Don't forget to assign the virtualbox 5.0.16 update to QA also, as that'll need to go out with this kernel update.

Tested both on one of our new Dell workstations that we just got, Mageia 5 x86_64.
Dell OptiPlex 9020
Xeon E3-1200 v3/4th Gen Core Processor PCI Express x16 Controller
Intel Corporation 8 Series/C220 Series Chipset Family
Radeon HD 8570 / R7 240/340 OEM
Broadcom Corporation NetXtreme BCM5722 Gigabit Ethernet PCI Express
Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz (4 cores)

On the host, I get this error printed at the top of the screen early in the boot:
radeon 0000:01:00.0: VCE init error (-110).

The display seems to work fine though, both in Plymouth and X.

The journal log also shows this critical error:
xt_addrtype: ipv6 does not support BROADCAST matching

but that one was there on the old kernels too.

I tested several VMs, which all worked fine:
Windows XP, 7
Solaris 10, 11
Fedora 23
Ubuntu 14.04 LTS
OpenSuSE 13.2
Debian 7
PC-BSD 10.2

On our older Dell workstations we've been using the past 4+ years, Mageia 5 i586, this update works fine too.

Also on our two big Dell servers, Mageia 5 i586, this works fine.  No issues there either.  One of them even has an XFS filesystem in /var.

Mageia 5 i586 VirtualBox guest works fine with the updated packages too.

Running the 4.4.6 desktop kernel on x86_64 UEFI machine with nvidia graphics.
Everything looks OK so far.

All the recent kernels have been tested on this machine and somewhere recently two of the nvidia libraries got lost.  This prevents nvidia-settings from running.  nvidia-settings is in the /usr/share tree but there is no link from /usr/bin.  Have no idea if this could be related to the nvidia340-kernel packages.

CC: (none) => tarazed25

Sad to say, on my 32-bit Dell Dimension, this breaks my wifi. Symptoms are exactly like Bug #17990 for M6. The entire wlan0 option just disappeared. It was working before the upgrade. In fact, I downloaded the upgrade using the wifi.

Other than that, looks fine if I use a wired connection.

CC: (none) => andrewsfarm

Mageia 5 i586, Dell Inspiron 600m works fine.  Haven't tried the wifi (will try to test that this weekend), but the wlp2s3 interface does show up and draknetcenter does show the available SSIDs just fine.

Some more testing references:
- Works good on a x86_64 Asus Zenbook Skylake
- Works good on a x86_64 Haswell workstation
- Works good on a couple of HP Proliant Servers

x86_64 server kernels gets heavy testing as whole Mageia buildsystem is now running 4.4.6-1 kernel

Testing on mga5-64 along with nvidia-current from bug#17604

Packages installed cleanly. System re-booted normally to the KDE desktop. No regressions noted.

I don't have wifi and so cannot test that.

OK for mga5-64 on this system

Machine:   Mobo: ECS model: GeForce7050M-M v: 1.0
CPU:       Quad core AMD Phenom 9500 (-MCP-)
Graphics:  Card: NVIDIA GF108 [GeForce GT 630]

Booting back into the current 4.1.x kernel restores wifi. It looks like the problem is due to changes in the requirements of b43-openfwwf. No idea if other firmware files would be affected, but I think it bears looking into.

My guess would be that if I went oldstyle and used b43-fwcutter on the appropriate file downloaded from somewhere else, as I used to have to do, then it would work again. But I'm sure we'd rather get the open firmware working if we can, even though this is an old card and probably not seen much any more.

(In reply to Thomas Andrews from comment #5)
> Sad to say, on my 32-bit Dell Dimension, this breaks my wifi. Symptoms are
> exactly like Bug #17990 for M6. The entire wlan0 option just disappeared. It
> was working before the upgrade. In fact, I downloaded the upgrade using the
> wifi.


The fact that you see the same issue in mga6-dev1 and this update is that they currently have same kernel & firmware :)

So we'll push the same fix as in #17990 to mga5 too as part of this update

I was able to get my wifi working again by downloading additional firmware from the b43-openfwwf site, and installing it manually. Explained more fully in Bug #17990, with the tar of the additional firmware attached there.

I should have noted in comment#8 that I used kernel-desktop.

Testing on another mga5-64 system, also with kernel-desktop

Packages installed cleanly. System re-booted normally to the KDE desktop. 
No regressions noted

OK for mga5-64 on this system

System: Hewlett-Packard product: CQ2925EA v: 1.00
Mobo: PEGATRON model: 2AE2 v: 1.02 Bios: AMI v: 8.08 date: 08/31/2012
Dual core Intel Pentium G645T (-MCP-) cache: 3072 KB           
Card:Intel 810 and later:Intel 2nd Generation Core Processor

Created attachment 7585 [details]
afternoon journal entries

CC: (none) => brtians1

Video Card: RS780L [Radeon 3000]

Installed:

kernel-desktop-4.4.6-1.mga5-1-1.mga5.x86_64.rpm


Rebooted - went to blank screen.  (See Journal)

Let me know what you need to help debug this.

Video Card:  C51 [GeForce 6150 LE]  --- AMD X2-3800

brian@localhost ~]$ uname -a
Linux localhost 4.4.6-desktop586-1.mga5 #1 SMP Wed Mar 16 20:11:36 UTC 2016 i686 i686 i686 GNU/Linux


Installed 

nvidia304-kernel-desktop-latest-304.131-3.mga5.nonfree.i586.rpm and it brought in the other pieces.

Working as designed.

Installed 32-bit server kernel with nvidia340 driver, in conjunction with the virtualbox upgrades. In addition, installed desktop kernel in a Mageia 5 32-bit guest.

No problems noted.

I will test on the same hardware in a 64-bit install, but because that is my main production install I prefer to wait until some of the other problems have been resolved, unless there is a compelling need.

Wireless testing i586
Vendor: âRalink corp.  Description: âRT2760 Wireless 802.11n 1T/2R


Working as designed.

On my newer PogoLinux machine at home, working fine for the most part, but I do have one significant (to me) issue.

I have an ATI TV Wonder VE card, and for audio, there's a 1/8" audio cable going out from the TV card and into the sound card, and then the "Line" input in the mixer (aumix -l at the command line) is where the sound coming in is.  With the updated kernel, I no longer get sound if Line is turned on.

hard boot or soft boot David? It may need a hard boot to replace card firmware.

(In reply to Brian Rockwell from comment #14)
> Video Card: RS780L [Radeon 3000]
> 
> Installed:
> 
> kernel-desktop-4.4.6-1.mga5-1-1.mga5.x86_64.rpm
> 
> 
> Rebooted - went to blank screen.  (See Journal)
> 
> Let me know what you need to help debug this.


/var/log/Xorg.0.log if the same boot, Xorg.0.log.old if not, cp it to your home directory and chown to your user.

You could add dmesg and perhaps journalctl -b > journal.txt too

I've tried with both a warm and cold boot multiple times.

Unfortunately snd_hda_intel is broken in 4.4.6 and not able to capture audio from Line In.  I don't think bt878 (TV card driver) is an issue, as the Line Out from that card is a hardware feature.

snd_hda_intel   : Intel Corporation|8 Series/C220 Series Chipset High Definition Audio Controller [MULTIMEDIA_AUDIO_DEV] (rev: 05)

Testing on mga5-32, with kernel-server and nouveau video driver

System re-booted normally. No regressions.

OK for mga5-32 on this system

System:    Host: mga-5-32 Kernel: 4.4.6-server-1.mga5 i686 (32 bit) 
Machine:   Mobo: ECS model: GeForce7050M-M v: 1.0 Bios: American Megatrends 
CPU:       Quad core AMD Phenom 9500 (-MCP-)
Graphics:  Card: NVIDIA GF108 [GeForce GT 630]
           Display Server: X.Org 1.16.4 drivers: v4l,nouveau

Created attachment 7587 [details]
Errors when trying to run 4.4.6 - xorg

Tried:

The following 2 packages are going to be installed:

- kernel-desktop-devel-4.4.6-1.mga5-1-1.mga5.x86_64
- kernel-desktop-devel-latest-4.4.6-1.mga5.x86_64

System hangs with no screen.  Cannot F2 to terminal.

In VirtualBox, M5, KDE, 32-bit

Package(s) under test:
kernel-desktop586-latest vboxadditions-kernel-desktop586-latest

default install of kernel-desktop586-latest

[root@localhost wilcal]# uname -a
Linux localhost 4.1.15-desktop586-2.mga5 #1 SMP Wed Jan 20 17:06:34 UTC 2016 i686 i686 i686 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop586-latest
Package kernel-desktop586-latest-4.1.15-2.mga5.i586 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop586-latest
Package vboxadditions-kernel-desktop586-latest-5.0.14-1.mga5.i586 is already installed

System boots to a working desktop. Common apps work. Screen dimensions are correct.

install kernel-desktop-latest from updates_testing

[root@localhost wilcal]# uname -a
Linux localhost 4.4.6-desktop586-1.mga5 #1 SMP Wed Mar 16 20:11:36 UTC 2016 i686 i686 i686 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop586-latest
Package kernel-desktop586-latest-4.4.6-1.mga5.i586 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop586-latest
Package vboxadditions-kernel-desktop586-latest-5.0.16-2.mga5.i586 is already installed

System boots to a working desktop. Common apps work. Screen dimensions are correct.

CC: (none) => wilcal.int

In VirtualBox, M5, KDE, 64-bit

Package(s) under test:
kernel-desktop-latest vboxadditions-kernel-desktop-latest

default install of kernel-desktop586-latest

[root@localhost wilcal]# uname -a
Linux localhost 4.1.15-desktop-2.mga5 #1 SMP Wed Jan 20 17:05:51 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.14-1.mga5.x86_64 is already installed

System boots to a working desktop. Common apps work. Screen dimensions are correct.

install kernel-desktop-latest & vboxadditions-kernel-desktop-latest from updates_testing

[root@localhost wilcal]# uname -a
Linux localhost 4.4.6-desktop-1.mga5 #1 SMP Wed Mar 16 20:11:06 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.4.6-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.16-2.mga5.x86_64 is already installed

System boots to a working desktop. Common apps work. Screen dimensions are correct.

Further to my comment#22, something very strange happened 

uname -r reported that I was running server-4.4.6, but I discovered that the symlinks for vmlinuz and initrd were pointing to non-existent files and rpm said that the package was not installed.

I keep a copy of the list of files that rpmdrake says are to be installed and kernel-server-4.4.6-1.mga5-1-1.mga5.i586 is on that list. No error was reported.

I installed the packages again and the 4.4.6 initrd and vmlinuz files were correctly created this time.

The system rebooted normally to the new kernel.

No regressions noticed.

Tested with KDE4/Plasma desktop, nvidia-settings, steam games, smplayer using vpdau, audio play, audio record (front mic, rear mic), etc.

$ uname -a
Linux marte 4.4.6-desktop-1.mga5 #1 SMP Wed Mar 16 20:11:06 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
$ cat /proc/driver/nvidia/version 
NVRM version: NVIDIA UNIX x86_64 Kernel Module  340.96  Sun Nov  8 22:33:28 PST 2015
GCC version:  gcc version 4.9.2 (GCC) 
$ lspcidrake | grep AUDIO
snd_hda_intel   : NVIDIA Corporation|High Definition Audio Controller [MULTIMEDIA_AUDIO_DEV] (rev: a1)
snd_hda_intel   : Intel Corporation|82801JI (ICH10 Family) HD Audio Controller [MULTIMEDIA_AUDIO_DEV]
$ lspcidrake | grep VGA
Card:NVIDIA GeForce 8100 to GeForce 415: NVIDIA Corporation|GT218 [GeForce 210] [DISPLAY_VGA] (rev: a2)
$ lspci | grep VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GT218 [GeForce 210] (rev a2)

CC: (none) => mageia

No regressions noticed.

Tested on Xfce, extreme tux racer, firefox+youtube+fc-solve.shlomifish.org, VLC, samba.

My laptop is:

<<<
 I also have an Acer Aspire 5738DZG laptop with the following specs:

    Intel Pentium(R) Dual-Core CPU T4300 @ 2.10GHz. (x86-64).
    ATI Mobility Radeon⢠HD 4570 (r700)
    15.6â³ 3D HD LCD Screen.
    3 GB Memory
    320 GB Hard Disk Drive.
    âDVD Super Multi DL driveâ
    Acer Nplify⢠802.11b/g/n.

>>>

(In reply to Shlomi Fish from comment #28)
> No regressions noticed.
> 

OK, I noticed one regression with my setup: I cannot disable the laptop's built-in trackpad (that mouse-board thingy). Previously, I was able to disable it once and no longer restore it (which was a buggy behaviour but tolerable), and now I cannot disable it at all. I may have to try if it happens with a vanilla kernel, an older kernel, on Windows , etc.

-- Shlomi

> Tested on Xfce, extreme tux racer, firefox+youtube+fc-solve.shlomifish.org,
> VLC, samba.
> 
> My laptop is:
> Ok
> <<<
>  I also have an Acer Aspire with the following specs:
> 
>     Intel Pentium(R) Dual-Core CPU T4300 @ 2.10GHz. (x86-64).
>     ATI Mobility Radeon⢠HD 4570 (r700)
>     15.6â³ 3D HD LCD Screen.
>     3 GB Memory
>     320 GB Hard Disk Drive.
>     âDVD Super Multi DL driveâ
>     Acer Nplify⢠802.11b/g/n.
> 
> >>>

(In reply to Shlomi Fish from comment #29)
> (In reply to Shlomi Fish from comment #28)
> > No regressions noticed.
> > 
> 
> OK, I noticed one regression with my setup: I cannot disable the laptop's
> built-in trackpad (that mouse-board thingy). Previously, I was able to
> disable it once and no longer restore it (which was a buggy behaviour but
> tolerable), and now I cannot disable it at all. I may have to try if it
> happens with a vanilla kernel, an older kernel, on Windows , etc.
> 

Well, it seems that this now happens in the laptop globally even on Windows 7 and while using the older kernel, so it is likely not the newer kernel's fault. Sorry for the noise.

Regards,

-- Shlomi  

> -- Shlomi

Ouch, serious regression just noticed on my Dell Inspiron 600m laptop.  The screen doesn't blank when closing the lid.  I use IceWM and this is controlled globally through ACPI via the following:

[david@marin ~]$ cat /etc/acpi/actions/lidbtn.sh 
#!/bin/sh
grep "open" /proc/acpi/button/lid/LID/state > /dev/null
if [ $? -eq 0 ]; then
    vbetool dpms on
else
    vbetool dpms off
fi
[david@marin ~]$ cat /etc/acpi/events/lidbtn 
event=button[ /]lid
action=/etc/acpi/actions/lidbtn.sh

So I'll have to check whether it's ACPI not triggering the event or vbetool not working that's the problem.

No regression on my 5 years old Dell Vostro 3500

Adding feedback marker for now due to regressions

Whiteboard: (none) => feedback

On real hardware, M5, KDE, 64-bit

Package(s) under test:
kernel-desktop-latest
virtualbox vboxadditions-kernel-desktop-latest dkms-virtualbox
virtualbox-guest-additions virtualbox-kernel-desktop-latest x11-driver-video-vboxvideo
nvidia-current-kernel-desktop-latest

default install of:
kernel-desktop-latest
virtualbox vboxadditions-kernel-desktop-latest dkms-virtualbox
virtualbox-guest-additions virtualbox-kernel-desktop-latest x11-driver-video-vboxvideo
nvidia-current-kernel-desktop-latest

[root@localhost wilcal]# uname -a
Linux localhost 4.1.15-desktop-2.mga5 #1 SMP Wed Jan 20 17:05:51 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox
Package virtualbox-5.0.14-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.14-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi dkms-virtualbox
Package dkms-virtualbox-5.0.14-1.mga5.noarch is already installed
[root@localhost wilcal]# urpmi virtualbox-guest-additions
Package virtualbox-guest-additions-5.0.14-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox-kernel-desktop-latest
Package virtualbox-kernel-desktop-latest-5.0.14-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi x11-driver-video-vboxvideo
Package x11-driver-video-vboxvideo-5.0.14-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi nvidia-current-kernel-desktop-latest
Package nvidia-current-kernel-desktop-latest-346.96-5.mga5.nonfree.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nvidia
        Kernel modules: nvidiafb, nouveau, nvidia_current

Created clients:
M4.1 i586 KDE Live-CD runs as a Vbox client. Screen sizes are correct. Sound ok
M5 x86_64 KDE CI installed and runs as a Vbox client. Screen sizes are correct. Sound ok

install from updates_testing:
kernel-desktop-latest
virtualbox vboxadditions-kernel-desktop-latest dkms-virtualbox
virtualbox-guest-additions virtualbox-kernel-desktop-latest x11-driver-video-vboxvideo
nvidia-current-kernel-desktop-latest
from updates_testing

Packages installed:

- btrfs-progs-4.4.1-1.mga5.x86_64
- dkms-nvidia-current-352.79-1.mga5.nonfree.x86_64
- dkms-virtualbox-5.0.16-1.mga5.noarch
- dracut-038-21.mga5.x86_64
- kernel-desktop-4.4.6-1.mga5-1-1.mga5.x86_64
- kernel-desktop-latest-4.4.6-1.mga5.x86_64
- kernel-firmware-20160219-1.mga5.noarch
- kernel-firmware-nonfree-20160311-1.mga5.nonfree.noarch
- lib64btrfs0-4.4.1-1.mga5.x86_64
- microcode-0.20160303-1.mga5.nonfree.noarch
- nvidia-current-kernel-4.4.6-desktop-1.mga5-352.79-3.mga5.nonfree.x86_64
- nvidia-current-kernel-desktop-latest-352.79-3.mga5.nonfree.x86_64
- radeon-firmware-20160219-1.mga5.nonfree.noarch
- vboxadditions-kernel-4.4.6-desktop-1.mga5-5.0.16-2.mga5.x86_64
- vboxadditions-kernel-desktop-latest-5.0.16-2.mga5.x86_64
- virtualbox-5.0.16-1.mga5.x86_64
- virtualbox-guest-additions-5.0.16-1.mga5.x86_64
- virtualbox-kernel-4.4.6-desktop-1.mga5-5.0.16-2.mga5.x86_64
- virtualbox-kernel-desktop-latest-5.0.16-2.mga5.x86_64
- x11-driver-video-vboxvideo-5.0.16-1.mga5.x86_64

Reboot results in an unrecoverable system. Even attempting to go back
to 4.1.15 it is still unrecoverable. Last boot screen is attached.

