Re: [Xen-devel] [PATCH 1/2] x86/HVM: fix preemption handling in do_hvm_op() (try 2)

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 09.04.14 at 07:32, <aravindp@xxxxxxxxx> wrote:
>>From: Jan Beulich <JBeulich@xxxxxxxx>
>>Date: Wed, Apr 2, 2014 at 1:09 AM
>>
>>Just like previously done for some mem-op hypercalls, undo preemption
>>using the interface structures (altering it in ways the caller may not
>>expect) and replace it by storing the continuation point in the high bits of 
>>sub-
>>operation argument.
> 
> As part of my PV mem_access work, I am moving HVMOP_set_mem_access in to 
> mem_access_memop() as XENMEM_access_op_set_access, to make it more generic. I 
> am not sure how to go about doing the preemption handling here. I am not able 
> to do something similar as the mem_access ops is a subarch_memory_op, so I 
> was thinking of using the interface structures but the above comment has me 
> worried. Is it not safe to do this for _set_mem_access op? I will send out an 
> RFC patch to make it clear as to what I am doing. Please take a look and 
> advise as to how I should proceed.

Without seeing what you refer to I can't really comment; in any event
I can't offhand see why you shouldn't be able to follow the existing
preemption model used for other mem ops.

Also please remember to strip quotes from the responded to mail that
aren't relevant for your reply.

Jan


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