When adding support for the DisplayPort part of the QMP PHY the binding
(and devicetree parser) for the (USB) child node was simply reused and
this has lead to some confusion.
The third DP register region is really the DP_PHY region, not "PCS" as
the binding claims, and lie at offset 0x2a00 (not 0x2c00).
Similarly, there likely are no "RX", "RX2" or "PCS_MISC" regions as
there are for the USB part of the PHY (and in any case the Linux driver
does not use them).
Note that the sixth "PCS_MISC" region is not even in the binding.
Fixes: 23737b9557 ("arm64: dts: qcom: sm6350: Add USB1 nodes")
Cc: stable@vger.kernel.org # 5.16
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Link: https://lore.kernel.org/r/20221111094729.11842-2-johan+linaro@kernel.org
This is not a fix on its own but more a cleanup. Phy qmp pcie driver
currently have a workaround to handle pcs_misc not declared and add
0x400 offset to the pcs reg if pcs_misc is not declared.
Correctly declare pcs_misc reg and reduce PCS size to the common value
of 0x1f0 as done for every other qmp based pcie phy device.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Reviewed-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Link: https://lore.kernel.org/r/20221103212125.17156-2-ansuelsmth@gmail.com
There are three UFS reference clocks on SC8280XP which are used as
follows:
- The GCC_UFS_REF_CLKREF_CLK clock is fed to any UFS device connected
to either controller.
- The GCC_UFS_1_CARD_CLKREF_CLK and GCC_UFS_CARD_CLKREF_CLK clocks
provide reference clocks to the two PHYs.
Note that this depends on first updating the clock driver to reflect
that all three clocks are sourced from CXO. Specifically, the UFS
controller driver expects the device reference clock to have a valid
frequency:
ufshcd-qcom 1d84000.ufs: invalid ref_clk setting = 0
Fixes: 152d1faf1e ("arm64: dts: qcom: add SC8280XP platform")
Fixes: 8d6b458ce6 ("arm64: dts: qcom: sc8280xp: fix ufs_card_phy ref clock")
Fixes: f3aa975e23 ("arm64: dts: qcom: sc8280xp: correct ref clock for ufs_mem_phy")
Link: https://lore.kernel.org/lkml/Y2OEjNAPXg5BfOxH@hovoldconsulting.com/
Cc: stable@vger.kernel.org # 5.20
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Reviewed-by: Brian Masney <bmasney@redhat.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@somainline.org>
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Link: https://lore.kernel.org/r/20221104092045.17410-2-johan+linaro@kernel.org
The devices on the SC8280XP PCIe buses are cache coherent and must be
marked as such to avoid data corruption.
A coherent device can, for example, end up snooping stale data from the
caches instead of using data written by the CPU through the
non-cacheable mapping which is used for consistent DMA buffers for
non-coherent devices.
Note that this is much more likely to happen since commit c44094eee3
("arm64: dma: Drop cache invalidation from arch_dma_prep_coherent()")
that was added in 6.1 and which removed the cache invalidation when
setting up the non-cacheable mapping.
Marking the PCIe devices as coherent specifically fixes the intermittent
NVMe probe failures observed on the Thinkpad X13s, which was due to
corruption of the submission and completion queues. This was typically
observed as corruption of the admin submission queue (with well-formed
completion):
could not locate request for tag 0x0
nvme nvme0: invalid id 0 completed on queue 0
or corruption of the admin or I/O completion queues (malformed
completion):
could not locate request for tag 0x45f
nvme nvme0: invalid id 25695 completed on queue 25965
presumably as these queues are small enough to not be allocated using
CMA which in turn make them more likely to be cached (e.g. due to
accesses to nearby pages through the cacheable linear map). Increasing
the buffer sizes to two pages to force CMA allocation also appears to
make the problem go away.
Fixes: 813e831570 ("arm64: dts: qcom: sc8280xp/sa8540p: add PCIe2-4 nodes")
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Link: https://lore.kernel.org/r/20221124142501.29314-1-johan+linaro@kernel.org
The 1-mic and 3-mic dtsi still had two minor cosmetic differences
after commit '3d11e7e120ee ("arm64: dts: qcom: sc7280: sort out the
"Status" to last property with
sc7280-herobrine-audio-rt5682.dtsi")'. Let's fix them so the two files
diff better. This is expected to have no effect though it will
slightly change the generated dtb by removing an unnecessary 'status =
"okay"' from the sound node.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Link: https://lore.kernel.org/r/20221114162807.1.I0900b97128f9bb03e5f96fcb3068c227a33f143a@changeid
The Qualcomm Snapdragon 670 has been out for a while. Add a device tree
for it and the Google Pixel 3a as the first device.
