| Age | Commit message (Collapse) | Author |
|---|
|
For QXL hw we really want the bits to be replaced as we change
the preferred mode on the fly, and the same goes for virgl when
I get to it, however the original fix for this seems to have caused
a wierd regression on Intel G33 that in a stunning display of failure
at opposition to his normal self, Daniel failed to diagnose.
So we are left doing this, ugly ugly ugly ugly, Daniel you fixed
that G33 yet?, ugly, ugly.
Tested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
Fix 4 spelling errors in help sections.
Signed-off-by: Stratos Karafotis <stratosk@semaphore.gr>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
In menu_select function we check for correction factor every time.
If it is zero we are initializing to unity. Hence move it to init function
and initialise by unity, hence avoid repeated comparisons.
Signed-off-by: Chander Kashyap <chander.kashyap@linaro.org>
Reviewed-by: Tuukka Tikkanen <tuukka.tikkanen@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
There has been confusion all the time about which mailing list to follow
for cpufreq activities, linux-pm@vger.kernel.org or cpufreq@vger.kernel.org.
Since patches sent to cpufreq@vger.kernel.org don't go to Patchwork
which is a maintenance workflow problem, make linux-pm@vger.kernel.org
the official mailing list for cpufreq stuff and remove all references
of cpufreq@vger.kernel.org from kernel source.
Later, we can request that the list be dropped entirely.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
[rjw: Changelog]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
git://anongit.freedesktop.org/drm-intel into drm-next
drm-intel-next-2014-04-16:
- vlv infoframe fixes from Jesse
- dsi/mipi fixes from Shobhit
- gen8 pageflip fixes for LRI/SRM from Damien
- cmd parser fixes from Brad Volkin
- some prep patches for CHV, DRRS, ...
- and tons of little things all over
drm-intel-next-2014-04-04:
- cmd parser for gen7 but only in enforcing and not yet granting mode - the
batch copying stuff is still missing. Also performance is a bit ... rough
(Brad Volkin + OACONTROL fix from Ken).
- deprecate UMS harder (i.e. CONFIG_BROKEN)
- interrupt rework from Paulo Zanoni
- runtime PM support for bdw and snb, again from Paulo
- a pile of refactorings from various people all over the place to prep for new
stuff (irq reworks, power domain polish, ...)
drm-intel-next-2014-04-04:
- cmd parser for gen7 but only in enforcing and not yet granting mode - the
batch copying stuff is still missing. Also performance is a bit ... rough
(Brad Volkin + OACONTROL fix from Ken).
- deprecate UMS harder (i.e. CONFIG_BROKEN)
- interrupt rework from Paulo Zanoni
- runtime PM support for bdw and snb, again from Paulo
- a pile of refactorings from various people all over the place to prep for new
stuff (irq reworks, power domain polish, ...)
Conflicts:
drivers/gpu/drm/i915/i915_gem_context.c
|
|
This patch uses dev_err/info function to show accurate log message
with device name instead of pr_err/info function.
Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
There are no users of pci_enable_msi_block() function left. Obsolete it in
favor of pci_enable_msi_range() and pci_enable_msi_exact() functions.
Previously, we called arch_setup_msi_irqs() once, requesting the same
vector count we passed to arch_msi_check_device(). Now we may call it
several times: if it returns failure, we may retry and request fewer
vectors.
We don't keep track of the vector count we initially passed to
arch_msi_check_device(). We only keep track of the number of vectors
successfully set up by arch_setup_msi_irqs(), and this is what we use to
clean things up when disabling MSI. Therefore, we assume that
arch_msi_check_device() does nothing that will have to be cleaned up later.
[bhelgaas: changelog]
Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
|
|
|
|
Sub page write doesn't work because of hw issue in controller found on
Keystone SOCs. AEMIF controller is also used on DaVinci SOCs which
don't seems to have any issue. So add "ti,keysone-nand" compatible
to nand driver in order to set NAND_NO_SUBPAGE_WRITE option.
