drm: xlnx: zynqmp: Disable a plane when the fb format changes

The drm core doesn't explicitly disable a plane when format changes.
So add a check in the plane update functions if the new framebuffer
format has changed, and disable the plane for the format change.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Tested-by: Kuldeep Dave <kuldeepd@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm64: zynqmp: Add support for zc1275 revB

This patch enables support zc1275 revB board. It has
SD added compared to revA. The same configuration will
work for RevC boards aswell.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm64: zynqmp: Disable WP on zcu111

On this board there is SD slot without WP connected.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>

serial: xuartps: Fix suspend functionality

The driver's suspend/resume functions were buggy.
If UART node contains any child node in the DT and
the child is established a communication path with
the parent UART. The relevant /dev/ttyPS* node will
be not available for other operations.
If the driver is trying to do any operations like
suspend/resume without checking the tty->dev status
it leads to the kernel crash/hang.

This patch fix this issue by call the device_may_wake()
with the generic parameter of type struct device.
in the uart suspend and resume paths.

It also fixes a race condition in the uart suspend
path(i.e uart_suspend_port() should be called at the
end of cdns_uart_suspend API this path updates the same)

Signed-off-by: Nava kishore Manne <navam@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drm: xlnx: drv: Shutdown the drm device properly

If a drm device is released immediately in shutdown callback by calling
drm_put_device(), it gives a warnning as below because there are some
resources, drm gem objects, that are still used by fbdev emulation and
attached to the drm device.

> reboot

or

> shutdown -h now

...
[   40.731016] [<ffffff800850adc0>] drm_gem_destroy+0x18/0x30
[   40.736485] [<ffffff800850d690>] drm_dev_fini+0x28/0x90
[   40.741693] [<ffffff800850d764>] drm_dev_unref+0x6c/0x78
[   40.746988] [<ffffff800850d7ac>] drm_put_dev+0x3c/0x70
[   40.752111] [<ffffff80085379b4>] xlnx_shutdown+0x14/0x20
[   40.757407] [<ffffff8008547b2c>] device_shutdown+0x114/0x1f0
[   40.763050] [<ffffff80080ba464>] kernel_restart_prepare+0x34/0x40
[   40.769125] [<ffffff80080ba544>] kernel_restart+0x14/0x78
[   40.774507] [<ffffff80080ba7e4>] SyS_reboot+0xbc/0x1e8
[   40.779627] Exception stack(0xffffff8013b13ec0 to 0xffffff8013b14000)
...

This fixes the warning by shutting down the drm device in complete
sequence.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

soc: xilinx: vcu: Provided API to get number of cores

Number of cores related information is provided
through the xvcu_get_num_cores API.

Signed-off-by: Dhaval Shah <dshah@xilinx.com>
Reviewed-by: Varunkumar Allagadapa <varunkumar.allagadapa@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

xilinx: v4l2: dma: Bug fix in single plane support

This patch fixes bugs in single plane handling.
 a. Passing v4l2 pixel format to framebuffer driver
 b. Making dst_icg is zero, though it does not matter
    in case of single plane formats
 c. Correcting bytes per line for single plane 10 bit
    formats

Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Setting FEC_CODE change

Setting of the FEC_Code (LDPC/Turbo) is decided at using the device tree
only (at design time only), remove the setting of the value in the function
xsdfec_set_turbo and xsdfec_add_ldpc to function xsdfec_parse_of to ensure
that this is OK.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Correct invalid order check

Correct the value used when checking order value in the function
xsdfec_start.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Use dev_dbg in xsdfec_regread

To be consistent with xsdfec_regwrite function.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Memory leak fix

In function xsdfec_add_ldpc and xsdfec_get_ldpc_code_params all failures
after successfully allocating "ldpc/ldpc_params" memory should free the
allocated memory before returning from the function.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Change struct for GET_STATUS IOCTL

Removes configuration and error stats information from the status.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Add IOCTL XSDFEC_IS_ACTIVE

Gives the user the ability to determine if the SDFEC is currently
progressing a block of data.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Add BYPASS IOCTLs

To allow user to the bypass the SDFEC block, where output data matches the
input data but the SDFEC performs the same operation(number of cycles).

