[solved] Sierra Toshiba SandyBridge Reboot in Sleep and Shutdown

[bisk]

Yes.
They are nothing like what you would end up with if you followed my guides.

Some example of the brokenness:
- no kexts in EFI/Clover/kexts/Other, so in the event of a system update, you won't be able to boot the installer to perform the update. Same goes for not being able to boot the recovery partition.
- dropping all SSDTs results in problems in DSDT when the code there references symbols that were in the SSDTs. For example, the PNOT method references CPU related identifiers that were dropped because the CPU related SSDTs were dropped. When these symbols are referenced, it will cause ACPI abort. And there are many places in DSDT that call PNOT... Best to not drop those SSDTs, such that the symbols are still present in ACPI namespace. Unless those SSDTs cause KP,.. then you must drop them but patch DSDT appropriately such that the symbols are no longer referenced (PNOT patch).
- panel backlight control not implemented correctly
- USB configuration not correct
- I didn't look at DSDT as there is no ACPI/origin/DSDT.aml to compare against



My laptop guide does not suggest using Unibeast.



100%
You attached the files in post #1.

Fair enough.

I don't have any problem installing Apple system updates or any other types of updates for that matter.
Didn't even realize kexts/Other had anything to do with that.

I had versions of my DSDT both with PNOT references commented out and not commented out.
I noticed no difference between the 2. For most of my hacks, I don't. I suppose that I could just always comment PNOT out but I prefer to leave things original when I notice no improvement from the change.

My LCD backlight is performing perfectly well so far as I can tell. I can change illumination by slider or FKeys smoothly plus with animation and it does not degrade after sleep. Not sure how it could be any better.

Other than to redefine the FKeys so that they actually match the manufacturers' intended function but that's another story.

USB ports all work just fine and they do not hamper sleep either.

I did need to do the _PRW mod however, comment it out in 7 places, which may not be in the version that I posted.

You never need multiple versions to track the mods in one of my DSDTs.
I leave all of the original code, commented out with annotations for every mod.
There isn't that much code to worry about weight here.
I work with enough files already !

I'm not kidding when I tell you that everything works on this laptop.

I never did bother worrying about the Recovery partition tho' so I'll give you that one.

Maybe it's high time that I looked into supporting the Recovery Partition.

I do believe that I have an unusual problem stemming from the ACPI however.

I need to perform an EC reset after every shutdown or the laptop will not turn on again.
I have never seen this happen before or since on any other hack that I've ever done and I do 'em all the same way

If it does it with Windows too than maybe it's just failing hardware.

I have unresolved references to an LCPB.H_EC structure in an SSDT that is defined in no other extracted DSDT/SSDT file. This is NOT the usual EC0 structure but something different.

So it makes no difference to the Embedded Controller whether I don't drop SSDTs or I do.

Maybe gotta track that structure down and define it.

Other than that ...

 

[RehabMan]

I don't have any problem installing Apple system updates or any other types of updates for that matter.
Didn't even realize kexts/Other had anything to do with that.

In 10.13, updates are done via the installer.
You will not be able to boot the installer without FakeSMC.kext.

I had versions of my DSDT both with PNOT references commented out and not commented out.
I noticed no difference between the 2. For most of my hacks, I don't. I suppose that I could just always comment PNOT out but I prefer to leave things original when I notice no improvement from the change.

To properly test, you would need to instrument DSDT (ACPIDebug.kext) such that you can get logs from all code that might reference the aforementioned CPU SSDT objects.

Or pay close attention to the ACPI kernel logs.

My LCD backlight is performing perfectly well so far as I can tell. I can change illumination by slider or FKeys smoothly plus with animation and it does not degrade after sleep. Not sure how it could be any better.

The way you have it, you will end up with different brightness levels before a sleep/wake cycle vs. after.
And likely not full brightness at the highest level, especially after a cold boot.

USB ports all work just fine and they do not hamper sleep either.

Hub ports connected to internal devices should be marked portType=2.
Ports that are not used should be eliminated.

You never need multiple versions to track the mods in one of my DSDTs.
I leave all of the original code, commented out with annotations for every mod.
There isn't that much code to worry about weight here.
I work with enough files already !

Impossible to determine what patches were made to ACPI/patched when ACPI/origin is not provided.

I'm not kidding when I tell you that everything works on this laptop.

Hmm... why post here regarding "Reboot in Sleep and Shutdown"?

I never did bother worrying about the Recovery partition tho' so I'll give you that one.

As I mentioned, for system updates you need to be able to boot the installer.
You can't expect to boot the macOS installer without FakeSMC.kext (and you'll likely need other kexts to interact with the installer... such as PS2 kexts).

I need to perform an EC reset after every shutdown or the laptop will not turn on again.

Your SystemMemory addresses in ACPI/patched are likely out-of-sync with ACPI/origin.
It is one of the first things I check when I look at problem reports,... but you didn't provide the data requested in the FAQ (ACPI/origin is missing).

I have unresolved references to an LCPB.H_EC structure in an SSDT that is defined in no other extracted DSDT/SSDT file. This is NOT the usual EC0 structure but something different.

If it is for PTID, it doesn't matter... macOS doesn't use it.
I think H_EC stands for "Host EC" and is likely implemented by Windows.

 

[IshtarDust]

Just for info.
I had the same problem (reboot instead of shutdown) on my H87N-Wifi with Haswell cpu (10.13.5) and it was solved after replacing ALXEthernet.kext with AtherosE2200Ethernet.kext.

 

[bisk]

Roger that. I've had ALXEthernet.kext crash me out on a couple of other Sierra/High Sierra builds since.

Looks like ALXEthernet.kext is a dead kext now going forward unless somebody is gonna step up and fix it.

But probably nobody will bother with the new Atheros2200Ethernet.kext doing the job on the newer LAN chips tho'.