IvyBridge: After DPMS, screen never turns back on

[kallisti5]

After the screen turns off via DPMS, the display will not come back on.

Asus Zenbook, Intel HD 4000 IvyBridge.

[mmlr]

On my Samsung np900x3e (also IvyBridge / Intel HD Graphics 4000) something curious happens when DPMS is used as well: Instead of turning off, the screen goes into some kind of test mode where it continuously cycles between red, green, blue, white and black. Moving the mouse gets back to Haiku just fine. Unless that is some kind of easter egg in our Blackness screen safer?

[jessicah]

I'm not sure if this issue is directly related to mine. It does the whole system init thing all the way to the rocket, Then the display just turns off. Brand new Intel setup: DH77DF with Core i5 3570K CPU (IvyBridge / Intel HD 4000).

[kallisti5]

I've seen the same issue recently on a Dell OptiPlex 3010 i5 system. Definitely still an issue.

[pulkomandy]

Can we get a syslog/serial log of this? There could be some relevant info from the driver, or a crash log there.

[kallisti5]

logs

[kallisti5]

These syslogs are from a DVI monitor that isn't turning on at boot. Just a black screen. I think it's related as it's ivy bridge and past monitors have had the DPMS issue.

[pulkomandy]

The original ticket is "After DPMS" (I think from screensaver), not turning off at boot. These are different issues (completely broken driver vs just the powersave/wakeup part).

I can reproduce the "no display at all" on the VGA port of my Haswell machine too (but HDMI works there). But it's a different problem.

[kallisti5]

I still think these two might be related. Pretty much the powerup sequence of the DAC / Encoder not working randomly. Sometimes it's at boot powerup, sometimes it's at DPMS power-up.

Issue still exists as of hrev48858 (after Pulkomandy's vsync fixes)

~> cat /var/log/syslog | grep intel_extreme
KERN: intel_extreme: CALLED status_t init_hardware()
KERN: intel_extreme: CALLED status_t init_driver()
KERN: intel_extreme: (0) IvyBridge Desktop GT1, revision = 0x9
KERN: intel_extreme: CALLED const char** publish_devices()
KERN: intel_extreme: CALLED device_hooks* find_device(const char*)
KERN: loaded driver /boot/system/add-ons/kernel/drivers/dev/graphics/intel_extreme
KERN: intel_extreme: CALLED status_t device_open(const char*, uint32, void**)
KERN: intel_extreme: intel_extreme: CALLED status_t intel_extreme_init(intel_info&)
KERN: intel_extreme:CALLED status_t intel_en_gating(intel_info&)
KERN: intel_extreme:IvyBridge clock gating
KERN: intel_extreme:CALLED status_t intel_en_downclock(intel_info&)
KERN: intel_extreme:KERN: intel_en_downclock: Skip GPU downclocking on non-mobile device.
KERN: intel_extreme: mapping VBIOS: 0xc0000 -> 0xffffffff80746000
KERN: intel_extreme: parsing BDB blocks
KERN: intel_extreme: using message signaled interrupts
KERN: intel_extreme: intel_extreme_init: completed successfully!
KERN: intel_extreme: accelerant: intel_extreme.accelerant
KERN: intel_extreme accelerant:CALLED status_t intel_init_accelerant(int)
KERN: intel_extreme accelerant:head detected: 0x1
KERN: intel_extreme accelerant:adpa: 83c40018, dova: 00000018, dovb: 00000018, lvds: 00000000
KERN: intel_extreme accelerant:found edid data on gpio 'vga'
KERN: intel_extreme accelerant:CALLED status_t intel_get_edid_info(void*, size_t, uint32*)
KERN: intel_extreme accelerant:intel_set_display_mode(1920x1080)
KERN: intel_extreme accelerant:PLL limits, min: p 5 (p1 1, p2 10), n 1, m 79 (m1 12, m2 5)
KERN: intel_extreme accelerant:PLL limits, max: p 80 (p1 8, p2 5), n 5, m 127 (m1 22, m2 9)
KERN: intel_extreme accelerant:compute_pll_divisors: required MHz: 172.51
KERN: intel_extreme accelerant:compute_pll_divisors: found: 172.8 MHz, p = 20 (p1 = 2, p2 = 10), n = 3, m = 108 (m1 = 20, m2 = 8)
KERN: intel_extreme accelerant:CALLED status_t intel_get_frame_buffer_config(frame_buffer_config*)

[pulkomandy]

What's the status on this? Did the many changes to the intel driver fix it?

[kallisti5]

I'll dig out the hardware and give it a try :-)

[markh]

I have an IvyBrigde Mobile GT2 chipset and the screen turns back on after DPMS turns it off. I'm on revision hrev51019. I used to have this problem as well, so I would say it got fixed somewhere along the way.

[waddlesplash]

Excellent. Closing as fixed then.