--- /dev/null
- smp_mb__before_clear_bit();
+ /*
+ * Device operations for the pnfs nfs4 file layout driver.
+ *
+ * Copyright (c) 2002
+ * The Regents of the University of Michigan
+ * All Rights Reserved
+ *
+ * Dean Hildebrand <dhildebz@umich.edu>
+ * Garth Goodson <Garth.Goodson@netapp.com>
+ *
+ * Permission is granted to use, copy, create derivative works, and
+ * redistribute this software and such derivative works for any purpose,
+ * so long as the name of the University of Michigan is not used in
+ * any advertising or publicity pertaining to the use or distribution
+ * of this software without specific, written prior authorization. If
+ * the above copyright notice or any other identification of the
+ * University of Michigan is included in any copy of any portion of
+ * this software, then the disclaimer below must also be included.
+ *
+ * This software is provided as is, without representation or warranty
+ * of any kind either express or implied, including without limitation
+ * the implied warranties of merchantability, fitness for a particular
+ * purpose, or noninfringement. The Regents of the University of
+ * Michigan shall not be liable for any damages, including special,
+ * indirect, incidental, or consequential damages, with respect to any
+ * claim arising out of or in connection with the use of the software,
+ * even if it has been or is hereafter advised of the possibility of
+ * such damages.
+ */
+
+ #include <linux/nfs_fs.h>
+ #include <linux/vmalloc.h>
+ #include <linux/module.h>
+ #include <linux/sunrpc/addr.h>
+
+ #include "../internal.h"
+ #include "../nfs4session.h"
+ #include "filelayout.h"
+
+ #define NFSDBG_FACILITY NFSDBG_PNFS_LD
+
+ static unsigned int dataserver_timeo = NFS4_DEF_DS_TIMEO;
+ static unsigned int dataserver_retrans = NFS4_DEF_DS_RETRANS;
+
+ /*
+ * Data server cache
+ *
+ * Data servers can be mapped to different device ids.
+ * nfs4_pnfs_ds reference counting
+ * - set to 1 on allocation
+ * - incremented when a device id maps a data server already in the cache.
+ * - decremented when deviceid is removed from the cache.
+ */
+ static DEFINE_SPINLOCK(nfs4_ds_cache_lock);
+ static LIST_HEAD(nfs4_data_server_cache);
+
+ /* Debug routines */
+ void
+ print_ds(struct nfs4_pnfs_ds *ds)
+ {
+ if (ds == NULL) {
+ printk("%s NULL device\n", __func__);
+ return;
+ }
+ printk(" ds %s\n"
+ " ref count %d\n"
+ " client %p\n"
+ " cl_exchange_flags %x\n",
+ ds->ds_remotestr,
+ atomic_read(&ds->ds_count), ds->ds_clp,
+ ds->ds_clp ? ds->ds_clp->cl_exchange_flags : 0);
+ }
+
+ static bool
+ same_sockaddr(struct sockaddr *addr1, struct sockaddr *addr2)
+ {
+ struct sockaddr_in *a, *b;
+ struct sockaddr_in6 *a6, *b6;
+
+ if (addr1->sa_family != addr2->sa_family)
+ return false;
+
+ switch (addr1->sa_family) {
+ case AF_INET:
+ a = (struct sockaddr_in *)addr1;
+ b = (struct sockaddr_in *)addr2;
+
+ if (a->sin_addr.s_addr == b->sin_addr.s_addr &&
+ a->sin_port == b->sin_port)
+ return true;
+ break;
+
+ case AF_INET6:
+ a6 = (struct sockaddr_in6 *)addr1;
+ b6 = (struct sockaddr_in6 *)addr2;
+
+ /* LINKLOCAL addresses must have matching scope_id */
+ if (ipv6_addr_src_scope(&a6->sin6_addr) ==
+ IPV6_ADDR_SCOPE_LINKLOCAL &&
+ a6->sin6_scope_id != b6->sin6_scope_id)
+ return false;
+
+ if (ipv6_addr_equal(&a6->sin6_addr, &b6->sin6_addr) &&
+ a6->sin6_port == b6->sin6_port)
+ return true;
+ break;
+
+ default:
+ dprintk("%s: unhandled address family: %u\n",
+ __func__, addr1->sa_family);
+ return false;
+ }
+
+ return false;
+ }
+
+ static bool
+ _same_data_server_addrs_locked(const struct list_head *dsaddrs1,
+ const struct list_head *dsaddrs2)
+ {
+ struct nfs4_pnfs_ds_addr *da1, *da2;
+
+ /* step through both lists, comparing as we go */
+ for (da1 = list_first_entry(dsaddrs1, typeof(*da1), da_node),
+ da2 = list_first_entry(dsaddrs2, typeof(*da2), da_node);
+ da1 != NULL && da2 != NULL;
+ da1 = list_entry(da1->da_node.next, typeof(*da1), da_node),
+ da2 = list_entry(da2->da_node.next, typeof(*da2), da_node)) {
+ if (!same_sockaddr((struct sockaddr *)&da1->da_addr,
+ (struct sockaddr *)&da2->da_addr))
+ return false;
+ }
+ if (da1 == NULL && da2 == NULL)
+ return true;
+
+ return false;
+ }
+
+ /*
+ * Lookup DS by addresses. nfs4_ds_cache_lock is held
+ */
+ static struct nfs4_pnfs_ds *
+ _data_server_lookup_locked(const struct list_head *dsaddrs)
+ {
+ struct nfs4_pnfs_ds *ds;
+
+ list_for_each_entry(ds, &nfs4_data_server_cache, ds_node)
+ if (_same_data_server_addrs_locked(&ds->ds_addrs, dsaddrs))
+ return ds;
+ return NULL;
+ }
+
+ /*
+ * Create an rpc connection to the nfs4_pnfs_ds data server
+ * Currently only supports IPv4 and IPv6 addresses
+ */
+ static int
+ nfs4_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds)
+ {
+ struct nfs_client *clp = ERR_PTR(-EIO);
+ struct nfs4_pnfs_ds_addr *da;
+ int status = 0;
+
+ dprintk("--> %s DS %s au_flavor %d\n", __func__, ds->ds_remotestr,
+ mds_srv->nfs_client->cl_rpcclient->cl_auth->au_flavor);
+
+ list_for_each_entry(da, &ds->ds_addrs, da_node) {
+ dprintk("%s: DS %s: trying address %s\n",
+ __func__, ds->ds_remotestr, da->da_remotestr);
+
+ clp = nfs4_set_ds_client(mds_srv->nfs_client,
+ (struct sockaddr *)&da->da_addr,
+ da->da_addrlen, IPPROTO_TCP,
+ dataserver_timeo, dataserver_retrans);
+ if (!IS_ERR(clp))
+ break;
+ }
+
+ if (IS_ERR(clp)) {
+ status = PTR_ERR(clp);
+ goto out;
+ }
+
+ status = nfs4_init_ds_session(clp, mds_srv->nfs_client->cl_lease_time);
+ if (status)
+ goto out_put;
+
+ smp_wmb();
+ ds->ds_clp = clp;
+ dprintk("%s [new] addr: %s\n", __func__, ds->ds_remotestr);
+ out:
+ return status;
+ out_put:
+ nfs_put_client(clp);
+ goto out;
+ }
+
+ static void
+ destroy_ds(struct nfs4_pnfs_ds *ds)
+ {
+ struct nfs4_pnfs_ds_addr *da;
+
+ dprintk("--> %s\n", __func__);
+ ifdebug(FACILITY)
+ print_ds(ds);
+
+ if (ds->ds_clp)
+ nfs_put_client(ds->ds_clp);
+
+ while (!list_empty(&ds->ds_addrs)) {
+ da = list_first_entry(&ds->ds_addrs,
+ struct nfs4_pnfs_ds_addr,
+ da_node);
+ list_del_init(&da->da_node);
+ kfree(da->da_remotestr);
+ kfree(da);
+ }
+
+ kfree(ds->ds_remotestr);
+ kfree(ds);
+ }
+
+ void
+ nfs4_fl_free_deviceid(struct nfs4_file_layout_dsaddr *dsaddr)
+ {
+ struct nfs4_pnfs_ds *ds;
+ int i;
+
+ nfs4_print_deviceid(&dsaddr->id_node.deviceid);
+
+ for (i = 0; i < dsaddr->ds_num; i++) {
+ ds = dsaddr->ds_list[i];
+ if (ds != NULL) {
+ if (atomic_dec_and_lock(&ds->ds_count,
+ &nfs4_ds_cache_lock)) {
+ list_del_init(&ds->ds_node);
+ spin_unlock(&nfs4_ds_cache_lock);
+ destroy_ds(ds);
+ }
+ }
+ }
+ kfree(dsaddr->stripe_indices);
+ kfree(dsaddr);
+ }
+
+ /*
+ * Create a string with a human readable address and port to avoid
+ * complicated setup around many dprinks.
+ */
+ static char *
+ nfs4_pnfs_remotestr(struct list_head *dsaddrs, gfp_t gfp_flags)
+ {
+ struct nfs4_pnfs_ds_addr *da;
+ char *remotestr;
+ size_t len;
+ char *p;
+
+ len = 3; /* '{', '}' and eol */
+ list_for_each_entry(da, dsaddrs, da_node) {
+ len += strlen(da->da_remotestr) + 1; /* string plus comma */
+ }
+
+ remotestr = kzalloc(len, gfp_flags);
+ if (!remotestr)
+ return NULL;
+
+ p = remotestr;
+ *(p++) = '{';
+ len--;
+ list_for_each_entry(da, dsaddrs, da_node) {
+ size_t ll = strlen(da->da_remotestr);
+
+ if (ll > len)
+ goto out_err;
+
+ memcpy(p, da->da_remotestr, ll);
+ p += ll;
+ len -= ll;
+
+ if (len < 1)
+ goto out_err;
+ (*p++) = ',';
+ len--;
+ }
+ if (len < 2)
+ goto out_err;
+ *(p++) = '}';
+ *p = '\0';
+ return remotestr;
+ out_err:
+ kfree(remotestr);
+ return NULL;
+ }
+
+ static struct nfs4_pnfs_ds *
+ nfs4_pnfs_ds_add(struct list_head *dsaddrs, gfp_t gfp_flags)
+ {
+ struct nfs4_pnfs_ds *tmp_ds, *ds = NULL;
+ char *remotestr;
+
+ if (list_empty(dsaddrs)) {
+ dprintk("%s: no addresses defined\n", __func__);
+ goto out;
+ }
+
+ ds = kzalloc(sizeof(*ds), gfp_flags);
+ if (!ds)
+ goto out;
+
+ /* this is only used for debugging, so it's ok if its NULL */
+ remotestr = nfs4_pnfs_remotestr(dsaddrs, gfp_flags);
+
+ spin_lock(&nfs4_ds_cache_lock);
+ tmp_ds = _data_server_lookup_locked(dsaddrs);
+ if (tmp_ds == NULL) {
+ INIT_LIST_HEAD(&ds->ds_addrs);
+ list_splice_init(dsaddrs, &ds->ds_addrs);
+ ds->ds_remotestr = remotestr;
+ atomic_set(&ds->ds_count, 1);
+ INIT_LIST_HEAD(&ds->ds_node);
+ ds->ds_clp = NULL;
+ list_add(&ds->ds_node, &nfs4_data_server_cache);
+ dprintk("%s add new data server %s\n", __func__,
+ ds->ds_remotestr);
+ } else {
+ kfree(remotestr);
+ kfree(ds);
+ atomic_inc(&tmp_ds->ds_count);
+ dprintk("%s data server %s found, inc'ed ds_count to %d\n",
+ __func__, tmp_ds->ds_remotestr,
+ atomic_read(&tmp_ds->ds_count));
+ ds = tmp_ds;
+ }
+ spin_unlock(&nfs4_ds_cache_lock);
+ out:
+ return ds;
+ }
+
+ /*
+ * Currently only supports ipv4, ipv6 and one multi-path address.
