topic title: 4.8 kernels in repos
Posts: 45
hobbyist7890
Joined: 24 Apr 2016
#46
Just edit the antix.list file in / etc / apt / sources.list.d and add the 'dev' after the 'nosystemd' part.
Then 'apt-get update' so the list of packages will find the 'dev' packages. Then 'apt-get dist-upgrade'.
The eudev package in the dev list will automatically get upgraded and replace the old udev causing the problem.
There might be a few other changes. They didn't cause any problems and then everything worked.

The kernel should be unchanged so the Grub should still work. The 'update-grub' command shouldn't hurt anything. The 'grub-install' command should always follow the update-grub when using UEFI so I made my own custom 'update-efi' command with the correct parameters:

Code: Select all

#!/bin/sh
grub-install --target=x86_64-efi --efi-directory=/boot/efi --bootloader-id=antix
That solved my core i3 problems so maybe that will solve your atom problems.
Posts: 48
myway_1
Joined: 26 Oct 2016
#47
hobbyist7890 wrote:Just edit the antix.list file in / etc / apt / sources.list.d and add the 'dev' after the 'nosystemd' part.
Then 'apt-get update' so the list of packages will find the 'dev' packages. Then 'apt-get dist-upgrade'.
The eudev package in the dev list will automatically get upgraded and replace the old udev causing the problem.
There might be a few other changes. They didn't cause any problems and then everything worked.

The kernel should be unchanged so the Grub should still work. The 'update-grub' command shouldn't hurt anything. The 'grub-install' command should always follow the update-grub when using UEFI so I made my own custom 'update-efi' command with the correct parameters:

Code: Select all

#!/bin/sh
grub-install --target=x86_64-efi --efi-directory=/boot/efi --bootloader-id=antix
That solved my core i3 problems so maybe that will solve your atom problems.
Thanks Hobbyist.

I did the steps in the 1st paragraph (not the second paragraph since it appears to be optional)

Still no-go with 4.8.10 gnu-antix. Exact same boot messages as shown in my picture posted previously (except for the UUID number)

I am glad to be using the eudev anyhow. It worked with 4.6 kernel.
Posts: 45
hobbyist7890
Joined: 24 Apr 2016
#48
I am running out of ideas.

Maybe the gnu kernel is missing some driver needed by the atom. I don't know exactly what might have changed between the 4.6 and 4.8 kernels to cause the problem.

I would install the linux-headers-4.8.10-antix and the linux-image-4.8.10-antix along with more firmware packages. Look in 'apt-cache search firmware | more' for example. There might be some needed firmware in the 'firmware-linux-nonfree' package for example.

Maybe somebody else knows what changed with the 4.8 kernels about the atom. I know the audio for the Broadwell Core M processors didn't work until the recent kernels. There is probably an answer on the internet someplace.
Posts: 48
myway_1
Joined: 26 Oct 2016
#49
I got the regular 4.8 antix amd64 kernel. I got Intel microcode.

Still no-go. Same messages.
Posts: 1,445
skidoo
Joined: 09 Feb 2012
#50
.
Seriously, in addition to the 'sticky post' we need a boilerplate reply template.
Help requests where folks don't show their system specs and explain
"Here's what I've searched, here's what I have already tried"
should be ignored or met with a boilerplate reply.

In this case, a 30-second glance at google search results indicates that processor (gleaned from myway's forum sig) requires a specially-patched kernel.


========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"https://www.google.com/#q=linux+gnu+kernel+atom+z3735f+processor"
linktext was:"https://www.google.com/#q=linux+gnu+ker ... +processor"
====================================

^---v

========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://linuxiumcomau.blogspot.com/2016/04/ubuntu-1604-iso-for-intel-atom-compute.html"
linktext was:"http://linuxiumcomau.blogspot.com/2016/ ... mpute.html"
====================================



To be clear:
I'm suggesting that"whoever gets there first" in replying to such"vague" posts would use the same boilerplate reply
Posts: 48
myway_1
Joined: 26 Oct 2016
#51
skidoo wrote:.
Seriously, in addition to the 'sticky post' we need a boilerplate reply template.
Help requests where folks don't show their system specs and explain
"Here's what I've searched, here's what I have already tried"
should be ignored or met with a boilerplate reply.

