Upgrading CPU from 32 to 64bit and solving elusive errors from X

Recently I have been upgrading a few of my laptops, where possible from 32bit to 64bit processors, mostly replacing 1.86Ghz Celeron M440 single core processors (27watt) with 2.0Ghz T7200 core2 processors (34watt) – both are socket M – so far the extra wattage required has not been an issue and for the most part there have been no major issues.

I did however find a kernel error/oops repeated 4 times in dmesg, each looked like this:

Dec 17 22:16:52 hostname kernel: i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=io+mem
Dec 17 22:16:52 hostname systemd[1]: Started Load/Save RF Kill Switch Status.
Dec 17 22:16:52 hostname systemd[1]: Reached target Sound Card.
Dec 17 22:16:52 hostname kernel: plane B assertion failure, should be off on pipe B but is still active
Dec 17 22:16:52 hostname kernel: ————[ cut here ]————
Dec 17 22:16:52 hostname kernel: WARNING: CPU: 0 PID: 239 at drivers/gpu/drm/i915/intel_display.c:1295 assert_planes_disabled+0x124/0x130 [i915]
Dec 17 22:16:52 hostname kernel: Modules linked in: i915(+) mac80211 coretemp cfg80211 i2c_algo_bit lpc_ich wmi_bmof drm_kms_helper snd_hda_codec_conexant
snd_hda_code
Dec 17 22:16:52 hostname kernel: CPU: 0 PID: 239 Comm: systemd-udevd Not tainted 4.14.6-1-ARCH #1
Dec 17 22:16:52 hostname kernel: Hardware name: Hewlett-Packard Presario C500 (GC193PA#ABG) /30C6, BIOS F.24 04/25/2007
Dec 17 22:16:52 hostname kernel: task: ffff8a07baa66c00 task.stack: ffffa2dc4069c000
Dec 17 22:16:52 hostname kernel: RIP: 0010:assert_planes_disabled+0x124/0x130 [i915]
Dec 17 22:16:52 hostname kernel: RSP: 0018:ffffa2dc4069f900 EFLAGS: 00010286
Dec 17 22:16:52 hostname kernel: RAX: 0000000000000046 RBX: ffff8a07bbaa8000 RCX: 0000000000000000
Dec 17 22:16:52 hostname kernel: RDX: 0000000000000000 RSI: ffff8a07bf40dc98 RDI: ffff8a07bf40dc98
Dec 17 22:16:52 hostname kernel: RBP: 0000000000000002 R08: 0000000000000001 R09: 000000000000027c
Dec 17 22:16:52 hostname kernel: R10: ffff8a07bbaa81a0 R11: 0000000000000000 R12: 0000000000000001
Dec 17 22:16:52 hostname kernel: R13: 0000000000000042 R14: ffff8a07baa91828 R15: ffff8a07baa91830
Dec 17 22:16:52 hostname kernel: FS: 00007ff48223d0c0(0000) GS:ffff8a07bf400000(0000) knlGS:0000000000000000
Dec 17 22:16:52 hostname kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Dec 17 22:16:52 hostname kernel: CR2: 0000560f26f35a50 CR3: 000000007aa5c000 CR4: 00000000000006f0
Dec 17 22:16:52 hostname kernel: Call Trace:
Dec 17 22:16:52 hostname kernel: intel_disable_pipe+0x4e/0x190 [i915]
Dec 17 22:16:52 hostname kernel: i9xx_crtc_disable+0x7b/0x410 [i915]
Dec 17 22:16:52 hostname kernel: ? drm_atomic_add_affected_connectors+0x101/0x120 [drm]
Dec 17 22:16:52 hostname kernel: intel_crtc_disable_noatomic+0x8c/0x270 [i915]
