Re: [PATCH 4.20 14/57] CIFS: Do not hide EINTR after sending network packets

2019-01-15 Thread Greg Kroah-Hartman
On Tue, Jan 15, 2019 at 07:22:38PM +, Pavel Shilovskiy wrote:
> Hi Greg,
> 
> I am wondering if it is possible to include exact stable kernel
> version (e.g. 4.20.3 in this case) in the email. This would help to
> quickly understand which kernel version should be installed in order
> to get the fix.

Given that this is not in a released kernel yet, it's hard to give you a
final version :)

We have had times when a stable release happens while a -rc is out for
review due to special circumstances, so you couldn't always rely on my
email for this, as well as the fact that sometimes we drop patches from
-rc releases before they are in a real release, so such an email would
lie and confuse people.

Just watch the kernel releases, and use git to see what commit is
in what release, it's not that difficult.  Especially as you are going
to want to track this across multiple stable releases, so any guess I
might provide here isn't going to help you out much.

greg k-h


RE: [PATCH 4.20 14/57] CIFS: Do not hide EINTR after sending network packets

2019-01-15 Thread Pavel Shilovskiy
Hi Greg,

I am wondering if it is possible to include exact stable kernel version (e.g. 
4.20.3 in this case) in the email. This would help to quickly understand which 
kernel version should be installed in order to get the fix.

Best regards,
Pavel Shilovsky

-Original Message-
From: Greg Kroah-Hartman  
Sent: Tuesday, January 15, 2019 8:36 AM
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman ; sta...@vger.kernel.org; 
Pavel Shilovskiy ; Jeff Layton ; 
Steven French 
Subject: [PATCH 4.20 14/57] CIFS: Do not hide EINTR after sending network 
packets

4.20-stable review patch.  If anyone has any objections, please let me know.

--

From: Pavel Shilovsky 

commit ee13919c2e8d1f904e035ad4b4239029a8994131 upstream.

Currently we hide EINTR code returned from sock_sendmsg() and return 0 instead. 
This makes a caller think that we successfully completed the network operation 
which is not true. Fix this by properly returning EINTR to callers.

Cc: 
Signed-off-by: Pavel Shilovsky 
Reviewed-by: Jeff Layton 
Signed-off-by: Steve French 
Signed-off-by: Greg Kroah-Hartman 

---
 fs/cifs/transport.c |2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/fs/cifs/transport.c
+++ b/fs/cifs/transport.c
@@ -385,7 +385,7 @@ smbd_done:
if (rc < 0 && rc != -EINTR)
cifs_dbg(VFS, "Error %d sending data on socket to server\n",
 rc);
-   else
+   else if (rc > 0)
rc = 0;
 
return rc;




[PATCH 4.20 14/57] CIFS: Do not hide EINTR after sending network packets

2019-01-15 Thread Greg Kroah-Hartman
4.20-stable review patch.  If anyone has any objections, please let me know.

--

From: Pavel Shilovsky 

commit ee13919c2e8d1f904e035ad4b4239029a8994131 upstream.

Currently we hide EINTR code returned from sock_sendmsg()
and return 0 instead. This makes a caller think that we
successfully completed the network operation which is not
true. Fix this by properly returning EINTR to callers.

Cc: 
Signed-off-by: Pavel Shilovsky 
Reviewed-by: Jeff Layton 
Signed-off-by: Steve French 
Signed-off-by: Greg Kroah-Hartman 

---
 fs/cifs/transport.c |2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/fs/cifs/transport.c
+++ b/fs/cifs/transport.c
@@ -385,7 +385,7 @@ smbd_done:
if (rc < 0 && rc != -EINTR)
cifs_dbg(VFS, "Error %d sending data on socket to server\n",
 rc);
-   else
+   else if (rc > 0)
rc = 0;
 
return rc;