Re: Linux 3.4-rc4

[Luca Tettamanti <kronos.it@gmail.com>]

On Mon, Apr 30, 2012 at 11:07 AM, Maarten Maathuis <madman2003@gmail.com> wrote:
> On Mon, Apr 30, 2012 at 12:37 AM, Dmitry Torokhov
> <dmitry.torokhov@gmail.com> wrote:
>> On Sat, Apr 28, 2012 at 11:33:50AM -0400, Nick Bowler wrote:
>>> On 2012-04-28 02:19 -0400, Alex Deucher wrote:
>>> > On Fri, Apr 27, 2012 at 8:39 PM, Nick Bowler <nbowler@elliptictech.com> wrote:
>>> > > Hi Ben,
>>> > >
>>> > > On 2012-04-27 15:20 +1000, Ben Skeggs wrote:
>>> > >> Does this patch help you at all?
>>> > >>
>>> > >> http://cgit.freedesktop.org/nouveau/linux-2.6/commit/?id=a3a285f17867f0018de798b5ee85731ec1268305
>>> > >
>>> > > Yes.  I cherry-picked this patch on top of Linus' master (3.4-rc4+) and
>>> > > this appears to solve the "black screen on VGA" problem described in the
>>> > > original report.  Thanks!
>>> > >
>>> > > Unfortunately, that's not the end of my VGA-related regressions. :(
>>> > >
>>> > > While tracking down the black screen issue, I've been having the monitor
>>> > > directly connected to the video card the whole time, but now when I'm
>>> > > connected through my KVM switch (an IOGear GCS1804), it appears that
>>> > > something's going wrong with reading the EDID, because the available
>>> > > modes are all screwed up (both console and X decide they want to drive
>>> > > the display at 1024x768).  Here's the output of xrandr on 3.2.15:
>>> > >
>>> > >  % xrandr
>>> > >  Screen 1: minimum 320 x 200, current 1600 x 1200, maximum 4096 x 4096
>>> > >  VGA-1 connected 1600x1200+0+0 (normal left inverted right x axis y axis) 352mm x 264mm
>>> > >     1600x1200      75.0*+   70.0     65.0     60.0
>>> > >     1280x1024      85.0 +   75.0     60.0
>>> > >     1920x1440      60.0
>>> > >     1856x1392      60.0
>>> > >     1792x1344      60.0
>>> > >     1920x1200      74.9     59.9
>>> > >     1680x1050      84.9     74.9     60.0
>>> > >     1400x1050      85.0     74.9     60.0
>>> > >     1440x900       84.8     75.0     59.9
>>> > >     1280x960       85.0     60.0
>>> > >     1360x768       60.0
>>> > >     1280x800       84.9     74.9     59.8
>>> > >     1152x864       75.0
>>> > >     1280x768       84.8     74.9     59.9
>>> > >     1024x768       85.0     75.1     75.0     70.1     60.0     43.5     43.5
>>> > >     832x624        74.6
>>> > >     800x600        85.1     72.2     75.0     60.3     56.2
>>> > >     848x480        60.0
>>> > >     640x480        85.0     75.0     72.8     72.8     66.7     60.0     59.9
>>> > >     720x400        85.0     87.8     70.1
>>> > >     640x400        85.1
>>> > >     640x350        85.1
>>> > >     320x200       165.1
>>> > >
>>> > > And on 3.4-rc4+ (with your patch cherry-picked):
>>> > >
>>> > >  % xrandr
>>> > >  Screen 1: minimum 320 x 200, current 1024 x 768, maximum 4096 x 4096
>>> > >  VGA-1 connected 1024x768+0+0 (normal left inverted right x axis y axis) 0mm x 0mm
>>> > >     1024x768       60.0*
>>> > >     800x600        60.3     56.2
>>> > >     848x480        60.0
>>> > >     640x480        59.9
>>> > >     320x200       165.1
>>> > >
>>> > > Running xrandr on 3.4-rc4+ also causes the screen to go black for a
>>> > > second when it does not on 3.2.15.  It also causes several messages of
>>> > > the form
>>> > >
>>> > >  [drm] nouveau 0000:01:00.0: Load detected on output B
>>> > >
>>> > > to be logged.  Also, looking at /sys/class/drm/card0-VGA-1/edid I see
>>> > > that it is empty on 3.4-rc4+ and it is correct on 3.2.15.  Things seem
>>> > > to work OK when the KVM is not involved.
>>> >
>>> > Were you ever able to fetch a EDID with the KVM involved?  KVMs are
>>> > notorious for not connecting the ddc pins.
>>>
>>> Yes, it works on 3.2.15 as described above.
>>
>> I have the same (or similar) KVM (not in the office at the moment) and I
>> can confirm that with newer kernels EDID fecthing in flaky. It's 50/50
>> if EDED retrieval succeeds or if it fails with:
>>
>> Apr 26 13:06:57 dtor-d630 kernel: [13464.936336] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>> Apr 26 13:06:57 dtor-d630 kernel: [13464.955317] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>> Apr 26 13:06:57 dtor-d630 kernel: [13464.973879] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>> Apr 27 09:13:03 dtor-d630 kernel: [44602.087659] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>> Apr 27 09:13:03 dtor-d630 kernel: [44602.107147] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>> Apr 27 09:13:03 dtor-d630 kernel: [44602.126908] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>> Apr 27 09:13:03 dtor-d630 kernel: [44602.146277] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>> Apr 27 09:13:03 dtor-d630 kernel: [44602.297659] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>> Apr 27 09:13:03 dtor-d630 kernel: [44602.317063] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 208
>>
>> Earlier kernels were able to retrieve EDEDs reliably.
>>
>> This is with:
>>
>> [    1.678392] [drm] nouveau 0000:01:00.0: Detected an NV50 generation card (0x086b00a2)
>
> Just a crazy thought, but didn't we change some timings related to
> EDID retrieval? To make it faster.

Hum, this commit:

commit 1849ecb22fb3b5d57b65e7369a3957adf9f26f39
Author: Jean Delvare <jdelvare@suse.de>
Date:   Sat Jan 28 11:07:09 2012 +0100

    drm/kms: Make i2c buses faster

doubled the data rate but only for radeon and intel drivers. nouveau
doesn't use the standard i2c-algo-bit helpers (BTW: the cond_resched()
has been removed), and AFAICS it's using 1us delay; the other drivers
are using 10us, 1us seems a bit too low...

Luca