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Fix open_count decrement when already open

In function xsdfec_dev_open if the atomic_dec_and_test detects that driver
is already opened a increment is required. This ensures the driver
instance can the reopened after it is released by the user.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Remove hardcoded Order setting

Occurs in function xsdfec_start.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Add IOCTL XSDFEC_SET_ORDER

Gives user ability to set the order, i.e. if data blocks sequence can
change from input to output.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Verify IOCTL arguments in every IOCTL

Previously the function xsdfec_dev_ioctl always checks if the ioctl
argument is not NULL.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm64: zynqmp: Set Dual Role Device Mode by default

This patch enables the dwc3 DRD mode by default. Doing so, will
make the same Image to be used for both Host & Device modes
based on the dr_mode dts parameter.

Signed-off-by: Anurag Kumar Vulisha <anuragku@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drm: xlnx: zynqmp_dp: Use the correct number of lanes

The number of lanes is configurable, thus the dynamically configured
value should be used instead of hard-coded max number of lanes. Use
the number of assigned lanes correctly.

While at it, don't print an error for phy probe deferral.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Tested-by: Parth Gajjar <parthg@xilinx.com>
Reviewed-by: Saurabh Sengar <saurabhs@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

v4l: xilinx: dma: Return from mplane handling for enum formats

In current driver implemenataion, it is always falling back to
single plane handling after mplane handling. This patch ensures
that mplane handling return gracefully.

Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dt: bindings: display: xlnx: Updated gpio reset entry

This patch updates reset gpio entry arguments.

Signed-off-by: Venkateshwar Rao G <vgannava@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drm: xlnx: csc: Adding vpss-csc reset through gpio

The color space converter reset is controlled by gpio.

Signed-off-by: Venkateshwar Rao G <vgannava@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dt: bindings: display: xlnx: Adding gpio reset entry

This patch adds the gpio reset entry for vpss color space converter.
Moved to bindings/display/xlnx folder.

Signed-off-by: Venkateshwar Rao G <vgannava@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

tty: xilinx_uartps: Enable automatic flow control

Enable automatic flow control which should ensure that there is no
mainteinance in connection for zcu100 BT case.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Nava kishore Manne <navam@xilinx.com>

irqchip: xilinx: Clean up error paths

This patch cleans error paths to ensure that
irqc->base is unmapped and intc_dev is being
freed properly.

Signed-off-by: Mubin Sayyed <mubinusm@xilinx.com>
Reviewed-by: Radhey Shyam Pandey <radheys@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Add IOCTL to get LDPC Params

Adding IOTCL XSDFEC_GET_LDPC_CODE_PARAMS to allow user verify LDPC
code params.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Add IOCTL to get Turbo Params

Adding IOCTL XSDFEC_GET_TURBO to allow user verify entries turbo param
values.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

soc: zynqmp: Fix PMU version comparison

Fix comparison of actual PMUFW version with expected version.

Signed-off-by: Rajan Vaja <rajanv@xilinx.com>
Acked-by: Jolly Shah <jollys@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm64: zynqmp: dt: Move zynqmp_dpsub under amba

The DP device node should be under amba. Currently, it's under root,
and this patch moves the node under amba.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drm: xlnx: scaler: Updating vpss-scaler reset logic

This patch updates the reset through gpio logic

Signed-off-by: Venkateshwar Rao G <vgannava@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

Bluetooth: hci_ll: Extend timeout between start and communication

Extend timeout between bt enable and communication with chip.
Flow control is automatic and driver is responding that device is ready
to accept messages but it doesn't need to be truth. Wait more time to
ensure that communication cat start.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Series-prefix: LINUX
Cover-letter:
zcu100 BT fixes
Hi,

I have cherry-picked mainline patches to make sure that we are using the
latest hci_ll code with fixes.

CR-993696

Thanks,
Michal
END

arm64: zynqmp: Enable serdev devices for zcu100 BT chip

Enable HCIuart and serdev.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm64: zynqmp: Use serdev for zcu100 BT

Mainline started to use serdev interface for uart attached devices.
Change description to reflect it.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>

Bluetooth: hci_ll: Add optional nvmem BD address source

This adds an optional nvmem consumer to get a BD address from an external
source. The BD address is then set in the Bluetooth chip after the
firmware has been loaded.

This has been tested working with a TI CC2560A chip (in a LEGO MINDSTORMS
EV3).

Signed-off-by: David Lechner <david@lechnology.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

Bluetooth: hci_ll: Add endianness conversion when setting baudrate

This adds an endianness conversion when setting the baudrate using a
vendor-specific command. Otherwise, bad things might happen on a big-
endian system.

