yaroslavros [Wed, 14 Aug 2019 14:22:55 +0000 (15:22 +0100)]
Ported pcie-brcmstb bounce buffer implementation to ARM64. (#3144)
Ported pcie-brcmstb bounce buffer implementation to ARM64.
This enables full 4G RAM usage on Raspberry Pi in 64-bit mode.
Signed-off-by: Yaroslav Rosomakho <yaroslavros@gmail.com>
Dave Stevenson [Thu, 28 Jun 2018 14:57:25 +0000 (15:57 +0100)]
staging: bcm2835_camera: Ensure all buffers are returned on disable
With the recent change to match MMAL and V4L2 buffers there
is a need to wait for all MMAL buffers to be returned during
stop_streaming.
Fixes: 9384167 "staging: bcm2835-camera: Remove V4L2/MMAL buffer remapping"
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Hermann Lauer [Thu, 8 Aug 2019 13:40:37 +0000 (15:40 +0200)]
According to 5713 pdf doc CLOCK_CTRL is a readonly status register, and it behaves so. Remove useless setting
Eric Anholt [Thu, 2 May 2019 20:22:53 +0000 (13:22 -0700)]
drm/v3d: Clock V3D down when not in use.
My various attempts at re-enabling runtime PM have failed, so just
crank the clock down when V3D is idle to reduce power consumption.
Signed-off-by: Eric Anholt <eric@anholt.net>
Eric Anholt [Mon, 14 Jan 2019 23:13:17 +0000 (15:13 -0800)]
drm/v3d: HACK: gut runtime pm for now.
Something is still unstable -- on starting a new glxgears from an idle
X11, I get an MMU violation in high addresses. The CTS also failed
quite quickly. With this, CTS progresses for an hour before OOMing
(allocating some big buffers when my board only has 600MB available to
Linux)
Signed-off-by: Eric Anholt <eric@anholt.net>
Hui Wang [Thu, 22 Aug 2019 15:32:46 +0000 (23:32 +0800)]
rtl8192cu: Let it support to build in the non-src folder
If we build the kernel with "-O=$non-src-folder", this driver will
introdcue a building error because of the header's location.
Signed-off-by: Hui Wang <hui.wang@canonical.com>
Phil Elwell [Thu, 15 Aug 2019 07:39:08 +0000 (08:39 +0100)]
drm/vc4: Prevent load tracking from breaking FKMS
Firmware KMS uses a mixture of VC4 processing and dedicated code. The
load tracking support in VC4 assumes it is dealing with vc4_plane_state
objects when up-casting with container_of, but FKMS uses unadorned
drm_plane_state structures causing the VC4 code to read off the end
into random portions of memory. Work around the problem in a minimally-
invasive way by over-allocating the FKMS plane state structures to be
large enough to contain a vc4_plane_state, filling the remainder with
zeroes.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Phil Elwell [Wed, 7 Aug 2019 16:19:33 +0000 (17:19 +0100)]
pcie-brcmstb: Don't set DMA ops for root complex
A change to arm_get_dma_map_ops has stopped get_dma_ops from working
on the root complex, causing an error to be logged. However, there is
no need to override the DMA ops in that case, so skip it and
eliminate the error message.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
popcornmix [Tue, 6 Aug 2019 14:23:14 +0000 (15:23 +0100)]
clk-bcm2835: Avoid null pointer exception
clk_desc_array[BCM2835_PLLB] doesn't exist so we dereference null when iterating
Signed-off-by: popcornmix <popcornmix@gmail.com>
Phil Elwell [Thu, 4 Apr 2019 12:33:47 +0000 (13:33 +0100)]
bcm2835-dma: Add proper 40-bit DMA support
The 40-bit additions are not fully tested, but it should be
capable of supporting both 40-bit memcpy on BCM2711 and regular
Lite channels on BCM2835.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Marek Behún [Sat, 3 Aug 2019 12:34:59 +0000 (14:34 +0200)]
staging: vc04_services: fix compiling in separate directory
The vc04_services Makefiles do not respect the O=path argument
correctly: include paths in CFLAGS are given relatively to object path,
not source path. Compiling in a separate directory yields #include
errors.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Andrei Gherzan [Tue, 16 Jul 2019 12:28:22 +0000 (13:28 +0100)]
arm64/mm: Limit the DMA zone for arm64
On RaspberryPi, only the first 1Gb can be used for DMA[1].
[1] http://lists.infradead.org/pipermail/linux-arm-kernel/2019-July/665986.html
Signed-off-by: Andrei Gherzan <andrei@balena.io>
Phil Elwell [Fri, 12 Jul 2019 14:38:35 +0000 (15:38 +0100)]
i2c: bcm2835: Set clock-stretch timeout to 35ms
The BCM2835 I2C blocks have a register to set the clock-stretch
timeout - how long the device is allowed to hold SCL low - in bus
cycles. The current driver doesn't write to the register, therefore
the default value of 64 cycles is being used for all devices.
