Jump to content

going

Members
  • Posts

    769
  • Joined

  • Last visited

Everything posted by going

  1. The patch is accepted in v6.12 drivers/of/property.c Many thanks from the author of megi
  2. This feature is necessary for developers. I mean those people who really understand the kernel code and will be able to draw conclusions and steps when they see the trace message. These people will be able to add any functionality to the kernel and debug it on their own. P.S. Thank you for not asking why the kernel is not being built with debugging symbols.
  3. PR: https://github.com/armbian/build/pull/7442
  4. Subject: Fix broken allwinner,sram dependency on h616, h618 Or, at your discretion, indicate in another way that this is relevant for these processors
  5. I will send this fix to the author of the original patch. We will wait for his opinion. P.S. @Sesse I remembered that this is relevant for h616 and h618 processors. Other users have informed me about the lack of a temperature interface on boards with these processors. With your permission, I will edit the message in the patch a little.
  6. Did I understand correctly? Did you mean this patches.megous/of-property-fw_devlink-Support-allwinner-sram-links.patch? You can disable any patch by simply putting a minus sign at the beginning of a line in the series file: diff --git a/patch/kernel/archive/sunxi-6.11/series.conf b/patch/kernel/archive/sunxi-6.11/series.conf index 3eaf6617..3a2067e0 100644 --- a/patch/kernel/archive/sunxi-6.11/series.conf +++ b/patch/kernel/archive/sunxi-6.11/series.conf @@ -171,7 +171,7 @@ patches.megous/usb-gadget-Fix-dangling-pointer-in-netdev-private-data.patch patches.megous/mmc-dw-mmc-rockchip-fix-sdmmc-after-soft-reboot.patch patches.megous/Revert-drm-sun4i-lvds-Invert-the-LVDS-polarity.patch - patches.megous/of-property-fw_devlink-Support-allwinner-sram-links.patch +- patches.megous/of-property-fw_devlink-Support-allwinner-sram-links.patch patches.megous/arm64-dts-rockchip-rk356x-Fix-PCIe-register-map-and-ranges.patch patches.megous/Fix-intptr_t-typedef.patch patches.megous/mmc-sunxi-mmc-Remove-runtime-PM.patch without making any other changes to the build system. The best solution is to make changes on top of all applied patches. Make a git commit and extract it as a patch to the root directory with patches. This can be done directly in the build system. sesion (1) cat <<- EOF > "userpatches/config-test.conf" display_alert "Settings from" "config-test.conf file" "info" BOARD=bigtreetech-cb1 BRANCH=edge RELEASE=bookworm KERNEL_GIT=shallow # Enter your github credentials here so that we know the author and how to contact you. declare -g MAINTAINERMAIL="48602507+The-going@users.noreply.github.com" declare -g MAINTAINER="The-going" declare -g SYNC_CLOCK="no" EOF ./compile.sh test kernel-patch # The build system will apply all patches, commit and stop. # Don't do anything in this terminal session (1). # The script will print the path to the directory where you want to make changes. # You will need another session (2) in the terminal to access and act on the kernel repository. session (2): cd /path/to/kernel-worktree sudo su # Make changes to the file in any convenient editor. For example: nano README # write and exit the editor session (1) Press <ENTER> And you'll see something like this: ───────┬─────────────────────────────────────────────────────────────────────────────────────────────────────────────── │ File: /home/leo/armbian/output/patch/kernel-rockchip-rk3588-edge.patch ───────┼─────────────────────────────────────────────────────────────────────────────────────────────────────────────── 1 │ diff --git a/README b/README 2 │ index fd903645e..fa0d6844d 100644 3 │ --- a/README 4 │ +++ b/README 5 │ @@ -1,5 +1,5 @@ 6 │ -Linux kernel 7 │ -============ 8 │ +Linux kernel Armbian framework patching 9 │ +======================================= 10 │ 11 │ There are several guides for kernel developers and users. These guides can 12 │ be rendered in a number of formats, like HTML and PDF. Please read ───────┴─────────────────────────────────────────────────────────────────────────────────────────────────────────────── [🚸] Are you happy with this patch? [ Type 'yes' to accept, 'stop' to stop patching, or anything else to keep patching ] Enter yes The script will make the correct patch assign it a strange name that does not match the "Subject:" field and put it in the "output/patch/" folder. leo@noble-vm:~/armbian$ batcat /home/leo/armbian/output/patch/kernel-rockchip-rk3588-edge.patch ───────┬───────────────────────────────────────────────────────────────────────── │ File: /home/leo/armbian/output/patch/kernel-rockchip-rk3588-edge.patch ───────┼───────────────────────────────────────────────────────────────────────── 1 │ From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001 2 │ From: The-going <48602507+The-going@users.noreply.github.com> 3 │ Date: Mon, 4 Nov 2024 12:21:22 +0000 4 │ Subject: Patching kernel rockchip-rk3588 files README 5 │ 6 │ Signed-off-by: The-going <48602507+The-going@users.noreply.github.com> 7 │ --- 8 │ README | 4 ++-- 9 │ 1 file changed, 2 insertions(+), 2 deletions(-) 10 │ 11 │ diff --git a/README b/README 12 │ index fd903645e..fa0d6844d 100644 13 │ --- a/README 14 │ +++ b/README 15 │ @@ -1,7 +1,7 @@ 16 │ -Linux kernel 17 │ -============ 18 │ +Linux kernel Armbian framework patching 19 │ +======================================= 20 │ 21 │ There are several guides for kernel developers and users. These guides can 22 │ be rendered in a number of formats, like HTML and PDF. Please read 23 │ Documentation/admin-guide/README.rst first. 