Re: [PATCH v3 2/2] x86/vpt: Simplify locking argument to write_{un}lock
On 3/30/21 8:49 AM, Jan Beulich wrote:
> On 30.03.2021 09:36, Roger Pau Monné wrote:
>> On Mon, Mar 29, 2021 at 05:15:02PM -0400, Boris Ostrovsky wrote:
>>> Make both create_periodic_time() and pt_adjust_vcpu() call
>>> write_{un}lock with similar arguments.
> This makes it sound like you adjust both functions, but really
> you bring the latter in line with the former. Would you mind me
> adjusting the wording along these lines while (and when)
> committing?
Yes, please.
>
>> Might be worth adding that this is not a functional change?
>>
>>> Requested-by: Jan Beulich
>>> Signed-off-by: Boris Ostrovsky
>> Either way:
>>
>> Reviewed-by: Roger Pau Monné
>>
>>> ---
>>> New patch.
>>>
>>> I ended up doing what Jan asked --- create_periodic_time() is also using
>>> different
>>> start pointers in lock() and unlock().
>> Hm, I'm not sure I'm following, create_periodic_time uses 'v' in both
>> write_{un}lock calls, which doesn't change across the function.
> I guess Boris merely meant to express that there's already precedent?
Yes, that's what I wanted to say. But clearly not what I actually said.
-boris
Re: [PATCH v3 2/2] x86/vpt: Simplify locking argument to write_{un}lock
On 30.03.2021 09:36, Roger Pau Monné wrote:
> On Mon, Mar 29, 2021 at 05:15:02PM -0400, Boris Ostrovsky wrote:
>> Make both create_periodic_time() and pt_adjust_vcpu() call
>> write_{un}lock with similar arguments.
This makes it sound like you adjust both functions, but really
you bring the latter in line with the former. Would you mind me
adjusting the wording along these lines while (and when)
committing?
> Might be worth adding that this is not a functional change?
>
>> Requested-by: Jan Beulich
>> Signed-off-by: Boris Ostrovsky
>
> Either way:
>
> Reviewed-by: Roger Pau Monné
>
>> ---
>> New patch.
>>
>> I ended up doing what Jan asked --- create_periodic_time() is also using
>> different
>> start pointers in lock() and unlock().
>
> Hm, I'm not sure I'm following, create_periodic_time uses 'v' in both
> write_{un}lock calls, which doesn't change across the function.
I guess Boris merely meant to express that there's already precedent?
Jan
Re: [PATCH v3 2/2] x86/vpt: Simplify locking argument to write_{un}lock
On Mon, Mar 29, 2021 at 05:15:02PM -0400, Boris Ostrovsky wrote:
> Make both create_periodic_time() and pt_adjust_vcpu() call
> write_{un}lock with similar arguments.
>
Might be worth adding that this is not a functional change?
> Requested-by: Jan Beulich
> Signed-off-by: Boris Ostrovsky
Either way:
Reviewed-by: Roger Pau Monné
> ---
> New patch.
>
> I ended up doing what Jan asked --- create_periodic_time() is also using
> different
> start pointers in lock() and unlock().
Hm, I'm not sure I'm following, create_periodic_time uses 'v' in both
write_{un}lock calls, which doesn't change across the function.
Thanks, Roger.
