On 7/27/2023 5:49 PM, Tengfei Fan wrote:
在 7/27/2023 9:08 PM, Konrad Dybcio 写道:
On 27.07.2023 10:49, Tengfei Fan wrote:
在 7/27/2023 3:59 PM, Konrad Dybcio 写道:
On 27.07.2023 08:56, Krzysztof Kozlowski wrote:
On 27/07/2023 04:35, Tengfei Fan wrote:
Add DTS for Qualcomm QRD platform which uses SM4450 SoC.
Signed-off-by: Tengfei Fan <quic_tengfan@xxxxxxxxxxx>
---
arch/arm64/boot/dts/qcom/Makefile | 1 +
arch/arm64/boot/dts/qcom/sm4450-qrd.dts | 18 ++++++++++++++++++
2 files changed, 19 insertions(+)
create mode 100644 arch/arm64/boot/dts/qcom/sm4450-qrd.dts
diff --git a/arch/arm64/boot/dts/qcom/Makefile
b/arch/arm64/boot/dts/qcom/Makefile
index 337abc4ceb17..db805d0929c8 100644
--- a/arch/arm64/boot/dts/qcom/Makefile
+++ b/arch/arm64/boot/dts/qcom/Makefile
@@ -186,6 +186,7 @@ dtb-$(CONFIG_ARCH_QCOM) +=
sdm850-lenovo-yoga-c630.dtb
dtb-$(CONFIG_ARCH_QCOM) += sdm850-samsung-w737.dtb
dtb-$(CONFIG_ARCH_QCOM) += sdx75-idp.dtb
dtb-$(CONFIG_ARCH_QCOM) += sm4250-oneplus-billie2.dtb
+dtb-$(CONFIG_ARCH_QCOM) += sm4450-qrd.dtb
dtb-$(CONFIG_ARCH_QCOM) += sm6115-fxtec-pro1x.dtb
dtb-$(CONFIG_ARCH_QCOM) += sm6115p-lenovo-j606f.dtb
dtb-$(CONFIG_ARCH_QCOM) += sm6125-sony-xperia-seine-pdx201.dtb
diff --git a/arch/arm64/boot/dts/qcom/sm4450-qrd.dts
b/arch/arm64/boot/dts/qcom/sm4450-qrd.dts
new file mode 100644
index 000000000000..04ad1dd4285a
--- /dev/null
+++ b/arch/arm64/boot/dts/qcom/sm4450-qrd.dts
@@ -0,0 +1,18 @@
+// SPDX-License-Identifier: BSD-3-Clause
+/*
+ * Copyright (c) 2023, Qualcomm Innovation Center, Inc. All
rights reserved.
+ */
+
+/dts-v1/;
+
+#include "sm4450.dtsi"
+/ {
+ model = "Qualcomm Technologies, Inc. SM4450 QRD";
+ compatible = "qcom,sm4450-qrd", "qcom,sm4450";
+
+ aliases { };
+
+ chosen {
+ bootargs = "console=hvc0 earlycon=hvc0 hvc_dcc.enable=1
cpuidle.off=1";
No earlycon, no hvc.enable (there is no such parameter), no
cpuidle.off
(again don't add fake stuff). So the only suitable argument is
console,
but this should be actually used via stdout path, although it seems
there is no device node for such usage?
hvc totally comes from Qualcomm downstream and all the Gunyah
shenanigans..
Tengfei, please ensure the patches are tested against linux-next
with no additional changes or modules, with a clean Linux userspace
(or at least a ramdisk).
Konrad
Hi Konrad,
All these patches are tesed on linux-next, and get DCC console.
For support this test, I did a temporal for enable
CONFIG_HVC_DCC(this change haven't push to upstream).
So what you've said is that you've not tested against linux-next +
this patchset.
You're supposed to `git checkout linux-next/master && (pick the series
you're sending)` and test just that. No less, no more. Preferably with
the upstream arm64 defconfig.
Konrad
Hi Konrad,
Sorry, my explain isn't enough before.
My tested against linux-next + this patchset, all my test code from
upstream linux-next, haven't downstream code related.
About the temporal patch which enable CONFIG_HVC_DCC, because we haven't
enable uart, so I enable CONFIG_HVC_DCC for verify if we can get DCC
console through "make ARCH=arm64 menuconfig" on linux-next/master, but
this patchset haven't contain this enable CONFIG_HVC_DCC patch, because
upstream seems don't want to enable CONFIG_HVC_DCC defaultly.
Besides these, I all do dtbs_check and checkpatch.pl check, all these
check passed.
CONFIG_HVC_DCC is at upstream. It is not necessary to have UART to test
the kernel w/ the console.
One can boot the kernel w/ JTAG console too. This will allow us to
submit the patches without submitting the UART / clocks dependencies.
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/tty/hvc/hvc_dcc.c
--
---Trilok Soni