Set the timeout to the value recommended for SMBus - 35ms.
See: https://github.com/raspberrypi/linux/issues/3064
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Jonathan Bell [Thu, 11 Jul 2019 16:55:43 +0000 (17:55 +0100)]
xhci: add quirk for host controllers that don't update endpoint DCS
Seen on a VLI VL805 PCIe to USB controller. For non-stream endpoints
at least, if the xHC halts on a particular TRB due to an error then
the DCS field in the Out Endpoint Context maintained by the hardware
is not updated with the current cycle state.
Using the quirk XHCI_EP_CTX_BROKEN_DCS and instead fetch the DCS bit
from the TRB that the xHC stopped on.
See: https://github.com/raspberrypi/linux/issues/3060
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.org>
Phil Elwell [Thu, 11 Jul 2019 12:13:39 +0000 (13:13 +0100)]
tty: amba-pl011: Make TX optimisation conditional
pl011_tx_chars takes a "from_irq" parameter to reduce the number of
register accesses. When from_irq is true the function assumes that the
FIFO is half empty and writes up to half a FIFO's worth of bytes
without polling the FIFO status register, the reasoning being that
the function is being called as a result of the TX interrupt being
raised. This logic would work were it not for the fact that
pl011_rx_chars, called from pl011_int before pl011_tx_chars, releases
the spinlock before calling tty_flip_buffer_push.
A user thread writing to the UART claims the spinlock and ultimately
calls pl011_tx_chars with from_irq set to false. This reverts to the
older logic that polls the FIFO status register before sending every
byte. If this happen on an SMP system during the section of the IRQ
handler where the spinlock has been released, then by the time the TX
interrupt handler is called, the FIFO may already be full, and any
further writes are likely to be lost.
The fix involves adding a per-port flag that is true iff running from
within the interrupt handler and the spinlock has not yet been released.
This flag is then used as the value for the from_irq parameter of
pl011_tx_chars, causing polling to be used in the unsafe case.
Fixes:
1e84d22322ce ("serial/amba-pl011: Refactor and simplify TX FIFO handling")
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Dave Stevenson [Mon, 24 Jun 2019 01:29:40 +0000 (02:29 +0100)]
drm/vc4: Add support for color encoding on YUV planes
Adds signalling for BT601/709/2020, and limited/full range
(on BT601).
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Chris Miller [Wed, 26 Jun 2019 09:40:30 +0000 (10:40 +0100)]
drm: vc4_dsi: Fix DMA channel and memory leak in vc4 (#3012)
Signed-off-by: Chris G Miller <chris@creative-electronics.net>
Phil Elwell [Fri, 19 Jul 2019 13:29:28 +0000 (14:29 +0100)]
drm/vc4: Ignore HVS unless initialised
An upstream commit to report HVS underruns causes VC4 in firmware KMS
mode to cross into the HVS side, where it crashes due to a NULL hvs
pointer.
Make the underrun masking conditional on the hvs pointer being
initialised.
Fixes:
531a1b622da9 ("drm/vc4: Report HVS underrun errors")
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Dave Stevenson [Tue, 18 Jun 2019 20:37:45 +0000 (21:37 +0100)]
drm/vc4: Limit fkms to modes <= 85Hz
Selecting 1080p100 and 120 has very limited gain, but don't want
to block VGA85 and similar.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 12 Jun 2019 16:13:21 +0000 (17:13 +0100)]
drm/vc4: In FKMS look at the modifiers correctly for SAND
Incorrect masking was used in the switch for the modifier,
therefore for SAND (which puts the column pitch in the
modifier) it didn't match.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 4 Jun 2019 11:14:30 +0000 (12:14 +0100)]
drm: vc4: Add status of which display is updated through vblank
Previously multiple displays were slaved off the same SMI
interrupt, triggered by HVS channel 1 (HDMI0).
This doesn't work if you only have a DPI or DSI screen (HVS channel
0), and gives slightly erroneous results with dual HDMI as the
events for HDMI1 are incorrect.
Use SMIDSW0 and SMIDSW1 registers to denote which display has
triggered the vblank.
Handling should be backwards compatible with older firmware.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Mon, 10 Jun 2019 15:32:51 +0000 (16:32 +0100)]
drm/vc4: Remove 340MHz clock limit from FKMS now scrambling issues resolved
Firmware TMDS scrambling is now being correctly configured, so
we can use it.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 7 Jun 2019 10:31:21 +0000 (11:31 +0100)]
drm/vc4: Fix T-format modifiers in FKMS.
