summaryrefslogtreecommitdiff
path: root/fatfs/pager.c
diff options
context:
space:
mode:
authorJeff Bailey <jbailey@gnu.org>2003-07-28 01:48:26 +0000
committerJeff Bailey <jbailey@gnu.org>2003-07-28 01:48:26 +0000
commit211ee30f24c0abf5fcd5df17451b738e32c2c2c5 (patch)
treee2281b39ab091d290441f038de849494e99a76ea /fatfs/pager.c
parent3099f96171c69d65a1b7add77e9a7356b073d92b (diff)
2003-07-14 Marco Gerards <metgerards@student.han.nl>
* fat.c (fat_read_sblock): Don't test if the root dir size is a multiple of sectors_per_cluster. Reported by Barry deFreese (bddebian@cox.net). * fatfs.h (LOG2_BLOCKS_PER_CLUSTER): New macro. (FAT_FIRST_CLUSTER_BLOCK): Likewise. (fat_first_cluster_byte): Macro removed. * inode.c (read_node): Correctly setup diskfs_root_node for FAT32 filesystems. * pager.c (fat_getcluster): Check for reading beyond allocsize correctly for file systems with a clustersize > vm_page_size. (file_pager_read_small_page): Don't use byte offsets when calculating the block. (file_pager_read_huge_page): Likewise. (pending_clusters_write): Likewise. (file_pager_write_small_page): Likewise. (STAT_INC): Cast to void to suppress warning.
Diffstat (limited to 'fatfs/pager.c')
-rw-r--r--fatfs/pager.c23
1 files changed, 13 insertions, 10 deletions
diff --git a/fatfs/pager.c b/fatfs/pager.c
index 606dc4d8..61cf7cf7 100644
--- a/fatfs/pager.c
+++ b/fatfs/pager.c
@@ -1,5 +1,5 @@
/* pager.c - Pager for fatfs.
- Copyright (C) 1997, 1999, 2002 Free Software Foundation, Inc.
+ Copyright (C) 1997, 1999, 2002, 2003 Free Software Foundation, Inc.
Written by Thomas Bushnell, n/BSG and Marcus Brinkmann.
This file is part of the GNU Hurd.
@@ -36,7 +36,7 @@ spin_lock_t node_to_page_lock = SPIN_LOCK_INITIALIZER;
#define MAY_CACHE 1
#endif
-#define STAT_INC(field) /* nop */0
+#define STAT_INC(field) (void) 0
#define MAX_FREE_PAGE_BUFS 32
@@ -178,9 +178,11 @@ file_pager_read_small_page (struct node *node, vm_offset_t page,
if (!err)
{
err = store_read (store,
- (fat_first_cluster_byte(cluster) +
- (page % bytes_per_cluster)) >> store->log2_block_size,
+ FAT_FIRST_CLUSTER_BLOCK(cluster)
+ + ((page % bytes_per_cluster)
+ >> store->log2_block_size),
vm_page_size, (void **) buf, &read);
+
if (read != vm_page_size)
err = EIO;
}
@@ -215,7 +217,7 @@ file_pager_read_huge_page (struct node *node, vm_offset_t page,
{
if (num_pending_clusters > 0)
{
- size_t dev_block = fat_first_cluster_byte(pending_clusters) >> store->log2_block_size;
+ size_t dev_block = FAT_FIRST_CLUSTER_BLOCK(pending_clusters);
size_t amount = num_pending_clusters << log2_bytes_per_cluster;
/* The buffer we try to read into; on the first read, we pass in a
size of zero, so that the read is guaranteed to allocate a new
@@ -317,7 +319,8 @@ pending_clusters_write (struct pending_clusters *pc)
if (pc->num > 0)
{
error_t err;
- size_t dev_block = fat_first_cluster_byte(pc->cluster) >> store->log2_block_size;
+ size_t dev_block = FAT_FIRST_CLUSTER_BLOCK(pc->cluster);
+
size_t length = pc->num << log2_bytes_per_cluster, amount;
if (pc->offs > 0)
@@ -477,10 +480,10 @@ file_pager_write_small_page (struct node *node, vm_offset_t offset, void *buf)
if (!err)
{
- err = store_write (store,
- (fat_first_cluster_byte(cluster) +
- (offset % bytes_per_cluster)) >> store->log2_block_size,
- (void **) buf, vm_page_size, &write);
+ err = store_write (store, FAT_FIRST_CLUSTER_BLOCK(cluster)
+ + ((offset % bytes_per_cluster)
+ >> store->log2_block_size),
+ (void **) buf, vm_page_size, &write);
if (write != vm_page_size)
err = EIO;
}