Suggested-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: David Lechner <david@lechnology.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

Bluetooth: hci_ll: remove \n from kernel messages

The bt_* printk macros include a \n already, so we don't need extra ones
here.

Signed-off-by: David Lechner <david@lechnology.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

Bluetooth: mark expected switch fall-throughs

In preparation to enabling -Wimplicit-fallthrough, mark switch cases
where we are expecting to fall through.

In this particular case, notice that I replaced the
"deliberate fall-through..." comment with a "fall through"
comment, which is what GCC is expecting to find.

Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

Bluetooth: hci_ll: add support for setting public address

This adds support for setting the public address on Texas Instruments
Bluetooth chips using a vendor-specific command.

This has been tested on a CC2560A chip. The TI wiki also indicates that
this command should work on TI WL17xx/WL18xx Bluetooth chips.

During review, there was some question as to the correctness of the byte
swapping since TI's documentation is not clear on this matter. This can
be tested with the btmgmt utility from bluez. The adapter must be powered
off to change the address. If the baswap() is omitted, address is reversed.

In case there is a issue in the future, here is the output of btmon during
the command `btmgmt public-addr 00:11:22:33:44:55`:

Bluetooth monitor ver 5.43
= Note: Linux version 4.15.0-rc2-08561-gcb132a1-dirty (armv5tejl)      0.707043
= Note: Bluetooth subsystem version 2.22                               0.707091
= New Index: 00:17:E7:BD:1C:8E (Primary,UART,hci0)              [hci0] 0.707106
@ MGMT Open: btmgmt (privileged) version 1.14                 {0x0002} 0.707124
@ MGMT Open: bluetoothd (privileged) version 1.14             {0x0001} 0.707137
@ MGMT Open: btmon (privileged) version 1.14                  {0x0003} 0.707540
@ MGMT Command: Set Public Address (0x0039) plen 6    {0x0002} [hci0] 11.167991
        Address: 00:11:22:33:44:55 (CIMSYS Inc)
@ MGMT Event: Command Complete (0x0001) plen 7        {0x0002} [hci0] 11.175681
      Set Public Address (0x0039) plen 4
        Status: Success (0x00)
        Missing options: 0x00000000
@ MGMT Event: Index Removed (0x0005) plen 0           {0x0003} [hci0] 11.175757
@ MGMT Event: Index Removed (0x0005) plen 0           {0x0002} [hci0] 11.175757
@ MGMT Event: Index Removed (0x0005) plen 0           {0x0001} [hci0] 11.175757
= Open Index: 00:17:E7:BD:1C:8E                                [hci0] 11.176807
< HCI Command: Vendor (0x3f|0x0006) plen 6                     [hci0] 11.176975
        00 11 22 33 44 55                                .."3DU
> HCI Event: Command Complete (0x0e) plen 4                    [hci0] 11.188260
      Vendor (0x3f|0x0006) ncmd 1
        Status: Success (0x00)
...
< HCI Command: Read Local Version Info.. (0x04|0x0001) plen 0  [hci0] 11.189859
> HCI Event: Command Complete (0x0e) plen 12                   [hci0] 11.190732
      Read Local Version Information (0x04|0x0001) ncmd 1
        Status: Success (0x00)
        HCI version: Bluetooth 2.1 (0x04) - Revision 0 (0x0000)
        LMP version: Bluetooth 2.1 (0x04) - Subversion 6431 (0x191f)
        Manufacturer: Texas Instruments Inc. (13)
< HCI Command: Read BD ADDR (0x04|0x0009) plen 0               [hci0] 11.191027
> HCI Event: Command Complete (0x0e) plen 10                   [hci0] 11.192101
      Read BD ADDR (0x04|0x0009) ncmd 1
        Status: Success (0x00)
        Address: 00:11:22:33:44:55 (CIMSYS Inc)
...

Signed-off-by: David Lechner <david@lechnology.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

Bluetooth: hci_ll: add constant for vendor-specific command

This adds a #define for the vendor-specific HCI command to set the
baudrate instead of using the bare 0xff36 multiple times.

Signed-off-by: David Lechner <david@lechnology.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

Bluetooth: serdev: hci_ll: Wait for CTS instead of using msleep

When a TI Bluetooth chip is reset, it takes about 100ms for the RTS line of
the chip to deassert. For my use case with a TI CC2560A chip, this delay
was not long enough and caused the local UART to never transmit at all (TI
AM1808 SoC UART2).

