From: Waiman Long <longman@redhat.com>
To: Will Deacon <will@kernel.org>, peterz@infradead.org
Cc: Michael Ellerman <mpe@ellerman.id.au>,
Nicholas Piggin <npiggin@gmail.com>,
linuxppc-dev@lists.ozlabs.org, Boqun Feng <boqun.feng@gmail.com>,
Ingo Molnar <mingo@redhat.com>,
Anton Blanchard <anton@ozlabs.org>,
linux-kernel@vger.kernel.org,
virtualization@lists.linux-foundation.org,
kvm-ppc@vger.kernel.org, linux-arch@vger.kernel.org
Subject: Re: [PATCH v3 5/6] powerpc/pseries: implement paravirt qspinlocks for SPLPAR
Date: Fri, 24 Jul 2020 23:02:58 -0400 [thread overview]
Message-ID: <ccf0c6a6-b7c3-8909-cc8f-0c5e7434c372@redhat.com> (raw)
In-Reply-To: <8532332b-85dd-661b-cf72-81a8ceb70747@redhat.com>
On 7/24/20 3:10 PM, Waiman Long wrote:
> On 7/24/20 4:16 AM, Will Deacon wrote:
>> On Thu, Jul 23, 2020 at 08:47:59PM +0200, peterz@infradead.org wrote:
>>> On Thu, Jul 23, 2020 at 02:32:36PM -0400, Waiman Long wrote:
>>>> BTW, do you have any comment on my v2 lock holder cpu info
>>>> qspinlock patch?
>>>> I will have to update the patch to fix the reported 0-day test
>>>> problem, but
>>>> I want to collect other feedback before sending out v3.
>>> I want to say I hate it all, it adds instructions to a path we spend an
>>> aweful lot of time optimizing without really getting anything back for
>>> it.
>>>
>>> Will, how do you feel about it?
>> I can see it potentially being useful for debugging, but I hate the
>> limitation to 256 CPUs. Even arm64 is hitting that now.
>
> After thinking more about that, I think we can use all the remaining
> bits in the 16-bit locked_pending. Reserving 1 bit for locked and 1
> bit for pending, there are 14 bits left. So as long as NR_CPUS < 16k
> (requirement for 16-bit locked_pending), we can put all possible cpu
> numbers into the lock. We can also just use smp_processor_id() without
> additional percpu data.
Sorry, that doesn't work. The extra bits in the pending byte won't get
cleared on unlock. That will have noticeable performance impact.
Clearing the pending byte on unlock will cause other performance
problem. So I guess we will have to limit the cpu number in the locked byte.
Regards,
Longman
WARNING: multiple messages have this Message-ID (diff)
From: Waiman Long <longman@redhat.com>
To: Will Deacon <will@kernel.org>, peterz@infradead.org
Cc: linux-arch@vger.kernel.org, Boqun Feng <boqun.feng@gmail.com>,
linux-kernel@vger.kernel.org, Nicholas Piggin <npiggin@gmail.com>,
virtualization@lists.linux-foundation.org,
Ingo Molnar <mingo@redhat.com>,
kvm-ppc@vger.kernel.org, linuxppc-dev@lists.ozlabs.org
Subject: Re: [PATCH v3 5/6] powerpc/pseries: implement paravirt qspinlocks for SPLPAR
Date: Fri, 24 Jul 2020 23:02:58 -0400 [thread overview]
Message-ID: <ccf0c6a6-b7c3-8909-cc8f-0c5e7434c372@redhat.com> (raw)
In-Reply-To: <8532332b-85dd-661b-cf72-81a8ceb70747@redhat.com>
On 7/24/20 3:10 PM, Waiman Long wrote:
> On 7/24/20 4:16 AM, Will Deacon wrote:
>> On Thu, Jul 23, 2020 at 08:47:59PM +0200, peterz@infradead.org wrote:
>>> On Thu, Jul 23, 2020 at 02:32:36PM -0400, Waiman Long wrote:
>>>> BTW, do you have any comment on my v2 lock holder cpu info
>>>> qspinlock patch?
