| Age | Commit message (Collapse) | Author |
|---|
|
PHTW register is selected based on default bit rate from Table[1].
for the bit rates less than or equal to 250. Currently first
value of default bit rate which is greater than or equal to
the caculated mbps is selected. This selection can be further
improved by selecting the default bit rate which is nearest to
the calculated value.
[1] specs r19uh0105ej0200-r-car-3rd-generation.pdf [Table 25.12]
Fixes: 769afd212b16 ("media: rcar-csi2: add Renesas R-Car MIPI CSI-2 receiver driver")
Signed-off-by: Suresh Udipi <sudipi@jp.adit-jv.com>
Signed-off-by: Michael Rodin <mrodin@de.adit-jv.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Add a warning message when the selected PHY speed is less
than supported minimum PHY speed given in the hsfreq table[1].
For raspberry pi camera capture on Kingfisher board with resolution
640x480, the calculated PHY speed is 48 mbps which is less than
the minimum PHY speed 80 Mbps from the table[1]. But in this cases
capture is successful.
[1] specs r19uh0105ej0200-r-car-3rd-generation.pdf [Table 25.9]
Signed-off-by: Suresh Udipi <sudipi@jp.adit-jv.com>
Signed-off-by: Michael Rodin <mrodin@de.adit-jv.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
hsfreqrange should be chosen based on the calculated mbps which
is closer to the default bit rate and within the range as per
table[1]. But current calculation always selects first value which
is greater than or equal to the calculated mbps which may lead
to chosing a wrong range in some cases.
For example for 360 mbps for H3/M3N
Existing logic selects
Calculated value 360Mbps : Default 400Mbps Range [368.125 -433.125 mbps]
This hsfreqrange is out of range.
The logic is changed to get the default value which is closest to the
calculated value [1]
Calculated value 360Mbps : Default 350Mbps Range [320.625 -380.625 mpbs]
[1] specs r19uh0105ej0200-r-car-3rd-generation.pdf [Table 25.9]
Please note that According to Renesas in Table 25.9 the range for
220 default value is corrected as below
|Range (Mbps) | Default Bit rate (Mbps) |
-----------------------------------------------
| 197.125-244.125 | 220 |
-----------------------------------------------
Fixes: 769afd212b16 ("media: rcar-csi2: add Renesas R-Car MIPI CSI-2 receiver driver")
Signed-off-by: Suresh Udipi <sudipi@jp.adit-jv.com>
Signed-off-by: Kazuyoshi Akiyama <akiyama@nds-osk.co.jp>
Signed-off-by: Michael Rodin <mrodin@de.adit-jv.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Convert the Renesas JPEG Processing Unit Device Tree binding
documentation to json-schema.
Document missing properties.
Update the example to match reality.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Mikhail Ulyanov <mikhail.ulyanov@cogentembedded.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Implement fwnode parsing at probe time.
Check if the bus and number of lanes are supported.
Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Simplify probe function by adding a local variable dev instead of using
&client->dev all the time.
Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com>
Reviewed-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
ACPI _HID INT347A represents the OV8865 sensor, the driver for which can
support the platforms that the cio2-bridge serves. Add it to the array
of supported sensors so the bridge will connect the sensor to the CIO2
device.
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The maximum gain figure in the v4l2 ctrl is wrong. The field is 12 bits
wide, which is where the 8191 figure comes from, but the datasheet is
specific that maximum gain is 16x (the low seven bits are fractional, so
16x gain is 2048)
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
There is a chance that regulator_get() returns -EPROBE_DEFER, in which
case printing an error message is undesirable. To avoid spurious messages
in dmesg in the event that -EPROBE_DEFER is returned, use dev_err_probe()
on error paths for regulator_get().
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The ov8865 driver currently has the unit of the V4L2_CID_EXPOSURE control
as 1/16th of a line. This is what the sensor expects, but isn't very
intuitive. Switch the control to be in units of a line and simply do the
16x multiplication before passing the value to the sensor.
The datasheet for this sensor gives minimum exposure as 2 lines, so take
the opportunity to correct the lower bounds of the control.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Add V4L2_CID_CAMERA_ORIENTATION and V4L2_CID_CAMERA_SENSOR_ROTATION
controls to the ov8865 driver by attempting to parse them from firmware.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Exposure limits depend on the total height; when vblank is altered (and
thus the total height is altered), change the exposure limits to reflect
the new cap.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The HTS values for some of the modes in the ov8865 driver are a bit
unusual, coming in lower than the output_size_x is set to. It seems
like they might be calculated to fit the desired framerate into a
configuration with just two data lanes. To bring this more in line
with expected behaviour, raise the HTS values above the output_size_x.