Cc: Warner Losh <imp@bsdimp.com>
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@ti.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
|
|
If there is a PM QoS latency limit and all of the sufficiently shallow
C-states are disabled, the cpuidle menu governor returns 0 which on
some systems is CPUIDLE_DRIVER_STATE_START and shouldn't be returned
if that C-state has been disabled.
Fix the issue by modifying the menu governor to return (-1) in such
situations.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Since both cpuidle_enabled() and cpuidle_select() are only called by
cpuidle_idle_call(), it is not really useful to keep them separate
and combining them will help to avoid complicating cpuidle_idle_call()
even further if governors are changed to return error codes sometimes.
This code modification shouldn't lead to any functional changes.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio into staging-linus
Jonathan writes:
Second set of fixes for IIO in the 3.15 cycle.
* ad2s1200 - Fix some missing parenthesis in a for statement that could have
led to an error being missed when getting gpios.
* Fix a null derefference issue in the mpu6050 when platform data is not
provided (or is provided via the device tree for example).
* exynos_adc bug on remove due to child devices having been added to the
parent of the IIO device rather than the IIO device itself. This caused an
issue with the IIO device removing itself in it's remove function.
* Make all ADC drivers buildable as modules to avoid dependency issues if
the IIO core is itself built as a module. The exynos adc bug became
apparently whilst this fix was being tested.
|
|
Fix two format string mismatch in ad7280a.c
Signed-off-by: Masanari Iida <standby24x7@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
|
|
Move the devspec OF attribute to PCI common code's set of device attributes
since it's not architecture dependent. As a side effect microblaze and
powerpc no longer need to use pcibios_add_platform_entries().
[bhelgaas: fold in #include for compile error]
Link: https://lkml.kernel.org/r/alpine.LFD.2.11.1404141101500.1529@denkbrett
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
|
|
Add reinit_completion() before the wait_for_completion_timeout in
raw_read() call.
Signed-off-by: Naveen Krishna Chatradhi <ch.naveen@samsung.com>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
|
|
Do a soft reset software if a timeout happens.
This is applicable only for ADC_V2.
Signed-off-by: Naveen Krishna Chatradhi <ch.naveen@samsung.com>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
|
|
ADC module on Exynos5 SoCs runs at 600KSPS. At this conversion rate,
waiting for 1000 msecs is wasteful (incase of h/w failure).
Hence, reduce the time out to 100msecs and use
wait_for_completion_timeout() instead of
wait_for_completion_interruptible_timeout()
Signed-off-by: Naveen Krishna Chatradhi <ch.naveen@samsung.com>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
|
|
This patch maintains the following order in
probe(), remove(), resume() and suspend() calls
regulator enable, clk prepare enable
...
clk disable unprepare, regulator disable
While at it,
1. enable the regulator before the iio_device_register()
2. handle the return values for enable/disable calls
Signed-off-by: Naveen Krishna Ch <ch.naveen@samsung.com>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
|
|
The ACPI PNP subsystem returns errors from pnpacpi_set_resources()
and pnpacpi_disable_resources() if the _SRS or _DIS methods are not
present, respectively, but it should not do that, because those
methods are optional. For this reason, modify pnpacpi_set_resources()
and pnpacpi_disable_resources(), respectively, to ignore missing _SRS
or _DIS.
This problem has been uncovered by commit 202317a573b2 (ACPI / scan:
Add acpi_device objects for all device nodes in the namespace) and
manifested itself by causing serial port suspend to fail on some
systems.
Fixes: 202317a573b2 (ACPI / scan: Add acpi_device objects for all device nodes in the namespace)
References: https://bugzilla.kernel.org/show_bug.cgi?id=74371
Reported-by: wxg4net <wxg4net@gmail.com>
Reported-and-tested-by: <nonproffessional@gmail.com>
Cc: 3.14+ <stable@vger.kernel.org> # 3.14+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Do checksum offload only if the client of the driver wants checksum to be
offloaded.