>>>> I will have to update the patch to fix the reported 0-day test
>>>> problem, but
>>>> I want to collect other feedback before sending out v3.
>>> I want to say I hate it all, it adds instructions to a path we spend an
>>> aweful lot of time optimizing without really getting anything back for
>>> it.
>>>
>>> Will, how do you feel about it?
>> I can see it potentially being useful for debugging, but I hate the
>> limitation to 256 CPUs. Even arm64 is hitting that now.
>
> After thinking more about that, I think we can use all the remaining
> bits in the 16-bit locked_pending. Reserving 1 bit for locked and 1
> bit for pending, there are 14 bits left. So as long as NR_CPUS < 16k
> (requirement for 16-bit locked_pending), we can put all possible cpu
> numbers into the lock. We can also just use smp_processor_id() without
> additional percpu data.
Sorry, that doesn't work. The extra bits in the pending byte won't get
cleared on unlock. That will have noticeable performance impact.
Clearing the pending byte on unlock will cause other performance
problem. So I guess we will have to limit the cpu number in the locked byte.
Regards,
Longman
WARNING: multiple messages have this Message-ID (diff)
From: Waiman Long <longman@redhat.com>
To: Will Deacon <will@kernel.org>, peterz@infradead.org
Cc: Michael Ellerman <mpe@ellerman.id.au>,
Nicholas Piggin <npiggin@gmail.com>,
linuxppc-dev@lists.ozlabs.org, Boqun Feng <boqun.feng@gmail.com>,
Ingo Molnar <mingo@redhat.com>,
Anton Blanchard <anton@ozlabs.org>,
linux-kernel@vger.kernel.org,
virtualization@lists.linux-foundation.org,
kvm-ppc@vger.kernel.org, linux-arch@vger.kernel.org
Subject: Re: [PATCH v3 5/6] powerpc/pseries: implement paravirt qspinlocks for SPLPAR
Date: Sat, 25 Jul 2020 03:02:58 +0000 [thread overview]
Message-ID: <ccf0c6a6-b7c3-8909-cc8f-0c5e7434c372@redhat.com> (raw)
In-Reply-To: <8532332b-85dd-661b-cf72-81a8ceb70747@redhat.com>
On 7/24/20 3:10 PM, Waiman Long wrote:
> On 7/24/20 4:16 AM, Will Deacon wrote:
>> On Thu, Jul 23, 2020 at 08:47:59PM +0200, peterz@infradead.org wrote:
>>> On Thu, Jul 23, 2020 at 02:32:36PM -0400, Waiman Long wrote:
>>>> BTW, do you have any comment on my v2 lock holder cpu info
>>>> qspinlock patch?
>>>> I will have to update the patch to fix the reported 0-day test
>>>> problem, but
>>>> I want to collect other feedback before sending out v3.
>>> I want to say I hate it all, it adds instructions to a path we spend an
>>> aweful lot of time optimizing without really getting anything back for
>>> it.
>>>
>>> Will, how do you feel about it?
>> I can see it potentially being useful for debugging, but I hate the
>> limitation to 256 CPUs. Even arm64 is hitting that now.
>
> After thinking more about that, I think we can use all the remaining
> bits in the 16-bit locked_pending. Reserving 1 bit for locked and 1
> bit for pending, there are 14 bits left. So as long as NR_CPUS < 16k
> (requirement for 16-bit locked_pending), we can put all possible cpu
> numbers into the lock. We can also just use smp_processor_id() without
> additional percpu data.
Sorry, that doesn't work. The extra bits in the pending byte won't get
cleared on unlock. That will have noticeable performance impact.
Clearing the pending byte on unlock will cause other performance
problem. So I guess we will have to limit the cpu number in the locked byte.
