563 replies to this topic

[crazybirdy]

kizwan, on Mar 22 2011, 03:52 PM, said:

I don't know. Probably. Can you use original kernel?

I have the same problem, get ./RadeonDump: Bad CPU type in executable.
But "sudo dmesg > ~/Desktop/Encoder.txt"  is ok to run. (enter password when asked.)

[checcco]

@atlee I'm not sure I've understood... With patched10.7  Uakari you get your LVDS working? Have you tried booting with every other monitor disconnected? Did you leave stock Transmitter and Encoder and just changed senseid?

[atlee]

checcco, on Mar 22 2011, 09:18 PM, said:

@atlee I'm not sure I've understood... With patched10.7  Uakari you get your LVDS working? Have you tried booting with every other monitor disconnected? Did you leave stock Transmitter and Encoder and just changed senseid?

10.6.6 Uakari

00 04 00 00 00 04 00 00 00 01 00 00 12 04 04 01
04 00 00 00 14 00 00 00 00 01 00 00 01 12 01 03
00 02 00 00 14 00 00 00 00 01 00 00 00 00 06 05
00 08 00 00 00 02 00 00 00 01 00 00 22 05 05 04

10.7 Uakari

00 04 00 00 00 04 00 00 00 71 00 00 12 04 04 01
04 00 00 00 14 00 00 00 00 71 00 00 01 12 01 03
00 02 00 00 14 00 00 00 00 71 00 00 00 00 06 05
00 08 00 00 00 02 00 00 00 71 00 00 22 05 05 04

10.7 Uakari - Patched

00 04 00 00 00 04 00 00 00 71 00 00 12 04 04 00 - DP not sure if works
02 00 00 00 14 00 00 00 00 71 00 00 11 02 01 07 - shows as CRT but works, change to 40 and it will not work properly, you will lose extended and it will force cloning also.
10 00 00 00 10 00 00 00 00 71 00 00 00 00 06 08 - shows as CRT, working as extended, can also do mirror if i wish.
00 08 00 00 00 02 00 00 00 71 00 00 20 01 05 01 - Still changing around, combos 22,05/02,04/12,04/21,03/20,01 do not work.

I have LVDS as CRT, cannot get it to change it's name no matter what i do. VGA as CRT. DP cannot test. Might try swapping HDMI to either 01,02/11,02 as they are the last combos left, don't have access to that monitor as my wife is still awake

And yes i have tested booting without the VGA plugged in, LVDS switches res and turns on. VGA switch on and off when plugging in and out so the behavior seems ok to me.

If you change the LVDS T/E sometimes you lose the Extended desktop and it will force cloning, So i have kept it on something that works. default also works. but i was playing around to see what i could do.

Don't know if i'm right but i'm enjoying the results i'm getting now anyway.

Also note i do not really consider DP port of any use so i will not intend to play around and fix it, as it is not a requirement for me.

[wmarsh]

kizwan, on Mar 22 2011, 12:56 AM, said:

jvilla have HD4670. So, only DIG1 & DIG2 are applicable. DIG5 is only for HD 5XXX Series and above.

That makes sense -- I did not know his hardware

CAUTION ON 10.6.7 UPGRADE

Has anyone gotten LVDS output on 10.6.7 yet?

I applied same patch that worked on 10.6.6 and have a blank screen.

Fortunately I took the precaution of doing the upgrade as a test on a USB drive clone.

Now, I need to recheck what I did. And I need to reset Terminal so I can get ioreg out of a black box. So maybe I am wrong.

But I do note the rather unusual fact that the new ATIFramebuffer is dated 2 days ago -- usually things are dated back when the previews went out -- and I am wondering if they did something to break this method.

[EDIT -- I updated one desktop to check. lipo -detailed_info and bcc9's ati_personality tool give identical output. So all offsets should be identical in 10.6.7]

[atlee]

wmarsh, on Mar 22 2011, 10:23 PM, said:

That makes sense -- I did not know his hardware

CAUTION ON 10.6.7 UPGRADE

But I do note the rather unusual fact that the new ATIFramebuffer is dated 2 days ago -- usually things are dated back when the previews went out -- and I am wondering if they did something to break this method.

Yes mwarsh there is definitely a new Framebuffer v1.6.26.31 i think and you could be right maybe they did something to break it because I have looked a few FB's with no changes, haven't bothered testing 10.6.7 files though as i am getting more success with 10.7.

If anyone needs the 10.6.6 files, i keep a backup of all

[jvilla]

atlee, on Mar 22 2011, 05:28 AM, said:

Yes mwarsh there is definitely a new Framebuffer v1.6.26.31 i think and you could be right maybe they did something to break it because I have looked a few FB's with no changes, haven't bothered testing 10.6.7 files though as i am getting more success with 10.7.

