During device mode initialization, lots of device information
are printed to console, see below. Change them as debug message.
cdns-usb3 5b130000.cdns3: Initialized ep0 support:
cdns-usb3 5b130000.cdns3: Initialized ep1out support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep2out support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep3out support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep4out support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep5out support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep6out support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep7out support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep1in support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep2in support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep3in support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep4in support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep5in support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep6in support: BULK, INT ISO
cdns-usb3 5b130000.cdns3: Initialized ep7in support: BULK, INT ISO
Reviewed-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
Signed-off-by: Felipe Balbi <balbi@kernel.org>
And delete cdsn3_hw_role_state_machine declare which doesn't
be needed.
Reviewed-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
Signed-off-by: Felipe Balbi <balbi@kernel.org>
In short, we have three kinds of role switches:
- Based on SoC: ID and VBUS
- Based on external connnctor, eg, Type-C or GPIO Connector
- Based on user choices through sysfs
Since HW handling and usb-role-switch handling are at
different places, we do not need role_override any more,
and this flag could not judge external connector case well.
With role_override deleted, We use cdns3_hw_role_switch for
the 1st use case, and usb-role-switch for the 2nd and 3rd cases.
Reviewed-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
Signed-off-by: Felipe Balbi <balbi@kernel.org>
After that, the role switch device (eg, Type-C device) could call
cdns3_role_set to finish the role switch.
Reviewed-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
Signed-off-by: Felipe Balbi <balbi@kernel.org>
If there are still pending requests because no TRB was available,
prepare more when started requests are completed.
Introduce dwc3_gadget_ep_should_continue() to check for incomplete and
pending requests to resume updating new TRBs to the controller's TRB
cache.
Signed-off-by: Thinh Nguyen <thinhn@synopsys.com>
Signed-off-by: Felipe Balbi <balbi@kernel.org>
Register vbus_draw to gadget ops and update corresponding vbus
draw current to usb_phy.
Signed-off-by: Nagarjuna Kristam <nkristam@nvidia.com>
Signed-off-by: Felipe Balbi <balbi@kernel.org>
Terratec Cinergy S2 PCIe Dual is a PCIe device with two tuners that
actually contains two USB devices. The devices are visible in the
lsusb printout.
Bus 004 Device 002: ID 153b:1182 TerraTec Electronic GmbH Cinergy S2 PCIe Dual Port 2
Bus 003 Device 002: ID 153b:1181 TerraTec Electronic GmbH Cinergy S2 PCIe Dual Port 1
The devices use the Montage M88DS3000/M88TS2022 demod/tuner.
Signed-off-by: Dirk Nehring <dnehring@gmx.net>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
When the MFD driver is a loadable module, the watchdog driver fails
to get linked into the kernel:
drivers/watchdog/iTCO_wdt.o: In function `update_no_reboot_bit_pmc':
iTCO_wdt.c:(.text+0x54f): undefined reference to `intel_pmc_gcr_update'
The code is written to support operation without the MFD driver, so
add a Kconfig dependency that allows this, while disallowing the watchdog
to be built-in when the MFD driver is a module.
Fixes: 25f1ca31e2 ("platform/x86: intel_pmc_ipc: Convert to MFD")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20200428212959.2993304-1-arnd@arndb.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
There is actually no need to ping the watchdog before disabling it
during timeout change. Disabling the watchdog already takes care of
resetting the counter.
This fixes an issue during boot when the userspace watchdog handler takes
over and the watchdog is already running. Opening the watchdog in this case
leads to the first ping and directly after that without the required
heartbeat delay a second ping issued by the set_timeout call. Due to the
missing delay this resulted in a reset.
Signed-off-by: Stefan Riedmueller <s.riedmueller@phytec.de>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Link: https://lore.kernel.org/r/20200403130728.39260-3-s.riedmueller@phytec.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
During probe try to set the timeout from device tree and fall back to
either the pre-configured timeout set by e.g. the bootloader in case the
watchdog is already running or the default value.
If the watchdog is already running make sure to update the timeout and
tell the framework about the running state to make sure the watchdog is
handled correctly until user space takes over. Updating the timeout also
removes the need for an additional manual ping so we can remove that as
well.
Signed-off-by: Stefan Riedmueller <s.riedmueller@phytec.de>
Reviewed-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20200403130728.39260-1-s.riedmueller@phytec.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
Oded writes:
This tag contains the following changes for kernel 5.8:
- Improve MMU cache invalidation code and handle case where the
invalidation doesn't finish in a reasonable time.
