4K at 60 YCbCr420 missing mode in usermode

Tue Jun 26 18:14:43 UTC 2018

Hi Mike:

It is my bad that I need to check DCE version > 112 to add this flag on. Since I am testing on Vega10, I didn’t realize that. Anything like Carrizo, Fiji doesn’t apply to the ycbcr420_only patch.

I’ll get the patch updated soon.

Regard,
Jerry

From: Mike Lothian [mailto:mike at fireburn.co.uk]
Sent: June 26, 2018 12:41 PM
To: Michel Dänzer <michel at daenzer.net>
Cc: Emil Velikov <emil.l.velikov at gmail.com>; Zuo, Jerry <Jerry.Zuo at amd.com>; xorg-devel at lists.x.org; Lipski, Mikita <Mikita.Lipski at amd.com>; Wentland, Harry <Harry.Wentland at amd.com>
Subject: Re: 4K at 60 YCbCr420 missing mode in usermode

With AMDGPU DDX I'm not seeing 4k at 60Hz with this patch and allowing >8bpc

With modesetting X doesn't start and I get the following in dmesg:

[  105.397875] ------------[ cut here ]------------
[  105.397877] kernel BUG at drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:4796!
[  105.397882] invalid opcode: 0000 [#1] PREEMPT SMP NOPTI
[  105.397884] CPU: 0 PID: 334 Comm: Xorg Tainted: G        W         4.18.0-rc1-agd5f+ #187
[  105.397885] Hardware name: System manufacturer System Product Name/ROG STRIX X470-I GAMING, BIOS 0601 04/19/2018
[  105.397890] RIP: 0010:dm_update_crtcs_state+0x45b/0x4e0
[  105.397890] Code: ff ff 48 85 db 0f 84 ea fe ff ff 48 c7 44 24 18 00 00 00 00 48 c7 44 24 08 00 00 00 00 48 c7 04 24 00 00 00 00 e9 34 fe ff ff <0f> 0b 48 83 c4 30 b8 ea ff ff ff 5b 5d 41 5c 41 5d 41 5e 41 5f c3
[  105.397908] RSP: 0018:ffffc9000063b8f8 EFLAGS: 00010246
[  105.397910] RAX: ffff880819942001 RBX: ffff88081b317000 RCX: 0000000000000000
[  105.397911] RDX: 0000000000007690 RSI: 0000000000007690 RDI: ffff880819943800
[  105.397911] RBP: ffff880805bd0a80 R08: 0000000000022ac0 R09: ffffffff815bf6f4
[  105.397912] R10: ffffea0020665000 R11: ffff88083e806e80 R12: 0000000000000000
[  105.397913] R13: ffff880819942400 R14: ffff880819940400 R15: ffff88081b32e800
[  105.397914] FS:  00007f0374519f80(0000) GS:ffff88083ec00000(0000) knlGS:0000000000000000
[  105.397914] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  105.397915] CR2: 0000561ac61e0e10 CR3: 000000081acc6000 CR4: 00000000003406b0
[  105.397916] Call Trace:
[  105.397920]  ? amdgpu_dm_atomic_check+0x1b7/0x3d0
[  105.397922]  ? _raw_spin_lock_irqsave+0x12/0x40
[  105.397923]  ? _raw_spin_unlock_irqrestore+0xf/0x30
[  105.397926]  ? drm_atomic_check_only+0x44d/0x640
[  105.397928]  ? drm_atomic_commit+0xe/0x50
[  105.397930]  ? restore_fbdev_mode_atomic+0x1b9/0x1d0
[  105.397932]  ? drm_fb_helper_restore_fbdev_mode_unlocked+0x40/0x90
[  105.397933]  ? drm_fb_helper_set_par+0x24/0x50
[  105.397936]  ? fb_set_var+0x20f/0x3e0
[  105.397938]  ? sugov_should_update_freq+0x42/0x60
[  105.397940]  ? sugov_update_single+0x6e/0x1f0
[  105.397942]  ? fbcon_blank+0x246/0x320
[  105.397944]  ? try_to_wake_up+0x1f1/0x370
[  105.397948]  ? do_unblank_screen+0x96/0x150
[  105.397949]  ? vt_ioctl+0x35e/0x10f0
[  105.397952]  ? path_parentat.isra.63+0x40/0x80
[  105.397954]  ? tty_ioctl+0x217/0x890
[  105.397956]  ? do_vfs_ioctl+0x9f/0x610
[  105.397957]  ? dput.part.36+0x9a/0x120
[  105.397959]  ? __fput+0x11a/0x1e0
[  105.397960]  ? ksys_ioctl+0x35/0x70
[  105.397962]  ? __x64_sys_ioctl+0x11/0x20
[  105.397963]  ? do_syscall_64+0x43/0xf0
[  105.397965]  ? entry_SYSCALL_64_after_hwframe+0x44/0xa9
[  105.397966] Modules linked in:
[  105.397967] ---[ end trace 82b2adbb705155c8 ]---

