[Guide] Airport - PCIe Half Mini v2

[jaymonkey]

Are you unloading the WiFi kext across sleep (with something like SleepWatcher)?

No

If the kext was unloaded and reloaded upon wake, that would explain the behavior you're seeing (Clover patches only work on kexts/cache that Clover loads).

Maybe you need to rebuild cache...

Was the first thing i tried (multiple times) ......

Like i said, everything was fine before 10.10.3 .....

Jay

 

[RehabMan]

No



Was the first thing i tried (multiple times) ......

Like i said, everything was fine before 10.10.3 .....

Jay

Run IORegistryExplorer prior to sleep. Leave it running, then sleep. After wake verify that the WiFi entries are not going away and coming back (they are shown in red when they go away, and new devices are shown in green).

 

[jaymonkey]

Run IORegistryExplorer prior to sleep. Leave it running, then sleep. After wake verify that the WiFi entries are not going away and coming back (they are shown in red when they go away, and new devices are shown in green).

@RM,

Interesting all the Airport entries look to be the same before and after sleep, however upon wake it looks like the Bluetooth side of the card is reset ... i wonder if that could cause the whole card to reset causing the stack to reset ...


As you can see the 'RemoteWake' key value has changed but thats the only thing i think .... apart from the link speed, country code ... etc but thats to be expected obviously.

I wonder is this is a firmware issue with the AW-CE123H as this is the same firmware version that i was using with BTFirmWareUploader before switching to BrcmPatchRAM ..... although different pkg source i assume ?

Before and After IOREG attached .. maybe you can spot something, I'm not up to speed on using BrcmPatchRAM & FakePCI so maybe i missed something in my setup ?

Cheers
Jay

 

[RehabMan]

@RM,

Interesting all the Airport entries look to be the same before and after sleep, however upon wake it looks like the Bluetooth side of the card is reset ... i wonder if that could cause the whole card to reset causing the stack to reset ...

That is normal. Bluetooth has nothing to do with WiFi. WiFi is on PCIe, and BT is on USB.

I wonder is this is a firmware issue with the AW-CE123H as this is the same firmware version that i was using with BTFirmWareUploader before switching to BrcmPatchRAM ..... although different pkg source i assume ?

Same comment as above. Separate devices on separate bus using separate drivers.

View attachment 133945
As you can see the 'RemoteWake' key value has changed but thats the only thing i think .... apart from the link speed, country code ... etc but thats to be expected obviously.

I notice you're doing a bunch of DSDT injections. It is not necessary (and I'm not using any). All you need is FakePCIID kexts.

I also find it interesting that the IO80211CountryCode changed from US to XT...


Off-topic, but thought I'd mention it: You should probably disable Radeon for power savings.

 

[jaymonkey]

I notice you're doing a bunch of DSDT injections. It is not necessary (and I'm not using any). All you need is FakePCIID kexts.

I also find it interesting that the IO80211CountryCode changed from US to XT...

Off-topic, but thought I'd mention it: You should probably disable Radeon for power savings.

@RM,

Thanks for taking a look, I suspect the injections are coming from clover as i got FixAPT DSDT patch checked .. will knock them out and see how things are:

Question: Is it necessary to inject Fake ID 0x43a014E4 for WiFi when using FakePCI kexts ?



Like i said , am new to FakePCI kexts .....

Good Spot with that pesky Radion GPU .. that was disabled ..... I switched from using DSDT code to SSDT a few weeks ago and everything was working, i wonder why it's not now ?, will look into that ....thanks.

Cheers
Jay

 

[RehabMan]

...
Question: Is it necessary to inject Fake ID 0x43a014E4 for WiFi when using FakePCI kexts ?

No. The IOKitPersonalities required are already in the FakePCIID_BCM*.kext.

Good Spot with that pesky Radion GPU .. that was disabled ..... I switched from using DSDT code to SSDT a few weeks ago and everything was working, i wonder why it's not now ?, will look into that ....thanks.