- Remove the option to perform soft-reset to GAUDI. Soft-reset is where the
driver only resets the compute and DMA engines of the ASIC. This is not
relevant to GAUDI as we must also reset the NIC ports. And when we reset
the NIC ports, we must also reset other stuff so we prefer to just do
hard-reset (where we reset the entire ASIC except for PCIe).
- Fail the hard-reset procedure in case we still have user processes which
have active file-descriptors on a device. Doing hard-reset in that case
can result in a kernel panic because of gen_pool checks
- Don't initialize the default wait callback of dma_buf with the default
wait function as that's the default...
* tag 'misc-habanalabs-next-2020-05-25' of git://people.freedesktop.org/~gabbayo/linux:
habanalabs: handle MMU cache invalidation timeout
habanalabs: don't allow hard reset with open processes
habanalabs: GAUDI does not support soft-reset
habanalabs: add print for soft reset due to event
habanalabs: improve MMU cache invalidation code
habanalabs: don't set default fence_ops->wait
MMU cache invalidation timeout indicates that the device is unstable and
therefore unusable.
Hence in such case do hard reset and return an error to the user if was
called from ioctl.
In addition, change the print to error level and rephrase its text.
Signed-off-by: Omer Shpigelman <oshpigelman@habana.ai>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
When the MMU is heavily used by the engines, unmapping might take a lot of
time due to a full MMU cache invalidation done as part of the unmap flow.
Hence we might not be able to kill all open processes before going to hard
reset the device, as it involves unmapping of all user memory.
In case of a failure in killing all open processes, we should stop the
hard reset flow as it might lead to a kernel crash - one thread (killing
of a process) is updating MMU structures that other thread (hard reset) is
freeing.
Stopping a hard reset flow leaves the device as nonoperational and the
user can then initiate a hard reset via sysfs to reinitialize the device.
Signed-off-by: Omer Shpigelman <oshpigelman@habana.ai>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
GAUDI does not support soft-reset as it leaves the NIC ports in an awkward
state, where their QMANs were reset but the NIC itself is still working.
In addition, there is not much sense in doing soft-reset when training is
done on multiple GAUDIs.
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Tomer Tayar <ttayar@habana.ai>
It's the default.
Also so much for "we're not going to tell the graphics people how to
review their code", dma_fence is a pretty core piece of gpu driver
infrastructure. And it's very much uapi relevant, including piles of
corresponding userspace protocols and libraries for how to pass these
around.
Would be great if habanalabs would not use this (from a quick look
it's not needed at all), since open source the userspace and playing
by the usual rules isn't on the table. If that's not possible (because
it's actually using the uapi part of dma_fence to interact with gpu
drivers) then we have exactly what everyone promised we'd want to
avoid.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
The previous commit:
c6e7bd7afa: ("sched/core: Optimize ttwu() spinning on p->on_cpu")
avoids spinning on p->on_rq when the task is descheduling, but only if the
wakee is on a CPU that does not share cache with the waker.
This patch offloads the activation of the wakee to the CPU that is about to
go idle if the task is the only one on the runqueue. This potentially allows
the waker task to continue making progress when the wakeup is not strictly
synchronous.
This is very obvious with netperf UDP_STREAM running on localhost. The
waker is sending packets as quickly as possible without waiting for any
reply. It frequently wakes the server for the processing of packets and
when netserver is using local memory, it quickly completes the processing
and goes back to idle. The waker often observes that netserver is on_rq
and spins excessively leading to a drop in throughput.
This is a comparison of 5.7-rc6 against "sched: Optimize ttwu() spinning
on p->on_cpu" and against this patch labeled vanilla, optttwu-v1r1 and
localwakelist-v1r2 respectively.
5.7.0-rc6 5.7.0-rc6 5.7.0-rc6
vanilla optttwu-v1r1 localwakelist-v1r2
Hmean send-64 251.49 ( 0.00%) 258.05 * 2.61%* 305.59 * 21.51%*
Hmean send-128 497.86 ( 0.00%) 519.89 * 4.43%* 600.25 * 20.57%*
Hmean send-256 944.90 ( 0.00%) 997.45 * 5.56%* 1140.19 * 20.67%*
Hmean send-1024 3779.03 ( 0.00%) 3859.18 * 2.12%* 4518.19 * 19.56%*
Hmean send-2048 7030.81 ( 0.00%) 7315.99 * 4.06%* 8683.01 * 23.50%*
Hmean send-3312 10847.44 ( 0.00%) 11149.43 * 2.78%* 12896.71 * 18.89%*
Hmean send-4096 13436.19 ( 0.00%) 13614.09 ( 1.32%) 15041.09 * 11.94%*
Hmean send-8192 22624.49 ( 0.00%) 23265.32 * 2.83%* 24534.96 * 8.44%*
Hmean send-16384 34441.87 ( 0.00%) 36457.15 * 5.85%* 35986.21 * 4.48%*
Note that this benefit is not universal to all wakeups, it only applies
to the case where the waker often spins on p->on_rq.
