On Thu, 16 Dec 2010 20:37:32 +0300
Pavel Shilovsky <piastry@xxxxxxxxxxx> wrote:
> From: Pavel Shilovsky <piastryyy@xxxxxxxxx>
>
> Read from the cache if we have at least Level II oplock - otherwise
> read from the server. Add cifs_user_readv to let the client read into
> several buffers.
>
> Signed-off-by: Pavel Shilovsky <piastryyy@xxxxxxxxx>
> ---
> fs/cifs/cifsfs.c | 38 ++++++++++++++++++-
> fs/cifs/cifsfs.h | 4 +-
> fs/cifs/file.c | 108 +++++++++++++++++++++++++++++++++++++++---------------
> 3 files changed, 117 insertions(+), 33 deletions(-)
>
> diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
> index a218afe..bcb4d2c 100644
> --- a/fs/cifs/cifsfs.c
> +++ b/fs/cifs/cifsfs.c
> @@ -572,6 +572,40 @@ cifs_do_mount(struct file_system_type *fs_type,
> return dget(sb->s_root);
> }
>
> +static ssize_t cifs_strict_read(struct kiocb *iocb, const struct iovec *iov,
> + unsigned long nr_segs, loff_t pos)
> +{
> + struct inode *inode;
> + struct cifs_sb_info *cifs_sb;
> + ssize_t read = 0;
> +
> + if (!nr_segs)
> + return read;
> +
> + inode = iocb->ki_filp->f_path.dentry->d_inode;
> + cifs_sb = CIFS_SB(iocb->ki_filp->f_path.dentry->d_sb);
> +
> + if (CIFS_I(inode)->clientCanCacheRead)
> + return generic_file_aio_read(iocb, iov, nr_segs, pos);
> +
> + /*
> + * In strict cache mode we need to read from the server all the time
> + * if we don't have level II oplock because the server can delay mtime
> + * change - so we can't make a decision about inode invalidating.
> + * And we can also fail with pagereading if there are mandatory locks
> + * on pages affected by this read but not on the region from pos to
> + * pos+len-1.
> + */
> +
> + read = cifs_user_readv(iocb->ki_filp, iov, nr_segs, &pos);
> + if (read < 0)
> + return read;
> +
> + iocb->ki_pos = pos;
> +
> + return read;
> +}
> +
> static ssize_t cifs_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
> unsigned long nr_segs, loff_t pos)
> {
> @@ -713,7 +747,7 @@ const struct file_operations cifs_file_ops = {
> const struct file_operations cifs_file_strict_ops = {
> .read = do_sync_read,
> .write = do_sync_write,
> - .aio_read = generic_file_aio_read,
> + .aio_read = cifs_strict_read,
> .aio_write = cifs_file_aio_write,
> .open = cifs_open,
> .release = cifs_close,
> @@ -769,7 +803,7 @@ const struct file_operations cifs_file_nobrl_ops = {
> const struct file_operations cifs_file_strict_nobrl_ops = {
> .read = do_sync_read,
> .write = do_sync_write,
> - .aio_read = generic_file_aio_read,
> + .aio_read = cifs_strict_read,
> .aio_write = cifs_file_aio_write,
> .open = cifs_open,
> .release = cifs_close,
> diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h
> index 657738f..f690d1c 100644
> --- a/fs/cifs/cifsfs.h
> +++ b/fs/cifs/cifsfs.h
> @@ -81,7 +81,9 @@ extern int cifs_open(struct inode *inode, struct file *file);
> extern int cifs_close(struct inode *inode, struct file *file);
> extern int cifs_closedir(struct inode *inode, struct file *file);
> extern ssize_t cifs_user_read(struct file *file, char __user *read_data,
> - size_t read_size, loff_t *poffset);
> + size_t read_size, loff_t *poffset);
> +extern ssize_t cifs_user_readv(struct file *file, const struct iovec *iov,
> + unsigned long nr_segs, loff_t *poffset);
> extern ssize_t cifs_user_write(struct file *file, const char __user *write_data,
> size_t write_size, loff_t *poffset);
> extern int cifs_lock(struct file *, int, struct file_lock *);
> diff --git a/fs/cifs/file.c b/fs/cifs/file.c
> index d48723c..bc53e9e 100644
> --- a/fs/cifs/file.c
> +++ b/fs/cifs/file.c
> @@ -1677,20 +1677,60 @@ int cifs_flush(struct file *file, fl_owner_t id)
> return rc;
> }
>
> -ssize_t cifs_user_read(struct file *file, char __user *read_data,
> - size_t read_size, loff_t *poffset)
> +static int
> +cifs_fill_iov(const struct iovec *iov, unsigned long nr_segs,
> + unsigned long *last_len, unsigned long *last_ind, char *data,
> + size_t datalen)
> {
I have to wonder if we already have generic helpers for this sort of thing?