Test platform:
Intel Core i7-2600K Sandy Bridge 3.4GHz
GIGABYTE GA-Z68X-UD3-B3 LGA 1155 MoBo
GIGABYTE GV-N440D3-1GI Nvidia GeForce GT 440 (Fermi) 1GB
RTL8111/8168B PCI Express 1Gbit Ethernet
DRAM 16GB (4 x 4GB)

Created attachment 7591 [details]
Unrecoverable boot screen

tried with a skylake laptop with optimus.  Everything seems working, but I had to change /etc/bumblebee/xorg.conf.nvidia adding
Section "Screen"
    Identifier "Default Screen"
    Device "DiscreteNvidia"
EndSection

Otherwise the nvidia card would not start, while working previously

CC: (none) => birrachiara

I now tried upgrading the kernel and booting in an x86-64 VBox VM. Everything seems to be in order: konsole, perl, firefox, youtube, HTML 5 demo, etc.

Now I've also tried in a Mageia 5 i586 (32-bit) VBox VM. Everything seems to be fine: konsole, perl, Firefox, YouTube, VLC, Firefox playing .webm, IceWM, KDE4.

(In reply to David Walser from comment #31)
> Ouch, serious regression just noticed on my Dell Inspiron 600m laptop.  The
> screen doesn't blank when closing the lid.  So I'll have to check whether it's ACPI not triggering the event or vbetool not working that's the problem.

The verdict is that VBE is broken, as is likely anything using libx86.

vbetool gives this output:
Error: something went wrong performing real mode interrupt

and strace shows:
vm86old(0xb77404cc)                     = -1 ENOSYS (Function not implemented)

monitor-get-edid-using-vbe gives no output, has an exit status of 1, and strace:
vm86old(0xb77974cc)                     = -1 ENOSYS (Function not implemented)

So it looks like a system call is missing in the 4.4.6 kernel build.

A longstanding issue with the i586 server kernel is that hard disk I/O is significantly slower than with the desktop kernel in some cases.  4.4.6 has exacerbated this issue quite a bit, and now just trying to rsync files from one disk to another on our big server or trying to extract a large tarball on my workstation are so slow as to be impractical.

Created attachment 7611 [details]
systemd journal showing kernel stack traces

Attaching my Hitachi Touro Mobile Pro to one of our new Optiplex 9020's with this kernel (kernel-desktop in this case) on an x86_64 install is causing the kernel to crash.  With i586 and kernel-server on the older Optiplex 990 it didn't crash.

(In reply to David Walser from comment #41)
> Created attachment 7611 [details]
> systemd journal showing kernel stack traces
> 
> Attaching my Hitachi Touro Mobile Pro to one of our new Optiplex 9020's with
> this kernel (kernel-desktop in this case) on an x86_64 install is causing
> the kernel to crash.  With i586 and kernel-server on the older Optiplex 990
> it didn't crash.

Thomas asked me to try installing the microcode package and recreating the initrd, which didn't help.

Then he helped me realize that the USB ports that say "SS" are USB3, and that's the kind of port I was using that was causing this issue.  Using USB2 works fine.

You can see if port is usb 3.0 other way too.
http://www.howtogeek.com/wp-content/uploads/2014/01/650x434xusb-2.0-vs-usb-3.0-blue-port.jpg.pagespeed.gp+jp+jw+pj+js+rj+rp+rw+ri+cp+md.ic.Hw2qDXl-UA.jpg

I get no crash with asus laptop usb 3.0 port with usb 3.0 stick,even usb 2.0 hub works with usb 3.0 in force legacy mode with usb 3.0 kingston stick plugged.

CC: (none) => otto.leipala

(In reply to Otto Leipälä from comment #43)
> You can see if port is usb 3.0 other way too.
> http://www.howtogeek.com/wp-content/uploads/2014/01/650x434xusb-2.0-vs-usb-3.
> 0-blue-port.jpg.pagespeed.gp+jp+jw+pj+js+rj+rp+rw+ri+cp+md.ic.Hw2qDXl-UA.jpg
> 
> I get no crash with asus laptop usb 3.0 port with usb 3.0 stick,even usb 2.0
> hub works with usb 3.0 in force legacy mode with usb 3.0 kingston stick
> plugged.

My PogoLinux computer at home has blue USB 3.0 ports, but on the Dells at work they are black and look exactly the same as the USB 2.0 ports, except for the little label next to it.

It also depends on what you plug into it.  I can plug my phone into it to charge and nothing bad happens.

Not all computer manufactors are following that color system in usb 3.0.
Your phone is not mounted to filesystem,it's only take power from usb and works even no os is booted.

True, I haven't tried using MTP through the USB3 port to see if that kills the kernel.  It actually won't charge when no OS is booted though.

My samsung galaxy trend plus charging it when i stop to boot order screen. If it's not getting power it will also affects to usb drive installed linux,they get power from usb port and it will make it fail to boot/work.

On my Intel (Skylake) i5 6500 machine, the 4.4 kernel fixes graphical and network issues I had with the default 3.19 kernel available from the Mageia 5 DVD. But the GUI still displays a lot of black lines, especially when reducing/expanding windows. I only have the default Intel HD Graphics 530 card on my machine.

Does someone know if installing https://01.org/linuxgraphics/downloads/intel-graphics-installer-linux-1.4.0 would help? Or is it mostly the same as x11-driver-video-intel? I also thought about installing vaapi-driver-intel-1.4.1-1.mga5, but this version doesn't support Skylake; version 1.6 or newer is required.

CC: (none) => LpSolit

(In reply to Frédéric Buclin from comment #48)
> On my Intel (Skylake) i5 6500 machine, the 4.4 kernel fixes graphical and
> network issues I had with the default 3.19 kernel available from the Mageia
> 5 DVD. But the GUI still displays a lot of black lines, especially when
> reducing/expanding windows. I only have the default Intel HD Graphics 530
> card on my machine.
> 

You may be seeing Bug 16986. I saw something similar to what you describe with my Dell Dimension E310, which has on-board Intel graphics. The fix for that seems to be to add a line to /etc/X11/xorg.conf as described in the report of that bug.

(In reply to Thomas Andrews from comment #49)
> You may be seeing Bug 16986. I saw something similar to what you describe
> with my Dell Dimension E310, which has on-board Intel graphics. The fix for
> that seems to be to add a line to /etc/X11/xorg.conf as described in the
> report of that bug.

Unfortunately, this didn't help, even after a system reboot. :(

(In reply to Frédéric Buclin from comment #50)
> (In reply to Thomas Andrews from comment #49)
> > You may be seeing Bug 16986. I saw something similar to what you describe
> > with my Dell Dimension E310, which has on-board Intel graphics. The fix for
> > that seems to be to add a line to /etc/X11/xorg.conf as described in the
> > report of that bug.
> 
> Unfortunately, this didn't help, even after a system reboot. :(

Yeah, uxa is not really useful for skylake class hw as it's only tested in sna mode upstream... I will look into rolling a new x11-driver-video-intel for mga5 too

(In reply to Frédéric Buclin from comment #50)
> (In reply to Thomas Andrews from comment #49)
> > You may be seeing Bug 16986. I saw something similar to what you describe
> > with my Dell Dimension E310, which has on-board Intel graphics. The fix for
> > that seems to be to add a line to /etc/X11/xorg.conf as described in the
> > report of that bug.
> 
> Unfortunately, this didn't help, even after a system reboot. :(

Sorry about that. Seemed like it was worth a try, though.

Fresh rpms to test:

SRPMS:
aufs-tools-4.4-0.git20160301.1.mga5.src.rpm
b43-openfwwf-5.2-6.2.mga5.src.rpm
btrfs-progs-4.4.1-1.mga5.src.rpm
dracut-038-21.mga5.src.rpm
iproute2-4.4.0-1.mga5.src.rpm
kernel-4.4.9-1.mga5.src.rpm
kernel-firmware-20160409-1.mga5.src.rpm
kernel-userspace-headers-4.4.9-1.mga5.src.rpm
kmod-vboxadditions-5.0.20-2.mga5.src.rpm
kmod-virtualbox-5.0.20-2.mga5.src.rpm
kmod-xtables-addons-2.10-4.mga5.src.rpm
xfsdump-3.1.6-1.mga5.src.rpm
xfsprogs-4.3.0-1.mga5.src.rpm
xstream-1.4.9-1.mga5.src.rpm
xtables-addons-2.10-1.mga5.src.rpm

broadcom-wl-6.30.223.271-4.mga5.nonfree.src.rpm
fglrx-15.302-4.mga5.nonfree.src.rpm
kernel-firmware-nonfree-20160503-1.mga5.nonfree.src.rpm
kmod-broadcom-wl-6.30.223.271-9.mga5.nonfree.src.rpm
kmod-fglrx-15.302-7.mga5.nonfree.src.rpm
kmod-nvidia-current-352.79-8.mga5.nonfree.src.rpm
kmod-nvidia304-304.131-5.mga5.nonfree.src.rpm
kmod-nvidia340-340.96-5.mga5.nonfree.src.rpm
radeon-firmware-20160429-1.mga5.nonfree.src.rpm



i586:
kernel-desktop-4.4.9-1.mga5-1-1.mga5.i586.rpm
kernel-desktop586-4.4.9-1.mga5-1-1.mga5.i586.rpm
kernel-desktop586-devel-4.4.9-1.mga5-1-1.mga5.i586.rpm
kernel-desktop586-devel-latest-4.4.9-1.mga5.i586.rpm
kernel-desktop586-latest-4.4.9-1.mga5.i586.rpm
kernel-desktop-devel-4.4.9-1.mga5-1-1.mga5.i586.rpm
kernel-desktop-devel-latest-4.4.9-1.mga5.i586.rpm
kernel-desktop-latest-4.4.9-1.mga5.i586.rpm
kernel-doc-4.4.9-1.mga5.noarch.rpm
kernel-server-4.4.9-1.mga5-1-1.mga5.i586.rpm
kernel-server-devel-4.4.9-1.mga5-1-1.mga5.i586.rpm
kernel-server-devel-latest-4.4.9-1.mga5.i586.rpm
kernel-server-latest-4.4.9-1.mga5.i586.rpm
kernel-source-4.4.9-1.mga5-1-1.mga5.noarch.rpm
kernel-source-latest-4.4.9-1.mga5.noarch.rpm
perf-4.4.9-1.mga5.i586.rpm

kernel-userspace-headers-4.4.9-1.mga5.i586.rpm

kernel-firmware-20160409-1.mga5.noarch.rpm
b43-openfwwf-5.2-6.2.mga5.noarch.rpm

aufs-tools-4.4-0.git20160301.1.mga5.i586.rpm

btrfs-progs-4.4.1-1.mga5.i586.rpm
libbtrfs0-4.4.1-1.mga5.i586.rpm
libbtrfs-devel-4.4.1-1.mga5.i586.rpm

dracut-038-21.mga5.i586.rpm

iproute2-4.4.0-1.mga5.i586.rpm
iproute2-doc-4.4.0-1.mga5.i586.rpm
libiproute2-static-devel-4.4.0-1.mga5.i586.rpm

xfsdump-3.1.6-1.mga5.i586.rpm
xfsprogs-4.3.0-1.mga5.i586.rpm
libxfs1-4.3.0-1.mga5.i586.rpm
libxfs-devel-4.3.0-1.mga5.i586.rpm
libxfs-static-devel-4.3.0-1.mga5.i586.rpm

vboxadditions-kernel-4.4.9-desktop-1.mga5-5.0.20-2.mga5.i586.rpm
vboxadditions-kernel-4.4.9-desktop586-1.mga5-5.0.20-2.mga5.i586.rpm
vboxadditions-kernel-4.4.9-server-1.mga5-5.0.20-2.mga5.i586.rpm
vboxadditions-kernel-desktop586-latest-5.0.20-2.mga5.i586.rpm
vboxadditions-kernel-desktop-latest-5.0.20-2.mga5.i586.rpm
vboxadditions-kernel-server-latest-5.0.20-2.mga5.i586.rpm

virtualbox-kernel-4.4.9-desktop-1.mga5-5.0.20-2.mga5.i586.rpm
virtualbox-kernel-4.4.9-desktop586-1.mga5-5.0.20-2.mga5.i586.rpm
virtualbox-kernel-4.4.9-server-1.mga5-5.0.20-2.mga5.i586.rpm
virtualbox-kernel-desktop586-latest-5.0.20-2.mga5.i586.rpm
virtualbox-kernel-desktop-latest-5.0.20-2.mga5.i586.rpm
virtualbox-kernel-server-latest-5.0.20-2.mga5.i586.rpm

dkms-xtables-addons-2.10-1.mga5.i586.rpm
iptaccount-2.10-1.mga5.i586.rpm
libaccount0-2.10-1.mga5.i586.rpm
libaccount-devel-2.10-1.mga5.i586.rpm
xtables-addons-2.10-1.mga5.i586.rpm
xtables-addons-kernel-4.4.9-desktop-1.mga5-2.10-4.mga5.i586.rpm
xtables-addons-kernel-4.4.9-desktop586-1.mga5-2.10-4.mga5.i586.rpm
xtables-addons-kernel-4.4.9-server-1.mga5-2.10-4.mga5.i586.rpm
xtables-addons-kernel-desktop586-latest-2.10-4.mga5.i586.rpm
xtables-addons-kernel-desktop-latest-2.10-4.mga5.i586.rpm
xtables-addons-kernel-server-latest-2.10-4.mga5.i586.rpm
xtables-geoip-2.10-1.mga5.noarch.rpm

iwlwifi-agn-ucode-20160503-1.mga5.nonfree.noarch.rpm
kernel-firmware-nonfree-20160503-1.mga5.nonfree.noarch.rpm
radeon-firmware-20160429-1.mga5.nonfree.noarch.rpm
ralink-firmware-20160503-1.mga5.nonfree.noarch.rpm
rtlwifi-firmware-20160503-1.mga5.nonfree.noarch.rpm

broadcom-wl-common-6.30.223.271-4.mga5.nonfree.i586.rpm
broadcom-wl-kernel-4.4.9-desktop-1.mga5-6.30.223.271-9.mga5.nonfree.i586.rpm
broadcom-wl-kernel-4.4.9-desktop586-1.mga5-6.30.223.271-9.mga5.nonfree.i586.rpm
broadcom-wl-kernel-4.4.9-server-1.mga5-6.30.223.271-9.mga5.nonfree.i586.rpm
broadcom-wl-kernel-desktop586-latest-6.30.223.271-9.mga5.nonfree.i586.rpm
broadcom-wl-kernel-desktop-latest-6.30.223.271-9.mga5.nonfree.i586.rpm
broadcom-wl-kernel-server-latest-6.30.223.271-9.mga5.nonfree.i586.rpm
dkms-broadcom-wl-6.30.223.271-4.mga5.nonfree.i586.rpm

dkms-fglrx-15.302-4.mga5.nonfree.i586.rpm
fglrx-control-center-15.302-4.mga5.nonfree.i586.rpm
fglrx-devel-15.302-4.mga5.nonfree.i586.rpm
fglrx-kernel-4.4.9-desktop-1.mga5-15.302-7.mga5.nonfree.i586.rpm
fglrx-kernel-4.4.9-desktop586-1.mga5-15.302-7.mga5.nonfree.i586.rpm
fglrx-kernel-4.4.9-server-1.mga5-15.302-7.mga5.nonfree.i586.rpm
fglrx-kernel-desktop586-latest-15.302-7.mga5.nonfree.i586.rpm
fglrx-kernel-desktop-latest-15.302-7.mga5.nonfree.i586.rpm
fglrx-kernel-server-latest-15.302-7.mga5.nonfree.i586.rpm
fglrx-opencl-15.302-4.mga5.nonfree.i586.rpm
x11-driver-video-fglrx-15.302-4.mga5.nonfree.i586.rpm

