HWSensors project

[Time2Retire]

try with just fakesmc and intel thermal - as M@dMac says

Oh dear. I completely missed his reply. I guess that we were writing our posts at the same time. With me being a bit slower pressing the Add Reply button. Sorry

 

Thank you both!

[DarwinX]

Where is the SuperIO.kext? I am currently using Slice's FakeSMC SuperIO plugin to load the NSC PC8739x

 

NSC chip ID=0xfc REVISION=0x66 on ADDRESS=0x1600

 

I have just noticed that yesterday Mozodojo has committed an NSC PC8739x one chip support to his Github! I am looking forward to testing it to determine if it is a better option than Slice's all inclusive SuperIO plugin.

Thank You, Mozodojo. Henceforth, I shall be keeping an eye on the commit history log.

[kozlek]

How it could be better or worse if this is the same code? To avoid all-in-one detection problems (in some cases cmos or time reset), I decided to use multiple drivers instead of one.

[FKA]

How it could be better or worse if this is the same code? To avoid all-in-one detection problems (in some cases cmos or time reset), I decided to use multiple drivers instead of one.

 

for me, it does seem to be quicker at boot from E/E.

last build i had from slices' thread hung for a little while when 'binding'.

 

D

[kozlek]

for me, it does seem to be quicker at boot from E/E.

last build i had from slices' thread hung for a little while when 'binding'.

 

D

 

Sure. Because of address verification driver waits one second while probing every supported chip family (in superio). But here you have installed only one specific to your hardware - driver waits only 1 second.

[blackosx]

Making the edits to NVClockX.kext as suggested above by THe KiNG and oldnapalm allows it to run from /E/E.. Thanks guys

 

But when running NVClockX from /E/E and idling in the Finder, although the GPU diode and GPU Heatsink readings look normal, I have noticed inconsistent GPU FAN readings of sometimes -4092rpm or 5380rpm? Comparing this to using the previous FakeSMCnVclockPort.kext in FakeSMC from projectOSX loading in to /S/L/E where I would see consistent readings of 1219rpm - 1229rpm

 

Could this be due to the removing of the dependency to com.apple.NVDAResman?

I've also seen the following NVClockX warning in the kernel.log:

Oct 18 06:43:14 localhost kernel[0]: sp78, size 2
Oct 18 06:43:14 localhost kernel[0]: FakeSMC: Adding key F0Ac with handler, type fpe2, size 2
Oct 18 06:43:14 localhost kernel[0]: FakeSMC: Adding key F1Ac with handler, type fpe2, size 2
Oct 18 06:43:14 localhost kernel[0]: FakeSMC: Adding key F2Ac with handler, type fpe2, size 2
Oct 18 06:43:14 localhost kernel[0]: NVClockX: [Warning] error updating FNum value
Oct 18 06:43:14 localhost kernel[0]: mbinit: done (64 MB memory set for mbuf pool)
Oct 18 06:43:14 localhost kernel[0]: From path: "uuid", 
…./snip/.…

[DarwinX]

How it could be better or worse if this is the same code? To avoid all-in-one detection problems (in some cases cmos or time reset), I decided to use multiple drivers instead of one.

 

What I meant was your branch versus Slice's as a best implementation for my machine as obviously they aren't interchangable despite the similarity of code. Individualizing drivers is the way to go, unfortunately there is no NSC support short of the SuperIO kext in the other branch.

 

Making the edits to NVClockX.kext as suggested above by THe KiNG and oldnapalm allows it to run from /E/E.. Thanks guys

 

But when running NVClockX from E/E and idling in the Finder, although the GPU diode and GPU Heatsink readings look normal, I have noticed inconsistent GPU FAN readings of sometimes -4092rpm or 5380rpm? Comparing this to using the previous FakeSMCnVclockPort.kext in FakeSMC from projectOSX loading in to /S/L/E where I would see consistent readings of 1219rpm - 1229rpm

 

Could this be due to the removing of the dependency to com.apple.NVDAResman?