The wrong vc_image formats were being checked for in the switch
statement. Correct these.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Thu, 30 May 2019 14:55:15 +0000 (15:55 +0100)]
drm/vc4: Max resolution of 7680 is conditional on being Pi4
The max resolution had been increased from 2048 to 7680 for all
platforms. This code is common with Pi0-3 which have a max render
target for GL of 2048, therefore the increased resolution has to
be conditional on the platform.
Switch based on whether the bcm2835-v3d node is found, as that is
not present on Pi4. (There is a potential configuration on Pi0-3
with no v3d, but this is very unlikely).
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Thu, 30 May 2019 12:56:15 +0000 (13:56 +0100)]
drm/vc4: fkms to query the VPU for HDMI clock limits
The VPU has configured clocks for 4k (or not) via config.txt,
and will limit the choice of video modes based on that.
Make fkms query it for these limits too to avoid selecting modes
that can not be handled by the current clock setup.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 29 May 2019 14:44:11 +0000 (15:44 +0100)]
drm/vc4: Correct SAND support for FKMS.
It was accepting NV21 which doesn't map through, but
also wasn't advertising the modifier so nothing would know
to request it.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 7 May 2019 14:00:02 +0000 (15:00 +0100)]
drm: vc4: Fixup typo when setting HDMI aspect ratio
Assignment was to the wrong structure.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 24 May 2019 16:59:01 +0000 (17:59 +0100)]
drm/vc4: Support the VEC in FKMS
Extends the DPI/DSI support to also report the VEC output
which supports interlacing too.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Jonathan Bell [Tue, 28 May 2019 12:56:06 +0000 (13:56 +0100)]
drm: vc4: handle the case where there are no available displays
It's reasonable for the firmware to return zero as the number of
attached displays. Handle this case as otherwise drm thinks that
the DSI panel is attached, which is nonsense.
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.org>
Dave Stevenson [Tue, 21 May 2019 10:50:00 +0000 (11:50 +0100)]
drm: vc4: Probe DPI/DSI timings from the firmware
For DPI and DSI displays query the firmware as to the configuration
and add it as the only mode for DRM.
In theory we can add plumbing for setting the DPI/DSI mode from
KMS, but this is not being added at present as the support frameworks
aren't present in the firmware.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Thu, 16 May 2019 16:49:42 +0000 (17:49 +0100)]
drm: vc4-firmware-kms: Fix DSI display support
The mode was incorrectly listed as interlaced, which was then
rejected.
Correct this and FKMS works with the DSI display.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 7 May 2019 11:13:34 +0000 (12:13 +0100)]
drm: vc4: Log flags in fkms mode set
The flags contain info such as limited/full range RGB, aspect
ratio, and a fwe other useful things.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 3 May 2019 12:58:03 +0000 (13:58 +0100)]
drm: vc4-firmware-kms: Remove incorrect overscan support.
The overscan support was required for the old mailbox API
in order to match up the cursor and frame buffer planes.
With the newer API directly talking to dispmanx there is no
difference, therefore FKMS does not need to make any
adjustments.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 9 Apr 2019 17:23:41 +0000 (18:23 +0100)]
drm: vc4: FKMS reads the EDID from fw, and supports mode setting
This extends FKMS to read the EDID from the display, and support
requesting a particular mode via KMS.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Mon, 29 Apr 2019 17:45:00 +0000 (18:45 +0100)]
drm: vc4: Increase max_width/height to 7680.
There are some limits still being investigated that stop
us going up to 8192, but 7680 is sufficient for dual 4k
displays.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 10 Apr 2019 16:43:57 +0000 (17:43 +0100)]
drm: vc4: Bring fkms into line with kms in blocking doublescan modes
Implement vc4_crtc_mode_valid so that it blocks doublescan modes
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 10 Apr 2019 16:42:37 +0000 (17:42 +0100)]
drm: vc4: Iterate over all planes in vc4_crtc_[dis|en]able
Fixes a FIXME where the overlay plane wouldn't be restored.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 10 Apr 2019 16:35:05 +0000 (17:35 +0100)]
drm: vc4: Remove unused vc4_fkms_cancel_page_flip function
"32a3dbe drm/vc4: Nuke preclose hook" removed vc4_cancel_page_flip,
but vc4_fkms_cancel_page_flip was still be added to with the
fkms driver, even though it was never called.
