HP Spectre x360 Bad Battery Life

[ddegrasse]

Hi,
I'm experiencing much worse battery life than I expected on my HP Spectre x360 (i5-7200U, Intel HD 620 Graphics). The battery lasts roughly 3-4 hours, whereas on windows it lasts much longer, in the range of 7+ hours. I have implemented power management via the PluginType=1 SSDT, As well as injecting custom frequency vectors to enable a lower idle speed. HWP is enabled and persists across sleep.
I have applied the Fix WAK v2, HPET, IRQ, Non-zero Mutex Synclevel, OS Check (reports windows 2012), RTC, SkyLake LPC, and SMBUS DSDT patches. I also have Clover set to Fix Shutdown, and Fix Regions.

AppleIntelInfo.kext shows the following P states:

CPU P-States [ 6 7 8 20 21 24 (26) 31 ] iGPU P-States [ 18 (21) ]

It also shows HWP as enabled:

IA32_PM_ENABLE...................(0x770) : 0x1 (HWP Supported and Enabled)

IOReg shows X86PlatformPlugin as loaded.

I'm using MacBookPro14,1 as my SMBIOS definition, which should provide native kaby lake support

I'm a bit stuck. The only other symptom I've seen of a possible bug is that disk writes, such as the OS X installation and App Store downloads, wildly underestimate the time remaining, for example reporting 12 minutes remaining at the start of the OS X installation, whereas the actual time was closer to 25 minutes.
Another reason for this might be that rather than using NVME hacker kexts, I formatted my NVME drive with 4k blocks.

Any ideas?? Problem reporting zip is attached, along with a verbose boot log in the same zip

Just a note for anyone looking at the CLOVER folder, loader.efi is CLOVERX64.efi. I had to rename it to work around a firmware bug with my pc

 

[P1LGRIM]

Moved to Laptop Support.

 

[RehabMan]

Hi,
I'm experiencing much worse battery life than I expected on my HP Spectre x360 (i5-7200U, Intel HD 620 Graphics). The battery lasts roughly 3-4 hours, whereas on windows it lasts much longer, in the range of 7+ hours. I have implemented power management via the PluginType=1 SSDT, As well as injecting custom frequency vectors to enable a lower idle speed. HWP is enabled and persists across sleep.
I have applied the Fix WAK v2, HPET, IRQ, Non-zero Mutex Synclevel, OS Check (reports windows 2012), RTC, SkyLake LPC, and SMBUS DSDT patches. I also have Clover set to Fix Shutdown, and Fix Regions.

AppleIntelInfo.kext shows the following P states:

CPU P-States [ 6 7 8 20 21 24 (26) 31 ] iGPU P-States [ 18 (21) ]

It also shows HWP as enabled:

IA32_PM_ENABLE...................(0x770) : 0x1 (HWP Supported and Enabled)

IOReg shows X86PlatformPlugin as loaded.

I'm using MacBookPro14,1 as my SMBIOS definition, which should provide native kaby lake support

I'm a bit stuck. The only other symptom I've seen of a possible bug is that disk writes, such as the OS X installation and App Store downloads, wildly underestimate the time remaining, for example reporting 12 minutes remaining at the start of the OS X installation, whereas the actual time was closer to 25 minutes.
Another reason for this might be that rather than using NVME hacker kexts, I formatted my NVME drive with 4k blocks.

Any ideas?? Problem reporting zip is attached, along with a verbose boot log in the same zip

Just a note for anyone looking at the CLOVER folder, loader.efi is CLOVERX64.efi. I had to rename it to work around a firmware bug with my pc

You should check Activity Monitor to see if you have anything using CPU cycles unnecessarily.

 

[ddegrasse]

You should check Activity Monitor to see if you have anything using CPU cycles unnecessarily.

Although to my knowledge this wasn't occurring last time I looked at activity monitor, Kernel task is currently eating up CPU cycles. I'm unsure what could be causing this, but the CPU usage is also over 100%. CPU is currently running around 2.5 GHz

I think this bug may not occur on every boot, but it is unclear.

 

[RehabMan]

Although to my knowledge this wasn't occurring last time I looked at activity monitor, Kernel task is currently eating up CPU cycles. I'm unsure what could be causing this, but the CPU usage is also over 100%. CPU is currently running around 2.5 GHz

I think this bug may not occur on every boot, but it is unclear.

Problem is in some kext you installed.

 

[grassmudhorse]

Although to my knowledge this wasn't occurring last time I looked at activity monitor, Kernel task is currently eating up CPU cycles. I'm unsure what could be causing this, but the CPU usage is also over 100%. CPU is currently running around 2.5 GHz

I think this bug may not occur on every boot, but it is unclear.

Have you figured out what is the problem? My battery life seems to be short as well.

 

[ddegrasse]

Have you figured out what is the problem? My battery life seems to be short as well.

If you are having 100% usage by kernel_task, remove patches to X86 Platform Plugin. That's what fixed it for me.

 

[grassmudhorse]

If you are having 100% usage by kernel_task, remove patches to X86 Platform Plugin. That's what fixed it for me.

Is the patch Plugin=1 in clover ? Can you have HWP after removing it?

Another question maybe slightly off topic. When you changed your WiFi card, is there a warning message Everyime when you boot up ? Also did you have trouble with Bluetooth in windows sometimes?

 

[ddegrasse]

Is the patch Plugin=1 in clover ? Can you have HWP after removing it?

Another question maybe slightly off topic. When you changed your WiFi card, is there a warning message Everyime when you boot up ? Also did you have trouble with Bluetooth in windows sometimes?

yes, use plugintype=1 ssdt. This ssdt will enable HWP. The frequency vectors patch simply lowers the idle frequency of HWP. I haven't seen any of those other issues, sorry