Somewhat Modern Kernel for the WM8850 (for GPU Acceleration)

[A Linux Enthusiast]

Hello everyone,

I have a CRAIG CLP281x with a WM8850 CPU and as you all know the WM8850 has a Mali 400 GPU.  My netbook is currently running postmarketOS thanks to this GitHub repo and the pmbootstrap tool for my rootfs.  I was hoping to achieve GPU acceleration since it is very well possible considering the existence of the open source Lima GPU driver.  The only problem is that it requires a newer kernel in order for it to work.  According to this website was it first upstreamed in Linux kernel 5.2 and so I do wonder why nobody has brought up the Lima driver in recent times around here.

As far as building ARM kernels go, I'm very familiar with building Linux kernels both for ARM and x86 so I do hope that I won't encounter too many difficulties with trying to build a newer kernel that works on the device.  I would like to ask if there's any kind of patches in particular that I need for for the WM8850?  Has anyone tried to build a newer kernel for the WM8850 and has anyone ever tried the Lima driver for the WM8850?  I hope to be able to build at least a bootable 5.4 Linux kernel soon.

Help would be greatly appreciated.  Have nice day everyone!

[A Linux Enthusiast]

Update:

I decided to use bookconfig by wh0 (with modifications) to build Linux kernel 5.15.64 (even though it's outdated).  I have posted the latest config that I used to build the kernel.  The last kernel source-provided dts file that I built and used was "wm8850-twd-pid7-8198-usa.dts" (so once finished was it "wm8850-twd-pid7-8198-usa.dtb").  The laptop unfortunately would just be at a black screen and stay there when trying to boot the kernel.  This is the exact same issue I was having when I was trying to use the postmarketOS boot setup (with the provided kernel).  I decided to extract the dtb file used in the only working kernel bin file found in this GitHub repo and got several files: 00_kernel, 01_dtbdump_*,00.dtb, and 02_dtbdump_wm,wm8850.dtb


I decided to have the bookconfig script put it in the kernel bin file; this didn't solve my issue.  A bit of time later, I have decided to decompile the extracted dtb files but only one of them actually decompiled successfully: wm8850.dts (I named it this intentionally so I could get the exact name "wm8850.dtb" after recompiling it; I'm about to upload said file as well).  The TechKnow Toolkit that comes with Uberoid gave me some information that made me believe that "wm8850-twd-pid7-8198-usa.dtb" was going to work (see 

wm_analyzer.txt).


Trying to build the newly decompiled dtb file gave these warnings when building it:

arch/arm/boot/dts/wm8850.dts:93.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/plla:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:102.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/pllb:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:111.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/pllc:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:118.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/plld:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:127.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/plle:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:134.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/pllf:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:141.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/pllg:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:148.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/arm:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:155.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/ahb:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:162.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/apb:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:169.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/ddr:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:176.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/uart0:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:186.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/uart1:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:196.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/uart2:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:206.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/uart3:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:216.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/pwm:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:227.6-22: Warning (clocks_property): /soc/pmc@d8130000/clocks/sdhc:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:271.4-20: Warning (clocks_property): /soc/pwm@d8220000:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:304.4-20: Warning (clocks_property): /soc/serial@d8200000:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:312.4-20: Warning (clocks_property): /soc/serial@d82b0000:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:320.4-20: Warning (clocks_property): /soc/serial@d8210000:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:328.4-20: Warning (clocks_property): /soc/serial@d82c0000:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:342.4-20: Warning (clocks_property): /soc/sdhc@d800a000:clocks: cell 0 is not a phandle reference
arch/arm/boot/dts/wm8850.dts:356.3-34: Warning (pwms_property): /backlight:pwms: cell 0 is not a phandle reference

I'll also be attaching the kernel.  Does anyone have any suggestions/recommendations?