In V1 version of this patch, I addressed comments from
Stephen Hemminger <stephen@networkplumber.org> and
Eric Dumazet <eric.dumazet@gmail.com>.
In this version of the patch I have addressed comments from
David Miller.
This patch fixes a bug that is exposed in gateway scenarios.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
For phy devices that don't issue interrupts upon link
state changes, phylib polls the link state resulting in
repeated calls to adjust_link(), even if the link state
didn't change. As a result, some mac registers are
repeatedly read and written with the same values, which
is not ok.
To fix this, adjust_link() has been refactored to check
first whether the link state has changed and to take action
only if needed, updating mac registers and local state
variables. The 'new_state' local flag, set if one of the
link params changed (link, speed or duplex), has been
rendered useless and removed by this refactoring.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This reverts commit 12a2856b604476c27d85a5f9a57ae1661fc46019.
The commit above doesn't appear to be necessary any more as the
checksums appear to be correctly computed/validated.
Additionally the above commit breaks kvm configurations where
one VM is using a device that support checksum offload (virtio) and
the other VM does not.
In this case, packets leaving virtio device will have CHECKSUM_PARTIAL
set. The packets is forwarded to a macvtap that has offload features
turned off. Since we use CHECKSUM_UNNECESSARY, the host does does not
update the checksum and thus a bad checksum is passed up to
the guest.
CC: Daniel Lezcano <daniel.lezcano@free.fr>
CC: Patrick McHardy <kaber@trash.net>
CC: Andrian Nord <nightnord@gmail.com>
CC: Eric Dumazet <eric.dumazet@gmail.com>
CC: Michael S. Tsirkin <mst@redhat.com>
CC: Jason Wang <jasowang@redhat.com>
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The following is a problematic configuration:
VM1: virtio-net device connected to macvtap0@eth0
VM2: e1000 device connect to macvtap1@eth0
The problem is is that virtio-net supports checksum offloading
and thus sends the packets to the host with CHECKSUM_PARTIAL set.
On the other hand, e1000 does not support any acceleration.
For small TCP packets (and this includes the 3-way handshake),
e1000 ends up receiving packets that only have a partial checksum
set. This causes TCP to fail checksum validation and to drop
packets. As a result tcp connections can not be established.
Commit 3e4f8b787370978733ca6cae452720a4f0c296b8
macvtap: Perform GSO on forwarding path.
fixes this issue for large packets wthat will end up undergoing GSO.
This commit adds a check for the non-GSO case and attempts to
compute the checksum for partially checksummed packets in the
non-GSO case.
CC: Daniel Lezcano <daniel.lezcano@free.fr>
CC: Patrick McHardy <kaber@trash.net>
CC: Andrian Nord <nightnord@gmail.com>
CC: Eric Dumazet <eric.dumazet@gmail.com>
CC: Michael S. Tsirkin <mst@redhat.com>
CC: Jason Wang <jasowang@redhat.com>
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This patch adds set_rx_int_on_com function for interrupt when
dma is completed.
Signed-off-by: Byungho An <bh74.an@samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This patch adds rxqueue enable function according to number of rxqueue
and adds rxqueue disable function for removing.
Signed-off-by: Byungho An <bh74.an@samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This patch moves sw reset to probe function because
sw reset is needed early stage before open function.
Signed-off-by: Byungho An <bh74.an@samsung.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add handling for " 8 GT/s" in print_port_info().
Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This fixes the following sparse warnings:
drivers/net/ethernet/altera/altera_msgdma.c:23:5: warning: symbol 'msgdma_initialize' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:28:6: warning: symbol 'msgdma_uninitialize' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:32:6: warning: symbol 'msgdma_reset' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:77:6: warning: symbol 'msgdma_disable_rxirq' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:83:6: warning: symbol 'msgdma_enable_rxirq' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:89:6: warning: symbol 'msgdma_disable_txirq' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:95:6: warning: symbol 'msgdma_enable_txirq' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:101:6: warning: symbol 'msgdma_clear_rxirq' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:107:6: warning: symbol 'msgdma_clear_txirq' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:114:5: warning: symbol 'msgdma_tx_buffer' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:129:5: warning: symbol 'msgdma_tx_completions' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:154:5: warning: symbol 'msgdma_add_rx_desc' was not declared. Should it be static?
drivers/net/ethernet/altera/altera_msgdma.c:181:5: warning: symbol 'msgdma_rx_status' was not declared. Should it be static?
