LU-5873 ldiskfs: osd_do_bio()) ASSERTION( iobuf->dr_rw == 0 ) failed

The bug happens when  16TB-4KB limit is exceeded during write.

Add check for maximum file size on client and server sides.

Xyratex-bug-id: MRP-2131
Change-Id: I73f0ee803670ada869c2618f275049948668848e
Signed-off-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Reviewed-on: http://review.whamcloud.com/12600
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>

diff --git a/lustre/llite/file.c b/lustre/llite/file.c
index 0ee0dde..6c8bc5c 100644 (file)
--- a/lustre/llite/file.c
+++ b/lustre/llite/file.c
@@ -1137,7 +1137,9 @@ ll_file_io_generic(const struct lu_env *env, struct vvp_io_args *args,
                   struct file *file, enum cl_io_type iot,
                   loff_t *ppos, size_t count)
 {
-       struct ll_inode_info *lli = ll_i2info(file->f_dentry->d_inode);
+       struct inode *inode = file->f_dentry->d_inode;
+       struct ll_inode_info *lli = ll_i2info(inode);
+       loff_t               end;
        struct ll_file_data  *fd  = LUSTRE_FPRIVATE(file);
        struct cl_io         *io;
        ssize_t               result;
@@ -1151,6 +1153,19 @@ restart:
         io = ccc_env_thread_io(env);
         ll_io_init(io, file, iot == CIT_WRITE);
 
+       /* The maximum Lustre file size is variable, based on the
+        * OST maximum object size and number of stripes.  This
+        * needs another check in addition to the VFS checks earlier. */
+       end = (io->u.ci_wr.wr_append ? i_size_read(inode) : *ppos) + count;
+       if (end > ll_file_maxbytes(inode)) {
+               result = -EFBIG;
+               CDEBUG(D_INODE, "%s: file "DFID" offset %llu > maxbytes "LPU64
+                      ": rc = %zd\n", ll_get_fsname(inode->i_sb, NULL, 0),
+                      PFID(&lli->lli_fid), end, ll_file_maxbytes(inode),
+                      result);
+               RETURN(result);
+       }
+
         if (cl_io_rw_init(env, io, iot, *ppos, count) == 0) {
                struct vvp_io *vio = vvp_env_io(env);
                bool range_locked = false;

diff --git a/lustre/osd-ldiskfs/osd_io.c b/lustre/osd-ldiskfs/osd_io.c
index d731d4a..e2e08e3 100644 (file)
--- a/lustre/osd-ldiskfs/osd_io.c
+++ b/lustre/osd-ldiskfs/osd_io.c
@@ -790,10 +790,17 @@ static int osd_ldiskfs_map_bm_inode_pages(struct inode *inode,
                                          unsigned long *blocks, int create)
 {
        int blocks_per_page = PAGE_CACHE_SIZE >> inode->i_blkbits;
+       pgoff_t bitmap_max_page_index;
        unsigned long *b;
        int rc = 0, i;
 
+       bitmap_max_page_index = LDISKFS_SB(inode->i_sb)->s_bitmap_maxbytes >>
+                               PAGE_SHIFT;
        for (i = 0, b = blocks; i < pages; i++, page++) {
+               if ((*page)->index + 1 >= bitmap_max_page_index) {
+                       rc = -EFBIG;
+                       break;
+               }
                rc = ldiskfs_map_inode_page(inode, *page, b, create);
                if (rc) {
                        CERROR("ino %lu, blk %lu create %d: rc %d\n",
@@ -814,6 +821,9 @@ static int osd_ldiskfs_map_ext_inode_pages(struct inode *inode,
        int rc = 0, i = 0;
        struct page *fp = NULL;
        int clen = 0;
+       pgoff_t extent_max_page_index;
+
+       extent_max_page_index = inode->i_sb->s_maxbytes >> PAGE_SHIFT;
 
        CDEBUG(D_OTHER, "inode %lu: map %d pages from %lu\n",
                inode->i_ino, pages, (*page)->index);
@@ -835,6 +845,9 @@ static int osd_ldiskfs_map_ext_inode_pages(struct inode *inode,
                        continue;
                }
 
+               if (fp->index + i >= extent_max_page_index)
+                       GOTO(cleanup, rc = -EFBIG);
+
                /* process found extent */
                rc = osd_ldiskfs_map_nblocks(inode, fp->index * blocks_per_page,
                                             clen * blocks_per_page, blocks,
@@ -879,6 +892,9 @@ static int osd_ldiskfs_map_inode_pages(struct inode *inode, struct page **page,
        int rc = 0, i = 0;
        struct page *fp = NULL;
        int clen = 0;
+       pgoff_t max_page_index;
+
+       max_page_index = inode->i_sb->s_maxbytes >> PAGE_SHIFT;
 
        CDEBUG(D_OTHER, "inode %lu: map %d pages from %lu\n",
                inode->i_ino, pages, (*page)->index);
@@ -902,6 +918,8 @@ static int osd_ldiskfs_map_inode_pages(struct inode *inode, struct page **page,
                        if (++i != pages)
                                continue;
                }
+               if (fp->index + i >= max_page_index)
+                       GOTO(cleanup, rc = -EFBIG);
                /* process found extent */
                map.m_lblk = fp->index * blocks_per_page;
                map.m_len = blen = clen * blocks_per_page;

diff --git a/lustre/tests/sanity.sh b/lustre/tests/sanity.sh
index 03c9cc9..1f06a98 100644 (file)
--- a/lustre/tests/sanity.sh
+++ b/lustre/tests/sanity.sh
@@ -12941,6 +12941,20 @@ test_243()
 }
 run_test 243 "various group lock tests"
 
+test_250() {
+       [ "$(facet_fstype ost$(($($GETSTRIPE -i $DIR/$tfile) + 1)))" = "zfs" ] \
+        && skip "no 16TB file size limit on ZFS" && return
+
+       $SETSTRIPE -c 1 $DIR/$tfile
+       # ldiskfs extent file size limit is (16TB - 4KB - 1) bytes
+       local size=$((16 * 1024 * 1024 * 1024 * 1024 - 4096 - 1))
+       $TRUNCATE $DIR/$tfile $size || error "truncate $tfile to $size failed"
+       dd if=/dev/zero of=$DIR/$tfile bs=10 count=1 oflag=append \
+               conv=notrunc,fsync && error "append succeeded"
+       return 0
+}
+run_test 250 "Write above 16T limit"
+
 cleanup_test_300() {
        trap 0
        umask $SAVE_UMASK