1 |
john <jdm@××××××××××××.uk> wrote: |
2 |
|
3 |
> On Mon, 28 Dec 2015 13:41:57 +0000 |
4 |
> Mick <michaelkintzios@×××××.com> wrote: |
5 |
> |
6 |
> > On Monday 28 Dec 2015 09:11:10 john wrote: |
7 |
> > > Hello, |
8 |
> > > |
9 |
> > > I have installed Gentoo on to another partition in order to use |
10 |
> > > radeon drivers for desktop instead of fglrx. |
11 |
> > > |
12 |
> > > When I start my resolution is very small (1280x600) and only uses |
13 |
> > > half the screen. My max resolution is 1920*1080 so I have a big |
14 |
> > > black border around my screen. I can startx which works but cannot |
15 |
> > > change the resolution to anything bigger. GRUB splash uses full |
16 |
> > > screen but when the kernel loads it switches to small resolution. |
17 |
> > > |
18 |
> > > The only thing I can think of which is causing this issue is UEFI. |
19 |
> > > In my other set (fglrx) up I have simple framebuffer enabled which |
20 |
> > > works when I set up resolution in grub default. |
21 |
> > > |
22 |
> > > Strangely enough when I start system rescue cd in UEFI mode I get |
23 |
> > > the same problem with small resolution but not if I do not boot in |
24 |
> > > UEFI mode. |
25 |
> > |
26 |
> > Hmm ... I don't get such problems here on 3 PCs running different |
27 |
> > amd/radeon cards and all with radeon drivers. |
28 |
> > |
29 |
> > You haven't said which card you're running. If it is a very recent |
30 |
> > model the radeon drivers may not have caught up with it yet? |
31 |
> > |
32 |
> > If you need to compare kernel settings let me know, although there |
33 |
> > are wiki pages that are quite detailed on this topic. |
34 |
> > |
35 |
> > |
36 |
> > > Anyone come across this or got this to work properly with |
37 |
> > > radeon/UEFI? |
38 |
> > > |
39 |
> > > Moral, don't use UEFI I guess but that's the future???????!!!!!! |
40 |
> > > |
41 |
> > > John |
42 |
> > |
43 |
> > On PC boots in UEFI (directly boots the efi binary, no boot |
44 |
> > managers) and it has no such problems on 2 x 1920*1080 monitors. |
45 |
> > |
46 |
> |
47 |
> Hmm, |
48 |
> thanks, I have now tried an old monitor and its screen |
49 |
> resolution was good (ie 1920x1080) so I think it's the monitor. |
50 |
> I have a R9 280 radeon card which is pretty new but the monitor is |
51 |
> very new (display port only, acer xb240h model) so I guess it's |
52 |
> something do do with that, perhaps EDID but not really sure. The |
53 |
> whole graphics stack is pretty bemusing and I would love to make |
54 |
> sense of it all. |
55 |
|
56 |
Several months ago I was thinking about buying a LG monitor with |
57 |
Cinema4k (4096x2160) resolution. The monitor that I received had some |
58 |
firmware bugs. One bug was that the monitor shows a broken (somewhat |
59 |
incomplete) EDID block. So it was not possible to use the full |
60 |
resolution with the radeon driver but only UHD resolution (3840x2160). |
61 |
|
62 |
I read somewhere that this monitor maybe would run fine with the fglrx |
63 |
driver. But because I don't like proprietary software and because the |
64 |
monitor also had some other bugs and IMHO also a poor picture quality, |
65 |
I decided to sent it back. |
66 |
|
67 |
There is a kernel option DRM_LOAD_EDID_FIRMWARE. It allows you to |
68 |
specify an EDID data set instead of probing for it. If your problem |
69 |
is caused by broken EDID data, this option maybe will help you to run |
70 |
the monitor at its full resolution. |
71 |
|
72 |
-- |
73 |
Regards |
74 |
wabe |