I realised there was no good reason for any of the three filesystems which actually use read_cache_page() to pass in something that wasn't a struct file pointer. Indeed, it made each of them more complex. These aren't filesystems I test regularly, so please scrutinise carefully. This is on top of the read_folio() patchset that I posted recently and can be found at git://git.infradead.org/users/willy/pagecache.git for-next Matthew Wilcox (Oracle) (3): jffs2: Pass the file pointer to jffs2_do_readpage_unlock() nfs: Pass the file pointer to nfs_symlink_filler() fs: Change the type of filler_t fs/gfs2/aops.c | 29 +++++++++++------------------ fs/jffs2/file.c | 9 ++++----- fs/jffs2/gc.c | 2 +- fs/jffs2/os-linux.h | 2 +- fs/nfs/symlink.c | 16 ++++++++-------- include/linux/pagemap.h | 6 +++--- mm/filemap.c | 40 ++++++++++++++++++++-------------------- 7 files changed, 48 insertions(+), 56 deletions(-) -- 2.34.1