Nuke it too.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 9 Apr 2019 16:19:51 +0000 (17:19 +0100)]
drm: vc4: Add support for H & V flips on each plane for FKMS
They are near zero cost options for the HVS, therefore they
may as well be implemented, and it allows us to invert the
DSI display.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 9 Apr 2019 17:14:44 +0000 (18:14 +0100)]
drm: vc4: Need to call drm_crtc_vblank_[on|off] from vc4_crtc_[en|dis]able
vblank needs to be enabled and disabled by the driver to avoid the
DRM framework complaining in the kernel log.
vc4_fkms_disable_vblank needs to signal that we don't want vblank
callbacks too.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 9 Apr 2019 13:00:07 +0000 (14:00 +0100)]
drm/vc4: Set the display number when querying the display resolution
Without this the two displays got set to the same resolution.
(Requires a firmware bug fix to work).
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 9 Apr 2019 11:37:28 +0000 (12:37 +0100)]
drm: vc4: Query the display ID for each display in FKMS
Replace the hard coded list of display IDs for a mailbox call
that returns the display ID for each display that has been
detected.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 5 Apr 2019 16:24:20 +0000 (17:24 +0100)]
drm: vc4: Remove now unused structure.
Cleaning up structure that was unused after
fbb59a2 drm: vc4: Add an overlay plane to vc4-firmware-kms
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 5 Apr 2019 16:23:15 +0000 (17:23 +0100)]
drm: vc4: Select display to blank during initialisation
Otherwise the rainbow splash screen remained in the display list
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 5 Apr 2019 16:21:56 +0000 (17:21 +0100)]
drm: vc4: Fix build warning
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 3 Apr 2019 16:15:45 +0000 (17:15 +0100)]
drm: vc4: Add support for multiple displays to fkms
There is a slightly nasty hack in that all crtcs share the
same SMI interrupt from the firmware. This seems to currently
work well enough, but ought to be fixed at a later date.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 3 Apr 2019 14:20:05 +0000 (15:20 +0100)]
drm: vc4: Increase max screen size to 4096x4096.
We now should support 4k screens, therefore this limit needs to
be increased.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 27 Mar 2019 17:45:01 +0000 (17:45 +0000)]
drm: vc4: Add an overlay plane to vc4-firmware-kms
This uses a new API that is exposed via the mailbox service
to stick an element straight on the screen using DispmanX.
The primary and cursor planes have also been switched to using
the new plane API, and it supports layering based on the DRM
zpos parameter.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 26 Mar 2019 14:43:06 +0000 (14:43 +0000)]
gpu: vc4-fkms: Switch to the newer mailbox frame buffer API.
The old mailbox FB API was ideally deprecated but still used by
the FKMS driver.
Update to the newer API.
NB This needs current firmware that accepts ARM allocated buffers
through the newer API.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Eric Anholt [Tue, 2 Apr 2019 20:29:00 +0000 (13:29 -0700)]
drm/vc4: Fix vblank timestamping for firmwarekms.
The core doesn't expect a false return from the scanoutpos function in
normal usage, so we were doing the precise vblank timestamping path
and thus "immediate" vblank disables (even though firmwarekms can't
actually disable vblanks interrupts, sigh), and the kernel would get
confused when getting timestamp info when also turning vblanks back
on.
Signed-off-by: Eric Anholt <eric@anholt.net>
Eric Anholt [Mon, 18 Mar 2019 23:38:32 +0000 (16:38 -0700)]
drm/vc4: Expose the format modifiers for firmware kms.
This should technically not expose VC4_T_TILED on pi4. However, if we
don't expose anything, then userspace will assume that display can
handle whatever modifiers 3d can do (UIF on 2711). By exposing a
list, that will get intersected with what 3D can do so that we get T
tiling for display on 2710 and linear on 2711.
Signed-off-by: Eric Anholt <eric@anholt.net>
Eric Anholt [Thu, 28 Mar 2019 18:58:51 +0000 (11:58 -0700)]
drm/vc4: Fix synchronization firmwarekms against GL rendering.
We would present the framebuffer immediately without waiting for
rendering to finish first, resulting in stuttering and flickering as a
window was dragged around when the GPU was busy enough to not just win
the race.
Signed-off-by: Eric Anholt <eric@anholt.net>
Eric Anholt [Mon, 14 Jan 2019 22:47:57 +0000 (14:47 -0800)]
drm/v3d: Hook up the runtime PM ops.
In translating the runtime PM code from vc4, I missed the ".pm"
assignment to actually connect them up. Fixes missing MMU setup if
runtime PM resets V3D.
Signed-off-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit
ca197699af29baa8236c74c53d4904ca8957ee06)
Eric Anholt [Mon, 14 Jan 2019 20:35:43 +0000 (12:35 -0800)]
drm/v3d: Skip MMU flush if the device is currently off.
If it's off, we know it will be reset on poweron, so the MMU won't
have any TLB cached from before this point. Avoids failed waits for
MMU flush to reply.
Signed-off-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit
3ee4e2e0a9e9587eacbb69b067bbc72ab2cdc47b)
Eric Anholt [Fri, 5 Oct 2018 00:22:43 +0000 (17:22 -0700)]
drm/v3d: Add support for 2711.