I've also seen the following NVClockX warning in the kernel.log:

Oct 18 06:43:14 localhost kernel[0]: sp78, size 2
Oct 18 06:43:14 localhost kernel[0]: FakeSMC: Adding key F0Ac with handler, type fpe2, size 2
Oct 18 06:43:14 localhost kernel[0]: FakeSMC: Adding key F1Ac with handler, type fpe2, size 2
Oct 18 06:43:14 localhost kernel[0]: FakeSMC: Adding key F2Ac with handler, type fpe2, size 2
Oct 18 06:43:14 localhost kernel[0]: NVClockX: [Warning] error updating FNum value
Oct 18 06:43:14 localhost kernel[0]: mbinit: done (64 MB memory set for mbuf pool)
Oct 18 06:43:14 localhost kernel[0]: From path: "uuid", 
…./snip/.…

 

I believe that Mozodojo has resolved this issue in his latest commit as of 10.16 - "Fixed FNum updating error message"

Reference: http://github.com/Mozodojo/FakeSMC/commits/master

[blackosx]

I believe that Mozodojo has resolved this issue in his latest commit as of 10.16

Thanks DarwinX - that indeed has removed the warning in the kernel.log

However, I still see inconsistent GPU FAN readings in istat menus 3 from -508rpm to 2308rpm?

[The Real Deal]

Those edits would be in the NVClockX.kext of course for those of us following along at home, thanks works great although don't know where the voltage for the CPU comes from as it does not seem to correspond to anything Voodoo Monitor shows...

 

 

 

I have just removed the dependency and it seems to work fine:

 

Oct 18 09:10:08 localhost kernel[0]: NVClockX: started
Oct 18 09:10:08 localhost kernel[0]: NVClockX: Vendor ID: 10de, Device ID: 6e4
Oct 18 09:10:08 localhost kernel[0]: NVClockX: Architecture: 80000
Oct 18 09:10:08 localhost kernel[0]: NVClockX: G98 [GeForce 8400 GS]
Oct 18 09:10:08 localhost kernel[0]: FireWire runtime power conservation disabled. (2)
Oct 18 09:10:08 localhost kernel[0]: FireWire (OHCI) TI ID 8024 PCI now active, GUID 00caf7380000241d; max speed s400.
Oct 18 09:10:08 localhost kernel[0]: BIOS successfully read

 

 

What is the purpose of the second edit?

[kozlek]

What I meant was your branch versus Slice's as a best implementation for my machine as obviously they aren't interchangable despite the similarity of code. Individualizing drivers is the way to go, unfortunately there is no NSC support short of the SuperIO kext in the other branch.

 

Ok, I'll add more descriptive names to Super I/O plugins for you to know there is NSC support in the new branch.

[clockT]

great work!

[sylph93]

Thanks for such a great project!

 

I used slice's 405 version for now. Today I changed to this version.

 

Everything is good except for disappearing CPU heatsink item.

 

I use gigabyte EP35-DS3P, ITE chipset.

 

plz check Ite kext.

 

thanks!

[DarwinX]

Ok, I'll add more descriptive names to Super I/O plugins for you to know there is NSC support in the new branch.

 

I don't think that's actually necessary. Short of personally compiling it, where can I download your NSC Super I/O plugin so I can finally test it out?

[kozlek]

Thanks for such a great project!

 

I used slice's 405 version for now. Today I changed to this version.

 

Everything is good except for disappearing CPU heatsink item.

 

I use gigabyte EP35-DS3P, ITE chipset.

 

plz check Ite kext.

 

thanks!

 

There is no auto-detection code for ITE. Try to open plugin's info.plist and move "Processor" to TEMPIN1.

 

I don't think that's actually necessary. Short of personally compiling it, where can I download your NSC Super I/O plugin so I can finally test it out?

 

PC8739x.kext from the github (link in the 1rst post)

[DarwinX]

PC8739x.kext from the github (link in the 1rst post)

 

Failed to load /System/Library/Extensions/PC8739x.kext - (libkern/kext) link error.

Load org.slice.PC8739x failed; removing personalities.

Failed to load executable for kext org.slice.PC8739x.

Kext org.slice.PC8739x failed to load (0xdc008016).

Failed to load kext org.slice.PC8739x (error 0xdc008016).

kxld[org.slice.PC8739x]: The following vtables were unpatchable because each one's parent vtable either was not found or also was not patchable:

kxld[org.slice.PC8739x]: '__ZTV7PC8739x' (super vtable '__ZTV14SuperIOMonitor')

Can't load kext org.slice.PC8739x - link failed.

Failed to load executable for kext org.slice.PC8739x.

Kext org.slice.PC8739x failed to load (0xdc008016).

Failed to load kext org.slice.PC8739x (error 0xdc008016).

