Intel-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Daniel Vetter <daniel@ffwll.ch>
To: Alfonso Fiore <alfonso.fiore@gmail.com>
Cc: Intel Graphics Development <intel-gfx@lists.freedesktop.org>
Subject: Re: Sandy Bridge Desktop - 1920x1080i interlace not working
Date: Thu, 19 Jan 2012 21:50:39 +0100	[thread overview]
Message-ID: <20120119205039.GK3779@phenom.ffwll.local> (raw)
In-Reply-To: <CAO=xpBHXYVOXYbDJaDxd7Z-r0nUWToYSpFr8eKQ5XEGw6iUtxg@mail.gmail.com>

On Thu, Jan 19, 2012 at 08:56:44PM +0100, Alfonso Fiore wrote:
> On Thu, Jan 19, 2012 at 4:37 PM, Daniel Vetter <daniel@ffwll.ch> wrote:
> >
> > Yeah, unfortunately Peter's interlaced patches have issues still on
> > Sandybridge. Meanwhile you could try out all the modes the kernel detected
> > from the edid with xrandr (also see whether picking a different refresh
> > rate with --rate changes anything). If any of them work correctly (instead
> > of the cut-in-half or tiny screen), that might be interesting to know.
> 
> Hi,
> 
> more tests with the drm-intel-fixes and no "nomoreset".
> 
> I boot (as usual) just with HDMI cable.
> 
> The PC boots in that weird way (I sent photos today) and then when I
> login the screen is split.
> 
> Interestingly, trying to connect the same screen over VGA (while
> running and already connected over HDMI) makes X hang BUT it also
> changes the way X shows: full screen, no flicker (i.e. as it should).
> 
> when I restart X (sudo restart lightdm) I have a FANTASTIC PERFECT SCREEN! :) :)

Ok, please make this _really_ precise, we need to know exactly when things
work and when not. I presume this was showing a perfect screen over the
HDMI cable?

Also please recheck for the split-screen issue whether that is some custom
mode still lingering around by pasting xrandr -q output. For the broken
case (i.e. only HDMI connected) also please try to change the resolution
to something else like 1024x768 and see what happens.

> $ xrandr -q
> Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> VGA1 connected 1024x768+0+0 (normal left inverted right x axis y axis)
> 400mm x 300mm
>    1024x768       60.0*
>    800x600        60.3     56.2
>    848x480        60.0
>    640x480        59.9
> HDMI1 disconnected (normal left inverted right x axis y axis)
> DP1 disconnected (normal left inverted right x axis y axis)
> HDMI2 connected 1024x768+0+0 (normal left inverted right x axis y
> axis) 640mm x 360mm
>    1280x720       50.0     60.0
>    1024x768       60.0*
>    800x600        60.3
>    720x576        50.0
>    720x480        59.9     59.9
>    640x480        60.0     59.9
> HDMI3 disconnected (normal left inverted right x axis y axis)
> DP2 disconnected (normal left inverted right x axis y axis)
> DP3 disconnected (normal left inverted right x axis y axis)
> 
> 2nd test: I reboot again just with HDMI connected.
> LightDM still looking "weird" (like today's picture) and after login
> still split screen.
> When I plug the VGA cable (over DVI-I) the screen goes black (not blue).
> If I restart X the screen goes blue and then black and from VNC I can
> see it's still 1024x768 (so why black?).
> 
> xrandr -q didn't change.
> 
> if I turn the TV over the VGA signal I get a very weird full green
> image flickering very fast.
> 
> 3rd test: I reboot again with both cables connected (VGA over DVI-I and HDMI).
> 
> LightDM and X looks PERFECT! :) Both login screen and session looks
> just like they should.
> 
> 4rd test: same as previous but I want to check if it will happen the same way.
> 
> Yes, again, LightDM and X looks PERFECT! Interestingly, if on the TV I
> switch to VGA input I see a black screen...
> 
> I attach the logs for this session, maybe you can figure out how to
> have this behavior without VGA cable plugged in!
> 
> about testing different resolutions. If I test them from THIS
> situation (e.g. both cables connected from boot) I report the
> following:
> 
> xrandr --output HDMI2 --mode 1280x720 ===> I get quite a bit overscan
> but the picture is stable
> xrandr --output HDMI2 --mode 1024x768 ===> fits like a glove. Not a
> pixel of overscan. Never had such a good fit over VGA with my laptop
> and XP.
> xrandr --output HDMI2 --mode 800x600 ===> this is quite hard to
> explain. The "real" screen is 1024x768 but I can only see 800x600 but
> X seems to know what's going on because it creates two sets of
> controls, one aligned at 800, one aligned at 1024.

