svn commit: r192535 - head/sys/kern

[Scott Long]

M. Warner Losh wrote:
> In message: <4A16AC32.2040507 at samsco.org>
>             Scott Long <scottl at samsco.org> writes:
> : John Baldwin wrote:
> : > On Thursday 21 May 2009 6:11:02 pm Attilio Rao wrote:
> : >> At this point I wonder what's the purpose of maintaining the sleeping
> : >> version for such functions?
> : > 
> : > Actually, I still very much do not like using M_NOWAIT needlessly.  I would 
> : > much rather the solution for make_dev() be that the 1 or 2 places that need 
> : > to do it with a mutex held instead queue a task to do the actual make_dev() 
> : > in a taskqueue when no locks are held.  This is basically what 
> : > destroy_dev_sched() is doing.  Perhaps a make_dev_sched() with a similar 
> : > callback to be called on completion would be better.  Having a device driver 
> : > do all the work to setup the hardware only to fail to create a node in /dev 
> : > so that userland can actually use it is pretty rediculous and useless.
> : > 
> : 
> : It's a lot easier for me to handle a failure of make_dev in CAM than it 
> : is to decouple the call to it.  Please don't dictate policy.
> 
> On the other hand, we do dictate policy in things like busdma where
> one has to do things in callbacks rather than inline.  This is for
> fairly good reasons, and I'm having trouble seeing why the reasons
> presented here for make_dev_sched() are any worse...
> 
> Warner

Busdma isn't a good example anymore.  I've tried to be very responsive
and accommodating to requests for change; see the
bus_dmamap_load_mbuf_sg() routine for example.  It also lets you break
the "normal" semantics without penalty via BUS_DMA_NOWAIT.  About the
only thing left in busdma that is cumbersome without an alternative is
allocating static memory.  Even then, I provided an alternative for a
number of years, and not a single person used it, so it eventually got
removed.

Scott