In this case, a 30-second glance at google search results indicates that processor (gleaned from myway's forum sig) requires a specially-patched kernel.


========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"https://www.google.com/#q=linux+gnu+kernel+atom+z3735f+processor"
linktext was:"https://www.google.com/#q=linux+gnu+ker ... +processor"
====================================

^---v

========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://linuxiumcomau.blogspot.com/2016/04/ubuntu-1604-iso-for-intel-atom-compute.html"
linktext was:"http://linuxiumcomau.blogspot.com/2016/ ... mpute.html"
====================================



To be clear:
I'm suggesting that"whoever gets there first" in replying to such"vague" posts would use the same boilerplate reply
Skidoo, I have added dongles to my PC so it will work with regular kernels. How on earth do you think I was able to post results from terminal commands otherwise? I already knew all about these issues and I implement the cstate fix soon after I install a linux distro. It has never prevented me from booting in any distro or kernel. It just causes my computer to freeze.

I am always reluctant to post in any computer-related forum because there are often mean-spirited people ready to jump on folks who don't do things the way they think they should be done.
Posts: 1,445
skidoo
Joined: 09 Feb 2012
#52
Joe, I think what you're doing (tackling challenge of booting from compute stick) is cool.
The kernel patch seems to be an important consideraton. If your earlier posts mentioned that you're already aware,
then I apologize for missing that and for my grumbling about finding it mentioned in googlesearch page1 search results.
Posts: 45
hobbyist7890
Joined: 24 Apr 2016
#53
My antix16 on an Asus Vivo works fine with 4.8.1 not with 4.8.10.
The only problem is the wireless does not connect to the router. The card is a Realtek rtl8821ae.
I have both kernels installed and the installation is upgraded to the latest 'testing' packages.

Code: Select all

$ cat inxi-4.8.1-3dec2016.txt
System:    Host: vivoax16 Kernel: 4.8.1-antix-sid.1-amd64-smp x86_64 (64 bit) Console: tty 2
           Distro: antiX-16-a1-xorg-sid_x64-core -FÁÍÔÁÑÓÕÁ 19 February 2016
Machine:   Device: desktop Mobo: ASUSTeK model: VC62B UEFI: American Megatrends v: 0405 date: 03/23/2015
CPU:       Dual core Intel Core i3-4030U (-HT-MCP-) cache: 3072 KB 
           clock speeds: max: 1800 MHz 1: 939 MHz 2: 837 MHz 3: 908 MHz 4: 891 MHz
Graphics:  Card: Intel Haswell-ULT Integrated Graphics Controller
           Display Server: X.org 1.18.4 drivers: modesetting (unloaded: fbdev,vesa)
           tty size: 120x37 Advanced Data: N/A out of X
Audio:     Card Intel Haswell-ULT HD Audio Controller driver: snd_hda_intel
           Sound: Advanced Linux Sound Architecture v: k4.8.1-antix-sid.1-amd64-smp
Network:   Card-1: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller driver: r8169
           IF: eth0 state: down mac: 1c:b7:2c:01:ca:84
           Card-2: Realtek RTL8821AE 802.11ac PCIe Wireless Network Adapter driver: rtl8821ae
           IF: wlan0 state: up mac: 40:e2:30:cc:51:c3
Drives:    HDD Total Size: 6001.2GB (41.5% used)
           ID-1: /dev/sda model: WDC_WD60EFRX size: 6001.2GB
Partition: ID-1: / size: 55G used: 3.1G (6%) fs: ext4 dev: /dev/sda6
           ID-2: swap-1 size: 8.00GB used: 0.00GB (0%) fs: swap dev: /dev/sda3
Sensors:   System Temperatures: cpu: 29.8C mobo: 27.8C
           Fan Speeds (in rpm): cpu: 0
Repos:     Active apt sources in file: / etc/apt/sources.list.d/antix.list
           deb http://antix.daveserver.info/testing testing main nosystemd dev
           Active apt sources in file: / etc/apt/sources.list.d/debian.list
           deb http://ftp.us.debian.org/debian/ testing main contrib non-free
Info:      Processes: 121 Uptime: 5 min Memory: 151.9/15959.4MB Init: SysVinit runlevel: 5
           Client: Shell (bash) inxi: 2.3.5 