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Acked-by: Vince Bridgers <vbridgers2013@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
sgdma_descrip is a function name as well as the name of a struct. In
sgdma_initialize(), we should initialize the descriptor length field
with the actual length of a descriptor not with the size of the
function. In order to prevent such things from happening in the future,
rename the function to sgdma_setup_descrip().
Found by sparse which yields the following warning:
drivers/net/ethernet/altera/altera_sgdma.c:74:30: warning: expression using sizeof on a function
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
If there are no peer_devices or connections, I'd rather have NULL
than some "arbitrary" address pretending to point to a struct.
Helps to avoid hard to debug symptoms, in case we ever try to use
and dereference a drbd_connection or drbd_peer_device
where we in fact don't have any connection at all.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
A newly created device was never exposed before, i.e. has a
exposed_data_uuid of 0. Then it is valid to attach to any current_uuid
of a backing device (of course also to a newly created one (4))
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
In case a connection transitions into C_TIMEOUT within the timer
function (request_timer_fn()) we need to make sure that the receiver
thread (potentially running on a different CPU) sees the updated
cstate later on.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
...instead directly assigning to q->limits.discard_zeroes_data
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
Just because it is the oldest not yet completed request
does not make it the oldest request waiting for disk.
Or waiting for the peer.
And we completely missed already completed requests
that would still hold references to activity log extents,
waiting only for the barrier ack.
Find two oldest not yet completely processed requests,
one that is still waiting for local completion,
and one that is still waiting for some response from the peer.
These may or may not be the same request object.
Then separately apply the network and disk timeouts, respectively.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
In the implementation as it was, the two peers sent each other
a challenge, and expects the challenge hashed with the shared
secret back.
A attacker could simply wait for the challenge of the peer, and
send the same challenge back. Then it waits for the response, and
sends the same response back.
Prevent this by not accepting a challenge from the peer that is
the same as the challenge sent to the peer.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
Once our sender thread needs to wait_for_work(),
and actually needs to schedule(), just before we do that,
we already check if it is useful to implicitly close the last epoch.
The condition was too strict: only implicitly close the epoch,
if there have been no new (write) requests at all.
The assumption was that if there were new requests, they would
always be communicated one way or another, and would send necessary
epoch separating barriers explicitly.
This is not always true, e.g. when becoming diskless,
or while explicitly starting a full resync.
The last communicated epoch could stay open for a long time,
locking down corresponding activity log extents.
It is safe to always implicitly send that last barrier, as soon as we
determin that there cannot be more requests in the last communicated
epoch, even if there have been (uncommunicated) new requests in new
epochs meanwhile.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
When batching more updates to the activity log into single transactions,
we lost the ability for new requests to force themselves into the active
set: all preparation steps became non-blocking, and if all currently
hot extents keep busy, they could starve out new incoming requests
to cold extents for quite a while.
This can only happen if your IO backend accepts more IO operations per
average DRBD replication round trip time than you have al-extents
configured.
If we have incoming requests to cold extents,
at least do one blocking update per transaction.
In an artificial worst-case workload on SSD with an asynchronous 600 ms
replication link, with al-extents = 7 (the minimum we allow), and
concurrent full resynch, without this patch, some write requests have
been observed to be starved for 40 seconds.
With this patch, application observed a worst case latency of twice the
replication round trip time.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
We want to store in persistent meta data what the peer DRBD can handle,
which, due to spreading requests to multiple bios,
may be more than its backing device can handle.
