netkas.org forum
September 24, 2017, 04:38:49 PM *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
News: Information for registering users http://forum.netkas.org/index.php/topic,2246.0.html
 
   Home   Help Search Login Register  
Pages: [1] 2 3 ... 23
  Print  
Author Topic: Radeon Port Mapping  (Read 120510 times)
Fl0r!an
Sr. Member
****
Offline Offline

Posts: 348


« on: April 22, 2015, 05:43:07 PM »

Hey Guys,

I'm quite new to MacPro VGA flashing, so I've got a few questions: Recently I upgraded the graphics card in my Hackintosh, so I had a spare HD 6870. Since my dad has some MacPros, I decided to flash it for him to replace one of his entry-level GPUs (GT120 and stuff like that). Thanks to this great forum, flashing was really simple, the card works fine, but a few questions have come to my mind:

1. Port Mapping in OSX
Before flashing the card, I put it in a MacPro and fired it up. It worked fine OOB in Yosemite (no EFI boot screen of course): Both monitors (hooked up to both DVI ports) were working and the Cinebench score looked fine, too. Everything was stable and smooth.
After flashing, I had an EFI boot screen on one of the ports (as it's supposed to be), but the 2nd DVI port was gone, even when OSX took over control. Beside that, everything stayed the same (e.g. Cinebench score).
From my understanding, this is because of the Duckweed framebuffer, which doesn't support this 2nd DVI port. I know that you could modify the Duckweed framebuffer to support that port (that's what some of the hackintosh guys do), but I want to keep the OSX install vanilla to prevent issues when updating.
I think that, before flashing, the "RadeonFramebuffer" personality was used (I didn't check it however because I didn't expect any problems after flashing). Since everything seemed to work fine unflashed (beside the lack of a boot screen), I thought that it might be possible to use "RadeonFramebuffer" with a flashed card?
Would that be as easy as decompressing the EFI, replacing all occurrences of "Duckweed" with "RadeonFramebuffer", compressing the EFI and building a ROM from it? I've looked through all Radeon EFIs I could find, but all of them were using a specific framebuffer personality, never the generic "RadeonFramebuffer", so I'm not sure if it's that simple...
As I've said, I don't have a MacPro to try that myself, so I'd have to visit my dad to mess around with one of his productive machines, which I don't want. It's not that important to get the 2nd DVI working (other ports work fine), but I'd like to know if my idea would work. I'd then try it later on his machine, when I'm sure I won't fuck things up.  Grin

2. Port Mapping in EFI
I know that the EFI boot screen only works on all ports if the flashed card uses the same design as the Apple one, where the EFI comes from; otherwise only some (or no) ports are working (e.g. only VGA on 5770 and only 1st DVI on my 6870). This port mapping is inside the EFI part of the ROM.
Since I've got a quite good programming background and stuff like a disassembler isn't foreign to me I wanted to take a look at what's going on in the EFI: I extracted and decompressed the EFI from different Mac ROMs (e.g. 5770 (EBC and EFI64), 6870, 7950) and threw it in IDA Pro. I found many familiar things (e.g. framebuffer personalities, device ID&name display types and even port names), but I did not find any kind of port mapping.
I expected to find a mapping which connects each instance of "ATY,EFIDisplay" to some index/address, which corresponds to a specific hardware port. Ideally this would be the same address as in the vBIOS (e.g. visible with radeon_bios_decode), so you would just have to make the EFI mapping fit the vBIOS mapping. Of course it's not that easy, I can't relate any of the "magic numbers" in the port setup part of the EFI (at least I think that's what it is) to any addresses or IDs in the corresponding vBIOS or framebuffers.
I couldn't find much information on this topic, I think the latest stuff was from 2012 or 2013, so I'd like to know if anyone has made any progress on modifying the EFI port mapping?
Again, it's not very important for me to have a boot screen on all ports, but it looks like a really interesting challenge to me, and I'd like to know if anyone has any hints.
Logged
Fl0r!an
Sr. Member
****
Offline Offline

Posts: 348


« Reply #1 on: April 24, 2015, 10:06:04 AM »

Hi, I've made a bit progress regarding the EFI port mapping. I'm not sure if it's something new, but most information on stuff like that is really wide scattered around the internet, so I'll try to collect something here.

