856 replies to this topic

[Pietruszka]

 kozlek, on 27 April 2012 - 07:25 AM, said:

Please try this version (activated temperature algo for earlier cards): GeForceX.kext.zip

no score

GPU TEMP 0C

THX

 kozlek, on 27 April 2012 - 09:12 AM, said:

This register is used by default temperature function for you GPU's arch... But you said it shows zero. Something wrong with GPU initialization in OS X?

Sorry kozlek

my mistake... to many tests

NV_20400 : 0000002f (this is hex) (DEC 47)

NV_20400 : 00000030 (this is hex) (DEC 48)

informations are from gt230 nvidia (the same G92 arch)
with my 8800gs NV_20400 I've got 00000000

need to find how Riva Tuner gets gpu temp

thank you

update:
riva tuner shows gpu temp because of use nvthermaldiode.dll
there is a source code but don't know if it can help kozlek

update2:
nvthermaldiode.dll is the only way to see gpu temp with 90% of 8800gs
there is something strange with built-in thermal diode so don't bother kozlek

[magnifico]

kozlek said:

Use installer from my branch and reinstall all the kexts and utilities: https://github.com/kozlek/HWSensors

Hi Friend , please i have installed your installer
for you work perfect on my pc...what do you need?
Vmemory dont work , and +12V


Thanks

[robertx]

hey eh!...i'm using HWsensor on an old lenovo thinkCentre...i get a warning from superio during boot

Apr 27 21:49:40 localhost kernel[0]: PFM64 (36 cpu) 0xf10000000, 0xf0000000
Apr 27 21:49:40 localhost kernel[0]: [ PCI configuration begin ]
Apr 27 21:49:40 localhost kernel[0]: console relocated to 0xf19000000
Apr 27 21:49:40 localhost kernel[0]: PCI configuration changed (bridge=1 device=2 cardbus=0)
Apr 27 21:49:40 localhost kernel[0]: [ PCI configuration end, bridges 3 devices 20 ]
Apr 27 21:49:40 localhost kernel[0]: AppleIntelCPUPowerManagement: (built 21:08:10 Aug  9 2011) initialization complete
Apr 27 21:49:40 localhost kernel[0]: RTC: Only single RAM bank (128 bytes)
Apr 27 21:49:40 localhost kernel[0]: FakeSMC: Opensource SMC device emulator. Copyright 2009 netkas. All rights reserved.
Apr 27 21:49:40 localhost kernel[0]: FakeSMC: Hardware monitoring plugins. Copyright 2012 kozlek, usr-sse2, slice, navi. All rights reserved.
Apr 27 21:49:40 localhost kernel[0]: FakeSMCDevice: 19 preconfigured key(s) added
Apr 27 21:49:40 localhost kernel[0]: SMC: successfully initialized
Apr 27 21:49:40 localhost kernel[0]: IntelThermal: CPU family 0x6, model 0xf, stepping 0xb, cores 2, threads 2, TJmax 90
Apr 27 21:49:40 localhost kernel[0]: GeForceX: VBIOS successfully read from PRAMIN
Apr 27 21:49:40 localhost kernel[0]: GeForceX: BIT VBIOS found
Apr 27 21:49:40 localhost kernel[0]: GeForceX: detected an NVD0 generation card (0x0d9000a1) with 1024Mb of DDR3 memory (6)
Apr 27 21:49:40 localhost kernel[0]:[b] SuperIO: [Warning] found unsupported ship ID=0xffff[/b]
Apr 27 21:49:40 localhost kernel[0]: mbinit: done [64 MB total pool size, (42/21) split]
Apr 27 21:49:40 localhost kernel[0]: Pthread support ABORTS when sync kernel primitives misused
Apr 27 21:49:40 localhost kernel[0]:[b] vendor:device: 0x8086:0x10bd.[/b]
Apr 27 21:49:40 localhost kernel[0]: AppleIntelE1000e(Info): changing MTU from 0 to 1500
Apr 27 21:49:40 localhost kernel[0]: com.apple.AppleFSCompressionTypeDataless kmod start
Apr 27 21:49:40 localhost kernel[0]: com.apple.AppleFSCompressionTypeZlib kmod start
Apr 27 21:49:40 localhost kernel[0]: com.apple.AppleFSCompressionTypeDataless load succeeded
Apr 27 21:49:40 localhost kernel[0]: com.apple.AppleFSCompressionTypeZlib load succeeded
Apr 27 21:49:40 localhost kernel[0]: AppleIntelCPUPowerManagementClient: ready
Apr 27 21:49:40 localhost kernel[0]: rooting via boot-uuid from /chosen: B05F5F9D-FDD2-3E59-80AD-21FD85D96A05
Apr 27 21:49:40 localhost kernel[0]: From path: "uuid", Waiting on <dict ID="0"><key>IOProviderClass</key><string ID="1">IOResources</string><key>IOResourceMatch</key><string ID="2">boot-uuid-media</string></dict>
Apr 27 21:49:40 localhost kernel[0]: Got boot device = IOService:/AppleACPIPlatformExpert/PCI0@0/AppleACPIPCI/IDE1@1F,2/AppleIntelPIIXATARoot/PRID@0/AppleIntelICHxSATA/ATADeviceNub@0/AppleATADiskDriver/IOATABlockStorageDevice/IOBlockStorageDriver/ST380815AS ST380815AS/IOFDiskPartitionScheme/Untitled 5@5
Apr 27 21:49:40 localhost kernel[0]: BSD root: disk0s5, major 14, minor 3
Apr 27 21:49:40 localhost kernel[0]: Kernel is LP64
Apr 27 21:49:44 localhost kernel[0]: Waiting for DSMOS...
Apr 27 21:49:44 localhost kernel[0]: macx_swapon SUCCESS
Apr 28 01:49:02 localhost kernel[0]: Apple16X50PCI3: Identified 1 Serial channels at PCI Bus=0 Dev=3 Func=3
Apr 28 01:49:02 localhost kernel[0]: Apple16X50UARTSync3: Detected 16550AF/C/CF FIFO=16 MaxBaud=115200
Apr 28 01:49:02 localhost kernel[0]: Previous Shutdown Cause: 3
Apr 28 01:49:02 localhost kernel[0]: Controller: Intel 82801I (vendor ID: 8086, device ID: 293e)
Apr 28 01:49:02 localhost kernel[0]: Controller: NVidia (Unknown) (vendor ID: 10de, device ID: 0e08)
Apr 28 01:49:02 localhost kernel[0]: NVDAGF100HAL loaded and registered.
Apr 28 01:49:03 localhost kernel[0]: DSMOS has arrived

...can this be fixed?

[kozlek]

 robertx, on 28 April 2012 - 08:32 PM, said:

hey eh!...i'm using HWsensor on an old lenovo thinkCentre...i get a warning from superio during boot

...can this be fixed?

Seems your hardware is unsupported.

 Pietruszka, on 27 April 2012 - 06:49 PM, said:

no score

GPU TEMP 0C

THX



Sorry kozlek

my mistake... to many tests

NV_20400 : 0000002f (this is hex) (DEC 47)

NV_20400 : 00000030 (this is hex) (DEC 48)

informations are from gt230 nvidia (the same G92 arch)
with my 8800gs NV_20400 I've got 00000000

need to find how Riva Tuner gets gpu temp

thank you

update:
riva tuner shows gpu temp because of use nvthermaldiode.dll
there is a source code but don't know if it can help kozlek

update2:
nvthermaldiode.dll is the only way to see gpu temp with 90% of 8800gs
there is something strange with built-in thermal diode so don't bother kozlek

Try latest rev. I have imported RivaTuner logic for G92

[Pietruszka]

 kozlek, on 29 April 2012 - 05:56 PM, said:

Try latest rev. I have imported RivaTuner logic for G92

GPU TEMP 239C

will try to open nvthermaldiode source code with Visual C++ and find
which part of code is for my GPU

update: VC++ 2010 can't open project (can't upgrade), will try older version

[kozlek]

 Pietruszka, on 29 April 2012 - 07:23 PM, said:

GPU TEMP 239C

This should show something better:  HWSensors.pkg.zip   557.39K   45 downloads

[magnifico]

This is a new release?
For all?

