Re: mdelay causes BUG, please use udelay

[Dave Hansen <haveblue@us.ibm.com>]

Martin J. Bligh wrote:
>>>-    msec_delay(10);
>>>+    usec_delay(10000);
>>
>>Jeff, 10000 seems on the border of what's OK.  If it's acceptable, 
>>then let's go for that.  Otherwise, we're going to have to chain 
>>several mod_timer callbacks together to do a controller reset.
> 
> Whilst this sort of delay in interrupt context is undoubtedly bad
> any way we do it, I'd question the context a little more before we
> make a decision. This is called from e1000_reset_hw - are we likely
> to ever actually call this except under initialisation?

It doesn't happen often, or under good circumstances.  In certain 
cases, the driver detects that something timed out and it assumes 
something on the card to be dead.  Instead of delaying the 
reinitialization of the dead card with a timer, they just do it during 
the interrupt where the problem was detected.


-- 
Dave Hansen
haveblue@us.ibm.com