Probably because of the rather unconventional order you are loading your SSDTs. From your bootlog:

6:798  0:000  Inserting SSDT.aml from EFI\CLOVER\ACPI\patched ... Success
6:799  0:000  Inserting SSDT-0.aml from EFI\CLOVER\ACPI\patched ... Success
6:799  0:000  Inserting SSDT-4.aml from EFI\CLOVER\ACPI\patched ... Success
6:803  0:003  Inserting SSDT-5.aml from EFI\CLOVER\ACPI\patched ... Success
6:806  0:002  Inserting SSDT-1.aml from EFI\CLOVER\ACPI\patched ... Success
6:807  0:000  Inserting SSDT-2.aml from EFI\CLOVER\ACPI\patched ... Success
6:807  0:000  Inserting SSDT-3.aml from EFI\CLOVER\ACPI\patched ... Success

 

[robi42]

I hace the Azurware AW-CE123H (BCM94352) card, when I apply the DSDT patch it comes up with error name already exists in scope (_DSM), anyone has an idea how to fix that, see my DSDT.dsl and DSDT.aml file and the patch below,


# Maintained by: toleda for: github.com/toleda/wifi
# File: arpt3-bcm4352_v2.0.txt
# Credit: PikeRAlpha
#
# Broadcom BCM4352 WiFi dsdt edits
# Verify IOReg/pci14e4 device @1C,3/RP04, otherwise, edit patch file:
# Find: RP04
# Replace All: name (name = IOReg/pci14e4 device name, RP01, BR01, PEX1, etc.)
#
# If no PXSX, manually add Device (ARPT) after Name (_ADR, 0x001C000x) to dsdt and
# edit patch file:
# Find: PXSX
# Replace All: ARPT
# dsdt edit looks like:
# Device (name)
# {
# Name (_ADR, 0x001C000x)
# Device (ARPT)
#
# 1. Insert device ARPT dsdt injection
into device Label PXSX parent_label RP04 replace_content begin
Name (_ADR, Zero)\n
Name (_SUN, One)\n
Name (_PRW, Package (0x02) {0x09,0x04})\n
Method (_DSM, 4, NotSerialized)\n
{\n
If (LEqual (Arg2, Zero)) { Return (Buffer() { 0x03 } ) }\n
Return (Package()\n
{\n
"AAPL,slot-name", "AirPort",\n
"built-in", Buffer () {0x00},\n
"device_type", "AirPort",\n
"model", "Broadcom BCM4352 802.11 a/b/g/n/ac Wireless Network Controller",\n
"name", "AirPort Extreme",\n
"compatible","pci14e4,43a0"\n
})\n
}\n
end;
into definitionblock code_regex . insert
begin Store ("arpt3-bcm4352_v2.0 dsdt edits, github.com/toleda", Debug) end

 

[Al-Muftah]

I could not make my card Atheros AR5B195 working
anyone help me pleassss in details, I am really new to this

 

[RehabMan]

I hace the Azurware AW-CE123H (BCM94352) card, when I apply the DSDT patch it comes up with error name already exists in scope (_DSM), anyone has an idea how to fix that, see my DSDT.dsl and DSDT.aml file and the patch below,


