Opened 16 years ago

Closed 3 years ago

Last modified 3 years ago

#1729 closed bug (not reproducible)

[radeon] monitor turns off after boot logo

Reported by: diver Owned by: euan
Priority: normal Milestone: R1
Component: Drivers/Graphics/radeon Version: R1/Development
Keywords: boot-failure Cc: Wanderer49@…, siarzhuk, jstressman@…
Blocked By: Blocking: #5348
Platform: All

Description

It's a bug from another user, but i CC'ed him.

With radeon driver his monitor turns off after haiku boot logo. Without radeon driver it works fine, in vesa of course.

listdev from BeOS:

PCI Devices VGA - vendor ATI Technologies (0*1002) card ID 5b60

Display Controller vendor ATI Technologies (0*1002) card ID 5b70

Attachments (2)

syslog (61.7 KB ) - added by diver 16 years ago.
syslog without radeon driver
softdc-07-2.log (93.5 KB ) - added by siarzhuk 14 years ago.
Debug output acquired during normal boot on system with ATI 5a61: RC410 [Radeon Xpress 200] Graphics.

Download all attachments as: .zip

Change History (23)

comment:1 by euan, 16 years ago

Can you get a syslog?

What type of monitor is it? What's it native resolution? What connection type? What output port is it connected to?

comment:2 by diver, 16 years ago

No syslog yet. Compaq 140 CRT.

Resolution up to 1024x768@72. Connected to analog port.

by diver, 16 years ago

Attachment: syslog added

syslog without radeon driver

comment:3 by diver, 16 years ago

with radeon driver it currently gives

PANIC: vm_page fault: unhandled page fault in kernel space ot 0*250409, ip 0*90571ddb

Don't know if it related, though.

comment:4 by lt_henry, 14 years ago

As I can see, it is a RV370 based chipset. I own a R300 one and It happens but not very often through. It also happens on any Linux distro I've test since 2005 (almost all Ubuntu releases). In fact, does the monitor really turns off? I get a black screen but LCD lamp still working.

comment:5 by axeld, 14 years ago

Version: R1/pre-alpha1R1/Development

What happens if you press "Command + Control + Escape" after a while?

Also, if that doesn't help, can you log into the system, and execute the following application?

#include <Application.h>
#include <Screen.h>

int
main()
{
	BApplication app("application/x-vnd.test");

	BScreen screen;
	screen.SetDPMS(B_DPMS_OFF);
	snooze(5000000);
	screen.SetDPMS(B_DPMS_ON);

	return 0;
}

comment:6 by siarzhuk, 14 years ago

Cc: siarzhuk added

comment:7 by siarzhuk, 14 years ago

Blocking: 5348 added

(In #5348) Blocked by #1729. Until that ticket will be not fixed on my hardware it is not possible to check audio functions in normal Haiku session. Using fail-safe video mode bring the audio hardware in "hda: no active codec" state.

in reply to:  5 comment:8 by siarzhuk, 14 years ago

Blocking: 5348 removed

Replying to axeld:

I have the following graphics hardware on one of our PCs:

device Display controller (VGA compatible controller, VGA controller) [3|0|0]
  vendor 1002: ATI Technologies Inc
  device 5a61: RC410 [Radeon Xpress 200]

This is normal PC with onboard graphics with external LCD Monitor attached to VGA output.

I observe the above described behavior with hrev35333. Just after the system finishes coloring all boot icons the monitor first display an "Invalid signal range" message for one second than goes to "No signal" state. It is only possible to boot with fail-safe video mode.

What happens if you press "Command + Control + Escape" after a while?

No effect on the screen but lot of tracing in the debug output. Look into attached softdc-07-2.log for details after "sync: device 3 couldn't sync: Operation not supported" line. Yes, I have pressed Alt-Ctrl-Esc and Ctrl-Alt-Esc twice. :-)

Also, if that doesn't help, can you log into the system, and execute the following application?

The specific of this PC is that there is the only possibility to load Haiku - booting from the Live-CD. I have compiled this "DPMS reset" program but with "return 1;" and replaced the /boot/system/bin/ReadOnlyBootPrompt with it's executable. ;-) I do this just to automaticaly force it to be started during loading Live-CD session. Unfortunately that doesn't help to switch the display on.

by siarzhuk, 14 years ago

Attachment: softdc-07-2.log added

Debug output acquired during normal boot on system with ATI 5a61: RC410 [Radeon Xpress 200] Graphics.

comment:9 by siarzhuk, 14 years ago

Blocking: 5348 added

Doesn't our Track handle mid-air collisions???? That's bad. :-(

comment:10 by euan, 14 years ago

Looks like the driver has detected an Svideo connection, probably due to the monitor routing being incorrect on that board. Can you test to see if there is a working picture outputing on Svideo, or TV-out (assuming it even has that connection).

in reply to:  10 comment:11 by siarzhuk, 14 years ago

Replying to euan:

Looks like the driver has detected an Svideo connection, probably due to the monitor routing being incorrect on that board. Can you test to see if there is a working picture outputing on Svideo, or TV-out (assuming it even has that connection).

Unfortunately it is a cheap onboard hardware, I see only VGA output on ot. :-(

comment:12 by scottmc, 13 years ago

Can you recheck this with a recent Haiku build?

comment:13 by scottmc, 13 years ago

Blocking: 7665 added

comment:14 by jstressman, 12 years ago

Cc: jstressman@… added

I have an S-video out on my motherboard, and I can confirm that this is in fact what it's doing... however, the video on the tv is so bad that you can't make anything out. :(

CTRL-ALT-ESC toggles it on and off apparently... but doesn't make it any more visible.

It's basically a thin line of scrambled colored pixels about 1/3rd of the way down from the top of the screen. They do change when you move the mouse, but you can't make anything out.

This is with the "RC410 [Radeon Xpress 200]" as well, and I'm still seeing this on hrev44335

comment:15 by waddlesplash, 6 years ago

Keywords: boot-failure added

comment:16 by waddlesplash, 6 years ago

Blocking: 7665 removed

comment:17 by X512, 3 years ago

Is it still actual? Anybody have hardware to test?

comment:18 by euan, 3 years ago

I would say no. This is quite possibly retro. Haiku-os would do good to clear out all these ancient issues and set a base level of modern hardware for primary support and attention. :)

Shall we put this one to bed?

Last edited 3 years ago by euan (previous) (diff)

comment:19 by jstressman, 3 years ago

I no longer have any of this hardware to test the bug. I'd lean toward retiring this as well, if it matters.

comment:20 by euan, 3 years ago

Resolution: not reproducible
Status: newclosed

comment:21 by euan, 3 years ago

Done

Note: See TracTickets for help on using tickets.