[Bug 112281] New: [i915][hdmi] Problems detecting/negotiating resolution and with sleeping

[bugzilla-daemon at bugzilla.kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=112281

            Bug ID: 112281
           Summary: [i915][hdmi]  Problems detecting/negotiating
                    resolution and with sleeping
           Product: Drivers
           Version: 2.5
    Kernel Version: 4.4.1
          Hardware: x86-64
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: high
          Priority: P1
         Component: Video(DRI - Intel)
          Assignee: intel-gfx-bugs at lists.freedesktop.org
          Reporter: j.e.vasquez.v at gmail.com
                CC: intel-gfx-bugs at lists.freedesktop.org
        Regression: No

Please notice on linux version 4.3.3 I had no problems, and now with version
4.4.1 the problems are there.

The problems are present both on plain terminal (no Xorg) or under X (Xorg).

I'm not setting any particular mode setting, and I just let KMS do its magic. 
On X, I have no configuration for the graphics at all, letting again KMS and
Xorg do their magic.

I'm using i915 driver, on a MintBox 2 Arch gnu/linux box, which comes with a
"Intel HD Graphics 4000", which according to lspci -v:

00:02.0 VGA compatible controller: Intel Corporation 3rd Gen Core processor
Graphics Controller (rev 09) (prog-if 00 [VGA controller])
Subsystem: Intel Corporation Device 2211
Flags: bus master, fast devsel, latency 0, IRQ 28
Memory at e0000000 (64-bit, non-prefetchable) [size=4M]
Memory at c0000000 (64-bit, prefetchable) [size=512M]
I/O ports at 4000 [size=64]
Expansion ROM at <unassigned> [disabled]
Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
Capabilities: [d0] Power Management version 2
Capabilities: [a4] PCI Advanced Features
Kernel driver in use: i915
Kernel modules: i915

I don't have any particular modprobe setting for the i915 module, just default.

Another important piece of information prior to describe the problems, is that
the display is hooked to the box through its HDMI port.

Now the problems...

1.-  Resolution problem @ boot
     =========================

If the box boots before the display has turned on and is waiting for signal on
the hdmi port, then the 2 never get to negotiate the right resolution and a
pretty low one gets picked.  The display has some manual setting, but the
"native" one from what it's hooked to is not an option, like there's no way for
the display to negotiate with the box...  BTW, the display is a 720p one...

To keep things totally automatic, the only work around is to turn the display
on, until its state of waiting for signal on the hdmi port, and then turn the
box on.  If this is missed, a reboot is required to get the box and the display
to negotiate the resolution.

2.-  Resolution problem @ non fully asleep power saving state
     ========================================================

When left turned on, but unused, at some point the box graphics get to a state
such that it blanks the display, but it's not totally asleep.  At that point
one can turn the display back on, but the same resolution problem as the boot
one shows up.  If leaving everything automatic, a reboot is required.

3.-  Graphics never recovering from fully asleep power saving state
     ==============================================================

When left turned on, but unused, at some point the box graphics get to a state
such that it blanks the display and the graphics get totally asleep.  At that
point, one can turn the display on, then off, then on again and so on, that the
graphics never wake up.  At that point, if not handled remotely, the option is
a hard power button (whether it turns off or reboot the box is a matter of
configuration).

To prevent getting into this state under Xorg, I'm using as a work around:

xset s off
xset -dpms

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.