I suspect you've had the other output still on with 1024x768. You can
always double-check with xrandr, '*' marks the currently active mode (and
at the top of each output you see which part of a possibly large desktop
it scans out).

> I guess the two screenshots will help you (one of the TV screen and
> then a screen grab).
> 
> > Maybe also attach the raw edid, you can find it in /sys/class/drm/*/edid
> > for your TV.
> 
> about my edid info:
> 
> $ find /sys/ | grep edid
> /sys/devices/pci0000:00/0000:00:02.0/drm/card0/card0-VGA-1/edid
> /sys/devices/pci0000:00/0000:00:02.0/drm/card0/card0-HDMI-A-1/edid
> /sys/devices/pci0000:00/0000:00:02.0/drm/card0/card0-DP-1/edid
> /sys/devices/pci0000:00/0000:00:02.0/drm/card0/card0-HDMI-A-2/edid
> /sys/devices/pci0000:00/0000:00:02.0/drm/card0/card0-HDMI-A-3/edid
> /sys/devices/pci0000:00/0000:00:02.0/drm/card0/card0-DP-2/edid
> /sys/devices/pci0000:00/0000:00:02.0/drm/card0/card0-DP-3/edid
> 
> Interestingly enough the EDID I get from VGA is 256 bytes while EDID
> from HDMI is 128 bytes...

Yeah, that's common TV edid madness, they give you different stuff
depending upon connector. Sometimes you get only TV-y modes on HDMI and
all the default *VGA resolutions only on the dvi/vga port.

Usually this is also good to explain the difference in overscan. For
standard TV resolutions you get overscan, for *VGA resolutions you don't.

Add to that the fact that some of the receivers in TV's are dirt-cheap and
_only_ work with exactly the mode they spec in the edid. The result is
what you've discovered ;-)

> They are both in the logs.

A few other things to test.
- Check whether the edid changes when you connect only one cable. Yep,
  that can happen.
- Check what happens when you only connect the VGA cable.
- There's likely something broken within our driver if HDMI+VGA works, but
  HDMI alone not. The kernel/X also switch on the VGA output which might
  influence (but shouldn't) how we set up HDMI. To check that out please:

Attach xrandr --verbose output for both configurations. Please ensure that
you use the same resolution for both, i.e. for the HDMI+VGA case switch to
1280x720 for the HDMI output.

Grab intel-gpu-tools v1.1 (hopefully your distro ships it) and run the
intel_reg_dumper tool for both configurations (again, please ensure you
use the same mode for HDMI and that you don't change anything between
grabbing the xrandr output and the intel_reg_dumper output).

Yours, Daniel
-- 
Daniel Vetter
Mail: daniel@ffwll.ch
Mobile: +41 (0)79 365 57 48

  reply	other threads:[~2012-01-19 20:50 UTC|newest]