Otherwise, if a disconnected Primary temporarily loses access to its local data
as well, we may accidentally shrink the max-bio setting, portentially causing
already assembled, but not yet processed, application bios to be spuriously
failed due to device limits.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
In the drbd make request function, specifically in
drbd_send_and_submit(), we decide whether we want to send the actual
write request, or only a "set this block out of sync" information.
We do so based on the current connection state, while holding the req_lock.
The connection state is not supposed to change while holding the req_lock.
But in drbd_start_resync, we did change that state anyways,
while only holding the global_state_lock, which is enough to change
sync-after dependencies (paused vs active resync), but
not good enough to change the connection state.
Fix: in drbd_start_resync, first grab the req_lock to serialize with
drbd_send_and_submit(), before grabbing the global_state_lock
to be able to evaluate the sync-after dependencies.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
Allow the user of REQ_DISCARD.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
Note that I do NOT call __drbd_chk_io_error for failed REQ_DISCARD.
That may be wrong, though, or needs to differ between EOPNOTSUPP and
other errors...
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
If the receiver needs to serve a discard request on a queue that does
not announce to be discard cabable, it falls back to do synchronous
blkdev_issue_zeroout().
We expect only "reasonably" large (up to one activity log extent?)
discard requests.
We do this to not to not block the receiver for too long in this
fallback code path, and to not set/clear too many bits inside one
spinlock_irq_save() in drbd_set_in_sync/drbd_set_out_of_sync,
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
We plan to use genl_family->parallel_ops = true in the future,
but need to review all possible interactions first.
For now, only selectively drop genl_lock() in drbd_set_role(),
instead serializing on our own internal resource->conf_update mutex.
We now can be promoted/demoted on many resources in parallel,
which may significantly improve cluster failover times
when fencing is required.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
Because all administrative requests via genetlink have been globally
serialized via genl_lock(), we used to have one static struct
drbd_config_context "admin context".
Move this on-stack to the respective callback functions.
This will allow us to selectively drop the genl_lock()
(or use genl_family->parallel_ops) in the future.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
When a 'cluster wide' disconnect executes, the result comes back
from the peer, and immediately after that the connection breaks
then _conn_rq_cond() reported back SS_CW_SUCCESS.
Therefore _conn_request_state() calls conn_set_state(), which
has a BUG() in it.
The BUG() is hit because conn_is_valid_transition() does not like
the transaction. Which goes back to is_valid_soft_transition()
returning SS_OUTDATE_WO_CONN.
This fix is to consider an error reported by is_valid_soft_transition()
even when the peer agreed to the transaction.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
Before, application IO could pre-empt resync activity
for up to hardcoded 20 seconds per resync request.
A very busy server could throttle the effective resync bandwidth
down to one request per 20 seconds.
Now, we only let application IO pre-empt resync traffic
while the current resync rate estimate is above c-min-rate.
If you disable the c-min-rate throttle feature (set c-min-rate = 0),
application IO will no longer pre-empt resync traffic at all.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
If max-buffers and socket buffer sizes are "too small" for the chosen
resync rate, this could lead potentially lead to a distributed deadlock,
which may or may not resolve itself via the "ko-count" and request
timeout mechanism, or could be resolved by forced disconnect.
One option to deal with this is proper configuration:
use larger max-buffer and socket buffers settings,
or reduce the resync rate.
But even with bad configuration we should not deadlock,
but "gracefully" recover.
The issue is avoided by using only up to max-buffers/2 for resync
requests, and by using max-buffers not as a hard limit for data buffer
allocations, but as a throttle threshold only.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
While merging adjacent dirty blocks into resync requests,
the resync rate throttle was disregarded.
For very low resync rates, the effective rate may have exceeded
the intended rate by a larger margin.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
If we don't make resync or verify progress for "too long",
we want to flag it as "stalled".
Since 2010, "use rolling marks for resync speed calculation"
this "too long" was wrong by a factor of HZ.
With HZ 250, it would have been flagged as stalled
after 100 minutes.
Hardcode 3 minutes instead.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
|
If a user forces the operation he takes the blame in case
the peer does not have enough space. No reason to dey this...
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
|