nvidia304-kernel-4.4.9-desktop-1.mga5-304.131-5.mga5.nonfree.i586.rpm
nvidia304-kernel-4.4.9-desktop586-1.mga5-304.131-5.mga5.nonfree.i586.rpm
nvidia304-kernel-4.4.9-server-1.mga5-304.131-5.mga5.nonfree.i586.rpm
nvidia304-kernel-desktop586-latest-304.131-5.mga5.nonfree.i586.rpm
nvidia304-kernel-desktop-latest-304.131-5.mga5.nonfree.i586.rpm
nvidia304-kernel-server-latest-304.131-5.mga5.nonfree.i586.rpm

nvidia340-kernel-4.4.9-desktop-1.mga5-340.96-5.mga5.nonfree.i586.rpm
nvidia340-kernel-4.4.9-desktop586-1.mga5-340.96-5.mga5.nonfree.i586.rpm
nvidia340-kernel-4.4.9-server-1.mga5-340.96-5.mga5.nonfree.i586.rpm
nvidia340-kernel-desktop586-latest-340.96-5.mga5.nonfree.i586.rpm
nvidia340-kernel-desktop-latest-340.96-5.mga5.nonfree.i586.rpm
nvidia340-kernel-server-latest-340.96-5.mga5.nonfree.i586.rpm

nvidia-current-kernel-4.4.9-desktop-1.mga5-352.79-8.mga5.nonfree.i586.rpm
nvidia-current-kernel-4.4.9-desktop586-1.mga5-352.79-8.mga5.nonfree.i586.rpm
nvidia-current-kernel-4.4.9-server-1.mga5-352.79-8.mga5.nonfree.i586.rpm
nvidia-current-kernel-desktop586-latest-352.79-8.mga5.nonfree.i586.rpm
nvidia-current-kernel-desktop-latest-352.79-8.mga5.nonfree.i586.rpm
nvidia-current-kernel-server-latest-352.79-8.mga5.nonfree.i586.rpm



x86_64:
cpupower-4.4.9-1.mga5.x86_64.rpm
cpupower-devel-4.4.9-1.mga5.x86_64.rpm
kernel-desktop-4.4.9-1.mga5-1-1.mga5.x86_64.rpm
kernel-desktop-devel-4.4.9-1.mga5-1-1.mga5.x86_64.rpm
kernel-desktop-devel-latest-4.4.9-1.mga5.x86_64.rpm
kernel-desktop-latest-4.4.9-1.mga5.x86_64.rpm
kernel-doc-4.4.9-1.mga5.noarch.rpm
kernel-server-4.4.9-1.mga5-1-1.mga5.x86_64.rpm
kernel-server-devel-4.4.9-1.mga5-1-1.mga5.x86_64.rpm
kernel-server-devel-latest-4.4.9-1.mga5.x86_64.rpm
kernel-server-latest-4.4.9-1.mga5.x86_64.rpm
kernel-source-4.4.9-1.mga5-1-1.mga5.noarch.rpm
kernel-source-latest-4.4.9-1.mga5.noarch.rpm
perf-4.4.9-1.mga5.x86_64.rpm

kernel-userspace-headers-4.4.9-1.mga5.x86_64.rpm

kernel-firmware-20160409-1.mga5.noarch.rpm
b43-openfwwf-5.2-6.2.mga5.noarch.rpm

aufs-tools-4.4-0.git20160301.1.mga5.x86_64.rpm

btrfs-progs-4.4.1-1.mga5.x86_64.rpm
lib64btrfs0-4.4.1-1.mga5.x86_64.rpm
lib64btrfs-devel-4.4.1-1.mga5.x86_64.rpm

dracut-038-21.mga5.x86_64.rpm

iproute2-4.4.0-1.mga5.x86_64.rpm
iproute2-doc-4.4.0-1.mga5.x86_64.rpm
iptaccount-2.10-1.mga5.x86_64.rpm
lib64iproute2-static-devel-4.4.0-1.mga5.x86_64.rpm

vboxadditions-kernel-4.4.9-desktop-1.mga5-5.0.20-2.mga5.x86_64.rpm
vboxadditions-kernel-4.4.9-server-1.mga5-5.0.20-2.mga5.x86_64.rpm
vboxadditions-kernel-desktop-latest-5.0.20-2.mga5.x86_64.rpm
vboxadditions-kernel-server-latest-5.0.20-2.mga5.x86_64.rpm

virtualbox-kernel-4.4.9-desktop-1.mga5-5.0.20-2.mga5.x86_64.rpm
virtualbox-kernel-4.4.9-server-1.mga5-5.0.20-2.mga5.x86_64.rpm
virtualbox-kernel-desktop-latest-5.0.20-2.mga5.x86_64.rpm
virtualbox-kernel-server-latest-5.0.20-2.mga5.x86_64.rpm

xfsdump-3.1.6-1.mga5.x86_64.rpm
xfsprogs-4.3.0-1.mga5.x86_64.rpm
lib64xfs1-4.3.0-1.mga5.x86_64.rpm
lib64xfs-devel-4.3.0-1.mga5.x86_64.rpm
lib64xfs-static-devel-4.3.0-1.mga5.x86_64.rpm

dkms-xtables-addons-2.10-1.mga5.x86_64.rpm
lib64account0-2.10-1.mga5.x86_64.rpm
lib64account-devel-2.10-1.mga5.x86_64.rpm
xtables-addons-2.10-1.mga5.x86_64.rpm
xtables-addons-kernel-4.4.9-desktop-1.mga5-2.10-4.mga5.x86_64.rpm
xtables-addons-kernel-4.4.9-server-1.mga5-2.10-4.mga5.x86_64.rpm
xtables-addons-kernel-desktop-latest-2.10-4.mga5.x86_64.rpm
xtables-addons-kernel-server-latest-2.10-4.mga5.x86_64.rpm
xtables-geoip-2.10-1.mga5.noarch.rpm

iwlwifi-agn-ucode-20160503-1.mga5.nonfree.noarch.rpm
kernel-firmware-nonfree-20160503-1.mga5.nonfree.noarch.rpm
radeon-firmware-20160429-1.mga5.nonfree.noarch.rpm
ralink-firmware-20160503-1.mga5.nonfree.noarch.rpm
rtlwifi-firmware-20160503-1.mga5.nonfree.noarch.rpm

broadcom-wl-common-6.30.223.271-4.mga5.nonfree.x86_64.rpm
broadcom-wl-kernel-4.4.9-desktop-1.mga5-6.30.223.271-9.mga5.nonfree.x86_64.rpm
broadcom-wl-kernel-4.4.9-server-1.mga5-6.30.223.271-9.mga5.nonfree.x86_64.rpm
broadcom-wl-kernel-desktop-latest-6.30.223.271-9.mga5.nonfree.x86_64.rpm
broadcom-wl-kernel-server-latest-6.30.223.271-9.mga5.nonfree.x86_64.rpm
dkms-broadcom-wl-6.30.223.271-4.mga5.nonfree.x86_64.rpm

dkms-fglrx-15.302-4.mga5.nonfree.x86_64.rpm
fglrx-control-center-15.302-4.mga5.nonfree.x86_64.rpm
fglrx-devel-15.302-4.mga5.nonfree.x86_64.rpm
fglrx-kernel-4.4.9-desktop-1.mga5-15.302-7.mga5.nonfree.x86_64.rpm
fglrx-kernel-4.4.9-server-1.mga5-15.302-7.mga5.nonfree.x86_64.rpm
fglrx-kernel-desktop-latest-15.302-7.mga5.nonfree.x86_64.rpm
fglrx-kernel-server-latest-15.302-7.mga5.nonfree.x86_64.rpm
fglrx-opencl-15.302-4.mga5.nonfree.x86_64.rpm
x11-driver-video-fglrx-15.302-4.mga5.nonfree.x86_64.rpm

nvidia304-kernel-4.4.9-desktop-1.mga5-304.131-5.mga5.nonfree.x86_64.rpm
nvidia304-kernel-4.4.9-server-1.mga5-304.131-5.mga5.nonfree.x86_64.rpm
nvidia304-kernel-desktop-latest-304.131-5.mga5.nonfree.x86_64.rpm
nvidia304-kernel-server-latest-304.131-5.mga5.nonfree.x86_64.rpm

nvidia340-kernel-4.4.9-desktop-1.mga5-340.96-5.mga5.nonfree.x86_64.rpm
nvidia340-kernel-4.4.9-server-1.mga5-340.96-5.mga5.nonfree.x86_64.rpm
nvidia340-kernel-desktop-latest-340.96-5.mga5.nonfree.x86_64.rpm
nvidia340-kernel-server-latest-340.96-5.mga5.nonfree.x86_64.rpm

nvidia-current-kernel-4.4.9-desktop-1.mga5-352.79-8.mga5.nonfree.x86_64.rpm
nvidia-current-kernel-4.4.9-server-1.mga5-352.79-8.mga5.nonfree.x86_64.rpm
nvidia-current-kernel-desktop-latest-352.79-8.mga5.nonfree.x86_64.rpm
nvidia-current-kernel-server-latest-352.79-8.mga5.nonfree.x86_64.rpm

Whiteboard: feedback => (none)

depend on virtualbox, nvidia-current, microcode bugs to go out before or at the same time as this update.

block kernel-linus and kernel-tmb as they depends on firmwares/drivers/utils pushed as part of this update

Blocks: 17604, 17174 => 18374, 18375
Depends on: (none) => 17604, 18236, 18372
Summary: Update request: kernel 4.4.6 => Update request: kernel 4.4.9

In VirtualBox, M5, KDE, 64-bit

Package(s) under test:
kernel-desktop-latest vboxadditions-kernel-desktop-latest

default install of kernel-desktop586-latest

[root@localhost wilcal]# uname -a
Linux localhost.localdomain 4.1.15-desktop-2.mga5 #1 SMP Wed Jan 20 17:05:51 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.16-3.mga5.x86_64 is already installed

System boots to a working desktop. Common apps work. Screen dimensions are correct.

install kernel-desktop-latest & vboxadditions-kernel-desktop-latest from updates_testing

[root@localhost wilcal]# uname -a
Linux localhost.localdomain 4.4.9-desktop-1.mga5 #1 SMP Tue May 3 20:38:36 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.4.9-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.20-2.mga5.x86_64 is already installed

System boots to a working desktop. Common apps work. Screen dimensions are correct.

In VirtualBox, M5, KDE, 32-bit

Package(s) under test:
kernel-desktop-latest vboxadditions-kernel-desktop-latest

default install of kernel-desktop586-latest

[root@localhost wilcal]# uname -a
Linux localhost.localdomain 4.1.15-desktop586-2.mga5 #1 SMP Wed Jan 20 17:06:34 UTC 2016 i686 i686 i686 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.1.15-2.mga5.i586 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.16-3.mga5.i586 is already installed

System boots to a working desktop. Common apps work. Screen dimensions are correct.

install kernel-desktop-latest & vboxadditions-kernel-desktop-latest from updates_testing

[root@localhost wilcal]# uname -a
Linux localhost.localdomain 4.1.15-desktop586-2.mga5 #1 SMP Wed Jan 20 17:06:34 UTC 2016 i686 i686 i686 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.4.9-1.mga5.i586 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.20-2.mga5.i586 is already installed

System boots to a working desktop. Common apps work. Screen dimensions are correct.

on this system:

<<<<<<

I also have an Acer Aspire 5738DZG laptop with the following specs:

    Intel Pentium(R) Dual-Core CPU T4300 @ 2.10GHz. (x86-64).
    ATI Mobility Radeon⢠HD 4570 (r700)
    15.6â³ 3D HD LCD Screen.
    3 GB Memory
    320 GB Hard Disk Drive.
    âDVD Super Multi DL driveâ
    Acer Nplify⢠802.11b/g/n.

>>>>>>

I updated from the 4.4.6 kernel:

* X works. (Xfce)
* Firefox works.
* HexChat works (send+receive)
* Konqueror works.
* VLC plays audio+video
* SMB/CIFS works.
* extremetuxracer works.

*uname -a
Linux localhost 4.4.9-desktop-1.mga5 #1 SMP Tue May 3 20:38:36 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
shlomif@lap:~$ 

* ssh works.
* tmux works.
* konsole works.

Seems fine here.

Video:
Identification
Vendor: âAdvanced Micro Devices, Inc. [AMD/ATI]

Description: âRS780L [Radeon 3000]

--------

Identification
Processor ID: â1

Vendor: âAuthenticAMD

Model name: âAMD Athlon(tm) II X3 450 Processor


----------------------------------------

During Install 4.4.9 Desktop, the following messages popped up.

To satisfy dependencies, the following package(s) also need to be installed:

- dracut-038-21.mga5.x86_64
- kernel-firmware-20160409-1.mga5.noarch

19MB of disk space will be freed.



The following packages have to be removed for others to be upgraded:
kernel-firmware-nonfree-20151018-1.mga5.nonfree.noarch
 (due to conflicts with kernel-firmware-20160409-1.mga5.noarch)
microcode-0.20140913-4.mga5.nonfree.noarch
 (due to conflicts with kernel-desktop-4.4.9-1.mga5-1-1.mga5.x86_64)
radeon-firmware-20150824-1.mga5.nonfree.noarch
 (due to conflicts with kernel-desktop-4.4.9-1.mga5-1-1.mga5.x86_64)

----------------------------------

Rebooted to blank screen no progress.  I collected JournalCTL and will save in next post.

Created attachment 7749 [details]
kernel4.4.9 - journal ctl - AMD/Radeon - no screen after install.

Created attachment 7750 [details]
kernel4.4.9 - dmesg - AMD/Radeon

(In reply to Brian Rockwell from comment #58)

> 
> During Install 4.4.9 Desktop, the following messages popped up.


> The following packages have to be removed for others to be upgraded:
> kernel-firmware-nonfree-20151018-1.mga5.nonfree.noarch
>  (due to conflicts with kernel-firmware-20160409-1.mga5.noarch)
> microcode-0.20140913-4.mga5.nonfree.noarch
>  (due to conflicts with kernel-desktop-4.4.9-1.mga5-1-1.mga5.x86_64)
> radeon-firmware-20150824-1.mga5.nonfree.noarch
>  (due to conflicts with kernel-desktop-4.4.9-1.mga5-1-1.mga5.x86_64)


You forgot to enable nonfree updates testing...

you need the updated kernel-firmware-nonfree, microcode, radeon-firmware from there ...

as it's now the removed radeon-firmware gave the "reboot to black screen"

[brian@localhost ~]$ uname -a
Linux localhost 4.4.9-desktop-1.mga5 #1 SMP Tue May 3 20:38:36 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux


Thanks Thomas - that worked. 


It is working as designed now.

Broadcom 4318 wireless still failing. It appears that the kernel is looking for a file named "pcm5.fw" in /lib/firmware/b43. The same thing is happening with Cauldron. If I download openfwwf-jam-5.3.tar.gz from http://netweb.ing.unibs.it/~openfwwf/firmware/ and copy the needed file from that archive into /lib/firmware/b43 the wifi then works OK.

However, unlike the other b43-openfwwf packages on the site, this one is just binaries, with no source code. The b43-openfwwf maintainer said in bug 17990 that he had asked about source code, but I haven't heard anything beyond that. It may be that it's proprietary, and we can't distribute it. I've looked with Google and so far have only found references to extracting it from downloaded proprietary drivers via b43-fwcutter. 

Other than that, the 4.4.9 kernel is working fine.

It does seem to be included in opensuse package

x86_64

Intel Core i7-4790K 4GHz
nvidia GTX 770
Atheros Killer E220x Gigabit Ethernet

Packages installed:
* cpupower-4.4.9-1.mga5.x86_64.rpm
* kernel-desktop-4.4.9-1.mga5-1-1.mga5.x86_64.rpm
* kernel-desktop-devel-4.4.9-1.mga5-1-1.mga5.x86_64.rpm
* kernel-desktop-devel-latest-4.4.9-1.mga5.x86_64.rpm
* kernel-desktop-latest-4.4.9-1.mga5.x86_64.rpm
* perf-4.4.9-1.mga5.x86_64.rpm
* kernel-userspace-headers-4.4.9-1.mga5.x86_64.rpm
* kernel-firmware-20160409-1.mga5.noarch.rpm
* aufs-tools-4.4-0.git20160301.1.mga5.x86_64.rpm
* dracut-038-21.mga5.x86_64.rpm
* iproute2-4.4.0-1.mga5.x86_64.rpm
* iptaccount-2.10-1.mga5.x86_64.rpm
* xfsdump-3.1.6-1.mga5.x86_64.rpm
* xfsprogs-4.3.0-1.mga5.x86_64.rpm
* lib64xfs1-4.3.0-1.mga5.x86_64.rpm
* dkms-xtables-addons-2.10-1.mga5.x86_64.rpm
* lib64account0-2.10-1.mga5.x86_64.rpm
* xtables-addons-2.10-1.mga5.x86_64.rpm
* xtables-addons-kernel-4.4.9-desktop-1.mga5-2.10-4.mga5.x86_64.rpm
* xtables-addons-kernel-desktop-latest-2.10-4.mga5.x86_64.rpm
* xtables-geoip-2.10-1.mga5.noarch.rpm
* iwlwifi-agn-ucode-20160503-1.mga5.nonfree.noarch.rpm
* broadcom-wl-common-6.30.223.271-4.mga5.nonfree.x86_64.rpm
* nvidia-current-kernel-4.4.9-desktop-1.mga5-352.79-8.mga5.nonfree.x86_64.rpm
* nvidia-current-kernel-desktop-latest-352.79-8.mga5.nonfree.x86_64.rpm

Booted to a working desktop.
Had to unprotect ethernet interface after reboot.

ldetect-lst-0.1.346.1-1.mga5.x86_64 is the latest installed.  ?