The corollary of that change is that the hardcoded frame intervals
against the modes no longer make sense, so remove those entirely.
Update the .g/s_frame_interval() callbacks to calculate the frame
interval based on the current mode and the vblank and hblank settings.
The implementation of the .enum_frame_interval() callback is no longer
suitable since the possible frame rate is now a continuous range depending
on the vblank control setting, so remove that callback entirely.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Add a V4L2_CID_HBLANK control to the ov8865 driver. This is read only
with timing control intended to be done via vblanking alone.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Add a V4L2_CID_VBLANK control to the ov8865 driver.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The V4L2_CID_GAIN control for this driver configures registers that
the datasheet specifies as analogue gain. Switch the control's ID
to V4L2_CID_ANALOGUE_GAIN.
Reviewed-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The ov8865 driver's v4l2_subdev_pad_ops currently does not include
.get_selection() - add support for that callback.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The ov8865 driver as written expects a 24MHz input clock, but the sensor
is sometimes found on x86 platforms with a 19.2MHz input clock supplied.
Add a set of PLL configurations to the driver to support that rate too.
As ACPI doesn't auto-configure the clock rate, check for a clock-frequency
during probe and set that rate if one is found.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The ov8865 driver is one of those that can be connected to a CIO2
device by the cio2-bridge code. This means that the absence of an
endpoint for this device is not necessarily fatal, as one might be
built by the cio2-bridge when it probes. Return -EPROBE_DEFER if no
endpoint is found rather than a fatal error.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The PLL configuration defined here sets 72MHz (which is correct), not
80MHz. Correct the comment.
Reviewed-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The ov8865 sensor is sometimes found on x86 platforms enumerated via ACPI.
Add an ACPI match table to the driver so that it's probed on those
platforms.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
ov8865_state_init() calls ov8865_state_mipi_configure() which uses
__v4l2_ctrl_s_ctrl[_int64](). This means that sensor->mutex (which
is also sensor->ctrls.handler.lock) must be locked before calling
ov8865_state_init().
Note ov8865_state_mipi_configure() is also used in other places where
the lock is already held so it cannot be changed itself.
This fixes the following lockdep kernel WARN:
[ 13.233421] WARNING: CPU: 0 PID: 8 at drivers/media/v4l2-core/v4l2-ctrls-api.c:833 __v4l2_ctrl_s_ctrl+0x4d/0x60 [videodev]
...
[ 13.234063] Call Trace:
[ 13.234074] ov8865_state_configure+0x98b/0xc00 [ov8865]
[ 13.234095] ov8865_probe+0x4b1/0x54c [ov8865]
[ 13.234117] i2c_device_probe+0x13c/0x2d0
Fixes: 11c0d8fdccc5 ("media: i2c: Add support for the OV8865 image sensor")
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The kerneldoc for pm_runtime_set_suspended() says:
"It is not valid to call this function for devices with runtime PM
enabled"
To satisfy that requirement, re-order the calls so that
pm_runtime_enable() is the last one.
Fixes: 11c0d8fdccc5 ("media: i2c: Add support for the OV8865 image sensor")
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Add a link frequency to the cio2-bridge table of supported sensors.
This means that the driver can parse supported link frequencies from
firmware in the usual way and validate that it can accommodate them.
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Tested-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The OV5693 is a 5 Mpx CMOS image sensor, connected via MIPI CSI-2. The
chip is capable of a single lane configuration, but currently only two
lanes are supported.
Most of the sensor's features are supported, with the main exception
being the lens correction algorithm.
The driver provides all mandatory, optional and recommended V4L2 controls
for maximum compatibility with libcamera.
[mchehab: fixed a coding style warning]
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Tested-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The .suspend() and .resume() runtime_pm operations for the ipu3-cio2
driver currently do not handle the sensor's stream. Setting .s_stream() on
or off for the sensor subdev means that sensors will pause and resume the
stream at the appropriate time even if their drivers don't implement those
operations.
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Daniel Scally <djrscally@gmail.com>
Tested-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Current driver supports only SRGGB 10 bit RAW bayer format.
Add the enum_mbus_code implementation to report this format supported.
# v4l2-ctl -d /dev/v4l-subdev3 --list-subdev-mbus-codes
ioctl: VIDIOC_SUBDEV_ENUM_MBUS_CODE (pad=0)
0x300f: MEDIA_BUS_FMT_SRGGB10_1X10
#
Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Fix typo obainted/obtained.
