kernel compiling in woof-ce

Message

01micko · #256 Post by **01micko** » Thu 20 Aug 2015, 12:59

Ah.. good on you mate! It's gpl so do as you wish

Mind posting your .config please?

Cheers

stemsee · #257 Post by **stemsee** » Thu 20 Aug 2015, 15:13

Here it is! DOTconfig-4.1.6-'no'pae

peebee · #258 Post by **peebee** » Fri 21 Aug 2015, 20:17

stemsee wrote:Successfully compiled 4.1.6-nopae using 01micko's fdrv kernel-kit. Configured with vfio vfio_pci vga iommu. untested...

https://mega.nz/#!oZ5wEbIR!7yGduSfDNIH6 ... heMHUp3orE

stemsee

Hi stemsee

Tried your new kernel in LxPup-15.06 - seems to work as per usual however I have a couple of observations:

1. the usb keyboard problem during the boot sequence seems to have made a re-appearance in this build

2. your file naming convention produces hidden files when using pcmanfm when the downloaded .tar.bz2 is unpacked (Rox shows the files)

Any chance of making the kernel-sources.sfs available for driver compilation?

Thanks for your builds
Cheers
peebee

stemsee · #259 Post by **stemsee** » Sun 23 Aug 2015, 12:44

Got no wifi at home for the moment!

@peebee

Thanks for feedback.

I was doing a test run with kernel-kit fdrv, and I simply used the config in that reconfiguring the qemu stuff I wanted but I forgot about that keyboard issue!! It will be ok when I have adjusted sukk, in fact I have already re-configured the DOTconfigs. But because of no interent connection, or such a slow one, I haven't been able to compie kernels!

Same for kernel sources, when I get a good connection for a prolonged amount of time I will upload the sources.

stemsee · #260 Post by **stemsee** » Thu 27 Aug 2015, 12:11

kernel-4.1.6-64 sources

https://drive.google.com/file/d/0B4jotf ... sp=sharing

stemsee · #261 Post by **stemsee** » Mon 31 Aug 2015, 08:55

So I seem to have discovered the major part of my aufs problems --- certificates! My system did not accept the git server's certificates!! But why what changed? Is it my system or the certificates? Is there some external interference? How to set up some tests?

stemsee · #262 Post by **stemsee** » Mon 31 Aug 2015, 15:31

kernel-4.1.6-pae

http://tinyurl.com/pe8zok4

sources

http://tinyurl.com/ogft7ly

x64

https://mega.nz/#!1RIgSSzD!RgXoXH7K6_JR ... BOH13LoSbg

sources

https://mega.nz/#!4I4SUC5L!cVM9Hn-Sbsvf ... YwkuTjCQt8

i486
https://mega.nz/#!VIJhCRCI!ybDnwSEgGJ1A ... WtJMpkWuUI

sources
https://mega.nz/#!BAI0WSTD!HKl4DKkZpMM7 ... 8TNGO-jkXw

WillM · #263 Post by **WillM** » Mon 31 Aug 2015, 23:32

Hi stemsee,

I used 4.1.6-64 in Fatdog64-700 and it is working very well. Thanks for uploading.

stemsee · #264 Post by **stemsee** » Tue 01 Sep 2015, 09:05

You're Welcome WillM

peebee · #265 Post by **peebee** » Sun 06 Sep 2015, 10:59

stemsee wrote:kernel-4.1.6-pae

http://tinyurl.com/pe8zok4

sources

http://tinyurl.com/ogft7ly

Thanks Stemsee

Tested with LxPup-15.06.

Confirmed - can now select from list of savefiles/folders on boot with usb keyboard - thanks

Drivers succesfully compiled: Nvidia-304.128, Broadcom wl-223.248, Ralink mt7601u and ndiswrapper-1.59.

Looking good here.

Cheers
peebee

Chili Dog · #266 Post by **Chili Dog** » Sat 12 Sep 2015, 17:03

Hey Stemsee, I know you've compiled working 4-series kernels, but have you managed to build a working 4-series initrd? Because mine keeps trying to find something at boot, giving me "-iname" errors.

Maybe there's some module or folder the init script is looking for that doesn't exist in the 4-series kernel modules?

All I know is - it's really frustrating.

stemsee · #267 Post by **stemsee** » Sun 13 Sep 2015, 11:58

The initrd is independent of the kernel. I use the same initrd with new kernels. Only in fatdog and lighthouse and quirky do you need to edit the initrd to pack in the new modules.

stemsee · #268 Post by **stemsee** » Mon 14 Sep 2015, 10:10

@Chilli Dog

If you have edited your initrd maybe something (init script, DISTRO_SPECS) got corrupted. Did you keep backup? Which pup are you trying to use it with? I can only imagine that you have maybe tried to change the kernel version in the DISTRO_SPECS or your initrd might be looking for the DISTRO_IDSTRING on the initrd or modules sfs or the initrd. I am not sure which files get the idstring appended, but maybe an older init script will need it. I am only guessing.

Chili Dog · #269 Post by **Chili Dog** » Mon 14 Sep 2015, 23:26

Thanks. I got it sorted. It was the naming feature in the kernel-kit, guess I was doing it wrong.

Dry Falls · #270 Post by **Dry Falls** » Sun 20 Sep 2015, 21:57

Stemsee, I've been playing with k4.1.6-EmSee-64 in lxpup64 and JL64 and noticed about 20 extra processes going on at booting X in either. It tapers off in lxpup64 but stays up there in jl64. Then I re-compiled it with the big module package (and your dotconfig64 for k4.1.6) but makes no difference. Still using suuk-universal since SUKK_Universal_GM (and GooMega) wouldn't patch aufs. Any idea where these extra processes are coming from? As near as I can tell, it's more instances of dbus daemon, but why?

Other than that, I like this kernel. It seems (x-ing fingers) to have cured my problem with rox crashing and losing the usb keyboard + monitor.

df

stemsee · #271 Post by **stemsee** » Mon 21 Sep 2015, 17:33

@Dry Falls

Interesting!

I will go through the configs thoroughly when i build 4.2.x, as i do not see the point in working on that kernel anymore. More instances of dbus kernel might derive from the fact that more features for passing hardware to a virtual machine have been enabled. I will research this a little before compiling the next batch.

Iguleder · #272 Post by **Iguleder** » Wed 23 Sep 2015, 12:10

I just updated kernel-kit to add support for deblobbed and longterm kernels.

stemsee - feel free to add your kernel configuration to woof-CE, so others can use it more easily.

stemsee · #273 Post by **stemsee** » Wed 23 Sep 2015, 13:39

Ok I will do. Just looked up deblobbing!

Dry Falls · #274 Post by **Dry Falls** » Sat 26 Sep 2015, 12:44

stemsee wrote:More instances of dbus kernel might derive from the fact that more features for passing hardware to a virtual machine have been enabled. I will research this a little before compiling the next batch.

I did a remaster with this k4.1.6 and for some reason, that settled down the persistent processes. Still more at boot (your explanation makes sense!), but now they taper off.

df

stemsee · #275 Post by **stemsee** » Sun 04 Oct 2015, 19:12

Update to SUKK-Universal with new deblobbing by Iguleder added.

https://mega.nz/#!EVxEDCCL!zjznzKKYgFAM ... N6EubBjDcM

(old)Puppy Linux Discussion Forum

(old)Puppy Linux Discussion Forum

kernel compiling in woof-ce