Kodi Community Forum
[LINUX] My Asrock 330 ION Tweaks, Tips, Tricks, and Fixes! - Printable Version

+- Kodi Community Forum (https://forum.kodi.tv)
+-- Forum: Support (https://forum.kodi.tv/forumdisplay.php?fid=33)
+--- Forum: General Support (https://forum.kodi.tv/forumdisplay.php?fid=111)
+---- Forum: Linux (https://forum.kodi.tv/forumdisplay.php?fid=52)
+---- Thread: [LINUX] My Asrock 330 ION Tweaks, Tips, Tricks, and Fixes! (/showthread.php?tid=61697)

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16


- rossoneri711 - 2010-02-02

Thanks a lot guys for the help.

ok so here is the output for lsmod below

http://iquik.pastebin.com/m6a2780e3

and here is the output for dmesg | grep lirc

Quote:[ 8.560719] lirc_dev: IR Remote Control driver registered, major 61
[ 8.696391] lirc_wb677: disagrees about version of symbol module_layout
[ 16.412766] lirc_wb677: disagrees about version of symbol module_layout



- monty0815 - 2010-02-02

To clarify my previous post:

Instead of issuing the modprobe command you can simply go ahead and add the correct line to /etc/modules or issue this command as root:
Code:
echo "w83627ehf force_id=0xa510" >> /etc/modules
Now the module should load on the next reboot.

You can find the chip_id that I am forcing onto the driver in the linux sources at drivers/hwmon/w83627ehf.c:
Code:
Chip        #vin    #fan    #pwm    #temp  chip IDs       man ID
    w83627ehf   10      5       4       3      0x8850 0x88    0x5ca3
                                               0x8860 0xa1
    w83627dhg    9      5       4       3      0xa020 0xc1    0x5ca3
    w83627dhg-p  9      5       4       3      0xb070 0xc1    0x5ca3
    w83667hg     9      5       3       3      0xa510 0xc1    0x5ca3
Last line shows w83667hg, which seems to be close enough to the one used in the 330HT (and which the new sensors-detect actually) finds.

As far as I'm concerned the entry in the lm-sensors FAQ does not bother me, as I am running the stock karmic kernel (2.6.31-17). the issue seems to affect 2.6.30 and below.

Also I don't think that you have to patch the module, support for this specific chip has been added quite some time ago.

Btw thanks for this link http://article.gmane.org/gmane.linux.drivers.sensors/21091
I'll try to figure out which temperatures are actually useful, but after a first test, all 3 temp readings seem to change during a small stress test.

regards,
monty


- andyblac - 2010-02-02

monty0815 Wrote:To clarify my previous post:

Instead of issuing the modprobe command you can simply go ahead and add the correct line to /etc/modules or issue this command as root:
Code:
echo "w83627ehf force_id=0xa510" >> /etc/modules
Now the module should load on the next reboot.

You can find the chip_id that I am forcing onto the driver in the linux sources at drivers/hwmon/w83627ehf.c:
Code:
Chip        #vin    #fan    #pwm    #temp  chip IDs       man ID
    w83627ehf   10      5       4       3      0x8850 0x88    0x5ca3
                                               0x8860 0xa1
    w83627dhg    9      5       4       3      0xa020 0xc1    0x5ca3
    w83627dhg-p  9      5       4       3      0xb070 0xc1    0x5ca3
    w83667hg     9      5       3       3      0xa510 0xc1    0x5ca3
Last line shows w83667hg, which seems to be close enough to the one used in the 330HT (and which the new sensors-detect actually) finds.

As far as I'm concerned the entry in the lm-sensors FAQ does not bother me, as I am running the stock karmic kernel (2.6.31-17). the issue seems to affect 2.6.30 and below.

Also I don't think that you have to patch the module, support for this specific chip has been added quite some time ago.

