Merge branch 'mw-3.1-jul25' of git://oss.oracle.com/git/smushran/linux-2.6 into ocfs2...

diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index ff98c169b631efb1a0e2d2467f9a31eec9cc5e85..78b68af3b0e32627b1874277d8ae58003501acb5 100644 (file)
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -290,7 +290,15 @@ static int ocfs2_readpage(struct file *file, struct page *page)
        }
 
        if (down_read_trylock(&oi->ip_alloc_sem) == 0) {
+               /*
+                * Unlock the page and cycle ip_alloc_sem so that we don't
+                * busyloop waiting for ip_alloc_sem to unlock
+                */
                ret = AOP_TRUNCATED_PAGE;
+               unlock_page(page);
+               unlock = 0;
+               down_read(&oi->ip_alloc_sem);
+               up_read(&oi->ip_alloc_sem);
                goto out_inode_unlock;
        }
 
@@ -563,6 +571,7 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
 {
        struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode;
        int level;
+       wait_queue_head_t *wq = ocfs2_ioend_wq(inode);
 
        /* this io's submitter should not have unlocked this before we could */
        BUG_ON(!ocfs2_iocb_is_rw_locked(iocb));
@@ -570,6 +579,15 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
        if (ocfs2_iocb_is_sem_locked(iocb))
                ocfs2_iocb_clear_sem_locked(iocb);
 
+       if (ocfs2_iocb_is_unaligned_aio(iocb)) {
+               ocfs2_iocb_clear_unaligned_aio(iocb);
+
+               if (atomic_dec_and_test(&OCFS2_I(inode)->ip_unaligned_aio) &&
+                   waitqueue_active(wq)) {
+                       wake_up_all(wq);
+               }
+       }
+
        ocfs2_iocb_clear_rw_locked(iocb);
 
        level = ocfs2_iocb_rw_locked_level(iocb);

diff --git a/fs/ocfs2/aops.h b/fs/ocfs2/aops.h
index 75cf3ad987a66d911c15234a803243185ccc5a94..ffb2da370a99d05dd4b919fc64a5483dbc2df7a3 100644 (file)
--- a/fs/ocfs2/aops.h
+++ b/fs/ocfs2/aops.h
@@ -78,6 +78,7 @@ enum ocfs2_iocb_lock_bits {
        OCFS2_IOCB_RW_LOCK = 0,
        OCFS2_IOCB_RW_LOCK_LEVEL,
        OCFS2_IOCB_SEM,
+       OCFS2_IOCB_UNALIGNED_IO,
        OCFS2_IOCB_NUM_LOCKS
 };
 
@@ -91,4 +92,17 @@ enum ocfs2_iocb_lock_bits {
        clear_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private)
 #define ocfs2_iocb_is_sem_locked(iocb) \
        test_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private)
+
+#define ocfs2_iocb_set_unaligned_aio(iocb) \
+       set_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+#define ocfs2_iocb_clear_unaligned_aio(iocb) \
+       clear_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+#define ocfs2_iocb_is_unaligned_aio(iocb) \
+       test_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+
+#define OCFS2_IOEND_WQ_HASH_SZ 37
+#define ocfs2_ioend_wq(v)   (&ocfs2__ioend_wq[((unsigned long)(v)) %\
+                                           OCFS2_IOEND_WQ_HASH_SZ])
+extern wait_queue_head_t ocfs2__ioend_wq[OCFS2_IOEND_WQ_HASH_SZ];
+
 #endif /* OCFS2_FILE_H */

diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index 7642d7ca73e523f3bc29b7b8a5d767e8f7c7cb83..da103f555a621e82f178ec6ae47dbb074f727d60 100644 (file)
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -1692,7 +1692,7 @@ int ocfs2_open_lock(struct inode *inode)
        mlog(0, "inode %llu take PRMODE open lock\n",
             (unsigned long long)OCFS2_I(inode)->ip_blkno);
 
-       if (ocfs2_mount_local(osb))
+       if (ocfs2_is_hard_readonly(osb) || ocfs2_mount_local(osb))
                goto out;
 
        lockres = &OCFS2_I(inode)->ip_open_lockres;
@@ -1718,6 +1718,12 @@ int ocfs2_try_open_lock(struct inode *inode, int write)
             (unsigned long long)OCFS2_I(inode)->ip_blkno,
             write ? "EXMODE" : "PRMODE");
 
+       if (ocfs2_is_hard_readonly(osb)) {
+               if (write)
+                       status = -EROFS;
+               goto out;
+       }
+
        if (ocfs2_mount_local(osb))
                goto out;
 