So, I had a look at different EFIs (both EBC and EFi64) of which I know the reference layout. This is what I found in the EBC HD 5770 EFI:
https://www.picflash.org/picture.php?key=Z1JME7&action=show
Basically it's just a loop running from 0 to 2 (I'm quite sure this index is the port-number you can see in IOReg) and assigning a "connector-type" for each port-number. The connector-type has the same format as in the different Framebuffers, see Hackintosh-Sites. For the EBC-HD5770 (MacPro card) this is the result:
#0: connector-type = 0x400 (DisplayPort)
#1: connector-type = 0x400 (DisplayPort)
#2: connector-type = 0x4 (DVI)
As you can see, this perfectly matches the IOReg entry of an official HD5770 Mac Edition:
https://www.picflash.org/picture.php?key=OOPVW3&action=show
To make it fit a flashed HD5770 (1x DP, 2x DVI, 1x HDMI), you'd have to remove on of the 0x400 connectors, and add another 0x4 (DVI) and an 0x800 (HDMI). For the correct order of the port, I'd first try the order in the PC-BIOS. If this doesn't work, you could still try different combinations.
I don't think that this simple modification is enough to enable EFI Bootscreens on all ports, but at least it seems to be a start. I think you'd also have to match stuff like Encoder, Transmitter, SenseID and/or ControlFlags (didn't find any of these yet).

On the older HD4870, things are even more obvious: Since this card only has two ports (DP & DVI, according to MacTracker), they didn't bother doing a loop but simple copy&pasted the setup code a second time in there:
https://www.picflash.org/picture.php?key=KXE9BS&action=show
This is the result:
#0: connector-type = 0x400 (DisplayPort)
#1: connector-type = 0x4 (DVI)
Again, match!

On the HD5770 EFI64 (from iMac) you'll get:
#0: connector-type = 0x400 (DisplayPort)
#1: connector-type = 0x400 (DisplayPort)
This makes sense, too: One internal Displayport to connect the screen and another external one.

As I've said above, I don't think that this is enough to get bootscreens working, but it's a start. Reason: DVI-I ports only appear once (1 connector for both digital and VGA), but flashed HD5770 have the bootscreen only on VGA. I think I'll have a look at this aapl,boot-display and ATY,EFIDisplay stuff. As before, hints from some smart people are always welcome...  Wink
Ah, one question: I couldn't find any information where the HD6870-EFI comes from. Is it an iMac EFI or was there any Mac Edition of that Radeon generation?

Update: The flashed HD 6870 is connected via 1st DVI (the one which gives bootscreen). IOReg says on this port "ATY,EFIDisplay" = "HDMID", which is obviously wrong. Since bootscreen still works, this seems to be purely cosmetic...
Also found out that "appl,boot-display" is hard-coded to 0x1, so this shouldn't be an issue either. Will have a look at "ATY,EFIOrientation" and "ATY,EFIDispConfig" now (anyone knows something 'bout them?).
« Last Edit: April 24, 2015, 12:35:37 PM by Fl0r!an » Logged
Rominator
Hero Member
*****
Offline Offline

Posts: 2131



« Reply #2 on: April 24, 2015, 11:30:02 PM »

Great job !!

I've always said that the AMD roms could be improved, but nobody ever spends much time on it.

Maybe 666sheep or Netkas will chime in, they are the only other people who do real work on EFI that I know of.

Everyone else does copy & paste code kiddy tricks

Nice to see someone make a serious effort.
Logged

Before asking a question, check your "Personal Settings" and be sure that you have "Brain Services" set to "On".
Fl0r!an
Sr. Member
****
Offline Offline

Posts: 348


« Reply #3 on: April 25, 2015, 05:32:27 PM »

Thanks for your post!
Yeah, I was hoping to meet someone with more experience than me here, maybe they'll see this thread.

Today I played a bit around with the flashed HD 6870: I wanted to investigate how IOReg differs when booting up with a port that doesn't show boot screen. So I hooked the display up via HDMI and startet the MacPro. I was quite surprised to see the gray Apple  Shocked
And I was even more surprised when the display turned black when OSX took over.  Grin Seems my Sapphire HD 6870 has a non-standard layout, which doesn't show anything on HDMI using Duckweed, but instead shows boot screen on HDMI...  Smiley
So I had a look at what the "eBay copy & past kiddies" (  Wink ) are selling: Their 6870s have 2x mDP, 2x DVI, 1x HDMI, mine has 1x DP, 2x DVI, 1x HDMI.

Here's my BIOS port layout:
Code:
Connector at index 0
Type [@offset 44283]: DisplayPort (10)
Encoder [@offset 44287]: INTERNAL_UNIPHY2 (0x21)
i2cid [@offset 44393]: 0x90, OSX senseid: 0x1
Connector at index 1
Type [@offset 44293]: HDMI-A (11)
Encoder [@offset 44297]: INTERNAL_UNIPHY2 (0x21)
i2cid [@offset 44420]: 0x93, OSX senseid: 0x4
Connector at index 2
Type [@offset 44303]: DVI-D (3)
Encoder [@offset 44307]: INTERNAL_UNIPHY1 (0x20)
i2cid [@offset 44447]: 0x95, OSX senseid: 0x6
Connector at index 3
Type [@offset 44313]: DVI-I (2)
Encoder [@offset 44317]: INTERNAL_UNIPHY (0x1e)
i2cid [@offset 44484]: 0x94, OSX senseid: 0x5
Connector at index 4
Type [@offset 44323]: DVI-I (2)
Encoder [@offset 44327]: INTERNAL_KLDSCP_DAC1 (0x15)
i2cid [@offset 44484]: 0x94, OSX senseid: 0x5

Here's the BIOS port layout of a standard card:
Code:
Connector at index 0
Type [@offset 44824]: DisplayPort (10)
Encoder [@offset 44828]: INTERNAL_UNIPHY2 (0x21)
i2cid [@offset 44952]: 0x90, OSX senseid: 0x1
Connector at index 1
Type [@offset 44834]: DisplayPort (10)
Encoder [@offset 44838]: INTERNAL_UNIPHY2 (0x21)
i2cid [@offset 44979]: 0x91, OSX senseid: 0x2
Connector at index 2
Type [@offset 44844]: HDMI-A (11)
Encoder [@offset 44848]: INTERNAL_UNIPHY1 (0x20)
i2cid [@offset 45006]: 0x93, OSX senseid: 0x4
Connector at index 3
Type [@offset 44854]: DVI-D (3)
Encoder [@offset 44858]: INTERNAL_UNIPHY1 (0x20)
i2cid [@offset 45033]: 0x95, OSX senseid: 0x6
Connector at index 4
Type [@offset 44864]: DVI-I (2)
Encoder [@offset 44868]: INTERNAL_UNIPHY (0x1e)
i2cid [@offset 45070]: 0x94, OSX senseid: 0x5
Connector at index 5
Type [@offset 44874]: DVI-I (2)
Encoder [@offset 44878]: INTERNAL_KLDSCP_DAC1 (0x15)
i2cid [@offset 45070]: 0x94, OSX senseid: 0x5

Basically the difference concerning HDMI is the connector index (2 vs. 1) and the encoder (0x20 vs. 0x21). Will have a look at 6870.efi if I can find an explanation for the working HDMI boot screen there. (Investigating black HDMI in OS X is less interesting, this'll definitively be because of Duckweed).

Update: Didn't find an explanation yet: 6870 EFI seems to contain 4x DP and 1x 0x2 (SL-DVI???), so DVI+HDMI should NOT work (but it does for me)... Also found a switch case statement containing 0x90 - 0x95 (looks like i2cid, see above), but I can't make sense of that yet...
« Last Edit: April 26, 2015, 12:56:22 AM by Fl0r!an » Logged
Fl0r!an
Sr. Member
****
Offline Offline

Posts: 348


« Reply #4 on: April 27, 2015, 11:51:30 AM »

Hehe, have a look at this: https://www.picflash.org/viewer.php?img=MappingIFQOMM.png
Screenshot is a bit cropped and incomplete, complete mapping (OEM 7950 EFI btw) looks like this:

Code:
Name    internal ID  txmit   enc

DP1     0x100        0x12   0x4
DP2     0x200        0x22   0x5
TMDSA   0x1          0x30   0x0(?)
HDMIC   0x1080       0x12   0x4
HDMID   0x2080       0x22   0x5
HDMIE   0x4080       0x11   0x2

