Skip to content

W
wr-switch-sw
Switch branch/tag
  1. 24 Aug, 2015 3 commits
    • Alessandro Rubini's avatar
      ppsi: use new master, with some timestamp fixes · 8c3baabe
      Alessandro Rubini authored 
      The bug we were facing was in wr-nic.ko (previous commit fixes it),
but we made a few changes to the better in ppsi handing of timestamps
(wrs-socket.c) while looking for that problem. So this picks those
4 commits of ppsi:
wrs-socket: decrease the timeout when waiting for a timestamp
wrs-socket: simplify the code retrieving timestamps
unix and wrs socket: open in non-blocking mode and report write errors
wrs-socket: use POLLERR for timestamps
Signed-off-by: Alessandro Rubini's avatarAlessandro Rubini <rubini@gnudd.com>
      8c3baabe
    • Alessandro Rubini's avatar
      kernel/wr-nic: fix a memory leakage under heavy load · ce7eb8a2
      Alessandro Rubini authored 
      When the switch is loaded, some PTP outgoing frames get discarded, so we
never get back the timestamp interrupt.  By round-robin'ing the slots,
we get back to the "busy" one, which holds an skb structure that must be
released.

Without this freeing, the skb remains in the output queue of the
socket, and after a while we get EAGAIN in send() because the buffer
is full (400 frames, more or less). And the ptp port is stuck until we
reopen the socket.

This diagnosis exposed the fact that it's not an overflow of pending
descriptors, so I change the message from

  descriptor overflow: tx timestamp pending

to

  discarding tx frame that got no timestamp

Finally, this is pr_warn, not pr_err as we now know it's expected.
Signed-off-by: Alessandro Rubini's avatarAlessandro Rubini <rubini@gnudd.com>
      ce7eb8a2
    • Alessandro Rubini's avatar
      kernel/wr-nic: trivial: fix a possible issue once every 64k timestamps · 5bd07e8e
      Alessandro Rubini authored 
      Signed-off-by: Alessandro Rubini's avatarAlessandro Rubini <rubini@gnudd.com>
      5bd07e8e
  3. 21 Aug, 2015 9 commits
  5. 20 Aug, 2015 28 commits