Lenovo T460s - El Capitan

[tluck]

guys - here is a fakesmc v6.21.1412 with some minor tweaks cpu identification for skyLake.

if you get a chance - please run the xrandr app from linux/ubutu to see if your display is eDP or LVDS. i would be grateful and its good to know if thats the difference on why my system is not running full gfx. it wedges on the IntelAccelerator. the SKLGraphicsFramebuffer kext loads but then the AppleIntelSKLGraphics.kext wedges when it tries to load/run.

$ kextstat |grep -i graph

78 10 0xffffff7f80e38000 0x3b000 0x3b000 com.apple.iokit.IOGraphicsFamily (2.4.1) A360453D-2050-3C49-A549-AC0DD5E87917 <12 7 5 4 3 1>

95 2 0xffffff7f82a77000 0x7000 0x7000 com.apple.AppleGraphicsDeviceControl (3.12.8) 81C2784E-285A-38A7-A16E-515DCB816E0A <7 6 5 4 3 1>

96 0 0xffffff7f82a7e000 0x63000 0x63000 com.apple.driver.AppleIntelSKLGraphicsFramebuffer (10.1.4) AB1AAEE9-2484-3A2B-A6FF-2B449CF938D7 <95 94 78 15 12 11 7 6 5 4 3 1>

 

[loppol]

guys - here is a fakesmc v6.21.1412 with some minor tweaks cpu identification for skyLake.

if you get a chance - please run the xrandr app from linux/ubutu to see if your display is eDP or LVDS. i would be grateful and its good to know if thats the difference on why my system is not running full gfx. it wedges on the IntelAccelerator. the SKLGraphicsFramebuffer kext loads but then the AppleIntelSKLGraphics.kext wedges when it tries to load/run.

$ kextstat |grep -i graph

78 10 0xffffff7f80e38000 0x3b000 0x3b000 com.apple.iokit.IOGraphicsFamily (2.4.1) A360453D-2050-3C49-A549-AC0DD5E87917 <12 7 5 4 3 1>

95 2 0xffffff7f82a77000 0x7000 0x7000 com.apple.AppleGraphicsDeviceControl (3.12.8) 81C2784E-285A-38A7-A16E-515DCB816E0A <7 6 5 4 3 1>

96 0 0xffffff7f82a7e000 0x63000 0x63000 com.apple.driver.AppleIntelSKLGraphicsFramebuffer (10.1.4) AB1AAEE9-2484-3A2B-A6FF-2B449CF938D7 <95 94 78 15 12 11 7 6 5 4 3 1>

here is mine:

kextstat |grep -i graph
   68   11 0xffffff7f80d40000 0x3b000    0x3b000    com.apple.iokit.IOGraphicsFamily (2.4.1) 172C2960-EDF5-382D-80A5-C13E97D74880 <12 7 5 4 3 1>
   70    2 0xffffff7f82a47000 0x7000     0x7000     com.apple.AppleGraphicsDeviceControl (3.12.8) 81C2784E-285A-38A7-A16E-515DCB816E0A <7 6 5 4 3 1>
   71    0 0xffffff7f82a4e000 0x63000    0x63000    com.apple.driver.AppleIntelSKLGraphicsFramebuffer (10.1.4) 17B4A4AC-D3E4-36B4-907C-98034096277E <70 69 68 15 12 11 7 6 5 4 3 1>
   99    0 0xffffff7f82ad3000 0xb7000    0xb7000    com.apple.driver.AppleIntelSKLGraphics (10.1.4) 2105CAB9-A7BE-3456-8422-F10D189DAF24 <98 69 68 12 7 6 5 4 3 1>

Thanks for your sharing !
By the way, I believe nowadays most of laptop's display panel is not using the lvds signal, I can verify it by looking at the panel model # from Aida and check it to see if it's a edp signal on cite panellook。

 

[loppol]

guys - here is a fakesmc v6.21.1412 with some minor tweaks cpu identification for skyLake.

if you get a chance - please run the xrandr app from linux/ubutu to see if your display is eDP or LVDS. i would be grateful and its good to know if thats the difference on why my system is not running full gfx. it wedges on the IntelAccelerator. the SKLGraphicsFramebuffer kext loads but then the AppleIntelSKLGraphics.kext wedges when it tries to load/run.

$ kextstat |grep -i graph

78 10 0xffffff7f80e38000 0x3b000 0x3b000 com.apple.iokit.IOGraphicsFamily (2.4.1) A360453D-2050-3C49-A549-AC0DD5E87917 <12 7 5 4 3 1>

95 2 0xffffff7f82a77000 0x7000 0x7000 com.apple.AppleGraphicsDeviceControl (3.12.8) 81C2784E-285A-38A7-A16E-515DCB816E0A <7 6 5 4 3 1>

96 0 0xffffff7f82a7e000 0x63000 0x63000 com.apple.driver.AppleIntelSKLGraphicsFramebuffer (10.1.4) AB1AAEE9-2484-3A2B-A6FF-2B449CF938D7 <95 94 78 15 12 11 7 6 5 4 3 1>

i just confirmed that my display uses edp signal interface.
http://www.panelook.com/LP140WF6-SPB1_LG Display_14.0_LCM_overview_27176.html

 

[tluck]

ok good to know. interesting and encouraging!

