Opened 9 years ago

Closed 4 months ago

Last modified 4 months ago

#4312 closed bug (fixed)

"Could not read block, operation timed out" at boot

Reported by: hiei Owned by: mmlr
Priority: normal Milestone: R1
Component: Drivers/USB Version: R1/pre-alpha1
Keywords: Cc:
Blocked By: Blocking:
Has a Patch: no Platform: x86

Description

Downloaded image "haiku-pre-alpha-hrev32591-raw" (gcc2) and flashed on a USB memory stick. Booting from USB, it starts correctly but after showing the blue background, it panics with the attached message.

Some time ago (a year or so) the same procedure worked correctly on the same hardware (Acer Aspire 5560, Intel Core Duo).

Attachments (4)

panic.jpg (263.6 KB) - added by hiei 9 years ago.
syslog.jpg (246.9 KB) - added by hiei 9 years ago.
ints.jpg (305.4 KB) - added by hiei 9 years ago.
20110612-output.jpg (938.2 KB) - added by hiei 7 years ago.

Download all attachments as: .zip

Change History (27)

Changed 9 years ago by hiei

Attachment: panic.jpg added

comment:1 Changed 9 years ago by mmlr

Component: System/KernelDrivers/USB
Owner: changed from axeld to mmlr
Status: newassigned

Can you please try that with a revision >= hrev32615. It's likely that during the heavy load of the initial MIME-setting the tiny 2 second timeout simply wasn't enough. I've enlarged it now, so it should only happen if there is a real underlying problem.

comment:2 Changed 9 years ago by hiei

I've tried hrev32619, now it doesn't panics but simply is stuck with the blue background and the mouse icon. Pressing F12 doesn't works.

comment:3 in reply to:  2 Changed 9 years ago by anevilyak

Replying to hiei:

I've tried hrev32619, now it doesn't panics but simply is stuck with the blue background and the mouse icon. Pressing F12 doesn't works.

F12 won't do anything anyways, the KDL shortcut was changed to alt+printscreen+d a few weeks ago.

comment:4 Changed 9 years ago by hiei

Ok, I tried hrev32619 again and, after waiting 15-20 seconds after USB key led stop flashing, I entered in KDL and typed the "syslog" command. I don't know if it is the right way, but at the end of the syslog there are some messages about interrupts that could be related with my problem. I've attached the screenshoot.

Changed 9 years ago by hiei

Attachment: syslog.jpg added

comment:5 Changed 9 years ago by anevilyak

That actually sounds like ticket #2247 ... can you try disabling Firewire in your BIOS?

comment:6 in reply to:  5 Changed 9 years ago by hiei

Replying to anevilyak:

That actually sounds like ticket #2247 ... can you try disabling Firewire in your BIOS?

Sorry, In my BIOS the only option is for the infrared port. No settings for Fireware...

comment:7 Changed 9 years ago by anevilyak

In that case, can you type "ints" at the kdebug> prompt and take a shot of the resulting output?

comment:8 in reply to:  7 Changed 9 years ago by hiei

Replying to anevilyak:

In that case, can you type "ints" at the kdebug> prompt and take a shot of the resulting output?

Here it is. It's not very readable, I hope you can read what you need.

Changed 9 years ago by hiei

Attachment: ints.jpg added

comment:9 Changed 9 years ago by mmlr

My guess would be that the firewire driver attaches to the wrong interrupt line here. I take it you don't have the option to remove the firewire driver somehow? If you can, you can either remove the user of the firewire stack, the fw_raw driver, or you can remove the firewire bus_manager itself.

comment:10 Changed 9 years ago by hiei

Currently I'm using the nightly images from haiku-files so I think I'm not able to do it. There is any way to disable it from the boot loader?

If needed I could try to compile an image sometime this week and then remove the driver.

comment:11 in reply to:  10 Changed 9 years ago by mmlr

Replying to hiei:

Currently I'm using the nightly images from haiku-files so I think I'm not able to do it. There is any way to disable it from the boot loader?

There isn't.

What I'd be interested in is if the PCI debug output actually shows the same interrupt line as the firewire bus_manager expected it. To find that out could you please enable on-screen debug output and take a look at the first few pages. There you'll see all the devices enumerated with the interesting information. If you search through that you'll see the OHCI Firewire controller somewhere and it has a "interrupt_line" attribute printed. If you could copy that info here, that'd be helpful. In doubt feel free to post a photo of said pages. Thanks!

comment:12 Changed 9 years ago by hiei

Ok, it says "interrupt_line 0b" and "interrupt_pin 01". I tried to make a photo but with the black background it wasn't readable at all.

comment:13 Changed 9 years ago by hiei

I've tried the alfa1 candidate image that doesn't have the fireware driver included and now haiku boots correctly.

comment:14 Changed 8 years ago by Coldfirex

Should this ticket be closed, or should the problem be retested with the firewire driver re-introduced?

comment:15 Changed 7 years ago by scottmc

Can you recheck this with a recent Haiku build?

comment:16 Changed 7 years ago by scottmc

Blocking: 7665 added

comment:17 Changed 7 years ago by hiei

I'm not following Haiku project from some time, if you can point me out where I can find an image to place on the USB to test I can recheck.

Changed 7 years ago by hiei

Attachment: 20110612-output.jpg added

comment:19 Changed 7 years ago by hiei

hrev42095 gives me the attached output, let me know if you need other tests.

comment:20 in reply to:  19 Changed 7 years ago by mmlr

Replying to hiei:

hrev42095 gives me the attached output, let me know if you need other tests.

That one isn't related. That one would be #3305. It probably doesn't get triggered always, so you could try booting a few times. One of those boots should then go through.

comment:21 Changed 7 years ago by hiei

Today I've tried to download the last nightly but there are no recent files on www.haiku-files.org/anyboot at the moment... I've tried Alpha3 and it starts without problems (I'm writing this comment with WebPositive!).

comment:22 Changed 4 months ago by waddlesplash

Resolution: fixed
Status: in-progressclosed

comment:23 Changed 4 months ago by waddlesplash

Blocking: 7665 removed
Note: See TracTickets for help on using tickets.