Regards,
Longman
next prev parent reply other threads:[~2020-07-25 3:03 UTC|newest]
Thread overview: 153+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-07-06 4:35 [PATCH v3 0/6] powerpc: queued spinlocks and rwlocks Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 4:35 ` [PATCH v3 1/6] powerpc/powernv: must include hvcall.h to get PAPR defines Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-09 10:05 ` Michael Ellerman
2020-07-09 10:05 ` Michael Ellerman
2020-07-09 10:05 ` Michael Ellerman
2020-07-09 10:05 ` Michael Ellerman
2020-07-09 10:05 ` Michael Ellerman
2020-07-06 4:35 ` [PATCH v3 2/6] powerpc/pseries: move some PAPR paravirt functions to their own file Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-09 10:11 ` Michael Ellerman
2020-07-09 10:11 ` Michael Ellerman
2020-07-09 10:11 ` Michael Ellerman
2020-07-09 10:11 ` Michael Ellerman
2020-07-09 10:11 ` Michael Ellerman
2020-07-06 4:35 ` [PATCH v3 3/6] powerpc: move spinlock implementation to simple_spinlock Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-09 10:15 ` Michael Ellerman
2020-07-09 10:15 ` Michael Ellerman
2020-07-09 10:15 ` Michael Ellerman
2020-07-09 10:15 ` Michael Ellerman
2020-07-09 10:15 ` Michael Ellerman
2020-07-06 4:35 ` [PATCH v3 4/6] powerpc/64s: implement queued spinlocks and rwlocks Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-09 10:20 ` Michael Ellerman
2020-07-09 10:20 ` Michael Ellerman
2020-07-09 10:20 ` Michael Ellerman
2020-07-09 10:20 ` Michael Ellerman
2020-07-09 10:20 ` Michael Ellerman
2020-07-09 10:33 ` Peter Zijlstra
2020-07-09 10:33 ` Peter Zijlstra
2020-07-09 10:33 ` Peter Zijlstra
2020-07-23 14:37 ` Michal Suchánek
2020-07-23 14:37 ` Michal Suchánek
2020-07-23 14:37 ` Michal Suchánek
2020-07-06 4:35 ` [PATCH v3 5/6] powerpc/pseries: implement paravirt qspinlocks for SPLPAR Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-09 10:53 ` Michael Ellerman
2020-07-09 10:53 ` Michael Ellerman
2020-07-09 10:53 ` Michael Ellerman
2020-07-09 10:53 ` Michael Ellerman
2020-07-09 10:53 ` Michael Ellerman
2020-07-09 11:03 ` Peter Zijlstra
2020-07-09 11:03 ` Peter Zijlstra
2020-07-09 11:03 ` Peter Zijlstra
2020-07-09 16:06 ` Waiman Long
2020-07-09 16:06 ` Waiman Long
2020-07-09 16:06 ` Waiman Long
2020-07-23 14:00 ` Peter Zijlstra
2020-07-23 14:00 ` Peter Zijlstra
2020-07-23 14:00 ` Peter Zijlstra
2020-07-23 18:32 ` Waiman Long
2020-07-23 18:32 ` Waiman Long
2020-07-23 18:32 ` Waiman Long
2020-07-23 18:47 ` peterz
2020-07-23 18:47 ` peterz
2020-07-23 18:47 ` peterz
2020-07-23 18:47 ` peterz
2020-07-23 19:04 ` Waiman Long
2020-07-23 19:04 ` Waiman Long
2020-07-23 19:04 ` Waiman Long
2020-07-23 19:58 ` peterz
2020-07-23 19:58 ` peterz
2020-07-23 19:58 ` peterz
2020-07-23 20:30 ` Segher Boessenkool
2020-07-23 20:30 ` Segher Boessenkool
2020-07-23 20:30 ` Segher Boessenkool
2020-07-23 21:58 ` Waiman Long
2020-07-23 21:58 ` Waiman Long
2020-07-23 21:58 ` Waiman Long
2020-07-24 8:16 ` Will Deacon
2020-07-24 8:16 ` Will Deacon
2020-07-24 8:16 ` Will Deacon
