diff options
author | Trond Myklebust <trond.myklebust@hammerspace.com> | 2022-02-22 18:39:26 +0300 |
---|---|---|
committer | Trond Myklebust <trond.myklebust@hammerspace.com> | 2022-02-28 18:11:33 +0300 |
commit | 0b2662b7e7fddaaf6d6c055763270765cdb97a0d (patch) | |
tree | d814b3052e607005cfc849e72b374425f02486dc | |
parent | 281f31b2e5a2f1c5ae82dbe0b14c9d57401e0967 (diff) | |
download | linux-0b2662b7e7fddaaf6d6c055763270765cdb97a0d.tar.xz |
NFS: Calculate page offsets algorithmically
Instead of relying on counting the page offsets as we walk through the
page cache, switch to calculating them algorithmically.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
-rw-r--r-- | fs/nfs/dir.c | 18 |
1 files changed, 13 insertions, 5 deletions
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 379f88b158fb..6f0a38db6c37 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -249,17 +249,20 @@ static const char *nfs_readdir_copy_name(const char *name, unsigned int len) return ret; } +static size_t nfs_readdir_array_maxentries(void) +{ + return (PAGE_SIZE - sizeof(struct nfs_cache_array)) / + sizeof(struct nfs_cache_array_entry); +} + /* * Check that the next array entry lies entirely within the page bounds */ static int nfs_readdir_array_can_expand(struct nfs_cache_array *array) { - struct nfs_cache_array_entry *cache_entry; - if (array->page_full) return -ENOSPC; - cache_entry = &array->array[array->size + 1]; - if ((char *)cache_entry - (char *)array > PAGE_SIZE) { + if (array->size == nfs_readdir_array_maxentries()) { array->page_full = 1; return -ENOSPC; } @@ -318,6 +321,11 @@ static struct page *nfs_readdir_page_get_locked(struct address_space *mapping, return page; } +static loff_t nfs_readdir_page_offset(struct page *page) +{ + return (loff_t)page->index * (loff_t)nfs_readdir_array_maxentries(); +} + static u64 nfs_readdir_page_last_cookie(struct page *page) { struct nfs_cache_array *array; @@ -448,7 +456,7 @@ static int nfs_readdir_search_for_cookie(struct nfs_cache_array *array, if (array->array[i].cookie == desc->dir_cookie) { struct nfs_inode *nfsi = NFS_I(file_inode(desc->file)); - new_pos = desc->current_index + i; + new_pos = nfs_readdir_page_offset(desc->page) + i; if (desc->attr_gencount != nfsi->attr_gencount || !nfs_readdir_inode_mapping_valid(nfsi)) { desc->duped = 0; |