@@ -2298,7 +2304,7 @@ int ocfs2_inode_lock_full_nested(struct inode *inode,
        if (ocfs2_is_hard_readonly(osb)) {
                if (ex)
                        status = -EROFS;
-               goto bail;
+               goto getbh;
        }
 
        if (ocfs2_mount_local(osb))
@@ -2356,7 +2362,7 @@ local:
                        mlog_errno(status);
                goto bail;
        }
-
+getbh:
        if (ret_bh) {
                status = ocfs2_assign_bh(inode, ret_bh, local_bh);
                if (status < 0) {
@@ -2628,8 +2634,11 @@ int ocfs2_dentry_lock(struct dentry *dentry, int ex)
 
        BUG_ON(!dl);
 
-       if (ocfs2_is_hard_readonly(osb))
-               return -EROFS;
+       if (ocfs2_is_hard_readonly(osb)) {
+               if (ex)
+                       return -EROFS;
+               return 0;
+       }
 
        if (ocfs2_mount_local(osb))
                return 0;
@@ -2647,7 +2656,7 @@ void ocfs2_dentry_unlock(struct dentry *dentry, int ex)
        struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
        struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
 
-       if (!ocfs2_mount_local(osb))
+       if (!ocfs2_is_hard_readonly(osb) && !ocfs2_mount_local(osb))
                ocfs2_cluster_unlock(osb, &dl->dl_lockres, level);
 }
 

diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index c0f015e11c28f2d45eecbcb62b3886108b3df228..5c4a74e04ab4b67e9e64a533f1968955b4d11eca 100644 (file)
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -2052,6 +2052,23 @@ out:
        return ret;
 }
 
+static void ocfs2_aiodio_wait(struct inode *inode)
+{
+       wait_queue_head_t *wq = ocfs2_ioend_wq(inode);
+
+       wait_event(*wq, (atomic_read(&OCFS2_I(inode)->ip_unaligned_aio) == 0));
+}
+
+static int ocfs2_is_io_unaligned(struct inode *inode, size_t count, loff_t pos)
+{
+       int blockmask = inode->i_sb->s_blocksize - 1;
+       loff_t final_size = pos + count;
+
+       if ((pos & blockmask) || (final_size & blockmask))
+               return 1;
+       return 0;
+}
+
 static int ocfs2_prepare_inode_for_refcount(struct inode *inode,
                                            struct file *file,
                                            loff_t pos, size_t count,
@@ -2230,6 +2247,7 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
        int full_coherency = !(osb->s_mount_opt &
                               OCFS2_MOUNT_COHERENCY_BUFFERED);
+       int unaligned_dio = 0;
 
        trace_ocfs2_file_aio_write(inode, file, file->f_path.dentry,
                (unsigned long long)OCFS2_I(inode)->ip_blkno,
@@ -2297,6 +2315,10 @@ relock:
                goto out;
        }
 
+       if (direct_io && !is_sync_kiocb(iocb))
+               unaligned_dio = ocfs2_is_io_unaligned(inode, iocb->ki_left,
+                                                     *ppos);
+
        /*
         * We can't complete the direct I/O as requested, fall back to
         * buffered I/O.
@@ -2311,6 +2333,18 @@ relock:
                goto relock;
        }
 
+       if (unaligned_dio) {
+               /*
+                * Wait on previous unaligned aio to complete before
+                * proceeding.
+                */
+               ocfs2_aiodio_wait(inode);
+
+               /* Mark the iocb as needing a decrement in ocfs2_dio_end_io */
+               atomic_inc(&OCFS2_I(inode)->ip_unaligned_aio);
+               ocfs2_iocb_set_unaligned_aio(iocb);
+       }
+
        /*
         * To later detect whether a journal commit for sync writes is
         * necessary, we sample i_size, and cluster count here.
@@ -2382,8 +2416,12 @@ out_dio:
        if ((ret == -EIOCBQUEUED) || (!ocfs2_iocb_is_rw_locked(iocb))) {
                rw_level = -1;
                have_alloc_sem = 0;
+               unaligned_dio = 0;
        }
 
+       if (unaligned_dio)
+               atomic_dec(&OCFS2_I(inode)->ip_unaligned_aio);
+
 out:
        if (rw_level != -1)
                ocfs2_rw_unlock(inode, rw_level);

diff --git a/fs/ocfs2/inode.h b/fs/ocfs2/inode.h
index 1c508b149b3ac1bd4325fd33a9aae6bdb70e024a..88924a3133fae7c15ca3f5a5259b64eecd97022e 100644 (file)
--- a/fs/ocfs2/inode.h
+++ b/fs/ocfs2/inode.h
@@ -43,6 +43,9 @@ struct ocfs2_inode_info
        /* protects extended attribute changes on this inode */
        struct rw_semaphore             ip_xattr_sem;
 