Dec 17 22:16:52 hostname kernel: intel_modeset_setup_hw_state+0xb65/0xd30 [i915]
Dec 17 22:16:52 hostname kernel: ? drm_modeset_lock+0x2a/0xa0 [drm]
Dec 17 22:16:52 hostname kernel: intel_modeset_init+0x6c5/0x1850 [i915]
Dec 17 22:16:52 hostname kernel: ? intel_setup_gmbus+0x20a/0x280 [i915]
Dec 17 22:16:52 hostname kernel: i915_driver_load+0xba8/0x1610 [i915]
Dec 17 22:16:52 hostname kernel: ? acpi_dev_found+0x5f/0x70
Dec 17 22:16:52 hostname kernel: local_pci_probe+0x3d/0x90
Dec 17 22:16:52 hostname kernel: ? pci_match_device+0xde/0x110
Dec 17 22:16:52 hostname kernel: pci_device_probe+0x146/0x1b0
Dec 17 22:16:52 hostname kernel: driver_probe_device+0x2f6/0x440
Dec 17 22:16:52 hostname kernel: __driver_attach+0x9d/0xe0
Dec 17 22:16:52 hostname kernel: ? driver_probe_device+0x440/0x440
Dec 17 22:16:52 hostname kernel: bus_for_each_dev+0x69/0xb0
Dec 17 22:16:52 hostname kernel: bus_add_driver+0x1c2/0x260
Dec 17 22:16:52 hostname kernel: ? 0xffffffffc0ab8000
Dec 17 22:16:52 hostname kernel: driver_register+0x57/0xc0
Dec 17 22:16:52 hostname kernel: ? 0xffffffffc0ab8000
Dec 17 22:16:52 hostname kernel: do_one_initcall+0x4b/0x190
Dec 17 22:16:52 hostname kernel: ? kmem_cache_alloc_trace+0x99/0x1b0
Dec 17 22:16:52 hostname kernel: do_init_module+0x5b/0x1ea
Dec 17 22:16:52 hostname kernel: load_module+0x2414/0x2bf0
Dec 17 22:16:52 hostname kernel: ? SYSC_init_module+0x15b/0x190
Dec 17 22:16:52 hostname kernel: SYSC_init_module+0x15b/0x190
Dec 17 22:16:52 hostname kernel: do_syscall_64+0x55/0x110
Dec 17 22:16:52 hostname kernel: entry_SYSCALL64_slow_path+0x25/0x25
Dec 17 22:16:52 hostname kernel: RIP: 0033:0x7ff481b65c7a
Dec 17 22:16:52 hostname kernel: RSP: 002b:00007ffd97e6c818 EFLAGS: 00000246 ORIG_RAX: 00000000000000af
Dec 17 22:16:52 hostname kernel: RAX: ffffffffffffffda RBX: 00005618ef476ce0 RCX: 00007ff481b65c7a
Dec 17 22:16:52 hostname kernel: RDX: 00007ff481437cb5 RSI: 000000000038e368 RDI: 00005618efddaed0
Dec 17 22:16:52 hostname kernel: RBP: 00007ff481437cb5 R08: 00005618ef4835e0 R09: 0000000000000050Dec 17 22:16:52 kernel: R10: 00005618ef451010 R11: 0000000000000246 R12: 00005618efddaed0
Dec 17 22:16:52 hostname kernel: R13: 00005618ef475be0 R14: 0000000000020000 R15: 00007ffd97e6d310
Dec 17 22:16:52 hostname kernel: Code: e9 72 a9 c0 ff 89 ce 48 c7 c7 f0 de 9f c0…
Dec 17 22:16:52 hostname kernel: —[ end trace 646da543156c974a ]—
Dec 17 22:16:52 hostname kernel: [drm] RC6 disabled, disabling runtime PM support
Dec 17 22:16:52 hostname kernel: [drm] initialized overlay support
Dec 17 22:16:52 hostname kernel: [drm] Initialized i915 1.6.0 20170818 for 0000:00:02.0 on minor 0
Dec 17 22:16:52 hostname kernel: WARN_ON(!connector_state->crtc)
Dec 17 22:16:52 hostname kernel: ————[ cut here ]————

