- Jun 18, 2020
-
-
Phil Elwell authored
Fixes: "dts: bcm2711: set #size-cells = <2>" Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Stefan Wahren authored
This adds the labels for all the SoC GPIOs on the Raspberry Pi 4. Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com> Reviewed-by: Linus Walleij <linus.walleij@linaro.org> Link: https://lore.kernel.org/r/1581166975-22949-5-git-send-email-stefan.wahren@i2se.com Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
-
Hristo Venev authored
The following addresses change: - 0xfc00_0000 -> 0x4_7c00_0000 - 0xff80_0000 -> 0x4_c000_0000 The range 0xfc00_0000-0xffff_ffff becomes available as system RAM on devices with >= 4 GiB of RAM. Firmware should initialize the memory node appropriately. Signed-off-by: Hristo Venev <hristo@venev.name>
-
Hristo Venev authored
There already is one 4 GiB range, and one more will appear when high peripheral mode is enabled. Signed-off-by: Hristo Venev <hristo@venev.name>
-
Hristo Venev authored
Physical addresses in the 0x0-0xfc00_0000 range correspond to SDRAM. They have no place here. Signed-off-by: Hristo Venev <hristo@venev.name>
-
Phil Elwell authored
Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Phil Elwell authored
Although 0xfc000000 looks like an inaccessible RAM address (due to the peripheral mappings), with RAM mapped at 0xc0000000 (as it is on the 30/32-bit VPU bus) this is actually 0x3c000000 in the ARM memory space, which is fine. This difference is potentially the cause of some warnings seen in sdhci_send_command. Fixes: "dts: bcm2711: Move emmc2 to its own 'bus'" Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Panagiotis Vasilopoulos authored
-
Jörg Schambacher authored
This patch fixes a problem of the re-calculation of i2s-clock and -parameter settings when only the ADC is activated. Signed-off-by: Joerg Schambacher <joerg@i2audio.com>
-
James Hilliard authored
If the vc4 hdmi driver loads before the pixel clock is available we see a spurious "*ERROR* Failed to get pixel clock" error. Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
-
Phil Elwell authored
Moving the EMMC2 controller under a dedicated bus allows the firmware to patch the dma-ranges property for different memory sizes without affecting anything else. Signed-off-by: Phil Elwell <phil@raspberrypi.org>
-
Phil Elwell authored
Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Phil Elwell authored
The downstream 32-bit 2711 kernel configuration enables HIGHMEM for access to more physical RAM. The HIGHMEM zone starts at 0x30000000 (768MB), and allowing the CMA zone to overlap that area causes a failure during CMA activation. Avoid the overlap by limiting CMA to the first 768MB. This is overly restrictive on a 64-bit kernel, but shouldn't cause any practical problems. Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
AMuszkat authored
Signed-off-by: AMuszkat <ariel.muszkat@gmail.com>
-
AMuszkat authored
Add 96KHz rate support to MA120X0P codec and make enable and mute gpio pins optional. Signed-off-by: AMuszkat <ariel.muszkat@gmail.com>
-
Matthias Reichl authored
Standard IR receivers use inverted / active-low signalling and the gpio-ir overlay configures the GPIO appropriately as GPIO_ACTIVE_LOW (1). In order to support (rather rare) non-inverted / active-high signalling the GPIO needs to be configured as GPIO_ACTIVE_HIGH (0). Add an "invert" parameter to override this like in the gpio-ir-tx overlay. Signed-off-by: Matthias Reichl <hias@horus.com>
-
Phil Elwell authored
The GCC plugin feature leads to different kernel configurations on what ought to be equivalent build systems because they depend on the build hosts native compilers rather than the cross compilers needed for the target. This causes problems with module symbol version mismatches. Disable GCC plugins for all build hosts. Advanced build script hackery borrowed from a patch by milhouse. Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Phil Elwell authored
Better late than never. Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Matt Flax authored
This patch adds support for the Audio Injector Isolated sound card. Signed-off-by: Matt Flax <flatmax@flatmax.org>
-
Phil Elwell authored
The filesystem-specific encryption options have been replaced by a generic FS_ENCRYPTION option. Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
popcornmix authored
For performance/power it is beneficial to adjust gpu clocks with arm clock. This is how the downstream cpufreq driver works Signed-off-by: popcornmix <popcornmix@gmail.com>
-
Phil Elwell authored
Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Phil Elwell authored
Because the upstream overlay applies vc4-kms-v3d, of which Pi 4 has its own version, there also needs to be a Pi 4 version - vc4-kms-v3d-pi4. Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Phil Elwell authored
Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Phil Elwell authored
Take advantage of the overlay_map to rename or deprecate some obsolete overlays. Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Phil Elwell authored
The overlay map permits platform-specific overlays, with deprecation and renaming. See: https://github.com/raspberrypi/linux/issues/3520 Signed-off-by: Phil Elwell <phil@raspberrypi.com>
-
Phil Elwell authored
-
popcornmix authored
The original firmware code worked with float and did offset = ((vco_freq / fref * 2) * (1 << 22)); offset >>= 2; In this code it's all integer so doing the integer divide before the shift loses lots of precision This fixes the issue of 1080p59.94 mode having 59.64 fps Signed-off-by: popcornmix <popcornmix@gmail.com>
-
Dave Stevenson authored
It was set to 390kHz, which is outside of the required spec for reading HDMI (max 100kHz). The i2c-brcmstb driver only supports a number of fixed bus speeds, of which 97.5kHz is the closest to 100kHz without exceeding it. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
If there is no interrupt defined then an error is logged due to the use of platform_get_irq. The driver handles not having the interrupt by falling back to polling, therefore make the appropriate call when claiming it. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
The debugfs usage within asoc gets confused if multiple interfaces have the same card name, therefore use unique names when initialising them. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
"drm/vc4: plane: Move additional planes creation to driver" moved overlay and cursor plane creation to a global function thata was unconditionally run, when it is not wanted in firmware KMS mode. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
Fix up "drm/vc4: Kick the core clock up during a mode change" for firmware KMS mode where we don't have the HVS or core clock configured. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
Experimental commit to kick the core clock up during mode switching. This makes mode switching far more reliable, and mimics what the firmware does. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
Using vc4-kms-v3d with X has worked for quite a while, and essentially required not using fbturbo and having an up to date MESA library. Remove the comment that says otherwise. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
Whilst the documentation for BCM2835 states that the HDMI state machine clock needs to be 108% of the pixel clock, other documentation says that it only has to be greater than the pixel clock. The firmware uses 101%, and that allows 1920x1200@60Hz to work within the constraint of the HSM clock being < 163.68MHz. Adopt 101%, and increase the maximum pixel clock for vc4 to 162MHz so that it too supports 1920x1200@60. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
This could be a revert of "drm/vc4: hdmi: Add an audio support flag" as it is no longer needed. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
The audio configuration has changed for Pi4, so support the configuration functions via the variant tables. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
HDMI audio configuration was using fixed index numbers to load in DT register settings. Switch to using reg-names for flexibility and to match Pi4. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-
Dave Stevenson authored
Adds the relevant DMA settings for HDMI audio to work. Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
-