+       /* Number of outstanding AIO's which are not page aligned */
+       atomic_t                        ip_unaligned_aio;
+
        /* These fields are protected by ip_lock */
        spinlock_t                      ip_lock;
        u32                             ip_open_count;

diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c
index bc91072b72196fd335c4b7cbc02ba08cb67254e6..d9a6ce7792669fedc66cfdbe26b6fe4d00976e59 100644 (file)
--- a/fs/ocfs2/ioctl.c
+++ b/fs/ocfs2/ioctl.c
@@ -381,7 +381,7 @@ int ocfs2_info_handle_freeinode(struct inode *inode,
        if (!oifi) {
                status = -ENOMEM;
                mlog_errno(status);
-               goto bail;
+               goto out_err;
        }
 
        if (o2info_from_user(*oifi, req))
@@ -431,7 +431,7 @@ bail:
                o2info_set_request_error(&oifi->ifi_req, req);
 
        kfree(oifi);
-
+out_err:
        return status;
 }
 
@@ -666,7 +666,7 @@ int ocfs2_info_handle_freefrag(struct inode *inode,
        if (!oiff) {
                status = -ENOMEM;
                mlog_errno(status);
-               goto bail;
+               goto out_err;
        }
 
        if (o2info_from_user(*oiff, req))
@@ -716,7 +716,7 @@ bail:
                o2info_set_request_error(&oiff->iff_req, req);
 
        kfree(oiff);
-
+out_err:
        return status;
 }
 

diff --git a/fs/ocfs2/move_extents.c b/fs/ocfs2/move_extents.c
index cd9427023d2e72dd1d5109113569f7dee6eb1066..d3433d60dbb945e50eac45adc2c1161e2de478e2 100644 (file)
--- a/fs/ocfs2/move_extents.c
+++ b/fs/ocfs2/move_extents.c
@@ -746,7 +746,7 @@ static int ocfs2_move_extent(struct ocfs2_move_extents_context *context,
         */
        ocfs2_probe_alloc_group(inode, gd_bh, &goal_bit, len, move_max_hop,
                                new_phys_cpos);
-       if (!new_phys_cpos) {
+       if (!*new_phys_cpos) {
                ret = -ENOSPC;
                goto out_commit;
        }

diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index 409285854f647e2357223bc7a8d24c36b376a6bb..bcde467acca396a1916831849bc495faa5d413b5 100644 (file)
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -836,13 +836,13 @@ static inline unsigned int ocfs2_clusters_to_megabytes(struct super_block *sb,
 
 static inline void _ocfs2_set_bit(unsigned int bit, unsigned long *bitmap)
 {
-       __test_and_set_bit_le(bit, bitmap);
+       __set_bit_le(bit, bitmap);
 }
 #define ocfs2_set_bit(bit, addr) _ocfs2_set_bit((bit), (unsigned long *)(addr))
 
 static inline void _ocfs2_clear_bit(unsigned int bit, unsigned long *bitmap)
 {
-       __test_and_clear_bit_le(bit, bitmap);
+       __clear_bit_le(bit, bitmap);
 }
 #define ocfs2_clear_bit(bit, addr) _ocfs2_clear_bit((bit), (unsigned long *)(addr))
 

diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 3e7850380d2b6145cd917a7baaec0b6bf06f94f7..938e2b2b0c9c9f8635a43ad4686dff675885361f 100644 (file)
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -54,6 +54,7 @@
 #include "ocfs1_fs_compat.h"
 
 #include "alloc.h"
+#include "aops.h"
 #include "blockcheck.h"
 #include "dlmglue.h"
 #include "export.h"
@@ -1616,12 +1617,17 @@ static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt)
        return 0;
 }
 
+wait_queue_head_t ocfs2__ioend_wq[OCFS2_IOEND_WQ_HASH_SZ];
+
 static int __init ocfs2_init(void)
 {
-       int status;
+       int status, i;
 
        ocfs2_print_version();
 
+       for (i = 0; i < OCFS2_IOEND_WQ_HASH_SZ; i++)
+               init_waitqueue_head(&ocfs2__ioend_wq[i]);
+
        status = init_ocfs2_uptodate_cache();
        if (status < 0) {
                mlog_errno(status);
@@ -1760,7 +1766,7 @@ static void ocfs2_inode_init_once(void *data)
        ocfs2_extent_map_init(&oi->vfs_inode);
        INIT_LIST_HEAD(&oi->ip_io_markers);
        oi->ip_dir_start_lookup = 0;
-
+       atomic_set(&oi->ip_unaligned_aio, 0);
        init_rwsem(&oi->ip_alloc_sem);
        init_rwsem(&oi->ip_xattr_sem);
        mutex_init(&oi->ip_io_mutex);
@@ -1974,7 +1980,8 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err)
         * If we failed before we got a uuid_str yet, we can't stop
         * heartbeat.  Otherwise, do it.
         */
-       if (!mnt_err && !ocfs2_mount_local(osb) && osb->uuid_str)
+       if (!mnt_err && !ocfs2_mount_local(osb) && osb->uuid_str &&
+           !ocfs2_is_hard_readonly(osb))
                hangup_needed = 1;
 
        if (osb->cconn)