Re: workqueue thing

[Stefan Richter]

Ingo Molnar wrote:
> Likewise, if there's a reduction in complexity, that is a tangible metric as 
> well: lets do a few conversions as part of the patch-set and see how much 
> simpler things have become as a result of it.

There are for example about 160 users of  create_singlethread_workqueue
(about 140 of them in drivers/.).  As has mentioned by others, many if
not most of them would be better served by either the existing slow-work
API or by the proposed worker thread pool.  Conversion to the former
takes a bit more effort than the latter (not much, but it matters).

The little driver subsystem which I maintain extensively uses the shared
workqueue and also one single-thread workqueue.  The facts that the
shared queue is used for several purposes and a single thread for some
are both just compromises which I would rather like to get rid of.  I
should have converted the create_singlethread_workqueue usage to David
Howells' slow-work infrastructure immediately when that was merged, but
I didn't do so yet because there is too much else on my to-do list for
that particular to-do item...

> We really are not forced to the space of Gedankenexperiments here.

...to actually leave Gedankenexperiment stage as quickly as I would
like.  Tejun's worker pool would make things easier for me.
-- 
Stefan Richter
-=====-==--= ==-- =-===
http://arcgraph.de/sr/
--
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/