On Wed, Jul 31, 2013 at 10:07:58AM -0400, Tejun Heo wrote:

On Wed, Jul 31, 2013 at 02:57:51PM +0100, Mark Brown wrote:

quoted

That's the only API I've ever heard of doing that.  Everything else is
just using it to do deallocation.

I'm not sure why or what you're trying to argue here but take a look
at devm_pwm_release() for example.  It calls back into low level
driver free routine.  Are you arguing that it'd be a good idea to

That the callback is into the driver providing the PWM, not into the
driver that's using the PWM and is releasing it.

release pci regions before this is complete?  It's just stupid to do
any differently.  There's nothing to argue about.

What I'm saying is that in essentially all the users I've seen devm is
only being used for things like kfree() or clk_put() which aren't really
connected in any way and can happen in any order.  This (coupled with
the lack of documentation that this is supported) is why people are
nervous about anything that relies on ordering with this stuff - aside
from ATA everything is just using this for straight frees.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20130731/5bc05f02/attachment.sig>