Add func intel_color_lut_equal() to compare hw/sw gamma
lut values. Since hw/sw gamma lut sizes and lut entries comparison
will be different for different gamma modes, add gamma mode dependent
checks.
v3: -Rebase
v4: -Renamed intel_compare_color_lut() to intel_color_lut_equal() [Jani]
-Added the default label above the correct label [Jani]
-Corrected smatch warn "variable dereferenced before check"
[Dan Carpenter]
v5: -Added condition (!blob1 && !blob2) return true [Jani]
v6: -Made patch11 as patch3 [Jani]
v8: -Split patch 3 into 4 patches
-Optimized blob check condition [Ville]
v9: -Exclude spilt gamma mode (bdw and ivb platforms)
as there is exception in way gamma values are written in
hardware [Ville]
-Added exception made in commit [Uma]
-Dropped else, character limit and indentation [Uma]
-Added multi segmented gama mode for icl+ platforms [Uma]
v10: -Dropped multi segmented mode for icl+ platforms [Jani]
-Removed references of sw and hw state in compare code [Jani]
-Dropped inline from func [Jani]
Signed-off-by: Swati Sharma <swati2.sharma@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1567538578-4489-4-git-send-email-swati2.sharma@intel.com
Each platform supports different gamma modes and each gamma mode
has different bit precision. Here bit precision corresponds
to number of bits the hw LUT supports.
Add func per platform to return bit precision corresponding to gamma mode
which will be later used as a parameter in lut comparison function
intel_color_lut_equal().
This is done for legacy, ilk, glk and their variant platforms.
v6: -Added func intel_color_get_bit_precision() to get bit precision for
gamma and degamma lut readout depending upon platform and
corresponding to load_luts() [Ankit]
-Made patch11 as patch3 [Jani]
v7: -Renamed func intel_color_get_bit_precision() to
intel_color_get_gamma_bit_precision()
-Added separate function/platform for gamma bit precision [Ville]
-Corrected checkpatch warnings
v8: -Split patch 3 into 4 separate patches
v9: -Changed commit message, gave more info [Uma]
-Added precision func for icl+ platform
v10: -Removed precision func for chv and icl+ platforms [Jani]
-Added gamma_enable check once [Jani]
Signed-off-by: Swati Sharma <swati2.sharma@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1567538578-4489-2-git-send-email-swati2.sharma@intel.com
Add debug log for color related parameters like gamma_mode, gamma_enable,
csc_enable, etc inside intel_dump_pipe_config().
v6: -Added debug log for color para in intel_dump_pipe_config [Jani]
v7: -Split patch 3 into 4 patches
v8: -Corrected alignment [Uma]
Signed-off-by: Swati Sharma <swati2.sharma@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1567538578-4489-3-git-send-email-swati2.sharma@intel.com
A collection of improvements and fixes for Intel pinctrl drivers
including:
- Converting drivers to use new devm_platform_ioremap_resource()
helper function.
- Make Interrupt Status (IS) register configurable.
- Allow locked pins to be requested and used as long as they are not
modified.
- Fix intel_pinctrl_should_save() to translate pin number to GPIO
number where needed. This fixes Asus X571GT touchpad
unresponsiveness issue after suspend/resume cycle.
-----BEGIN PGP SIGNATURE-----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=owt6
-----END PGP SIGNATURE-----
Merge tag 'intel-pinctrl-v5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/pinctrl/intel into devel
intel-pinctrl for v5.4
A collection of improvements and fixes for Intel pinctrl drivers
including:
- Converting drivers to use new devm_platform_ioremap_resource()
helper function.
- Make Interrupt Status (IS) register configurable.
- Allow locked pins to be requested and used as long as they are not
modified.
- Fix intel_pinctrl_should_save() to translate pin number to GPIO
number where needed. This fixes Asus X571GT touchpad
unresponsiveness issue after suspend/resume cycle.
Time has come to get rid of the old eeprom driver. The at24 driver
should be used instead. So mark the eeprom driver as deprecated and
give users some time to migrate. Then we can remove the legacy
eeprom driver completely.
Signed-off-by: Jean Delvare <jdelvare@suse.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20190902104838.058725c2@endymion
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch adds bindings for Soundwire Slave devices that includes how
SoundWire enumeration address and Link ID are used to represented in
SoundWire slave device tree nodes.