[SuperstarGoddess]

Why does Vertex 3 ALWAYS show up as 128 degrees?  Is there something wrong with that or is it always the exact same temperature?  ><

[Luthraz]

Thanks for the great work Kozlek. I can view the temperature for just about everything BUT my CPU. I have a Core 2 Quad Q6600 @ 2.40Ghz. Any ideas on how to fix this?

[magnifico]

 SuperstarGoddess, on 29 April 2012 - 11:21 PM, said:

Why does Vertex 3 ALWAYS show up as 128 degrees?  Is there something wrong with that or is it always the exact same temperature?  ><

This is a bug firmware vertex ...relax .
Go on forum oczvertex

[deckert]

Finally got around to adding a profile for my motherboard to the Nuvoton sensor configurations in the SuperIOSensors.kext Info.plist
I just did a copy/paste from the default profile and made some changes. I didn't change any of the TEMPINX entries as I wasn't sure if they needed changing or not.

<key>ASRock</key>
<dict>
<key>Z68 Pro3-M</key>
<dict>
  <key>FANIN0</key>
  <string>CPU Fan</string>
  <key>FANIN1</key>
  <string>Chassis Fan #1</string>
  <key>FANIN2</key>
  <string>Chassis Fan #2</string>
  <key>FANIN3</key>
  <string>Power Fan</string>
  <key>TEMPIN0</key>
  <string>CPU</string>
  <key>TEMPIN1</key>
  <string></string>
  <key>TEMPIN2</key>
  <string></string>
  <key>TEMPIN3</key>
  <string>System</string>
  <key>TEMPIN4</key>
  <string></string>
  <key>TEMPIN5</key>
  <string></string>
  <key>TEMPIN6</key>
  <string></string>
  <key>TEMPIN7</key>
  <string></string>
  <key>TEMPIN8</key>
  <string></string>
  <key>VIN0</key>
  <string>CPU</string>
  <key>VIN1</key>
  <dict>
   <key>gain</key>
   <integer>5597</integer>
   <key>name</key>
   <string>Main 12V</string>
   <key>offset</key>
   <integer>0</integer>
   <key>reference</key>
   <integer>0</integer>
  </dict>
  <key>VIN2</key>
  <string>Auxiliary 3V</string>
  <key>VIN3</key>
  <string>Main 3V</string>
  <key>VIN4</key>
  <string></string>
  <key>VIN5</key>
  <dict>
   <key>gain</key>
   <integer>2000</integer>
   <key>name</key>
   <string>Main 5V</string>
   <key>offset</key>
   <integer>0</integer>
   <key>reference</key>
   <integer>0</integer>
  </dict>
  <key>VIN6</key>
  <string></string>
  <key>VIN7</key>
  <string>Standby 3V</string>
</dict>
</dict>

[Mieze]

Hello Kozlek,

here is a patch to enable voltage monitoring of the Sandybridge CPUs IGPU and a configuration for MSI B75MA-P45 mainboard.

Regards,

Mausi

Attached Files

•  diff-B75MA-P45.txt   3.47K   6 downloads

[Pietruszka]

 kozlek, on 29 April 2012 - 08:00 PM, said:

This should show something better: HWSensors.pkg.zip

still 239C

no change

[Greggen]

Love it! Going to check all values and names and post any changes here.

One thing though. With FakeSMC 4.0 and IntelCPUMonitor 1.0 I seem to get an extra P-state. Got a 2500k @ 4,2GHz and with that combo I had the normal x16 and x42 but also x27. Can't seem to get this state with 4.2 and IntelThermal 2.0. Was this a bug thats fixed or vise-versa? Or something completely different?

EDIT: Just realized it's not even the same branch. Maybe someone knows something about it anyway?

EDIT 2: Ok, seems like a lot more differences with the multipliers. At least thats what I noticed after playing around with BIOS-settings and such. If i set different multipliers for the number of cores (testing with 42, 43, 44 and 45) and test the different FakeSMCs (the latest here and the one from slice) I get totally different results.

Testing with "yes > /dev/null", values from MSRDumper.

FakeSMC from this thread:
4 cores in use, x31
3 cores, x32
2 cores, x43
1 core, x44
Other states: x16, x42

Slice FakeSMC:
4 cores in use, x42
3 cores, x43
2 cores, x44
1 core, x45
Other states: x16, x30

So the later one seems to be the one thats fully functional. And I can confirm that this just isn't cosmetic, with Geekbench score and temperatures.

