *)&timeo, sizeof(timeo)); + write_lock_bh(&sk->sk_callback_lock); xs_save_old_callbacks(transport, sk); @@ -2125,9 +2191,6 @@ out: /** * xs_tcp_setup_socket - create a TCP socket and connect to a remote endpoint - * @xprt: RPC transport to connect - * @transport: socket transport to connect - * @create_sock: function to create a socket of the correct type * * Invoked by a work queue tasklet. */ @@ -2463,6 +2526,8 @@ static struct rpc_xprt_ops xs_local_ops = { .close = xs_close, .destroy = xs_destroy, .print_stats = xs_local_print_stats, + .enable_swap = xs_enable_swap, + .disable_swap = xs_disable_swap, }; static struct rpc_xprt_ops xs_udp_ops = { @@ -2482,6 +2547,9 @@ static struct rpc_xprt_ops xs_udp_ops = { .close = xs_close, .destroy = xs_destroy, .print_stats = xs_udp_print_stats, + .enable_swap = xs_enable_swap, + .disable_swap = xs_disable_swap, + .inject_disconnect = xs_inject_disconnect, }; static struct rpc_xprt_ops xs_tcp_ops = { @@ -2498,6 +2566,9 @@ static struct rpc_xprt_ops xs_tcp_ops = { .close = xs_tcp_shutdown, .destroy = xs_destroy, .print_stats = xs_tcp_print_stats, + .enable_swap = xs_enable_swap, + .disable_swap = xs_disable_swap, + .inject_disconnect = xs_inject_disconnect, }; /* @@ -2515,6 +2586,9 @@ static struct rpc_xprt_ops bc_tcp_ops = { .close = bc_close, .destroy = bc_destroy, .print_stats = xs_tcp_print_stats, + .enable_swap = xs_enable_swap, + .disable_swap = xs_disable_swap, + .inject_disconnect = xs_inject_disconnect, }; static int xs_init_anyaddr(const int family, struct sockaddr *sap) @@ -2982,7 +3056,7 @@ static int param_set_portnr(const char *val, const struct kernel_param *kp) RPC_MAX_RESVPORT); } -static struct kernel_param_ops param_ops_portnr = { +static const struct kernel_param_ops param_ops_portnr = { .set = param_set_portnr, .get = param_get_uint, }; @@ -3001,7 +3075,7 @@ static int param_set_slot_table_size(const char *val, RPC_MAX_SLOT_TABLE); } -static struct kernel_param_ops param_ops_slot_table_size = { +static const struct kernel_param_ops param_ops_slot_table_size = { .set = param_set_slot_table_size, .get = param_get_uint, }; @@ -3017,7 +3091,7 @@ static int param_set_max_slot_table_size(const char *val, RPC_MAX_SLOT_TABLE_LIMIT); } -static struct kernel_param_ops param_ops_max_slot_table_size = { +static const struct kernel_param_ops param_ops_max_slot_table_size = { .set = param_set_max_slot_table_size, .get = param_get_uint, }; |