If anyone needs the 10.6.6 files, i keep a backup of all

atlee how do you get your cpu to work without a legacy kernel?  i have the same cpu but i need legacy kernel to boot...

jvilla, on Mar 22 2011, 09:26 AM, said:

atlee how do you get your cpu to work without a legacy kernel?  i have the same cpu but i need legacy kernel to boot...

  i managed to boot with mach_kernel arch=i386.  when i run ./RadeonDump > myencoder.txt i get this error still:  error: IORegistryEntryGetPath returned 0x10000003

Do i need to install something to get ioregistry working?

[kizwan]

jvilla, on Mar 22 2011, 11:53 PM, said:

i managed to boot with mach_kernel arch=i386.  when i run ./RadeonDump > myencoder.txt i get this error still:  error: IORegistryEntryGetPath returned 0x10000003

Do i need to install something to get ioregistry working?

I suspect either RadeonHD.kext is not installed or not loaded.

hjs89, on Mar 17 2011, 08:35 AM, said:

And the radeon's log:

RHDAtomOutputAllocFree
Mapping DIG1 encoder to KLDSKP_UNIPHYE
Crtc[0]: found native mode from Monitor[LVDS Panel]: 
Modeline "1366x768"  72  1366 1414 1446 1486  768 771 775 806
Connector "PANEL" uses Monitor "LVDS Panel":
RHDValidateScaledToMode
rhdModeValidateCrtc
DxModeValid: ATOM CRTC 1
rhdAtomOutputModeValid
Listing modesetting layout:
ATOM CRTC 1: tied to Atom PLL 1 and LUT A:
	Outputs: AtomOutputtainCountUniphyE (PANEL)
ATOM CRTC 2: unused
	Unused Outputs: AtomOutputDACA, AtomOutputUniphyA

Can you post RadeonHD's log with external monitor connected to HDMI?

[mojtabazarei]

atlee, on Mar 22 2011, 03:25 PM, said:

10.6.6 Uakari

00 04 00 00 00 04 00 00 00 01 00 00 12 04 04 01
04 00 00 00 14 00 00 00 00 01 00 00 01 12 01 03
00 02 00 00 14 00 00 00 00 01 00 00 00 00 06 05
00 08 00 00 00 02 00 00 00 01 00 00 22 05 05 04

10.7 Uakari

00 04 00 00 00 04 00 00 00 71 00 00 12 04 04 01
04 00 00 00 14 00 00 00 00 71 00 00 01 12 01 03
00 02 00 00 14 00 00 00 00 71 00 00 00 00 06 05
00 08 00 00 00 02 00 00 00 71 00 00 22 05 05 04

10.7 Uakari - Patched

00 04 00 00 00 04 00 00 00 71 00 00 12 04 04 00 - DP not sure if works
02 00 00 00 14 00 00 00 00 71 00 00 11 02 01 07 - shows as CRT but works, change to 40 and it will not work properly, you will lose extended and it will force cloning also.
10 00 00 00 10 00 00 00 00 71 00 00 00 00 06 08 - shows as CRT, working as extended, can also do mirror if i wish.
00 08 00 00 00 02 00 00 00 71 00 00 20 01 05 01 - Still changing around, combos 22,05/02,04/12,04/21,03/20,01 do not work.

I have LVDS as CRT, cannot get it to change it's name no matter what i do. VGA as CRT. DP cannot test. Might try swapping HDMI to either 01,02/11,02 as they are the last combos left, don't have access to that monitor as my wife is still awake

And yes i have tested booting without the VGA plugged in, LVDS switches res and turns on. VGA switch on and off when plugging in and out so the behavior seems ok to me.

If you change the LVDS T/E sometimes you lose the Extended desktop and it will force cloning, So i have kept it on something that works. default also works. but i was playing around to see what i could do.

Don't know if i'm right but i'm enjoying the results i'm getting now anyway.

Also note i do not really consider DP port of any use so i will not intend to play around and fix it, as it is not a requirement for me.

Hi i did this crt work fully LVDS is recognized as CRT but black screen should install switch res to get it?

did you use all 4 kexts from lion or 3 of them?

[hjs89]

kizwan, on Mar 22 2011, 05:35 PM, said:

Can you post RadeonHD's log with external monitor connected to HDMI?

 ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
Validating Modes from Monitor "Philips 190CW" on "DVI-D 1"
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
[color="#FF0000"][b]Output->ModeValid failed[/b][/color]
Rejected mode "1440x1440" (1440x1440:175Mhz): Unknown status.
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
[color="#FF0000"][b]Output->ModeValid failed[/b][/color]
Rejected mode "1440x1440" (1440x1440:223Mhz): Unknown status.
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 1
rhdModeValidateCrtc
FUNCTION: DxFBValid: ATOM CRTC 2
DxModeValid: ATOM CRTC 2
rhdAtomOutputModeValid
RHDGetVirtualFromModesAndFilter
FUNCTION: DxFBValid: ATOM CRTC 1
FUNCTION: DxFBValid: ATOM CRTC 2
Using 105x92 DPI.
Using 1440x900 Framebuffer with 1472 pitch
Free FB offset 0x00000000 (size = 0x10000000)
Display resolutions detected: 
1440 X 900 @ 59Hz
1152 X 864 @ 75Hz
1024 X 768 @ 75Hz
832 X 624 @ 74Hz
800 X 600 @ 75Hz
640 X 480 @ 75Hz
720 X 400 @ 70Hz

:S It was conected hdmi->dvi cable to my external display (I know that you said that It's better to work on VGA first, but I don't have a VGA monitor close to test)

When I loaded the RadeonHD there wasn't any signal in HDMI display. The LVDS the screen was corrupted (like when I boot without hdmi connected) but it was detected like the HDMI display (Philips 190CW).

With the last RadeonHD HDMI was "working" as a mirror screen and the colors were incorrect (like happens to me when I use Lion's fb in the LVDS screen) and only the HDMI display is recognized. The LVDS works but with HDMI's display resolution.

[jvilla]

do i need to remove any kexts from S/L/E to make radeonhd.kext work?  and do i need to add graphicsenabler=no to boot options?

[hjs89]

jvilla, on Mar 22 2011, 07:47 PM, said:

do i need to remove any kexts from S/L/E to make radeonhd.kext work?  and do i need to add graphicsenabler=no to boot options?

You've to boot with GraphicsEnabler=Yes and usually you need to remove ATI's kext from S/L/E. With some versions they conflict one each other.

[jvilla]

Heh.  I had to set graphicsenabler=no and remove atiframebuffer.kext and ati4600controller.kext from s/le.

Kizwan,  if you have the time could you take a look at my radeon encoder.txt and fix the framebuffer for me?  i tried hex editing but it still over my head.

Attached Files

•  Encoder.txt   13.7K   14 downloads

[atlee]

jvilla, on Mar 23 2011, 02:53 AM, said:

atlee how do you get your cpu to work without a legacy kernel?  i have the same cpu but i need legacy kernel to boot...

Hi Jvilla,

You have two options:
legacy_kernel (Busratio=12) 4 cores, 8 threads
mach_kernel (cpus=1 busratio=12)
or meklort's RC5 branch has the apic cpu panic fixed inside of the code so it patches mach_kernel before it loads.
this will allow you to run mach_kernel but it's really a modified mach_kernel

@ mojtabazarei, I did have to VNC initially to setup the extended desktop properly but after that everything worked.
example lvds was not at a recognised reslution so it didn't show properly but after changing that switched on, then changed monitor res and that made everything good.

It's still probably not a perfect fix but i have two outputs now, both working at two different resolutions both mirror and extended so i'm happy at this stage. Better then using Eulemur anyway.

[checcco]

atlee, on Mar 22 2011, 11:24 PM, said:

Hi Jvilla,

You have two options:
legacy_kernel (Busratio=12) 4 cores, 8 threads
mach_kernel (cpus=1 busratio=12)
or meklort's RC5 branch has the apic cpu panic fixed inside of the code so it patches mach_kernel before it loads.
this will allow you to run mach_kernel but it's really a modified mach_kernel

@ mojtabazarei, I did have to VNC initially to setup the extended desktop properly but after that everything worked.
example lvds was not at a recognised reslution so it didn't show properly but after changing that switched on, then changed monitor res and that made everything good.

It's still probably not a perfect fix but i have two outputs now, both working at two different resolutions both mirror and extended so i'm happy at this stage. Better then using Eulemur anyway.

atlee does dvd player works? I remember in my first tests dvd player didn't work with 10.7 kexts but only with 10.6.6 ones

[atlee]

checcco, on Mar 23 2011, 09:49 AM, said:

atlee does dvd player works? I remember in my first tests dvd player didn't work with 10.7 kexts but only with 10.6.6 ones

QE/CI, DVD Player, Frontrow...yeah seems good. I'm guessing maybe it might be the new ATIRadeonX3000 from 10.6.7. Because i remember testing 10.7 Uakari and Vervet and they gave me no QE/CI. But now I have it again.

