Hi Ben-
A few review comments below.
> On Sep 12, 2019, at 1:07 PM, Benjamin Coddington <bcodding@xxxxxxxxxx> wrote:
>
> The GSS Message Integrity Check data for krb5i may lie partially in the XDR
> reply buffer's pages and tail. If so, we try to copy the entire MIC into
> free space in the tail. But as the estimations of the slack space required
> for authentication and verification have improved there may be less free
> space in the tail to complete this copy -- see commit 2c94b8eca1a2
> ("SUNRPC: Use au_rslack when computing reply buffer size"). In fact, there
> may only be room in the tail for a single copy of the MIC, and not part of
> the MIC and then another complete copy.
>
> The real world failure reported is that `ls` of a directory on NFS may
> sometimes return -EIO, which can be traced back to xdr_buf_read_netobj()
> failing to find available free space in the tail to copy the MIC.
>
> Fix this by checking for the case of the MIC crossing the boundaries of
> head, pages, and tail. If so, shift the buffer until the MIC is contained
> completely within the pages or tail. This allows the remainder of the
> function to create a sub buffer that directly address the complete MIC.
>
> Signed-off-by: Benjamin Coddington <bcodding@xxxxxxxxxx>
> Cc: stable@xxxxxxxxxxxxxxx
> ---
> net/sunrpc/xdr.c | 45 +++++++++++++++++++++++++++------------------
> 1 file changed, 27 insertions(+), 18 deletions(-)
>
> diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c
> index 48c93b9e525e..6e05a9693568 100644
> --- a/net/sunrpc/xdr.c
> +++ b/net/sunrpc/xdr.c
> @@ -1237,39 +1237,48 @@ xdr_encode_word(struct xdr_buf *buf, unsigned int base, u32 obj)
> EXPORT_SYMBOL_GPL(xdr_encode_word);
>
> /* If the netobj starting offset bytes from the start of xdr_buf is contained
> - * entirely in the head or the tail, set object to point to it; otherwise
> - * try to find space for it at the end of the tail, copy it there, and
> - * set obj to point to it. */
> + * entirely in the head, pages, or tail, set object to point to it; otherwise
> + * shift the buffer until it is contained entirely within the pages or tail.
> + */
Nit: I would explicitly note in this comment that there is no need
for the caller to free @obj, and perhaps it should be noted that
the organization of @buf can be changed by this function.
Maybe more appropriate for 2/2.
> int xdr_buf_read_netobj(struct xdr_buf *buf, struct xdr_netobj *obj, unsigned int offset)
> {
> struct xdr_buf subbuf;
> + unsigned int len_to_boundary;
>
> if (xdr_decode_word(buf, offset, &obj->len))
> return -EFAULT;
> - if (xdr_buf_subsegment(buf, &subbuf, offset + 4, obj->len))
> +
> + offset += 4;
Nit: No blank line before "offset += 4;" would help me understand
how the offset bump is related to xdr_decode_word(). It took me
a few blinks to see.
> +
> + /* Is the obj partially in the head? */
> + len_to_boundary = buf->head->iov_len - offset;
> + if (len_to_boundary > 0 && len_to_boundary < obj->len)
I'm not especially excited about the integer underflow when offset
is larger than buf->head->iov_len. This might be more explicit:
if (offset < buf->head[0].iov_len &&
offset + obj->len > buf->head[0].iov_len)
> + xdr_shift_buf(buf, len_to_boundary);
> +
> + /* Is the obj partially in the pages? */
> + len_to_boundary = buf->head->iov_len + buf->page_len - offset;
> + if (len_to_boundary > 0 && len_to_boundary < obj->len)
Ditto.
> + xdr_shrink_pagelen(buf, len_to_boundary);
> +
> + if (xdr_buf_subsegment(buf, &subbuf, offset, obj->len))
> return -EFAULT;
>
> - /* Is the obj contained entirely in the head? */
> - obj->data = subbuf.head[0].iov_base;
> - if (subbuf.head[0].iov_len == obj->len)
> - return 0;
> - /* ..or is the obj contained entirely in the tail? */
> + /* Most likely: is the obj contained entirely in the tail? */
> obj->data = subbuf.tail[0].iov_base;
> if (subbuf.tail[0].iov_len == obj->len)
> return 0;
>
> - /* use end of tail as storage for obj:
> - * (We don't copy to the beginning because then we'd have
> - * to worry about doing a potentially overlapping copy.
> - * This assumes the object is at most half the length of the
> - * tail.) */
> + /* ..or is the obj contained entirely in the head? */
> + obj->data = subbuf.head[0].iov_base;
> + if (subbuf.head[0].iov_len == obj->len)
> + return 0;
It looks like you're reversing these two tests as a micro-optimization?
Maybe that should be left for another patch, since this is supposed to
be a narrow fix.
Also, I found the new comments confusing: here they refer to the head
and tail of @subbuf; above they refer to head and tail of @buf. Note for
2/2, I guess.
> +
> + /* obj is in the pages: move to tail */
> if (obj->len > buf->buflen - buf->len)
> return -ENOMEM;
> - if (buf->tail[0].iov_len != 0)
> - obj->data = buf->tail[0].iov_base + buf->tail[0].iov_len;
> - else
> - obj->data = buf->head[0].iov_base + buf->head[0].iov_len;
> + obj->data = buf->head[0].iov_base + buf->head[0].iov_len;
Not sure this is a safe change. It's possible that the head buffer
and tail buffer are not contiguous, which is what the buf->tail.iov_len
check is looking for, IMO. Can this hunk be left out?
> __read_bytes_from_xdr_buf(&subbuf, obj->data, obj->len);
> +
> return 0;
> }
> EXPORT_SYMBOL_GPL(xdr_buf_read_netobj);
> --
> 2.20.1
>
--
Chuck Lever
