Re: softlockup interaction with slow consoles

[David S. Miller]

From: Matt Mackall <mpm@xxxxxxxxxxx>
Date: Tue, 21 Feb 2006 13:23:40 -0600

> I don't like it. We should instead just have printk tickle the watchdog.

You can't, interrupts are disabled the entire time and thus
jiffies aren't advancing.  The "touch" just sets the local
cpu timestamp to whatever jiffies is.

That's the problem.

I agree that it would be nice to fix this, but the desire for
synchronous console output makes this very hard to solve.

Like I said, maybe some local chunking in release_console_sem()
but since interrupts will come in we can trigger more and more
lengthy console output and eventually overflow the printk log
buffer before the chunking loop can catch up and we'll thus
lose messages.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/