Re: Debug: sleeping function called from invalid context

[Matt Mackall <mpm@selenic.com>]

On Mon, Aug 18, 2003 at 08:35:13PM -0700, Andrew Morton wrote:
> "Randy.Dunlap" <rddunlap@osdl.org> wrote:
> >
> > Debug: sleeping function called with interrupts disabled at
> >  include/asm/uaccess.h:473
> 
> OK, now my vague understanding of what's going on is that the app has
> chosen to disable local interupts (via iopl()) and has taken a vm86 trap. 

Are you suggesting that whatever's calling sys_vm86 has disabled
interrupts beforehand? I don't see why that's necessary at all. The
vm86 fault handler is called via do_general_protection in any case and
as such is in_interrupt() by definition. And a vm86 general protection
fault can be caused by any of cli, sti, pushf, popf, intx, or iret. In
fact, typical usage of vm86 mode is to setup a call to a 16-bit
software interrupt handler and return via fault on the iret.

I'm increasingly convinced it's actually broken.

-- 
Matt Mackall : http://www.selenic.com : of or relating to the moon