+ */
+ static struct nfs4_pnfs_ds_addr *
+ decode_ds_addr(struct net *net, struct xdr_stream *streamp, gfp_t gfp_flags)
+ {
+ struct nfs4_pnfs_ds_addr *da = NULL;
+ char *buf, *portstr;
+ __be16 port;
+ int nlen, rlen;
+ int tmp[2];
+ __be32 *p;
+ char *netid, *match_netid;
+ size_t len, match_netid_len;
+ char *startsep = "";
+ char *endsep = "";
+
+
+ /* r_netid */
+ p = xdr_inline_decode(streamp, 4);
+ if (unlikely(!p))
+ goto out_err;
+ nlen = be32_to_cpup(p++);
+
+ p = xdr_inline_decode(streamp, nlen);
+ if (unlikely(!p))
+ goto out_err;
+
+ netid = kmalloc(nlen+1, gfp_flags);
+ if (unlikely(!netid))
+ goto out_err;
+
+ netid[nlen] = '\0';
+ memcpy(netid, p, nlen);
+
+ /* r_addr: ip/ip6addr with port in dec octets - see RFC 5665 */
+ p = xdr_inline_decode(streamp, 4);
+ if (unlikely(!p))
+ goto out_free_netid;
+ rlen = be32_to_cpup(p);
+
+ p = xdr_inline_decode(streamp, rlen);
+ if (unlikely(!p))
+ goto out_free_netid;
+
+ /* port is ".ABC.DEF", 8 chars max */
+ if (rlen > INET6_ADDRSTRLEN + IPV6_SCOPE_ID_LEN + 8) {
+ dprintk("%s: Invalid address, length %d\n", __func__,
+ rlen);
+ goto out_free_netid;
+ }
+ buf = kmalloc(rlen + 1, gfp_flags);
+ if (!buf) {
+ dprintk("%s: Not enough memory\n", __func__);
+ goto out_free_netid;
+ }
+ buf[rlen] = '\0';
+ memcpy(buf, p, rlen);
+
+ /* replace port '.' with '-' */
+ portstr = strrchr(buf, '.');
+ if (!portstr) {
+ dprintk("%s: Failed finding expected dot in port\n",
+ __func__);
+ goto out_free_buf;
+ }
+ *portstr = '-';
+
+ /* find '.' between address and port */
+ portstr = strrchr(buf, '.');
+ if (!portstr) {
+ dprintk("%s: Failed finding expected dot between address and "
+ "port\n", __func__);
+ goto out_free_buf;
+ }
+ *portstr = '\0';
+
+ da = kzalloc(sizeof(*da), gfp_flags);
+ if (unlikely(!da))
+ goto out_free_buf;
+
+ INIT_LIST_HEAD(&da->da_node);
+
+ if (!rpc_pton(net, buf, portstr-buf, (struct sockaddr *)&da->da_addr,
+ sizeof(da->da_addr))) {
+ dprintk("%s: error parsing address %s\n", __func__, buf);
+ goto out_free_da;
+ }
+
+ portstr++;
+ sscanf(portstr, "%d-%d", &tmp[0], &tmp[1]);
+ port = htons((tmp[0] << 8) | (tmp[1]));
+
+ switch (da->da_addr.ss_family) {
+ case AF_INET:
+ ((struct sockaddr_in *)&da->da_addr)->sin_port = port;
+ da->da_addrlen = sizeof(struct sockaddr_in);
+ match_netid = "tcp";
+ match_netid_len = 3;
+ break;
+
+ case AF_INET6:
+ ((struct sockaddr_in6 *)&da->da_addr)->sin6_port = port;
+ da->da_addrlen = sizeof(struct sockaddr_in6);
+ match_netid = "tcp6";
+ match_netid_len = 4;
+ startsep = "[";
+ endsep = "]";
+ break;
+
+ default:
+ dprintk("%s: unsupported address family: %u\n",
+ __func__, da->da_addr.ss_family);
+ goto out_free_da;
+ }
+
+ if (nlen != match_netid_len || strncmp(netid, match_netid, nlen)) {
+ dprintk("%s: ERROR: r_netid \"%s\" != \"%s\"\n",
+ __func__, netid, match_netid);
+ goto out_free_da;
+ }
+
+ /* save human readable address */
+ len = strlen(startsep) + strlen(buf) + strlen(endsep) + 7;
+ da->da_remotestr = kzalloc(len, gfp_flags);
+
+ /* NULL is ok, only used for dprintk */
+ if (da->da_remotestr)
+ snprintf(da->da_remotestr, len, "%s%s%s:%u", startsep,
+ buf, endsep, ntohs(port));
+
+ dprintk("%s: Parsed DS addr %s\n", __func__, da->da_remotestr);
+ kfree(buf);
+ kfree(netid);
+ return da;
+
+ out_free_da:
+ kfree(da);
+ out_free_buf:
+ dprintk("%s: Error parsing DS addr: %s\n", __func__, buf);
+ kfree(buf);
+ out_free_netid:
+ kfree(netid);
+ out_err:
+ return NULL;
+ }
+
+ /* Decode opaque device data and return the result */
+ static struct nfs4_file_layout_dsaddr*
+ decode_device(struct inode *ino, struct pnfs_device *pdev, gfp_t gfp_flags)
+ {
+ int i;
+ u32 cnt, num;
+ u8 *indexp;
+ __be32 *p;
+ u8 *stripe_indices;
+ u8 max_stripe_index;
+ struct nfs4_file_layout_dsaddr *dsaddr = NULL;
+ struct xdr_stream stream;
+ struct xdr_buf buf;
+ struct page *scratch;
+ struct list_head dsaddrs;
+ struct nfs4_pnfs_ds_addr *da;
+
+ /* set up xdr stream */
+ scratch = alloc_page(gfp_flags);
+ if (!scratch)
+ goto out_err;
+
+ xdr_init_decode_pages(&stream, &buf, pdev->pages, pdev->pglen);
+ xdr_set_scratch_buffer(&stream, page_address(scratch), PAGE_SIZE);
+
+ /* Get the stripe count (number of stripe index) */
+ p = xdr_inline_decode(&stream, 4);
+ if (unlikely(!p))
+ goto out_err_free_scratch;
+
+ cnt = be32_to_cpup(p);
+ dprintk("%s stripe count %d\n", __func__, cnt);
+ if (cnt > NFS4_PNFS_MAX_STRIPE_CNT) {
+ printk(KERN_WARNING "NFS: %s: stripe count %d greater than "
+ "supported maximum %d\n", __func__,
+ cnt, NFS4_PNFS_MAX_STRIPE_CNT);
+ goto out_err_free_scratch;
+ }
+
+ /* read stripe indices */
+ stripe_indices = kcalloc(cnt, sizeof(u8), gfp_flags);
+ if (!stripe_indices)
+ goto out_err_free_scratch;
+
+ p = xdr_inline_decode(&stream, cnt << 2);
+ if (unlikely(!p))
+ goto out_err_free_stripe_indices;
+
+ indexp = &stripe_indices[0];
+ max_stripe_index = 0;
+ for (i = 0; i < cnt; i++) {
+ *indexp = be32_to_cpup(p++);
+ max_stripe_index = max(max_stripe_index, *indexp);
+ indexp++;
+ }
+
+ /* Check the multipath list count */
+ p = xdr_inline_decode(&stream, 4);
+ if (unlikely(!p))
+ goto out_err_free_stripe_indices;
+
+ num = be32_to_cpup(p);
+ dprintk("%s ds_num %u\n", __func__, num);
+ if (num > NFS4_PNFS_MAX_MULTI_CNT) {
+ printk(KERN_WARNING "NFS: %s: multipath count %d greater than "
+ "supported maximum %d\n", __func__,
+ num, NFS4_PNFS_MAX_MULTI_CNT);
+ goto out_err_free_stripe_indices;
+ }
+
+ /* validate stripe indices are all < num */
+ if (max_stripe_index >= num) {
+ printk(KERN_WARNING "NFS: %s: stripe index %u >= num ds %u\n",
+ __func__, max_stripe_index, num);
+ goto out_err_free_stripe_indices;
+ }
+
+ dsaddr = kzalloc(sizeof(*dsaddr) +
+ (sizeof(struct nfs4_pnfs_ds *) * (num - 1)),
+ gfp_flags);
+ if (!dsaddr)
+ goto out_err_free_stripe_indices;
+
+ dsaddr->stripe_count = cnt;
+ dsaddr->stripe_indices = stripe_indices;
+ stripe_indices = NULL;
+ dsaddr->ds_num = num;
+ nfs4_init_deviceid_node(&dsaddr->id_node,
+ NFS_SERVER(ino)->pnfs_curr_ld,
+ NFS_SERVER(ino)->nfs_client,
+ &pdev->dev_id);
+
+ INIT_LIST_HEAD(&dsaddrs);
+
+ for (i = 0; i < dsaddr->ds_num; i++) {
+ int j;
+ u32 mp_count;
+
+ p = xdr_inline_decode(&stream, 4);
+ if (unlikely(!p))
+ goto out_err_free_deviceid;
+
+ mp_count = be32_to_cpup(p); /* multipath count */
+ for (j = 0; j < mp_count; j++) {
+ da = decode_ds_addr(NFS_SERVER(ino)->nfs_client->cl_net,
+ &stream, gfp_flags);
+ if (da)
+ list_add_tail(&da->da_node, &dsaddrs);
+ }
+ if (list_empty(&dsaddrs)) {
+ dprintk("%s: no suitable DS addresses found\n",
+ __func__);
+ goto out_err_free_deviceid;
+ }
+
+ dsaddr->ds_list[i] = nfs4_pnfs_ds_add(&dsaddrs, gfp_flags);
+ if (!dsaddr->ds_list[i])
+ goto out_err_drain_dsaddrs;
+
+ /* If DS was already in cache, free ds addrs */
+ while (!list_empty(&dsaddrs)) {
+ da = list_first_entry(&dsaddrs,
+ struct nfs4_pnfs_ds_addr,
+ da_node);
+ list_del_init(&da->da_node);
+ kfree(da->da_remotestr);
+ kfree(da);
+ }
+ }
+
+ __free_page(scratch);
+ return dsaddr;
+
+ out_err_drain_dsaddrs:
+ while (!list_empty(&dsaddrs)) {
+ da = list_first_entry(&dsaddrs, struct nfs4_pnfs_ds_addr,
+ da_node);
+ list_del_init(&da->da_node);
+ kfree(da->da_remotestr);
+ kfree(da);
+ }
+ out_err_free_deviceid:
+ nfs4_fl_free_deviceid(dsaddr);
+ /* stripe_indicies was part of dsaddr */
+ goto out_err_free_scratch;
+ out_err_free_stripe_indices:
+ kfree(stripe_indices);
+ out_err_free_scratch:
+ __free_page(scratch);
+ out_err:
+ dprintk("%s ERROR: returning NULL\n", __func__);
+ return NULL;
+ }
+
+ /*
+ * Decode the opaque device specified in 'dev' and add it to the cache of
+ * available devices.
+ */
+ static struct nfs4_file_layout_dsaddr *
+ decode_and_add_device(struct inode *inode, struct pnfs_device *dev, gfp_t gfp_flags)
+ {
+ struct nfs4_deviceid_node *d;
+ struct nfs4_file_layout_dsaddr *n, *new;
+
+ new = decode_device(inode, dev, gfp_flags);
+ if (!new) {
+ printk(KERN_WARNING "NFS: %s: Could not decode or add device\n",
+ __func__);
+ return NULL;
+ }
+
+ d = nfs4_insert_deviceid_node(&new->id_node);
+ n = container_of(d, struct nfs4_file_layout_dsaddr, id_node);
+ if (n != new) {
+ nfs4_fl_free_deviceid(new);
+ return n;
+ }
+
+ return new;
+ }
+
+ /*
+ * Retrieve the information for dev_id, add it to the list
+ * of available devices, and return it.
+ */
+ struct nfs4_file_layout_dsaddr *
+ filelayout_get_device_info(struct inode *inode,
+ struct nfs4_deviceid *dev_id,
+ struct rpc_cred *cred,
+ gfp_t gfp_flags)
+ {
+ struct pnfs_device *pdev = NULL;
+ u32 max_resp_sz;
+ int max_pages;
+ struct page **pages = NULL;
+ struct nfs4_file_layout_dsaddr *dsaddr = NULL;
+ int rc, i;
+ struct nfs_server *server = NFS_SERVER(inode);
+
+ /*
+ * Use the session max response size as the basis for setting
+ * GETDEVICEINFO's maxcount
+ */
+ max_resp_sz = server->nfs_client->cl_session->fc_attrs.max_resp_sz;
+ max_pages = nfs_page_array_len(0, max_resp_sz);
+ dprintk("%s inode %p max_resp_sz %u max_pages %d\n",
+ __func__, inode, max_resp_sz, max_pages);
+
+ pdev = kzalloc(sizeof(struct pnfs_device), gfp_flags);
+ if (pdev == NULL)
+ return NULL;
+
+ pages = kzalloc(max_pages * sizeof(struct page *), gfp_flags);
+ if (pages == NULL) {
+ kfree(pdev);
+ return NULL;
+ }
+ for (i = 0; i < max_pages; i++) {
+ pages[i] = alloc_page(gfp_flags);
+ if (!pages[i])
+ goto out_free;
+ }
+
+ memcpy(&pdev->dev_id, dev_id, sizeof(*dev_id));
+ pdev->layout_type = LAYOUT_NFSV4_1_FILES;
+ pdev->pages = pages;
+ pdev->pgbase = 0;
+ pdev->pglen = max_resp_sz;
+ pdev->mincount = 0;
+ pdev->maxcount = max_resp_sz - nfs41_maxgetdevinfo_overhead;
+
+ rc = nfs4_proc_getdeviceinfo(server, pdev, cred);
+ dprintk("%s getdevice info returns %d\n", __func__, rc);
+ if (rc)
+ goto out_free;
+
+ /*
+ * Found new device, need to decode it and then add it to the
+ * list of known devices for this mountpoint.