$ cat inxi-4.8.10-3dec2016.txt
System:    Host: vivoax16 Kernel: 4.8.10-antix-sid.1-amd64-smp x86_64 (64 bit) Console: tty 2
           Distro: antiX-16-a1-xorg-sid_x64-core -FÁÍÔÁÑÓÕÁ 19 February 2016
Machine:   Device: desktop Mobo: ASUSTeK model: VC62B UEFI: American Megatrends v: 0405 date: 03/23/2015
CPU:       Dual core Intel Core i3-4030U (-HT-MCP-) cache: 3072 KB 
           clock speeds: max: 1800 MHz 1: 826 MHz 2: 1209 MHz 3: 800 MHz 4: 822 MHz
Graphics:  Card: Intel Haswell-ULT Integrated Graphics Controller
           Display Server: X.org 1.18.4 drivers: modesetting (unloaded: fbdev,vesa)
           tty size: 120x37 Advanced Data: N/A out of X
Audio:     Card Intel Haswell-ULT HD Audio Controller driver: snd_hda_intel
           Sound: Advanced Linux Sound Architecture v: k4.8.10-antix-sid.1-amd64-smp
Network:   Card-1: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller driver: r8169
           IF: eth0 state: down mac: 1c:b7:2c:01:ca:84
           Card-2: Realtek RTL8821AE 802.11ac PCIe Wireless Network Adapter driver: rtl8821ae
           IF: wlan0 state: down mac: 40:e2:30:cc:51:c3
Drives:    HDD Total Size: 6001.2GB (41.5% used)
           ID-1: /dev/sda model: WDC_WD60EFRX size: 6001.2GB
Partition: ID-1: / size: 55G used: 3.0G (6%) fs: ext4 dev: /dev/sda6
           ID-2: swap-1 size: 8.00GB used: 0.00GB (0%) fs: swap dev: /dev/sda3
Sensors:   System Temperatures: cpu: 29.8C mobo: 27.8C
           Fan Speeds (in rpm): cpu: 0
Repos:     Active apt sources in file: / etc/apt/sources.list.d/antix.list
           deb http://antix.daveserver.info/testing testing main nosystemd dev
           Active apt sources in file: / etc/apt/sources.list.d/debian.list
           deb http://ftp.us.debian.org/debian/ testing main contrib non-free
Info:      Processes: 133 Uptime: 2 min Memory: 120.1/15959.5MB Init: SysVinit runlevel: 5
           Client: Shell (bash) inxi: 2.3.5
I was switching back and forth between the kernels and tried to 'ifup wlan0' several times with the 4.8.10 is the reason my uptime is low just for running inxi. The 4.8.10 never connected to the router when the 4.8.1 kernel connected every time. Sorry about still running the Feb alpha dist except I doubt that is the reason.
Posts: 521
Shay
Joined: 20 Apr 2015
#54
Do both pick up the wireles correctly?
Run

Code: Select all

$ lspci
Find it. Mine reads:
0c:00.0 Network controller: Broadcom Corporation BCM4321
Plug in your information and replace my 0c:00.0

Code: Select all

$ lspci -vnn -s 0c:00.0
Post it from both 4.8.1and 4.8.10.
Hopefully someone can spot the problem.
Posts: 45
hobbyist7890
Joined: 24 Apr 2016
#55
I can try that next chance I get. Looking in the logs shows the only difference is the 4.8.10 saying"no dhcpcd offers received" after trying to connect while the 4.8.1 connects normally. Everything else is the same for both kernels, only the kernels are different.