ATI5000Controller is 10.7
ATIFramebuffer is 10.7
ATISupport is 10.7
ATIRadeonX3000 is v1.6.26.31 stock new from 10.6.7 upgrade.

I haven't tried to patch 10.6.7 files, i'm happy with the extended and mirroring, qe/ci under Uakari. I can work on swapping T/E until HDMI is working but during my uni work i will eventually get around to it today

[kizwan]

hjs89, on Mar 23 2011, 03:04 AM, said:

:S It was conected hdmi->dvi cable to my external display (I know that you said that It's better to work on VGA first, but I don't have a VGA monitor close to test)

When I loaded the RadeonHD there wasn't any signal in HDMI display. The LVDS the screen was corrupted (like when I boot without hdmi connected) but it was detected like the HDMI display (Philips 190CW).

With the last RadeonHD HDMI was "working" as a mirror screen and the colors were incorrect (like happens to me when I use Lion's fb in the LVDS screen) and only the HDMI display is recognized. The LVDS works but with HDMI's display resolution.

You're using HDMI to DVI converter to connect to the external monitor? In that case, you can ignore my request earlier if you don't have monitor with HDMI port. I was looking for "pure" HDMI connection between notebook & external monitor.

Even though there is no signal to external monitor & LVDS showed corrupted screen, RadeonHD should be able to detect the external monitor & LVDS links.

[kizwan]

jvilla, on Mar 23 2011, 04:44 AM, said:

Heh.  I had to set graphicsenabler=no and remove atiframebuffer.kext and ati4600controller.kext from s/le.

Kizwan,  if you have the time could you take a look at my radeon encoder.txt and fix the framebuffer for me?  i tried hex editing but it still over my head.

Your log file look strange. It doesn't contains EDID from the internal monitor (LVDS) & external monitor (VGA).

________________________________________________________________________


There are 3 people with 3 different Dell notebook & all 3 have Full HD internal LCD:-
almuras - Dell Studio 1747 with Radeon Mobility 4650
sydlix - Dell Studio XPS 1640 with Mobility 4670
jvilla - Dell Studio XPS 1645 with Mobility 4670

So, the LVDS link should use DUAL_LINK for transmitter. This link for LVDS should work theoretically:-
CRTC0 > DIG1 > UNIPHY_EF > LVDS: Encoder=0x00 Transmitter=0x02
This should be test without external monitor connected to HDMI & DisplayPort. Even though the LVDS link (transmitter & encoder) is correct but HDMI's/DP's is not, interference can happen if LVDS & HDMI/DP links are conflicting with each other which can make the LVDS seems doesn't work.

[jvilla]

kizwan, on Mar 22 2011, 06:43 PM, said:

Your log file look strange. It doesn't contains EDID from the internal monitor (LVDS) & external monitor (VGA).

________________________________________________________________________


There are 3 people with 3 different Dell notebook & all 3 have Full HD internal LCD:-
almuras - Dell Studio 1747 with Radeon Mobility 4650
sydlix - Dell Studio XPS 1640 with Mobility 4670
jvilla - Dell Studio XPS 1645 with Mobility 4670

So, the LVDS link should use DUAL_LINK for transmitter. This link for LVDS should work theoretically:-
CRTC0 > DIG1 > UNIPHY_EF > LVDS: Encoder=0x00 Transmitter=0x02
This should be test without external monitor connected to HDMI & DisplayPort. Even though the LVDS link (transmitter & encoder) is correct but HDMI's/DP's is not, interference can happen if LVDS & HDMI/DP links are conflicting with each other which can make the LVDS seems doesn't work.

i i tried again with debug to true and verbose level 2.

Attached Files

•  Encoder.debug.true.verbose.2.txt   30.31K   8 downloads

[kizwan]

jvilla, on Mar 23 2011, 09:12 AM, said:

i i tried again with debug to true and verbose level 2.

Do you have external monitor which have HDMI port?

[hjs89]

kizwan, on Mar 22 2011, 11:53 PM, said:

You're using HDMI to DVI converter to connect to the external monitor? In that case, you can ignore my request earlier if you don't have monitor with HDMI port. I was looking for "pure" HDMI connection between notebook & external monitor.

OK. If you want I can repeat the test after. My parents have a TV with HDMI and I can test it again but I'm not going to be in their house for a least in a month

kizwan, on Mar 22 2011, 11:53 PM, said:

Even though there is no signal to external monitor & LVDS showed corrupted screen, RadeonHD should be able to detect the external monitor & LVDS links.

I think that RadeonHD has never detected well any external monitor in my laptop.