Skip to contentSkip to conversation

macvtap: update file current position

PATCH In review Zhi Yong Wu 3 participants 7 messages Dec 6, 2013 – Dec 7, 2013

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH v2 1/2 macvtap: update file current position

From: Zhi Yong Wu <[email protected]>

---
 drivers/net/macvtap.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)
drivers/net/macvtap.c +2 -0
index 9093004..e6e2dd6 100644
@@ -876,6 +876,8 @@ static ssize_t macvtap_aio_read(struct kiocb *iocb, const struct iovec *iv,
 
 	ret = macvtap_do_read(q, iocb, iv, len, file->f_flags & O_NONBLOCK);
 	ret = min_t(ssize_t, ret, len); /* XXX copied from tun.c. Why? */
+	if (ret > 0)
+		iocb->ki_pos += ret;
 out:
 	return ret;
 }
Signature · 1 line
1.7.6.5
#1 lore raw

PATCH v2 2/2 tun: update file current position

From: Zhi Yong Wu <[email protected]>

---
 drivers/net/tun.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)
drivers/net/tun.c +2 -0
index 782e38b..c8ddbd0 100644
@@ -1356,6 +1356,8 @@ static ssize_t tun_chr_aio_read(struct kiocb *iocb, const struct iovec *iv,
 	ret = tun_do_read(tun, tfile, iocb, iv, len,
 			  file->f_flags & O_NONBLOCK);
 	ret = min_t(ssize_t, ret, len);
+	if (ret > 0)
+		iocb->ki_pos += ret;
 out:
 	tun_put(tun);
 	return ret;
Signature · 1 line
1.7.6.5
maintainer
#2 lore raw

Re: [PATCH v2 2/2] tun: update file current position

In reply to Zhi Yong Wu

From: Zhi Yong Wu <[email protected]>
Date: Fri, 6 Dec 2013 17:08:50 +0800

Quoted 3 lines

From: Zhi Yong Wu <[email protected]>

Also applied and queued up for -stable, thanks.

I noticed in these two cases that that min_t() adjustment of 'ret' seems strange. I can't understand why it's needed.

If, for example, tun_do_read() really did read more than 'len' bytes:

1) That would write past the end of the buffer.

2) Writing a different value to the ->ki_pos would mean that ->ki_pos is now inaccurate.

Unless someone can explain why the min_t() is needed, we should remove it.

#3 lore raw

Re: [PATCH v2 2/2] tun: update file current position

Quoted 15 lines · David Miller

From: Zhi Yong Wu <[email protected]>
Date: Fri, 6 Dec 2013 17:08:50 +0800

Quoted 3 lines

From: Zhi Yong Wu <[email protected]>

Also applied and queued up for -stable, thanks.

I noticed in these two cases that that min_t() adjustment of 'ret' seems strange. I can't understand why it's needed.

If, for example, tun_do_read() really did read more than 'len' bytes:

Quoted 7 lines

1) That would write past the end of the buffer.

2) Writing a different value to the ->ki_pos would mean that ->ki_pos is now inaccurate.

Unless someone can explain why the min_t() is needed, we should remove it.

Yes, from my side, it seems to be impossible that ret is bigger than let or total_len.
So we also remove the branch "if (ret > total_len) {...}" in xxx_rcvmsg(). If you hope to submit the patch for this, please let me know, thanks.

Signature · 3 lines
Regards,

Zhi Yong Wu
maintainer
#4 lore raw

Re: [PATCH v2 2/2] tun: update file current position

From: Zhi Yong Wu <[email protected]>
Date: Sat, 7 Dec 2013 04:32:20 +0800

Quoted 1 line

If you hope to submit the patch for this, please let me know, thanks.

Someone should :-)

#5 lore raw

Re: [PATCH v2 2/2] tun: update file current position

In reply to David Miller
Quoted 15 lines · David Miller

From: Zhi Yong Wu <[email protected]>
Date: Fri, 6 Dec 2013 17:08:50 +0800

Quoted 3 lines

From: Zhi Yong Wu <[email protected]>

Also applied and queued up for -stable, thanks.

I noticed in these two cases that that min_t() adjustment of 'ret' seems strange. I can't understand why it's needed.

If, for example, tun_do_read() really did read more than 'len' bytes:

Quoted 7 lines

1) That would write past the end of the buffer.

2) Writing a different value to the ->ki_pos would mean that ->ki_pos is now inaccurate.

Unless someone can explain why the min_t() is needed, we should remove it.

So, back when that code was added, it was actually possible for the tun_do_read to return a value larger then user specified length, but the copy would only be done the length bytes. This was used to signal MSG_TRUNC.

Specifically we had the following code in tun_put_user()

      len = min_t(int, skb->len, len);

      skb_copy_datagram_const_iovec(skb, 0, iv, total, len);
      total += skb->len;

This has since changed to:

      len = min_t(int, skb->len, len);
      ...
      skb_copy_datagram_const_iovec(skb, vlan_offset, iv, total, len);
      total += len;

So, no it seems impossible to return a value larger then user specified length, so MSG_TRUNC will never be set. It probably makes sense to signal message truncation, but that seems broken right now.

-vlad

Quoted 5 lines

--
To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html