Jump to content

HP Envy 17t-j000 Quad (Haswell) + 10.8.5/10.9.5/10.10.x/10.11.x/10.12.x/10.13.x/10.14.x


gygabyte666
 Share

1,321 posts in this topic

Recommended Posts

@gygabyte666Can you boot (using Clover) "without caches"? I'm having some issues where I can only boot if I have a kernel cache (and use it).See here for some discussion/background: http://www.projectosx.com/forum/index.php?showtopic=2562&st=11800&p=37514&&do=findComment&comment=37514Since your computer is very similar... wondering if you have the same issue (note you must have FSInject-64.efi to be able force cacheless boot).

Yeah, I've had no problems using that option. I use it rather frequently since I make so many changes in S/L/E. I will admit though that I haven't checked Clover's boot logs since I haven't had problems booting using that option, I never felt the need to.

Link to comment
Share on other sites

Yeah, I've had no problems using that option. I use it rather frequently since I make so many changes in S/L/E. I will admit though that I haven't checked Clover's boot logs since I haven't had problems booting using that option, I never felt the need to.

Thanks for the info.

 

You're using UEFI Clover boot, right? I ask because this problem only happens with UEFI Clover, not legacy boot Clover. It is workable for me to install both UEFI Clover and legacy Clover and boot using legacy when I need to boot w/o caches, but it is kind of an ugly solution (when updating Clover, I need to run the Clover installer twice, be careful with the config, etc). I'm going to put the HDD for Windows back in this machine and do a BIOS upgrade as I think there is one available (or maybe use USB stick). I wonder if there is a difference between BIOS versions here. And I've got some other things to try wrt HFSPlus.efi (trying to extract the one from my MacBookAir6,2).

Link to comment
Share on other sites

Thanks for the info.

 

You're using UEFI Clover boot, right? I ask because this problem only happens with UEFI Clover, not legacy boot Clover. It is workable for me to install both UEFI Clover and legacy Clover and boot using legacy when I need to boot w/o caches, but it is kind of an ugly solution (when updating Clover, I need to run the Clover installer twice, be careful with the config, etc). I'm going to put the HDD for Windows back in this machine and do a BIOS upgrade as I think there is one available (or maybe use USB stick). I wonder if there is a difference between BIOS versions here. And I've got some other things to try wrt HFSPlus.efi (trying to extract the one from my MacBookAir6,2).

No problem and yes, mine is setup as UEFI. I don't have the legacy version installed at all. That is very strange I must admit. You're not the only Envy user I've spoken with who had issues with Clover UEFI. Hell, initially, I had several myself but I'm pretty sure those related to Clover not having support for my newer hardware yet. Hopefully, your issues get resolved soon.

 

It's worth mentioning that my system is set to boot in legacy mode and not in UEFI as it defaults. Since UEFI takes priority during bootup, Clover loads first like I want it to anyway so it works for me. This setup allows me to boot my non-UEFI Windows install. You might want to look into testing it if you haven't already.

 

I need to try a BIOS upgrade too since I noticed an irritating bug that causes my new Azurewave WiFI/BT card to fail to start Bluetooth properly if it is disabled using the F12 switch during a session. I don't believe you've done much BT testing last I checked but could you test out yours if possible? It occurs with OS X and with Windows. In Mac, the BT icon is visible at boot and all options are available but if you try to connect a paired device, it will never connect. The only fix I've found involves resetting the BIOS to it's default settings and manually changing them back to preference and booting again. It is really annoying.

Link to comment
Share on other sites

No problem and yes, mine is setup as UEFI. I don't have the legacy version installed at all. That is very strange I must admit. You're not the only Envy user I've spoken with who had issues with Clover UEFI. Hell, initially, I had several myself but I'm pretty sure those related to Clover not having support for my newer hardware yet. Hopefully, your issues get resolved soon.

 

It's worth mentioning that my system is set to boot in legacy mode and not in UEFI as it defaults. Since UEFI takes priority during bootup, Clover loads first like I want it to anyway so it works for me. This setup allows me to boot my non-UEFI Windows install. You might want to look into testing it if you haven't already.

That's exactly what I'm doing now. I have both Clover UEFI and Clover legacy installed. UEFI takes priority, so in the normal case, when booting using caches everything is fine. If things should go south... cache damaged, bad cache rebuild, or missing cache, then I can use F9 to force legacy boot to the HDD, which loads legacy Clover, which works to boot without caches.

 

I need to try a BIOS upgrade too since I noticed an irritating bug that causes my new Azurewave WiFI/BT card to fail to start Bluetooth properly if it is disabled using the F12 switch during a session. I don't believe you've done much BT testing last I checked but could you test out yours if possible? It occurs with OS X and with Windows. In Mac, the BT icon is visible at boot and all options are available but if you try to connect a paired device, it will never connect. The only fix I've found involves resetting the BIOS to it's default settings and manually changing them back to preference and booting again. It is really annoying.