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20190829163514.11221-2-srinivas.kandagatla@linaro.org
Signed-off-by: Vinod Koul <vkoul@kernel.org>
This fixes an issue in which key down events for function keys would be
repeatedly emitted even after the user has raised the physical key. For
example, the driver fails to emit the F5 key up event when going through
the following steps:
- fnmode=1: hold FN, hold F5, release FN, release F5
- fnmode=2: hold F5, hold FN, release F5, release FN
The repeated F5 key down events can be easily verified using xev.
Signed-off-by: Joao Moreno <mail@joaomoreno.com>
Co-developed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
This ThinkCentre machine has a new realtek codec alc222, it is not
in the support list, we add it in the realtek.c then this machine
can apply FIXUPs for the realtek codec.
And this machine has two front mics which can't be handled
by PA so far, it uses the pin 0x18 and 0x19 as the front mics, as
a result the existing FIXUP ALC294_FIXUP_LENOVO_MIC_LOCATION doesn't
work on this machine. Fortunately another FIXUP
ALC283_FIXUP_HEADSET_MIC also can change the location for one of the
two mics on this machine.
Link: https://lore.kernel.org/r/20190904055327.9883-1-hui.wang@canonical.com
Signed-off-by: Hui Wang <hui.wang@canonical.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
port_status[port_num] are assigned for each port_num in some if
conditions. So some of the port_status may not be initialized.
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190829181135.16049-1-yung-chuan.liao@linux.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
while filling the linux inode, using switch-case statement to check
the type of inode.
switch-case statement looks more clean here.
Signed-off-by: Pratik Shinde <pratikshinde320@gmail.com>
Reviewed-by: Gao Xiang <gaoxiang25@huawei.com>
Link: https://lore.kernel.org/r/20190830095615.10995-1-pratikshinde320@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fixes checkpatch.pl warnings:
CHECK: Comparison to NULL could be written "expr"
CHECK: Comparison to NULL could be written "!expr"
Signed-off-by: Valentin Vidic <vvidic@valentin-vidic.from.hr>
Link: https://lore.kernel.org/r/20190903205659.18856-1-vvidic@valentin-vidic.from.hr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The `ar_usb` field of `ath6kl_usb_pipe_usb_pipe` objects
are initialized to point to the containing `ath6kl_usb` object
according to endpoint descriptors read from the device side, as shown
below in `ath6kl_usb_setup_pipe_resources`:
for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) {
endpoint = &iface_desc->endpoint[i].desc;
// get the address from endpoint descriptor
pipe_num = ath6kl_usb_get_logical_pipe_num(ar_usb,
endpoint->bEndpointAddress,
&urbcount);
......
// select the pipe object
pipe = &ar_usb->pipes[pipe_num];
// initialize the ar_usb field
pipe->ar_usb = ar_usb;
}
The driver assumes that the addresses reported in endpoint
descriptors from device side to be complete. If a device is
malicious and does not report complete addresses, it may trigger
NULL-ptr-deref `ath6kl_usb_alloc_urb_from_pipe` and
`ath6kl_usb_free_urb_to_pipe`.
This patch fixes the bug by preventing potential NULL-ptr-deref
(CVE-2019-15098).
Signed-off-by: Hui Peng <benquike@gmail.com>
Reported-by: Hui Peng <benquike@gmail.com>
Reported-by: Mathias Payer <mathias.payer@nebelwelt.net>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
clang triggers a warning about oversized stack frames that gcc does not
notice because of slightly different inlining decisions:
ath/wcn36xx/smd.c:1409:5: error: stack frame size of 1040 bytes in function 'wcn36xx_smd_config_bss' [-Werror,-Wframe-larger-than=]
ath/wcn36xx/smd.c:640:5: error: stack frame size of 1032 bytes in function 'wcn36xx_smd_start_hw_scan' [-Werror,-Wframe-larger-than=]
Basically the wcn36xx_hal_start_scan_offload_req_msg,
wcn36xx_hal_config_bss_req_msg_v1, and wcn36xx_hal_config_bss_req_msg
structures are too large to be put on the kernel stack, but small
enough that gcc does not warn about them.