The time is getting late now. Here is what I found in the logs. There wasn't any difference as shown:

Code: Select all

[    0.000000] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-4.8.1-antix-sid.1-amd64-smp root=UUID=7a2d2169-6cc2-46b4-a8c3-8575baf81e13 ro quiet
[    4.959573] rtl8821ae 0000:04:00.0: enabling device (0000 -> 0003)
[    5.000439] rtl8821ae: Using firmware rtlwifi/rtl8821aefw.bin
[    5.000446] rtl8821ae: Using firmware rtlwifi/rtl8821aefw_wowlan.bin
[    5.041425] ieee80211 phy0: Selected rate control algorithm 'rtl_rc'
[    5.042420] rtlwifi: rtlwifi: wireless switch is on

[    0.000000] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-4.8.10-antix-sid.1-amd64-smp root=UUID=7a2d2169-6cc2-46b4-a8c3-8575baf81e13 ro quiet
[    4.332111] rtl8821ae 0000:04:00.0: enabling device (0000 -> 0003)
[    4.374569] rtl8821ae: Using firmware rtlwifi/rtl8821aefw.bin
[    4.374573] rtl8821ae: Using firmware rtlwifi/rtl8821aefw_wowlan.bin
[    4.468117] ieee80211 phy0: Selected rate control algorithm 'rtl_rc'
[    4.468809] rtlwifi: rtlwifi: wireless switch is on
This isn't too big of a problem yet because I can stay with 4.8.1 when I need the network and see if the 4.9 kernels fix the problem. I have another project starting soon so this can easily wait for me. Maybe I am the only one with this problem.

Edit: I just checked and the kernel config is exactly the same for both kernels and the kernel sizes are the same and the initrd.img files are about the same size for both kernels. I think there must be something changed within the kernels. I might look into the changelogs for the kernels between 4.8.1 and 4.8.10 to see what happened. I expect this is a temporary problem.
Posts: 45
hobbyist7890
Joined: 24 Apr 2016
#56
I just checked with lspci and both kernels produced exactly the same output. There was no diff at all. I plan to let this problem ride out a few more kernel changes before worrying.

Code: Select all

04:00.0 Network controller: Realtek Semiconductor Co., Ltd. RTL8821AE 802.11ac PCIe Wireless Network Adapter
    Subsystem: ASUSTeK Computer Inc. RTL8821AE 802.11ac PCIe Wireless Network Adapter
    Flags: bus master, fast devsel, latency 0, IRQ 46
    I/O ports at d000 [size=256]
    Memory at f7c00000 (64-bit, non-prefetchable) [size=16K]
    Capabilities: [40] Power Management version 3
    Capabilities: [50] MSI: Enable+ Count=1/1 Maskable- 64bit+
    Capabilities: [70] Express Endpoint, MSI 00
    Capabilities: [100] Advanced Error Reporting
    Capabilities: [140] Device Serial Number 00-e0-4c-ff-fe-87-2b-01
    Capabilities: [150] Latency Tolerance Reporting
    Capabilities: [158] L1 PM Substates
    Kernel driver in use: rtl8821ae
    Kernel modules: rtl8821ae
Posts: 45
hobbyist7890
Joined: 24 Apr 2016
#57
The rtl8821ae must be an oddball wifi chip. The 3.16 kernel had some drivers except they didn't work completely. Then the driver quit working until the 4.4 kernel. There is still work now on the driver and firmware so maybe the chip will start working again in the future.
I have several dists installed with several kernels and everything still works if I choose the working kernels so not really a problem now.
Posts: 521
Shay
Joined: 20 Apr 2015
#58
It must be, but I did find this.

========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://unix.stackexchange.com/questions/261609/rtl8821ae-wifi-driver"
linktext was:"http://unix.stackexchange.com/questions ... ifi-driver"
====================================


Maybe of help.