> - int rc = -EACCES;
> - unsigned int bytes_read = 0;
> - unsigned int total_read = 0;
> - unsigned int current_read_size;
> + const struct iovec *ciov;
> + unsigned long cur_len;
> +
> + while (datalen) {
> + cur_len = min_t(unsigned long, datalen,
> + *last_len);
^^^^^
This can be on a single line...
> + ciov = &iov[*last_ind];
> +
> + if (copy_to_user(ciov->iov_base + (ciov->iov_len - *last_len),
> + data, cur_len))
> + return -EFAULT;
> +
> + datalen -= cur_len;
> + data += cur_len;
> +
> + *last_len -= cur_len;
> + if (*last_len == 0) {
> + (*last_ind)++;
> + *last_len = iov[*last_ind].iov_len;
> + }
> + }
> +
> + return 0;
> +}
> +
> +ssize_t cifs_user_readv(struct file *file, const struct iovec *iov,
> + unsigned long nr_segs, loff_t *poffset)
> +{
> + int rc;
> + int xid;
> + unsigned int bytes_read = 0, total_read = 0;
> + unsigned long last_len = 0, len = 0, last_ind = 0;
> + unsigned long cur_len, i;
> struct cifs_sb_info *cifs_sb;
> struct cifsTconInfo *pTcon;
> - int xid;
> struct cifsFileInfo *open_file;
> - char *smb_read_data;
> - char __user *current_offset;
> struct smb_com_read_rsp *pSMBr;
> + char *read_data;
> +
> + if (!nr_segs)
> + return 0;
> +
> + last_len = iov->iov_len;
> +
> + for (i = 0; i < nr_segs; i++)
> + len += iov[i].iov_len;
> +
> + if (!len)
> + return 0;
>
> xid = GetXid();
> cifs_sb = CIFS_SB(file->f_path.dentry->d_sb);
> @@ -1700,19 +1740,19 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data,
> FreeXid(xid);
> return rc;
> }
> +
> open_file = file->private_data;
> pTcon = tlink_tcon(open_file->tlink);
>
> if ((file->f_flags & O_ACCMODE) == O_WRONLY)
> cFYI(1, "attempting read on write only file instance");
>
> - for (total_read = 0, current_offset = read_data;
> - read_size > total_read;
> - total_read += bytes_read, current_offset += bytes_read) {
> - current_read_size = min_t(const int, read_size - total_read,
> + for (total_read = 0; total_read < len; total_read += bytes_read) {
> + cur_len = min_t(const int, len - total_read,
> cifs_sb->rsize);
> rc = -EAGAIN;
> - smb_read_data = NULL;
> + read_data = NULL;
> +
> while (rc == -EAGAIN) {
> int buf_type = CIFS_NO_BUFFER;
> if (open_file->invalidHandle) {
> @@ -1721,26 +1761,24 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data,
> break;
> }
> rc = CIFSSMBRead(xid, pTcon,
> - open_file->netfid,
> - current_read_size, *poffset,
> - &bytes_read, &smb_read_data,
> - &buf_type);
> - pSMBr = (struct smb_com_read_rsp *)smb_read_data;
> - if (smb_read_data) {
> - if (copy_to_user(current_offset,
> - smb_read_data +
> - 4 /* RFC1001 length field */ +
> - le16_to_cpu(pSMBr->DataOffset),
> - bytes_read))
> + open_file->netfid, cur_len, *poffset,
> + &bytes_read, &read_data, &buf_type);
> + pSMBr = (struct smb_com_read_rsp *)read_data;
> + if (read_data) {
> + char *data_offset = read_data + 4 +
> + le16_to_cpu(pSMBr->DataOffset);
> + if (cifs_fill_iov(iov, nr_segs,
> + &last_len, &last_ind,
> + data_offset, bytes_read))
> rc = -EFAULT;
^^^^^^^^^^^^^
nit: it's generally better not to indent the inside of a block along with the conditions.
> -
> if (buf_type == CIFS_SMALL_BUFFER)
> - cifs_small_buf_release(smb_read_data);
> + cifs_small_buf_release(read_data);
> else if (buf_type == CIFS_LARGE_BUFFER)
> - cifs_buf_release(smb_read_data);
> - smb_read_data = NULL;
> + cifs_buf_release(read_data);
> + read_data = NULL;
> }
> }
> +
> if (rc || (bytes_read == 0)) {
> if (total_read) {
> break;
> @@ -1753,13 +1791,23 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data,
> *poffset += bytes_read;
> }
> }
> +
> FreeXid(xid);
> return total_read;
> }
>
> +ssize_t cifs_user_read(struct file *file, char __user *read_data,
> + size_t read_size, loff_t *poffset)
> +{
> + struct iovec iov;
> + iov.iov_base = read_data;
> + iov.iov_len = read_size;
> +
> + return cifs_user_readv(file, &iov, 1, poffset);
> +}
>
> static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size,
> - loff_t *poffset)
> + loff_t *poffset)
> {
> int rc = -EACCES;
> unsigned int bytes_read = 0;
Other than the minor nits above, it looks good to me.
Reviewed-by: Jeff Layton <jlayton@xxxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe linux-cifs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