Signed-off-by: Eric Anholt <eric@anholt.net>
Eric Anholt [Mon, 4 Mar 2019 19:59:34 +0000 (11:59 -0800)]
drm/vc4: Fix oops at boot with firmwarekms on 4.19.
Signed-off-by: Eric Anholt <eric@anholt.net>
Phil Elwell [Tue, 11 Jun 2019 16:38:28 +0000 (17:38 +0100)]
arm: bcm2835: Add bcm2838 compatible string.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Jonathan Bell [Tue, 11 Jun 2019 10:42:03 +0000 (11:42 +0100)]
usbhid: call usb_fixup_endpoint after mangling intervals
Lets the mousepoll override mechanism work with xhci.
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.org>
Jonathan Bell [Tue, 11 Jun 2019 10:33:39 +0000 (11:33 +0100)]
xhci: implement xhci_fixup_endpoint for interval adjustments
Must be called in a non-atomic context, after the endpoint
has been registered with the hardware via xhci_add_endpoint
and before the first URB is submitted for the endpoint.
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.org>
Jonathan Bell [Tue, 11 Jun 2019 09:55:00 +0000 (10:55 +0100)]
usb: add plumbing for updating interrupt endpoint interval state
xHCI caches device and endpoint data after the interface is configured,
so an explicit command needs to be issued for any device driver wanting
to alter the polling interval of an endpoint.
Add usb_fixup_endpoint() to allow drivers to do this. The fixup must be
called after calculating endpoint bandwidth requirements but before any
URBs are submitted.
If polling intervals are shortened, any bandwidth reservations are no
longer valid but in practice polling intervals are only ever relaxed.
Limit the scope to interrupt transfers for now.
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.org>
Eric Anholt [Thu, 2 May 2019 22:24:04 +0000 (15:24 -0700)]
clk: bcm2835: Allow reparenting leaf clocks while they're running.
This falls under the same "we can reprogram glitch-free as long as we
pause generation" rule as updating the div/frac fields. This can be
used for runtime reclocking of V3D to manage power leakage.
Signed-off-by: Eric Anholt <eric@anholt.net>
Eric Anholt [Thu, 2 May 2019 22:11:05 +0000 (15:11 -0700)]
clk: bcm2835: Add support for setting leaf clock rates while running.
As long as you wait for !BUSY, you can do glitch-free updates of clock
rate while the clock is running.
Signed-off-by: Eric Anholt <eric@anholt.net>
Phil Elwell [Tue, 4 Sep 2018 10:50:25 +0000 (11:50 +0100)]
config: Permit LPAE and PCIE_BRCMSTB on BCM2835
Eric Anholt [Sat, 12 Jan 2019 01:31:07 +0000 (17:31 -0800)]
soc: bcm: bcm2835-pm: Add support for 2711.
Without the actual power management part any more, there's a lot less
to set up for V3D. We just need to clear the RSTN field for the power
domain, and expose the reset controller for toggling it again.
This is definitely incomplete -- the old ISP and H264 is in the old
bridge, but since we have no consumers of it I've just done the
minimum to get V3D working.
Signed-off-by: Eric Anholt <eric@anholt.net>
Phil Elwell [Wed, 23 Jan 2019 16:11:50 +0000 (16:11 +0000)]
clk-bcm2835: Don't wait for pllh lock
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Martin Sperl [Mon, 13 May 2019 11:05:27 +0000 (11:05 +0000)]
spi: bcm2835: enable shared interrupt support
Add shared interrupt support for this driver.
Signed-off-by: Martin Sperl <kernel@martin.sperl.org>
Phil Elwell [Thu, 23 May 2019 14:08:30 +0000 (15:08 +0100)]
usb: xhci: Show that the VIA VL805 supports LPM
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Tim Gover [Fri, 22 Mar 2019 09:47:14 +0000 (09:47 +0000)]
usb: xhci: Disable the XHCI 5 second timeout
If the VL805 EEPROM has not been programmed then boot will hang for five
seconds. The timeout seems to be arbitrary and is an unecessary
delay on the first boot. Remove the timeout.
This is common code and probably can't be upstreamed unless the timeout
can be configurable somehow or perhaps the XHCI driver can be skipped
on the first boot.
Jonathan Bell [Fri, 17 May 2019 12:31:21 +0000 (13:31 +0100)]
phy: bcm54213pe: configure the LED outputs to be more user-friendly
The default state was both LEDs indicating link speed.
