Control: close -1
Hi Ben,
thanks for caring.
Ben Hutchings wrote:
> > * X is only shown on two of three screens and both show the same
> > contents instead of different content.
> >
> > * Resolution is only (1x) 1024x768 instead of the expected 1920x1080 +
> > 1280x720 (+ 2048x1152)
>
> This looks like X is having to use a generic vesa driver.
*nod*
> > * As soon as X starts, neither USB keyboard nor mouse are working
> > anymore under X. Pressing Ctrl-Alt-F1 to get to a text console has no
> > effect. (I do get a text console if I login remotely via SSH as root
> > and call "chvt 1". On the text console the USB keyboard is working
> > again.)
>
> I don't know why that would happen, though.
I have a vague idea now: Hardware failure. :-/
Today I replaced an PCI-based USB 3.0 controller which wasn't working
for a while. I got the replacement card (cheap noname instead of
Delock) only today and put in before I booted back into 4.9 to answer
your questions.
I properly mounted the replacement card and then noticed that I don't
have a spare power cable left (the old one had a SATA power socket,
the new one had an old Molex power socket), so I started it without
the required power cable for now. It still throws similar error
messages (which suggest that maybe the power cable I used with the old
adapter and not the adapter was broken) and additional ones about
being underpowered, but to my surprise …
* X works again as expected.
* USB input devices (as before not connected to the underpowered
adapter but to the primary onboard controller) work fine again, too.
Actually I'm writing this mail under 4.9.2-2 and X.
And since nothing kernel- or X-wise changed since I reported the
issue, I can only assume that this hardware replacement fixed it.
I nevertheless have no idea why a non working USB controller would
reproducibly break the graphics card recognition depending on the
kernel version. (I tried at least 5 times: 4.8 worked, 4.9 didn't
work, 4.8-grsec worked, 4.9 didn't work, 4.8 worked.)
> This device is still listed in i915's device ID table, so it should
> have been loaded and would then have probed the device (even though
> that may have failed). Yet it's not even on the loaded module list.
>
> Does 'modprobe i915' have any effect?
Since I (luckily :-) can no more reproduce the issue, I can't answer
that question anymore.
I'm hence closing this bug report and hopefully don't have to reopen
it again. :-)
Sorry for the noise.
Regards, Axel
--
,''`. | Axel Beckert <a...@debian.org>, http://people.debian.org/~abe/
: :' : | Debian Developer, ftp.ch.debian.org Admin
`. `' | 4096R: 2517 B724 C5F6 CA99 5329 6E61 2FF9 CD59 6126 16B5
`- | 1024D: F067 EA27 26B9 C3FC 1486 202E C09E 1D89 9593 0EDE
