ceph: perform lazy reads when file mode and caps permit

If the file mode is marked as "lazy," perform cached/buffered reads when
the caps permit it.  Adjust the rdcache_gen and invalidation logic
accordingly so that we manage our cache based on the FILE_CACHE -or-
FILE_LAZYIO cap bits.

Signed-off-by: Sage Weil <sage@newdream.net>

4 files changed, 22 insertions, 13 deletions

diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index d9c60b84949a..e00797eed29c 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -552,7 +552,7 @@ static void writepages_finish(struct ceph_osd_request *req,
                  * page truncation thread, possibly losing some data that
                  * raced its way in
                  */
-                if ((issued & CEPH_CAP_FILE_CACHE) == 0)
+                if ((issued & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) == 0)
                         generic_error_remove_page(inode->i_mapping, page);
 
                 unlock_page(page);
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index 1a70a3ebf013..b28915d5f404 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -482,8 +482,8 @@ static void __check_cap_issue(struct ceph_inode_info *ci, struct ceph_cap *cap,
          * Each time we receive FILE_CACHE anew, we increment
          * i_rdcache_gen.
          */
-        if ((issued & CEPH_CAP_FILE_CACHE) &&
-            (had & CEPH_CAP_FILE_CACHE) == 0)
+        if ((issued & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) &&
+            (had & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) == 0)
                 ci->i_rdcache_gen++;
 
         /*
@@ -1509,11 +1509,13 @@ retry_locked:
             ci->i_wrbuffer_ref == 0 &&               /* no dirty pages... */
             ci->i_rdcache_gen &&                     /* may have cached pages */
             (file_wanted == 0 ||                     /* no open files */
-             (revoking & CEPH_CAP_FILE_CACHE)) &&     /*  or revoking cache */
+             (revoking & (CEPH_CAP_FILE_CACHE|
+                          CEPH_CAP_FILE_LAZYIO))) && /*  or revoking cache */
             !tried_invalidate) {
                 dout("check_caps trying to invalidate on %p\n", inode);
                 if (try_nonblocking_invalidate(inode) < 0) {
-                        if (revoking & CEPH_CAP_FILE_CACHE) {
+                        if (revoking & (CEPH_CAP_FILE_CACHE|
+                                        CEPH_CAP_FILE_LAZYIO)) {
                                 dout("check_caps queuing invalidate\n");
                                 queue_invalidate = 1;
                                 ci->i_rdcache_revoking = ci->i_rdcache_gen;
@@ -2276,7 +2278,8 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
          * try to invalidate (once).  (If there are dirty buffers, we
          * will invalidate _after_ writeback.)
          */
-        if (((cap->issued & ~newcaps) & CEPH_CAP_FILE_CACHE) &&
+        if (((cap->issued & ~newcaps) & (CEPH_CAP_FILE_CACHE|
+                                         CEPH_CAP_FILE_LAZYIO)) &&
             !ci->i_wrbuffer_ref) {
                 if (try_nonblocking_invalidate(inode) == 0) {
                         revoked_rdcache = 1;
@@ -2374,7 +2377,8 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
                         writeback = 1; /* will delay ack */
                 else if (dirty & ~newcaps)
                         check_caps = 1;  /* initiate writeback in check_caps */
-                else if (((used & ~newcaps) & CEPH_CAP_FILE_CACHE) == 0 ||
+                else if (((used & ~newcaps) & (CEPH_CAP_FILE_CACHE|
+                                               CEPH_CAP_FILE_LAZYIO)) == 0 ||
                            revoked_rdcache)
                         check_caps = 2;     /* send revoke ack in check_caps */
                 cap->issued = newcaps;
diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 85c86ed5f4c0..2329244f427b 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -740,28 +740,32 @@ static ssize_t ceph_aio_read(struct kiocb *iocb, const struct iovec *iov,
                              unsigned long nr_segs, loff_t pos)
 {
         struct file *filp = iocb->ki_filp;
+        struct ceph_file_info *fi = filp->private_data;
         loff_t *ppos = &iocb->ki_pos;
         size_t len = iov->iov_len;
         struct inode *inode = filp->f_dentry->d_inode;
         struct ceph_inode_info *ci = ceph_inode(inode);
         void *base = iov->iov_base;
         ssize_t ret;
-        int got = 0;
+        int want, got = 0;
         int checkeof = 0, read = 0;
 
         dout("aio_read %p %llx.%llx %llu~%u trying to get caps on %p\n",
              inode, ceph_vinop(inode), pos, (unsigned)len, inode);
 again:
         __ceph_do_pending_vmtruncate(inode);
-        ret = ceph_get_caps(ci, CEPH_CAP_FILE_RD, CEPH_CAP_FILE_CACHE,
-                            &got, -1);
+        if (fi->fmode & CEPH_FILE_MODE_LAZY)
+                want = CEPH_CAP_FILE_CACHE | CEPH_CAP_FILE_LAZYIO;
+        else
+                want = CEPH_CAP_FILE_CACHE;
+        ret = ceph_get_caps(ci, CEPH_CAP_FILE_RD, want, &got, -1);
         if (ret < 0)
                 goto out;
         dout("aio_read %p %llx.%llx %llu~%u got cap refs on %s\n",
              inode, ceph_vinop(inode), pos, (unsigned)len,
              ceph_cap_string(got));
 
-        if ((got & CEPH_CAP_FILE_CACHE) == 0 ||
+        if ((got & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) == 0 ||
             (iocb->ki_filp->f_flags & O_DIRECT) ||
             (inode->i_sb->s_flags & MS_SYNCHRONOUS))
                 /* hmm, this isn't really async... */
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 389f9dbd9949..5d893d31e399 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -442,8 +442,9 @@ int ceph_fill_file_size(struct inode *inode, int issued,
                          * the file is either opened or mmaped
                          */
                         if ((issued & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_RD|
-                                      CEPH_CAP_FILE_WR|CEPH_CAP_FILE_BUFFER|
-                                      CEPH_CAP_FILE_EXCL)) ||
+                                       CEPH_CAP_FILE_WR|CEPH_CAP_FILE_BUFFER|
+                                       CEPH_CAP_FILE_EXCL|
+                                       CEPH_CAP_FILE_LAZYIO)) ||
                             mapping_mapped(inode->i_mapping) ||
                             __ceph_caps_file_wanted(ci)) {
                                 ci->i_truncate_pending++;