2020-07-24 19:10 ` Waiman Long
2020-07-24 19:10 ` Waiman Long
2020-07-24 19:10 ` Waiman Long
2020-07-25 3:02 ` Waiman Long [this message]
2020-07-25 3:02 ` Waiman Long
2020-07-25 3:02 ` Waiman Long
2020-07-25 17:26 ` Peter Zijlstra
2020-07-25 17:26 ` Peter Zijlstra
2020-07-25 17:26 ` Peter Zijlstra
2020-07-25 17:26 ` Peter Zijlstra
2020-07-25 17:36 ` Waiman Long
2020-07-25 17:36 ` Waiman Long
2020-07-25 17:36 ` Waiman Long
2020-07-23 14:09 ` Nicholas Piggin
2020-07-23 14:09 ` Nicholas Piggin
2020-07-23 14:09 ` Nicholas Piggin
2020-07-06 4:35 ` [PATCH v3 6/6] powerpc/qspinlock: optimised atomic_try_cmpxchg_lock that adds the lock hint Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 4:35 ` Nicholas Piggin
2020-07-06 18:39 ` [PATCH v3 0/6] powerpc: queued spinlocks and rwlocks Waiman Long
2020-07-06 18:39 ` Waiman Long
2020-07-06 18:39 ` Waiman Long
2020-07-07 5:57 ` Nicholas Piggin
2020-07-07 5:57 ` Nicholas Piggin
2020-07-07 5:57 ` Nicholas Piggin
2020-07-08 3:33 ` Waiman Long
2020-07-08 3:33 ` Waiman Long
2020-07-08 3:33 ` Waiman Long
2020-07-08 5:10 ` Nicholas Piggin
2020-07-08 5:10 ` Nicholas Piggin
2020-07-08 5:10 ` Nicholas Piggin
2020-07-08 23:50 ` Waiman Long
2020-07-08 23:50 ` Waiman Long
2020-07-08 23:50 ` Waiman Long
2020-07-08 23:58 ` Waiman Long
2020-07-08 23:58 ` Waiman Long
2020-07-08 23:58 ` Waiman Long
2020-07-08 8:32 ` Peter Zijlstra
2020-07-08 8:32 ` Peter Zijlstra
2020-07-08 8:32 ` Peter Zijlstra
2020-07-08 8:32 ` Peter Zijlstra
2020-07-08 23:53 ` Waiman Long
2020-07-08 23:53 ` Waiman Long
2020-07-08 23:53 ` Waiman Long
2020-07-08 8:41 ` Peter Zijlstra
2020-07-08 8:41 ` Peter Zijlstra
2020-07-08 8:41 ` Peter Zijlstra
2020-07-08 23:54 ` Waiman Long
2020-07-08 23:54 ` Waiman Long
2020-07-08 23:54 ` Waiman Long
2020-07-09 8:31 ` Peter Zijlstra
2020-07-09 8:31 ` Peter Zijlstra
2020-07-09 8:31 ` Peter Zijlstra
2020-07-09 8:31 ` Peter Zijlstra
2020-07-21 11:20 ` Nicholas Piggin
2020-07-21 11:20 ` Nicholas Piggin
2020-07-21 11:20 ` Nicholas Piggin
2020-07-21 11:08 ` Nicholas Piggin
2020-07-21 11:08 ` Nicholas Piggin
2020-07-21 11:08 ` Nicholas Piggin
2020-07-21 11:08 ` Nicholas Piggin
2020-07-21 14:36 ` Waiman Long
2020-07-21 14:36 ` Waiman Long
2020-07-21 14:36 ` Waiman Long
2020-07-23 13:30 ` Nicholas Piggin
2020-07-23 13:30 ` Nicholas Piggin
2020-07-23 13:30 ` Nicholas Piggin
2020-07-23 14:29 ` Waiman Long
2020-07-23 14:29 ` Waiman Long
2020-07-23 14:29 ` Waiman Long
2020-07-23 16:12 ` Nicholas Piggin
2020-07-23 16:12 ` Nicholas Piggin
2020-07-23 16:12 ` Nicholas Piggin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ccf0c6a6-b7c3-8909-cc8f-0c5e7434c372@redhat.com \
--to=longman@redhat.com \
--cc=anton@ozlabs.org \
--cc=boqun.feng@gmail.com \
--cc=kvm-ppc@vger.kernel.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mingo@redhat.com \
--cc=mpe@ellerman.id.au \
--cc=npiggin@gmail.com \
--cc=peterz@infradead.org \
--cc=virtualization@lists.linux-foundation.org \
--cc=will@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.