Nothing breaks, yet.
Leaving virtualbox until later.

(In reply to Len Lawrence from comment #65)

> Had to unprotect ethernet interface after reboot.

What do you mean Len?

Shorewall, uncheck eth*

Has the interface been renamed back to eth* from eg. enp3s4 ?
(similar to cauldron problem)

Check with..
# ifconfig 

and also check /etc/shorewall/interfaces

ifconfig shows enp3s0.  Sorry I should have been more specific.

/etc/shorewall/interfaces

net	wlp6s0	detect
loc	docker0	detect	bridge
loc	enp3s0	detect

MGA5-32 on AcerD620 Xfce (Radeon Xpress 1200/1250/1270 video and Atheros AR242x / AR542x Wireless Network Adapter ).
No installation issues (added ldetect).
Al seems OK after reboot.
No testing of vbox on this machine.

CC: (none) => herman.viaene

MGA5-32 on AcerD620 Xfce (Radeon Xpress 1200/1250/1270 video and Atheros AR242x / AR542x Wireless Network Adapter ).
No installation issues (added ldetect).
Al seems OK after reboot with wifi. Plugging in ethernet cable immediately blocks internet connection.
No testing of vbox on this machine.

Stopping shorewall6 (shorewall was not active) in MCC allows ethernet connection to work properly.

FWIW, I use kernel 4.4.9 x86_64 with Intel Skylake i5-6500 + Intel HD Graphics 530 on real hardware. Except GUI problems already mentioned in comment 48 (due to Mesa 10.5.9? Looks like Mesa 11.0 is required for Skylake), everything else is working fine, including using Windows 10 in virtualbox.

x86_64 platform.
Updated virtualbox/i586 working fine with the kernel 4.4.9-desktop-1.mga5.
Bridged network and openssl functioning.  Transferred USB TV tuner from host to vbox - vlc able to stream live HD TV.  Moved it back to the host OK.

mga5 64

So far so good Haswell i3 UEFI laptop.

No problems installing the kernel. Works fine here. Also bumblebee is working.

System:    Host: kre010.fritz.box Kernel: 4.4.9-desktop-1.mga5 x86_64 (64 bit gcc: 4.9.2) 
           Desktop: KDE 4.14.5 (Qt 4.8.6) info: plasma-desktop dm: kdm Distro: Mageia 5 thornicroft 
Machine:   System: Acer product: Aspire V3-571G v: V1.06 Chassis: type: 10 v: V1.06
           Mobo: Acer model: VA50_HC_CR v: Type2 - Board Version Bios: Acer v: V1.06 date: 04/20/2012
CPU:       Dual core Intel Core i5-3210M (-HT-MCP-) cache: 3072 KB
           flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx) bmips: 9977 
           Clock Speeds: 1: 1397 MHz 2: 1235 MHz 3: 1303 MHz 4: 1221 MHz
Graphics:  Card-1: Intel 3rd Gen Core processor Graphics Controller bus-ID: 00:02.0 chip-ID: 8086:0166
           Card-2: NVIDIA GF108M [GeForce GT 630M] bus-ID: 01:00.0 chip-ID: 10de:0de9
           Display Server: X.Org 1.16.4 drivers: v4l,intel (unloaded: nouveau) Resolution: 1366x768@60.01hz
           GLX Renderer: Mesa DRI Intel Ivybridge Mobile GLX Version: 3.0 Mesa 10.5.9 Direct Rendering: Yes

CC: (none) => alfred.kretschmer

A glutton for punishment, I decided to try the following updates, all at once:

- cpupower-4.4.9-1.mga5.x86_64
- dkms-virtualbox-5.0.20-1.mga5.noarch
- dracut-038-21.mga5.x86_64
- iproute2-4.4.0-1.mga5.x86_64
- iwlwifi-agn-ucode-20160503-1.mga5.nonfree.noarch
- kernel-firmware-20160409-1.mga5.noarch
- kernel-firmware-nonfree-20160503-1.mga5.nonfree.noarch
- kernel-server-4.4.9-1.mga5-1-1.mga5.x86_64
- kernel-server-devel-4.4.9-1.mga5-1-1.mga5.x86_64
- kernel-server-devel-latest-4.4.9-1.mga5.x86_64
- kernel-server-latest-4.4.9-1.mga5.x86_64
- kernel-userspace-headers-4.4.9-1.mga5.x86_64
- ldetect-lst-0.1.346.4-1.mga5.x86_64
- microcode-0.20160409-1.mga5.nonfree.noarch
- nvidia340-kernel-4.4.9-server-1.mga5-340.96-5.mga5.nonfree.x86_64
- nvidia340-kernel-server-latest-340.96-5.mga5.nonfree.x86_64
- radeon-firmware-20160429-1.mga5.nonfree.noarch
- ralink-firmware-20160503-1.mga5.nonfree.noarch
- rtlwifi-firmware-20160503-1.mga5.nonfree.noarch
- virtualbox-kernel-4.4.9-server-1.mga5-5.0.20-2.mga5.x86_64
- virtualbox-kernel-server-latest-5.0.20-2.mga5.x86_64
- wireless-regdb-20160502-1.mga5.noarch
- x11-driver-video-intel-2.99.917-14.2.mga5.x86_64
- xfsprogs-4.3.0-1.mga5.x86_64

Somehow I missed selecting these two:
- virtualbox-5.0.20-1.mga5.x86_64
- virtualbox-kernel-4.1.15-server-2.mga5-5.0.20-1.mga5.x86_64
so I installed them after the reboot.

Everything seems to be working just as it should, though I haven't tried starting my Mageia guest with virtualbox yet. I did, however, successfully download and install the extpack, played a game of Klondike, and opened Firefox to access this bug report.

Kernel update successful in 64-bit Mageia 5 guest in virtualbox 5.0.20.

No regressions noticed for kernel 4.4.9.

Tested with KDE4/Plasma desktop, nvidia-settings, steam games, smplayer using vpdau, audio play, audio record (front mic, rear mic), etc.

$ uname -a
Linux marte 4.4.9-desktop-1.mga5 #1 SMP Tue May 3 20:38:36 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
$ cat /proc/driver/nvidia/version 
NVRM version: NVIDIA UNIX x86_64 Kernel Module  340.96  Sun Nov  8 22:33:28 PST 2015
GCC version:  gcc version 4.9.2 (GCC) 
$ lspcidrake | grep AUDIO
snd_hda_intel   : NVIDIA Corporation|High Definition Audio Controller [MULTIMEDIA_AUDIO_DEV] (rev: a1)
snd_hda_intel   : Intel Corporation|82801JI (ICH10 Family) HD Audio Controller [MULTIMEDIA_AUDIO_DEV]
$ lspcidrake | grep VGA
Card:NVIDIA GeForce 8100 to GeForce 415: NVIDIA Corporation|GT218 [GeForce 210] [DISPLAY_VGA] (rev: a2)
$ lspci | grep VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GT218 [GeForce 210] (rev a2)

Same hardware as Comment 78, 32-bit install, also server kernel with Virtualbox. Same procedure as with the 64-bit install, update, except that I took care to select ALL the appropriate packages the first time.

Updates were all successful, and everything appears to be working normally.

Tested on several hardware :

32 bit for Dell D600, and a 2007 Fujitsu laptops. No regression noted with sensors, sound, opengl and wifi.

64 bit for a 2009 AMD Athlon X2 and a 2008 Intel Core. All looks good.

I will report if any problem appears in the next days.

CC: (none) => lists.jjorge

On real hardware, M5, KDE, 64-bit

Start with a completely blank drive, nouveau driver.

default install of:
kernel-desktop-latest
virtualbox vboxadditions-kernel-desktop-latest dkms-virtualbox
virtualbox-guest-additions virtualbox-kernel-desktop-latest x11-driver-video-vboxvideo

The following 20 packages are going to be installed:

- dkms-2.0.19-34.mga5.noarch
- dkms-minimal-2.0.19-34.mga5.noarch
- dkms-virtualbox-5.0.16-1.mga5.noarch
- gcc-4.9.2-4.1.mga5.x86_64
- gcc-cpp-4.9.2-4.1.mga5.x86_64
- glibc-devel-2.20-21.mga5.x86_64
- kernel-desktop-devel-4.1.15-2.mga5-1-1.mga5.x86_64
- kernel-desktop-devel-latest-4.1.15-2.mga5.x86_64
- kernel-userspace-headers-4.1.15-2.mga5.x86_64
- lib64mpc3-1.0.2-4.mga5.x86_64
- lib64ncurses-devel-5.9-21.mga5.x86_64
- libstdc++5-3.3.6-11.mga5.x86_64
- libstdc++5-devel-3.3.6-11.mga5.x86_64
- vboxadditions-kernel-4.1.15-desktop-2.mga5-5.0.16-3.mga5.x86_64
- vboxadditions-kernel-desktop-latest-5.0.16-3.mga5.x86_64
- virtualbox-5.0.16-1.mga5.x86_64
- virtualbox-guest-additions-5.0.16-1.mga5.x86_64
- virtualbox-kernel-4.1.15-desktop-2.mga5-5.0.16-3.mga5.x86_64
- virtualbox-kernel-desktop-latest-5.0.16-3.mga5.x86_64
- x11-driver-video-vboxvideo-5.0.16-1.mga5.x86_64

[root@localhost wilcal]# uname -a
Linux localhost.localdomain 4.1.15-desktop-2.mga5 #1 SMP Wed Jan 20 17:05:51 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox
Package virtualbox-5.0.16-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.16-3.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi dkms-virtualbox
Package dkms-virtualbox-5.0.16-1.mga5.noarch is already installed
[root@localhost wilcal]# urpmi virtualbox-guest-additions
Package virtualbox-guest-additions-5.0.16-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox-kernel-desktop-latest
Package virtualbox-kernel-desktop-latest-5.0.16-3.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi x11-driver-video-vboxvideo
Package x11-driver-video-vboxvideo-5.0.16-1.mga5.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nouveau
        Kernel modules: nvidiafb, nouveau

Created client:
M5 i586 KDE Live-DVD runs as a Vbox client. Screen sizes are correct. Sound ok

install from updates_testing:
kernel-desktop-latest
virtualbox vboxadditions-kernel-desktop-latest dkms-virtualbox
virtualbox-guest-additions virtualbox-kernel-desktop-latest x11-driver-video-vboxvideo
from updates_testing

The following 18 packages are going to be installed:

- btrfs-progs-4.4.1-1.mga5.x86_64
- dkms-virtualbox-5.0.20-1.mga5.noarch
- dracut-038-21.mga5.x86_64
- kernel-desktop-4.4.9-1.mga5-1-1.mga5.x86_64
- kernel-desktop-latest-4.4.9-1.mga5.x86_64
- kernel-firmware-20160409-1.mga5.noarch
- kernel-firmware-nonfree-20160503-1.mga5.nonfree.noarch
- lib64btrfs0-4.4.1-1.mga5.x86_64
- microcode-0.20160409-1.mga5.nonfree.noarch
- radeon-firmware-20160429-1.mga5.nonfree.noarch
- vboxadditions-kernel-4.1.15-desktop-2.mga5-5.0.20-1.mga5.x86_64
- vboxadditions-kernel-desktop-latest-5.0.20-1.mga5.x86_64
- virtualbox-5.0.20-1.mga5.x86_64
- virtualbox-guest-additions-5.0.20-1.mga5.x86_64
- virtualbox-kernel-4.1.15-desktop-2.mga5-5.0.20-1.mga5.x86_64
- virtualbox-kernel-4.4.9-desktop-1.mga5-5.0.20-2.mga5.x86_64
- virtualbox-kernel-desktop-latest-5.0.20-2.mga5.x86_64
- x11-driver-video-vboxvideo-5.0.20-1.mga5.x86_64

[root@localhost wilcal]# uname -a
Linux localhost 4.4.9-desktop-1.mga5 #1 SMP Tue May 3 20:38:36 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.4.9-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox
Package virtualbox-5.0.20-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.20-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi dkms-virtualbox
Package dkms-virtualbox-5.0.20-1.mga5.noarch is already installed
[root@localhost wilcal]# urpmi virtualbox-guest-additions
Package virtualbox-guest-additions-5.0.20-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox-kernel-desktop-latest
Package virtualbox-kernel-desktop-latest-5.0.20-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi x11-driver-video-vboxvideo
Package x11-driver-video-vboxvideo-5.0.20-1.mga5.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nouveau
        Kernel modules: nvidiafb, nouveau
        
Reboot system.
        
Create client:
M6 x86_64 Plasma Live-DVD runs as a Vbox client. Screen sizes are correct. Sound ok
Still runs correctly:
M5 i586 KDE Live-DVD runs as a Vbox client. Screen sizes are correct. Sound ok

Now install the nvidia driver:

default install of:
nvidia-current-kernel-desktop-latest

reboot system successful.

[root@localhost wilcal]# urpmi nvidia-current-kernel-desktop-latest
Package nvidia-current-kernel-desktop-latest-352.79-8.mga5.nonfree.x86_64 is already installed
[root@localhost wilcal]# uname -a
Linux localhost 4.4.9-desktop-1.mga5 #1 SMP Tue May 3 20:38:36 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.4.9-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox
Package virtualbox-5.0.20-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.20-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi dkms-virtualbox
Package dkms-virtualbox-5.0.20-1.mga5.noarch is already installed
[root@localhost wilcal]# urpmi virtualbox-guest-additions
Package virtualbox-guest-additions-5.0.20-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox-kernel-desktop-latest
Package virtualbox-kernel-desktop-latest-5.0.20-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi x11-driver-video-vboxvideo
Package x11-driver-video-vboxvideo-5.0.20-1.mga5.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nvidia
        Kernel modules: nvidiafb, nouveau, nvidia_current
        
Created clients:
M6 x86_64 Gnome Live-DVD runs as a Vbox client. Screen sizes are correct. Sound ok
M6 x86_64 Plasma Live-DVD, installed, runs as a Vbox client. Screen sizes are correct. Sound ok
Still runs correctly:
M6 x86_64 Plasma Live-DVD runs as a Vbox client. Screen sizes are correct. Sound ok
M5 i586 KDE Live-DVD runs as a Vbox client. Screen sizes are correct. Sound ok

System reboots back to a working desktop.

(In reply to David Walser from comment #18)
> On my newer PogoLinux machine at home, working fine for the most part, but I
> do have one significant (to me) issue.
> 
> I have an ATI TV Wonder VE card, and for audio, there's a 1/8" audio cable
> going out from the TV card and into the sound card, and then the "Line"
> input in the mixer (aumix -l at the command line) is where the sound coming
> in is.  With the updated kernel, I no longer get sound if Line is turned on.

Line In audio is still broken on 4.4.9.

CC: (none) => luigiwalser
Whiteboard: (none) => feedback

(In reply to David Walser from comment #39)
> The verdict is that VBE is broken, as is likely anything using libx86.

Still the case with 4.4.9 :o(

About to try x64 real EFI h/w with ATI/Radeon/AMD video

Taking mainly note of David's (+ earlier) qa-discuss posts: "so testing with the updated microcode, dmidecode, wireless-regdb, kernel-firmware, kernel-firmware-nonfree, radeon-firmware, aufs-tools, btrfs-progs, dracut, iproute2, xfsdump, and xfsprogs packages (as relevant to your system) is still a good idea" - here is what I updated:-
 btrfs-progs-4.4.1-1.mga5.x86_64
 cpupower-4.4.9-1.mga5.x86_64
 dkms-fglrx-15.302-4.mga5.nonfree.x86_64
 dmidecode-3.0-1.mga5.x86_64
 dracut-038-21.mga5.x86_64
 fglrx-control-center-15.302-4.mga5.nonfree.x86_64
 fglrx-kernel-4.4.9-desktop-1.mga5-15.302-7.mga5.nonfree.x86_64
 fglrx-kernel-desktop-latest-15.302-7.mga5.nonfree.x86_64
 iproute2-4.4.0-1.mga5.x86_64
 kernel-desktop-4.4.9-1.mga5-1-1.mga5.x86_64
 kernel-desktop-devel-4.4.9-1.mga5-1-1.mga5.x86_64
 kernel-desktop-devel-latest-4.4.9-1.mga5.x86_64
 kernel-desktop-latest-4.4.9-1.mga5.x86_64
 kernel-firmware-20160409-1.mga5.noarch
 kernel-userspace-headers-4.4.9-1.mga5.x86_64
 ldetect-lst-0.1.346.4-1.mga5.x86_64
 lib64btrfs0-4.4.1-1.mga5.x86_64
 microcode-0.20160409-1.mga5.nonfree.noarch
 radeon-firmware-20160429-1.mga5.nonfree.noarch
 wireless-regdb-20160502-1.mga5.noarch
 x11-driver-video-fglrx-15.302-4.mga5.nonfree.x86_64
 x11-driver-video-intel-2.99.917-14.2.mga5.x86_64
 xfsprogs-4.3.0-1.mga5.x86_64

Having no Intel video hardware, that X11 driver is enigmatic; ditto for wireless.
Writing this before re-booting...