Fixes: 0985dd306f72 ("media: imx274: V4l2 driver for Sony imx274 CMOS sensor")
Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Fix typo expsoure/exposure
Fixes: 0985dd306f72 ("media: imx274: V4l2 driver for Sony imx274 CMOS sensor")
Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
pm_runtime_resume_and_get should be called when the s_frame_interval
is called.
The driver will try to access device registers to configure VMAX, coarse
time and exposure.
Currently if the runtime is not resumed, this fails:
# media-ctl -d /dev/media0 --set-v4l2 '"IMX274 1-001a":0[fmt:SRGGB10_1X10/3840x2
160@1/10]'
IMX274 1-001a: imx274_binning_goodness: ask 3840x2160, size 3840x2160, goodness 0
IMX274 1-001a: imx274_binning_goodness: ask 3840x2160, size 1920x1080, goodness -3000
IMX274 1-001a: imx274_binning_goodness: ask 3840x2160, size 1280x720, goodness -4000
IMX274 1-001a: imx274_binning_goodness: ask 3840x2160, size 1280x540, goodness -4180
IMX274 1-001a: __imx274_change_compose: selected 1x1 binning
IMX274 1-001a: imx274_set_frame_interval: input frame interval = 1 / 10
IMX274 1-001a: imx274_read_mbreg : addr 0x300e, val=0x1 (2 bytes)
IMX274 1-001a: imx274_set_frame_interval : register SVR = 1
IMX274 1-001a: imx274_read_mbreg : addr 0x30f6, val=0x6a8 (2 bytes)
IMX274 1-001a: imx274_set_frame_interval : register HMAX = 1704
IMX274 1-001a: imx274_set_frame_length : input length = 2112
IMX274 1-001a: imx274_write_mbreg : i2c bulk write failed, 30f8 = 884 (3 bytes)
IMX274 1-001a: imx274_set_frame_length error = -121
IMX274 1-001a: imx274_set_frame_interval error = -121
Unable to setup formats: Remote I/O error (121)
The device is not resumed thus the remote I/O error.
Setting the frame interval works at streaming time, because
pm_runtime_resume_and_get is called at s_stream time before sensor setup.
The failure happens when only the s_frame_interval is called separately
independently on streaming time.
Fixes: ad97bc37426c ("media: i2c: imx274: Add IMX274 power on and off sequence")
Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The MAX9286 has not explicitly declared a dependency upon VIDEO_V4L2.
While this dependency has likely always been met by configurations
including it, the device does use V4L2 core, and should depend upon it.
Add VIDEO_V4L2 as a dependency to match other drivers and prevent
failures when compile testing.
Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
dmaengine_terminate_all() is deprecated in favor of explicitly saying if
it should be sync or async. Here, we use dmaengine_terminate_sync in
i2c_xfer and i2c_smbus_xfer handlers and rely on
dmaengine_terminate_async within interrupt handlers
(transmission error cases).
dmaengine_synchronize is added within i2c_xfer and i2c_smbus_xfer handler
to finalize terminate started in interrupt handlers.
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Reviewed-by: Pierre-Yves MORDRET <pierre-yves.mordret@foss.st.com>
Signed-off-by: Wolfram Sang <wsa@kernel.org>
|
|
In case of receiving a NACK, the dma transfer should be stopped
to avoid feeding data into the FIFO.
Also ensure to properly return the proper error code and avoid
waiting for the end of the dma completion in case of
error happening during the transmission.
Fixes: 7ecc8cfde553 ("i2c: i2c-stm32f7: Add DMA support")
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Reviewed-by: Pierre-Yves MORDRET <pierre-yves.mordret@foss.st.com>
Signed-off-by: Wolfram Sang <wsa@kernel.org>
|
|
When getting an access timeout, ensure that the bus is in a proper
state prior to returning the error.
Fixes: aeb068c57214 ("i2c: i2c-stm32f7: add driver")
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Reviewed-by: Pierre-Yves MORDRET <pierre-yves.mordret@foss.st.com>
Signed-off-by: Wolfram Sang <wsa@kernel.org>
|
|
Generally, kvm->lock is not taken for a long time, but
sev_lock_two_vms is different: it takes vCPU locks
inside, so userspace can hold it back just by calling
a vCPU ioctl. Play it safe and use mutex_lock_killable.
Message-Id: <20211123005036.2954379-13-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Taking the lock is useless since there are no other references,
and there are already accesses (e.g. to sev->enc_context_owner)
that do not take it. So get rid of it.