# Maintained by: toleda for: github.com/toleda/wifi
# File: arpt3-bcm4352_v2.0.txt
# Credit: PikeRAlpha
#
# Broadcom BCM4352 WiFi dsdt edits
# Verify IOReg/pci14e4 device @1C,3/RP04, otherwise, edit patch file:
# Find: RP04
# Replace All: name (name = IOReg/pci14e4 device name, RP01, BR01, PEX1, etc.)
#
# If no PXSX, manually add Device (ARPT) after Name (_ADR, 0x001C000x) to dsdt and
# edit patch file:
# Find: PXSX
# Replace All: ARPT
# dsdt edit looks like:
# Device (name)
# {
# Name (_ADR, 0x001C000x)
# Device (ARPT)
#
# 1. Insert device ARPT dsdt injection
into device Label PXSX parent_label RP04 replace_content begin
Name (_ADR, Zero)\n
Name (_SUN, One)\n
Name (_PRW, Package (0x02) {0x09,0x04})\n
Method (_DSM, 4, NotSerialized)\n
{\n
If (LEqual (Arg2, Zero)) { Return (Buffer() { 0x03 } ) }\n
Return (Package()\n
{\n
"AAPL,slot-name", "AirPort",\n
"built-in", Buffer () {0x00},\n
"device_type", "AirPort",\n
"model", "Broadcom BCM4352 802.11 a/b/g/n/ac Wireless Network Controller",\n
"name", "AirPort Extreme",\n
"compatible","pci14e4,43a0"\n
})\n
}\n
end;
into definitionblock code_regex . insert
begin Store ("arpt3-bcm4352_v2.0 dsdt edits, github.com/toleda", Debug) end

Remove the existing (OEM) _DSM.

into method label _DSM parent_label _SB.PCI0.RP04.PXSX remove_entry;
# optionally, remove the now empty Scope
into scope label _SB.PCI0.RP04.PXSX remove_entry;

Note: You do not need DSDT patches. Simply using FakePCIID.kext + FakePCIID_BCM*.kext will work.

 

[jaymonkey]

Probably because of the rather unconventional order you are loading your SSDTs. From your bootlog:

6:798  0:000  Inserting SSDT.aml from EFI\CLOVER\ACPI\patched ... Success
6:799  0:000  Inserting SSDT-0.aml from EFI\CLOVER\ACPI\patched ... Success
6:799  0:000  Inserting SSDT-4.aml from EFI\CLOVER\ACPI\patched ... Success
6:803  0:003  Inserting SSDT-5.aml from EFI\CLOVER\ACPI\patched ... Success
6:806  0:002  Inserting SSDT-1.aml from EFI\CLOVER\ACPI\patched ... Success
6:807  0:000  Inserting SSDT-2.aml from EFI\CLOVER\ACPI\patched ... Success
6:807  0:000  Inserting SSDT-3.aml from EFI\CLOVER\ACPI\patched ... Success

@RM,

Not sure whats going there ... how on earth did you find that log from the two IOREG files i sent ?

Looking at my Clover config.plist it seems that the 'SortedOrder'' key and array has disappeared ... have put back in place and SSDT load order is now ok but ATI GPU still showing up ...

6:924 0:000 Inserting table[0]:SSDT.aml from EFI\CLOVER\ACPI\patched ... Success
6:925 0:000 Inserting table[1]:SSDT-0.aml from EFI\CLOVER\ACPI\patched ... Success
6:927 0:001 Inserting table[2]:SSDT-1.aml from EFI\CLOVER\ACPI\patched ... Success
6:927 0:000 Inserting table[3]:SSDT-2.aml from EFI\CLOVER\ACPI\patched ... Success
6:927 0:000 Inserting table[4]:SSDT-3.aml from EFI\CLOVER\ACPI\patched ... Success
6:928 0:000 Inserting table[8]:SSDT-7.aml from EFI\CLOVER\ACPI\patched ... Success
6:928 0:000 Inserting table[9]:SSDT-8.aml from EFI\CLOVER\ACPI\patched ... Success
6:929 0:000 CPUBase=0 and ApicCPUBase=1 ApicCPUNum=8
6:929 0:000 EdidDiscovered size=0
6:929 0:000 ATI injection not set
6:929 0:000 Intel Intel HD Graphics 3000 [8086:0126] :: PcieRoot(0x0)\Pci(0x2,0x0)
6:929 0:000 Intel GFX revision =0x9
6:930 0:000 stringlength = 566
6:930 0:000 CurrentMode: Width=1600 Height=900

Note: that i keep original SSDT names now rather than renumbering/renaming them ..

Will continue to work on problem ...

Cheers
Jay