Now have a close look at redsock_bios_decode output for the same ROM file:
Code:
Connector Object Id [19] which is [DISPLAY_PORT]
encoder obj id [0x21] which is [INTERNAL_UNIPHY2 (osx txmit 0x12 [duallink 0x2] enc 0x4)] linkb: false
Connector Object Id [19] which is [DISPLAY_PORT]
encoder obj id [0x21] which is [INTERNAL_UNIPHY2 (osx txmit 0x22 [duallink 0x2] enc 0x5)] linkb: true
Connector Object Id [12] which is [HDMI_TYPE_A]
encoder obj id [0x20] which is [INTERNAL_UNIPHY1 (osx txmit 0x11 [duallink 0x1] enc 0x2)] linkb: false
Connector Object Id [2] which is [DVI_I]
encoder obj id [0x1e] which is [INTERNAL_UNIPHY (osx txmit 0x10 [duallink 0x0] enc 0x0)] linkb: false
Connector Object Id [2] which is [DVI_I]
encoder obj id [0x15] which is [INTERNAL_KLDSCP_DAC1 (osx txmit 0x00 enc 0x10?)] linkb: false

Everything except DVI-I port matches perfectly (maybe information for both digital and analog output is somehow encoded in the same port in EFI). Coincidence? Maybe, but not very likely I think...

Of course I can't be sure that these bytes in the EFI are really values for txmit and enc, but it looks quite obvious to me.
Btw, in the EFI this isn't really a table but more a somehow dynamic assignment scattered all over the EFI. Everything is linked by an identifier which I call internal ID (e.g. 0x100 for DP1). Why dynamic? The EFI contains information for 3 HDMI ports, but 2 of them are getting mapped to the same txmit & enc as DP1 and DP2. I'd assume that happens when using an DP->HDMI adapter.
All other OEM EFIs I had a look at contain a similar mapping: Again, everything matches except DVI.

Think I'll have to do some flashing now. Would be a lot easier if I had a MacPro at home  Grin
Hope I'll get around to do this this week. If s.o. wants to volunteer as tester, feel free to contact me  Smiley
« Last Edit: April 27, 2015, 02:28:56 PM by Fl0r!an » Logged
Rominator
Hero Member
*****
Offline Offline

Posts: 2131



« Reply #5 on: April 27, 2015, 10:05:12 PM »

Very interesting stuff you are finding.

Not sure if you knew, but the 6870 EFI was done by Netkas using an iMac EFI. That is why it is 64bit instead of EBC and why it may list some odd outputs in EFI.

One reason we have kind of given up on AMD is that there are likely to no longer be frame buffer personalities for future cards. They will likely only be created for cards in nMP or iMac so nothing with outputs that exist in Newegg choices.

So even if you write the perfect EFI, you will have to go back to "RadeonFramebuffer" or you will have to create or mod an existing frame buffer and update with each OS update.

Nvidia cards allow display outputs to be assigned in the rom and driver works around that, so much easier to end up with "perfect" result.

If you are able to make flashed 5770s and 5870s show output on all ports you will have a lot of fans.
Logged

Before asking a question, check your "Personal Settings" and be sure that you have "Brain Services" set to "On".
Fl0r!an
Sr. Member
****
Offline Offline

Posts: 348


« Reply #6 on: April 28, 2015, 08:59:59 AM »

What's wrong with RadeonFramebuffer? On my hackintosh I've been using it for quite some time without any problems (HD 5770, HD 6870, R9 280).
For me as a hackintosh guy, AMDs appear a lot easier at the moment, because you don't have to care about any web drivers which might/will get f*cked up on the next OS update.

I know that I'm a bit late with what I do (HD 5770/5870 has been around for about 5 years now and even the 7xxx/R9 EFI is 2 years old) and certainly the classic MacPros best time is over, but I'll still see what I can do. Maybe something useful comes up in the end, maybe not.  Smiley
Logged
Fl0r!an
Sr. Member
****
Offline Offline

Posts: 348


« Reply #7 on: April 28, 2015, 06:39:13 PM »

Today I played a bit around with my R9 280. I think it's a good card to start messing around because of its dual BIOS. It's a non-reference card (Sapphire Dual-X with 1x DP, 2x DVI, 1x HDMI), so the 2nd DVI port is not working.

The good news: My observations mentioned above (regarding txmit & enc) seem to be correct: For testing purpose, I've changed the txmit value of my (working) HDMI port in the EFI to an arbitrary value (e.g. 0x23, which does not appear in my graphic cards BIOS). As expected, this leads to a garbled screen on HDMI while other ports (e.g. DVI) remain fully functional.

The bad news: Above information is not enough to introduce a completely new port (e.g. 2nd DVI) to the system. I've tried to make the 2nd DVI work instead of 1st DVI by changing mentioned IDs, but this simply killed both DVIs. I think the main problem is that the 7xxx EFI only contains information for one DVI port, so it takes a lot more than just fixing some mismatched mappings to make both DVIs work.