 

After a lot of searching to what causes this I found out it was due to a “Phantom” TV0ut either called TV1 or SVIDEO1

lspci:

00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03)
00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller (rev 03)

 

xrandr:

Screen 0: minimum 8 x 8, current 1280 x 800, maximum 32767 x 32767
LVDS1 connected primary 1280×800+0+0 (normal left inverted right x axis y axis) 330mm x 210mm
1280×800 60.00*+
1024×768 60.00
800×600 60.32 56.25
640×480 59.94
640×400 60.00
TV1 disconnected (normal left inverted right x axis y axis)
VGA1 disconnected (normal left inverted right x axis y axis)
VIRTUAL1 disconnected (normal left inverted right x axis y axis)


 
I disabled the TV-1 via grub with no change, to fully fix the error you need to disable KMS, disable the “phantom” SVIDEO1 and implicitly set the display
resolution (mine is 1280×800) to do this I added the following kernal parameters to my bootloader

nomodeset i915.modeset=1 video=SVIDEO-1:d vga=0x0361

I hope that this helps someone somewhere, oh! and happy new year!

Advertisements
Upgrading CPU from 32 to 64bit and solving elusive errors from X

SSD’s , booting too fast and clamd service fails

a brief post to serve as a reminder to myself in future (and hopfully help out others too)

** Edited on 1st July 2017 , the edit is below the original post, which I have kept intact. I may later repost a fully integrated edit.**

I recently got a second user SSD for what was a great price ( from PCBitz) , so I took time out to reinstall ArchLinux  all went smoothly and eventually I was booting up to my chosen DE (desktop environment) Plasma-Desktop by KDE  and I was booting very quickly

systemd-bootime

which is of course one of the reasons for having a SSD

There was one issue however, one of my services clamd which I use to scan downloads in Firefox with fireclam and email in Thunderbird using clamdrib

ERROR: TCP: getaddrinfo failed: Address family for hostname not supported

leading me to start the service manually at which case it woud work perfectly, this led me to the conclusion that although launching after the network services clamd was completing prior to my router issuing an address to my PC.

(and yes I switched back to Tunderbird after many months of using Evolution.)

quick recap The issue was I was booting too quickly and clamd was starting before the PC had received an address from the router.

the solution , my friend Andy over at andrzejl.eu suggested using chronie to delay the start up of the clamd service , which made sense as follows
andys-clamd-solution

However me being me… looked elsewhere for a solution

systemd ( yes obligatory boo hiss from some quarters) has a delay timer function built in all you need to do is set up a timer value and call for a .timer file instead of a .service file , all it takes is a small change to the .service file and save it as a .timer instead

first thing was to disable the systemd service file

sudo systemctl disable clamd.service

then edit the .service file adding

[Timer]
OnBootSec=10

and as it is launching as a time , ther is no need for the instruction to launch after  the network service so comment out that line

#After=network.target

so that the service file looks like

service.timer

then save as

/usr/lib/systemd/system/clamd.timer

all that is then needed is to enable the timer service , using

sudo systemctl enable clamd.timer

and reboot , on all subequent boots the clamd service is sucessfull as it delays its start by a period of 10 seconds , long after the network has started.

I’m hoping this helps someone somewhere , and shoud it not , try Andy’s solution noted above.

Jase

 

EDIT: (Saturday 1st July 2017) troubleshooting addendum

Everything was working great , then I updated . not sure why but my solution presented started to fail to work.

inorder to get this solution working again I did the following

sudo systemctl disable clamd.timer

sudo mv /usr/lib/systemd/system/clamd.timer /etc/systemd/system/

cd /etc/systemd/system

sudo nano clamd.timer

change OnBootSec=10 to OnActiveSec=10 and uncomment the #After line

so that the .timer file looks like

clamd.timer2

Then enable the timer with

sudo systemctl enable clamd.timer

I wil leave this rather disjointed post as it is , as it may help troubleshoot an issue later on.