Unfortunately, I don't have any BT devices. Well, I have an iPhone but I'm not sure that will work for anything...

Link to comment
Share on other sites

There is no update to boot.efi in 10.9.1 (yet). But maybe boot.efi just reads PlatformSupport.plist. Hmmm...

FYI: No update to PlatformSupport.plist (yet) either...

 

Do you have access to the 10.9.1 (13B40) pre-release? If so, i'd really love to try out the boot.efi/PlatformSupport.plist from it with Clover. As a general note, i'd also like to try out the versions included with builds (13A2093) and (13A3017) from the Haswell MBP11,x.

 

Otherwise, it would be really awesome if someone PM'd me with an attachment present.  :whistle:

Link to comment
Share on other sites

I might know how.

 

Can you do me a favor, i will really appreciate!

 

I have a Haswell Laptop, whose Graphic infomations exist in SSDT.

Here is my DSDT + SSDT 

dsdt+ssdt.zip

First of all i have fixed DSDT and inject the HD4400's ig-platform-id and it works well, but havn't removed errors of SSDT(i've tried but failed)

 

and also tried editing the DSDT to disable the graphic card, add the _OFF Method in my DSDT and call it, it worked indeed only once , but after a rebooting it didn't work anymore, i tried many many times it just doesn't work.  it's really weird!

Here is the DSDT i have edited(add _OFF method):

dsdt.aml.zip

 

thanks in advance!!!

Link to comment
Share on other sites

Can you do me a favor, i will really appreciate!

 

I have a Haswell Laptop, whose Graphic infomations [/size]exist [/size]in SSDT.[/size]

Here is my DSDT + SSDT [/size]

attachicon.gifdsdt+ssdt.zip

First of all i have fixed DSDT and inject the HD4400's ig-platform-id and it works well, but havn't removed errors of SSDT(i've tried but failed)[/size]

 

and also tried editing the DSDT to disable the graphic card, add the _OFF Method in my DSDT and call it, it worked indeed only once , but after a rebooting it didn't work anymore, i tried many many times it just doesn't work.  it's really weird!

Here is the DSDT i have edited(add _OFF method):

attachicon.gifdsdt.aml.zip

 

thanks in advance!!!

You need to provide unpatched DSDT + SSDTs. The DSDT in the first ZIP has been modified.

Link to comment
Share on other sites

attachicon.gifdsdt+ssdt.zip

 

but this is the totally original version, which i haven't fix errors, i couldn't find the removed error one... :'( my dist is now totally chaos

Here's something to try. First off, it is helpful to disassemble your ssdt using the dsdt...

iasl -e dsdt.aml -d ssdt.aml
If you have MaciASL installed, you can find iasl in /Applications/MaciASL.app/Contents/MacOS/iasl5.

 

Note that the same technique can be applied to disassembly of your DSDT:

iasl -e ssdt.aml -d dsdt.aml
Resulting in quite a few less errors to fix in your DSDT... But that is off-topic as you do not need to touch the DSDT for this.

 

 

After that, open ssdt.dsl. It will almost compile cleanly. You can fix it with this one patch:

into method label _BCM parent_label DD02 code_regex Return\s\((.*)\)\n\s+(.*) replace_matched begin Return(%1(%2)) end;
into definitionblock code_regex External\s\((.*\._BCM),\s+IntObj\) replace_matched begin External(%1,MethodObj) end;
You can also do the edits manually (they are trivial), but the patch is nice for automation.

 

Next, the idea is to call _OFF from _INI. So far, I see no reason to do it in DSDT.

 

into method label _INI parent_label \_SB.PCI0.RP05.PEGP code_regex . insert begin _OFF()\n end;
The resulting ssdt is ready to try (Save As AML, of course). Place it in /Extra/ssdt-1.aml or in EFI/CLOVER/patched/ssdt-1.aml depending on what bootloader you're using. Verify that it loads by looking at your bootlog.

 

The only question is whether you need to eliminate the loop waiting for LNKS. If it doesn't work, try this additional patch:

into method label _OFF parent_label \_SB.PCI0.RP05.PEGP code_regex While\s+\([^\}]*\} remove_matched;
Link to comment
Share on other sites

Here's something to try. First off, it is helpful to disassemble your ssdt using the dsdt...

iasl -e dsdt.aml -d ssdt.aml
If you have MaciASL installed, you can find iasl in /Applications/MaciASL.app/Contents/MacOS/iasl5.

 

Note that the same technique can be applied to disassembly of your DSDT:

iasl -e ssdt.aml -d dsdt.aml
Resulting in quite a few less errors to fix in your DSDT... But that is off-topic as you do not need to touch the DSDT for this.