+ */
+ dsaddr = decode_and_add_device(inode, pdev, gfp_flags);
+ out_free:
+ for (i = 0; i < max_pages; i++)
+ __free_page(pages[i]);
+ kfree(pages);
+ kfree(pdev);
+ dprintk("<-- %s dsaddr %p\n", __func__, dsaddr);
+ return dsaddr;
+ }
+
+ void
+ nfs4_fl_put_deviceid(struct nfs4_file_layout_dsaddr *dsaddr)
+ {
+ nfs4_put_deviceid_node(&dsaddr->id_node);
+ }
+
+ /*
+ * Want res = (offset - layout->pattern_offset)/ layout->stripe_unit
+ * Then: ((res + fsi) % dsaddr->stripe_count)
+ */
+ u32
+ nfs4_fl_calc_j_index(struct pnfs_layout_segment *lseg, loff_t offset)
+ {
+ struct nfs4_filelayout_segment *flseg = FILELAYOUT_LSEG(lseg);
+ u64 tmp;
+
+ tmp = offset - flseg->pattern_offset;
+ do_div(tmp, flseg->stripe_unit);
+ tmp += flseg->first_stripe_index;
+ return do_div(tmp, flseg->dsaddr->stripe_count);
+ }
+
+ u32
+ nfs4_fl_calc_ds_index(struct pnfs_layout_segment *lseg, u32 j)
+ {
+ return FILELAYOUT_LSEG(lseg)->dsaddr->stripe_indices[j];
+ }
+
+ struct nfs_fh *
+ nfs4_fl_select_ds_fh(struct pnfs_layout_segment *lseg, u32 j)
+ {
+ struct nfs4_filelayout_segment *flseg = FILELAYOUT_LSEG(lseg);
+ u32 i;
+
+ if (flseg->stripe_type == STRIPE_SPARSE) {
+ if (flseg->num_fh == 1)
+ i = 0;
+ else if (flseg->num_fh == 0)
+ /* Use the MDS OPEN fh set in nfs_read_rpcsetup */
+ return NULL;
+ else
+ i = nfs4_fl_calc_ds_index(lseg, j);
+ } else
+ i = j;
+ return flseg->fh_array[i];
+ }
+
+ static void nfs4_wait_ds_connect(struct nfs4_pnfs_ds *ds)
+ {
+ might_sleep();
+ wait_on_bit(&ds->ds_state, NFS4DS_CONNECTING,
+ nfs_wait_bit_killable, TASK_KILLABLE);
+ }
+
+ static void nfs4_clear_ds_conn_bit(struct nfs4_pnfs_ds *ds)
+ {
- smp_mb__after_clear_bit();
++ smp_mb__before_atomic();
+ clear_bit(NFS4DS_CONNECTING, &ds->ds_state);
++ smp_mb__after_atomic();
+ wake_up_bit(&ds->ds_state, NFS4DS_CONNECTING);
+ }
+
+
+ struct nfs4_pnfs_ds *
+ nfs4_fl_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx)
+ {
+ struct nfs4_file_layout_dsaddr *dsaddr = FILELAYOUT_LSEG(lseg)->dsaddr;
+ struct nfs4_pnfs_ds *ds = dsaddr->ds_list[ds_idx];
+ struct nfs4_deviceid_node *devid = FILELAYOUT_DEVID_NODE(lseg);
+ struct nfs4_pnfs_ds *ret = ds;
+
+ if (ds == NULL) {
+ printk(KERN_ERR "NFS: %s: No data server for offset index %d\n",
+ __func__, ds_idx);
+ filelayout_mark_devid_invalid(devid);
+ goto out;
+ }
+ smp_rmb();
+ if (ds->ds_clp)
+ goto out_test_devid;
+
+ if (test_and_set_bit(NFS4DS_CONNECTING, &ds->ds_state) == 0) {
+ struct nfs_server *s = NFS_SERVER(lseg->pls_layout->plh_inode);
+ int err;
+
+ err = nfs4_ds_connect(s, ds);
+ if (err)
+ nfs4_mark_deviceid_unavailable(devid);
+ nfs4_clear_ds_conn_bit(ds);
+ } else {
+ /* Either ds is connected, or ds is NULL */
+ nfs4_wait_ds_connect(ds);
+ }
+ out_test_devid:
+ if (filelayout_test_devid_unavailable(devid))
+ ret = NULL;
+ out:
+ return ret;
+ }
+
+ module_param(dataserver_retrans, uint, 0644);
+ MODULE_PARM_DESC(dataserver_retrans, "The number of times the NFSv4.1 client "
+ "retries a request before it attempts further "
+ " recovery action.");
+ module_param(dataserver_timeo, uint, 0644);
+ MODULE_PARM_DESC(dataserver_timeo, "The time (in tenths of a second) the "
+ "NFSv4.1 client waits for a response from a "
+ " data server before it retries an NFS request.");
trace_nfs_invalidate_mapping_exit(inode, ret);
clear_bit_unlock(NFS_INO_INVALIDATING, bitlock);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
wake_up_bit(bitlock, NFS_INO_INVALIDATING);
out:
return ret;
inode->i_version = fattr->change_attr;
}
} else if (server->caps & NFS_CAP_CHANGE_ATTR)
- invalid |= save_cache_validity;
+ nfsi->cache_validity |= save_cache_validity;
if (fattr->valid & NFS_ATTR_FATTR_MTIME) {
memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime));
} else if (server->caps & NFS_CAP_MTIME)
- invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_ATTR
| NFS_INO_REVAL_FORCED);
if (fattr->valid & NFS_ATTR_FATTR_CTIME) {
memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime));
} else if (server->caps & NFS_CAP_CTIME)
- invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_ATTR
| NFS_INO_REVAL_FORCED);
/* Check if our cached file size is stale */
(long long)new_isize);
}
} else
- invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_ATTR
| NFS_INO_REVAL_PAGECACHE
| NFS_INO_REVAL_FORCED);
if (fattr->valid & NFS_ATTR_FATTR_ATIME)
memcpy(&inode->i_atime, &fattr->atime, sizeof(inode->i_atime));
else if (server->caps & NFS_CAP_ATIME)
- invalid |= save_cache_validity & (NFS_INO_INVALID_ATIME
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_ATIME
| NFS_INO_REVAL_FORCED);
if (fattr->valid & NFS_ATTR_FATTR_MODE) {
invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
}
} else if (server->caps & NFS_CAP_MODE)
- invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_ATTR
| NFS_INO_INVALID_ACCESS
| NFS_INO_INVALID_ACL
| NFS_INO_REVAL_FORCED);
inode->i_uid = fattr->uid;
}
} else if (server->caps & NFS_CAP_OWNER)
- invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_ATTR
| NFS_INO_INVALID_ACCESS
| NFS_INO_INVALID_ACL
| NFS_INO_REVAL_FORCED);
inode->i_gid = fattr->gid;
}
} else if (server->caps & NFS_CAP_OWNER_GROUP)
- invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_ATTR
| NFS_INO_INVALID_ACCESS
| NFS_INO_INVALID_ACL
| NFS_INO_REVAL_FORCED);
set_nlink(inode, fattr->nlink);
}
} else if (server->caps & NFS_CAP_NLINK)
- invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_ATTR
| NFS_INO_REVAL_FORCED);
if (fattr->valid & NFS_ATTR_FATTR_SPACE_USED) {
return status;
}
if (!(o_res->f_attr->valid & NFS_ATTR_FATTR))
- _nfs4_proc_getattr(server, &o_res->fh, o_res->f_attr, o_res->f_label);
+ nfs4_proc_getattr(server, &o_res->fh, o_res->f_attr, o_res->f_label);
return 0;
}
#define FATTR4_WORD1_NFS40_MASK (2*FATTR4_WORD1_MOUNTED_ON_FILEID - 1UL)
#define FATTR4_WORD2_NFS41_MASK (2*FATTR4_WORD2_SUPPATTR_EXCLCREAT - 1UL)
-#define FATTR4_WORD2_NFS42_MASK (2*FATTR4_WORD2_CHANGE_SECURITY_LABEL - 1UL)
+#define FATTR4_WORD2_NFS42_MASK (2*FATTR4_WORD2_SECURITY_LABEL - 1UL)
static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle)
{
return false;
}
- void __nfs4_read_done_cb(struct nfs_read_data *data)
+ void __nfs4_read_done_cb(struct nfs_pgio_data *data)
{
nfs_invalidate_atime(data->header->inode);
}
- static int nfs4_read_done_cb(struct rpc_task *task, struct nfs_read_data *data)
+ static int nfs4_read_done_cb(struct rpc_task *task, struct nfs_pgio_data *data)
{
struct nfs_server *server = NFS_SERVER(data->header->inode);
}
static bool nfs4_read_stateid_changed(struct rpc_task *task,
- struct nfs_readargs *args)
+ struct nfs_pgio_args *args)
{
if (!nfs4_error_stateid_expired(task->tk_status) ||
return true;
}
- static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data)
+ static int nfs4_read_done(struct rpc_task *task, struct nfs_pgio_data *data)
{
dprintk("--> %s\n", __func__);
return -EAGAIN;
if (nfs4_read_stateid_changed(task, &data->args))
return -EAGAIN;
- return data->read_done_cb ? data->read_done_cb(task, data) :
+ return data->pgio_done_cb ? data->pgio_done_cb(task, data) :
nfs4_read_done_cb(task, data);
}
- static void nfs4_proc_read_setup(struct nfs_read_data *data, struct rpc_message *msg)
+ static void nfs4_proc_read_setup(struct nfs_pgio_data *data, struct rpc_message *msg)
{
data->timestamp = jiffies;
- data->read_done_cb = nfs4_read_done_cb;
+ data->pgio_done_cb = nfs4_read_done_cb;
msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ];
nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0);
}
- static int nfs4_proc_read_rpc_prepare(struct rpc_task *task, struct nfs_read_data *data)
+ static int nfs4_proc_pgio_rpc_prepare(struct rpc_task *task, struct nfs_pgio_data *data)
{
if (nfs4_setup_sequence(NFS_SERVER(data->header->inode),
&data->args.seq_args,
task))
return 0;
if (nfs4_set_rw_stateid(&data->args.stateid, data->args.context,
- data->args.lock_context, FMODE_READ) == -EIO)
+ data->args.lock_context, data->header->rw_ops->rw_mode) == -EIO)
return -EIO;
if (unlikely(test_bit(NFS_CONTEXT_BAD, &data->args.context->flags)))
return -EIO;
return 0;
}
- static int nfs4_write_done_cb(struct rpc_task *task, struct nfs_write_data *data)
+ static int nfs4_write_done_cb(struct rpc_task *task, struct nfs_pgio_data *data)
{
struct inode *inode = data->header->inode;
}
static bool nfs4_write_stateid_changed(struct rpc_task *task,
- struct nfs_writeargs *args)
+ struct nfs_pgio_args *args)
{
if (!nfs4_error_stateid_expired(task->tk_status) ||
return true;
}
- static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data)
+ static int nfs4_write_done(struct rpc_task *task, struct nfs_pgio_data *data)
{
if (!nfs4_sequence_done(task, &data->res.seq_res))
return -EAGAIN;
if (nfs4_write_stateid_changed(task, &data->args))
return -EAGAIN;
- return data->write_done_cb ? data->write_done_cb(task, data) :
+ return data->pgio_done_cb ? data->pgio_done_cb(task, data) :
nfs4_write_done_cb(task, data);
}
static
- bool nfs4_write_need_cache_consistency_data(const struct nfs_write_data *data)
+ bool nfs4_write_need_cache_consistency_data(const struct nfs_pgio_data *data)
{
const struct nfs_pgio_header *hdr = data->header;
return nfs4_have_delegation(hdr->inode, FMODE_READ) == 0;
}
- static void nfs4_proc_write_setup(struct nfs_write_data *data, struct rpc_message *msg)
+ static void nfs4_proc_write_setup(struct nfs_pgio_data *data, struct rpc_message *msg)
{
struct nfs_server *server = NFS_SERVER(data->header->inode);
} else
data->args.bitmask = server->cache_consistency_bitmask;
- if (!data->write_done_cb)
- data->write_done_cb = nfs4_write_done_cb;
+ if (!data->pgio_done_cb)
+ data->pgio_done_cb = nfs4_write_done_cb;
data->res.server = server;
data->timestamp = jiffies;
nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1);
}
- static int nfs4_proc_write_rpc_prepare(struct rpc_task *task, struct nfs_write_data *data)
- {
- if (nfs4_setup_sequence(NFS_SERVER(data->header->inode),
- &data->args.seq_args,
- &data->res.seq_res,
- task))
- return 0;
- if (nfs4_set_rw_stateid(&data->args.stateid, data->args.context,
- data->args.lock_context, FMODE_WRITE) == -EIO)
- return -EIO;
- if (unlikely(test_bit(NFS_CONTEXT_BAD, &data->args.context->flags)))
- return -EIO;
- return 0;
- }
-
static void nfs4_proc_commit_rpc_prepare(struct rpc_task *task, struct nfs_commit_data *data)
{
nfs4_setup_sequence(NFS_SERVER(data->inode),
.pathconf = nfs4_proc_pathconf,
.set_capabilities = nfs4_server_capabilities,
.decode_dirent = nfs4_decode_dirent,
+ .pgio_rpc_prepare = nfs4_proc_pgio_rpc_prepare,
.read_setup = nfs4_proc_read_setup,
- .read_pageio_init = pnfs_pageio_init_read,
- .read_rpc_prepare = nfs4_proc_read_rpc_prepare,
.read_done = nfs4_read_done,
.write_setup = nfs4_proc_write_setup,
- .write_pageio_init = pnfs_pageio_init_write,
- .write_rpc_prepare = nfs4_proc_write_rpc_prepare,
.write_done = nfs4_write_done,
.commit_setup = nfs4_proc_commit_setup,
.commit_rpc_prepare = nfs4_proc_commit_rpc_prepare,
static void nfs4_clear_state_manager_bit(struct nfs_client *clp)
{
- smp_mb__before_clear_bit();
+ smp_mb__before_atomic();
clear_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
wake_up_bit(&clp->cl_state, NFS4CLNT_MANAGER_RUNNING);
rpc_wake_up(&clp->cl_rpcwaitq);
}
* server that doesn't support a grace period.
*/
spin_lock(&sp->so_lock);
- write_seqcount_begin(&sp->so_reclaim_seqcount);
+ raw_write_seqcount_begin(&sp->so_reclaim_seqcount);
restart:
list_for_each_entry(state, &sp->so_states, open_states) {
if (!test_and_clear_bit(ops->state_flag_bit, &state->flags))
spin_lock(&sp->so_lock);
goto restart;
}
- write_seqcount_end(&sp->so_reclaim_seqcount);
+ raw_write_seqcount_end(&sp->so_reclaim_seqcount);
spin_unlock(&sp->so_lock);
return 0;
out_err:
nfs4_put_open_state(state);
spin_lock(&sp->so_lock);
- write_seqcount_end(&sp->so_reclaim_seqcount);
+ raw_write_seqcount_end(&sp->so_reclaim_seqcount);
spin_unlock(&sp->so_lock);
return status;
}
#include "internal.h"
#include "pnfs.h"
+ #define NFSDBG_FACILITY NFSDBG_PAGECACHE
+
static struct kmem_cache *nfs_page_cachep;
+ static const struct rpc_call_ops nfs_pgio_common_ops;
+
+ static void nfs_free_request(struct nfs_page *);
- bool nfs_pgarray_set(struct nfs_page_array *p, unsigned int pagecount)
+ static bool nfs_pgarray_set(struct nfs_page_array *p, unsigned int pagecount)
{
p->npages = pagecount;
if (pagecount <= ARRAY_SIZE(p->page_array))
{
if (atomic_dec_and_test(&c->io_count)) {
clear_bit(NFS_IO_INPROGRESS, &c->flags);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
wake_up_bit(&c->flags, NFS_IO_INPROGRESS);
}
}
return __nfs_iocounter_wait(c);
}
- smp_mb__before_clear_bit();
+ static int nfs_wait_bit_uninterruptible(void *word)
+ {
+ io_schedule();
+ return 0;
+ }
+
+ /*
+ * nfs_page_group_lock - lock the head of the page group
+ * @req - request in group that is to be locked
+ *
+ * this lock must be held if modifying the page group list
+ */
+ void
+ nfs_page_group_lock(struct nfs_page *req)
+ {
+ struct nfs_page *head = req->wb_head;
+
+ WARN_ON_ONCE(head != head->wb_head);
+
+ wait_on_bit_lock(&head->wb_flags, PG_HEADLOCK,
+ nfs_wait_bit_uninterruptible,
+ TASK_UNINTERRUPTIBLE);
+ }
+
+ /*
+ * nfs_page_group_unlock - unlock the head of the page group
+ * @req - request in group that is to be unlocked
+ */
+ void
+ nfs_page_group_unlock(struct nfs_page *req)
+ {
+ struct nfs_page *head = req->wb_head;
+
+ WARN_ON_ONCE(head != head->wb_head);
+
- smp_mb__after_clear_bit();
++ smp_mb__before_atomic();
+ clear_bit(PG_HEADLOCK, &head->wb_flags);
++ smp_mb__after_atomic();
+ wake_up_bit(&head->wb_flags, PG_HEADLOCK);
+ }
+
+ /*
+ * nfs_page_group_sync_on_bit_locked
+ *
+ * must be called with page group lock held
+ */
+ static bool
+ nfs_page_group_sync_on_bit_locked(struct nfs_page *req, unsigned int bit)
+ {
+ struct nfs_page *head = req->wb_head;
+ struct nfs_page *tmp;
+
+ WARN_ON_ONCE(!test_bit(PG_HEADLOCK, &head->wb_flags));
+ WARN_ON_ONCE(test_and_set_bit(bit, &req->wb_flags));
+
+ tmp = req->wb_this_page;
+ while (tmp != req) {
+ if (!test_bit(bit, &tmp->wb_flags))
+ return false;
+ tmp = tmp->wb_this_page;
+ }
+
+ /* true! reset all bits */
+ tmp = req;
+ do {
+ clear_bit(bit, &tmp->wb_flags);
+ tmp = tmp->wb_this_page;
+ } while (tmp != req);
+
+ return true;
+ }
+
+ /*
+ * nfs_page_group_sync_on_bit - set bit on current request, but only
+ * return true if the bit is set for all requests in page group
+ * @req - request in page group
+ * @bit - PG_* bit that is used to sync page group
+ */
+ bool nfs_page_group_sync_on_bit(struct nfs_page *req, unsigned int bit)
+ {
+ bool ret;
+
+ nfs_page_group_lock(req);
+ ret = nfs_page_group_sync_on_bit_locked(req, bit);
+ nfs_page_group_unlock(req);
+
+ return ret;
+ }
+
+ /*
+ * nfs_page_group_init - Initialize the page group linkage for @req
+ * @req - a new nfs request
+ * @prev - the previous request in page group, or NULL if @req is the first
+ * or only request in the group (the head).
