[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v6 3/5] IOMMU: Make the pcidevs_lock a recursive one



>>> On 04.03.16 at 14:59, <dario.faggioli@xxxxxxxxxx> wrote:
> On Fri, 2016-03-04 at 11:54 +0000, Xu, Quan wrote:
>> On March 04, 2016 5:29pm, <JBeulich@xxxxxxxx> wrote:
>> > To be honest, changes like this would better not be buried in a big
>> > rework like
>> > the one here. Make it a prereq patch, where you then will kind of
>> > automatically
>> > describe why it's correct. (I agree current code is bogus, and
>> > we're not hitting
>> > the respective
>> > BUG_ON() in check_lock() only because spin_debug gets set to true
>> > only after
>> > most __init code has run.)
>> Agreed, I would make a prereq patch in v7.
>> 
> Ok. In general, I agree with Jan.
> 
> In this case, I suggested just mentioning in changelog as we curently
> basically have a bug, and I think it would be fine to have something
> like "Doing what we do also serves as a fix for a bug found in xxx
> yyy".
> 
> But it's indeed Jan's call, and his solution is certainly cleaner.

Well, one of the reasons to separate out bug fixes is to make them
backportable.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.