On Tue, 26 Jun 2018 at 17:27 Mike Lothian <mike at fireburn.co.uk<mailto:mike at fireburn.co.uk>> wrote:
Hi

I'm happy to test this when I get home

I'm currently reverting that patch that enables >8bit colour to get 60Hz back

Cheers

Mike

On Tue, 26 Jun 2018 at 17:23 Michel Dänzer <michel at daenzer.net<mailto:michel at daenzer.net>> wrote:
On 2018-06-26 05:43 PM, Emil Velikov wrote:
> Hi Jerry,
>
> On 25 June 2018 at 22:45, Zuo, Jerry <Jerry.Zuo at amd.com<mailto:Jerry.Zuo at amd.com>> wrote:
>> Hello all:
>>
>>
>>
>> We are working on an issue affecting 4K at 60 HDMI display not to light up, but
>> only showing up 4K at 30 from:
>> https://bugs.freedesktop.org/show_bug.cgi?id=106959 and others.
>>
>>
>>
>> Some displays (e.g., ASUS PA328) HDMI port shows YCbCr420 CEA extension
>> block with 4K at 60 supported. Such HDMI 4K at 60 is not real HDMI 2.0, but still
>> following HDMI 1.4 spec. with maximum TMDS clock of 300MHz instead of
>> 600MHz.
>>
>> To get such 4K at 60 supported, it needs to limit the bandwidth by reducing the
>> color space to YCbCr420 only. We’ve already raised YCbCr420 only flag
>> (attached patch) from kernel side to pass the mode validation, and expose it
>> to user space.
>>
>>
>>
>>     We think that one of the issues that causes this problem is due to
>> usermode pruning the 4K at 60 mode from the modelist (attached Xorg.0.log). It
>> seems like when usermode receives all the modes, it doesn't take in account
>> the 4K at 60 YCbCr4:2:0 specific mode. In order to pass validation of being
>> added to usermode modelist, its pixel clk needs to be divided by 2 so that
>> it won't exceed TMDS max physical pixel clk (300MHz). That might explain the
>> difference in modes between our usermode and modeset.
>>
>>
>>
>>     Such YCbCr4:2:0 4K at 60 special mode is marked in DRM by raising a flag
>> (y420_vdb_modes) inside connector's display_info which can be seen in
>> do_y420vdb_modes(). Usermode could rely on that flag to pick up such mode
>> and halve the required pclk to prevent such mode getting pruned out.
>>
>>
>>
>> We were hoping for someone helps to look at it from usermode perspective.
>> Thanks a lot.
>>
> Just some observations, while going through some coffee. Take them
> with a pinch of salt.
>
> Currently the kernel edid parser (in drm core) handles the
> EXT_VIDEO_DATA_BLOCK_420 extended block.
> Additionally, the kernel allows such modes only as the (per connector)
> ycbcr_420_allowed bool is set by the driver.
>
> Quick look shows that it's only enabled by i915 on gen10 && geminilake hardware.
>
> At the same time, X does it's own fairly partial edid parsing and
> doesn't handle any(?) extended blocks.
>
> One solution is to update the X parser, although it seems like a
> endless game of cat and mouse.
> IMHO a much better approach is to not use edid codepaths for KMS
> drivers (where AMDGPU is one).
> On those, the supported modes is advertised by the kernel module via
> drmModeGetConnector.

We are getting the modes from the kernel; the issue is they are then
pruned (presumably by xf86ProbeOutputModes => xf86ValidateModesClocks)
due to violating the clock limits, as described by Jerry above.

--
Earthling Michel Dänzer               |               http://www.amd.com
Libre software enthusiast             |             Mesa and X developer
_______________________________________________
xorg-devel at lists.x.org<mailto:xorg-devel at lists.x.org>: X.Org development
Archives: http://lists.x.org/archives/xorg-devel
Info: https://lists.x.org/mailman/listinfo/xorg-devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.x.org/archives/xorg-devel/attachments/20180626/de67aed4/attachment-0001.html>