Change the default configuration to
- Amber: 1000/100 link speed indication
- Green: link present + activity indication
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.org>
Jonathan Bell [Tue, 14 May 2019 16:00:41 +0000 (17:00 +0100)]
phy: broadcom: split out the BCM54213PE from the BCM54210E IDs
The last nibble is a revision ID, and the 54213pe is a later rev
than the 54210e. Running the 54210e setup code on a 54213pe results
in a broken RGMII interface.
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.org>
Jonathan Bell [Tue, 14 May 2019 16:17:59 +0000 (17:17 +0100)]
net: genet: enable link energy detect powerdown for external PHYs
There are several warts surrounding bcmgenet_mii_probe() as this
function is called from ndo_open, but it's calling registration-type
functions. The probe should be called at probe time and refactored
such that the PHY device data can be extracted to limit the scope
of this flag to Broadcom PHYs.
For now, pass this flag in as it puts our attached PHY into a low-power
state when disconnected.
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.org>
Phil Elwell [Wed, 27 Mar 2019 13:45:46 +0000 (13:45 +0000)]
bcmgenet: Better coalescing parameter defaults
Set defaults for TX and RX packet coalescing to be equivalent to:
# ethtool -C eth0 tx-frames 10
# ethtool -C eth0 rx-usecs 50
This may be something we want to set via DT parameters in the
future.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Jonathan Bell [Wed, 12 Sep 2018 13:44:53 +0000 (14:44 +0100)]
bcmgenet: constrain max DMA burst length
popcornmix [Tue, 30 Apr 2019 18:15:30 +0000 (19:15 +0100)]
bcm2835-pcm.c: Support multichannel audio
Phil Elwell [Thu, 1 Nov 2018 17:31:37 +0000 (17:31 +0000)]
vchiq: Add 36-bit address support
Conditional on a new compatible string, change the pagelist encoding
such that the top 24 bits are the pfn, leaving 8 bits for run length
(-1).
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Stefan Wahren [Sat, 18 May 2019 10:26:11 +0000 (12:26 +0200)]
thermal: brcmstb_thermal: Add BCM2838 support
The BCM2838 has an AVS TMON hardware block. This adds the necessary
support to the brcmstb_thermal driver ( no trip handling ).
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Stefan Wahren [Sat, 4 May 2019 15:06:15 +0000 (17:06 +0200)]
hwrng: iproc-rng200: Add BCM2838 support
The HWRNG on the BCM2838 is compatible to iproc-rng200, so add the
support to this driver instead of bcm2835-rng.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Phil Elwell [Fri, 28 Sep 2018 15:24:05 +0000 (16:24 +0100)]
mmc: sdhci: Mask "spurious" interrupts
Add a filter for "spurious" Transfer Complete interrupts, attempting
to make it as specific as possible:
* INT_DATA_END (transfer complete) is set
* There is a stop command in progress
* There is no data transfer in progress
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Phil Elwell [Wed, 29 May 2019 14:47:42 +0000 (15:47 +0100)]
arm: bcm2835: DMA can only address 1GB
The legacy peripherals can only address the first gigabyte of RAM, so
ensure that DMA allocations are restricted to that region.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Phil Elwell [Tue, 19 Feb 2019 22:06:59 +0000 (22:06 +0000)]
pcie-brcmstb: Changes for BCM2711
The initial brcmstb PCIe driver - originally taken from the V3(?)
patch set - has been modified significantly for the BCM2711.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Jim Quinlan [Mon, 15 Jan 2018 23:28:39 +0000 (18:28 -0500)]
dt-bindings: pci: Add DT docs for Brcmstb PCIe device
The DT bindings description of the Brcmstb PCIe device is described. This
node can be used by almost all Broadcom settop box chips, using
ARM, ARM64, or MIPS CPU architectures.
Signed-off-by: Jim Quinlan <jim2101024@gmail.com>
Phil Elwell [Tue, 19 Feb 2019 22:06:59 +0000 (22:06 +0000)]
PCI: brcmstb: Add MSI capability
This commit adds MSI to the Broadcom STB PCIe host controller. It does
not add MSIX since that functionality is not in the HW. The MSI
controller is physically located within the PCIe block, however, there
is no reason why the MSI controller could not be moved elsewhere in
the future.
Since the internal Brcmstb MSI controller is intertwined with the PCIe
controller, it is not its own platform device but rather part of the
PCIe platform device.
Signed-off-by: Jim Quinlan <jim2101024@gmail.com>
Phil Elwell [Tue, 19 Feb 2019 22:06:59 +0000 (22:06 +0000)]
PCI: brcmstb: Add dma-range mapping for inbound traffic
The Broadcom STB PCIe host controller is intimately related to the
memory subsystem. This close relationship adds complexity to how cpu
system memory is mapped to PCIe memory. Ideally, this mapping is an
identity mapping, or an identity mapping off by a constant. Not so in
this case.
