Switch to a single tunnel for all mappings, this removes the limitations
on how many mappings each tunnel can handle, and therefore how many Fan
slices each local address may hold.
BugLink: http://bugs.launchpad.net/bugs/1470091
Signed-off-by: Jay Vosburgh <email address hidden>
Signed-off-by: Andy Whitcroft <email address hidden>
Signed-off-by: Tim Gardner <email address hidden>
UBUNTU: SAUCE: fan: Proof of concept implementation (v2)
Modification to ipip tunnel driver to accept a new netlink option,
IFLA_IPTUN_FAN_UNDERLAY, which provides a /16 network prefix and enables
TX side destination address remapping for traffic entering the tunnel
(to be encapsulated).
For an overlay (inner) address Y.A.B.C, the transformation is F.G.A.B,
where "F" and "G" are the first two octets of the underlay network (the
network portion of a /16), "A" and "B" are the low order two octets of the
underlay network host (the host portion of a /16), and "Y" is a configured
first octet of the overlay network.
E.g., underlay host 10.88.3.4 with an overlay of 99 would host overlay
subnet 99.3.4.0/24. An overlay network datagram from 99.3.4.5 to 99.6.7.8
would be directed to underlay host 10.88.6.7, which hosts overlay network
99.6.7.0/24.
Includes net.fan.version sysctl as a sentinel for availability of the
fan functionality.
NOTE: this requires an updated iproute2 to facilitate configuration of
the fan.
BugLink: http://bugs.launchpad.net/bugs/1439706
Signed-off-by: Jay Vosburgh <email address hidden>
[<email address hidden>: move IFLA_IPTUN_FAN_UNDERLAY up to avoid clashing
with future feature additions.]
Signed-off-by: Andy Whitcroft <email address hidden>
multipath-tools 0.5.0 selects service-time as the default algorithm instead of
round-robin. Add both modules into the multipath-modules udeb to smoothly handle
the transition.
Signed-off-by: Chris J Arges <email address hidden>
Acked-by: Stefan Bader <email address hidden>
Acked-by: Andy Whitcroft <email address hidden>
Signed-off-by: Andy Whitcroft <email address hidden>
Signed-off-by: Tim Gardner <email address hidden>
Users of non-extent ext4 filesystems (ext4 ^extents, or ext3 w/
CONFIG_EXT4_USE_FOR_EXT23=y) can encounter data corruption when using fallocate
with FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE flags.
This patch returns EOPNOTSUPP when ext4_punch_hole is called from a non-extents
filesystem.
Signed-off-by: Chris J Arges <email address hidden>
Acked-by: Stefan Bader <email address hidden>
Signed-off-by: Andy Whitcroft <email address hidden>
rtsx_usb_ms creates a task that mostly sleeps, but tasks in
uninterruptible sleep still contribute to the load average (for
bug-compatibility with Unix). A load average of ~1 on a system that
should be idle is somewhat alarming.
Change the sleep to be interruptible, but still ignore signals.
A better fix might be to replace this loop with a delayed work item.
References: https://bugs.debian.org/765717
Signed-off-by: Ben Hutchings <email address hidden>
Signed-off-by: Chris J Arges <email address hidden>
Acked-by: Andy Whitcroft <email address hidden>
Signed-off-by: Leann Ogasawara <email address hidden>
Signed-off-by: Andy Whitcroft <email address hidden>
This patch fixes a RTC wakealarm issue, namely, the event fires during
hibernate and is not cleared from the list, causing hwlock to block.
The original fix has been sent to LKML twice [1] but has had no response
and not yet been picked up. Rather than wait for ever for the maintainter
to pick it up, I'm sending it as a fix for Vivid.
Note that the skipped expired timer events get reaped later on, so we
don't bother reaping them in this fix, instead ignoring them is the best
strategy.
The current enqueuing does not trigger an alarm if any expired timers
already exist on the timerqueue. This can occur when a RTC wake alarm
is used to wake a machine out of hibernate and the resumed state has
old expired timers that have not been removed from the timer queue.
This fix skips over any expired timers and triggers an alarm if there
are no pending timers on the timerqueue.
The bug was found running the example RTC timer program from
Documentation/rtc.txt; it runs fine before a hibernate but will block
forever on RTC reads after a resume from a hibernate that is woken
up using a RTC wakealarm.
Signed-off-by: Colin Ian King <email address hidden>
Acked-by: Andy Whitcroft <email address hidden>
Acked-by: Chris J Arges <email address hidden>
Signed-off-by: Leann Ogasawara <email address hidden>