+ */
+ static inline void
+ nfs_page_group_init(struct nfs_page *req, struct nfs_page *prev)
+ {
+ WARN_ON_ONCE(prev == req);
+
+ if (!prev) {
+ req->wb_head = req;
+ req->wb_this_page = req;
+ } else {
+ WARN_ON_ONCE(prev->wb_this_page != prev->wb_head);
+ WARN_ON_ONCE(!test_bit(PG_HEADLOCK, &prev->wb_head->wb_flags));
+ req->wb_head = prev->wb_head;
+ req->wb_this_page = prev->wb_this_page;
+ prev->wb_this_page = req;
+
+ /* grab extra ref if head request has extra ref from
+ * the write/commit path to handle handoff between write
+ * and commit lists */
+ if (test_bit(PG_INODE_REF, &prev->wb_head->wb_flags))
+ kref_get(&req->wb_kref);
+ }
+ }
+
+ /*
+ * nfs_page_group_destroy - sync the destruction of page groups
+ * @req - request that no longer needs the page group
+ *
+ * releases the page group reference from each member once all
+ * members have called this function.
+ */
+ static void
+ nfs_page_group_destroy(struct kref *kref)
+ {
+ struct nfs_page *req = container_of(kref, struct nfs_page, wb_kref);
+ struct nfs_page *tmp, *next;
+
+ if (!nfs_page_group_sync_on_bit(req, PG_TEARDOWN))
+ return;
+
+ tmp = req;
+ do {
+ next = tmp->wb_this_page;
+ /* unlink and free */
+ tmp->wb_this_page = tmp;
+ tmp->wb_head = tmp;
+ nfs_free_request(tmp);
+ tmp = next;
+ } while (tmp != req);
+ }
+
/**
* nfs_create_request - Create an NFS read/write request.
* @ctx: open context to use
- * @inode: inode to which the request is attached
* @page: page to write
+ * @last: last nfs request created for this page group or NULL if head
* @offset: starting offset within the page for the write
* @count: number of bytes to read/write
*
* User should ensure it is safe to sleep in this function.
*/
struct nfs_page *
- nfs_create_request(struct nfs_open_context *ctx, struct inode *inode,
- struct page *page,
- unsigned int offset, unsigned int count)
+ nfs_create_request(struct nfs_open_context *ctx, struct page *page,
+ struct nfs_page *last, unsigned int offset,
+ unsigned int count)
{
struct nfs_page *req;
struct nfs_lock_context *l_ctx;
req->wb_bytes = count;
req->wb_context = get_nfs_open_context(ctx);
kref_init(&req->wb_kref);
+ nfs_page_group_init(req, last);
return req;
}
printk(KERN_ERR "NFS: Invalid unlock attempted\n");
BUG();
}
- smp_mb__before_clear_bit();
+ smp_mb__before_atomic();
clear_bit(PG_BUSY, &req->wb_flags);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
wake_up_bit(&req->wb_flags, PG_BUSY);
}
}
}
-
/**
* nfs_release_request - Release the count on an NFS read/write request
* @req: request to release
*
* Note: Should never be called with the spinlock held!
*/
- static void nfs_free_request(struct kref *kref)
+ static void nfs_free_request(struct nfs_page *req)
{
- struct nfs_page *req = container_of(kref, struct nfs_page, wb_kref);
+ WARN_ON_ONCE(req->wb_this_page != req);
+
+ /* extra debug: make sure no sync bits are still set */
+ WARN_ON_ONCE(test_bit(PG_TEARDOWN, &req->wb_flags));
+ WARN_ON_ONCE(test_bit(PG_UNLOCKPAGE, &req->wb_flags));
+ WARN_ON_ONCE(test_bit(PG_UPTODATE, &req->wb_flags));
+ WARN_ON_ONCE(test_bit(PG_WB_END, &req->wb_flags));
+ WARN_ON_ONCE(test_bit(PG_REMOVE, &req->wb_flags));
/* Release struct file and open context */
nfs_clear_request(req);
void nfs_release_request(struct nfs_page *req)
{
- kref_put(&req->wb_kref, nfs_free_request);
- }
-
- static int nfs_wait_bit_uninterruptible(void *word)
- {
- io_schedule();
- return 0;
+ kref_put(&req->wb_kref, nfs_page_group_destroy);
}
/**
TASK_UNINTERRUPTIBLE);
}
- bool nfs_generic_pg_test(struct nfs_pageio_descriptor *desc, struct nfs_page *prev, struct nfs_page *req)
+ /*
+ * nfs_generic_pg_test - determine if requests can be coalesced
+ * @desc: pointer to descriptor
+ * @prev: previous request in desc, or NULL
+ * @req: this request
+ *
+ * Returns zero if @req can be coalesced into @desc, otherwise it returns
+ * the size of the request.
+ */
+ size_t nfs_generic_pg_test(struct nfs_pageio_descriptor *desc,
+ struct nfs_page *prev, struct nfs_page *req)
{
- /*
- * FIXME: ideally we should be able to coalesce all requests
- * that are not block boundary aligned, but currently this
- * is problematic for the case of bsize < PAGE_CACHE_SIZE,
- * since nfs_flush_multi and nfs_pagein_multi assume you
- * can have only one struct nfs_page.
- */
- if (desc->pg_bsize < PAGE_SIZE)
+ if (desc->pg_count > desc->pg_bsize) {
+ /* should never happen */
+ WARN_ON_ONCE(1);
return 0;
+ }
- return desc->pg_count + req->wb_bytes <= desc->pg_bsize;
+ return min(desc->pg_bsize - desc->pg_count, (size_t)req->wb_bytes);
}
EXPORT_SYMBOL_GPL(nfs_generic_pg_test);
+ static inline struct nfs_rw_header *NFS_RW_HEADER(struct nfs_pgio_header *hdr)
+ {
+ return container_of(hdr, struct nfs_rw_header, header);
+ }
+
+ /**
+ * nfs_rw_header_alloc - Allocate a header for a read or write
+ * @ops: Read or write function vector
+ */
+ struct nfs_rw_header *nfs_rw_header_alloc(const struct nfs_rw_ops *ops)
+ {
+ struct nfs_rw_header *header = ops->rw_alloc_header();
+
+ if (header) {
+ struct nfs_pgio_header *hdr = &header->header;
+
+ INIT_LIST_HEAD(&hdr->pages);
+ spin_lock_init(&hdr->lock);
+ atomic_set(&hdr->refcnt, 0);
+ hdr->rw_ops = ops;
+ }
+ return header;
+ }
+ EXPORT_SYMBOL_GPL(nfs_rw_header_alloc);
+
+ /*
+ * nfs_rw_header_free - Free a read or write header
+ * @hdr: The header to free
+ */
+ void nfs_rw_header_free(struct nfs_pgio_header *hdr)
+ {
+ hdr->rw_ops->rw_free_header(NFS_RW_HEADER(hdr));
+ }
+ EXPORT_SYMBOL_GPL(nfs_rw_header_free);
+
+ /**
+ * nfs_pgio_data_alloc - Allocate pageio data
+ * @hdr: The header making a request
+ * @pagecount: Number of pages to create
+ */
+ static struct nfs_pgio_data *nfs_pgio_data_alloc(struct nfs_pgio_header *hdr,
+ unsigned int pagecount)
+ {
+ struct nfs_pgio_data *data, *prealloc;
+
+ prealloc = &NFS_RW_HEADER(hdr)->rpc_data;
+ if (prealloc->header == NULL)
+ data = prealloc;
+ else
+ data = kzalloc(sizeof(*data), GFP_KERNEL);
+ if (!data)
+ goto out;
+
+ if (nfs_pgarray_set(&data->pages, pagecount)) {
+ data->header = hdr;
+ atomic_inc(&hdr->refcnt);
+ } else {
+ if (data != prealloc)
+ kfree(data);
+ data = NULL;
+ }
+ out:
+ return data;
+ }
+
+ /**
+ * nfs_pgio_data_release - Properly free pageio data
+ * @data: The data to release
+ */
+ void nfs_pgio_data_release(struct nfs_pgio_data *data)
+ {
+ struct nfs_pgio_header *hdr = data->header;
+ struct nfs_rw_header *pageio_header = NFS_RW_HEADER(hdr);
+
+ put_nfs_open_context(data->args.context);
+ if (data->pages.pagevec != data->pages.page_array)
+ kfree(data->pages.pagevec);
+ if (data == &pageio_header->rpc_data) {
+ data->header = NULL;
+ data = NULL;
+ }
+ if (atomic_dec_and_test(&hdr->refcnt))
+ hdr->completion_ops->completion(hdr);
+ /* Note: we only free the rpc_task after callbacks are done.
+ * See the comment in rpc_free_task() for why
+ */
+ kfree(data);
+ }
+ EXPORT_SYMBOL_GPL(nfs_pgio_data_release);
+
+ /**
+ * nfs_pgio_rpcsetup - Set up arguments for a pageio call
+ * @data: The pageio data
+ * @count: Number of bytes to read
+ * @offset: Initial offset
+ * @how: How to commit data (writes only)
+ * @cinfo: Commit information for the call (writes only)
+ */
+ static void nfs_pgio_rpcsetup(struct nfs_pgio_data *data,
+ unsigned int count, unsigned int offset,
+ int how, struct nfs_commit_info *cinfo)
+ {
+ struct nfs_page *req = data->header->req;
+
+ /* Set up the RPC argument and reply structs
+ * NB: take care not to mess about with data->commit et al. */
+
+ data->args.fh = NFS_FH(data->header->inode);
+ data->args.offset = req_offset(req) + offset;
+ /* pnfs_set_layoutcommit needs this */
+ data->mds_offset = data->args.offset;
+ data->args.pgbase = req->wb_pgbase + offset;
+ data->args.pages = data->pages.pagevec;
+ data->args.count = count;
+ data->args.context = get_nfs_open_context(req->wb_context);
+ data->args.lock_context = req->wb_lock_context;
+ data->args.stable = NFS_UNSTABLE;
+ switch (how & (FLUSH_STABLE | FLUSH_COND_STABLE)) {
+ case 0:
+ break;
+ case FLUSH_COND_STABLE:
+ if (nfs_reqs_to_commit(cinfo))
+ break;
+ default:
+ data->args.stable = NFS_FILE_SYNC;
+ }
+
+ data->res.fattr = &data->fattr;
+ data->res.count = count;
+ data->res.eof = 0;
+ data->res.verf = &data->verf;
+ nfs_fattr_init(&data->fattr);
+ }
+
+ /**
+ * nfs_pgio_prepare - Prepare pageio data to go over the wire
+ * @task: The current task
+ * @calldata: pageio data to prepare
+ */
+ static void nfs_pgio_prepare(struct rpc_task *task, void *calldata)
+ {
+ struct nfs_pgio_data *data = calldata;
+ int err;
+ err = NFS_PROTO(data->header->inode)->pgio_rpc_prepare(task, data);
+ if (err)
+ rpc_exit(task, err);
+ }
+
+ int nfs_initiate_pgio(struct rpc_clnt *clnt, struct nfs_pgio_data *data,
+ const struct rpc_call_ops *call_ops, int how, int flags)
+ {
+ struct rpc_task *task;
+ struct rpc_message msg = {
+ .rpc_argp = &data->args,
+ .rpc_resp = &data->res,
+ .rpc_cred = data->header->cred,
+ };
+ struct rpc_task_setup task_setup_data = {
+ .rpc_client = clnt,
+ .task = &data->task,
+ .rpc_message = &msg,
+ .callback_ops = call_ops,
+ .callback_data = data,
+ .workqueue = nfsiod_workqueue,
+ .flags = RPC_TASK_ASYNC | flags,
+ };
+ int ret = 0;
+
+ data->header->rw_ops->rw_initiate(data, &msg, &task_setup_data, how);
+
+ dprintk("NFS: %5u initiated pgio call "
+ "(req %s/%llu, %u bytes @ offset %llu)\n",
+ data->task.tk_pid,
+ data->header->inode->i_sb->s_id,
+ (unsigned long long)NFS_FILEID(data->header->inode),
+ data->args.count,
+ (unsigned long long)data->args.offset);
+
+ task = rpc_run_task(&task_setup_data);
+ if (IS_ERR(task)) {
+ ret = PTR_ERR(task);
+ goto out;
+ }
+ if (how & FLUSH_SYNC) {
+ ret = rpc_wait_for_completion_task(task);
+ if (ret == 0)
+ ret = task->tk_status;
+ }
+ rpc_put_task(task);
+ out:
+ return ret;
+ }
+ EXPORT_SYMBOL_GPL(nfs_initiate_pgio);
+
+ /**
+ * nfs_pgio_error - Clean up from a pageio error
+ * @desc: IO descriptor
+ * @hdr: pageio header
+ */
+ static int nfs_pgio_error(struct nfs_pageio_descriptor *desc,
+ struct nfs_pgio_header *hdr)
+ {
+ set_bit(NFS_IOHDR_REDO, &hdr->flags);
+ nfs_pgio_data_release(hdr->data);
+ hdr->data = NULL;
+ desc->pg_completion_ops->error_cleanup(&desc->pg_list);
+ return -ENOMEM;
+ }
+
+ /**
+ * nfs_pgio_release - Release pageio data
+ * @calldata: The pageio data to release
+ */
+ static void nfs_pgio_release(void *calldata)
+ {
+ struct nfs_pgio_data *data = calldata;
+ if (data->header->rw_ops->rw_release)
+ data->header->rw_ops->rw_release(data);
+ nfs_pgio_data_release(data);
+ }
+
/**
* nfs_pageio_init - initialise a page io descriptor
* @desc: pointer to descriptor
struct inode *inode,
const struct nfs_pageio_ops *pg_ops,
const struct nfs_pgio_completion_ops *compl_ops,
+ const struct nfs_rw_ops *rw_ops,
size_t bsize,
int io_flags)
{
desc->pg_inode = inode;
desc->pg_ops = pg_ops;
desc->pg_completion_ops = compl_ops;
+ desc->pg_rw_ops = rw_ops;
desc->pg_ioflags = io_flags;
desc->pg_error = 0;
desc->pg_lseg = NULL;
}
EXPORT_SYMBOL_GPL(nfs_pageio_init);
+ /**
+ * nfs_pgio_result - Basic pageio error handling
+ * @task: The task that ran
+ * @calldata: Pageio data to check
+ */
+ static void nfs_pgio_result(struct rpc_task *task, void *calldata)
+ {
+ struct nfs_pgio_data *data = calldata;
+ struct inode *inode = data->header->inode;
+
+ dprintk("NFS: %s: %5u, (status %d)\n", __func__,
+ task->tk_pid, task->tk_status);
+
+ if (data->header->rw_ops->rw_done(task, data, inode) != 0)
+ return;
+ if (task->tk_status < 0)
+ nfs_set_pgio_error(data->header, task->tk_status, data->args.offset);
+ else
+ data->header->rw_ops->rw_result(task, data);
+ }
+
+ /*
+ * Create an RPC task for the given read or write request and kick it.