[A Linux Enthusiast]

[A Linux Enthusiast]

Here's the kernel: https://drive.google.com/file/d/1ESZ8tgPCkHTx6BcLJ8UGprAtmE8zqVTp/view?usp=sharing

On Friday, June 23, 2023 at 2:45:33 AM UTC-5 A Linux Enthusiast wrote:

[Franco Palombo]

Hey! It's nice to see that this little "family" of laptops is still interesting for some people. I sadly can't help you because I've never did a kernel recompile or anything else like it. The best I could do, is let you know that I have a wm8650, with a hardware serial port available, and link you to some of the recent posts of people that have tried to run stable and latest branches (I think) of the Linux kernel on this little laptop:

prebuilt kernel

so, i got a 6.1 kernel running, now wheres my keyboard?

Getting the latest kernel onto WM8650

In any case, I'm looking forward on what you can do. You seem to have quite some experience, and I would really like to have the default android 2.something out of the nand and replace it by some nice, and preferably recent Linux distro.

[A Linux Enthusiast]

I'm going to rebase the 4.19 patched kernel (that has the Lima driver) with patches from this kernel linux-vtwm and going to do a few other things such as using the loaded modules and logs (from postmarketOS) as guide for setting up my config file (of course along with the appropriate defconfig that will also be adjusted based on on the wiki instructions found in linux-vtwm).  Unfortunately I was unable to get a config file from the working kernel.

I could also try to set up the proprietary Mali drivers using WonderMedia's kernel.  I also found this Linux distrubtion recently? (or if not, I'm revisiting it now) and it appears to have a similar kernel (this one being 3.4.5).  It has a bit of useful information in the "README" and I've attached that file to this post. 

I think I could definitely get the Mali drivers working on here with a bit of time and effort.  I took a look at the old Ubuntu rootfs and it looks like the userspace was already configured for the Mali proprietary drivers.  I'm about to try the distribution and see if it works out well.  I have an old enough toolchain for cross compilation I believe (if anyone reading would like a copy of it, here's a link to download it but if you have an x86_64 Linux system will it need to be able to run 32-bit applications).

Even if things don't go out as planned with getting the Lima driver running, what's helpful about that distribution is that we have working Mali kernel modules and a working userspace rootfs.  At this point if we just wanted GPU acceleration in general (regardless of the Linux setup utilizing the open source Lima drivers or Mali proprietary drivers) I believe all that would need to be done is to use copy the userspace files over and then build a bootable kernel with the kernel space driver.  It might be possible to just copy over the precompiled Mali driver modules over to be honest and I'm about to try this as well with the only kernel that has worked so far that I brought up at the beginning.

[A Linux Enthusiast]

Trying to use the copied-over modules did not work.  I looked up the error I got and it looks like I cannot simply copy over modules and expect them to work most of the time.

[A Linux Enthusiast]

I found kernel version 3.16 rc6, the same version as the working kernel.  The kernel build date is pretty close to the date of the last commit made before the Makefile was bumped to rc7 (at least in terms of the Makefile identifying the kernel as rc6, there is at least one commit that mentions rc7 but that appears to be just a fix for sound related things).  Theoretically should I be able to build this kernel and boot it successfully.  If so, I should be able to use either DX910-SW-99002-r4p0-00rel1 or DX910-SW-99002-r4p1-01rel0 and get myself a working kernel with working Mali kernel drivers; if both are successful do I think I should try the old Ubuntu setup from that previously aforementioned distribution as the userspace stuff is already configured.  Obviously would I prefer to be using a new operating system and kernel, but getting GPU acceleration working with a working kernel (even with older software) would definitely be progress.  I'll be using this for reference when building the kernel modules.

[A Linux Enthusiast]

I might have to try building slightly newer kernels over time until I get to something at to an at least somewhat more recent kernel.

[Alexey Charkov]

Hi there!

Very interesting initiative! Indeed, the Lima driver should technically work with these devices.

Have you tried starting the vanilla upstream kernel on your machine? Basic things should work, and it would be an easier starting place to tinker with Lima.