Jase

 

 

 

 

 

 

 

 

 

SSD’s , booting too fast and clamd service fails

Gargoyle Router and PIA OpenVPN config

OK, its been quite sometime since by last post, lots has happened in the world , stupid is on the rise , evil continues to try to get in the way of our day today,and everyone seems to want to know what YOU are doing (google, crackers , governments, and others)

One way to mitigate this is to use a VPN service , like PIA (Private Internet Access) and the most useful way to do this is by using a router configured to use the service exclusively.

There are many 3rd party router firmwares , however not all routers are compatible, so do check compatibility if you choose to use DD-Wrt , Advanced Tomato, Tomato – by Shibby, OpenWrt, LEDE or as I have for my TP-Link router Gargoyle.

OK for the most part it is REAL easy to set up Gargoyle-Router as a VPN client. the following assume you have Gargoyle-Router  set up and in use ( my network uses the Gargole-Router as a Subnet to another router that handles the ISP connection) . It also assumes that you have ssh access to the router.

before doing anything with the GUI , first ssh into the router and get to the OpenVPN folder

ssh root@192.168.x.x

root@Gargoyle:~# cd /etc/openvpn

root@Gargoyle:/etc/openvpn#

then use the echo command to create the auth.txt file ( contains username and password) and confirm it using the cat command.

root@Gargoyle:/etc/openvpn#echo “<username>” >> auth.txt

root@Gargoyle:/etc/openvpn#echo “<password>” >> auth.txt

root@Gargoyle:/etc/openvpn#cat auth.txt

<username>

<password>

It then get s little tedious as you also have to have a “cat crl.rsa.XXXX.pem” file, if you know how to use vim, good on you , cuz do not except how to delete lines and then save ( see later)

to set up the .pem file , I used the echo command with the >> switch to add each of the 15 lines individually like

root@Gargoyle:/etc/openvpn#echo “—–BEGIN X509 CRL—–” >> crl.rsa.2048.pem
root@Gargoyle:/etc/openvpn#echo “MIICWDCCAUAwDQYJKoZIhvcNAQENBQAwgegxCzAJBgNVBAYTAlVTMQswCQYDVQQI” >> crl.rsa.2048.pem
root@Gargoyle:/etc/openvpn#echo “HgYDVQQDExdQcml2YXRlIEludGVybmV0IEFjY2VzczEgMB4GA1UEKRMXUHJpdmF0” >> crl.rsa.2048.pem
root@Gargoyle:/etc/openvpn#echo “each line in turn” >> crl.rsa.2048.pem

ending with:
root@Gargoyle:/etc/openvpn#echo “—–END X509 CRL—–” >> crl.rsa.2048.pem

then confirming it with

root@Gargoyle:/etc/openvpn#cat crl.rsa.2048.pem

now you can ( for a while go to the GUI and log in.

Once logged in, go to Connections >OpenVPN  and select OpenVPN client. then complete all the required boxes.

vpnsetup2

One this is done Save changes, on doing so it will advise that the configuration has been saved but failed to connect (bummer) , at which point go back to your terminal that you ssh’d into the router with.

The reason why it fails is that Gargoyle writes its own OpenVPN configuration , that points to two files that PIA does not use , some will get round this by making phantom config files ( although as the files are keys, and certs) , this does not seem wise to me.