+ * The page must have been locked by the caller.
+ *
+ * It may happen that the page we're passed is not marked dirty.
+ * This is the case if nfs_updatepage detects a conflicting request
+ * that has been written but not committed.
+ */
+ int nfs_generic_pgio(struct nfs_pageio_descriptor *desc,
+ struct nfs_pgio_header *hdr)
+ {
+ struct nfs_page *req;
+ struct page **pages;
+ struct nfs_pgio_data *data;
+ struct list_head *head = &desc->pg_list;
+ struct nfs_commit_info cinfo;
+
+ data = nfs_pgio_data_alloc(hdr, nfs_page_array_len(desc->pg_base,
+ desc->pg_count));
+ if (!data)
+ return nfs_pgio_error(desc, hdr);
+
+ nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq);
+ pages = data->pages.pagevec;
+ while (!list_empty(head)) {
+ req = nfs_list_entry(head->next);
+ nfs_list_remove_request(req);
+ nfs_list_add_request(req, &hdr->pages);
+ *pages++ = req->wb_page;
+ }
+
+ if ((desc->pg_ioflags & FLUSH_COND_STABLE) &&
+ (desc->pg_moreio || nfs_reqs_to_commit(&cinfo)))
+ desc->pg_ioflags &= ~FLUSH_COND_STABLE;
+
+ /* Set up the argument struct */
+ nfs_pgio_rpcsetup(data, desc->pg_count, 0, desc->pg_ioflags, &cinfo);
+ hdr->data = data;
+ desc->pg_rpc_callops = &nfs_pgio_common_ops;
+ return 0;
+ }
+ EXPORT_SYMBOL_GPL(nfs_generic_pgio);
+
+ static int nfs_generic_pg_pgios(struct nfs_pageio_descriptor *desc)
+ {
+ struct nfs_rw_header *rw_hdr;
+ struct nfs_pgio_header *hdr;
+ int ret;
+
+ rw_hdr = nfs_rw_header_alloc(desc->pg_rw_ops);
+ if (!rw_hdr) {
+ desc->pg_completion_ops->error_cleanup(&desc->pg_list);
+ return -ENOMEM;
+ }
+ hdr = &rw_hdr->header;
+ nfs_pgheader_init(desc, hdr, nfs_rw_header_free);
+ atomic_inc(&hdr->refcnt);
+ ret = nfs_generic_pgio(desc, hdr);
+ if (ret == 0)
+ ret = nfs_initiate_pgio(NFS_CLIENT(hdr->inode),
+ hdr->data, desc->pg_rpc_callops,
+ desc->pg_ioflags, 0);
+ if (atomic_dec_and_test(&hdr->refcnt))
+ hdr->completion_ops->completion(hdr);
+ return ret;
+ }
+
static bool nfs_match_open_context(const struct nfs_open_context *ctx1,
const struct nfs_open_context *ctx2)
{
struct nfs_page *req,
struct nfs_pageio_descriptor *pgio)
{
- if (!nfs_match_open_context(req->wb_context, prev->wb_context))
- return false;
- if (req->wb_context->dentry->d_inode->i_flock != NULL &&
- !nfs_match_lock_context(req->wb_lock_context, prev->wb_lock_context))
- return false;
- if (req->wb_pgbase != 0)
- return false;
- if (prev->wb_pgbase + prev->wb_bytes != PAGE_CACHE_SIZE)
- return false;
- if (req_offset(req) != req_offset(prev) + prev->wb_bytes)
- return false;
- return pgio->pg_ops->pg_test(pgio, prev, req);
+ size_t size;
+
+ if (prev) {
+ if (!nfs_match_open_context(req->wb_context, prev->wb_context))
+ return false;
+ if (req->wb_context->dentry->d_inode->i_flock != NULL &&
+ !nfs_match_lock_context(req->wb_lock_context,
+ prev->wb_lock_context))
+ return false;
+ if (req_offset(req) != req_offset(prev) + prev->wb_bytes)
+ return false;
+ }
+ size = pgio->pg_ops->pg_test(pgio, prev, req);
+ WARN_ON_ONCE(size > req->wb_bytes);
+ if (size && size < req->wb_bytes)
+ req->wb_bytes = size;
+ return size > 0;
}
/**
static int nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc,
struct nfs_page *req)
{
+ struct nfs_page *prev = NULL;
if (desc->pg_count != 0) {
- struct nfs_page *prev;
-
prev = nfs_list_entry(desc->pg_list.prev);
- if (!nfs_can_coalesce_requests(prev, req, desc))
- return 0;
} else {
if (desc->pg_ops->pg_init)
desc->pg_ops->pg_init(desc, req);
desc->pg_base = req->wb_pgbase;
}
+ if (!nfs_can_coalesce_requests(prev, req, desc))
+ return 0;
nfs_list_remove_request(req);
nfs_list_add_request(req, &desc->pg_list);
desc->pg_count += req->wb_bytes;
* @desc: destination io descriptor
* @req: request
*
+ * This may split a request into subrequests which are all part of the
+ * same page group.
+ *
* Returns true if the request 'req' was successfully coalesced into the
* existing list of pages 'desc'.
*/
static int __nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
struct nfs_page *req)
{
- while (!nfs_pageio_do_add_request(desc, req)) {
- desc->pg_moreio = 1;
- nfs_pageio_doio(desc);
- if (desc->pg_error < 0)
- return 0;
- desc->pg_moreio = 0;
- if (desc->pg_recoalesce)
- return 0;
- }
+ struct nfs_page *subreq;
+ unsigned int bytes_left = 0;
+ unsigned int offset, pgbase;
+
+ nfs_page_group_lock(req);
+
+ subreq = req;
+ bytes_left = subreq->wb_bytes;
+ offset = subreq->wb_offset;
+ pgbase = subreq->wb_pgbase;
+
+ do {
+ if (!nfs_pageio_do_add_request(desc, subreq)) {
+ /* make sure pg_test call(s) did nothing */
+ WARN_ON_ONCE(subreq->wb_bytes != bytes_left);
+ WARN_ON_ONCE(subreq->wb_offset != offset);
+ WARN_ON_ONCE(subreq->wb_pgbase != pgbase);
+
+ nfs_page_group_unlock(req);
+ desc->pg_moreio = 1;
+ nfs_pageio_doio(desc);
+ if (desc->pg_error < 0)
+ return 0;
+ desc->pg_moreio = 0;
+ if (desc->pg_recoalesce)
+ return 0;
+ /* retry add_request for this subreq */
+ nfs_page_group_lock(req);
+ continue;
+ }
+
+ /* check for buggy pg_test call(s) */
+ WARN_ON_ONCE(subreq->wb_bytes + subreq->wb_pgbase > PAGE_SIZE);
+ WARN_ON_ONCE(subreq->wb_bytes > bytes_left);
+ WARN_ON_ONCE(subreq->wb_bytes == 0);
+
+ bytes_left -= subreq->wb_bytes;
+ offset += subreq->wb_bytes;
+ pgbase += subreq->wb_bytes;
+
+ if (bytes_left) {
+ subreq = nfs_create_request(req->wb_context,
+ req->wb_page,
+ subreq, pgbase, bytes_left);
+ if (IS_ERR(subreq))
+ goto err_ptr;
+ nfs_lock_request(subreq);
+ subreq->wb_offset = offset;
+ subreq->wb_index = req->wb_index;
+ }
+ } while (bytes_left > 0);
+
+ nfs_page_group_unlock(req);
return 1;
+ err_ptr:
+ desc->pg_error = PTR_ERR(subreq);
+ nfs_page_group_unlock(req);
+ return 0;
}
static int nfs_do_recoalesce(struct nfs_pageio_descriptor *desc)
kmem_cache_destroy(nfs_page_cachep);
}
+ static const struct rpc_call_ops nfs_pgio_common_ops = {
+ .rpc_call_prepare = nfs_pgio_prepare,
+ .rpc_call_done = nfs_pgio_result,
+ .rpc_release = nfs_pgio_release,
+ };
+
+ const struct nfs_pageio_ops nfs_pgio_rw_ops = {
+ .pg_test = nfs_generic_pg_test,
+ .pg_doio = nfs_generic_pg_pgios,
+ };
WARN_ON_ONCE(pgio->pg_lseg != NULL);
- if (req->wb_offset != req->wb_pgbase) {
- nfs_pageio_reset_read_mds(pgio);
- return;
- }
-
if (pgio->pg_dreq == NULL)
rd_size = i_size_read(pgio->pg_inode) - req_offset(req);
else
{
WARN_ON_ONCE(pgio->pg_lseg != NULL);
- if (req->wb_offset != req->wb_pgbase) {
- nfs_pageio_reset_write_mds(pgio);
- return;
- }
-
pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
req->wb_context,
req_offset(req),
}
EXPORT_SYMBOL_GPL(pnfs_generic_pg_init_write);
- void
- pnfs_pageio_init_read(struct nfs_pageio_descriptor *pgio, struct inode *inode,
- const struct nfs_pgio_completion_ops *compl_ops)
- {
- struct nfs_server *server = NFS_SERVER(inode);
- struct pnfs_layoutdriver_type *ld = server->pnfs_curr_ld;
-
- if (ld == NULL)
- nfs_pageio_init_read(pgio, inode, compl_ops);
- else
- nfs_pageio_init(pgio, inode, ld->pg_read_ops, compl_ops, server->rsize, 0);
- }
-
- void
- pnfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, struct inode *inode,
- int ioflags,
- const struct nfs_pgio_completion_ops *compl_ops)
- {
- struct nfs_server *server = NFS_SERVER(inode);
- struct pnfs_layoutdriver_type *ld = server->pnfs_curr_ld;
-
- if (ld == NULL)
- nfs_pageio_init_write(pgio, inode, ioflags, compl_ops);
- else
- nfs_pageio_init(pgio, inode, ld->pg_write_ops, compl_ops, server->wsize, ioflags);
- }
-
- bool
+ /*
+ * Return 0 if @req cannot be coalesced into @pgio, otherwise return the number
+ * of bytes (maximum @req->wb_bytes) that can be coalesced.
+ */
+ size_t
pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
struct nfs_page *req)
{
- if (pgio->pg_lseg == NULL)
- return nfs_generic_pg_test(pgio, prev, req);
+ unsigned int size;
+ u64 seg_end, req_start, seg_left;
+
+ size = nfs_generic_pg_test(pgio, prev, req);
+ if (!size)
+ return 0;
/*
- * Test if a nfs_page is fully contained in the pnfs_layout_range.
- * Note that this test makes several assumptions:
- * - that the previous nfs_page in the struct nfs_pageio_descriptor
- * is known to lie within the range.
- * - that the nfs_page being tested is known to be contiguous with the
- * previous nfs_page.
- * - Layout ranges are page aligned, so we only have to test the
- * start offset of the request.
+ * 'size' contains the number of bytes left in the current page (up
+ * to the original size asked for in @req->wb_bytes).
+ *
+ * Calculate how many bytes are left in the layout segment
+ * and if there are less bytes than 'size', return that instead.
*
* Please also note that 'end_offset' is actually the offset of the
* first byte that lies outside the pnfs_layout_range. FIXME?