Reviewed-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211123005036.2954379-12-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
VMs that mirror an encryption context rely on the owner to keep the
ASID allocated. Performing a KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM
would cause a dangling ASID:
1. copy context from A to B (gets ref to A)
2. move context from A to L (moves ASID from A to L)
3. close L (releases ASID from L, B still references it)
The right way to do the handoff instead is to create a fresh mirror VM
on the destination first:
1. copy context from A to B (gets ref to A)
[later] 2. close B (releases ref to A)
3. move context from A to L (moves ASID from A to L)
4. copy context from L to M
So, catch the situation by adding a count of how many VMs are
mirroring this one's encryption context.
Fixes: 0b020f5af092 ("KVM: SEV: Add support for SEV-ES intra host migration")
Message-Id: <20211123005036.2954379-11-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Now that we have a facility to lock two VMs with deadlock
protection, use it for the creation of mirror VMs as well. One of
COPY_ENC_CONTEXT_FROM(dst, src) and COPY_ENC_CONTEXT_FROM(src, dst)
would always fail, so the combination is nonsensical and it is okay to
return -EBUSY if it is attempted.
This sidesteps the question of what happens if a VM is
MOVE_ENC_CONTEXT_FROM'd at the same time as it is
COPY_ENC_CONTEXT_FROM'd: the locking prevents that from
happening.
Cc: Peter Gonda <pgonda@google.com>
Cc: Sean Christopherson <seanjc@google.com>
Reviewed-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211123005036.2954379-10-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
I am putting the tests in sev_migrate_tests because the failure conditions are
very similar and some of the setup code can be reused, too.
The tests cover both successful creation of a mirror VM, and error
conditions.
Cc: Peter Gonda <pgonda@google.com>
Cc: Sean Christopherson <seanjc@google.com>
Message-Id: <20211123005036.2954379-9-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Allow intra-host migration of a mirror VM; the destination VM will be
a mirror of the same ASID as the source.
Fixes: b56639318bb2 ("KVM: SEV: Add support for SEV intra host migration")
Reviewed-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211123005036.2954379-8-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
This was broken before the introduction of KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM,
but technically harmless because the region list was unused for a mirror
VM. However, it is untidy and it now causes a NULL pointer access when
attempting to move the encryption context of a mirror VM.
Fixes: 54526d1fd593 ("KVM: x86: Support KVM VMs sharing SEV context")
Message-Id: <20211123005036.2954379-7-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Encapsulate the handling of the migration_in_progress flag for both VMs in
two functions sev_lock_two_vms and sev_unlock_two_vms. It does not matter
if KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM locks the destination struct kvm a bit
later, and this change 1) keeps the cleanup chain of labels smaller 2)
makes it possible for KVM_CAP_VM_COPY_ENC_CONTEXT_FROM to reuse the logic.
Cc: Peter Gonda <pgonda@google.com>
Cc: Sean Christopherson <seanjc@google.com>
Message-Id: <20211123005036.2954379-6-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
list_replace_init cannot be used if the source is an empty list,
because "new->next->prev = new" will overwrite "old->next":
new old
prev = new, next = new prev = old, next = old
new->next = old->next prev = new, next = old prev = old, next = old
new->next->prev = new prev = new, next = old prev = old, next = new
new->prev = old->prev prev = old, next = old prev = old, next = old
new->next->prev = new prev = old, next = old prev = new, next = new
The desired outcome instead would be to leave both old and new the same
as they were (two empty circular lists). Use list_cut_before, which
already has the necessary check and is documented to discard the
previous contents of the list that will hold the result.
Fixes: b56639318bb2 ("KVM: SEV: Add support for SEV intra host migration")
Reviewed-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211123005036.2954379-5-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Currently, checks for whether VT-d PI can be used refer to the current
status of the feature in the current vCPU; or they more or less pick
vCPU 0 in case a specific vCPU is not available.
However, these checks do not attempt to synchronize with changes to
the IRTE. In particular, there is no path that updates the IRTE when
APICv is re-activated on vCPU 0; and there is no path to wakeup a CPU
that has APICv disabled, if the wakeup occurs because of an IRTE
that points to a posted interrupt.
To fix this, always go through the VT-d PI path as long as there are
assigned devices and APICv is available on both the host and the VM side.
Since the relevant condition was copied over three times, take the hint
and factor it into a separate function.