Btw thanks for this link http://article.gmane.org/gmane.linux.drivers.sensors/21091
I'll try to figure out which temperatures are actually useful, but after a first test, all 3 temp readings seem to change during a small stress test.

regards,
monty

yup, works without building Smile

here is the new info for advancedsettings.xml

Code:
<cputempcommand>echo "$(sensors -u | tail -n64 | grep temp1_input | awk '{print $2 }' |awk '{printf("%d\n",$1 + 0.5);}') C"</cputempcommand>



- X3lectric - 2010-02-02

rossoneri711 Wrote:Thanks a lot guys for the help.

ok so here is the output for lsmod below

http://iquik.pastebin.com/m6a2780e3

and here is the output for dmesg | grep lirc

ok your drivers has not been properly installed, if you were following the read me that comes with the drivers though it shouldnt differ much try it with terminal.

Again you must do sudo su and enter your password.

Follow this post exactly.... except the dpkg line

http://forum.xbmc.org/showpost.php?p=462900&postcount=60

only difference is the driver version.

so you do

dpkg -i lirc-nct677x-1.0.4-ubuntu9.10.deb
Intead.

now grep lirc again and do lsmod

on grep lirc you should read
Code:
[    9.191974] lirc_dev: IR Remote Control driver registered, major 61
[    9.200647] lirc_wb677 w677hga: chip id high: 0xb4
[    9.200661] lirc_wb677 w677hga: chip id low: 0x73 expect:0x73
[    9.201144] lirc_dev: lirc_register_driver: sample_rate: 0

and lsmod should show the driver on output... Which on your current is missing, whe you start XBMC the remote should work if not you have a bad install and lirc is not working ok.


gave up on IR - dlbogdan - 2010-02-03

Finally I gave up on the IR part of the Asrock (simply too picky about range and angle, too slow with the harmony... gave me a headache) and bought a Sony PS3 Bluetooth Remote. Got it working perfectly (extremely responsive and a 4 story building range! ) with this : http://code.google.com/p/bdremote-ng/downloads/list

The only problem is waking up from suspend with the bluetooth remote.
Any ideas?


- BoyBastos - 2010-02-03

hi X3lectric,

first of all would like to thank you for this wonderful guide. i just have queston regarding sensor on your #2 post. would this instruction fit my ion foxconn nt-330i? or is this just for asrock ion 330? thanks in advance Smile


- X3lectric - 2010-02-03

andyblac Wrote:yup, works without building Smile

here is the new info for advancedsettings.xml

Code:
<cputempcommand>echo "$(sensors -u | tail -n64 | grep temp1_input | awk '{print $2 }' |awk '{printf("%d\n",$1 + 0.5);}') C"</cputempcommand>

Im concerned because the issues which should have been resolved have really not with the kernel 2.6.31-xx nnot even with the very latest 2.6.32 unless your compiling it yourselves with the proposed solution.

Further reading https://bugs.launchpad.net/ubuntu/+source/linux/+bug/440470

Whatever kernel you guys use while the result of
Code:
dmesg | grep SMB
is
Code:
[    0.472150] ACPI: PCI Interrupt Link [LSMB] (IRQs 20 21 22 23) *11
[    5.610005] i2c-adapter i2c-0: nForce2 SMBus adapter at 0x4d00
[    5.610098] nForce2_smbus 0000:00:03.2: Error probing SMB2.
This means there are still unresolved issues which may be correlated.

However despite alarm bells ringing, I think this is indeed good progress and valuable posts and contributions to this topic and other users for that thank you monty0815 and andyblac


- X3lectric - 2010-02-03

dlbogdan Wrote:Finally I gave up on the IR part of the Asrock (simply too picky about range and angle, too slow with the harmony... gave me a headache) and bought a Sony PS3 Bluetooth Remote. Got it working perfectly (extremely responsive and a 4 story building range! ) with this : http://code.google.com/p/bdremote-ng/downloads/list

The only problem is waking up from suspend with the bluetooth remote.
Any ideas?

no idea personally about bluetooth remotes and no issues with angles or distance within reason.