*
*/
- return req_offset(req) < end_offset(pgio->pg_lseg->pls_range.offset,
- pgio->pg_lseg->pls_range.length);
+ if (pgio->pg_lseg) {
+ seg_end = end_offset(pgio->pg_lseg->pls_range.offset,
+ pgio->pg_lseg->pls_range.length);
+ req_start = req_offset(req);
+ WARN_ON_ONCE(req_start > seg_end);
+ /* start of request is past the last byte of this segment */
+ if (req_start >= seg_end)
+ return 0;
+
+ /* adjust 'size' iff there are fewer bytes left in the
+ * segment than what nfs_generic_pg_test returned */
+ seg_left = seg_end - req_start;
+ if (seg_left < size)
+ size = (unsigned int)seg_left;
+ }
+
+ return size;
}
EXPORT_SYMBOL_GPL(pnfs_generic_pg_test);
LIST_HEAD(failed);
/* Resend all requests through the MDS */
- nfs_pageio_init_write(&pgio, inode, FLUSH_STABLE, compl_ops);
+ nfs_pageio_init_write(&pgio, inode, FLUSH_STABLE, true, compl_ops);
pgio.pg_dreq = dreq;
while (!list_empty(head)) {
struct nfs_page *req = nfs_list_entry(head->next);
}
EXPORT_SYMBOL_GPL(pnfs_write_done_resend_to_mds);
- static void pnfs_ld_handle_write_error(struct nfs_write_data *data)
+ static void pnfs_ld_handle_write_error(struct nfs_pgio_data *data)
{
struct nfs_pgio_header *hdr = data->header;
/*
* Called by non rpc-based layout drivers
*/
- void pnfs_ld_write_done(struct nfs_write_data *data)
+ void pnfs_ld_write_done(struct nfs_pgio_data *data)
{
struct nfs_pgio_header *hdr = data->header;
static void
pnfs_write_through_mds(struct nfs_pageio_descriptor *desc,
- struct nfs_write_data *data)
+ struct nfs_pgio_data *data)
{
struct nfs_pgio_header *hdr = data->header;
nfs_pageio_reset_write_mds(desc);
desc->pg_recoalesce = 1;
}
- nfs_writedata_release(data);
+ nfs_pgio_data_release(data);
}
static enum pnfs_try_status
- pnfs_try_to_write_data(struct nfs_write_data *wdata,
+ pnfs_try_to_write_data(struct nfs_pgio_data *wdata,
const struct rpc_call_ops *call_ops,
struct pnfs_layout_segment *lseg,
int how)
}
static void
- pnfs_do_multiple_writes(struct nfs_pageio_descriptor *desc, struct list_head *head, int how)
+ pnfs_do_write(struct nfs_pageio_descriptor *desc,
+ struct nfs_pgio_header *hdr, int how)
{
- struct nfs_write_data *data;
+ struct nfs_pgio_data *data = hdr->data;
const struct rpc_call_ops *call_ops = desc->pg_rpc_callops;
struct pnfs_layout_segment *lseg = desc->pg_lseg;
+ enum pnfs_try_status trypnfs;
desc->pg_lseg = NULL;
- while (!list_empty(head)) {
- enum pnfs_try_status trypnfs;
-
- data = list_first_entry(head, struct nfs_write_data, list);
- list_del_init(&data->list);
-
- trypnfs = pnfs_try_to_write_data(data, call_ops, lseg, how);
- if (trypnfs == PNFS_NOT_ATTEMPTED)
- pnfs_write_through_mds(desc, data);
- }
+ trypnfs = pnfs_try_to_write_data(data, call_ops, lseg, how);
+ if (trypnfs == PNFS_NOT_ATTEMPTED)
+ pnfs_write_through_mds(desc, data);
pnfs_put_lseg(lseg);
}
static void pnfs_writehdr_free(struct nfs_pgio_header *hdr)
{
pnfs_put_lseg(hdr->lseg);
- nfs_writehdr_free(hdr);
+ nfs_rw_header_free(hdr);
}
EXPORT_SYMBOL_GPL(pnfs_writehdr_free);
int
pnfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc)
{
- struct nfs_write_header *whdr;
+ struct nfs_rw_header *whdr;
struct nfs_pgio_header *hdr;
int ret;
- whdr = nfs_writehdr_alloc();
+ whdr = nfs_rw_header_alloc(desc->pg_rw_ops);
if (!whdr) {
desc->pg_completion_ops->error_cleanup(&desc->pg_list);
pnfs_put_lseg(desc->pg_lseg);
nfs_pgheader_init(desc, hdr, pnfs_writehdr_free);
hdr->lseg = pnfs_get_lseg(desc->pg_lseg);
atomic_inc(&hdr->refcnt);
- ret = nfs_generic_flush(desc, hdr);
+ ret = nfs_generic_pgio(desc, hdr);
if (ret != 0) {
pnfs_put_lseg(desc->pg_lseg);
desc->pg_lseg = NULL;
} else
- pnfs_do_multiple_writes(desc, &hdr->rpc_list, desc->pg_ioflags);
+ pnfs_do_write(desc, hdr, desc->pg_ioflags);
if (atomic_dec_and_test(&hdr->refcnt))
hdr->completion_ops->completion(hdr);
return ret;
LIST_HEAD(failed);
/* Resend all requests through the MDS */
- nfs_pageio_init_read(&pgio, inode, compl_ops);
+ nfs_pageio_init_read(&pgio, inode, true, compl_ops);
pgio.pg_dreq = dreq;
while (!list_empty(head)) {
struct nfs_page *req = nfs_list_entry(head->next);
}
EXPORT_SYMBOL_GPL(pnfs_read_done_resend_to_mds);
- static void pnfs_ld_handle_read_error(struct nfs_read_data *data)
+ static void pnfs_ld_handle_read_error(struct nfs_pgio_data *data)
{
struct nfs_pgio_header *hdr = data->header;
/*
* Called by non rpc-based layout drivers
*/
- void pnfs_ld_read_done(struct nfs_read_data *data)
+ void pnfs_ld_read_done(struct nfs_pgio_data *data)
{
struct nfs_pgio_header *hdr = data->header;
static void
pnfs_read_through_mds(struct nfs_pageio_descriptor *desc,
- struct nfs_read_data *data)
+ struct nfs_pgio_data *data)
{
struct nfs_pgio_header *hdr = data->header;
nfs_pageio_reset_read_mds(desc);
desc->pg_recoalesce = 1;
}
- nfs_readdata_release(data);
+ nfs_pgio_data_release(data);
}
/*
* Call the appropriate parallel I/O subsystem read function.
*/
static enum pnfs_try_status
- pnfs_try_to_read_data(struct nfs_read_data *rdata,
+ pnfs_try_to_read_data(struct nfs_pgio_data *rdata,
const struct rpc_call_ops *call_ops,
struct pnfs_layout_segment *lseg)
{
}
static void
- pnfs_do_multiple_reads(struct nfs_pageio_descriptor *desc, struct list_head *head)
+ pnfs_do_read(struct nfs_pageio_descriptor *desc, struct nfs_pgio_header *hdr)
{
- struct nfs_read_data *data;
+ struct nfs_pgio_data *data = hdr->data;
const struct rpc_call_ops *call_ops = desc->pg_rpc_callops;
struct pnfs_layout_segment *lseg = desc->pg_lseg;
+ enum pnfs_try_status trypnfs;
desc->pg_lseg = NULL;
- while (!list_empty(head)) {
- enum pnfs_try_status trypnfs;
-
- data = list_first_entry(head, struct nfs_read_data, list);
- list_del_init(&data->list);
-
- trypnfs = pnfs_try_to_read_data(data, call_ops, lseg);
- if (trypnfs == PNFS_NOT_ATTEMPTED)
- pnfs_read_through_mds(desc, data);
- }
+ trypnfs = pnfs_try_to_read_data(data, call_ops, lseg);
+ if (trypnfs == PNFS_NOT_ATTEMPTED)
+ pnfs_read_through_mds(desc, data);
pnfs_put_lseg(lseg);
}
static void pnfs_readhdr_free(struct nfs_pgio_header *hdr)
{
pnfs_put_lseg(hdr->lseg);
- nfs_readhdr_free(hdr);
+ nfs_rw_header_free(hdr);
}
EXPORT_SYMBOL_GPL(pnfs_readhdr_free);
int
pnfs_generic_pg_readpages(struct nfs_pageio_descriptor *desc)
{
- struct nfs_read_header *rhdr;
+ struct nfs_rw_header *rhdr;
struct nfs_pgio_header *hdr;
int ret;
- rhdr = nfs_readhdr_alloc();
+ rhdr = nfs_rw_header_alloc(desc->pg_rw_ops);
if (!rhdr) {
desc->pg_completion_ops->error_cleanup(&desc->pg_list);
ret = -ENOMEM;
nfs_pgheader_init(desc, hdr, pnfs_readhdr_free);
hdr->lseg = pnfs_get_lseg(desc->pg_lseg);
atomic_inc(&hdr->refcnt);
- ret = nfs_generic_pagein(desc, hdr);
+ ret = nfs_generic_pgio(desc, hdr);
if (ret != 0) {
pnfs_put_lseg(desc->pg_lseg);
desc->pg_lseg = NULL;
} else
- pnfs_do_multiple_reads(desc, &hdr->rpc_list);
+ pnfs_do_read(desc, hdr);
if (atomic_dec_and_test(&hdr->refcnt))
hdr->completion_ops->completion(hdr);
return ret;
unsigned long *bitlock = &NFS_I(inode)->flags;
clear_bit_unlock(NFS_INO_LAYOUTCOMMITTING, bitlock);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
wake_up_bit(bitlock, NFS_INO_LAYOUTCOMMITTING);
}
EXPORT_SYMBOL_GPL(pnfs_set_lo_fail);
void
- pnfs_set_layoutcommit(struct nfs_write_data *wdata)
+ pnfs_set_layoutcommit(struct nfs_pgio_data *wdata)
{
struct nfs_pgio_header *hdr = wdata->header;
struct inode *inode = hdr->inode;
* Return PNFS_ATTEMPTED to indicate the layout code has attempted
* I/O, else return PNFS_NOT_ATTEMPTED to fall back to normal NFS
*/
- enum pnfs_try_status (*read_pagelist) (struct nfs_read_data *nfs_data);
- enum pnfs_try_status (*write_pagelist) (struct nfs_write_data *nfs_data, int how);
+ enum pnfs_try_status (*read_pagelist) (struct nfs_pgio_data *nfs_data);
+ enum pnfs_try_status (*write_pagelist) (struct nfs_pgio_data *nfs_data, int how);
void (*free_deviceid_node) (struct nfs4_deviceid_node *);
void pnfs_get_layout_hdr(struct pnfs_layout_hdr *lo);
void pnfs_put_lseg(struct pnfs_layout_segment *lseg);
- void pnfs_pageio_init_read(struct nfs_pageio_descriptor *, struct inode *,
- const struct nfs_pgio_completion_ops *);
- void pnfs_pageio_init_write(struct nfs_pageio_descriptor *, struct inode *,
- int, const struct nfs_pgio_completion_ops *);
-
void set_pnfs_layoutdriver(struct nfs_server *, const struct nfs_fh *, u32);
void unset_pnfs_layoutdriver(struct nfs_server *);
void pnfs_generic_pg_init_read(struct nfs_pageio_descriptor *, struct nfs_page *);
void pnfs_generic_pg_init_write(struct nfs_pageio_descriptor *pgio,
struct nfs_page *req, u64 wb_size);
int pnfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc);
- bool pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev, struct nfs_page *req);
+ size_t pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio,
+ struct nfs_page *prev, struct nfs_page *req);
void pnfs_set_lo_fail(struct pnfs_layout_segment *lseg);
struct pnfs_layout_segment *pnfs_layout_process(struct nfs4_layoutget *lgp);
void pnfs_free_lseg_list(struct list_head *tmp_list);
void pnfs_roc_release(struct inode *ino);
void pnfs_roc_set_barrier(struct inode *ino, u32 barrier);
bool pnfs_roc_drain(struct inode *ino, u32 *barrier, struct rpc_task *task);
- void pnfs_set_layoutcommit(struct nfs_write_data *wdata);
+ void pnfs_set_layoutcommit(struct nfs_pgio_data *wdata);
void pnfs_cleanup_layoutcommit(struct nfs4_layoutcommit_data *data);
int pnfs_layoutcommit_inode(struct inode *inode, bool sync);
int _pnfs_return_layout(struct inode *);
int pnfs_commit_and_return_layout(struct inode *);
- void pnfs_ld_write_done(struct nfs_write_data *);
- void pnfs_ld_read_done(struct nfs_read_data *);
+ void pnfs_ld_write_done(struct nfs_pgio_data *);
+ void pnfs_ld_read_done(struct nfs_pgio_data *);
struct pnfs_layout_segment *pnfs_update_layout(struct inode *ino,
struct nfs_open_context *ctx,
loff_t pos,
{
if (lseg) {
atomic_inc(&lseg->pls_refcount);
- smp_mb__after_atomic_inc();
+ smp_mb__after_atomic();
}
return lseg;
}
{
}
- static inline void pnfs_pageio_init_read(struct nfs_pageio_descriptor *pgio, struct inode *inode,
- const struct nfs_pgio_completion_ops *compl_ops)
- {
- nfs_pageio_init_read(pgio, inode, compl_ops);
- }
-
- static inline void pnfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, struct inode *inode, int ioflags,
- const struct nfs_pgio_completion_ops *compl_ops)
- {
- nfs_pageio_init_write(pgio, inode, ioflags, compl_ops);
- }
-
static inline int
pnfs_commit_list(struct inode *inode, struct list_head *mds_pages, int how,
struct nfs_commit_info *cinfo)
* Local function declarations
*/
static void nfs_redirty_request(struct nfs_page *req);
- static const struct rpc_call_ops nfs_write_common_ops;
static const struct rpc_call_ops nfs_commit_ops;
static const struct nfs_pgio_completion_ops nfs_async_write_completion_ops;
static const struct nfs_commit_completion_ops nfs_commit_completion_ops;
+ static const struct nfs_rw_ops nfs_rw_write_ops;
static struct kmem_cache *nfs_wdata_cachep;
static mempool_t *nfs_wdata_mempool;
}
EXPORT_SYMBOL_GPL(nfs_commit_free);
- struct nfs_write_header *nfs_writehdr_alloc(void)
+ static struct nfs_rw_header *nfs_writehdr_alloc(void)
{
- struct nfs_write_header *p = mempool_alloc(nfs_wdata_mempool, GFP_NOIO);
-
- if (p) {
- struct nfs_pgio_header *hdr = &p->header;
+ struct nfs_rw_header *p = mempool_alloc(nfs_wdata_mempool, GFP_NOIO);
+ if (p)
memset(p, 0, sizeof(*p));
- INIT_LIST_HEAD(&hdr->pages);
- INIT_LIST_HEAD(&hdr->rpc_list);
- spin_lock_init(&hdr->lock);
- atomic_set(&hdr->refcnt, 0);
- hdr->verf = &p->verf;
- }
return p;
}
- EXPORT_SYMBOL_GPL(nfs_writehdr_alloc);
-
- static struct nfs_write_data *nfs_writedata_alloc(struct nfs_pgio_header *hdr,
- unsigned int pagecount)
- {
- struct nfs_write_data *data, *prealloc;
-
- prealloc = &container_of(hdr, struct nfs_write_header, header)->rpc_data;
- if (prealloc->header == NULL)
- data = prealloc;
- else
- data = kzalloc(sizeof(*data), GFP_KERNEL);
- if (!data)
- goto out;
-
- if (nfs_pgarray_set(&data->pages, pagecount)) {
- data->header = hdr;
- atomic_inc(&hdr->refcnt);
- } else {
- if (data != prealloc)
- kfree(data);
- data = NULL;
- }
- out:
- return data;
- }
- void nfs_writehdr_free(struct nfs_pgio_header *hdr)
+ static void nfs_writehdr_free(struct nfs_rw_header *whdr)
{
- struct nfs_write_header *whdr = container_of(hdr, struct nfs_write_header, header);
mempool_free(whdr, nfs_wdata_mempool);
}
- EXPORT_SYMBOL_GPL(nfs_writehdr_free);
-
- void nfs_writedata_release(struct nfs_write_data *wdata)
- {
- struct nfs_pgio_header *hdr = wdata->header;
- struct nfs_write_header *write_header = container_of(hdr, struct nfs_write_header, header);
-
- put_nfs_open_context(wdata->args.context);
- if (wdata->pages.pagevec != wdata->pages.page_array)
- kfree(wdata->pages.pagevec);
- if (wdata == &write_header->rpc_data) {
- wdata->header = NULL;
- wdata = NULL;
- }
- if (atomic_dec_and_test(&hdr->refcnt))
- hdr->completion_ops->completion(hdr);
- /* Note: we only free the rpc_task after callbacks are done.
- * See the comment in rpc_free_task() for why
- */
- kfree(wdata);
- }
- EXPORT_SYMBOL_GPL(nfs_writedata_release);
static void nfs_context_set_write_error(struct nfs_open_context *ctx, int error)
{
nfs_zap_mapping(page_file_mapping(page)->host, page_file_mapping(page));
}
+ /*
+ * nfs_page_group_search_locked
+ * @head - head request of page group
+ * @page_offset - offset into page
+ *
+ * Search page group with head @head to find a request that contains the
+ * page offset @page_offset.