This isn't too bad however: I think you could still enable some non-functional ports on other cards using this method as long as the EFI offers basic support for this port type. My problem is currently that I can't build a working ROM file for anything else than HD 7xxx/R9: I've tried using netkas' scripts from the 7xxx testing thread with my HD 6870, but it didn't work. Of course I changed the device ID and modified the efiromheader.rom (see attachment), but it didn't work. The Mac just completely ignored the EFI part in the rom (=>no boot screen, RadeonFramebuffer in OSX). Sadly I couldn't find any information on how to build a ROM file, so I guess I need your advice here...

Thanks in advance!  Smiley

* efiromheader.rom (44 KB - downloaded 235 times.)
Logged
blacksheep
Sr. Member
****
Offline Offline

Posts: 337


« Reply #8 on: April 28, 2015, 08:18:51 PM »

Hi!
I'm glad that someone with programming skills joined to the party Smiley I'm mostly trial and error guy.
Your findings are great. Thank you for sharing. This is something I have missed in my experiments, because I was looking for sense ID, which determines if port works at all (see radeon_bios_decode output).


The good news: My observations mentioned above (regarding txmit & enc) seem to be correct: For testing purpose, I've changed the txmit value of my (working) HDMI port in the EFI to an arbitrary value (e.g. 0x23, which does not appear in my graphic cards BIOS). As expected, this leads to a garbled screen on HDMI while other ports (e.g. DVI) remain fully functional.

The bad news: Above information is not enough to introduce a completely new port (e.g. 2nd DVI) to the system. I've tried to make the 2nd DVI work instead of 1st DVI by changing mentioned IDs, but this simply killed both DVIs. I think the main problem is that the 7xxx EFI only contains information for one DVI port, so it takes a lot more than just fixing some mismatched mappings to make both DVIs work.

Did you patch the framebuffer to match 2x DVI layout, or you tested only EFI loading stage?

Quote
This isn't too bad however: I think you could still enable some non-functional ports on other cards using this method as long as the EFI offers basic support for this port type. My problem is currently that I can't build a working ROM file for anything else than HD 7xxx/R9: I've tried using netkas' scripts from the 7xxx testing thread with my HD 6870, but it didn't work. Of course I changed the device ID and modified the efiromheader.rom (see attachment), but it didn't work. The Mac just completely ignored the EFI part in the rom (=>no boot screen, RadeonFramebuffer in OSX). Sadly I couldn't find any information on how to build a ROM file, so I guess I need your advice here...

Thanks in advance!  Smiley

After you put the EFI in right place inside the BIOS, you need to correct OpROM indicators and checksum of each part. Fixrom script will do that for you.
Logged
Fl0r!an
Sr. Member
****
Offline Offline

Posts: 348


« Reply #9 on: April 28, 2015, 09:51:28 PM »

Hi!
Nice to see that you've found this thread! I've seen some of your posts where you investigated similar stuff (regarding 6870 EBC EFI), so you should roughly know what I'm currently doing and might be able to help me out  Smiley

I was looking for sense ID, which determines if port works at all (see radeon_bios_decode output).
Yeah, was looking for it, too, but didn't find it so far. Not quite sure if it's needed at all at EFI-level though...

Did you patch the framebuffer to match 2x DVI layout, or you tested only EFI loading stage?
Only EFI stage, didn't care about OS X yet since framebuffer patching seems to be a quite well documented topic.

After you put the EFI in right place inside the BIOS, you need to correct OpROM indicators and checksum of each part. Fixrom script will do that for you.
Yeah I know fixrom, it's executed at the end of netkas' HD 7xxx script, so that was definitely done. If it's really that easy, the "makerom.sh" might simply have messed up some things in the 6870 rom; I'll try it manually next time. So copying the modified, compressed EFI in it's old place will just work (after applying fixrom.py)?. Is there anything I have to consider if the length of the EFI part changes (e.g. some header which contains the total length), or is this all done by fixrom?

Would be cool if you could test some stuff, too (assuming you have the correct hardware, e.g. HD 6870 or 5770/5870). I think I've written down most stuff I've figured out so far (or at least think I've figured out  Grin ), but of course I could also provide some EFI files for testing.
Logged
Bunga-Bunga
Newbie

Offline Offline

Posts: 47


« Reply #10 on: April 28, 2015, 11:28:55 PM »

I have a MP4.1 with HD5870 (flashed) - and i hope i get a MP1.1 this week.