CC: (none) => lewyssmith

Further to Comment 87.
Untested x64 real h/w

Curses: unbootable result.
I shall have to re-install/Update Mageia 5. I saw too late that a few of the updates had their own Bug numbers:
 microcode
 dmidecode
 wirelessregdb
 X11Intelvideo
which I could presumably have installed alone to try just *them*.

(In reply to David Walser from comment #85)
> (In reply to David Walser from comment #18)
> > On my newer PogoLinux machine at home, working fine for the most part, but I
> > do have one significant (to me) issue.
> > 
> > I have an ATI TV Wonder VE card, and for audio, there's a 1/8" audio cable
> > going out from the TV card and into the sound card, and then the "Line"
> > input in the mixer (aumix -l at the command line) is where the sound coming
> > in is.  With the updated kernel, I no longer get sound if Line is turned on.
> 
> Line In audio is still broken on 4.4.9.

Does it happend with kernel-linus ?
Have you checked if PulseAudio have muted it?

(In reply to Thomas Backlund from comment #89)
> (In reply to David Walser from comment #85)
> > (In reply to David Walser from comment #18)
> > > On my newer PogoLinux machine at home, working fine for the most part, but I
> > > do have one significant (to me) issue.
> > > 
> > > I have an ATI TV Wonder VE card, and for audio, there's a 1/8" audio cable
> > > going out from the TV card and into the sound card, and then the "Line"
> > > input in the mixer (aumix -l at the command line) is where the sound coming
> > > in is.  With the updated kernel, I no longer get sound if Line is turned on.
> > 
> > Line In audio is still broken on 4.4.9.
> 
> Does it happend with kernel-linus ?

I have not tried kernel-linus yet.  I'll hopefully have time tomorrow evening.

> Have you checked if PulseAudio have muted it?

I can't use PulseAudio precisely because I use Line In, so no, it's not an issue.

x64 real EFI hardware with AMD processor & video

After a dicey 'update' re-installation (the final bootloader installation failed once, worked with no change on a repeat), my again working system shows:
 $ uname -r
 4.4.9-desktop-1.mga5
BTW I was not aware of any dkms re-compilation. Did I overlook something?
I have:
 dkms-fglrx-15.302-4.mga5.nonfree
 dkms-minimal-2.0.19-34.mga5
 dkms-2.0.19-34.mga5

fglrx-kernel-3.19.8-desktop-3.mga5-15.200.1046-1.1.mga5.nonfree
fglrx-kernel-4.1.15-desktop-2.mga5-15.302-1.mga5.nonfree
fglrx-kernel-4.4.9-desktop-1.mga5-15.302-7.mga5.nonfree
fglrx-kernel-desktop-latest-15.302-7.mga5.nonfree
kernel-desktop-3.19.8-3.mga5-1-1.mga5
kernel-desktop-4.1.15-2.mga5-1-1.mga5
kernel-desktop-4.4.9-1.mga5-1-1.mga5
kernel-desktop-devel-4.1.12-1.mga5-1-1.mga5
kernel-desktop-devel-4.1.15-1.mga5-1-1.mga5
kernel-desktop-devel-4.1.15-2.mga5-1-1.mga5
kernel-desktop-devel-4.4.9-1.mga5-1-1.mga5
kernel-desktop-devel-latest-4.4.9-1.mga5
kernel-desktop-latest-4.4.9-1.mga5
kernel-firmware-20160409-1.mga5
kernel-userspace-headers-4.4.9-1.mga5

Is anyone prepared to comment on what bit of the Comment 87 list might have caused the aggro? One has to assume that all the peripheral bits & pieces updates are inherently OK, because the now-bootable system works.

(In reply to David Walser from comment #40)
> A longstanding issue with the i586 server kernel is that hard disk I/O is
> significantly slower than with the desktop kernel in some cases.  4.4.6 has
> exacerbated this issue quite a bit, and now just trying to rsync files from
> one disk to another on our big server or trying to extract a large tarball
> on my workstation are so slow as to be impractical.

Anything interesting in logs ?
I dont have any issue... so I guess something hw related...

(In reply to David Walser from comment #86)
> (In reply to David Walser from comment #39)
> > The verdict is that VBE is broken, as is likely anything using libx86.
> 
> Still the case with 4.4.9 :o(

Heh, vbetool is ancient code... (since ~2008)... I guess a bisect between working 4.1 and 4.4 is needed...

Does it work in cauldron ?

(In reply to Thomas Backlund from comment #92)
> (In reply to David Walser from comment #40)
> > A longstanding issue with the i586 server kernel is that hard disk I/O is
> > significantly slower than with the desktop kernel in some cases.  4.4.6 has
> > exacerbated this issue quite a bit, and now just trying to rsync files from
> > one disk to another on our big server or trying to extract a large tarball
> > on my workstation are so slow as to be impractical.
> 
> Anything interesting in logs ?
> I dont have any issue... so I guess something hw related...

It's noticeable on more than one class of hardware, but on the big Dell servers it was particularly severe.  It became unbearable with the 4.4 kernels, so I reinstalled one server as x86_64, which fixed the issue, and the other one is running the desktop kernel now.  There's nothing interesting in logs, it's just super slow on the i586 server kernel.  The servers use a hardware RAID, that's the only interesting difference I can think of for them vs. other computers.

(In reply to Thomas Backlund from comment #93)
> (In reply to David Walser from comment #86)
> > (In reply to David Walser from comment #39)
> > > The verdict is that VBE is broken, as is likely anything using libx86.
> > 
> > Still the case with 4.4.9 :o(
> 
> Heh, vbetool is ancient code... (since ~2008)... I guess a bisect between
> working 4.1 and 4.4 is needed...

Probably a bisect is needed, but are you sure you didn't disable some CONFIG option that disabled that system call (see Comment 39)?

> Does it work in cauldron ?

I have no idea.

(In reply to David Walser from comment #90)
> (In reply to Thomas Backlund from comment #89)
> > (In reply to David Walser from comment #85)
> > > (In reply to David Walser from comment #18)
> > > > On my newer PogoLinux machine at home, working fine for the most part, but I
> > > > do have one significant (to me) issue.
> > > > 
> > > > I have an ATI TV Wonder VE card, and for audio, there's a 1/8" audio cable
> > > > going out from the TV card and into the sound card, and then the "Line"
> > > > input in the mixer (aumix -l at the command line) is where the sound coming
> > > > in is.  With the updated kernel, I no longer get sound if Line is turned on.
> > > 
> > > Line In audio is still broken on 4.4.9.
> > 
> > Does it happend with kernel-linus ?
> 
> I have not tried kernel-linus yet.  I'll hopefully have time tomorrow
> evening.
> 

An other thing to help debug this is to download: http://www.alsa-project.org/alsa-info.sh

then run it with: 
sh alsa-info.sh --no-upload

once with working 4.1 and once with non-working 4.4

and compare the output from both tests to see what it detects differently

and hopefully figure out why

(In reply to David Walser from comment #95)
> (In reply to Thomas Backlund from comment #93)
> > (In reply to David Walser from comment #86)
> > > (In reply to David Walser from comment #39)
> > > > The verdict is that VBE is broken, as is likely anything using libx86.
> > > 
> > > Still the case with 4.4.9 :o(
> > 
> > Heh, vbetool is ancient code... (since ~2008)... I guess a bisect between
> > working 4.1 and 4.4 is needed...
> 
> Probably a bisect is needed, but are you sure you didn't disable some CONFIG
> option that disabled that system call (see Comment 39)?
> 

Ah, indeed I missed that comment...

And yes, thats it...

in 4.3-rc1 this landed upstream:

commit 5aef51c340cb50ed9a3997dc5d782324372078bd
Author: Andy Lutomirski <luto@kernel.org>
Date:   Fri Jul 10 08:34:23 2015 -0700

    x86/kconfig/32: Rename CONFIG_VM86 and default it to 'n'
    
    VM86 is entirely broken if ptrace, syscall auditing, or
    NOHZ_FULL is in use.  The code is a big undocumented mess, it's
    a real PITA to test, and it looks like a big chunk of vm86_32.c
    is dead code.  It also plays awful games with the entry asm.
    
    No one should be using it anyway. Use DOSBOX or KVM instead.
    
    Let's accelerate its slow death.  Remove it from EXPERT and
    default it to n.  Distros should not enable it.  In the unlikely
    event that some user needs it, they can easily re-enable it.
    
    While we're at it, rename it to CONFIG_X86_LEGACY_VM86 so that 'make
    oldconfig' users will be prompted again.  I left CONFIG_VM86 as
    an alias to avoid a treewide replacement of the names.  We can
    clean that up once the current asm and vm86 code churn settles
    down.


So this most likely killed of vbetool support

Interestingly the kernel kconfig suggests any "recent" vbetool should work without it, but.... 

I guess I need to re-activate this one for mga5 to not regress a running install, but for mga6 I'd like to keep it disabled and try to goet vbetool updated/fixed/....

(In reply to David Walser from comment #94)
> (In reply to Thomas Backlund from comment #92)
> > (In reply to David Walser from comment #40)
> > > A longstanding issue with the i586 server kernel is that hard disk I/O is
> > > significantly slower than with the desktop kernel in some cases.  4.4.6 has
> > > exacerbated this issue quite a bit, and now just trying to rsync files from
> > > one disk to another on our big server or trying to extract a large tarball
> > > on my workstation are so slow as to be impractical.
> > 
> > Anything interesting in logs ?
> > I dont have any issue... so I guess something hw related...
> 
> It's noticeable on more than one class of hardware, but on the big Dell
> servers it was particularly severe.  It became unbearable with the 4.4
> kernels, so I reinstalled one server as x86_64, which fixed the issue, and
> the other one is running the desktop kernel now.  There's nothing
> interesting in logs, it's just super slow on the i586 server kernel.  The
> servers use a hardware RAID, that's the only interesting difference I can
> think of for them vs. other computers.

Ah, there might be some possible fixes coming upstream then... iirc I saw some possible fixes for longstanding bugs that could maybe explain some of it...

I guess most devs run 64bit nowdays so no-one really care about 32bit anymore...

OK, I ran into a problem with the 4.4.9 kernel which isn't present in the 3.12.x kernel: my ath9k wifi driver uses a lower bitrate ( 5.5 Mbps instead of 54 Mbps) and supports lower bandwidth (~600 KBytes/s instead of 1.6 MBytes/s for internet downloads and possibly more for local ones) when on the new kernel. This seems to be a known problem based on some DuckDuckGo searches and a regression in the kernel , and the only workaround that worked was downgrading the kernel.

(In reply to Thomas Backlund from comment #96)
> An other thing to help debug this is to download:
> http://www.alsa-project.org/alsa-info.sh
> 
> then run it with: 
> sh alsa-info.sh --no-upload
> 
> once with working 4.1 and once with non-working 4.4
> 
> and compare the output from both tests to see what it detects differently
> 
> and hopefully figure out why

I'll post both outputs.  I see there are some "stream=" numbers that are different and some "Amp-In vals:" that are different and some missing "Control:" lines, but none of it means anything to me.

Created attachment 7776 [details]
alsa-info.txt for 4.1.15

Created attachment 7777 [details]
alsa-info.txt for 4.4.9

(In reply to Thomas Backlund from comment #89)
> Does it happend with kernel-linus ?

Yes, same behavior confirmed for kernel-linus as kernel-server.  Works in 4.1.15-2, doesn't in 4.4.9.

(In reply to David Walser from comment #102)
> Created attachment 7777 [details]
> alsa-info.txt for 4.4.9


Was this with "Line is turned on" or "Line is turned off"

if it was "turned on", can you rerun the script with "turned off"
(or the other way, depending on on/off"

looks like it renumbers several Controls... 
what if you rename /var/lib/alsa/asound.state and reboot under new kernel, does it behave any different ?

Created attachment 7792 [details]
alsa-info.txt for 4.1.15 with Line In turned on (aumix -l 60)

Created attachment 7793 [details]
alsa-info.txt for 4.4.9 with Line In turned on (aumix -l 60)

(In reply to Thomas Backlund from comment #104)
> (In reply to David Walser from comment #102)
> > Created attachment 7777 [details]
> > alsa-info.txt for 4.4.9
> 
> 
> Was this with "Line is turned on" or "Line is turned off"

Turned off.

> if it was "turned on", can you rerun the script with "turned off"
> (or the other way, depending on on/off"

Done.

> looks like it renumbers several Controls... 
> what if you rename /var/lib/alsa/asound.state and reboot under new kernel,
> does it behave any different ?

No.

(In reply to David Walser from comment #107)
> (In reply to Thomas Backlund from comment #104)

 > if it was "turned on", can you rerun the script with "turned off"
> > (or the other way, depending on on/off"
> 
> Done.

Thanks.

Can you test 4.4.10-1 in testing ?
I pulled in an upstream hd audio reconf bugfix there that fixes a problem when more Controls gets exposed in 4.1+

(In reply to Thomas Backlund from comment #108)
> Can you test 4.4.10-1 in testing ?

Isn't virtualbox supposed to be recompiled? After installing 4.4.10, I wanted to uninstall 4.4.9, but it wants to uninstall virtualbox too.

(In reply to Frédéric Buclin from comment #109)
> (In reply to Thomas Backlund from comment #108)
> > Can you test 4.4.10-1 in testing ?
> 
> Isn't virtualbox supposed to be recompiled? After installing 4.4.10, I
> wanted to uninstall 4.4.9, but it wants to uninstall virtualbox too.

The virtualbox update is being tabled for now, as we need to wait for upstream to release another version that hopefully fixes vboxvideo/guest-additions, but yeah we should generate a kmod package for kernel-4.4.10+vbox-5.0.16.  For now, you can install dkms-virtualbox instead if you don't want to wait.

(In reply to Shlomi Fish from comment #28)
> No regressions noticed.
> 
> Tested on Xfce, extreme tux racer, firefox+youtube+fc-solve.shlomifish.org,
> VLC, samba.
> 
> My laptop is:
> 
> <<<
>  I also have an Acer Aspire 5738DZG laptop with the following specs:
> 
>     Intel Pentium(R) Dual-Core CPU T4300 @ 2.10GHz. (x86-64).
>     ATI Mobility Radeon⢠HD 4570 (r700)
>     15.6â³ 3D HD LCD Screen.
>     3 GB Memory
>     320 GB Hard Disk Drive.
>     âDVD Super Multi DL driveâ
>     Acer Nplify⢠802.11b/g/n.
> 
> >>>

Same thing with kernel 4.4.10 and this time it seems my Atheros wifi bandwidth is better.

vbetool works with 4.4.10 on my laptop.  Line In on my workstation still doesn't.

(In reply to David Walser from comment #107)
> (In reply to Thomas Backlund from comment #104)
> > (In reply to David Walser from comment #102)
> > > Created attachment 7777 [details]
> > > alsa-info.txt for 4.4.9
> > 
> > 
> > Was this with "Line is turned on" or "Line is turned off"
> 
> Turned off.
> 
> > if it was "turned on", can you rerun the script with "turned off"
> > (or the other way, depending on on/off"
> 
> Done.
> 
> > looks like it renumbers several Controls... 
> > what if you rename /var/lib/alsa/asound.state and reboot under new kernel,
> > does it behave any different ?
> 
> No.

Ok, comparing 4.1.15 "off" vs "on" to 4.4.9 "off" vs "on" shows they flip exactly the same bits, and sets exactly same levels.

Only difference is the Control numbering...

4.1.15 uses:
        control.17 {
                iface MIXER
                name 'Line Playback Volume'
        control.18 {
                iface MIXER
                name 'Line Playback Switch'


4.4.9 uses:
        control.18 {
                iface MIXER
                name 'Line Playback Volume'
        control.19 {
                iface MIXER
                name 'Line Playback Switch'


So the Controls are bumped by one...
I still wonder if that confuses your setup...

What if you try to reset alsa with:

mv /var/lib/alsa/asound.state /var/lib/alsa/asound.state.bak

then try:
alsactl init


Does it change anything  ?

If not I guess an upstream report is needed as I cant see any specific commits in upstream git history that stands out either :/

Assigning Thomas til it's ready

CC: (none) => qa-bugs
Assignee: qa-bugs => tmb

(In reply to Thomas Backlund from comment #113)
> What if you try to reset alsa with:
> 
> mv /var/lib/alsa/asound.state /var/lib/alsa/asound.state.bak
> 
> then try:
> alsactl init
> 
> 
> Does it change anything  ?

OK, now we're getting somewhere.  If I log into KDE, run alsactl init, and then run my TV, I get sound.  When I stop it, it turns off.

There are still problems though.  Although the Line volume turning on shows in Kmix when I run my TV, and shows going back to 0/mute when I turn it off, when it's running if I click the button to mute that channel in Kmix, the display shows mute but the sound keeps going.  Playing with that channel in Kmix has no effect at all.

Also, when I log out and log back in, I have to run alsactl init again before running the TV will give sound.

Stranger still, even with it working, it's working differently.