+ *
+ * Returns a pointer to the first matching nfs request, or NULL if no
+ * match is found.
+ *
+ * Must be called with the page group lock held
+ */
+ static struct nfs_page *
+ nfs_page_group_search_locked(struct nfs_page *head, unsigned int page_offset)
+ {
+ struct nfs_page *req;
+
+ WARN_ON_ONCE(head != head->wb_head);
+ WARN_ON_ONCE(!test_bit(PG_HEADLOCK, &head->wb_head->wb_flags));
+
+ req = head;
+ do {
+ if (page_offset >= req->wb_pgbase &&
+ page_offset < (req->wb_pgbase + req->wb_bytes))
+ return req;
+
+ req = req->wb_this_page;
+ } while (req != head);
+
+ return NULL;
+ }
+
+ /*
+ * nfs_page_group_covers_page
+ * @head - head request of page group
+ *
+ * Return true if the page group with head @head covers the whole page,
+ * returns false otherwise
+ */
+ static bool nfs_page_group_covers_page(struct nfs_page *req)
+ {
+ struct nfs_page *tmp;
+ unsigned int pos = 0;
+ unsigned int len = nfs_page_length(req->wb_page);
+
+ nfs_page_group_lock(req);
+
+ do {
+ tmp = nfs_page_group_search_locked(req->wb_head, pos);
+ if (tmp) {
+ /* no way this should happen */
+ WARN_ON_ONCE(tmp->wb_pgbase != pos);
+ pos += tmp->wb_bytes - (pos - tmp->wb_pgbase);
+ }
+ } while (tmp && pos < len);
+
+ nfs_page_group_unlock(req);
+ WARN_ON_ONCE(pos > len);
+ return pos == len;
+ }
+
/* We can set the PG_uptodate flag if we see that a write request
* covers the full page.
*/
- static void nfs_mark_uptodate(struct page *page, unsigned int base, unsigned int count)
+ static void nfs_mark_uptodate(struct nfs_page *req)
{
- if (PageUptodate(page))
- return;
- if (base != 0)
+ if (PageUptodate(req->wb_page))
return;
- if (count != nfs_page_length(page))
+ if (!nfs_page_group_covers_page(req))
return;
- SetPageUptodate(page);
+ SetPageUptodate(req->wb_page);
}
static int wb_priority(struct writeback_control *wbc)
}
}
- static void nfs_end_page_writeback(struct page *page)
+ static void nfs_end_page_writeback(struct nfs_page *req)
{
- struct inode *inode = page_file_mapping(page)->host;
+ struct inode *inode = page_file_mapping(req->wb_page)->host;
struct nfs_server *nfss = NFS_SERVER(inode);
- end_page_writeback(page);
+ if (!nfs_page_group_sync_on_bit(req, PG_WB_END))
+ return;
+
+ end_page_writeback(req->wb_page);
if (atomic_long_dec_return(&nfss->writeback) < NFS_CONGESTION_OFF_THRESH)
clear_bdi_congested(&nfss->backing_dev_info, BLK_RW_ASYNC);
}
struct nfs_pageio_descriptor pgio;
int err;
- NFS_PROTO(page_file_mapping(page)->host)->write_pageio_init(&pgio,
- page->mapping->host,
- wb_priority(wbc),
- &nfs_async_write_completion_ops);
+ nfs_pageio_init_write(&pgio, page->mapping->host, wb_priority(wbc),
+ false, &nfs_async_write_completion_ops);
err = nfs_do_writepage(page, wbc, &pgio);
nfs_pageio_complete(&pgio);
if (err < 0)
nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGES);
- NFS_PROTO(inode)->write_pageio_init(&pgio, inode, wb_priority(wbc), &nfs_async_write_completion_ops);
+ nfs_pageio_init_write(&pgio, inode, wb_priority(wbc), false,
+ &nfs_async_write_completion_ops);
err = write_cache_pages(mapping, wbc, nfs_writepages_callback, &pgio);
nfs_pageio_complete(&pgio);
clear_bit_unlock(NFS_INO_FLUSHING, bitlock);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
wake_up_bit(bitlock, NFS_INO_FLUSHING);
if (err < 0)
{
struct nfs_inode *nfsi = NFS_I(inode);
+ WARN_ON_ONCE(req->wb_this_page != req);
+
/* Lock the request! */
nfs_lock_request(req);
set_page_private(req->wb_page, (unsigned long)req);
}
nfsi->npages++;
+ set_bit(PG_INODE_REF, &req->wb_flags);
kref_get(&req->wb_kref);
spin_unlock(&inode->i_lock);
}
{
struct inode *inode = req->wb_context->dentry->d_inode;
struct nfs_inode *nfsi = NFS_I(inode);
+ struct nfs_page *head;
- spin_lock(&inode->i_lock);
- if (likely(!PageSwapCache(req->wb_page))) {
- set_page_private(req->wb_page, 0);
- ClearPagePrivate(req->wb_page);
- clear_bit(PG_MAPPED, &req->wb_flags);
+ if (nfs_page_group_sync_on_bit(req, PG_REMOVE)) {
+ head = req->wb_head;
+
+ spin_lock(&inode->i_lock);
+ if (likely(!PageSwapCache(head->wb_page))) {
+ set_page_private(head->wb_page, 0);
+ ClearPagePrivate(head->wb_page);
+ clear_bit(PG_MAPPED, &head->wb_flags);
+ }
+ nfsi->npages--;
+ spin_unlock(&inode->i_lock);
}
- nfsi->npages--;
- spin_unlock(&inode->i_lock);
nfs_release_request(req);
}
}
static inline
- int nfs_write_need_commit(struct nfs_write_data *data)
+ int nfs_write_need_commit(struct nfs_pgio_data *data)
{
if (data->verf.committed == NFS_DATA_SYNC)
return data->header->lseg == NULL;
}
static inline
- int nfs_write_need_commit(struct nfs_write_data *data)
+ int nfs_write_need_commit(struct nfs_pgio_data *data)
{
return 0;
}
{
struct nfs_commit_info cinfo;
unsigned long bytes = 0;
+ bool do_destroy;
if (test_bit(NFS_IOHDR_REDO, &hdr->flags))
goto out;
goto next;
}
if (test_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags)) {
- memcpy(&req->wb_verf, &hdr->verf->verifier, sizeof(req->wb_verf));
+ memcpy(&req->wb_verf, &hdr->verf.verifier, sizeof(req->wb_verf));
nfs_mark_request_commit(req, hdr->lseg, &cinfo);
goto next;
}
nfs_inode_remove_request(req);
next:
nfs_unlock_request(req);
- nfs_end_page_writeback(req->wb_page);
+ nfs_end_page_writeback(req);
+ do_destroy = !test_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags);
nfs_release_request(req);
}
out:
}
#if IS_ENABLED(CONFIG_NFS_V3) || IS_ENABLED(CONFIG_NFS_V4)
- static unsigned long
+ unsigned long
nfs_reqs_to_commit(struct nfs_commit_info *cinfo)
{
return cinfo->mds->ncommit;
}
#else
- static unsigned long nfs_reqs_to_commit(struct nfs_commit_info *cinfo)
+ unsigned long nfs_reqs_to_commit(struct nfs_commit_info *cinfo)
{
return 0;
}
if (req == NULL)
goto out_unlock;
+ /* should be handled by nfs_flush_incompatible */
+ WARN_ON_ONCE(req->wb_head != req);
+ WARN_ON_ONCE(req->wb_this_page != req);
+
rqend = req->wb_offset + req->wb_bytes;
/*
* Tell the caller to flush out the request if
req = nfs_try_to_update_request(inode, page, offset, bytes);
if (req != NULL)
goto out;
- req = nfs_create_request(ctx, inode, page, offset, bytes);
+ req = nfs_create_request(ctx, page, NULL, offset, bytes);
if (IS_ERR(req))
goto out;
nfs_inode_add_request(inode, req);
return PTR_ERR(req);
/* Update file length */
nfs_grow_file(page, offset, count);
- nfs_mark_uptodate(page, req->wb_pgbase, req->wb_bytes);
+ nfs_mark_uptodate(req);
nfs_mark_request_dirty(req);
nfs_unlock_and_release_request(req);
return 0;
return 0;
l_ctx = req->wb_lock_context;
do_flush = req->wb_page != page || req->wb_context != ctx;
+ /* for now, flush if more than 1 request in page_group */
+ do_flush |= req->wb_this_page != req;
if (l_ctx && ctx->dentry->d_inode->i_flock != NULL) {
do_flush |= l_ctx->lockowner.l_owner != current->files
|| l_ctx->lockowner.l_pid != current->tgid;
return RPC_PRIORITY_NORMAL;
}
- int nfs_initiate_write(struct rpc_clnt *clnt,
- struct nfs_write_data *data,
- const struct rpc_call_ops *call_ops,
- int how, int flags)
+ static void nfs_initiate_write(struct nfs_pgio_data *data, struct rpc_message *msg,
+ struct rpc_task_setup *task_setup_data, int how)
{
struct inode *inode = data->header->inode;
int priority = flush_task_priority(how);
- struct rpc_task *task;
- struct rpc_message msg = {
- .rpc_argp = &data->args,
- .rpc_resp = &data->res,
- .rpc_cred = data->header->cred,
- };
- struct rpc_task_setup task_setup_data = {
- .rpc_client = clnt,
- .task = &data->task,
- .rpc_message = &msg,
- .callback_ops = call_ops,
- .callback_data = data,
- .workqueue = nfsiod_workqueue,
- .flags = RPC_TASK_ASYNC | flags,
- .priority = priority,
- };
- int ret = 0;
-
- /* Set up the initial task struct. */
- NFS_PROTO(inode)->write_setup(data, &msg);
- dprintk("NFS: %5u initiated write call "
- "(req %s/%llu, %u bytes @ offset %llu)\n",
- data->task.tk_pid,
- inode->i_sb->s_id,
- (unsigned long long)NFS_FILEID(inode),
- data->args.count,
- (unsigned long long)data->args.offset);
+ task_setup_data->priority = priority;
+ NFS_PROTO(inode)->write_setup(data, msg);
nfs4_state_protect_write(NFS_SERVER(inode)->nfs_client,
- &task_setup_data.rpc_client, &msg, data);
-
- task = rpc_run_task(&task_setup_data);
- if (IS_ERR(task)) {
- ret = PTR_ERR(task);
- goto out;
- }
- if (how & FLUSH_SYNC) {
- ret = rpc_wait_for_completion_task(task);
- if (ret == 0)
- ret = task->tk_status;
- }
- rpc_put_task(task);
- out:
- return ret;
- }
- EXPORT_SYMBOL_GPL(nfs_initiate_write);
-
- /*
- * Set up the argument/result storage required for the RPC call.
- */
- static void nfs_write_rpcsetup(struct nfs_write_data *data,
- unsigned int count, unsigned int offset,
- int how, struct nfs_commit_info *cinfo)
- {
- struct nfs_page *req = data->header->req;
-
- /* Set up the RPC argument and reply structs
- * NB: take care not to mess about with data->commit et al. */
-
- data->args.fh = NFS_FH(data->header->inode);
- data->args.offset = req_offset(req) + offset;
- /* pnfs_set_layoutcommit needs this */
- data->mds_offset = data->args.offset;
- data->args.pgbase = req->wb_pgbase + offset;
- data->args.pages = data->pages.pagevec;
- data->args.count = count;
- data->args.context = get_nfs_open_context(req->wb_context);
- data->args.lock_context = req->wb_lock_context;
- data->args.stable = NFS_UNSTABLE;
- switch (how & (FLUSH_STABLE | FLUSH_COND_STABLE)) {
- case 0:
- break;
- case FLUSH_COND_STABLE:
- if (nfs_reqs_to_commit(cinfo))
- break;
- default:
- data->args.stable = NFS_FILE_SYNC;
- }
-
- data->res.fattr = &data->fattr;
- data->res.count = count;
- data->res.verf = &data->verf;
- nfs_fattr_init(&data->fattr);
- }
-
- static int nfs_do_write(struct nfs_write_data *data,
- const struct rpc_call_ops *call_ops,
- int how)
- {
- struct inode *inode = data->header->inode;
-
- return nfs_initiate_write(NFS_CLIENT(inode), data, call_ops, how, 0);
- }
-
- static int nfs_do_multiple_writes(struct list_head *head,
- const struct rpc_call_ops *call_ops,
- int how)
- {
- struct nfs_write_data *data;
- int ret = 0;
-
- while (!list_empty(head)) {
- int ret2;
-
- data = list_first_entry(head, struct nfs_write_data, list);
- list_del_init(&data->list);
-
- ret2 = nfs_do_write(data, call_ops, how);
- if (ret == 0)
- ret = ret2;
- }
- return ret;
+ &task_setup_data->rpc_client, msg, data);
}
/* If a nfs_flush_* function fails, it should remove reqs from @head and
{
nfs_mark_request_dirty(req);
nfs_unlock_request(req);
- nfs_end_page_writeback(req->wb_page);
+ nfs_end_page_writeback(req);
nfs_release_request(req);
}
.completion = nfs_write_completion,
};
- static void nfs_flush_error(struct nfs_pageio_descriptor *desc,
- struct nfs_pgio_header *hdr)
- {
- set_bit(NFS_IOHDR_REDO, &hdr->flags);
- while (!list_empty(&hdr->rpc_list)) {
- struct nfs_write_data *data = list_first_entry(&hdr->rpc_list,
- struct nfs_write_data, list);
- list_del(&data->list);
- nfs_writedata_release(data);
- }
- desc->pg_completion_ops->error_cleanup(&desc->pg_list);
- }
-
- /*
- * Generate multiple small requests to write out a single
- * contiguous dirty area on one page.
- */
- static int nfs_flush_multi(struct nfs_pageio_descriptor *desc,
- struct nfs_pgio_header *hdr)
- {
- struct nfs_page *req = hdr->req;
- struct page *page = req->wb_page;
- struct nfs_write_data *data;
- size_t wsize = desc->pg_bsize, nbytes;
- unsigned int offset;
- int requests = 0;
- struct nfs_commit_info cinfo;
-
- nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq);
-
- if ((desc->pg_ioflags & FLUSH_COND_STABLE) &&
- (desc->pg_moreio || nfs_reqs_to_commit(&cinfo) ||
- desc->pg_count > wsize))
- desc->pg_ioflags &= ~FLUSH_COND_STABLE;
-
-
- offset = 0;
- nbytes = desc->pg_count;
- do {
- size_t len = min(nbytes, wsize);
-
- data = nfs_writedata_alloc(hdr, 1);
- if (!data) {
- nfs_flush_error(desc, hdr);
- return -ENOMEM;
- }
- data->pages.pagevec[0] = page;
- nfs_write_rpcsetup(data, len, offset, desc->pg_ioflags, &cinfo);
- list_add(&data->list, &hdr->rpc_list);
- requests++;
- nbytes -= len;
- offset += len;
- } while (nbytes != 0);
- nfs_list_remove_request(req);
- nfs_list_add_request(req, &hdr->pages);
- desc->pg_rpc_callops = &nfs_write_common_ops;
- return 0;
- }
-
- /*
- * Create an RPC task for the given write request and kick it.