Did you have issues with cold start (start from total shutdown) with the asrock remote? If the answer is yes you may have a bummed automatic voltage regulator which can account for some other remote issues. Though Asrock 1st line support will try fob off and dismiss problems.

Usual response is that they installed 9.10 tested it on their systems and found no issues.

They ignore the fact most ppl are installing minimal ubuntu and do not take that into consideration though they should. Also I doubt they actually test anything since it's nuvoton who is doing the testing for them and the machines are or have had repairs done to fix poor hardware issues.

Personally I had 2 other HT's and both were suffering from that now on my 3rd it finally seems a better part as resolved this issue.

BoyBastos Wrote:hi X3lectric,

first of all would like to thank you for this wonderful guide. i just have queston regarding sensor on your #2 post. would this instruction fit my ion foxconn nt-330i? or is this just for asrock ion 330? thanks in advance Smile

Yes that is system independent as long as your system relies on coretemp and complains it cant be inserted so compiling is a solution.

it would be worth you looking closer at your hardware and following what is going on in these latest posts regarding temps. Specially because coretemp is not accurate due to various issues.

if you want to know how close your system is in relation to a ION HT (which I'm willing to bet my scrotum is manufactured by Pegatron even if yours is foxconn and this one is Asrock) pastebin the outputs of this you may need to install any commands not found. In meanwhile I will do same for comparison with yours and other systems.

One thing that must be kept in mind is that these ION platforms are relatively new and some or other parts may not be yet fully supported by Linux. That doesn't mean it cant be fixed whatever is not working quite right. This is the main purpose of this Topic.


- BoyBastos - 2010-02-03

Thanks for your quick reply X3lectric.

yes my system relies on coretemp as i have it default on screenlet. Smile

here is the pastebin of my hardware :

Hardware List

One last thing mate, i am planning to get a usb ir receiver for remote in xbmc. any recommendation that is compatible with ubuntu 9.10? i just need a receiver as i have a harmony remote ready on my home theater.

Thanks again.


- X3lectric - 2010-02-03

Any MCE remote should work ok I have a spare one going for £20 + shipping just used once for testing.

It's that one there only they come in black not silver. if your interested drop a pm. Im in UK BTW.

On first look it seems your system is sporting pretty much same components. What happens when you do

dmesg | grep SMB


- BoyBastos - 2010-02-03

here is what i get (dmesg | grep SMB)

[sudo] password for jeff:
[ 0.508640] ACPI: PCI Interrupt Link [LSMB] (IRQs 20 21 22 23) *10
[ 16.780081] i2c-adapter i2c-0: nForce2 SMBus adapter at 0x4d00
[ 16.780120] nForce2_smbus 0000:00:03.2: Error probing SMB2.

i havent followed your instructions yet coz am worried am new to linux came from windows/mac hehehe just making sure before i do this Smile


- andyblac - 2010-02-03

X3lectric Wrote:
Code:
dmesg | grep SMB
is
Code:
[    0.472150] ACPI: PCI Interrupt Link [LSMB] (IRQs 20 21 22 23) *11
[    5.610005] i2c-adapter i2c-0: nForce2 SMBus adapter at 0x4d00
[    5.610098] nForce2_smbus 0000:00:03.2: Error probing SMB2.
This means there are still unresolved issues which may be correlated.


think i have found a solution Smile

heres my:
Code:
dmesg | grep SMB
Code:
[    0.472164] ACPI: PCI Interrupt Link [LSMB] (IRQs 20 21 22 23) *11
[    5.929393] i2c-adapter i2c-0: nForce2 SMBus adapter at 0x4d00
[    5.929491] i2c-adapter i2c-1: nForce2 SMBus adapter at 0x4e00
here what you neeed to do
Code:
sudo sed -i 's/splash quiet/splash quiet acpi_enforce_resources=lax/g' /etc/default/grub
sudo update-grade
sudo reboot



- dlbogdan - 2010-02-03

I've found my remote workaround.
I'm using a sony bd remote paired with a PC104 computer that listens with bdremote daemon for buttons and sends them through network to the lirc daemon that runs on the Asrock ION. For poweron/wake-up the PC104 sends a WOL magic packet to the Asrock.