 

[nabisco]

if you get a chance - please run the xrandr app from linux/ubutu to see if your display is eDP or LVDS. i would be grateful and its good to know if thats the difference on why my system is not running full gfx. it wedges on the IntelAccelerator. the SKLGraphicsFramebuffer kext loads but then the AppleIntelSKLGraphics.kext wedges when it tries to load/run.

$ kextstat |grep -i graph

Here's mine:

makrolon:workdir bookian$ kextstat |grep -i graph
   38   11 0xffffff7f81701000 0x47000    0x47000    com.apple.iokit.IOGraphicsFamily (2.4.1) A360453D-2050-3C49-A549-AC0DD5E87917 <12 7 5 4 3 1>
   87    2 0xffffff7f826e3000 0x7000     0x7000     com.apple.AppleGraphicsDeviceControl (3.12.8) 81C2784E-285A-38A7-A16E-515DCB816E0A <7 6 5 4 3 1>
   88    0 0xffffff7f826ea000 0x63000    0x63000    com.apple.driver.AppleIntelSKLGraphicsFramebuffer (10.1.4) AB1AAEE9-2484-3A2B-A6FF-2B449CF938D7 <87 86 38 15 12 11 7 6 5 4 3 1>
  103    0 0xffffff7f8276f000 0xb7000    0xb7000    com.apple.driver.AppleIntelSKLGraphics (10.1.4) 6925A1F6-0381-30C9-9AC8-B0197206D546 <102 86 38 12 7 6 5 4 3 1>

BTW, do you guys have intermittent issue of 2 bcm20702A0 appearing after wake?

 

[tluck]

thanks for verifying your gfx. still no full gfx for me. i even tried your dsdt and config - which is very similar my non-s version.

BT - well yes i did - had that exact problem - it was driving me frickin crazy - worked sometimes and then not.

... but i just figured out what is/was issue. there is a race condition (conflict) as the WiFi starts back up it messes with the BT FW reload. so my workaround/fix is the use the "sleep enabler" tool to turn off wifi when it is going to sleep and then turn it on again after. this seems to let the FW get loaded without crashing. along the way, i got some advice on a making well defined SSDT-XHC too for the USB ports.


$ cat /etc/rc.sleep
#!/bin/bash

# turn off both networks
ifconfig en0 down
networksetup -setairportpower en1 off

------------------------------------------
$ cat /etc/rc.wakeup
#!/bin/bash

# en0 comes up automatically
# turn on WiFi after sleep
sleep 5
networksetup -setairportpower en1 on

 

[tluck]

and if you can capture ioreg -l and post maybe i can spot something different?

 

[nabisco]

and if you can capture ioreg -l and post maybe i can spot something different?

Here's my ioreg -l output.
I'll try the rc.sleep/wake when I get the chance later.
Thanks for the tip!

 

[loppol]

thanks for verifying your gfx. still no full gfx for me. i even tried your dsdt and config - which is very similar my non-s version.

BT - well yes i did - had that exact problem - it was driving me frickin crazy - worked sometimes and then not.

... but i just figured out what is/was issue. there is a race condition (conflict) as the WiFi starts back up it messes with the BT FW reload. so my workaround/fix is the use the "sleep enabler" tool to turn off wifi when it is going to sleep and then turn it on again after. this seems to let the FW get loaded without crashing. along the way, i got some advice on a making well defined SSDT-XHC too for the USB ports.


$ cat /etc/rc.sleep
#!/bin/bash

# turn off both networks
ifconfig en0 down
networksetup -setairportpower en1 off

------------------------------------------
$ cat /etc/rc.wakeup
#!/bin/bash

# en0 comes up automatically
# turn on WiFi after sleep
sleep 5
networksetup -setairportpower en1 on

Hi truck,thanks for you tips!
once set en0 to down then en0 is still on inactive mode after wake up and i also tried commend 'ifconfig en0 up' but it didn't work.
then I tried only by changing networksetup

# for sleep
#!/bin/bash
# get current wifi device
CURRENT_DEVICE=$(networksetup -listallhardwareports | awk '$3=="Wi-Fi" {getline; print $2}')
# turn off wifi
networksetup -setairportpower $CURRENT_DEVICE off


# for wake
#!/bin/bash
CURRENT_DEVICE=$(networksetup -listnetworkserviceorder | grep -o 'Wi-Fi.*' | grep -o 'en[[:digit:]]')
sleep 2
networksetup -setairportpower $CURRENT_DEVICE on

now it seems working without a problem

 

[tluck]

nice. yeah good idea to not hard code the devices as in my barebones script.

nabisco - i was bringing down en0 (ethernet) too for the heck of. my wifi (not BT) is on en1. but i will update my script to determine the devices! Note: with the combo card, the wifi shows ups as a PCI device and BT shows up as a USB