- * The page must have been locked by the caller.
- *
- * It may happen that the page we're passed is not marked dirty.
- * This is the case if nfs_updatepage detects a conflicting request
- * that has been written but not committed.
- */
- static int nfs_flush_one(struct nfs_pageio_descriptor *desc,
- struct nfs_pgio_header *hdr)
- {
- struct nfs_page *req;
- struct page **pages;
- struct nfs_write_data *data;
- struct list_head *head = &desc->pg_list;
- struct nfs_commit_info cinfo;
-
- data = nfs_writedata_alloc(hdr, nfs_page_array_len(desc->pg_base,
- desc->pg_count));
- if (!data) {
- nfs_flush_error(desc, hdr);
- return -ENOMEM;
- }
-
- nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq);
- pages = data->pages.pagevec;
- while (!list_empty(head)) {
- req = nfs_list_entry(head->next);
- nfs_list_remove_request(req);
- nfs_list_add_request(req, &hdr->pages);
- *pages++ = req->wb_page;
- }
-
- if ((desc->pg_ioflags & FLUSH_COND_STABLE) &&
- (desc->pg_moreio || nfs_reqs_to_commit(&cinfo)))
- desc->pg_ioflags &= ~FLUSH_COND_STABLE;
-
- /* Set up the argument struct */
- nfs_write_rpcsetup(data, desc->pg_count, 0, desc->pg_ioflags, &cinfo);
- list_add(&data->list, &hdr->rpc_list);
- desc->pg_rpc_callops = &nfs_write_common_ops;
- return 0;
- }
-
- int nfs_generic_flush(struct nfs_pageio_descriptor *desc,
- struct nfs_pgio_header *hdr)
- {
- if (desc->pg_bsize < PAGE_CACHE_SIZE)
- return nfs_flush_multi(desc, hdr);
- return nfs_flush_one(desc, hdr);
- }
- EXPORT_SYMBOL_GPL(nfs_generic_flush);
-
- static int nfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc)
- {
- struct nfs_write_header *whdr;
- struct nfs_pgio_header *hdr;
- int ret;
-
- whdr = nfs_writehdr_alloc();
- if (!whdr) {
- desc->pg_completion_ops->error_cleanup(&desc->pg_list);
- return -ENOMEM;
- }
- hdr = &whdr->header;
- nfs_pgheader_init(desc, hdr, nfs_writehdr_free);
- atomic_inc(&hdr->refcnt);
- ret = nfs_generic_flush(desc, hdr);
- if (ret == 0)
- ret = nfs_do_multiple_writes(&hdr->rpc_list,
- desc->pg_rpc_callops,
- desc->pg_ioflags);
- if (atomic_dec_and_test(&hdr->refcnt))
- hdr->completion_ops->completion(hdr);
- return ret;
- }
-
- static const struct nfs_pageio_ops nfs_pageio_write_ops = {
- .pg_test = nfs_generic_pg_test,
- .pg_doio = nfs_generic_pg_writepages,
- };
-
void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio,
- struct inode *inode, int ioflags,
+ struct inode *inode, int ioflags, bool force_mds,
const struct nfs_pgio_completion_ops *compl_ops)
{
- nfs_pageio_init(pgio, inode, &nfs_pageio_write_ops, compl_ops,
- NFS_SERVER(inode)->wsize, ioflags);
+ struct nfs_server *server = NFS_SERVER(inode);
+ const struct nfs_pageio_ops *pg_ops = &nfs_pgio_rw_ops;
+
+ #ifdef CONFIG_NFS_V4_1
+ if (server->pnfs_curr_ld && !force_mds)
+ pg_ops = server->pnfs_curr_ld->pg_write_ops;
+ #endif
+ nfs_pageio_init(pgio, inode, pg_ops, compl_ops, &nfs_rw_write_ops,
+ server->wsize, ioflags);
}
EXPORT_SYMBOL_GPL(nfs_pageio_init_write);
void nfs_pageio_reset_write_mds(struct nfs_pageio_descriptor *pgio)
{
- pgio->pg_ops = &nfs_pageio_write_ops;
+ pgio->pg_ops = &nfs_pgio_rw_ops;
pgio->pg_bsize = NFS_SERVER(pgio->pg_inode)->wsize;
}
EXPORT_SYMBOL_GPL(nfs_pageio_reset_write_mds);
- void nfs_write_prepare(struct rpc_task *task, void *calldata)
- {
- struct nfs_write_data *data = calldata;
- int err;
- err = NFS_PROTO(data->header->inode)->write_rpc_prepare(task, data);
- if (err)
- rpc_exit(task, err);
- }
-
void nfs_commit_prepare(struct rpc_task *task, void *calldata)
{
struct nfs_commit_data *data = calldata;
NFS_PROTO(data->inode)->commit_rpc_prepare(task, data);
}
- /*
- * Handle a write reply that flushes a whole page.
- *
- * FIXME: There is an inherent race with invalidate_inode_pages and
- * writebacks since the page->count is kept > 1 for as long
- * as the page has a write request pending.
- */
- static void nfs_writeback_done_common(struct rpc_task *task, void *calldata)
- {
- struct nfs_write_data *data = calldata;
-
- nfs_writeback_done(task, data);
- }
-
- static void nfs_writeback_release_common(void *calldata)
+ static void nfs_writeback_release_common(struct nfs_pgio_data *data)
{
- struct nfs_write_data *data = calldata;
struct nfs_pgio_header *hdr = data->header;
int status = data->task.tk_status;
if (test_bit(NFS_IOHDR_NEED_RESCHED, &hdr->flags))
; /* Do nothing */
else if (!test_and_set_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags))
- memcpy(hdr->verf, &data->verf, sizeof(*hdr->verf));
- else if (memcmp(hdr->verf, &data->verf, sizeof(*hdr->verf)))
+ memcpy(&hdr->verf, &data->verf, sizeof(hdr->verf));
+ else if (memcmp(&hdr->verf, &data->verf, sizeof(hdr->verf)))
set_bit(NFS_IOHDR_NEED_RESCHED, &hdr->flags);
spin_unlock(&hdr->lock);
}
- nfs_writedata_release(data);
}
- static const struct rpc_call_ops nfs_write_common_ops = {
- .rpc_call_prepare = nfs_write_prepare,
- .rpc_call_done = nfs_writeback_done_common,
- .rpc_release = nfs_writeback_release_common,
- };
+ /*
+ * Special version of should_remove_suid() that ignores capabilities.
+ */
+ static int nfs_should_remove_suid(const struct inode *inode)
+ {
+ umode_t mode = inode->i_mode;
+ int kill = 0;
+
+ /* suid always must be killed */
+ if (unlikely(mode & S_ISUID))
+ kill = ATTR_KILL_SUID;
+ /*
+ * sgid without any exec bits is just a mandatory locking mark; leave
+ * it alone. If some exec bits are set, it's a real sgid; kill it.
+ */
+ if (unlikely((mode & S_ISGID) && (mode & S_IXGRP)))
+ kill |= ATTR_KILL_SGID;
+
+ if (unlikely(kill && S_ISREG(mode)))
+ return kill;
+
+ return 0;
+ }
/*
* This function is called when the WRITE call is complete.
*/
- void nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data)
+ static int nfs_writeback_done(struct rpc_task *task, struct nfs_pgio_data *data,
+ struct inode *inode)
{
- struct nfs_writeargs *argp = &data->args;
- struct nfs_writeres *resp = &data->res;
- struct inode *inode = data->header->inode;
int status;
- dprintk("NFS: %5u nfs_writeback_done (status %d)\n",
- task->tk_pid, task->tk_status);
-
/*
* ->write_done will attempt to use post-op attributes to detect
* conflicting writes by other clients. A strict interpretation
*/
status = NFS_PROTO(inode)->write_done(task, data);
if (status != 0)
- return;
- nfs_add_stats(inode, NFSIOS_SERVERWRITTENBYTES, resp->count);
+ return status;
+ nfs_add_stats(inode, NFSIOS_SERVERWRITTENBYTES, data->res.count);
#if IS_ENABLED(CONFIG_NFS_V3) || IS_ENABLED(CONFIG_NFS_V4)
- if (resp->verf->committed < argp->stable && task->tk_status >= 0) {
+ if (data->res.verf->committed < data->args.stable && task->tk_status >= 0) {
/* We tried a write call, but the server did not
* commit data to stable storage even though we
* requested it.
dprintk("NFS: faulty NFS server %s:"
" (committed = %d) != (stable = %d)\n",
NFS_SERVER(inode)->nfs_client->cl_hostname,
- resp->verf->committed, argp->stable);
+ data->res.verf->committed, data->args.stable);
complain = jiffies + 300 * HZ;
}
}
#endif
- if (task->tk_status < 0)
- nfs_set_pgio_error(data->header, task->tk_status, argp->offset);
- else if (resp->count < argp->count) {
+
+ /* Deal with the suid/sgid bit corner case */
+ if (nfs_should_remove_suid(inode))
+ nfs_mark_for_revalidate(inode);
+ return 0;
+ }
+
+ /*
+ * This function is called when the WRITE call is complete.
+ */
+ static void nfs_writeback_result(struct rpc_task *task, struct nfs_pgio_data *data)
+ {
+ struct nfs_pgio_args *argp = &data->args;
+ struct nfs_pgio_res *resp = &data->res;
+
+ if (resp->count < argp->count) {
static unsigned long complain;
/* This a short write! */
- nfs_inc_stats(inode, NFSIOS_SHORTWRITE);
+ nfs_inc_stats(data->header->inode, NFSIOS_SHORTWRITE);
/* Has the server at least made some progress? */
if (resp->count == 0) {
static void nfs_commit_clear_lock(struct nfs_inode *nfsi)
{
clear_bit(NFS_INO_COMMIT, &nfsi->flags);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
wake_up_bit(&nfsi->flags, NFS_INO_COMMIT);
}
int __init nfs_init_writepagecache(void)
{
nfs_wdata_cachep = kmem_cache_create("nfs_write_data",
- sizeof(struct nfs_write_header),
+ sizeof(struct nfs_rw_header),
0, SLAB_HWCACHE_ALIGN,
NULL);
if (nfs_wdata_cachep == NULL)
kmem_cache_destroy(nfs_wdata_cachep);
}
+ static const struct nfs_rw_ops nfs_rw_write_ops = {
+ .rw_mode = FMODE_WRITE,
+ .rw_alloc_header = nfs_writehdr_alloc,
+ .rw_free_header = nfs_writehdr_free,
+ .rw_release = nfs_writeback_release_common,
+ .rw_done = nfs_writeback_done,
+ .rw_result = nfs_writeback_result,
+ .rw_initiate = nfs_initiate_write,
+ };
#define RPC_MAX_SLOT_TABLE_LIMIT (65536U)
#define RPC_MAX_SLOT_TABLE RPC_MAX_SLOT_TABLE_LIMIT
+ #define RPC_CWNDSHIFT (8U)
+ #define RPC_CWNDSCALE (1U << RPC_CWNDSHIFT)
+ #define RPC_INITCWND RPC_CWNDSCALE
+ #define RPC_MAXCWND(xprt) ((xprt)->max_reqs << RPC_CWNDSHIFT)
+ #define RPCXPRT_CONGESTED(xprt) ((xprt)->cong >= (xprt)->cwnd)
+
/*
* This describes a timeout strategy
*/
static inline void xprt_clear_connecting(struct rpc_xprt *xprt)
{
- smp_mb__before_clear_bit();
+ smp_mb__before_atomic();
clear_bit(XPRT_CONNECTING, &xprt->state);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
}
static inline int xprt_connecting(struct rpc_xprt *xprt)
static inline void xprt_clear_binding(struct rpc_xprt *xprt)
{
- smp_mb__before_clear_bit();
+ smp_mb__before_atomic();
clear_bit(XPRT_BINDING, &xprt->state);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
}
static inline int xprt_test_and_set_binding(struct rpc_xprt *xprt)
static DEFINE_SPINLOCK(xprt_list_lock);
static LIST_HEAD(xprt_list);
- /*
- * The transport code maintains an estimate on the maximum number of out-
- * standing RPC requests, using a smoothed version of the congestion
- * avoidance implemented in 44BSD. This is basically the Van Jacobson
- * congestion algorithm: If a retransmit occurs, the congestion window is
- * halved; otherwise, it is incremented by 1/cwnd when
- *
- * - a reply is received and
- * - a full number of requests are outstanding and
- * - the congestion window hasn't been updated recently.
- */
- #define RPC_CWNDSHIFT (8U)
- #define RPC_CWNDSCALE (1U << RPC_CWNDSHIFT)
- #define RPC_INITCWND RPC_CWNDSCALE
- #define RPC_MAXCWND(xprt) ((xprt)->max_reqs << RPC_CWNDSHIFT)
-
- #define RPCXPRT_CONGESTED(xprt) ((xprt)->cong >= (xprt)->cwnd)
-
/**
* xprt_register_transport - register a transport implementation
* @transport: transport to register
{
xprt->snd_task = NULL;
if (!test_bit(XPRT_CLOSE_WAIT, &xprt->state)) {
- smp_mb__before_clear_bit();
+ smp_mb__before_atomic();
clear_bit(XPRT_LOCKED, &xprt->state);
- smp_mb__after_clear_bit();
+ smp_mb__after_atomic();
} else
queue_work(rpciod_workqueue, &xprt->task_cleanup);
}
* @task: recently completed RPC request used to adjust window
* @result: result code of completed RPC request
*
- * We use a time-smoothed congestion estimator to avoid heavy oscillation.
+ * The transport code maintains an estimate on the maximum number of out-
+ * standing RPC requests, using a smoothed version of the congestion
+ * avoidance implemented in 44BSD. This is basically the Van Jacobson
+ * congestion algorithm: If a retransmit occurs, the congestion window is
+ * halved; otherwise, it is incremented by 1/cwnd when
+ *
+ * - a reply is received and
+ * - a full number of requests are outstanding and
+ * - the congestion window hasn't been updated recently.
*/
void xprt_adjust_cwnd(struct rpc_xprt *xprt, struct rpc_task *task, int result)
{
projects / mirror_ubuntu-artful-kernel.git / commitdiff