Here's the script I use to run my TV:
aumix -l 60
DISPLAY=":0.0" mplayer -tv driver=v4l2:width=640:height=480:adevice=/dev/dsp:amode=1:immediatemode=0:norm=ntsc:input=0 tv:// > /dev/null 2>&1
aumix -l 0

In 4.1.x and earlier kernels, the aumix is what actually turns on and off the sound.  Running the mplayer command by itself only gives the video.  In 4.4.11 now, the aumix command actually does nothing but change the display in Kmix, but it does not change the sound.  The mplayer command by itself brings the sound with it when it turns the video on, and stops it itself when it exits.

(In reply to David Walser from comment #115)
> In 4.1.x and earlier kernels, the aumix is what actually turns on and off
> the sound.

If I remember well, aumix only touches the OSS emulation layer. You should use alsamixer to touch the real hardware controls of the sound card. Pressing F6 you can choose the hardware instead of pulseaudio layer.

(In reply to José Jorge from comment #116)
> (In reply to David Walser from comment #115)
> > In 4.1.x and earlier kernels, the aumix is what actually turns on and off
> > the sound.
> 
> If I remember well, aumix only touches the OSS emulation layer. You should
> use alsamixer to touch the real hardware controls of the sound card.
> Pressing F6 you can choose the hardware instead of pulseaudio layer.

aumix is a command-line program, and a very important one for using the Line In channel of the sound card, as is required with this TV card.  If the kernel breaks aumix, that would be a problem that needs to be fixed.

aumix is also critical in recording from the TV card.  Here is the script for that:
aumix -i 60
mencoder -tv driver=v4l2:width=640:height=480:adevice=/dev/dsp:norm=ntsc:input=0 tv:// -o tv.avi -oac mp3lame -ovc lavc -lavcopts acodec=libmp3lame:abitrate=64:vbitrate=1800:vhq -lameopts vbr=3:br=64 -srate 44100 -vf crop=608:456:14:10 -noskip > /dev/null 2>&1
aumix -i 0

(input=0 for recording from coax, input=1 for recording from RCA video input)


For that, which *is* still working on 4.4.11, aumix is doing what it's supposed to and enabling recording from that channel.  The mencoder command alone does not capture the sound.

So, the association between aumix -l and the Line In channel for playback seems to be the only issue here.  Also keep in mind that Kmix seems to work the same as aumix in the backend.

Another problem I've mentioned multiple times before, but the i586 server kernel is absolutely incapable of reading /var/lib/rpm/Packages.  This is a long-standing problem, but it has gotten significantly worse with 4.4.x to the point of unusability, now reproduced on multiple different computers.  On my workstation at home, all I was trying to do was urpmi readline-devel (which would have installed libreadline-devel and libncursesw-devel) and it sounded like it was trying to eat my hard drive for lunch, got stuck in the Preparing stage, and was reading the Packages file so slowly, it looked like it would take weeks to complete.

I don't think I'll ever be able to validate a 4.4.x kernel update if this issue isn't solved on the i586 server kernel.

I also think that for Mageia 6 we should just drop the i586 server kernel variant and say you should move to x86_64 if you have 4GB+ of RAM.

This update is now splitted up as far as it can go:

Stuff to push first:
# kernel/radeon firmware (needs to go out before)
https://bugs.mageia.org/show_bug.cgi?id=18529

# btrfs-progs (needs to go out before)
https://bugs.mageia.org/show_bug.cgi?id=18531

# xtables-addons (needs to go out before so I can rebuild kmods for this update)
https://bugs.mageia.org/show_bug.cgi?id=18523

# broadcom-wl (needs to go out before)
https://bugs.mageia.org/show_bug.cgi?id=18528

# fglrx (needs to go out before)
https://bugs.mageia.org/show_bug.cgi?id=18525

# nvidia304 (needs to go out before)
https://bugs.mageia.org/show_bug.cgi?id=18526

# nvidia340 (needs to go out before)
https://bugs.mageia.org/show_bug.cgi?id=18527

# nvidia-current (needs to go out before)
https://bugs.mageia.org/show_bug.cgi?id=17604


Stuff that can go out after:
# aufs-tools
https://bugs.mageia.org/show_bug.cgi?id=18530

# xfs filesystem tools
https://bugs.mageia.org/show_bug.cgi?id=18532

# iproute2
https://bugs.mageia.org/show_bug.cgi?id=18533

Blocks: 17990, 18374, 18375 => (none)
Depends on: 18236, 18372 => 18523, 18525, 18526, 18527, 18528, 18529, 18531
Assignee: tmb => qa-bugs
Summary: Update request: kernel 4.4.9 => Update request: kernel 4.4.11
Whiteboard: feedback => (none)

So that leaves for this update:

SRPMS:
dracut-038-21.mga5.src.rpm

kernel-4.4.11-1.mga5.src.rpm
kernel-userspace-headers-4.4.11-1.mga5.src.rpm

kmod-vboxadditions-5.0.16-5.mga5.src.rpm
kmod-virtualbox-5.0.16-5.mga5.src.rpm

(+ kmod-xtables-addons-* after #18523 is pushed)


i586:
dracut-038-21.mga5.i586.rpm

cpupower-4.4.11-1.mga5.i586.rpm
cpupower-devel-4.4.11-1.mga5.i586.rpm
kernel-desktop-4.4.11-1.mga5-1-1.mga5.i586.rpm
kernel-desktop586-4.4.11-1.mga5-1-1.mga5.i586.rpm
kernel-desktop586-devel-4.4.11-1.mga5-1-1.mga5.i586.rpm
kernel-desktop586-devel-latest-4.4.11-1.mga5.i586.rpm
kernel-desktop586-latest-4.4.11-1.mga5.i586.rpm
kernel-desktop-devel-4.4.11-1.mga5-1-1.mga5.i586.rpm
kernel-desktop-devel-latest-4.4.11-1.mga5.i586.rpm
kernel-desktop-latest-4.4.11-1.mga5.i586.rpm
kernel-doc-4.4.11-1.mga5.noarch.rpm
kernel-server-4.4.11-1.mga5-1-1.mga5.i586.rpm
kernel-server-devel-4.4.11-1.mga5-1-1.mga5.i586.rpm
kernel-server-devel-latest-4.4.11-1.mga5.i586.rpm
kernel-server-latest-4.4.11-1.mga5.i586.rpm
kernel-source-4.4.11-1.mga5-1-1.mga5.noarch.rpm
kernel-source-latest-4.4.11-1.mga5.noarch.rpm
kernel-userspace-headers-4.4.11-1.mga5.i586.rpm
perf-4.4.11-1.mga5.i586.rpm

vboxadditions-kernel-4.4.11-desktop-1.mga5-5.0.16-5.mga5.i586.rpm
vboxadditions-kernel-4.4.11-desktop586-1.mga5-5.0.16-5.mga5.i586.rpm
vboxadditions-kernel-4.4.11-server-1.mga5-5.0.16-5.mga5.i586.rpm
vboxadditions-kernel-desktop586-latest-5.0.16-5.mga5.i586.rpm
vboxadditions-kernel-desktop-latest-5.0.16-5.mga5.i586.rpm
vboxadditions-kernel-server-latest-5.0.16-5.mga5.i586.rpm

virtualbox-kernel-4.4.11-desktop-1.mga5-5.0.16-5.mga5.i586.rpm
virtualbox-kernel-4.4.11-desktop586-1.mga5-5.0.16-5.mga5.i586.rpm
virtualbox-kernel-4.4.11-server-1.mga5-5.0.16-5.mga5.i586.rpm
virtualbox-kernel-desktop586-latest-5.0.16-5.mga5.i586.rpm
virtualbox-kernel-desktop-latest-5.0.16-5.mga5.i586.rpm
virtualbox-kernel-server-latest-5.0.16-5.mga5.i586.rpm


x86_64:
dracut-038-21.mga5.x86_64.rpm

cpupower-4.4.11-1.mga5.x86_64.rpm
cpupower-devel-4.4.11-1.mga5.x86_64.rpm
kernel-desktop-4.4.11-1.mga5-1-1.mga5.x86_64.rpm
kernel-desktop-devel-4.4.11-1.mga5-1-1.mga5.x86_64.rpm
kernel-desktop-devel-latest-4.4.11-1.mga5.x86_64.rpm
kernel-desktop-latest-4.4.11-1.mga5.x86_64.rpm
kernel-doc-4.4.11-1.mga5.noarch.rpm
kernel-server-4.4.11-1.mga5-1-1.mga5.x86_64.rpm
kernel-server-devel-4.4.11-1.mga5-1-1.mga5.x86_64.rpm
kernel-server-devel-latest-4.4.11-1.mga5.x86_64.rpm
kernel-server-latest-4.4.11-1.mga5.x86_64.rpm
kernel-source-4.4.11-1.mga5-1-1.mga5.noarch.rpm
kernel-source-latest-4.4.11-1.mga5.noarch.rpm
kernel-userspace-headers-4.4.11-1.mga5.x86_64.rpm
perf-4.4.11-1.mga5.x86_64.rpm

vboxadditions-kernel-4.4.11-desktop-1.mga5-5.0.16-5.mga5.x86_64.rpm
vboxadditions-kernel-4.4.11-server-1.mga5-5.0.16-5.mga5.x86_64.rpm
vboxadditions-kernel-desktop-latest-5.0.16-5.mga5.x86_64.rpm
vboxadditions-kernel-server-latest-5.0.16-5.mga5.x86_64.rpm

virtualbox-kernel-4.4.11-desktop-1.mga5-5.0.16-5.mga5.x86_64.rpm
virtualbox-kernel-4.4.11-server-1.mga5-5.0.16-5.mga5.x86_64.rpm
virtualbox-kernel-desktop-latest-5.0.16-5.mga5.x86_64.rpm
virtualbox-kernel-server-latest-5.0.16-5.mga5.x86_64.rpm

(In reply to David Walser from comment #118)
> Another problem I've mentioned multiple times before, but the i586 server
> kernel is absolutely incapable of reading /var/lib/rpm/Packages.  This is a
> long-standing problem, but it has gotten significantly worse with 4.4.x to
> the point of unusability, now reproduced on multiple different computers. 
> On my workstation at home, all I was trying to do was urpmi readline-devel
> (which would have installed libreadline-devel and libncursesw-devel) and it
> sounded like it was trying to eat my hard drive for lunch, got stuck in the
> Preparing stage, and was reading the Packages file so slowly, it looked like
> it would take weeks to complete.
> 
> I don't think I'll ever be able to validate a 4.4.x kernel update if this
> issue isn't solved on the i586 server kernel.
> 
> I also think that for Mageia 6 we should just drop the i586 server kernel
> variant and say you should move to x86_64 if you have 4GB+ of RAM.


Nope. this is something specific to your hw or setup...

I spent a lot of time last nite rebuilding a server and installed 32bit Mageia 5 on it and kept hammering on it with all kernels, both desktop and server kernel and both 4.1 and 4.4 series.

Yes, server kernel is a little slower than desktop because of HZ_100 and the overhead of doing full PAE on 32bit, but it works and is provided for those that for some reason must run a 32bit install with lots of memory... it's a trade-off, but it's known.

On my testserver the 4.4 series kernel is faster than 4.1 series (and that's even with spinning rust, not ssds)

Can you open up a separate bugreport for the slowdown (as I assume it's an upstream one and we can try to figure it out)

But this update really needs to get going for both better skylake support, 
and several CVEs ...

(In reply to Thomas Backlund from comment #121)
> Nope. this is something specific to your hw or setup...

No, it isn't.  I have seen this problem with the i586 server kernel on multiple different systems with completely different hardware.  Considering the desktop kernel doesn't exhibit this behavior, I'd surmise this must be a bad kernel CONFIG option that needs to be changed.

I understand that this kernel needs to get out, but it's still badly broken in multiple ways.  We can't have systems rendered completely unusable with the i586 server kernel because RPM doesn't work anymore, because it can't even remotely read /var/lib/rpm/Packages in a timely manner.

As for my Line In issue, hopefully I can bisect it at some point.  If you could make me kernel-linus 4.2 and 4.3 builds so I can narrow down the testing range, that would help.

(In reply to David Walser from comment #122)
> (In reply to Thomas Backlund from comment #121)
> > Nope. this is something specific to your hw or setup...
> 
> No, it isn't.  I have seen this problem with the i586 server kernel on
> multiple different systems with completely different hardware.  Considering
> the desktop kernel doesn't exhibit this behavior, I'd surmise this must be a
> bad kernel CONFIG option that needs to be changed.
> 
> I understand that this kernel needs to get out, but it's still badly broken
> in multiple ways.  We can't have systems rendered completely unusable with
> the i586 server kernel because RPM doesn't work anymore, because it can't
> even remotely read /var/lib/rpm/Packages in a timely manner.

For you that is... 
I havent Heard / seen any other complaints yet about this...

Since you say it's not hw specific, then there must be something specific on how you set up your systems.

Can you give a detailed description of how it's set up /installed so I can try to reproduce...

(In reply to Thomas Backlund from comment #123)
> For you that is... 

Nothing special about me.  I'm pretty sure there isn't a piece of my soul living in all of these machines.

> I havent Heard / seen any other complaints yet about this...

Maybe nobody's using the i586 server kernel.  I seem to be one of the only i586 fool diehards as it is.

> Since you say it's not hw specific, then there must be something specific on
> how you set up your systems.

These are different systems set up for different purposes.  They're not the same.  There is nothing exotic about how they are configured.

> Can you give a detailed description of how it's set up /installed so I can
> try to reproduce...

Is there any specific kind of information you would find helpful?

The only thing I can think of that may be uncommon is having /usr and /var on distinct partitions.  With the two big Dell servers, one has far as ext4 and the other has /var as XFS, and even both of them displayed the problem.  The ext4 one has since been reinstalled as x86_64 as is fine now.

Sadly, b43-openfwwf we supply still needs changes before it will work with this kernel, at least on i586 versions. (See Comment 63.)

Before getting the updates for kernel 4.4.11, I deleted the pcm5.fw file that I had manually added to /lib/firmware/b43/. I then got the updates, and rebooted. Broadcom 4318 wifi was not working. After restoring pcm5.fw manually, and rebooting, wifi again worked perfectly.

(In reply to David Walser from comment #124)
> (In reply to Thomas Backlund from comment #123)
> > For you that is... 
> 
> Nothing special about me.  I'm pretty sure there isn't a piece of my soul
> living in all of these machines.
> 
> > I havent Heard / seen any other complaints yet about this...
> 
> Maybe nobody's using the i586 server kernel.  I seem to be one of the only
> i586 fool diehards as it is.
> 
The Mageia installer has always installed the server kernel on my old motherboard with the Sempron 64 processor, even though it maxed out at 2GB of RAM. Since I saw no difference in using it, I just left it there. And when I passed the motherboard along to my brother, I left the server kernel on the new MGA5 install I put on his machine. So there is at least one more soul using it - but he wouldn't be doing anything like you describe. And I didn't notice any problems when I had it, but then I tend to stay away from such things myself. 

> > Since you say it's not hw specific, then there must be something specific on
> > how you set up your systems.
> 
> These are different systems set up for different purposes.  They're not the
> same.  There is nothing exotic about how they are configured.
> 
> > Can you give a detailed description of how it's set up /installed so I can
> > try to reproduce...
> 
> Is there any specific kind of information you would find helpful?
> 
> The only thing I can think of that may be uncommon is having /usr and /var
> on distinct partitions.  With the two big Dell servers, one has far as ext4
> and the other has /var as XFS, and even both of them displayed the problem. 
> The ext4 one has since been reinstalled as x86_64 as is fine now.

FWIW, all of my server kernel systems have /usr and /var in the / partition.

(In reply to David Walser from comment #124)
> (In reply to Thomas Backlund from comment #123)
> 
> > Can you give a detailed description of how it's set up /installed so I can
> > try to reproduce...
> 
> Is there any specific kind of information you would find helpful?
> 
> The only thing I can think of that may be uncommon is having /usr and /var
> on distinct partitions.  With the two big Dell servers, one has far as ext4
> and the other has /var as XFS, and even both of them displayed the problem. 
> The ext4 one has since been reinstalled as x86_64 as is fine now.

ok, /usr and /war on distinct partitions should not affect anything... unless you have some special mount options... ar the partitions on the same disk ?

I was thinking for example on possible xfs issue as I tested ext4, but that is apparently not the case either if you can reproduc it on ext4 too 

how old hw, are th ~same age? chipset ? intel based ? amd based ? amount of ram ?
what cpu ? core count, ht ?

Can you provide dmesg, journalctl -b, and /proc/meminfo from 4.4.11-1 desktop and server kernels from one of the affected systems so I maybe can spot what triggers different behaviour on them ?