The Pixel 3a has the same bootloader issue as the Pixel 3 and will not work
on Android 10 bootloaders or later until it gets fixed for the Pixel 3.
SoC Initial Features:
- power management
- clocks
- pinctrl
- eMMC
- USB 2.0
- GENI I2C
- IOMMU
- RPMh
- interrupts
Device-Specific Initial Features:
- side buttons (keys)
- regulators
- touchscreen
Signed-off-by: Richard Acayan <mailingradian@gmail.com>
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Link: https://lore.kernel.org/r/20221111001818.124901-5-mailingradian@gmail.com
Add support for Sony Xperia 10 IV, a.k.a PDX225. This device is a part
of the SoMC SM6375 Murray platform and currently it is the only
device based on that board, so no -common DTSI is created until (if?)
other Murray devices appear.
This commit brings support for:
* USB (only USB2 for now)
* Display via simplefb
To create a working boot image, you need to run:
cat arch/arm64/boot/Image.gz arch/arm64/boot/dts/qcom/sm6375-sony-xperia-\
murray-pdx225.dtb > .Image.gz-dtb
mkbootimg \
--kernel .Image.gz-dtb \
--ramdisk some_initrd.img \
--pagesize 4096 \
--base 0x0 \
--kernel_offset 0x8000 \
--ramdisk_offset 0x1000000 \
--tags_offset 0x100 \
--cmdline "SOME_CMDLINE" \
--dtb_offset 0x1f00000 \
--header_version 1 \
--os_version 12 \
--os_patch_level 2022-04 \ # or newer
-o boot.img-sony-xperia-pdx225
Then, you need to flash it on the device and get rid of all the
vendor_boot/dtbo mess:
First, you need to get rid of vendor_boot. However, the bootloader
is utterly retarded and it will not let you neither flash nor erase it.
There are a couple ways to handle this: you can either dd /dev/zero to
it from Android (if you have root) or a custom recovery or from fastbootd
(fastboot/adb reboot fastboot). You will not be able to boot Android
images on your phone unless you lock the bootloader (fastboot oem lock)
and restore the factory image with Xperia Companion
Windows-and-macOS-only software.
The best way so far is probably to use the second (_b) slot and flash
mainline there. This will however require you to flash some partitions
manually, as they are not populated from factory:
(boot_b, dtbo_b, vendor_boot_b, vbmeta_b, vbmeta_system_b) - these we
don't really care about as we nuke/replace them
(dsp_b, imagefv_b, modem_b, oem_b, rdimage_b) - these you NEED to populate
to get a successful boot on slot B, otherwise you will have limited / no
functionality.
To switch slots, simply run:
fastboot --set-active=a //or =b
The rest assumes you are on slot A.
// You have to either pull vbmeta{"","_system"} from
// /dev/block/bootdevice/by-name/ or build one as a part of AOSP
fastboot --disable-verity --disable-verification flash vbmeta_b vbmeta.img
fastboot --disable-verity --disable-verification flash vbmeta_system_b \
vbmeta_system.img
fastboot flash boot_b boot.img-sony-xperia-pdx225
fastboot reboot fastboot // entering fastbootd
fastboot flash vendor_boot_b emptything.img
fastboot flash dtbo_b emptything.img
fastboot reboot bootloader // entering bootloader fastboot
fastboot --set-active=b
fastboot reboot // mainline time!
Where emptything.img is a tiny file that consists of 2 bytes (all zeroes),
doing a "fastboot erase" won't cut it, the bootloader will go crazy and
things will fall apart when it tries to overlay random bytes from an empty
partition onto a perfectly good appended DTB.
From there on you can flash new mainline builds by simply flashing
boot.img that you create after each kernel rebuild.
Signed-off-by: Konrad Dybcio <konrad.dybcio@somainline.org>
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Link: https://lore.kernel.org/r/20221107120920.12593-4-konrad.dybcio@linaro.org
