Contribute
Register

Screen distorted on boot HD5500, UEFI native (no CSM), 1920x1080 HP 450 G2

Status
Not open for further replies.
Joined
May 14, 2013
Messages
38
Motherboard
I didn't read
CPU
the RULES
Graphics
so ignore me
EDIT: Look like it's not related to your problem. Can you take a screenshot of garbled screen?

So you have to set 1024x768 in config.plist or enable CSM to fix the garbled screen?

Yes exactly, both workarounds do not cause the garbled screen. Here is a screenshot:
IMG_20150923_164655.jpg

Another workaround is going to standby and wake-up :cool:

/edit: Actually a simple screen-off / on fixes it as well...
 
Joined
May 14, 2013
Messages
38
Motherboard
I didn't read
CPU
the RULES
Graphics
so ignore me
It's the untouched CLOVER from the HP post installer (Version 6.4.2 (September 14th 2015)), except that I changed to OsxAptioFixDrv-64.efi per your advice.
View attachment 152097

Let me know if you want me to re-upload that.



Jep I understand. Let me double check fast boot.

Disabling fast boot does _not_ solve the issue.
 
Joined
Oct 21, 2012
Messages
3,464
Motherboard
HP Elitebook 8470p
CPU
i5 3320M
Graphics
HD 4000, 1600x900
Mobile Phone
  1. iOS
Yes exactly, both workarounds do not cause the garbled screen. Here is a screenshot:
View attachment 152131

Another workaround is going to standby and wake-up :cool:

/edit: Actually a simple screen-off / on fixes it as well...
Same problem on Dell Broadwell. If you have to set 1024x768 or CSM enabled, use "Yosemite S" theme instead.
 

RehabMan

Moderator
Joined
May 2, 2012
Messages
183,951
Motherboard
Intel DH67BL
CPU
i7-2600K
Graphics
HD 3000
Mac
  1. MacBook Air
Mobile Phone
  1. iOS
Yes exactly, both workarounds do not cause the garbled screen. Here is a screenshot:
View attachment 152131

Another workaround is going to standby and wake-up :cool:

/edit: Actually a simple screen-off / on fixes it as well...

Experiment with a couple of settings in the PNLF patch.

Code:
                    Name (LMAX, 0xad9)
                    // KMAX: defines the unscaled range in the _BCL table below
                    Name (KMAX, 0xad9)
                    // _INI deals with differences between native setting and desired
                    Method (_INI, 0, NotSerialized)
                    {
                        // This 0xC value comes from looking what OS X initializes this
                        // register to after display sleep (using ACPIDebug/ACPIPoller)
                        Store(0xC0000000, LEVW)

Try with LMAX=0.

Try removing the Store(0xC*, LEVW) line.

It could be additional debugging of the IGPU registers is required (I originally figured out the 0xC0000000 for Haswell using ACPIDebug).
 

RehabMan

Moderator
Joined
May 2, 2012
Messages
183,951
Motherboard
Intel DH67BL
CPU
i7-2600K
Graphics
HD 3000
Mac
  1. MacBook Air
Mobile Phone
  1. iOS
Same problem on Dell Broadwell. If you have to set 1024x768 or CSM enabled, use "Yosemite S" theme instead.

Seems like a reasonable work around as long as these settings are available.

Unless someone wants to determine the true cause with ACPIDebug...
 
Joined
May 11, 2015
Messages
261
Motherboard
Dell 7348-Clover
CPU
i7-5500U
Graphics
HD 5500
Mac
  1. MacBook Pro
Mobile Phone
  1. Android
Seems like a reasonable work around as long as these settings are available.

Unless someone wants to determine the true cause with ACPIDebug...


I can help..
you can give me the steps to do ACPIDebug
or can release remote access chrome remote desktop and then posted the experience here
 

RehabMan

Moderator
Joined
May 2, 2012
Messages
183,951
Motherboard
Intel DH67BL
CPU
i7-2600K
Graphics
HD 3000
Mac
  1. MacBook Air
Mobile Phone
  1. iOS
I can help..
you can give me the steps to do ACPIDebug
or can release remote access chrome remote desktop and then posted the experience here

I used ACPIPoller.kext to run an ACPI method every second. You probably already have such a method in DSDT for the fan patches (eg. SMCD.TCPU).

From that method I dumped various IGPU backlight registers with ACPIDebug, so that I see the values in system.log each second. Then... you sleep the display, wake it, and you'll see if there is a related difference in the registers that might be useful in solving the problem. The registers can then be fixed in the PNLF._INI method, just as I did with the 0xC0000000 value for LEVW.
 
Joined
May 14, 2013
Messages
38
Motherboard
I didn't read
CPU
the RULES
Graphics
so ignore me
Experiment with a couple of settings in the PNLF patch.

Code:
                    Name (LMAX, 0xad9)
                    // KMAX: defines the unscaled range in the _BCL table below
                    Name (KMAX, 0xad9)
                    // _INI deals with differences between native setting and desired
                    Method (_INI, 0, NotSerialized)
                    {
                        // This 0xC value comes from looking what OS X initializes this
                        // register to after display sleep (using ACPIDebug/ACPIPoller)
                        Store(0xC0000000, LEVW)

Try with LMAX=0.

Try removing the Store(0xC*, LEVW) line.

It could be additional debugging of the IGPU registers is required (I originally figured out the 0xC0000000 for Haswell using ACPIDebug).

I tried both by editing the DSDT.aml (in CLOVER/ACPI/patched) with MaciASL, but no luck yet. I guess we'd need to debug that - it's not a major issue actually, but I might be willing to try.
 

RehabMan

Moderator
Joined
May 2, 2012
Messages
183,951
Motherboard
Intel DH67BL
CPU
i7-2600K
Graphics
HD 3000
Mac
  1. MacBook Air
Mobile Phone
  1. iOS
I tried both by editing the DSDT.aml (in CLOVER/ACPI/patched) with MaciASL, but no luck yet. I guess we'd need to debug that - it's not a major issue actually, but I might be willing to try.

Are you certain you did it correctly? You would have difficulty editing the DSDT.aml directly as it won't compile clean (need to edit the DSDT.dsl in the same directory, Save, then Save As DSDT.aml, format ACPI Machine Language Binary). Something for you to double check and clarify.
 
Status
Not open for further replies.
Top