Use kzalloc() to allocate them all. There are similar structures in other
parts of this driver, but they are all smaller, with the next largest
stack frame at 480 bytes for wcn36xx_smd_send_beacon.
Fixes: 8e84c25821 ("wcn36xx: mac80211 driver for Qualcomm WCN3660/WCN3680 hardware")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
In ath6kl_htc_mbox_create(), when kzalloc() on line 2855 fails,
target->dev is assigned to NULL, and ath6kl_htc_mbox_cleanup(target) is
called on line 2885.
In ath6kl_htc_mbox_cleanup(), target->dev is used on line 2895:
ath6kl_hif_cleanup_scatter(target->dev->ar);
Thus, a null-pointer dereference may occur.
To fix this bug, kfree(target) is called and NULL is returned when
kzalloc() on line 2855 fails.
This bug is found by a static analysis tool STCheck written by us.
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Initialize acktimeout to the maximum configurable value in
ath_dynack_reset in order to not disconnect long distance static links
enabling dynack and even to take care of possible errors configuring
a static timeout. Moreover initialize station timeout value to the current
acktimeout value
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Compute maximum configurable ackimeout/ctstimeout according to channel
width (clockrate)
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Add compute timeout to last computation timestamp in
ath_dynack_reset in order to not run ath_dynack_compute_ackto
immediately
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Introduce ath_dynack_set_timeout routine to configure slottime/ack/cts
timeouts and remove duplicated code
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Atheros cards with a AR92XX generation (and older) chip usually
store their pci(e) initialization vectors on an external eeprom chip.
However these chips technically don't need the eeprom chip attached,
the AR9280 Datasheet in section "6.1.2 DEVICE_ID" describes that
"... if the EEPROM content is not valid, a value of 0xFF1C returns
when read from the register". So, they will show up on the system's
pci bus. However in that state, ath9k can't load, since it relies
on having the correct pci-id, otherwise it doesn't know what chip it
actually is. This happens on many embedded devices like routers
and accesspoint since they want to keep the BOM low and store the
pci(e) initialization vectors together with the calibration data
on the system's FLASH, which is out of reach of the ath9k chip.
Furthermore, Some devices (like the Cisco Meraki Z1 Cloud Managed
Teleworker Gateway) need to be able to initialize the PCIe wifi device.
Normally, this should be done as a pci quirk during the early stages of
booting linux. However, this isn't possible for devices which have the
init code for the Atheros chip stored on NAND in an UBI volume.
Hence, this module can be used to initialize the chip when the
user-space is ready to extract the init code.
Martin Blumenstingl prodived the following fixes:
owl-loader: add support for OWL emulation PCI devices
owl-loader: don't re-scan the bus when ath9k_pci_fixup failed
owl-loader: use dev_* instead of pr_* logging functions
owl-loader: auto-generate the eeprom filename as fallback
owl-loader: add a debug message when swapping the eeprom data
owl-loader: add missing newlines in log messages
Reviewed-by: Julian Calaby <julian.calaby@gmail.com>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/net/wireless/ath/carl9170/usb.c: In function carl9170_usb_disconnect:
drivers/net/wireless/ath/carl9170/usb.c:1110:21:
warning: variable udev set but not used [-Wunused-but-set-variable]
It is not use since commit feb09b2933 ("carl9170:
fix misuse of device driver API")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Christian Lamparter <chunkeey@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
A null pointer would be passed to a call of the function “kfree”
directly after a call of the function “kcalloc” failed at one place.
Remove this superfluous function call.
This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Reviewed-by: Maya Erez <merez@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Add support for Enhanced Directional Multi-Gigabit (EDMG) channels 9-11.
wil6210 reports it's EDMG capabilities (that are also based on FW
capability) to cfg80211 by filling
wiphy->bands[NL80211_BAND_60GHZ]->edmg_cap.
wil6210 handles edmg.channels and edmg.bw_config requested in connect
and start_ap operations.
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
The commit 20c169aceb ("dmaengine: rcar-dmac: clear pertinence
number of channels") forgets to clear the last channel by
DMACHCLR in rcar_dmac_init() (and doesn't need to clear the first
channel) if iommu is mapped to the device. So, this patch fixes it
by using "channels_mask" bitfield.
Note that the hardware and driver don't support more than 32 bits
in DMACHCLR register anyway, so this patch should reject more than
32 channels in rcar_dmac_parse_of().