Consider the Broadcom reference board BCM97445LCC_4X8 which has 6 GB
of system memory. Here is how the PCIe controller maps the
system memory to PCIe memory:
memc0-a@[ 0....
3fffffff] <=> pci@[ 0....
3fffffff]
memc0-b@[
100000000...
13fffffff] <=> pci@[
40000000....
7fffffff]
memc1-a@[
40000000....
7fffffff] <=> pci@[
80000000....
bfffffff]
memc1-b@[
300000000...
33fffffff] <=> pci@[
c0000000....
ffffffff]
memc2-a@[
80000000....
bfffffff] <=> pci@[
100000000...
13fffffff]
memc2-b@[
c00000000...
c3fffffff] <=> pci@[
140000000...
17fffffff]
Although there are some "gaps" that can be added between the
individual mappings by software, the permutation of memory regions for
the most part is fixed by HW. The solution of having something close
to an identity mapping is not possible.
The idea behind this HW design is that the same PCIe module can
act as an RC or EP, and if it acts as an EP it concatenates all
of system memory into a BAR so anything can be accessed. Unfortunately,
when the PCIe block is in the role of an RC it also presents this
"BAR" to downstream PCIe devices, rather than offering an identity map
between its system memory and PCIe space.
Suppose that an endpoint driver allocs some DMA memory. Suppose this
memory is located at 0x6000_0000, which is in the middle of memc1-a.
The driver wants a dma_addr_t value that it can pass on to the EP to
use. Without doing any custom mapping, the EP will use this value for
DMA: the driver will get a dma_addr_t equal to 0x6000_0000. But this
won't work; the device needs a dma_addr_t that reflects the PCIe space
address, namely 0xa000_0000.
So, essentially the solution to this problem must modify the
dma_addr_t returned by the DMA routines routines. There are two
ways (I know of) of doing this:
(a) overriding/redefining the dma_to_phys() and phys_to_dma() calls
that are used by the dma_ops routines. This is the approach of
arch/mips/cavium-octeon/dma-octeon.c
In ARM and ARM64 these two routines are defiend in asm/dma-mapping.h
as static inline functions.
(b) Subscribe to a notifier that notifies when a device is added to a
bus. When this happens, set_dma_ops() can be called for the device.
This method is mentioned in:
http://lxr.free-electrons.com/source/drivers/of/platform.c?v=3.16#L152
where it says as a comment
"In case if platform code need to use own special DMA
configuration, it can use Platform bus notifier and
handle BUS_NOTIFY_ADD_DEVICE event to fix up DMA
configuration."
Solution (b) is what this commit does. It uses its own set of
dma_ops which are wrappers around the arch_dma_ops. The
wrappers translate the dma addresses before/after invoking
the arch_dma_ops, as appropriate.
Signed-off-by: Jim Quinlan <jim2101024@gmail.com>
Phil Elwell [Tue, 19 Feb 2019 22:06:59 +0000 (22:06 +0000)]
PCI: brcmstb: Add Broadcom STB PCIe host controller driver
This commit adds the basic Broadcom STB PCIe controller. Missing is
the ability to process MSI and also handle dma-ranges for inbound
memory accesses. These two functionalities are added in subsequent
commits.
The PCIe block contains an MDIO interface. This is a local interface
only accessible by the PCIe controller. It cannot be used or shared
by any other HW. As such, the small amount of code for this
controller is included in this driver as there is little upside to put
it elsewhere.
Signed-off-by: Jim Quinlan <jim2101024@gmail.com>
Tim Gover [Thu, 14 Mar 2019 10:16:02 +0000 (10:16 +0000)]
Fix copy_from_user if BCM2835_FAST_MEMCPY=n
The change which introduced CONFIG_BCM2835_FAST_MEMCPY unconditionally
changed the behaviour of arm_copy_from_user. The page pinning code
is not safe on ARMv7 if LPAE & high memory is enabled and causes
crashes which look like PTE corruption.
Make __copy_from_user_memcpy conditional on CONFIG_2835_FAST_MEMCPY=y
which is really an ARMv6 / Pi1 optimization and not necessary on newer
ARM processors.
Phil Elwell [Wed, 20 Feb 2019 08:49:39 +0000 (08:49 +0000)]
arm: bcm2835: Fix FIQ early ioremap
The ioremapping creates mappings within the vmalloc area. The
equivalent early function, create_mapping, now checks that the
requested explicit virtual address is between VMALLOC_START and
VMALLOC_END. As there is no reason to have any correlation between
the physical and virtual addresses, put the required mappings at
VMALLOC_START and above.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Phil Elwell [Wed, 12 Jun 2019 16:15:05 +0000 (17:15 +0100)]
w1: w1-gpio: Make GPIO an output for strong pullup
The logic to drive the data line high to implement a strong pullup
assumed that the pin was already an output - setting a value does
not change an input.