Last time I checked there was a regression from the time upstream maintainers dropped the empty “chosen” nodes from all .dtsi files, as the “appended dtb” scheme failed to create those on the fly for some reason, resulting in inability to process the kernel boot arguments. So you may want to re-add the “chosen” node as a workaround.

Best regards,

Alexey

--
You received this message because you are subscribed to the Google Groups "VT8500/WM8505 Linux Kernel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vt8500-wm8505-linux...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/vt8500-wm8505-linux-kernel/0ee86b23-955a-4266-8a8f-0706a9fdd253n%40googlegroups.com.

[Alexey Charkov]

Here’s a patch/workaround which Brigham Campbell prepared back then:

https://www.spinics.net/lists/devicetree/msg382725.html

[A Linux Enthusiast]

Hey, Alexey, thank you so much for the information.  I really appreciate the response! 

I accidentally downloaded the wrong Lima kernel that I was intending to rebase so I ended up rebasing their last (and also default branch) non-upstream kernel (which is Linux kernel version 5.0; the dts regression is not present there).  I've attached the config below (which is slightly different than what I used in newly built non-working kernel.  This config file has additional filesystem features but shouldn't be an actual factor between it booting and not booting.  Yes, my root filesystem is ext4 and ext4 has been enabled in the config this entire time). 

This is the GitHub repo to the patched sources (the default branch of linux-vtwm).  If you or anyone wants to try the kernel bin file, the modules are integrated into the kernel and there shouldn't be need for an initrd/initramfs of any kind, and dtb file is "wm8850-twd-pid7-8198-usa.dtb" so from the looks of it, you only need the kernel bin file for it to boot.  I'm unfortunately unable to get it booting on my laptop though.  I just recreated the kernel bin file (though without those filesystem additions).  My toolcain is "gcc-arm-none-eabi" running on Debian 12.

My WM8850 device has refused to actually boot anything that isn't that 3.16 rc6 kernel so far.

I'll give an attempt at building vanilla Linux kernel version 6.3.9 and see if it boots up.

Kind regards,
Nathan

[A Linux Enthusiast]

Unfortunately, the laptop doesn't like the new kernel either and I get the same exact issue as before.  I have provided both the config file and the kernel bin file.

[A Linux Enthusiast]

Its using this dtb file: "wm8850-twd-pid7-8198-usa.dtb"

[A Linux Enthusiast]

Here is also the patch file.

[A Linux Enthusiast]

This is an exported dmesg log from that working 3.16 rc6 kernel.  Should I try to get a dmesg log from Android?

[A Linux Enthusiast]

I have determined that the working kernel was a vendor kernel for a WM8850 device.  I'm not sure what else to do now unfortunately.  My device is just really odd when it comes to running non-vendor kernels.

[Alexey Charkov]

Have you tried increasing your bootdelay or replacing it with rootwait
in your kernel command line arguments?

Also, I trust you've appended the correct dtb file to your zImage
before packing it with mkimage? Are you replacing the working
uzImage.bin file in-place with your rebuilt one? Did you by any chance
mount this FAT32 filesystem on any recent Windows machine (or use
Windows to create the filesystem in the first place)?

3.16-rc6 doesn't really sound like a vendor kernel, so there must be
something fishy with the build process if none of the kernels you
build start successfully but that one does.

Best,
Alexey

On Sun, Jun 25, 2023 at 11:45 PM A Linux Enthusiast

> To view this discussion on the web visit https://groups.google.com/d/msgid/vt8500-wm8505-linux-kernel/84304b75-ddb1-4e20-891d-7fe0b59e54dan%40googlegroups.com.

[A Linux Enthusiast]

I have only been using Linux systems throughout the duration of this project.  The reason why I found that other kernel to be a vendor kernel is because of the device tree source file that I was able to extract while it was running that kernel.  It looks like it is utilizing a dtb file that is targeting the WM8850 netbook "x6-10v" and I have yet to find any other dts file target that laptop in particular; I also found from WonderMedia files as well from that GitHub repo.

I have attached a u-boot file that wh0's bookconfig tool made.  I renamed "scriptcmd" to "autorun.1.wmt" and the laptop still remains at a blackscreen. 

Here's the kernel command line arguments for that one: "setenv bootargs root=/dev/mmcblk0p2 rw rootwait"

I've attached my boot script file below.  Maybe it's because the first line has "imagedlcdinit" and the lack of "display init" found in the other boot script file?  I'm going to make a new boot script file that will very similar to the one from that GitHub repo.  I'll also add that what usually happens is that this laptop refuses to display text when it is set in the boot script file.  I don't really know why but this is why the Uberoid installer behaved oddly on that laptop.  I have only ever had something display whenever the laptop actually successfully initializes the kernel.

[A Linux Enthusiast]

The stock Android installation that the laptop came with did not have this issue though; for some reason did the original Android installation was able to output something to the display prior to loading the kernel.  The installer files for the original vendor-provided Android build are unavailable and nobody ever archived them unfortunately.

Also, that autorun.1.wmt that has "roowait" does work with the original kernel.  It's my not boot script file that's causing the problem unfortunately.

[A Linux Enthusiast]

I could also guess that perhaps the person to built that 3.16 rc6 kernel applied patches from the WM8850 kernel.  I could guess that this is how it was done:

- Git clone the 3.0.8 regular Linux kernel

- Git clone the 3.0.8 WM8850 vendor kernel
- Use Git to the differences between the both

- Apply those patches using Git to the 3.16 rc6 kernel

The stock Android installation that the laptop came with did not have the issue with it not being able to display anything until the kernel has loaded; for some reason did the original Android installation was able to output something to the display prior to loading the kernel.  The installer files for the original vendor-provided Android build are unavailable and nobody ever archived them unfortunately.

Also, that autorun.1.wmt that has "roowait" does work with the original kernel.  It's my not boot script file that's causing the problem unfortunately.

[Alexey Charkov]

Sounds like something off with your build process then.

 - Does any other non-WonderMedia prebuilt kernel work with your current setup?

 - Do you put your newly built uzImage.bin in the same location as the one that worked, if any?

 - Which cross-compiler do you use to build the kernel?

 - What are the steps you took to generate your uzImage.bin?

The vendor-provided Android kernel is of limited use here unfortunately, because it doesn't even use a device tree (so the early boot process is different). The "lcdinit" part should be irrelevant, because it only affects the bootloader and not the kernel (which should really be only a fraction of a second in the boot process). But you can use it to verify that your boot script gets executed at all by including some message to be printed to the display.

It might be easier to figure out what's wrong if you get access to the serial console.

If you only get a blank screen this most likely means that the kernel doesn't load at all, which might be due to the boot script (you can rule this out if the same boot script works with a different kernel), or due to the kernel image format (like using vmlinux instead of zImage or similar), or due to the dtb not loading correctly for the 'appended dtb' boot (like if you appended the dtb to the uImage after packing it, instead of to the zImage before packing it), or due to dtb parsing failures (like that 'chosen' node regression I mentioned before). The next likely explanation is that the kernel panics before the display driver gets time to initialize, such as if you don't have support for your root device or root filesystem compiled in.

Best,

Alexey

[A Linux Enthusiast]

- No, only that 3.16 rc6 kernel boots at all; I installed postmarketOS on it.  It was made "targeting" (not necessarily that device in particular) the Tokio Techbook, an a WM8850 device.  The postmarketOS build that the pmbootstrap tool gives you is a version of their 4.3.0 Linux kernel.  Unfortunately, that kernel did not work on my laptop either.  The 3.16 rc6 kernel is the only kernel that has ever worked from my testing.  If I had to make a guess however, maybe this person applied WonderMedia patches to the 3.16 rc6 kernel either from your GitHub repo or the regular Linux kernel because there is a dts file that I was able to extract successfully.  That kernel does come with some WonderMedia components though but I did also notice a lack of Mali proprietary driver  Oddly enough does it have its Linux headers but no config file nor the module needed to be able to create one with the options used to build said kernel.  I created a GitHub issue on that person's repo to see if I could get that kernel's config as currently I have no way of obtaining it.

- I put the newly build uzImage.bin file in the same place as the old one and ensured that the old one was not going to be creating any conflicts with the new image.

- The cross-compiler that I have been using is "gcc-arm-none-eabi" (on the stable release of Debian 12).  This is a compiler that should work as well but I could always be wrong; wh0's bookconfig (a WM8505 Linux image builder) makes use of that same cross-compiler and the on the repo, wh0 claims to have been using Debian unstable (which was the only version of Debian 12 at that time).

- I followed the guide from your wiki that's apart of your GitHub repo but firstly used  vt8500_v6_v7_defconfig as my base config (which turns out to have most of everything that I needed enabled).  The config that I've attached to this post is a slightly more updated config file that adds like one or two additional filesystems but otherwise, nothing more than that.  EXT4 is what my root filesystem is and it has been enabled in my kernels this entire time.

Unfortunately for some reason the u-boot implementation on this device is rather odd; the original Android ROM somehow was able to give me anything on the display prior to loading the kernel.  If I overwhelm the Uberoid installation and make it crash, it does show me an Android Jellybean animation; an animation that is supposed to be play prior to loading the kernel.  I'm able to rule out that it is my script as I've been using it this entire time and it boots that vendor(?) kernel.  I've been setting up my kernel images just as the wiki guide has told me to do; also, there's an error with the dtb portion of the guides' there's a "dts" folder with compiled dtbs; maybe older Linux kernels didn't have a "dts" folder.

How do I exactly access the serial connector of this laptop?  Would I need to solder anything?  How many pins is the serial connector?

Kind regards,
Nathan

[Alexey Charkov]

On Tue, Jun 27, 2023 at 5:11 PM A Linux Enthusiast
<randomte...@gmail.com> wrote:
>
> - No, only that 3.16 rc6 kernel boots at all; I installed postmarketOS on it. It was made "targeting" (not necessarily that device in particular) the Tokio Techbook, an a WM8850 device. The postmarketOS build that the pmbootstrap tool gives you is a version of their 4.3.0 Linux kernel. Unfortunately, that kernel did not work on my laptop either. The 3.16 rc6 kernel is the only kernel that has ever worked from my testing. If I had to make a guess however, maybe this person applied WonderMedia patches to the 3.16 rc6 kernel either from your GitHub repo or the regular Linux kernel because there is a dts file that I was able to extract successfully. That kernel does come with some WonderMedia components though but I did also notice a lack of Mali proprietary driver Oddly enough does it have its Linux headers but no config file nor the module needed to be able to create one with the options used to build said kernel. I created a GitHub issue on that person's repo to see if I could get that kernel's config as currently I have no way of obtaining it.
>
> - I put the newly build uzImage.bin file in the same place as the old one and ensured that the old one was not going to be creating any conflicts with the new image.
>
> - The cross-compiler that I have been using is "gcc-arm-none-eabi" (on the stable release of Debian 12). This is a compiler that should work as well but I could always be wrong; wh0's bookconfig (a WM8505 Linux image builder) makes use of that same cross-compiler and the on the repo, wh0 claims to have been using Debian unstable (which was the only version of Debian 12 at that time).
>
> - I followed the guide from your wiki that's apart of your GitHub repo but firstly used vt8500_v6_v7_defconfig as my base config (which turns out to have most of everything that I needed enabled). The config that I've attached to this post is a slightly more updated config file that adds like one or two additional filesystems but otherwise, nothing more than that. EXT4 is what my root filesystem is and it has been enabled in my kernels this entire time.

Okay, that's a tricky one then. Definitely need more info to figure
out what's wrong - serial console it is.

> Unfortunately for some reason the u-boot implementation on this device is rather odd; the original Android ROM somehow was able to give me anything on the display prior to loading the kernel. If I overwhelm the Uberoid installation and make it crash, it does show me an Android Jellybean animation; an animation that is supposed to be play prior to loading the kernel. I'm able to rule out that it is my script as I've been using it this entire time and it boots that vendor(?) kernel. I've been setting up my kernel images just as the wiki guide has told me to do; also, there's an error with the dtb portion of the guides' there's a "dts" folder with compiled dtbs; maybe older Linux kernels didn't have a "dts" folder.

Yep, the dtbs got moved in more recent kernel versions. As for the
u-boot implementation, they are all rather peculiar (based on ancient
v1 code, prior to device trees, prior to the driver model, with vendor
modifications scattered unpredictably around the source tree)

> How do I exactly access the serial connector of this laptop? Would I need to solder anything? How many pins is the serial connector?

It's usually a 3- or 4-pin unpopulated header on the motherboard, with
GND, Rx, Tx and +3.3V/+5V in one order or another. The Rx and Tx pads
would have thin traces on the PCB going in parallel straight to the
SOC. If it's not labeled on the silkscreen, one way to find the
required pads is just to take your USB-serial adapter, connect its
ground wire to the shielding of one of the USB ports, start the
terminal application on the PC with 115200n8 serial port settings,
then take the Rx wire and randomly poke it into "suspicious" looking
pins while rebooting the poor little WM8850. At some point text will
start pouring into your terminal, which would mean you found the Tx
pad :-) Rx is most likely right next to it.

No need to solder anything if you have a steady hand and/or Dupont
clips or something similar. Soldering a connector might be more
convenient though in the longer term.

Best,
Alexey

[A Linux Enthusiast]

I don't actually currently own a USB-serial adapter.  Would this adapter work?

Here's the dts file that I was able to get from the 3.16 rc6 kernel when it was booted; it was before I tried the "rootwait" kernel argument.  I did get a lot of warnings when trying to build that dtb in particular and even after compiling it and adding it to the zImage did it still not work.

[Alexey Charkov]

On Tue, 27 Jun 2023 at 19:13, A Linux Enthusiast <randomte...@gmail.com> wrote:

I don't actually currently own a USB-serial adapter.  Would this adapter work?

Yes, that should work. You may also find spring loaded probe pins helpful, something like this: 

https://a.co/d/gDUDYx2

Here's the dts file that I was able to get from the 3.16 rc6 kernel when it was booted; it was before I tried the "rootwait" kernel argument.  I did get a lot of warnings when trying to build that dtb in particular and even after compiling it and adding it to the zImage did it still not work.

Looks legit, and it’s definitely not a WonderMedia kernel :)

Best,

Alexey

[A Linux Enthusiast]

I guess that person decided to use WonderMedia patches in the kernel or something?  Not sure!  Do you think that I may have setup my config file incorrectly?  I didn't really make too many changes to the kernel besides removing the need for an initrd/initramfs and kernel modules by having them integrated into the kernel (by adjusting the config file appropriately).

[Matt Mason]

Hi all, apologies that I have nothing constructive to add to the discussion here at the moment. 

Only wanted to register my interest as I have recently gotten my WM8850 out of the drawer again. Keenly following any outcomes from the work here. Thanks in advance

[kez...@gmail.com]

Hi Matt. Do you have anything left on your hdd for this machine?

[Matthew Mason]

It does seem I kept some files, are you able to be more specific about what you're looking for?

--
You received this message because you are subscribed to a topic in the Google Groups "VT8500/WM8505 Linux Kernel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/vt8500-wm8505-linux-kernel/_vA1RYVsZdM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to vt8500-wm8505-linux...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/vt8500-wm8505-linux-kernel/59688b6d-74b3-47ec-8f50-3ce31203d942n%40googlegroups.com.