Without dealing with VirtualBox:

On real hardware, M5, KDE, 64-bit

Package(s) under test:
kernel-desktop-latest, nvidia-current-kernel-desktop-latest

default install of:
kernel-desktop-latest nvidia-current-kernel-desktop-latest

[root@localhost wilcal]# uname -a
Linux localhost 4.1.15-desktop-2.mga5 #1 SMP Wed Jan 20 17:05:51 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi nvidia-current-kernel-desktop-latest
Package nvidia-current-kernel-desktop-latest-346.96-5.mga5.nonfree.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nvidia
        Kernel modules: nvidiafb, nouveau, nvidia_current
        
Boots to a working desktop. All applications work normally.

install from updates_testing:

kernel-desktop-latest nvidia-current-kernel-desktop-latest

reboot system

[root@localhost wilcal]# uname -a
Linux localhost 4.4.11-desktop-1.mga5 #1 SMP Thu May 19 10:26:26 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi nvidia-current-kernel-desktop-latest
Package nvidia-current-kernel-desktop-latest-352.79-10.mga5.nonfree.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nouveau
        Kernel modules: nvidiafb, nouveau

Boots to a working desktop but ends up booting with nouveau not nvidia.
I can't seem to get it to go to nvidia even though I try to load it with:

MCC -> Hardware -> Set up the graphical server

One thing that makes this difficult is the difficult in disk I/O and especially in reading Packages is that it typically isn't so bad after a fresh boot, but gets progressively worse the longer the system is running (a few days is enough to make it bad).

When I did the urpmi readline-devel on my workstation at home, I'm thinking that one had been up for a few days.  I'll have to try it again from a fresh boot when I get back home (mid-weekend) with both kernels.

On the remaining big server at work still running i586 (with the XFS /var), I did some collections from a fresh boot, but the I/O seems fine seconds after boot.

For example...desktop
rpm -qa
real    0m1.604s
user    0m0.227s
sys     0m0.075s
urpmi --auto readline-devel
real    0m3.014s
user    0m1.513s
sys     0m0.196s
server
rpm -qa
real    0m1.544s
user    0m0.190s
sys     0m0.070s
urpmi --auto readline-devel
real    0m2.877s
user    0m1.270s
sys     0m0.260s

I'll leave it booted into the server kernel for now and can take a sample again on Friday morning if you want.

I'll still post what I collected from this server, and perhaps I can repeat this on my workstation at home this weekend.

As far as the partitions, yes, on the same disk in all cases (all systems).  Servers are hardware RAID, workstations I've seen this problem on are just regular "spinning rust" hard disks.  No exotic mount options (defaults in fact).

As for ages, my old workstation at work (Dell OptiPlex 990, recently replaced) was bought I believe in December 2011, the big servers probably also December 2011, and my workstation at home was purchased almost exactly two years ago (PogoLinux Verona 931H).

Created attachment 7858 [details]
Dell server collection with desktop 4.4.11 kernel

rpm -qa
/proc/meminfo
lspcidrake
journalctl -b
dmidecode
dmesg

Created attachment 7859 [details]
Dell server collection with server 4.4.11 kernel

rpm -qa
/proc/meminfo
lspcidrake
journalctl -b
dmidecode
dmesg

(In reply to William Kenney from comment #128)
> Without dealing with VirtualBox:
> 
> On real hardware, M5, KDE, 64-bit
> 
> Package(s) under test:
> kernel-desktop-latest, nvidia-current-kernel-desktop-latest
> 
> default install of:
> kernel-desktop-latest nvidia-current-kernel-desktop-latest
> 
> [root@localhost wilcal]# uname -a
> Linux localhost 4.1.15-desktop-2.mga5 #1 SMP Wed Jan 20 17:05:51 UTC 2016
> x86_64 x86_64 x86_64 GNU/Linux
> [root@localhost wilcal]# urpmi kernel-desktop-latest
> Package kernel-desktop-latest-4.1.15-2.mga5.x86_64 is already installed
> [root@localhost wilcal]# urpmi nvidia-current-kernel-desktop-latest
> Package nvidia-current-kernel-desktop-latest-346.96-5.mga5.nonfree.x86_64 is
> already installed
> [root@localhost wilcal]# lspci -k
> 01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440]
> (rev a1)
>         Subsystem: Gigabyte Technology Co., Ltd Device 3518
>         Kernel driver in use: nvidia
>         Kernel modules: nvidiafb, nouveau, nvidia_current
>         
> Boots to a working desktop. All applications work normally.
> 
> install from updates_testing:
> 
> kernel-desktop-latest nvidia-current-kernel-desktop-latest
> 
> reboot system
> 
> [root@localhost wilcal]# uname -a
> Linux localhost 4.4.11-desktop-1.mga5 #1 SMP Thu May 19 10:26:26 UTC 2016
> x86_64 x86_64 x86_64 GNU/Linux
> [root@localhost wilcal]# urpmi kernel-desktop-latest
> Package kernel-desktop-latest-4.4.11-1.mga5.x86_64 is already installed
> [root@localhost wilcal]# urpmi nvidia-current-kernel-desktop-latest
> Package nvidia-current-kernel-desktop-latest-352.79-10.mga5.nonfree.x86_64
> is already installed
> [root@localhost wilcal]# lspci -k
> 01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440]
> (rev a1)
>         Subsystem: Gigabyte Technology Co., Ltd Device 3518
>         Kernel driver in use: nouveau
>         Kernel modules: nvidiafb, nouveau
> 
> Boots to a working desktop but ends up booting with nouveau not nvidia.
> I can't seem to get it to go to nvidia even though I try to load it with:
> 
> MCC -> Hardware -> Set up the graphical server


Not exactly sure what you are asking here.

All the pre-built nvidia module pkgs (nvidia-current-kernel-*.rpm) are dead.
There will be no more:
"tmb <tmb> 352.79-10.mga5:
+ Revision: 1017586
- switch from prebuilt to dkms drivers for GPL compliance"


With info given do not know why nouveau is used instead of nvidia.

Logout.
Switch to tty and run dm stop
rmmod nouveau and then modprobe nvidia-current 
If no errors
Change driver in xorg.conf to nvidia
run alternatives --config gl_conf  (needs to be /etc/nvidia-current/ld.so.con)
re-start the dm and log-in
You will be using the nvidia driver.

CC: (none) => cae

(In reply to Charles Edwards from comment #132)
> With info given do not know why nouveau is used instead of nvidia.
> 
> Logout.
> Switch to tty and run dm stop
> rmmod nouveau and then modprobe nvidia-current 
> If no errors
> Change driver in xorg.conf to nvidia
> run alternatives --config gl_conf  (needs to be
> /etc/nvidia-current/ld.so.con)
> re-start the dm and log-in
> You will be using the nvidia driver.

I'm not looking to use it just trying to get it to properly install.
So are you saying we no longer use an nvidia proprietary driver?

(In reply to William Kenney from comment #133)
> (In reply to Charles Edwards from comment #132)
> > With info given do not know why nouveau is used instead of nvidia.
> > 
> > Logout.
> > Switch to tty and run dm stop
> > rmmod nouveau and then modprobe nvidia-current 
> > If no errors
> > Change driver in xorg.conf to nvidia
> > run alternatives --config gl_conf  (needs to be
> > /etc/nvidia-current/ld.so.con)
> > re-start the dm and log-in
> > You will be using the nvidia driver.
> 
> I'm not looking to use it just trying to get it to properly install.
> So are you saying we no longer use an nvidia proprietary driver?

No.

The nvidia proprietary driver can still be used, I'm using it here and on my test machine.

What I said is that there will no more nvidia-current-kernel rpms that contained
the pre-compiled ko modules.
It was decided that including those was in violation of the GPL.

The modules will now only be built locally using dkms-nvidia-current 


The same applies to nvidia304 and nvidia340

(In reply to Charles Edwards from comment #134)
> The modules will now only be built locally using dkms-nvidia-current 
> 
> The same applies to nvidia304 and nvidia340

OK lets make this as simple as possible. We've probably got lots of
users who have the present nvidia driver setup. Along comes this whatever
and the user has no conception of how to

"only be built locally using dkms-nvidia-current"

he/she simply says go ahead and update my system. What's gonna happen?

When the user originally answered Yes to the prompt about using proprietary drivers, Mageia will have installed not only the binary module but also dkms and everything that dkms needs to build the kernel module. So other than perhaps a short delay while dkms builds the module, the change should not be visible to the user.

CC: (none) => jkerr82508

OK, it seems I'm having problems getting:

x11-driver-video-nvidia-current-352.79-3.mga5.nonfree.x86_64.rpm

It is not installing from either the MCC or from a urpmi command line.

It's in core updates_testing. Even if I copy it to local Downloads and attempt
an install from there it's no go. It appears to go through the install process
but the system still falls back to nouveau.

It does not seem possible to test this update separately from the nvidia-current update. kernel-desktop-latest requires the latest dkms-nvidia-current-352.79.3. If one proceeds then dkms-nvidia-current is removed. Presumably this is why the system reverted to nouveau - see comment#128.

I can not duplicate the issue you are having.

Unless you can find something in demsg or other logs solving your problem may not be possible.

I did test on my Mga5 system.
Activated core-update-testing and nonfree-update testing

Installed:
dkms-nvidia-current-352.79-3.mga5.nonfree
x11-driver-video-nvidia-current-352.79-3.mga5.nonfree
nvidia-current-doc-html-352.79-3.mga5.nonfree

Module built without problem and was easily able to use the new nvidia driver with kernel-server-4.1.15-2.mga5-1-1.mga5

Perhaps the issue is specific to 4.4.11.
I installed:

[root@TestBed ~]# urpmi kernel-server-4.4.11-1.mga5-1-1.mga5 kernel-server-devel-4.4.11-1.mga5-1-1.mga5 kernel-source-4.4.11-1.mga5-1-1.mga5
To satisfy dependencies, the following packages are going to be installed:
  Package                        Version      Release       Arch    
(medium "Core Updates Testing (distrib5)")
  btrfs-progs                    4.4.1        1.mga5        x86_64  
  dracut                         038          21.mga5       x86_64  
  kernel-firmware                20160409     1.mga5        noarch  
  kernel-server-4.4.11-1.mga5    1            1.mga5        x86_64  
  kernel-server-devel-4.4.11-1.> 1            1.mga5        x86_64  
  kernel-source-4.4.11-1.mga5    1            1.mga5        noarch  
  lib64btrfs0                    4.4.1        1.mga5        x86_64  
(medium "Nonfree Updates Testing (distrib15)")
  kernel-firmware-nonfree        20160516     1.mga5.nonfr> noarch  
  radeon-firmware                20160516     1.mga5.nonfr> noarch  
630MB of additional disk space will be used.
165MB of packages will be retrieved.
Proceed with the installation of the 9 packages? (Y/n) y


On 1st boot of the 4.4.11 kernel the nvidia module successfully built 
and am able to use it without problem

uname -r && glxinfo
4.4.11-server-1.mga5
name of display: :0.0
display: :0  screen: 0
direct rendering: Yes
server glx vendor string: NVIDIA Corporation
server glx version string: 1.4

The solution to the problem that I described in comment#138 is to test the nvidia-current update bug #17604 before attempting to test this update.

Testing on mga5-64, after updating nvidia-current:

Packages installed from core-updates-testing:

kernel-userspace-headers-4.4.11-1
cpupower-4.4.11-1.mga5.x86_64
kernel-desktop-latest-4.4.11-1
kernel-desktop-devel-latest-4.4.11-1
kernel-desktop-4.4.11-1.mga5-1-1
dracut-038-21.mga5.x86_64
kernel-desktop-devel-4.4.11-1

Packages installed cleanly and the system re-booted normally.

OK for mga5-64 on this system:

CPU: Quad core AMD Phenom 9500 (-MCP-)
Mobo: ECS model: GeForce7050M-M v: 1.0

(In reply to James Kerr from comment #140)

> The solution to the problem that I described in comment#138 is to test the
> nvidia-current update bug #17604 before attempting to test this update.

Tomorrow ( Sat California time ) I will again start with a blank drive
and do an M5 64-bit boot.iso install to a blank drive bringing that install
up to date as we are now. Per your request I will first install the nvidia
driver, reboot, and the system should then be at desktop using the new
nvidia driver. After that if it is successful I will run the kernel update,
reboot, and if that is successful I'll install VirtualBox to see if that
installs, and reboots back to a working desktop.

Many thanks to all for what has become a long and arduous task.
Here's my test system:

Intel Core i7-2600K Sandy Bridge 3.4GHz
GIGABYTE GA-Z68X-UD3-B3 LGA 1155 MoBo
GIGABYTE GV-N440D3-1GI Nvidia GeForce GT 440 (Fermi) 1GB
RTL8111/8168B PCI Express 1Gbit Ethernet
DRAM 16GB (4 x 4GB)
Mageia 5 64-bit

(In reply to William Kenney from comment #141)
> (In reply to James Kerr from comment #140)
> 
> > The solution to the problem that I described in comment#138 is to test the
> > nvidia-current update bug #17604 before attempting to test this update.
> 
> Tomorrow ( Sat California time ) I will again start with a blank drive
> and do an M5 64-bit boot.iso install to a blank drive bringing that install
> up to date as we are now. Per your request I will first install the nvidia
> driver, reboot, and the system should then be at desktop using the new
> nvidia driver. After that if it is successful I will run the kernel update,
> reboot, and if that is successful I'll install VirtualBox to see if that
> installs, and reboots back to a working desktop.

That is what tmb is suggesting in comment#119

Earlier, I had failed to grasp the import of that comment, or had missed it entirely. Finding what is relevant in this bug report is not easy.  :)

On real hardware, M5, KDE, 64-bit

default install of:

- kernel-userspace-headers-4.4.11-1
- cpupower-4.4.11-1.mga5.x86_64
- kernel-desktop-latest-4.4.11-1
- kernel-desktop-devel-latest-4.4.11-1
- kernel-desktop-4.4.11-1.mga5-1-1
- dracut-038-21.mga5.x86_64
- kernel-desktop-devel-4.4.11-1

[wilcal@localhost ~]$ uname -a
Linux localhost 4.1.15-desktop-2.mga5 #1 SMP Wed Jan 20 17:05:51 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-userspace-headers
Package kernel-userspace-headers-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi cpupower
Package cpupower-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-devel-latest
Package kernel-desktop-devel-latest-4.1.15-2.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-4.1.15-2.mga5
Package kernel-desktop-4.1.15-2.mga5-1-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi dracut
Package dracut-038-19.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-devel
Package kernel-desktop-devel-4.1.15-2.mga5-1-1.mga5.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nvidia
        Kernel modules: nvidiafb, nouveau, nvidia_current
        
Boots to a working desktop. Screen sizes are correct. Applications work.

Install:
https://bugs.mageia.org/show_bug.cgi?id=17604
first

install:

- kernel-userspace-headers-4.4.11-1
- cpupower-4.4.11-1.mga5.x86_64
- kernel-desktop-latest-4.4.11-1
- kernel-desktop-devel-latest-4.4.11-1
- kernel-desktop-4.4.11-1.mga5-1-1
- dracut-038-21.mga5.x86_64
- kernel-desktop-devel-4.4.11-1

from updates_testing

[root@localhost wilcal]# uname -a
Linux localhost 4.4.11-desktop-1.mga5 #1 SMP Thu May 19 10:26:26 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-userspace-headers
Package kernel-userspace-headers-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi cpupower
Package cpupower-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-devel-latest
Package kernel-desktop-devel-latest-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-4.1.15-2.mga5
Package kernel-desktop-4.1.15-2.mga5-1-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi dracut
Package dracut-038-21.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-devel-4.1.15-2.mga5
Package kernel-desktop-devel-4.1.15-2.mga5-1-1.mga5.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nvidia
        Kernel modules: nvidiafb, nouveau, nvidia_current
        
Boots to a working desktop. Screen sizes are correct. Applications work.