See: https://github.com/raspberrypi/firmware/issues/1143
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Dave Stevenson [Fri, 10 May 2019 13:13:11 +0000 (14:13 +0100)]
staging: bcm2835-codec: Add support for setting S_PARM and G_PARM
Video encode can use the frame rate for rate control calculations,
therefore plumb it through from V4L2's [S|G]_PARM ioctl.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 10 May 2019 13:11:58 +0000 (14:11 +0100)]
staging: bcm2835-codec: Convert V4L2 nsec timestamps to MMAL usec
V4L2 uses nsecs, whilst MMAL uses usecs, but the code wasn't converting
between them. This upsets video encode rate control.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Marcel Holtmann [Wed, 22 May 2019 07:05:40 +0000 (09:05 +0200)]
Bluetooth: Check key sizes only when Secure Simple Pairing is enabled
The encryption is only mandatory to be enforced when both sides are using
Secure Simple Pairing and this means the key size check makes only sense
in that case.
On legacy Bluetooth 2.0 and earlier devices like mice the encryption was
optional and thus causing an issue if the key size check is not bound to
using Secure Simple Pairing.
Fixes:
d5bb334a8e17 ("Bluetooth: Align minimum encryption key size for LE and BR/EDR connections")
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Cc: stable@vger.kernel.org
Dave Stevenson [Thu, 2 May 2019 14:50:01 +0000 (15:50 +0100)]
staging: mmal-vchiq: Fix memory leak in error path
On error, vchiq_mmal_component_init could leave the
event context allocated for ports.
Clean them up in the error path.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 1 May 2019 12:27:23 +0000 (13:27 +0100)]
staging: mmal-vchiq: Free the event context for control ports
vchiq_mmal_component_init calls init_event_context for the
control port, but vchiq_mmal_component_finalise didn't free
it, causing a memory leak..
Add the free call.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Thu, 2 May 2019 13:32:21 +0000 (14:32 +0100)]
staging: bcm2835-codec: Remove height padding for ISP role
The ISP has no need for heights to be a multiple of macroblock
sizes, therefore doesn't require the align on the height.
Remove it for the ISP role. (It is required for the codecs).
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Thu, 2 May 2019 13:30:24 +0000 (14:30 +0100)]
staging: bcm2835-codec: Correct port width calc for truncation
The calculation converting from V4L2 bytesperline to MMAL
width had an operator ordering issue that lead to Bayer raw 10
(and 12 and 14) setting an incorrect stride for the buffer.
Correct this operation ordering issue.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Wed, 20 Mar 2019 10:06:51 +0000 (10:06 +0000)]
staging: bcm2835-codec: Refactor default resolution code
The default resolution code was different for each role
as compressed formats need to pass bytesperline as 0 and
set up customised buffer sizes.
This is common setup, therefore amend get_sizeimage and
get_bytesperline to do the correct thing whether compressed
or uncompressed.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 8 Mar 2019 11:26:00 +0000 (11:26 +0000)]
staging: bcm2835_codec: Clean up logging on unloading the driver
The log line was missing a closing \n, so wasn't added to the
log immediately.
Adds the function of the V4L2 device that is being unregistered
too.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Tue, 19 Mar 2019 17:55:09 +0000 (17:55 +0000)]
staging: bcm2835-codec: NULL component handle on queue_setup failure
queue_setup tries creating the relevant MMAL component and configures
the input and output ports as we're expecting to start streaming.
If the port configuration failed then it destroyed the component,
but failed to clear the component handle, therefore release tried
destroying the component again.
Adds some logging should the port config fail as well.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Mon, 18 Feb 2019 15:56:42 +0000 (15:56 +0000)]
staging: bcm2835_codec: Include timing info in SPS headers
Inserting timing information into the VUI block of the SPS is
optional with the VPU encoder.
GStreamer appears to require them when using V4L2 M2M, therefore
set the option to enable them from the encoder.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Mon, 18 Feb 2019 15:52:29 +0000 (15:52 +0000)]
staging: mmal-vchiq: Update mmal_parameters.h with recently defined params
mmal_parameters.h hasn't been updated to reflect additions made
over the last few years. Update it to reflect the currently
supported parameters.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Dave Stevenson [Fri, 15 Feb 2019 11:38:45 +0000 (11:38 +0000)]
staging: bcm2835_codec: Fix handling of VB2_MEMORY_DMABUF buffers
If the queue is configured as VB2_MEMORY_DMABUF then vb2_core_expbuf
fails as it ensures the queue is defined as VB2_MEMORY_MMAP.
Correct the handling so that we unmap the buffer from vcsm and the
VPU on cleanup, and then correctly get the dma buf of the new buffer.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>