Suggested-by: Sean Christopherson <seanjc@google.com>
Cc: stable@vger.kernel.org
Reviewed-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: David Matlack <dmatlack@google.com>
Message-Id: <20211123004311.2954158-5-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
The IRTE for an assigned device can trigger a POSTED_INTR_VECTOR even
if APICv is disabled on the vCPU that receives it. In that case, the
interrupt will just cause a vmexit and leave the ON bit set together
with the PIR bit corresponding to the interrupt.
Right now, the interrupt would not be delivered until APICv is re-enabled.
However, fixing this is just a matter of always doing the PIR->IRR
synchronization, even if the vCPU has temporarily disabled APICv.
This is not a problem for performance, or if anything it is an
improvement. First, in the common case where vcpu->arch.apicv_active is
true, one fewer check has to be performed. Second, static_call_cond will
elide the function call if APICv is not present or disabled. Finally,
in the case for AMD hardware we can remove the sync_pir_to_irr callback:
it is only needed for apic_has_interrupt_for_ppr, and that function
already has a fallback for !APICv.
Cc: stable@vger.kernel.org
Co-developed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: David Matlack <dmatlack@google.com>
Message-Id: <20211123004311.2954158-4-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
If APICv is disabled for this vCPU, assigned devices may still attempt to
post interrupts. In that case, we need to cancel the vmentry and deliver
the interrupt with KVM_REQ_EVENT. Extend the existing code that handles
injection of L1 interrupts into L2 to cover this case as well.
vmx_hwapic_irr_update is only called when APICv is active so it would be
confusing to add a check for vcpu->arch.apicv_active in there. Instead,
just use vmx_set_rvi directly in vmx_sync_pir_to_irr.
Cc: stable@vger.kernel.org
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: David Matlack <dmatlack@google.com>
Reviewed-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211123004311.2954158-3-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
A kvm_page_table_test run with its default settings fails on VMX due to
memory region add failure:
> ==== Test Assertion Failure ====
> lib/kvm_util.c:952: ret == 0
> pid=10538 tid=10538 errno=17 - File exists
> 1 0x00000000004057d1: vm_userspace_mem_region_add at kvm_util.c:947
> 2 0x0000000000401ee9: pre_init_before_test at kvm_page_table_test.c:302
> 3 (inlined by) run_test at kvm_page_table_test.c:374
> 4 0x0000000000409754: for_each_guest_mode at guest_modes.c:53
> 5 0x0000000000401860: main at kvm_page_table_test.c:500
> 6 0x00007f82ae2d8554: ?? ??:0
> 7 0x0000000000401894: _start at ??:?
> KVM_SET_USER_MEMORY_REGION IOCTL failed,
> rc: -1 errno: 17
> slot: 1 flags: 0x0
> guest_phys_addr: 0xc0000000 size: 0x40000000
This is because the memory range that this test is trying to add
(0x0c0000000 - 0x100000000) conflicts with LAPIC mapping at 0x0fee00000.
Looking at the code it seems that guest_test_*phys*_mem variable gets
mistakenly overwritten with guest_test_*virt*_mem while trying to adjust
the former for alignment.
With the correct variable adjusted this test runs successfully.
Signed-off-by: Maciej S. Szmigiero <maciej.szmigiero@oracle.com>
Message-Id: <52e487458c3172923549bbcf9dfccfbe6faea60b.1637940473.git.maciej.szmigiero@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Account for the '0' being a default, "let KVM choose" period, when
determining whether or not the recovery worker needs to be awakened in
response to userspace reducing the period. Failure to do so results in
the worker not being awakened properly, e.g. when changing the period
from '0' to any small-ish value.
Fixes: 4dfe4f40d845 ("kvm: x86: mmu: Make NX huge page recovery period configurable")
Cc: stable@vger.kernel.org
Cc: Junaid Shahid <junaids@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211120015706.3830341-1-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Initialize the mask for PKU permissions as if CR4.PKE=0, avoiding
incorrect interpretations of the nested hypervisor's page tables.
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Drop the "flush" param and return values to/from the TDP MMU's helper for
zapping collapsible SPTEs. Because the helper runs with mmu_lock held
for read, not write, it uses tdp_mmu_zap_spte_atomic(), and the atomic
zap handles the necessary remote TLB flush.
Similarly, because mmu_lock is dropped and re-acquired between zapping
legacy MMUs and zapping TDP MMUs, kvm_mmu_zap_collapsible_sptes() must
handle remote TLB flushes from the legacy MMU before calling into the TDP
MMU.
Fixes: e2209710ccc5d ("KVM: x86/mmu: Skip rmap operations if rmaps not allocated")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211120045046.3940942-4-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
