| Message ID | 20220919033747.461583-2-lsahlber@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | cifs: invalidate and destage pages before re-reading them for cache=none | expand |
diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 6f38b134a346..31ceef22ffe4 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -4314,6 +4314,9 @@ static ssize_t __cifs_readv( ssize_t cifs_direct_readv(struct kiocb *iocb, struct iov_iter *to) { + struct file *file = iocb->ki_filp; + + cifs_revalidate_mapping(file->f_inode); return __cifs_readv(iocb, to, true); }
This is the opposite case of kernel bugzilla 216301. If we mmap a file using cache=none and then proceed to update the mmapped area these updates are not reflected in a later pread() of that part of the file. To fix this we must invalidate the mapping, forcing the changed pages to be destaged to the server before we allow the pread() to proceed. Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com> --- fs/cifs/file.c | 3 +++ 1 file changed, 3 insertions(+)