 

The iStat only shows 2 CPU Core and GPU temps. Also, the TjMax=X plist override failed to implement, reason(s) unknown - the default TjMax is simpy too high. The same for both e/e/ + s/l/e/

[stefano.85]

For me NVClockX.kext doesn't work both in E/E and S/L/E. I've made the modifications written in above posts to plist with no results anyway. During boot seems that the kext isn't viewed by OS... no strings of error neither. i've attached the kernel.log for debug. my Hack is in rig (the first one

P.S. I've got istat widget pro installed.... not the "menu" one.

P.P.S. sorry if my english sux

kernel_log.rtf

[kozlek]

Failed to load /System/Library/Extensions/PC8739x.kext - (libkern/kext) link error.

Load org.slice.PC8739x failed; removing personalities.

Failed to load executable for kext org.slice.PC8739x.

Kext org.slice.PC8739x failed to load (0xdc008016).

Failed to load kext org.slice.PC8739x (error 0xdc008016).

kxld[org.slice.PC8739x]: The following vtables were unpatchable because each one's parent vtable either was not found or also was not patchable:

kxld[org.slice.PC8739x]: '__ZTV7PC8739x' (super vtable '__ZTV14SuperIOMonitor')

Can't load kext org.slice.PC8739x - link failed.

Failed to load executable for kext org.slice.PC8739x.

Kext org.slice.PC8739x failed to load (0xdc008016).

Failed to load kext org.slice.PC8739x (error 0xdc008016).

 

The iStat only shows 2 CPU Core and GPU temps. Also, the TjMax=X plist override failed to implement, reason(s) unknown - the default TjMax is simpy too high. The same for both e/e/ + s/l/e/

 

Check the new version in repo. I've made IntelThermal fix but didn't uploaded binary.

[MacKonsti]

Dear mozodojo, I also have a question. Everybody seems to be getting a thermal reading for a Northbridge chip; I am sure that I should also get one, since I am running on a ICH7 (G31) chipset! I know that Core i7 chipsets don't include Northbridge anymore, instead. Can you imagine why I don't have such a reading? I don't get any errors in kernel log. Slice/rubik's plugin produced such a reading... Thank you.

[kozlek]

Dear mozodojo, I also have a question. Everybody seems to be getting a thermal reading for a Northbridge chip; I am sure that I should also get one, since I am running on a ICH7 (G31) chipset! I know that Core i7 chipsets don't include Northbridge anymore, instead. Can you imagine why I don't have such a reading? I don't get any errors in kernel log. Slice/rubik's plugin produced such a reading... Thank you.

 

I have no idea before you'll show me your smc keys dump. Use atMonitor to do that.

[aleister Crowley]

Can you tell me what are these new kexts for ?

 

PC8739x.kext

F718x.kext

[omedia]

I can't found cpu sensors of iSat menus.

What use kext on my system ?

Please tell me.

 

my system :

- i5 760 / ga p55a ud3r (rev.2.0 / bios 12) / GTX260

- Chameleon 2 RC5 rev. 594

- 10.6.5(10H555)

- /Extra/Extensions ( FakeSMC.kext IOAHCIBlockStorageInjector.kext Legacy889HDA.kext)

[The Real Deal]

Mozodojo, radeon support please?

[DarwinX]

Check the new version in repo. I've made IntelThermal fix but didn't uploaded binary.

 

With the latest version, I lose the CPU voltage. The SuperIO.kext does detect my NSC chip now.

 

SuperIO: NSC: Found supported chip ID=0xfc REVISION=0x66 on ADDRESS=0x1600

Warning /: no supported helper partitions to update.

 

 

[kozlek]

With the latest version, I lose the CPU voltage. The SuperIO.kext does detect my NSC chip now.

 

SuperIO: NSC: Found supported chip ID=0xfc REVISION=0x66 on ADDRESS=0x1600

Warning /: no supported helper partitions to update.

 

 

Seems I forget to add voltage sensor.

[MacKonsti]

I have no idea before you'll show me your smc keys dump. Use atMonitor to do that.