We can wait for the CTS signal using serdev_device_wait_for_cts() instead
of trying to guess using msleep().

Also changed the comment to be more informative while we are touching this
code.

Signed-off-by: David Lechner <david@lechnology.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

Revert "of: overlay: add overlay symbols to live device tree"

This reverts commit d1651b03c2df75db8eda3fbcd3a07adb337ee8b0.

This commit must be reverted, else loading overlays
from the configfs interface will be fails with the
following errors:
OF: changeset: add_property failed @/__symbols__/...
OF: Error applying changeset (-17)

Signed-off-by: Nava kishore Manne <navam@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Rename XSDFEC_ADD_LDPC IOCTL

To be more explicit.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Rename xsd_ldpc

To be more explicit.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Remove lat_ctrl from xsd_ldpc

lat_ctrl is not required with the removal of the Latency Control field
from Reg2.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Remove XSDFEC_REG2_LAT_CTRL_*

Remove any references to Latency Control field of Reg2 register, the
register has been removed from the SDFEC Block.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

microblaze: Enable SPARSE_IRQ

Enabling SPARSE_IRQ to use dynamically allocated
irq descriptors.

Signed-off-by: Mubin Sayyed <mubinusm@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

irqchip: xilinx: Add support for multiple instances

This patch adds support for multiple instances of
xilinx interrupt controller. Below configurations are
supported by driver,

- peripheral->xilinx-intc->xilinx-intc->gic
- peripheral->xilinx-intc->xilinx-intc

Signed-off-by: Anirudha Sarangi <anirudh@xilinx.com>
Signed-off-by: Mubin Sayyed <mubinusm@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm64: zynqmp: dt: Add new ZynqMP DP changes

This removes old dt contents from all dts files for ZynqMP DisplayPort
and updates all with new bindings. Please note, some of these changes
are only build-tested.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm64: zynqmp: config: Enable soft IP DRM drivers

This enables DSI, SDI, Mixer, and the PL display drivers in defconfig.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Acked-by: Saurabh Sengar <saurabhs@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm64: zynqmp: config: Enable new Xilinx drm configs

Enable the new DRM configs. Keep the old ones for now,
and the DT can be used to selct one.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drm: xilinx: plane: Use drm_format_plane_width_bytes

This will calculate the number of bytes correctly for macropixel
formats.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Reviewed-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drm: xilinx: dp_sub: Add 10 bit YUV formats

This patch add support for 10 bit YUV formats (XV15 and XV20)
in ZynqMP dp driver.

Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Reviewed-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

uapi: drm: Add AVUY fourcc

AVUY is a 32bit single plane YUV format, with 2 bit alpha,
and in order of A:Cr:Cb:Y in little endian.

Signed-off-by: Jeffrey Mouroux <jmouroux@xilinx.com>
Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Reviewed-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

device-tree: serdes: Remove unused lpd register address mapping

This patch removes the unused lpd register address mapping from
serdes dts node.

Signed-off-by: Anurag Kumar Vulisha <anuragku@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

phy: zynqmp: Use eemi framework for performing ULPI reset sequence

This patch modifies the phy-zynqmp.c driver to use eemi framework
for performing ULPI reset sequence.

Signed-off-by: Anurag Kumar Vulisha <anuragku@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

zynqmp: firmware: Add IOCTL enum for performing ULPI reset

This patch adds IOCTL_ULPI_RESET IOCTL enum value that is
passed to eemi framework for performing ULPI reset

Signed-off-by: Anurag Kumar Vulisha <anuragku@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: DRIVER_VERSION Update

Updated DRIVER_VERSION to "0.3" for 2018.1 SD-FEC drivers updates.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Max # Instances display update

Display 7 as the MAX number of instances. In function xsdfec_probe during
check of "xsdfec_ndevs" to ensure Human Readable.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Fix enum xsdfec_order values

Add XSDFEC_ prefix to enum xsdfec_order values. To be consistent.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

misc: xilinx-sdfec: Added GET_CONFIG IOCTL

To allow validation of Core/Driver configuration.

Signed-off-by: Derek Kiernan <Derek.Kiernan@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

clk: clk-fixed-factor: Use new macro CLK_OF_DECLARE_DRIVER

Fixed factor clock has two initialization at of_clk_init()
time and also during platform driver probe. So declare the
fixed factor clock with CLK_OF_DECLARE_DRIVER instead of
CLK_OF_DECLARE.

See below commit for reference:
"clk: sunxi: apb0: Use new macro CLK_OF_DECLARE_DRIVER"
(sha1: 915128b621a05c63fa58ca9e4cbdf394bbe592f3)

Signed-off-by: Rajan Vaja <rajanv@xilinx.com>
Suggested-by: Michal Simek <michal.simek@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

uvcvideo: Prevent new URBs being processed at stream stop

With asynchronous handling of the URBs from the USB Complete handler, we
get a continual stream of packets being received while we are attempting
to shutdown the stream.

Packets that have already been received and processed are on a
work-queue, but during stream shutdown the URBs that those packets
belong to are killed and free'd.

To prevent this race from causing invalid memory accesses, prevent new
URBs from being processed when uvc_stop_streaming() is called by
introducing a new flag "UVC_QUEUE_STOPPING" into the queue, and checking
this when processing the URB to be queued.

With this, we can flush the work queue, and commence a normal pipe
shutdown. Work tasks that are already queued are processed, but the URBs
are prevented from being re-submitted to the USB stack.

Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Christian Kohn <christian.kohn@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

staging: apf: Add synchronization for DMA-BUF add/remove

DMA-BUF attachment in the APF driver lacked appropriate locking
guarantees. This patch adds those protections to allow for
concurrent attach/detach calls.

Signed-off-by: Michael Gill <gill@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

net: macb: Cleanup empty lines

Remove two extra empty lines.

Signed-off-by: Harini Katakam <harinik@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

networking: introduce and use skb_put_data()

A common pattern with skb_put() is to just want to memcpy()
some data into the new space, introduce skb_put_data() for
this.

An spatch similar to the one for skb_put_zero() converts many
of the places using it:

    @@
    identifier p, p2;
    expression len, skb, data;
    type t, t2;
    @@
    (
    -p = skb_put(skb, len);
    +p = skb_put_data(skb, data, len);
    |
    -p = (t)skb_put(skb, len);
    +p = skb_put_data(skb, data, len);
    )
    (
    p2 = (t2)p;
    -memcpy(p2, data, len);
    |
    -memcpy(p, data, len);
    )

    @@
    type t, t2;
    identifier p, p2;
    expression skb, data;
    @@
    t *p;
    ...
    (
    -p = skb_put(skb, sizeof(t));
    +p = skb_put_data(skb, data, sizeof(t));
    |
    -p = (t *)skb_put(skb, sizeof(t));
    +p = skb_put_data(skb, data, sizeof(t));
    )
    (
    p2 = (t2)p;
    -memcpy(p2, data, sizeof(*p));
    |
    -memcpy(p, data, sizeof(*p));
    )

    @@
    expression skb, len, data;
    @@
    -memcpy(skb_put(skb, len), data, len);
    +skb_put_data(skb, data, len);

(again, manually post-processed to retain some comments)a

Picked change for macb driver only.

Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Harini Katakam <harinik@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

lib/vsprintf.c: remove %Z support

Now that %z is standartised in C99 there is no reason to support %Z.
Unlike %L it doesn't even make format strings smaller.

Use BUILD_BUG_ON in a couple ATM drivers.

In case anyone didn't notice lib/vsprintf.o is about half of SLUB which
is in my opinion is quite an achievement.  Hopefully this patch inspires
someone else to trim vsprintf.c more.

Picked up change for macb driver only.

Link: http://lkml.kernel.org/r/20170103230126.GA30170@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Harini Katakam <harinik@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

net: macb: Adding Support for Jumbo Frames up to 10240 Bytes in SAMA5D3

As per the SAMA5D3 device specification it supports Jumbo frames.
But the suggested flag and length of bytes it supports was not updated
in this driver config_structure.
The maximum jumbo frames the device supports :
10240 bytes as per the device spec.

While changing the MTU value greater than 1500, it threw error:
sudo ifconfig eth1 mtu 9000
SIOCSIFMTU: Invalid argument

Add this support to driver so that it works as expected and designed.

Signed-off-by: vishnuvardhan <vardhanraj4143@gmail.com>
[nicolas.ferre@microchip.com: modify slightly commit msg]
Signed-off-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drivers: net: generalize napi_complete_done()

napi_complete_done() allows to opt-in for gro_flush_timeout,
added back in linux-3.19, commit 3b47d30396ba
("net: gro: add a per device gro flush timer")

This allows for more efficient GRO aggregation without
sacrifying latencies.

Picked up change for macb driver only.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Harini Katakam <harinik@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

net: macb: fix the RX queue reset in macb_rx()

On macb only (not gem), when a RX queue corruption was detected from
macb_rx(), the RX queue was reset: during this process the RX ring
buffer descriptor was initialized by macb_init_rx_ring() but we forgot
to also set bp->rx_tail to 0.

Indeed, when processing the received frames, bp->rx_tail provides the
macb driver with the index in the RX ring buffer of the next buffer to
process. So when the whole ring buffer is reset we must also reset
bp->rx_tail so the driver is synchronized again with the hardware.

Since macb_init_rx_ring() is called from many locations, currently from
macb_rx() and macb_init_rings(), we'd rather add the "bp->rx_tail = 0;"
line inside macb_init_rx_ring() than add the very same line after each
call of this function.

Without this fix, the rx queue is not reset properly to recover from
queue corruption and connection drop may occur.

Signed-off-by: Cyrille Pitchen <cyrille.pitchen@atmel.com>
Fixes: 9ba723b081a2 ("net: macb: remove BUG_ON() and reset the queue to handle RX errors")
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

net: macb: Use net_device_stats from struct net_device

Instead of using a private copy of struct net_device_stats in struct
macb, use stats from struct net_device.

Cc: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

macb: Remove bogus reference to NETIF_F_UFO.

This driver doesn't actually support UFO explicitly yet
it advertises this in netdev->features.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

net: macb: Add ethtool get_ringparam and set_ringparam functionality

Some applications want to tune the size of the macb rx/tx ring buffers.
The ethtool set_ringparam function is the standard way of doing it.

Signed-off-by: Zach Brown <zach.brown@ni.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

cadence: Add LSO support.

New Cadence GEM hardware support Large Segment Offload (LSO):
TCP segmentation offload (TSO) as well as UDP fragmentation
offload (UFO). Support for those features was added to the driver.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dmaengine: xilinx: dma: Program hardware supported buffer length

AXI-DMA IP supports configurable (c_sg_length_width) buffer length
register width, hence read buffer length (xlnx,sg-length-width) DT
property and ensure that driver doesn't program buffer length
exceeding the supported limit. For VDMA and CDMA there is no change.

Signed-off-by: Radhey Shyam Pandey <radheys@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

Documentation: DT: dma: Add axi-dma buffer length property

Buffer length is an optional DMA node property. DMA driver will
uses this property to ensure that programmed DMA length doesn't
exceed IP supported buffer length(c_sg_length_width).

Signed-off-by: Radhey Shyam Pandey <radheys@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dma: Add support for XV15 and XV20 contiguous formats

This patch updates the framebuffer driver to add support for
XV15 and XV20 contiguous formats.

Signed-off-by: Devarsh Thakkar <devarsht@xilinx.com>
Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

Documentation: devicetree: bindings: dma: Update fourcc codes for xv15/20

Support for xv15 and xv20 formats have been added to the Xilinx video
framebuffer driver.

Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: vsagar@xilinx.com
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

v4l2-core: Update entries for XV15 and XV20 contiguous formats

This patch adds missing entries XV20 and XV15 contiguous formats
and fix the warning as format description size is going beyond
32 byte limit.

Signed-off-by: Devarsh Thakkar <devarsht@xilinx.com>
Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

v4l: xilinx: dma: Add XV15 and XV20 contiguous format support

This patch adds support for XV15 and XV20 contiguous formats
support. Updated with right bitsperpixel for YUV 422 format.

Signed-off-by: Devarsh Thakkar <devarsht@xilinx.com>
Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dt-bindings: clock: zynqmp: Correct clock names

Some clocks names were intentionally kept incorrect to
match with Linux clock names for backward compatibility.
As now, clock is complete new driver, this is not required
so use proper names for the clocks.

Signed-off-by: Rajan Vaja <rajanv@xilinx.com>
Acked-by: Jolly Shah <jollys@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

gpio: xilinx: Add reset support

This patch updates xlate api and gpio cells value to
provide reset support for other IP's using axi gpio.

Signed-off-by: Swapna Manupati <swapnam@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dt-bindings: gpio-xilinx: Update gpio-cells information

This patch adds extra cell for flags in gpio-cells to reset
other IP's using axi gpio.

Signed-off-by: Swapna Manupati <swapnam@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

usb: dwc3: otg: don't clear event buffers when changing to Host mode from Device mode

Clearing event buffers when changing from Device mode -> Host mode
may require allocating event buffers again during role swap from
Host mode -> Device mode. Changing the event buffers at runtime are
not triggering events in device mode and also the role swap can happen
at any time. So, better not to clear the event buffers during role
swap instead reuse the previously allocated event buffers.
This patch does the same.

Signed-off-by: Anurag Kumar Vulisha <anuragku@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

arm: zynq: pl310_edac_l2: Update header file

Changed the header file from edac_core.h, which no
longer exists, to edac_module.h.

Signed-off-by: Jordan Anderson <anderson.jordan59@gmail.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drm: xlnx: mixer: reset the mixer IP before programming

Resetting the HLS IP as soon as gpio driver is probed.

Signed-off-by: Saurabh Sengar <saurabhs@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

drm: xlnx: mixer: remove deprecated dmaengine_terminate_all

Replacing dmaengine_terminate_all with dmaengine_terminate_sync,
as the former one is deprecated.

Signed-off-by: Saurabh Sengar <saurabhs@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dt: bindings: media: xilinx: Remove unimplemented TPG compatible string

TPG compatible string "xlnx,v-tpg-6.0" is not supported in the driver.
Hence remove unimplemented compatible string.

Signed-off-by: Radhey Shyam Pandey <radheys@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

media: xilinx: vip: Correct the fourcc code for BGRX8 format.

Correct the V4L2_PIX_FMT fourcc code for BGRX8 format
to V4L2_PIX_FMT_XBGR32.

Signed-off-by: Vishal Sagar <vsagar@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dma: xilinx: Correct the V4L fourcc for BGRX8 in framebuffer driver

Correct the V4L2_PIX_FMT fourcc code for BGRX8 to V4L2_PIX_FMT_XBGR32.

Signed-off-by: Vishal Sagar <vsagar@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

Documentation: devicetree: bindings: dma: Correct the BGRX8 format fourcc code

The BGRX8 framebuffer format is corrected for V4L2 Fourcc by changing
V4L2_PIX_FMT_XRGB32 to V4L2_PIX_FMT_XBGR32 in the table.

Signed-off-by: Vishal Sagar <vsagar@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

v4l: xilinx: Add BGR color format support.

Add support for the new BGR color format supported by Framebuffer
driver.

Signed-off-by: Vishal Sagar <vsagar@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

dma: xilinx: Add support for BGR8 color format in Framebuffer

Add support for BGR8 color format (Blue in LSB, 8 bits per component)
in Framebuffer.

Signed-off-by: Vishal Sagar <vsagar@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

Documentation: devicetree: bindings: dma: Add support for BGR8 color format

Add support for BGR8 color format (Blue in LSB, 8 bits per component)
to the Xilinx Video Framebuffer driver.

Signed-off-by: Vishal Sagar <vsagar@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

v4l: xilinx: dma: Fix bytes per line calculation

In current implementation there is a bug where min_bpl value
is not satisfying dma alignment. This patch will fix the issue
by aligning the value properly.

Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

v4l: xilinx: dma: Handle 10 bit format calucations for XV20

This patch handle 10 bit calculations for YUV 422 10 bit format

Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

v4l: xilinx: dma: Fix sizeimage calculation

sizeimage should be calculated using bytesperline instead
of width

Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

media: v4l2-core: Update V4L2 framework with new fourcc codes

New fourcc codes have been added to support additional video
memory layout supported by Xilinx Video IP.  These have been
added to the V4L2 framework with this patch.

Signed-off-by: Jeffrey Mouroux <jmouroux@xilinx.com>
Signed-off-by: Satish Kumar Nagireddy <satishna@xilinx.com>
Reviewed-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

net: macb: Remove unnecessary DBW read back from NWCFG

The data bus width field is already set based on MAC capability
using macb_dbw function. Remove this unnecessary read back of that
field which might interfere with the correct implementation.

Signed-off-by: Harini Katakam <harinik@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>

net: macb: Update macb RX tie-off descriptors

Assign correct descriptor size for RX tie-off descriptors and
use the updated functions for size and bit field to prepare them.

Signed-off-by: Harini Katakam <harinik@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>