Moderboard is Asus P8Z68-V LX and the only changes to DSDT is LPC device-id, added BUS0 and DTGP, IRQ fixes. No GenerateXState in plist. FakeSMC is also the only kext I use what-so-ever.

[GhaleonX]

 deckert, on 01 May 2012 - 07:26 AM, said:

Finally got around to adding a profile for my motherboard to the Nuvoton sensor configurations in the SuperIOSensors.kext Info.plist
I just did a copy/paste from the default profile and made some changes. I didn't change any of the TEMPINX entries as I wasn't sure if they needed changing or not.

This also seems to work nicely for NCT677F (same chip you have) on my ASRock X58 Extreme6. I did multiple readings from my BIOS and the voltage values seem to fall within the range I obtained from there, so if they're not exact, they're at least accurate. My BIOS doesn't give me Aux 3.3v however, so I'm not sure if that seems right or not. Temperatures might need adjusting, unless my CPU/NB gains at least 10C from the BIOS until OSX loads - we should probably check with an external sensor of some kind, and I may even try and fine-tune voltage readings with a multimeter (if I find the time).

Thanx for sharing!

[kozlek]

 Pietruszka, on 01 May 2012 - 11:54 AM, said:

still 239C

no change

Huh, I think It's ok.

The value only needs to be tweaked. If I'll disable RivaTuner's offset addition (offset + 187) and will use default SmartDoctor's offset you'll get 239 - 187 = 52C

 Mausi, on 01 May 2012 - 08:44 AM, said:

Hello Kozlek,

here is a patch to enable voltage monitoring of the Sandybridge CPUs IGPU and a configuration for MSI B75MA-P45 mainboard.

Regards,

Mausi

Added you profile but need to think about iGPU because we could have multiple GPUs. Thank you for sharing.

 deckert, on 01 May 2012 - 07:26 AM, said:

Finally got around to adding a profile for my motherboard to the Nuvoton sensor configurations in the SuperIOSensors.kext Info.plist
I just did a copy/paste from the default profile and made some changes. I didn't change any of the TEMPINX entries as I wasn't sure if they needed changing or not.

Added you profile too. Thank you.

[Mieze]

 kozlek, on 01 May 2012 - 07:02 PM, said:

Added you profile but need to think about iGPU because we could have multiple GPUs. Thank you for sharing.

Ok! In case you don't want to add the iGPU-patch please rename the GPU entry in the profile so that the voltage gets displayed anyway.

Mieze

[Greggen]

HWMonitor seems to be frozen almost all the time. The stats updates when I start the app but after that it's completely frozen. Works fine in iStat Menus.

[kozlek]

 Greggen, on 01 May 2012 - 09:06 PM, said:

HWMonitor seems to be frozen almost all the time. The stats updates when I start the app but after that it's completely frozen. Works fine in iStat Menus.

Fixed!

Seems it's also fixed overinfluence on CPU multiplier (abnormal CPU usage while updating sensors values).

[Greggen]

 kozlek, on 02 May 2012 - 06:25 PM, said:

Fixed!

Seems it's also fixed overinfluence on CPU multiplier (abnormal CPU usage while updating sensors values).

Perfect! Now it updates =)

But the speed stepping still doesn't seem 100% correct. I have all the values (16, 31, 32, 33, 42, 43, 44, 45) but when the CPU uses all cores it goes down to 31 instead of 42. Im fully aware that the multipliers over 42 might not work in OSX (but I want them on other OSes and they do show with MSRDumper) but with the other FakeSMC (Slice's branch) it still behaves correctly. When using just one core with your FakeSMC it behaves correctly (sits at 45/44) but from there it goes to 33 with 2 cores, 32 with 3 and 31 with all 4. State 42 and 43 are there though, they pop up occasionally, but not when they should.

Going to try out what happens if I set the highest value to 42 in BIOS. I'll be back with that info.

EDIT: Ok, it' still the same so it had nothing to do with the values being over 42. Now they are set to 42, 41, 40 and 39. With one core it goes to 41/42 and then its the same behavior as above. Steps down to 33, 32 and stops at 31.