Test platform:
Intel, P4 530J 3.0 GHz, 800MHz FSB, 1MB L2, LGA 775
GigaByte  GA-81915G Pro F4  i915G  LGA 775  MoBo
 Marvel Yukon 88E8001 Gigabit LAN
 Intel High Def Audio, Azalia (C-Media 9880) (snd-hda-intel)
 Intel Graphics Media Accelerator 900 (Intel 82915G)
Kingston 4GB (2 x 2GB) DDR400 PC-3200
1TB Seagate HD

On real hardware, M5, KDE, 64-bit

Installed after:

https://bugs.mageia.org/show_bug.cgi?id=18031

https://bugs.mageia.org/show_bug.cgi?id=17604

default install of:
- dkms-virtualbox-5.0.16-1.mga5.noarch
- vboxadditions-kernel-4.1.15-desktop-2.mga5-5.0.16-3.mga5.x86_64
- vboxadditions-kernel-desktop-latest-5.0.16-3.mga5.x86_64
- virtualbox-5.0.16-1.mga5.x86_64
- virtualbox-guest-additions-5.0.16-1.mga5.x86_64
- virtualbox-kernel-4.1.15-desktop-2.mga5-5.0.16-3.mga5.x86_64
- virtualbox-kernel-desktop-latest-5.0.16-3.mga5.x86_64
- x11-driver-video-vboxvideo-5.0.16-1.mga5.x86_64

virtualbox vboxadditions-kernel-desktop-latest dkms-virtualbox
virtualbox-guest-additions virtualbox-kernel-desktop-latest x11-driver-video-vboxvideo

[wilcal@localhost ~]$ uname -a
Linux localhost 4.4.11-desktop-1.mga5 #1 SMP Thu May 19 10:26:26 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi nvidia-current-kernel-desktop-latest
Package nvidia-current-kernel-desktop-latest-352.79-10.mga5.nonfree.x86_64 is already installed
[root@localhost wilcal]# urpmi dkms-virtualbox
Package dkms-virtualbox-5.0.16-1.mga5.noarch is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-4.1.15-desktop-2.mga5
Package vboxadditions-kernel-4.1.15-desktop-2.mga5-5.0.16-3.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi vboxadditions-kernel-desktop-latest
Package vboxadditions-kernel-desktop-latest-5.0.16-3.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox
Package virtualbox-5.0.16-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox-guest-additions
Package virtualbox-guest-additions-5.0.16-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox-kernel-4.1.15-desktop-2.mga5
Package virtualbox-kernel-4.1.15-desktop-2.mga5-5.0.16-3.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi virtualbox-kernel-desktop-latest
Package virtualbox-kernel-desktop-latest-5.0.16-3.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi x11-driver-video-vboxvideo
Package x11-driver-video-vboxvideo-5.0.16-1.mga5.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nvidia
        Kernel modules: nvidiafb, nouveau, nvidia_current
        
Virtualbox installs properly.
Mageia-5-LiveDVD-KDE4-x86_64-DVD.iso opens and displays properly as a Vbox client
Mageia-5-x86_64-DVD.iso installs, displays properly, updates and then reboots properly

Testing M5 x64 real h/w with AMD/ATI/Radeon video.
Prior to the 4.4.11 kernel update, I had pre-updated the following (Comment 119):
 kernel-firmware-20160409-1.mga5
 radeon-firmware-20160516-1.mga5.nonfree
 btrfs-progs-4.4.1-1.mga5
 lib64btrfs0-4.4.1-1.mga5
 dkms-fglrx-15.302-4.mga5.nonfree
 fglrx-control-center-15.302-4.mga5.nonfree
 fglrx-kernel-desktop-latest-15.302-10.mga5.nonfree
 x11-driver-video-fglrx-15.302-4.mga5.nonfree

 cpupower-4.4.11-1.mga5
$ uname -r
 4.4.11-desktop-1.mga5
No sign of the 4.4.11 fglrx equivalent kernel (only earlier ones). Not sure whether this matters. BTAIM The system runs OK so far.

CC: (none) => lewyssmith

On real hardware, M5, KDE, 64-bit

Vbox installed before

https://bugs.mageia.org/show_bug.cgi?id=18031

https://bugs.mageia.org/show_bug.cgi?id=17604

updates from updates_testing.

- kernel-userspace-headers-4.4.11-1
- cpupower-4.4.11-1.mga5.x86_64
- kernel-desktop-latest-4.4.11-1
- kernel-desktop-devel-latest-4.4.11-1
- kernel-desktop-4.4.11-1.mga5-1-1
- dracut-038-21.mga5.x86_64
- kernel-desktop-devel-4.4.11-1

[root@localhost wilcal]# uname -a
Linux localhost 4.4.11-desktop-1.mga5 #1 SMP Thu May 19 10:26:26 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost wilcal]# urpmi kernel-userspace-headers
Package kernel-userspace-headers-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi cpupower
Package cpupower-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-latest
Package kernel-desktop-latest-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-devel-latest
Package kernel-desktop-devel-latest-4.4.11-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-4.4.11-1.mga5
Package kernel-desktop-4.4.11-1.mga5-1-1.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi dracut
Package dracut-038-21.mga5.x86_64 is already installed
[root@localhost wilcal]# urpmi kernel-desktop-devel-4.4.11-1.mga5
Package kernel-desktop-devel-4.4.11-1.mga5-1-1.mga5.x86_64 is already installed
[root@localhost wilcal]# lspci -k
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 440] (rev a1)
        Subsystem: Gigabyte Technology Co., Ltd Device 3518
        Kernel driver in use: nvidia
        Kernel modules: nvidiafb, nouveau, nvidia_current
        
Mageia-5-LiveDVD-KDE4-x86_64-DVD.iso opens and displays properly as a Vbox client
Mageia-5-LiveDVD-GNOME-x86_64-DVD.iso opens and displays properly as a Vbox client
Mageia-5-x86_64-DVD.iso KDE installs, updates and runs as a Vbox client. Screen sizes are correct

(In reply to Thomas Andrews from comment #125)
> Sadly, b43-openfwwf we supply still needs changes before it will work with
> this kernel, at least on i586 versions. (See Comment 63.)
> 
> Before getting the updates for kernel 4.4.11, I deleted the pcm5.fw file
> that I had manually added to /lib/firmware/b43/. I then got the updates, and
> rebooted. Broadcom 4318 wifi was not working. After restoring pcm5.fw
> manually, and rebooting, wifi again worked perfectly.

Fixed in b43-openfwwf-5.2-6.3.mga5, which does now replace b43-openfwwf-5.2-6.2.mga5 in this big update...

(In reply to José Jorge from comment #148)
> (In reply to Thomas Andrews from comment #125)
> > Sadly, b43-openfwwf we supply still needs changes before it will work with
> > this kernel, at least on i586 versions. (See Comment 63.)
> > 
> > Before getting the updates for kernel 4.4.11, I deleted the pcm5.fw file
> > that I had manually added to /lib/firmware/b43/. I then got the updates, and
> > rebooted. Broadcom 4318 wifi was not working. After restoring pcm5.fw
> > manually, and rebooting, wifi again worked perfectly.
> 
> Fixed in b43-openfwwf-5.2-6.3.mga5, which does now replace
> b43-openfwwf-5.2-6.2.mga5 in this big update...

Close, but not quite. One more change is needed for best operation: The firmware files need to be in /lib/firmware/b43 instead of /lib/firmware/b43-open or there is an exceedingly long delay before the wireless connection comes up when booting. See Bug 17990.

Testing on mga5-32

Packages installed from testing:

- cpupower-4.4.11-1.mga5.i586
- dracut-038-21.mga5.i586
- kernel-desktop-4.4.11-1.mga5-1-1.mga5.i586
- kernel-desktop-devel-4.4.11-1.mga5-1-1.mga5.i586
- kernel-desktop-devel-latest-4.4.11-1.mga5.i586
- kernel-desktop-latest-4.4.11-1.mga5.i586
- kernel-firmware-20160409-1.mga5.noarch
- kernel-firmware-nonfree-20160516-1.mga5.nonfree.noarch
- kernel-userspace-headers-4.4.11-1.mga5.i586
- ldetect-lst-0.1.346.4-1.mga5.i586

Packages installed cleanly. System rebooted normally. No regressions noted.

OK for mga5-32 on this system:

Mobo: ECS model: GeForce7050M-M v: 1.0
CPU:  Quad core AMD Phenom 9500
Card: NVIDIA GF108 [GeForce GT 630]
      Display Server: X.Org 1.16.4 drivers: v4l,nouveau

(In reply to Thomas Andrews from comment #149)
> (In reply to José Jorge from comment #148)
> > (In reply to Thomas Andrews from comment #125)
> > > Sadly, b43-openfwwf we supply still needs changes before it will work with
> > > this kernel, at least on i586 versions. (See Comment 63.)
> > > 
> > > Before getting the updates for kernel 4.4.11, I deleted the pcm5.fw file
> > > that I had manually added to /lib/firmware/b43/. I then got the updates, and
> > > rebooted. Broadcom 4318 wifi was not working. After restoring pcm5.fw
> > > manually, and rebooting, wifi again worked perfectly.
> > 
> > Fixed in b43-openfwwf-5.2-6.3.mga5, which does now replace
> > b43-openfwwf-5.2-6.2.mga5 in this big update...
> 
> Close, but not quite. One more change is needed for best operation: The
> firmware files need to be in /lib/firmware/b43 instead of
> /lib/firmware/b43-open or there is an exceedingly long delay before the
> wireless connection comes up when booting. See Bug 17990.

It looks like changing the path for b43-openfwwf-5.2-6.3.mga5 is not a good thing to do, because of potential conflicts with proprietary b43 firmware that might have been or might be installed. The delay in making the connection is because of upstream kernel changes, and must be dealt with there, not by us.

Therefore, since BCM4318 wifi DOES work with this latest open firmware, I will no longer consider it a problem with this update. 

Until upstream fixes the problem, I would suggest that users who find the delay intolerable should be advised to install the proprietary firmware. I have done that, and the kernel seems to work fine with it.

Adding feedback marker for non re-distributable file in b43

Whiteboard: (none) => feedback

Created attachment 7893 [details]
Dell server collection with server 4.4.11 kernel (slow)

When I last checked early Saturday morning (6:49 am), rpm -qa still ran fine.

Now, the system's update is just over 8 days and it's dog slow again.

rpm -qa
real    4m39.431s
user    0m0.000s
sys     0m0.800s

I've attached the current meminfo, dmesg, and journal.

(In reply to claire robinson from comment #152)
> Adding feedback marker for non re-distributable file in b43

Huh? The file pcm5.fw is redistributable, as it comes from the upstream open-source project!

It was discussed in the QA meeting tonight & tmb thought it was non-redistributable. I'll let him check/comment.

(In reply to José Jorge from comment #154)
> (In reply to claire robinson from comment #152)
> > Adding feedback marker for non re-distributable file in b43
> 
> Huh? The file pcm5.fw is redistributable, as it comes from the upstream
> open-source project!


NO! It isn't!

It's a binary extracted from broadcom drivers and rolled up in a binary-only tarball on that site.

There is _no_ official license from Broadcom (the only ones that legally can approve distributing it) for that file, and no source available. 

So as such we can not legally re-distribute it in Mageia. 
It sucks, but it is a fact we have to live with.

(In reply to David Walser from comment #153)
> Created attachment 7893 [details]
> Dell server collection with server 4.4.11 kernel (slow)
> 
> When I last checked early Saturday morning (6:49 am), rpm -qa still ran fine.
> 
> Now, the system's update is just over 8 days and it's dog slow again.
> 
> rpm -qa
> real    4m39.431s
> user    0m0.000s
> sys     0m0.800s
> 
> I've attached the current meminfo, dmesg, and journal.

urpmi readline-devel
real    3m28.030s
user    0m1.920s
sys     0m0.050s

(In reply to Thomas Backlund from comment #156)
> (In reply to José Jorge from comment #154)
> > (In reply to claire robinson from comment #152)
> > > Adding feedback marker for non re-distributable file in b43
> > 
> > Huh? The file pcm5.fw is redistributable, as it comes from the upstream
> > open-source project!
> 
> 
> NO! It isn't!
> 
> It's a binary extracted from broadcom drivers and rolled up in a binary-only
> tarball on that site.
> 
> There is _no_ official license from Broadcom (the only ones that legally can
> approve distributing it) for that file, and no source available. 
> 
> So as such we can not legally re-distribute it in Mageia. 
> It sucks, but it is a fact we have to live with.

The more I look at this, the more I have to agree with Thomas. When I first looked for newer open source firmware, I failed to look at the dates on the openfwwf site. That was dumb.

The original openfwwf file is dated Jan. 2009. There was quite a bit of development for six months, then nothing until 2014, and the files posted at that time are of a different form, as if they were some kind of work in progress that was never finished. The site's "official" "latest" version is 5.2, dated July 2009. Essentially, there's been zero development since then. Jose, you had nothing but silence from the developers.

It's time to face facts. This open firmware no longer works. Unless one of us wants to pick up the development of it, those of us who have been using b43-openfwwf need to go back to the old way of getting needed firmware - using b43-fwcutter.

(In reply to Thomas Andrews from comment #158)
> It's time to face facts. This open firmware no longer works. Unless one of
> us wants to pick up the development of it, those of us who have been using
> b43-openfwwf need to go back to the old way of getting needed firmware -
> using b43-fwcutter.

This is Ok for me. It seems confirmed by this :

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=513974

On x64
I have only just noticed 'no sound'; but this has been complained about already.

The b43-openfwwf issue is of course not for just Jose and me to decide on our own, but I see no alternative to dropping it with this kernel update.

Assuming that is done, is there anything we can do for the users of the package to make the transition to proprietary firmware, or must we just leave them hanging? The procedure I followed, and it worked to perfection, was to boot into kernel 4.1.15, using the open firmware to establish a wifi connection, (the oly Internet connection) then in one MCC operation remove b43-openfwwf and install firmware-b43-installer. Upon rebooting, the computer was using the proprietary firmware, and the transition was relatively seamless. 

For my own part, I know what's coming, so I can prepare the computers in my charge that use the b43 driver. But what about the unsuspecting user whose wifi suddenly stops functioning simply because he updated the kernel? Yes, it's old hardware and most likely there are only a few of us, but still...

The trouble is, each individual can elect to perform my procedure on his own and not violate any license that I know anything about, but I don't know if we can do something as a distro to do it for him, without his express permission, so he doesn't get blindsided.

As xtables 2.10 got pushed for the 4.1.15 kernel, here are the kmods for 4.4.11 kernel:

SRPM:
kmod-xtables-addons-2.10-6.mga5.src.rpm

i586:
xtables-addons-kernel-4.4.11-desktop-1.mga5-2.10-6.mga5.i586.rpm
xtables-addons-kernel-4.4.11-desktop586-1.mga5-2.10-6.mga5.i586.rpm
xtables-addons-kernel-4.4.11-server-1.mga5-2.10-6.mga5.i586.rpm
xtables-addons-kernel-desktop586-latest-2.10-6.mga5.i586.rpm
xtables-addons-kernel-desktop-latest-2.10-6.mga5.i586.rpm
xtables-addons-kernel-server-latest-2.10-6.mga5.i586.rpm


x86_64:
xtables-addons-kernel-4.4.11-desktop-1.mga5-2.10-6.mga5.x86_64.rpm
xtables-addons-kernel-4.4.11-server-1.mga5-2.10-6.mga5.x86_64.rpm
xtables-addons-kernel-desktop-latest-2.10-6.mga5.x86_64.rpm
xtables-addons-kernel-server-latest-2.10-6.mga5.x86_64.rpm

Closing this one down and start over to get clean relevant info for the update

Status: NEW => RESOLVED
Resolution: (none) => OLD

(In reply to David Walser from comment #157)
> (In reply to David Walser from comment #153)
> > Created attachment 7893 [details]
> > Dell server collection with server 4.4.11 kernel (slow)
> > 
> > When I last checked early Saturday morning (6:49 am), rpm -qa still ran fine.
> > 
> > Now, the system's update is just over 8 days and it's dog slow again.
> > 
> > rpm -qa
> > real    4m39.431s
> > user    0m0.000s
> > sys     0m0.800s
> > 
> > I've attached the current meminfo, dmesg, and journal.
> 
> urpmi readline-devel
> real    3m28.030s
> user    0m1.920s
> sys     0m0.050s

Hello David

I'm experiencing exactly the same slowness as you described  when running 32 bit kernel server. I have 2 PC desktop (rather powerful ones) :

 - one DELL T3400 that worked fine for years with all the mageia versions, until I recently noticed some *huge* slow down in everything. Same as you even running urpmi to update takes between 10-20 minutes just to get the list of packages that need to be updated. Running strace on the rpm process shows it's reading /var/lib/Packages at an extremely slow speed.
Sometimes, I even had to kill rpm because it took too long, then rpm database was corrupt, so I had to run rebuilddb (which took much longer than expected too). This PC runs kernel 4.4.16

- So I thought this aging PC was maybe slowly dying. And I just installed a more recent HP Z240 (32 Go RAM, 2 disks, not using HW raid, but using mdadm) using mga5. I updated to kernel 4.4.26, and same problem : disk access are extremely slow. It took 5 minutes to copy /var/lib/rpm when this dir barely contains 100-200 MB.

Before using 4.4.16 on the Dell PC, I used kernel-server-4.1.15-2.mga5-1-1.mga5. That was a long time ago, but I think this kernel worked better IO wise.

Can you check kernel 4.1.15 if it's better ? Or did you find some kernel settings/parameters that fixed the IO performance in your case ?

CC: (none) => npomarede

This problem has existed for a long time, but yes, it became worse with 4.4, to the point where I had to stop using the server kernel just to get by temporarily, then had to switch to x86_64 to get back to normal.

In the meantime, I found some similar bugs with possible solution ; it seems it's another PAE problem, the more RAM you have, the slower the PC will get in 32 bit mode.
I can see this in my case, the 8 GB DELL gets slow, but the 32 GB HP is even slower (despite having a better cpu).

some people told to flush vm cache in /proc/sys/vm/drop_caches :
https://bugs.launchpad.net/ubuntu/+source/linux-meta-lts-trusty/+bug/1333294
But that's not a real fix.

It seems a real solution is here :
http://flaterco.com/kb/PAE_slowdown.html

-> kernel needs to be build with CONFIG_VMSPLIT_2G=y instead of CONFIG_VMSPLIT_3G=y  . This ways kernel will have 2 GB instead of 1 GB for its internal data and the impact in performance is supposed to be minimal.

I could verify that if I boot the HP with "mem=4G" then it runs as fast as it should, so the problem is really related to the amount of memory and the "VM split" build config.

Could we have an updated mga5 32 bit kernel with CONFIG_VMSPLIT_2G=y to test this ?

Nice findings. Please create a different bug report for that, because this one here is closed (or add these information to an existing one about that slowness if there's one).

A new BR https://bugs.mageia.org/show_bug.cgi?id=19675 was opened to keep track of this problem