The solution is not perfect (i.e. have to run the PC104 computer all the time, though it only drains about 5 Watts) but works perfectly (complete with remote poweron/wakeup).

The IR works perfect for poweron/wake-up but fails miserably for usual commands. (installed 1.0.4) it simply ignores most of the single button presses. You really have to find the perfect angle and i've just had enough. (i'm so pissed that i'll be selling the harmony 525 and i'll be thinking for a way to make my PC104 send IR commands for the rest of the AV setup and make the SONY BD Remote -- universal).

cheers!


- X3lectric - 2010-02-03

andyblac Wrote:think i have found a solution Smile

heres my:
Code:
dmesg | grep SMB
Code:
[    0.472164] ACPI: PCI Interrupt Link [LSMB] (IRQs 20 21 22 23) *11
[    5.929393] i2c-adapter i2c-0: nForce2 SMBus adapter at 0x4d00
[    5.929491] i2c-adapter i2c-1: nForce2 SMBus adapter at 0x4e00
here what you neeed to do
Code:
sudo sed -i 's/splash quiet/splash quiet acpi_enforce_resources=lax/g' /etc/default/grub
sudo update-grade
sudo reboot

Yea I have posted end of this post that you can lax the kernel what that means is that it will default to old behaviour before the acpi/hwmon was changed so that the IO ports are not being probed by both.

That resolves the error indeed but if both acpi and hwmon are using the same IO port for this you are just allowing what I am trying to avoid.

That is IF I still haven't determined if in this instance this particular IO port is in use by ACPI, still looking into it.

Better be safe then sorry. However your posts are still emparting with very usefull information until such determination is found.

BoyBastos Wrote:Thanks for your quick reply X3lectric.

yes my system relies on coretemp as i have it default on screenlet. Smile

here is the pastebin of my hardware :

Hardware List

One last thing mate, i am planning to get a usb ir receiver for remote in xbmc. any recommendation that is compatible with ubuntu 9.10? i just need a receiver as i have a harmony remote ready on my home theater.

Thanks again.

Please look at the end of the 3rd post for the HT HW list.


- X3lectric - 2010-02-03

andy this is what I have pulled off to see if ACPI is using the IO is question and to see if the IO for
Code:
modprobe w83627ehf force_id=0xa510
is not being used by acpi
http://iquik.pastebin.com/fd1310c7
That is the output of
Code:
cat /proc/ioports
if neither IO is actually being used for either of this assuming they are unrelated then what you and monty0815 have started is VERY useful.

Further information may need to be extracted off the system to fully determine if this is all safe. Also as bug report suggests a proper fix is already in the works.

You do understand why I'm making a "fuss" over this? Don't ya?