Thread overview: 55+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <4f15cfbd.c39dec0a.71a1.07a8SMTPIN_ADDED@mx.google.com>
2012-01-17 20:14 ` Sandy Bridge Desktop - 1920x1080i interlace not working Alfonso Fiore
2012-01-17 20:25   ` Daniel Vetter
2012-01-17 20:25     ` Daniel Vetter
2012-01-17 20:59       ` Alfonso Fiore
2012-01-17 21:34         ` Daniel Vetter
2012-01-17 21:53           ` Alfonso Fiore
2012-01-17 22:53             ` Angela Schmid
     [not found]             ` <4f15fbec.82620e0a.1fb6.ffff8297SMTPIN_ADDED@mx.google.com>
2012-01-17 23:12               ` Daniel Vetter
2012-01-18  5:54                 ` Angela Schmid
     [not found]                 ` <4f165e84.12960e0a.569b.ffffff51SMTPIN_ADDED@mx.google.com>
2012-01-18 10:56                   ` Daniel Vetter
2012-01-18 16:08           ` Alfonso Fiore
2012-01-18 20:12             ` Daniel Vetter
2012-01-18 20:44               ` Alfonso Fiore
2012-01-18 20:53                 ` Daniel Vetter
2012-01-18 20:55                   ` Alfonso Fiore
2012-01-19  7:05               ` Alfonso Fiore
2012-01-19  9:09                 ` Alfonso Fiore
2012-01-19 11:21                   ` Peter Ross
     [not found]                     ` <CAO=xpBEM=sMmsp7K3eE3Wh+xFnbJqCeHxkkxqsb=Bt4B5KWvNw@mail.gmail.com>
     [not found]                       ` <20120119115144.GB1909@9ba4542ee290bef7a68a5eff03328b0e>
     [not found]                         ` <CAO=xpBFU-WbrLJ9Y_JX=FKaR69N_pR+5RNDMCgHrCB8NY1cJGQ@mail.gmail.com>
     [not found]                           ` <CAO=xpBH+w=cmQ7VWxQPvByBdq145ku7n+2_rXw-PUnR-n6Hd6g@mail.gmail.com>
     [not found]                             ` <20120120124029.GC14963@9ba4542ee290bef7a68a5eff03328b0e>
     [not found]                               ` <CAO=xpBGRG7Zk5DXbKLVjuiSAci6C6qa1M04MT2ZJd7RP6f0QHQ@mail.gmail.com>
2012-01-21 10:25                                 ` Angela Schmid
     [not found]                                 ` <4f1a9294.8a1c0e0a.6f33.fffffbd0SMTPIN_ADDED@mx.google.com>
2012-01-25  1:03                                   ` Alfonso Fiore
2012-01-25 12:07                                     ` Peter Ross
2012-01-19 11:23               ` Alfonso Fiore
2012-01-19 11:32                 ` Daniel Vetter
2012-01-19 11:51                   ` Alfonso Fiore
2012-01-19 12:12                     ` Daniel Vetter
     [not found]                       ` <CAO=xpBG460JGP=b05yqnbK6+-F1LtudwekQ7H-AiarOcOxRqCg@mail.gmail.com>
2012-01-19 13:15                         ` Daniel Vetter
2012-01-19 15:16                           ` Alfonso Fiore
2012-01-19 15:37                             ` Daniel Vetter
2012-01-19 19:56                               ` Alfonso Fiore
2012-01-19 20:50                                 ` Daniel Vetter [this message]
2012-01-19 21:44                                   ` Alfonso Fiore
2012-01-19 21:54                                     ` Daniel Vetter
2012-01-19 22:10                                       ` Alfonso Fiore
2012-01-19 22:20                                         ` Daniel Vetter
2012-01-19 23:37                                   ` Alfonso Fiore
2012-01-21 21:09                                     ` Daniel Vetter
2012-01-21 21:13                                       ` Daniel Vetter
2012-01-22  1:55                                       ` Alfonso Fiore
2012-01-22 10:40                                         ` Daniel Vetter
2012-01-25  1:00                                           ` Alfonso Fiore
2012-01-25  7:55                                             ` Daniel Vetter
2012-01-25  8:12                                               ` Alfonso Fiore
2012-01-25  8:25                                                 ` Alfonso Fiore
2012-01-25  8:41                                                   ` Daniel Vetter
2012-01-25 12:17                                               ` Peter Ross
2012-01-25 12:21                                                 ` Alfonso Fiore
2012-01-25 12:51                                                 ` Daniel Vetter
2012-01-17 21:32       ` Angela Schmid
     [not found]       ` <4f15e8e5.0bbb0e0a.3dfb.655fSMTPIN_ADDED@mx.google.com>
2012-01-17 21:38         ` Daniel Vetter
2012-01-17 20:25   ` Eugeni Dodonov
2012-01-17 21:01     ` Alfonso Fiore
     [not found]       ` <20120118135526.GA16113@9ba4542ee290bef7a68a5eff03328b0e>
     [not found]         ` <CAO=xpBFUpu7vRL+1-1=vd9wDgm25+3Fk-iZSZGNe5L-SZY5gMQ@mail.gmail.com>
2012-01-18 16:47           ` Angela Schmid
     [not found]           ` <4f16f788.8e87cd0a.4c25.488fSMTPIN_ADDED@mx.google.com>
2012-01-18 20:14             ` Daniel Vetter
2012-01-18 20:47               ` Alfonso Fiore
2012-01-17 19:43 Angela Schmid

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20120119205039.GK3779@phenom.ffwll.local \
    --to=daniel@ffwll.ch \
    --cc=alfonso.fiore@gmail.com \
    --cc=intel-gfx@lists.freedesktop.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox