Hi Yixun Lan
On Wed, 2024-12-04 at 08:58 +0800, Yixun Lan wrote:
Hi Marcel, Huan
On 18:52 Tue 03 Dec , Marcel Ziswiler wrote:
Hi Huan Zhou
On Fri, 2024-11-29 at 13:37 +0800, Huan Zhou wrote:
From: Kongyang Liu seashell11234455@gmail.com
Add document for Banana Pi F3 board which based on SpacemiT's K1 SoC.
Signed-off-by: Kongyang Liu seashell11234455@gmail.com Signed-off-by: Huan Zhou pericycle.cc@gmail.com
doc/board/index.rst | 1 + doc/board/spacemit/bananapi-f3.rst | 86 ++++++++++++++++++++++++++++++++++++++ doc/board/spacemit/index.rst | 8 ++++ 3 files changed, 95 insertions(+)
diff --git a/doc/board/index.rst b/doc/board/index.rst index 417c128c7af6ad2267ef1bc743c7f10ae70b6de7..367da2d62316d4cab25ecc53f852ef742eba13dd 100644 --- a/doc/board/index.rst +++ b/doc/board/index.rst @@ -51,6 +51,7 @@ Board-specific doc sipeed/index socionext/index sophgo/index + spacemit/index st/index starfive/index ste/index diff --git a/doc/board/spacemit/bananapi-f3.rst b/doc/board/spacemit/bananapi-f3.rst new file mode 100644 index 0000000000000000000000000000000000000000..1e6d6ef19be7e25684966bd4a727bd21b147330a --- /dev/null +++ b/doc/board/spacemit/bananapi-f3.rst @@ -0,0 +1,86 @@ +.. SPDX-License-Identifier: GPL-2.0-or-later
+Banana Pi F3
Officially they call it Banana Pi BPI-F3.
+============
+Building +~~~~~~~~ +1. Install the spacemit riscv cross compile toolchain_, or skip it if riscv toolchain is installed.
I would not use a specific such toolchain but rather a regular RISC-V one e.g.
I think any riscv toolchain should work fine here, as I'm using Gentoo specific riscv toolchain.. so consider it's merely an example provided to demonstrate whole process working
Okay.
+.. _toolchain: https://archive.spacemit.com/toolchain/
+2. Setup cross compilation environment variable:
+.. code-block:: console
+ export CROSS_COMPILE=<riscv64 toolchain prefix, e.g /opt/spacemit/bin/riscv64-unknown-linux-gnu->
+3. Before building U-Boot, OpenSBI should be built first. OpenSBI can be +built for SpacemiT K1 SoC as below:
+.. code-block:: console
+ git clone https://github.com/cyyself/opensbi%C2%A0-b k1-opensbi + cd opensbi + make PLATFORM=generic
What speaks against using regular master opensbi?
no idea if master of opensbi would work here, we can certainly try..
I tried and it seems to work fine, resp. the only difference I could spot is that instead of 'crc32+ OK' it prints some random ASCII characters in the last line before any U-Boot output:
[ 0.490] ## Checking hash(es) for Image opensbi ... crc32+ OK
but I think it's another story, as we are focusing on mainlining uboot
Yes, exactly. I am just curious if it needs combining with some obscure vendor stuff.
FWIK, this branch is generally same version as vendor opensbi, but slightly rebase to more recent opensbi version
+4. Then build U-Boot as following:
+.. code-block:: console
+ cd <U-Boot-dir> + make bananapi-f3_defconfig + make OPENSBI=<OpenSBI-dir>/build/platform/generic/firmware/fw_dynamic.bin
+This will generate u-boot.itb
+Booting +~~~~~~~ +Actually, we can replace the uboot part from bianbu linux which is the bsp_ to validate this patch, +use `balena etcher` to burn the bianbu-minimal.img to the sd card,
trailing whitespace
+and replace the /dev/sdx4 where places the uboot_ with the `u-boot.itb` generated from this patch.
+.. _bsp: https://archive.spacemit.com/image/k1/version/bianbu/v2.0/ +.. _uboot: https://bianbu-linux.spacemit.com/en/device/boot#21-firmware-layout
+Sample boot log from Banana Pi F3 board
BPI-F3
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +.. code-block:: none
+ try sd... + bm:3 + j...
+ U-Boot SPL 2022.10spacemit-dirty (Oct 21 2024 - 09:01:13 +0000) + [ 0.279] DDR type LPDDR4X + [ 0.292] lpddr4_silicon_init consume 13ms + [ 0.293] Change DDR data rate to 2400MT/s + [ 0.430] ## Checking hash(es) for config conf-1 ... OK + [ 0.432] ## Checking hash(es) for Image opensbi ... OK + [ 0.437] ## Checking hash(es) for Image uboot ... OK + [ 0.443] ## Checking hash(es) for Image fdt-1 ... OK + [ 0.488] ## Checking hash(es) for config config_1 ... OK + [ 0.490] ## Checking hash(es) for Image opensbi ... crc32+ OK
+ U-Boot 2024.10-rc4-00462-g5b138cfcc587-dirty (Nov 28 2024 - 14:56:49 +0800)
+ DRAM: 4 GiB
I guess it does not yet detect the amount of RAM available.
I would be fine with this, as currently the memory size is hardcoded, we should work on follow-up patch to auto probe DDR size, and corret it
Yes, I agree.
+ Core: 19 devices, 8 uclasses, devicetree: separate + Loading Environment from nowhere... OK + In: serial@d4017000 + Out: serial@d4017000 + Err: serial@d4017000 + Net: No ethernet found. + => cpu list + 0: cpu@0 spacemit,x60 + 1: cpu@1 spacemit,x60 + 2: cpu@2 spacemit,x60 + 3: cpu@3 spacemit,x60 + 4: cpu@4 spacemit,x60 + 5: cpu@5 spacemit,x60 + 6: cpu@6 spacemit,x60 + 7: cpu@7 spacemit,x60 + => test + =>
trailing whitespace
diff --git a/doc/board/spacemit/index.rst b/doc/board/spacemit/index.rst new file mode 100644 index 0000000000000000000000000000000000000000..3fb7d804ac8fc8dd4c7ee67ffc877f9ad323162d --- /dev/null +++ b/doc/board/spacemit/index.rst @@ -0,0 +1,8 @@ +.. SPDX-License-Identifier: GPL-2.0-or-later
+SpacemiT +======== +.. toctree:: + :maxdepth: 1
+ bananapi-f3
new blank line at EOF
Thanks!
Cheers
Marcel