Bug 104660 - DC R9 285 dual link DVI modes not shown.
Summary: DC R9 285 dual link DVI modes not shown.
Status: RESOLVED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/AMDgpu (show other bugs)
Version: DRI git
Hardware: Other All
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-01-16 18:18 UTC by Andy Furniss
Modified: 2018-01-29 23:46 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg with error (56.21 KB, text/plain)
2018-01-16 18:18 UTC, Andy Furniss
no flags Details
xrandr with missing > 60Hz modes (7.86 KB, text/plain)
2018-01-16 18:19 UTC, Andy Furniss
no flags Details
Fix for dual-link DVI modes (3.04 KB, patch)
2018-01-16 18:54 UTC, Harry Wentland
no flags Details | Splinter Review

Description Andy Furniss 2018-01-16 18:18:19 UTC
Created attachment 136779 [details]
dmesg with error

R9 285 Tonga, I see in agd5f 4.17-wip there are commits to do with DVI dual link. 

I was hoping "Don't block dual-link DVI modes" would enable 120Hz my monitor supports, which has always worked with without dc.

There is an error in dmesg - don't know if it's related. dmesg and xrandr attached.

[drm:dal_bios_parser_init_cmd_tbl [amdgpu]] *ERROR* Don't have set_dce_clock for v0
Comment 1 Andy Furniss 2018-01-16 18:19:22 UTC
Created attachment 136780 [details]
xrandr with missing > 60Hz modes
Comment 2 Harry Wentland 2018-01-16 18:54:14 UTC
Created attachment 136782 [details] [review]
Fix for dual-link DVI modes

The error is unrelated. This patch should fix your dual-link DVI modes. It's hitting the mailing list soon, just took a bit longer than the rest of the DVI patches. Note that we also see issues with non-native modes on one display with this patch, but that's likely a monitor-specific issue.
Comment 3 Andy Furniss 2018-01-16 19:27:47 UTC
Thanks, that works.

This is the same with non dc, but is there a reason they are a bit low?

  1920x1080 (0x5b) 285.500MHz +HSync -VSync
        h: width  1920 start 1968 end 2000 total 2080 skew    0 clock 137.26KHz
        v: height 1080 start 1083 end 1088 total 1144           clock 119.98Hz

 1920x1080 (0x5c) 260.250MHz +HSync -VSync
        h: width  1920 start 1968 end 2000 total 2080 skew    0 clock 125.12KHz
        v: height 1080 start 1083 end 1088 total 1138           clock 109.95Hz

 1920x1080 (0x5d) 235.500MHz +HSync -VSync
        h: width  1920 start 1968 end 2000 total 2080 skew    0 clock 113.22KHz
        v: height 1080 start 1083 end 1088 total 1133           clock  99.93Hz

I was hoping DC would differ from the old driver in giving exact Hz for these.
Comment 4 Alex Deucher 2018-01-16 19:31:16 UTC
(In reply to Andy Furniss from comment #3)
> Thanks, that works.
> 
> This is the same with non dc, but is there a reason they are a bit low?
> 
>   1920x1080 (0x5b) 285.500MHz +HSync -VSync
>         h: width  1920 start 1968 end 2000 total 2080 skew    0 clock
> 137.26KHz
>         v: height 1080 start 1083 end 1088 total 1144           clock
> 119.98Hz
> 
>  1920x1080 (0x5c) 260.250MHz +HSync -VSync
>         h: width  1920 start 1968 end 2000 total 2080 skew    0 clock
> 125.12KHz
>         v: height 1080 start 1083 end 1088 total 1138           clock
> 109.95Hz
> 
>  1920x1080 (0x5d) 235.500MHz +HSync -VSync
>         h: width  1920 start 1968 end 2000 total 2080 skew    0 clock
> 113.22KHz
>         v: height 1080 start 1083 end 1088 total 1133           clock 
> 99.93Hz
> 
> I was hoping DC would differ from the old driver in giving exact Hz for
> these.

Those are the modes from the EDID from your monitor:

CEA extension block
Extension version: 1
0 8-byte timing descriptors
Detailed mode: Clock 235.500 MHz, 521 mm x 293 mm
               1920 1968 2000 2080 hborder 0
               1080 1083 1088 1133 vborder 0
               +hsync -vsync 
Detailed mode: Clock 260.250 MHz, 521 mm x 293 mm
               1920 1968 2000 2080 hborder 0
               1080 1083 1088 1138 vborder 0
               +hsync -vsync 
Detailed mode: Clock 285.500 MHz, 521 mm x 293 mm
               1920 1968 2000 2080 hborder 0
               1080 1083 1088 1144 vborder 0
               +hsync -vsync
Comment 5 Andy Furniss 2018-01-16 19:38:03 UTC
Thanks, stupid monitor rather than some rounding then :-)
Comment 6 Andy Furniss 2018-01-29 23:46:59 UTC
Working with latest agd5f kernels


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.