Further on this

Normal output of cat /proc/ioports
Code:
0080-008f : dma page reg
00a0-00a1 : pic2
00c0-00df : dma2
00f0-00ff : fpu
[color=Red]0290-029f : pnp 00:09[/color]  [color=SeaGreen]##This is the address of the modprobed chip[/color]
03c0-03df : vga+
04d0-04d1 : pnp 00:05
0800-080f : pnp 00:05
0810-0815 : ACPI CPU throttle
0cf8-0cff : PCI conf1
4000-407f : pnp 00:05
  4000-4003 : ACPI PM1a_EVT_BLK
  4004-4005 : ACPI PM1a_CNT_BLK
  4008-400b : ACPI PM_TMR
  4020-4027 : ACPI GPE0_BLK
4080-40ff : pnp 00:05
4400-447f : pnp 00:05
4480-44ff : pnp 00:05
  4484-4484 : ACPI PM2_CNT_BLK
  44a0-44af : ACPI GPE1_BLK
4800-487f : pnp 00:05
4880-48ff : pnp 00:05
4900-493f : 0000:00:03.2
4d00-4d3f : 0000:00:03.2
  [color=SeaGreen][COLOR=Red]4d00-4d3f : nForce2_smbus[/color]   ##This is the SMBUS error address without laxing kernel.
[color=Red]4e00-4e3f : 0000:00:03.2[/color][/COLOR]
4f00-4fff : 0000:00:03.0
c480-c48f : 0000:00:0b.0
  c480-c48f : ahci
c800-c803 : 0000:00:0b.0
  c800-c803 : ahci
c880-c887 : 0000:00:0b.0
  c880-c887 : ahci
cc00-cc03 : 0000:00:0b.0
  cc00-cc03 : ahci
d000-d007 : 0000:00:0b.0
  d000-d007 : ahci
d080-d087 : 0000:00:0a.0
  d080-d087 : forcedeth
e000-efff : PCI Bus 0000:01
  ec00-ec7f : 0000:01:00.0
output After modprobe w83627ehf force_id=0xa510
Code:
0000-001f : dma1
0020-0021 : pic1
0040-0043 : timer0
0050-0053 : timer1
0060-0060 : keyboard
0064-0064 : keyboard
0070-0071 : rtc0
0080-008f : dma page reg
00a0-00a1 : pic2
00c0-00df : dma2
00f0-00ff : fpu
[color=Red]0290-029f : pnp 00:09        [COLOR=Green]##after modprobing[/color]
  0295-0296 : w83627ehf    [/COLOR][color=Red][COLOR=Green]##after modprobing[/color][/COLOR]
[color=Red]     0295-0296 : w83627ehf[/color]  [color=Red][COLOR=Green]##after modprobing[/color][/COLOR]
03c0-03df : vga+
04d0-04d1 : pnp 00:05
0800-080f : pnp 00:05
0810-0815 : ACPI CPU throttle
0cf8-0cff : PCI conf1
4000-407f : pnp 00:05
  4000-4003 : ACPI PM1a_EVT_BLK
  4004-4005 : ACPI PM1a_CNT_BLK
  4008-400b : ACPI PM_TMR
  4020-4027 : ACPI GPE0_BLK
4080-40ff : pnp 00:05
4400-447f : pnp 00:05
4480-44ff : pnp 00:05
  4484-4484 : ACPI PM2_CNT_BLK
  44a0-44af : ACPI GPE1_BLK
4800-487f : pnp 00:05
4880-48ff : pnp 00:05
4900-493f : 0000:00:03.2
4d00-4d3f : 0000:00:03.2
  [color=Red]4d00-4d3f : nForce2_smbus   [COLOR=Green]##No acpi activity being shown[/color]
4e00-4e3f : 0000:00:03.2       [/COLOR][color=Red][COLOR=Green]##No acpi activity being shown[/color][/COLOR]
[color=Red] 4f00-4fff : 0000:00:03.0[/color]         [color=Red][COLOR=Green]##No acpi activity being shown[/color][/COLOR]
c480-c48f : 0000:00:0b.0
  c480-c48f : ahci
c800-c803 : 0000:00:0b.0
  c800-c803 : ahci
c880-c887 : 0000:00:0b.0
  c880-c887 : ahci
cc00-cc03 : 0000:00:0b.0
  cc00-cc03 : ahci
d000-d007 : 0000:00:0b.0
  d000-d007 : ahci
d080-d087 : 0000:00:0a.0
  d080-d087 : forcedeth
e000-efff : PCI Bus 0000:01
  ec00-ec7f : 0000:01:00.0

At first glance it looks like the proposed fixes both by monty0815 and andyblac who has driven the proper resolution and implementation of do not have any issues with acpi/hwmon conflicting.

I am fairly confident that what has been posted by andy and monty should now be relatively safe to implement though there is no apparent need to lax the kernel other then to silence the error.

Further on this I will still be looking at proper control of fan speeds and overclocking via the W83677HG-I

I see this as a small triumph for the Asrock ION users thx again guys for your incessant work I shall be updating the main posts to reflect the latest changes.

Please use any of the contents of this full topic at your own risk!!!