When back at the terminal you can check the gaining config directly by using

root@Gargoyle:/etc/openvpn# openvpn grouter_client_<randonstringybit>.conf

which will confirm the error, to fix it we need to remove two lines from the .conf file, and the only editor onboard is .. vi

The two lines you need to remove are

cert /etc/openvpn/grouter_client_<randonstringybit>.crt
key /etc/openvpn/grouter_client_<randonstringybit>.keyroot

to do this use vi on the config file

root@Gargoyle:/etc/openvpn# vi grouter_client_<randonstringybit>.conf

then move the cursor to the fist line and tap d twice , it wil remove the line , and then repeat again for the second line, after removing both lines press <esc> :wq <enter>.

you can the retry the config manually again

root@Gargoyle:/etc/openvpn# openvpn grouter_client_<randonstringybit>.conf &

which should now get to the initialisation confirmation and start the VPN client.

when you  rebook you shoudl see a confirmation that the VPN is connceted on the login page , and on the Connections > OpenVPN page.

Mine has now been up for about 2 hours without issue, please post back if you have any success with this.

Jase

 

 

 

 

Gargoyle Router and PIA OpenVPN config

Evolution – unified inbox

This is a little bit of a OMG! moment for at least one person that will shit themselves that not only have I written an entry for WordPress , but two in one day ( gotta be the world is ending.. )

As noted before my recent past involved searching for the ideal Email client  and after trying Thunderbird ( great but potentially EOL and was becoming unstable for me ) , I tried (among other solutions) Sylpheed, Claws, Geary, Seamonkey,  the best alternative I found was Mailpile – which was fantastic , however it suffers from “Beta testing syndrome” – it will be ideal once its worked its kinks out , but for now its not for me.

I settled on Evolution as it has a good combination of  functionality and if you have “gnome-control-center” installed , uses Oauth for google accounts, and serves as a client for the google calendars also ( this for me was the selling point).

After setting up my 5 email accounts and messing about with the configuration , I realised it was lacking one vital function for anyone with multiple accounts , a unified inbox , or so I thought..

To configure a unified inbox you have to set up a search folder ( a virtual folder to see various mails within your email accounts that is filtered by rules). a search folder could be set up to see all the emails sent by a particular person over a specific period of time , or all emails that contain a phrase , I think (hope) you get the idea.

To set up a Search folder you need to navigate (within Evolution) to  Edit > Search folders which brings up a box like this ( I already have configured a unified inbox – yours will most likely be blank)

183844_11_10_2016

To set up a new search folder click “Add” which will present you with another box

184256_11_10_2016

It is then a matter to complete a Name for your search folder and use the rules set to narrow down the criteria of what you wish to look at.

I could make you guess what the rules for a unified inbox are , however as I am feeling nice, here you are

181102_11_10_2016

and there you go all sorted .

If you are wondering where your search folder is  in the GUI of Evolution you can arrange its location in Edit > Preferences > Mail Accounts by dragging and dropping it to the top

Hoping this helps someone, somewhere at sometime.

Jase

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Evolution – unified inbox

of Plasma Fails and Gnome madness

First blog post in a LONG time , but I guess that as I am using it now , it must have some use.

This is mainly a reminder to myself so that I can roll back after trying out various applications and desktop environments .

As Thunderbird (email) and Lightning (calendar) were becoming unstable for me ( + the fact that Thunderbirds days are numbered) I decided to look at alternatives .

The native choice on plasma would be  KDE’s Korganizer and Kmail which after trying for a couple of weeks I realised were basically rubbish (oh yes they are!)  as they were unstable and did not theme nicely as the other KDE/Plasma5 applications

after looking at other options I decided to switch to using Evolution as my email and Gnome-Calendar as my… calendar ( yes there are lots of gnome dependencies and you need gnome-control-center too for Oauth2 on gmail accounts) , despite being Gnome apps , Plasma5 themes them nicely and garnishes them with matching window title bars ( and they look more native than Kmail and Korganizer).

This peaked my attention – the fact that Gnome apps were as good as this , so I thought “lets try Gnome”, having been burned in the past ( the fact that EVERY  time I try the Gnome3 DE that I get frustrated as to how user unfriendly it is)

So prior to doing anything else did the following

pacman -Qnq > before.txt

 

Then installed the remaining part of the Gnome3 environment with

sudo pacman -S gnome gnome-tweak-tool

 

which installed the Gnome3 environment, which I really wanted to get configured and even after about  3 hours , was no where near

Once again  Gnome3 proved to be a ghastly experience (for me) , even with the gnome-tweak-tool.