Fixes: 20c169aceb ("dmaengine: rcar-dmac: clear pertinence number of channels")
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/1567424643-26629-1-git-send-email-yoshihiro.shimoda.uh@renesas.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
There are several occasions where a negative cid value is passed
into wil_cid_valid and this is converted into a u8 causing the
range check of cid >= 0 to always succeed. Fix this by making
the cid argument an int to handle any -ve error value of cid.
An example of this behaviour is in wil_cfg80211_dump_station,
where cid is assigned -ENOENT if the call to wil_find_cid_by_idx
fails, and this -ve value is passed to wil_cid_valid. I believe
that the conversion of -ENOENT to the u8 value 254 which is
greater than wil->max_assoc_sta causes wil_find_cid_by_idx to
currently work fine, but I think is by luck and not the
intended behaviour.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Maya Erez <merez@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
For the Spreadtrum DMA link-list mode, when the DMA engine got a slave
hardware request, which will trigger the DMA engine to load the DMA
configuration from the link-list memory automatically. But before the
slave hardware request, the slave will get an incorrect residue due
to the first node used to trigger the link-list was configured as the
last source address and destination address.
Thus we should make sure the first node was configured the start source
address and destination address, which can fix this issue.
Fixes: 4ac6954647 ("dmaengine: sprd: Support DMA link-list mode")
Signed-off-by: Baolin Wang <baolin.wang@linaro.org>
Link: https://lore.kernel.org/r/77868edb7aff9d5cb12ac3af8827ef2e244441a6.1567150471.git.baolin.wang@linaro.org
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Fix printk format warning in iop-adma.c (seen on x86_64) by using
%pad:
../drivers/dma/iop-adma.c:118:12: warning: format ‘%x’ expects argument of type ‘unsigned int’, but argument 6 has type ‘dma_addr_t {aka long long unsigned int}’ [-Wformat=]
Fixes: c211092313 ("dmaengine: driver for the iop32x, iop33x, and iop13xx raid engines")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Dan Williams <dan.j.williams@intel.com>
Link: https://lore.kernel.org/r/1803541f-98a6-7cce-b050-ff1e9a333ab2@infradead.org
Signed-off-by: Vinod Koul <vkoul@kernel.org>
This pull-request contains the FPGA DFL changes for 5.4
- The first three patches are cleanup patches making use of dev_groups and
making the init callback optional.
- One patch adds userclock sysfs entries that are DFL specific
- One patch exposes AFU port disable/enable functions
- One patch adds error reporting
- One patch adds AFU SignalTap support
- One patch adds FME global error reporting
- The final patch is a documentation patch that decribes the
virtualization interfaces
This patchset requires the 'dev_groups_all_drivers' tag from drivers
core for the dev_groups refactoring as well as the DFL changes already
in char-misc-next.
Signed-off-by: Moritz Fischer <mdf@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQS/ea/a56fFi9QbQeJ+E8eWOj6VqQUCXW8mjQAKCRB+E8eWOj6V
qVbHAQDnhqNNyMPY93VseWIyJ/9OpkSzunc820XfmVehPZkw+gD+KLByUWy6vZN0
lh/LgdvIKYeUSm1ZZXj+DszemmaPiwk=
=xtf/
-----END PGP SIGNATURE-----
Merge tag 'fpga-dfl-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mdf/linux-fpga into char-misc-next
Moritz writes:
FPGA DFL Changes for 5.4
This pull-request contains the FPGA DFL changes for 5.4
- The first three patches are cleanup patches making use of dev_groups and
making the init callback optional.
- One patch adds userclock sysfs entries that are DFL specific
- One patch exposes AFU port disable/enable functions
- One patch adds error reporting
- One patch adds AFU SignalTap support
- One patch adds FME global error reporting
- The final patch is a documentation patch that decribes the
virtualization interfaces
This patchset requires the 'dev_groups_all_drivers' tag from drivers
core for the dev_groups refactoring as well as the DFL changes already
in char-misc-next.