So i would be glad to test some .roms; enough experience with bricked cards and automated/blind flashing under DOS  Wink
Logged
Fl0r!an
Sr. Member
****
Offline Offline

Posts: 348


« Reply #11 on: April 29, 2015, 09:57:47 AM »

That's nice! Okay, here we go:

I'll assume your HD 5870 looks like this (please verify with redsock_bios_decoder):
Code:
Connector Object Id [19] which is [DISPLAY_PORT]
encoder obj id [0x21] which is [INTERNAL_UNIPHY2 (osx txmit 0x12 [duallink 0x2] enc 0x4)] linkb: false
Connector Object Id [12] which is [HDMI_TYPE_A]
encoder obj id [0x21] which is [INTERNAL_UNIPHY2 (osx txmit 0x22 [duallink 0x2] enc 0x5)] linkb: true
Connector Object Id [2] which is [DVI_I]
encoder obj id [0x20] which is [INTERNAL_UNIPHY1 (osx txmit 0x11 [duallink 0x1] enc 0x2)] linkb: false
Connector Object Id [2] which is [DVI_I]
encoder obj id [0x16] which is [INTERNAL_KLDSCP_DAC2] linkb: false
Connector Object Id [2] which is [DVI_I]
encoder obj id [0x1e] which is [INTERNAL_UNIPHY (osx txmit 0x10 [duallink 0x0] enc 0x0)] linkb: false
Connector Object Id [2] which is [DVI_I]
encoder obj id [0x15] which is [INTERNAL_KLDSCP_DAC1 (osx txmit 0x00 enc 0x10?)] linkb: false

I've made the following changes (12 bytes changed in total):
- DP1: 0x11/0x2 => 0x12/0x4
- HDMI: 0x21/0x3 => 0x22/0x5 (the EFI doesn't really list a discrete HDMI port but only something that looks like support for DP=>HDMI adapter, maybe this will still change something...)
- DVI: 0x12/0x4 => 0x11/0x2 (only support for one single DVI port found)

I'd try DisplayPort first as things look most obvious here. I've attached both a compressed EFI file with EFI header and an uncompressed EFI file (without header). Normally you should be able to build a .rom by appending the compressed EFI (5870mod.efi.rom)to your PC BIOS (see HD 5xxx/6xxx guide here). If the EFI is not recognized by the system (e.g. RadeonFrame buffer and name "HD 5000 series") then I've done something wrong when I've attached the compressed EFI to the EFI header. In this case, someone with more experience than me should try to build a ROM from the uncompressed EFI file (5870mod.efi).

EDIT: Whoops, uploded wrong file. I've also added another version where I've flipped the mapping for DP and HDMI. Please test this, too, if the original version doesn't change anything.

* 5870mod.efi (108.5 KB - downloaded 250 times.)
* 5870mod2.efi.rom (42 KB - downloaded 301 times.)
* 5870mod2.efi (108.5 KB - downloaded 281 times.)
* 5870mod.efi.rom (42 KB - downloaded 315 times.)
« Last Edit: April 29, 2015, 11:45:40 AM by Fl0r!an » Logged
Bunga-Bunga
Newbie

Offline Offline

Posts: 47


« Reply #12 on: April 29, 2015, 10:31:28 AM »

I'll test it in about few hours!  Wink

Got a MP1.1 today, so i have a second machine for testing.
Logged
Bunga-Bunga
Newbie

Offline Offline

Posts: 47


« Reply #13 on: April 29, 2015, 07:43:52 PM »

Both files are working fine. I have a XFX 5870, i guess it is reference design.

Perhaps i am the first one, who saw a bootscreen on DVI-port with flashed pc-card - very, very good work!  Shocked

Unfortunately i can't find a DP-cable for testing, sorry.
Both roms, same result, bootscreen on the second DVI-port. First DVI-port (lower one) only via VGA-adapter or blackscreen until login.
Logged
Rominator
Hero Member
*****
Offline Offline

Posts: 2131



« Reply #14 on: April 29, 2015, 08:05:02 PM »

Excellent work !

I have a 5870 here and will test.

Btw, there is a problem with most 5870 DP ports such that they can run a Dell DP port but not an Apple LED Cinema Display.

I have both and will let you know.
Logged

Before asking a question, check your "Personal Settings" and be sure that you have "Brain Services" set to "On".
Pages: [1] 2 3 ... 23
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
SMFAds for Free Forums
Valid XHTML 1.0! Valid CSS!