24 │ 25 │ -- 26 │ Created with Armbian build tools https://github.com/armbian/build 27 │ ───────┴────────────────────────────────────────────────────────────────────────────── This strange behavior can be circumvented. session (2) cd /path/to/kernel-worktree sudo su # Tell who you are directly to git. git config --global user.email "48602507+The-going@users.noreply.github.com" git config --global user.name "The-going" # Check the git settings. cat ~/.gitconfig # Make the necessary changes to the file and commit. # The main thing is to write a good explanation in the message to the commit about the changes made. # End the superuser session exit # If it's a single commit then just send the patch to the destination (regular user session): git format-patch --keep-subject --filename-max-length=75 --abbrev=12 -1 -o /home/user/path/to/armbian-build/patch/kernel/archive/sunxi-6.11/ # If the changes are significant and you need several commits, # the best solution is to make a separate series. # Create an empty patches.addons folder in the target directory TDIR="/home/user/path/to/armbian-build/patch/kernel/archive/sunxi-6.11/patches.addons" mkdir $TDIR # If 5 commits revision_runge="HEAD~5..HEAD" # run script mk_format_patch.sh /home/user/path/to/armbian-build/tools/mk_format_patch . $revision_runge $TDIR # Add a list of new patches to the series.conf file cat ${TDIR}/../series.addons >>${TDIR}/../series.conf I wrote it in such detail because other users will read it. @Sesse You have very successfully found this rotten apple. I express my deep gratitude for the work done. Please post a properly designed patch here and I will send it to the megous upstream project so that it can be reviewed and the names included in v6.12. With respect
  7. Please understand me correctly. I only have an A64 processor. And I can only check the real work on it. For the rest, only compilation. When moving patches to a new version, something always breaks and the work of fixing it falls on someone else who has boards with one processor or another. I can only give you hints.
  8. In order to make it easier to view and analyze the source code, I support the kernel with patches already applied. Try to compare it with others sun50i-h616 trips It definitely works. sun50i-a64 trips
  9. Try rolling back the build system to this commit: 90637986eef70e5478105b64039a8ba1bfd85062 All subsequent commits are irrelevant to sunxi. Check the correctness of the temperature triplet in the DTS, fix it and the sun8i_thermal temperature driver will work correctly.
  10. Perhaps the problem is not in the configuration file. Provide more information about the OS such as: uname -r awk '{gsub(/ /, "\n\t", $0); print "\n\t" $0 "\n"}' /proc/cmdline dmesg | egrep -i 'drm|hdmi|gpu|model' And anything else that seems suspicious to you.
  11. Thank you very much! Now we can assume that I will disappear from the horizon for several months of study.🤗
  12. Thank you both so much for these kind answers. I've been wanting to learn how to use NPU for a long time. But unfortunately I haven't figured out where I can start yet. As far as I know from the documentation on the rk3588 SOC, there is a controller in it that is designed for I/O with NPU. I still don't understand how I can program it and how to interact with the Linux kernel. Can you tell me the direction to study?
  13. Good. I think I got it. You have already taken a ready-made rt core. Using the headers to this kernel, you compiled several drivers and got a kernel panic when running the test. Did I understand correctly? If this is the case, then it will be quite difficult for me to give any advice. I usually use full control. Build the kernel yourself. Preliminary application of patches. If there are problems with the kernel, then add a trace function. With respect.
  14. If you installed the kernel from the repository as apt install, then simply copy the required DTB file from the folder "/usr/lib/kernel-version" to the folder "/boot/dtb-kernel-version/rockchip". For some reason, the linux-dtb package for this kernel does not contain DTB files.
  15. Will you be able to post here, under spoiler, a list of files that are being modified by your RT PREEMPT patch? I mean the output of the "lsdiff" command.
  16. command help: mount --help full help page: man mount /dev/mmcblk0p2 29G 28G 0 100% / du --help create a file dirsize.sh: #!/bin/bash for d in /* do case $d in /dev|/proc|/sys) echo "Skipped $d ; continue ;; esac if [ -d $d ];then echo "$(du $d)";fi done chmod +x dirsize.sh - make it executable sudo ./dirsize.sh - Run with superuser rights
  17. First, you need to connect the pins correctly. They will be different for your board: Hardware Connection 232 PIN Raspberry Pi (BCM) Description VCC 5V 3.3/5V Power Input GND GND Ground SCLK P21 (SPI1 SCLK) SPI Clock Signal Input MOSI P20 (SPI1 MOSI) SPI Data Input MISO P19 (SPI1 MISO) SPI Data Output CS P18 (SPI1 CS) SPI Chip Select IRQ P24 Interrupt Output . You need to write the correct overlay file and compile it. Then this instruction.
  18. Question. Can you publish any technical documentation for this device?
  19. Can you explain what these numbers mean? And what alternative can they be compared with? It seems to me that this is probably a wrong translation.
  20. In fact, I found a lot of problems, these two were not on my list. The build system does some steps more correctly, but this utility lags behind. I think we need to teach her how to split empty space into sections and mount it the way the user wanted. The work is slow, but it is progressing.
  21. Thank you very much. I collect in a notebook all possible bugs when using this utility. I want to redo it.
  22. Can you describe the selection scenario (order) in the "armbian-install" interface?
  23. I wonder if the source codes for this loader exist?
  24. There are several ways to force a WIFI-USB device to be detected in the OS. You can write and add a node to the DTS. That is, to use an overlay. You can write a rule and add it to the initialization scripts. Or a combination of these methods. Please reach out to people who have already succeeded in this. There are publications on this topic here on the forum.
×
×
  • Create New...

Important Information

Terms of Use - Privacy Policy - Guidelines