The impact can be seen from a "perf sched latency" report generated from
a single iteration of one packet size:
-----------------------------------------------------------------------------------------------------------------
Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | Maximum delay at |
-----------------------------------------------------------------------------------------------------------------
vanilla
netperf:4337 | 21709.193 ms | 2932 | avg: 0.002 ms | max: 0.041 ms | max at: 112.154512 s
netserver:4338 | 14629.459 ms | 5146990 | avg: 0.001 ms | max: 1615.864 ms | max at: 140.134496 s
localwakelist-v1r2
netperf:4339 | 29789.717 ms | 2460 | avg: 0.002 ms | max: 0.059 ms | max at: 138.205389 s
netserver:4340 | 18858.767 ms | 7279005 | avg: 0.001 ms | max: 0.362 ms | max at: 135.709683 s
-----------------------------------------------------------------------------------------------------------------
Note that the average wakeup delay is quite small on both the vanilla
kernel and with the two patches applied. However, there are significant
outliers with the vanilla kernel with the maximum one measured as 1615
milliseconds with a vanilla kernel but never worse than 0.362 ms with
both patches applied and a much higher rate of context switching.
Similarly a separate profile of cycles showed that 2.83% of all cycles
were spent in try_to_wake_up() with almost half of the cycles spent
on spinning on p->on_rq. With the two patches, the percentage of cycles
spent in try_to_wake_up() drops to 1.13%
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Jirka Hladky <jhladky@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: valentin.schneider@arm.com
Cc: Hillf Danton <hdanton@sina.com>
Cc: Rik van Riel <riel@surriel.com>
Link: https://lore.kernel.org/r/20200524202956.27665-3-mgorman@techsingularity.net
Shutdown test is somtimes hung, since it keeps trying to flush dirty node pages
in an inifinite loop. Let's drop dirty pages at umount in that case.
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Ido Schimmel says:
====================
mlxsw: Various trap changes - part 1
This patch set contains various changes in mlxsw trap configuration.
Another set will perform similar changes before exposing control traps
(e.g., IGMP query, ARP request) via devlink-trap.
Tested with existing devlink-trap selftests. Please see individual
patches for a detailed changelog.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The rate with which packets are sampled is determined by user space, so
there is no need to associate such packets with a policer.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Both packet types are needed for the same reason (neighbour discovery),
so associate them with the same trap group.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ARP trap group will be used for IPv6 ND traps in the next patch, so
rename it to "NEIGH_DISCOVERY" which is more appropriate.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now that traffic class (TC) and priority are set to the same value,
there is no need to store both. Remove the first.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The traffic class (TC) attribute of packet traps determines through which
TC a packet trap will be scheduled through the CPU port.
The priority attribute determines which trap will be triggered in case
several packet traps match a packet.
We try to configure these attributes to the same value for all packet
traps as there is little reason not to.
Some packet traps did not use the same value, so rectify that now.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As explained in commit 9ffcc3725f ("mlxsw: spectrum: Allow packets to
be trapped from any PG"), incoming packets can be admitted to the shared
buffer and forwarded / trapped, if:
(Ingress{Port}.Usage < Thres && Ingress{Port,PG}.Usage < Thres &&
Egress{Port}.Usage < Thres && Egress{Port,TC}.Usage < Thres)
||
(Ingress{Port}.Usage < Min || Ingress{Port,PG} < Min ||
Egress{Port}.Usage < Min || Egress{Port,TC}.Usage < Min)
Trapped packets are scheduled to transmission through the CPU port.
Currently, the minimum and maximum quotas of traffic class (TC) 0 of the
CPU port are 0, which means it is not usable.
Assign non-zero quotas to TC 0 of the CPU port, so that it could be
utilized by subsequent patches.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Reduce the default acceptable rate of DHCP packets to 128 packets per
second and reduce their priority. This is reasonable given the Spectrum
ASICs are limited to 128 ports at the moment.
These are only the default values. Users will be able to modify them via
devlink-trap.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, IPv4 DHCP packets are trapped during L2 forwarding, which
means that packets might be trapped unnecessarily. Instead, only trap
the DHCP packets that reach the router. Either because they were flooded
to the router port or forwarded to it by the FDB. This is consistent
with the corresponding IPv6 trap.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Both packet types are needed for the same reason (multicast snooping),
so associate them with the same trap group.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The IGMP trap group will be used for MLD traps in the next patch, so
rename it to "MC_SNOOPING" which is more appropriate.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