Dear mozodojo, I don't know what atMonitor output you require, here is the kernel log during 32-bit boot (for atMonitor to work):

Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Opensource SMC device emulator by netkas © 2009
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Modified for plugins support by mozodojo © 2010
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Idea of FakeSMC plugins and code sample by usr-sse2
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Thanks to slice for help with hardware support code and plugins
Oct 19 14:22:12 localhost kernel[0]: FakeSMCDevice: 16 keys added
Oct 19 14:22:12 localhost kernel[0]: FireWire runtime power conservation disabled. (2)
Oct 19 14:22:12 localhost kernel[0]: FireWire (OHCI) TI ID 8023 built-in now active, GUID 00301b000048628f; max speed s400.
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key TC0D with handler, type sp78, size 2
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key TC1D with handler, type sp78, size 2
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key TC2D with handler, type sp78, size 2
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key TC3D with handler, type sp78, size 2
Oct 19 14:22:12 localhost kernel[0]: IntelThermal: CPU family 0x6, model 0x17, stepping 0xa, cores 4, threads 0
Oct 19 14:22:12 localhost kernel[0]: IntelThermal: CPU Tjmax 100
Oct 19 14:22:12 localhost kernel[0]: IT87x: found ITE IT8718F
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key TN0P with handler, type sp78, size 2
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key Th0H with handler, type sp78, size 2
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key VC0C with handler, type fp2e, size 2
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key F0Ac with handler, type fpe2, size 2
Oct 19 14:22:12 localhost kernel[0]: NVClockX: started
Oct 19 14:22:12 localhost kernel[0]: NVClockX: Vendor ID: 10de, Device ID: 6e4
Oct 19 14:22:12 localhost kernel[0]: NVClockX: Architecture: 80000
Oct 19 14:22:12 localhost kernel[0]: NVClockX: G98 [GeForce 8400 GS]
Oct 19 14:22:12 localhost kernel[0]: BIOS successfully read
Oct 19 14:22:12 localhost kernel[0]: FakeSMC: Adding key TG0D with handler, type sp78, size 2
Oct 19 14:22:12 localhost kernel[0]: AppleIntelCPUPowerManagement: initialization complete

I think you need a copy of the output from atMonitor's Utility > SMC Sensors below. Remember I have ICH7 and Q9550:

sensor: [#KEY]  type: [ui8 ]  value: [                              28]  data: [0000001c]
sensor: [MSSD]  type: [si8 ]  value: [                               3]  data: [03]
sensor: [OSK0]  type: [ch8*]  value: [ourhardworkbythesewordsguardedpl]  data: [6f757268617264776f726b62797468657365776f72647367756172646564706c]
sensor: [MSPS]  type: [{msp]  value: [                                ]  data: [00]
sensor: [LSSB]  type: [{lso]  value: [                               ?]  data: [0101]
sensor: [NVPR]  type: [ui8 ]  value: [                               0]  data: [00]
sensor: [CLKT]  type: [ui32]  value: [                           29211]  data: [0000721b]
sensor: [LSOF]  type: [flag]  value: [                               1]  data: [01]
sensor: [REV ]  type: [{rev]  value: [                           ?????]  data: [01300f000003]
sensor: [CLKH]  type: [{clh]  value: [                         ???????]  data: [0000708000011940]
sensor: [MOST]  type: [ui16]  value: [                           32771]  data: [8003]
sensor: [FNum]  type: [ui8 ]  value: [                               1]  data: [01]
sensor: [ACID]  type: [ch8*]  value: [                          <E   C]  data: [babe3c45c0031043]
sensor: [NTOK]  type: [ui8 ]  value: [                               1]  data: [01]
sensor: [NATJ]  type: [ui8 ]  value: [                               0]  data: [00]
sensor: [OSK1]  type: [ch8*]  value: [easedontsteal©AppleComputerInc]  data: [65617365646f6e74737465616c2863294170706c65436f6d7075746572496e63]
sensor: [TC0D]  type: [sp78]  value: [                           46.00]  data: [2e00]
sensor: [TC1D]  type: [sp78]  value: [                           37.00]  data: [2500]
sensor: [TC2D]  type: [sp78]  value: [                           38.00]  data: [2600]
sensor: [TC3D]  type: [sp78]  value: [                           40.00]  data: [2800]
sensor: [TN0P]  type: [sp78]  value: [                          201.00]  data: [c900]
sensor: [Th0H]  type: [sp78]  value: [                           35.00]  data: [2300]
sensor: [VC0C]  type: [fp2e]  value: [                           70.50]  data: [4680]
sensor: [F0Ac]  type: [fpe2]  value: [                         1136.00]  data: [11c0]
sensor: [F0ID]  type: [ch8*]  value: [                         Chassis]  data: [43686173736973]
sensor: [TG0D]  type: [sp78]  value: [                           49.00]  data: [3100]
sensor: [NATi]  type: [ui16]  value: [                               0]  data: [0000]
sensor: [MSDW]  type: [ui8 ]  value: [                               1]  data: [01]

Thank you for your time.