Signed-off-by: Moritz Fischer <mdf@kernel.org>
* tag 'fpga-dfl-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mdf/linux-fpga:
Documentation: fpga: dfl: add descriptions for virtualization and new interfaces.
fpga: dfl: fme: add global error reporting support
fpga: dfl: afu: add STP (SignalTap) support
fpga: dfl: afu: add error reporting support.
fpga: dfl: afu: expose __afu_port_enable/disable function.
fpga: dfl: afu: add userclock sysfs interfaces.
fpga: dfl: afu: convert platform_driver to use dev_groups
fpga: dfl: fme: convert platform_driver to use dev_groups
fpga: dfl: make init callback optional
driver core: add dev_groups to all drivers
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct stm32_dma_desc {
...
struct stm32_dma_sg_req sg_req[];
};
Make use of the struct_size() helper instead of an open-coded version
in order to avoid any potential type mistakes.
So, replace the following function:
static struct stm32_dma_desc *stm32_dma_alloc_desc(u32 num_sgs)
{
return kzalloc(sizeof(struct stm32_dma_desc) +
sizeof(struct stm32_dma_sg_req) * num_sgs, GFP_NOWAIT);
}
with:
kzalloc(struct_size(desc, sg_req, num_sgs), GFP_NOWAIT)
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Link: https://lore.kernel.org/r/20190830161423.GA3483@embeddedor
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Move object release into a separate worker. Releasing objects requires
sending commands to the host. Doing that in the dequeue worker will
cause deadlocks in case the command queue gets filled up, because the
dequeue worker is also the one which will free up slots in the command
queue.
Reported-by: Chia-I Wu <olvaffe@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Tested-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20190830060116.10476-1-kraxel@redhat.com
virtio-gpu basically needs a sg_table for the bo, to tell the host where
the backing pages for the object are. So the gem shmem helpers are a
perfect fit. Some drm_gem_object_funcs need thin wrappers to update the
host state, but otherwise the helpers handle everything just fine.
Once the fencing was sorted the switch was surprisingly easy and for the
most part just removing the ttm code.
v4: fix drm_gem_object_funcs name.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20190829103301.3539-15-kraxel@redhat.com
Rework fencing workflow. Stop using ttm helpers, use the
virtio_gpu_array_* helpers instead.
Due to using the gem reservation object it is initialized and ready for
use before calling ttm_bo_init. So we can simply use the standard
fencing workflow and drop the tricky logic which checks whenever the
command is in flight still.
v6: rewrite most of the patch.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20190829103301.3539-10-kraxel@redhat.com
Rework fencing workflow, starting with virtio_gpu_execbuffer_ioctl.
Stop using ttm helpers, use the virtio_gpu_array_* helpers (which work
on the reservation objects directly) instead.
Also store the object array in struct virtio_gpu_vbuffer, so we
explicitly keep a reference of all buffers used instead of depending
on ttm_bo_put() checking whenever the object is actually idle before
releasing it.
New workflow:
(1) All gem objects needed by a command are added to a
virtio_gpu_object_array.
(2) All reservation objects will be locked (virtio_gpu_array_lock_resv).
(3) virtio_gpu_fence_emit() completes fence initialization.
(4) fence gets added to the objects, reservation objects are unlocked
(virtio_gpu_array_add_fence, virtio_gpu_array_unlock_resv).
(5) virtio command is submitted to the host.
(6) The completion callback (virtio_gpu_dequeue_ctrl_func)
will drop object references and free virtio_gpu_object_array.
v6: rewrite most of the patch.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20190829103301.3539-9-kraxel@redhat.com
Some helper functions to manage an array of gem objects.
v9: use dma_resv_lock_interruptible.
v6:
- add ticket to struct virtio_gpu_object_array.
- add virtio_gpu_array_{lock,unlock}_resv helpers.
- add virtio_gpu_array_add_fence helper.
v5: some small optimizations (Chia-I Wu).
v4: make them virtio-private instead of generic helpers.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20190829103301.3539-8-kraxel@redhat.com
Use drm_gem_reservation_object_wait() in virtio_gpu_wait_ioctl().
This also makes the ioctl run lockless.
v9: fix return value.
v5: handle lookup failure.
v2: use reservation_object_test_signaled_rcu for VIRTGPU_WAIT_NOWAIT.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20190829103301.3539-3-kraxel@redhat.com
With this gem and ttm will use the same reservation object,
so mixing and matching ttm / gem reservation helpers should
work fine.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20190829103301.3539-2-kraxel@redhat.com