 

 

After that, open ssdt.dsl. It will almost compile cleanly. You can fix it with this one patch:

into method label _BCM parent_label DD02 code_regex Return\s\((.*)\)\n\s+(.*) replace_matched begin Return(%1(%2)) end;
into definitionblock code_regex External\s\((.*\._BCM),\s+IntObj\) replace_matched begin External(%1,MethodObj) end;
You can also do the edits manually (they are trivial), but the patch is nice for automation.

 

Next, the idea is to call _OFF from _INI. So far, I see no reason to do it in DSDT.

 

into method label _INI parent_label \_SB.PCI0.RP05.PEGP code_regex . insert begin _OFF()\n end;
The resulting ssdt is ready to try (Save As AML, of course). Place it in /Extra/ssdt-1.aml or in EFI/CLOVER/patched/ssdt-1.aml depending on what bootloader you're using. Verify that it loads by looking at your bootlog.

 

The only question is whether you need to eliminate the loop waiting for LNKS. If it doesn't work, try this additional patch:

into method label _OFF parent_label \_SB.PCI0.RP05.PEGP code_regex While\s+\([^\}]*\} remove_matched;

 

Thanks for your advice! i have tried, but at first it worked, the graphic card was sucessfully disabled! but after several times rebooting, it became invalid, it's really weird! i can't figur out what's the reason... Do you probably know why? and how can i check if the method _OFF is called while the System inits?

Link to comment
Share on other sites

Thanks for your advice! i have tried, but at first it worked, the graphic card was sucessfully disabled! but after several times rebooting, it became invalid, it's really weird! i can't figur out what's the reason... Do you probably know why? and how can i check if the method _OFF is called while the System inits?

Does it always work on cold boot?

 

For debugging DSDT/SSDTs see here: https://github.com/RehabMan/OS-X-ACPI-Debug

Link to comment
Share on other sites

Originally you reported it working intermittently. Which is it? Not working at all, or working sometimes?

 

correctly, for example, last time when i installed the new OS X and loaded the DSDT SSDT, the graphic was suceessfully disabled. and after some configurations(System configuration, kexts installation and so on..) and several times reboots, it just doesn't work anymore.... :( so weird

Link to comment
Share on other sites

correctly, for example, last time when i installed the new OS X and loaded the DSDT SSDT, the graphic was suceessfully disabled. and after some configurations(System configuration, kexts installation and so on..) and several times reboots, it just doesn't work anymore.... :( so weird

Check your work. Verify SSDTs/DSDT as you expect is loaded in ioreg. Post ioreg if you want me to look.

 

Note that you need to be careful with DSDT patches that are not reflected in the SSDT. For example, it is common to rename GFX0 -> IGPU to enable GPU PM. You must be sure to do that to both the DSDT and the SSDT or, OS X will throw away the SSDT when the system loads, and the _INI method in SSDT will not be called (because the whole SSDT was thrown out for being invalid).

Link to comment
Share on other sites

Check your work. Verify SSDTs/DSDT as you expect is loaded in ioreg. Post ioreg if you want me to look.

 

Note that you need to be careful with DSDT patches that are not reflected in the SSDT. For example, it is common to rename GFX0 -> IGPU to enable GPU PM. You must be sure to do that to both the DSDT and the SSDT or, OS X will throw away the SSDT when the system loads, and the _INI method in SSDT will not be called (because the whole SSDT was thrown out for being invalid).

got it! i'll try it again, but if it's possible, could you help me to edit it and send to me, then i can diff the two versions and find out where the problem hides. really appreciate your help!

Link to comment
Share on other sites

Hey RehabMan, I recall you saying you got USB 3.0 working properly with your machine. Would you mind sharing that method? I'm hoping it will fix my USB resets after wakeup. Thanks! :)

I haven't tested it extensively, but I'm using GenericUSBXHCI with -gux_defer_usb2 flag. Mostly I used this because without it (eg. AppleUSBXHCI) my USB ports didn't work after wake from sleep.

got it! i'll try it again, but if it's possible, could you help me to edit it and send to me, then i can diff the two versions and find out where the problem hides. really appreciate your help!

Like I said, post ioreg, if you want me to look...

Link to comment
Share on other sites

I haven't tested it extensively, but I'm using GenericUSBXHCI with -gux_defer_usb2 flag. Mostly I used this because without it (eg. AppleUSBXHCI) my USB ports didn't work after wake from sleep.

 

Like I said, post ioreg, if you want me to look...

 

Really sorry that required you to do that, i will try it myself, but i'm not sure, how to check if the DSDT/SSDT was loaded or a methode was called by using IOReg?

Link to comment
Share on other sites

 Share

×
×
  • Create New...