I will say that I like a traditional desktop which Gnome3 most certainly is not.

Having decided to ditch Gnome3 but keep the part of Gnome I was happy with ( Evolution and Gnome-Calendar) I did the following to make a list of the then -installed applications

pacman -Qnq > after.txt

Now all I needed to do extract the differences between before.txt and after.txt to create appstoberemoved.txt

comm -13 <(sort before.txt) <(sort after.txt) > appstoberemoved.txt

and then use pacman to remove them and any other unneeded packages

sudo pacman -Rnu – < appstoberemoved.txt

once completed , I was left as if the installation of  Gnome3 was like a bad dream.

Hope this helps others as it helped me.

Jase

 

 

 

 

 

 

 

 

 

 

of Plasma Fails and Gnome madness

Migrating Back to ArchLinux from Parabola

migrate back to archlinux from parabola

OK these are notes (not full instructions) I generated on moving back to Arch from testing out Parabola If you mess your system up using these notes , its your own fault! dont use these and moan, I will laugh at you, you have been warned …

First chain yourself up again

 pacman -R your-freedom 

Disable signature verification manually by modifying the line in /etc/pacman.conf:

 RemoteFileSigLevel = Never 

Grab the arch keyring and mirrorlist packages

pacman -U http://fooo.biz/archlinux/core/os/i686/archlinux-keyring-20151206-1-any.pkg.tar.xz
pacman -U http://fooo.biz/archlinux/core/os/i686/pacman-mirrorlist-20151122-1-any.pkg.tar.xz 

Replace the Parabola mirrorlist

cp /etc/pacman.d/mirrorlist.pacnew /etc/pacman.d/mirrorlist

If the installation of the mirrorlist package does not generate a .pacnew you will need to extract the mirrorlist and place it /etc/pacman.d/ dont forget to edit /etc/pacman.d/mirrorlist to activate a mirror.
Remove the Libre repo from /etc/pacman.conf by removing the following two lines

[libre]
Include = /etc/pacman.d/mirrorlist 

re-enable signature verification manually by modifying the line in /etc/pacman.conf:

 RemoteFileSigLevel = Required 

Clear the pacman Cache and force the database update

 
pacman -Scc
pacman -Syy 

remove conflicting firmware (or move them if you prefer)

rm /usr/lib/firmware/htc_7010.fw
rm /usr/lib/firmware/htc_9271.fw

update to arch pacman and update the system to non-libre

pacman -S pacman
pacman -Suu linux linux-firmware firefox thunderbird

before rebooting update your bootloader remove /boot/syslinux/syslinux.cfg

rm /boot/syslinux/syslinux.cfg
pacman -S syslinux
syslinux-install_update -iam 

reboot use pacman to identify non archlinux packages for removal

 pacman -Qem 

remove packages using pacman -R

Migrating Back to ArchLinux from Parabola

is kmailservice5 ruining your Plasma5 experience?

After a good friend had a whinge about “mailto:” links in Konsole not opening up in Thunderbird, I checked my install to , and yes when you click an email address in Konsole it launched “kmailservice5” which just sat there consuming CPU cycles and not much else, seemingly without a timeout too.

Its important to note that I have an ArchLinux install , with a “minimal”ish Plasma5 desktop environment.

After much searching (xdg-open xdg-email mimetypes etc), I found the solution.

It is NOT a Plasma5 issue (…Surprise!!…)

The solution (well, my solution) is to start up Thunderbird, click on edit, then preferences, and check that Thunderbird is set to being the default email client… yeah, its that simples, should have looked there much earlier

Hopefully when someone searches for “disable kmailservice5” this post will be of use

*** click for further update ***

Jase

is kmailservice5 ruining your Plasma5 experience?