| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <225e5fba-f763-3a1e-062b-c74cdc5f83af@codeaurora.org>
Date: Sat, 28 Nov 2020 09:42:55 +0530
From: Neeraj Upadhyay <neeraju@...eaurora.org>
To: paulmck@...nel.org
Cc: rcu@...r.kernel.org, linux-kernel@...r.kernel.org,
kernel-team@...com, mingo@...nel.org, jiangshanlai@...il.com,
akpm@...ux-foundation.org, mathieu.desnoyers@...icios.com,
josh@...htriplett.org, tglx@...utronix.de, peterz@...radead.org,
rostedt@...dmis.org, dhowells@...hat.com, edumazet@...gle.com,
fweisbec@...il.com, oleg@...hat.com, joel@...lfernandes.org,
kent.overstreet@...il.com
Subject: Re: [PATCH v2 tip/core/rcu 1/6] srcu: Make Tiny SRCU use multi-bit
grace-period counter
On 11/28/2020 7:46 AM, Paul E. McKenney wrote:
> On Wed, Nov 25, 2020 at 10:03:26AM +0530, Neeraj Upadhyay wrote:
>>
>>
>> On 11/24/2020 10:48 AM, Neeraj Upadhyay wrote:
>>>
>>>
>>> On 11/24/2020 1:25 AM, Paul E. McKenney wrote:
>>>> On Mon, Nov 23, 2020 at 10:01:13AM +0530, Neeraj Upadhyay wrote:
>>>>> On 11/21/2020 6:29 AM, paulmck@...nel.org wrote:
>>>>>> From: "Paul E. McKenney" <paulmck@...nel.org>
>>>>>>
>>>>>> There is a need for a polling interface for SRCU grace periods. This
>>>>>> polling needs to distinguish between an SRCU instance being idle on the
>>>>>> one hand or in the middle of a grace period on the other. This commit
>>>>>> therefore converts the Tiny SRCU srcu_struct structure's srcu_idx from
>>>>>> a defacto boolean to a free-running counter, using the bottom bit to
>>>>>> indicate that a grace period is in progress. The second-from-bottom
>>>>>> bit is thus used as the index returned by srcu_read_lock().
>>>>>>
>>>>>> Link:
>>>>>> https://lore.kernel.org/rcu/20201112201547.GF3365678@moria.home.lan/
>>>>>> Reported-by: Kent Overstreet <kent.overstreet@...il.com>
>>>>>> [ paulmck: Fix __srcu_read_lock() idx computation Neeraj per
>>>>>> Upadhyay. ]
>>>>>> Signed-off-by: Paul E. McKenney <paulmck@...nel.org>
>>>>>> ---
>>>>>> include/linux/srcutiny.h | 4 ++--
>>>>>> kernel/rcu/srcutiny.c | 5 +++--
>>>>>> 2 files changed, 5 insertions(+), 4 deletions(-)
>>>>>>
>>>>>> diff --git a/include/linux/srcutiny.h b/include/linux/srcutiny.h
>>>>>> index 5a5a194..d9edb67 100644
>>>>>> --- a/include/linux/srcutiny.h
>>>>>> +++ b/include/linux/srcutiny.h
>>>>>> @@ -15,7 +15,7 @@
>>>>>> struct srcu_struct {
>>>>>> short srcu_lock_nesting[2]; /* srcu_read_lock()
>>>>>> nesting depth. */
>>>>>> - short srcu_idx; /* Current reader array element. */
>>>>>> + unsigned short srcu_idx; /* Current reader array
>>>>>> element in bit 0x2. */
>>>>>> u8 srcu_gp_running; /* GP workqueue running? */
>>>>>> u8 srcu_gp_waiting; /* GP waiting for readers? */
>>>>>> struct swait_queue_head srcu_wq;
>>>>>> @@ -59,7 +59,7 @@ static inline int __srcu_read_lock(struct
>>>>>> srcu_struct *ssp)
>>>>>> {
>>>>>> int idx;
>>>>>> - idx = READ_ONCE(ssp->srcu_idx);
>>>>>> + idx = ((READ_ONCE(ssp->srcu_idx) + 1) & 0x2) >> 1;
>>>>>> WRITE_ONCE(ssp->srcu_lock_nesting[idx],
>>>>>> ssp->srcu_lock_nesting[idx] + 1);
>>>>>> return idx;
>>>>>> }
>>>>>
>>>>> Need change in idx calcultion in srcu_torture_stats_print() ?
>>>>>
>>>>> static inline void srcu_torture_stats_print(struct srcu_struct *ssp,
>>>>> idx = READ_ONCE(ssp->srcu_idx) & 0x1;
>>>>
>>>> Excellent point! It should match the calculation in __srcu_read_lock(),
>>>> shouldn't it? I have updated this, thank you!
>>>>
>>>> Thanx, Paul
>>>>
>>>
>>> Updated version looks good!
>>>
>>>
>>> Thanks
>>> Neeraj
>>>
>>
>> For the version in rcu -dev:
>>
>> Reviewed-by: Neeraj Upadhyay <neeraju@...eaurora.org>
>
> I applied all of these, thank you very much!
>
Welcome :)
>> Only minor point which I have is, the idx calculation can be made an inline
>> func (though srcu_drive_gp() does not require a READ_ONCE for ->srcu_idx):
>>
>> __srcu_read_lock() and srcu_torture_stats_print() are using
>>
>> idx = ((READ_ONCE(ssp->srcu_idx) + 1) & 0x2) >> 1;
>>
>> whereas srcu_drive_gp() uses:
>>
>> idx = (ssp->srcu_idx & 0x2) / 2;
>
> They do work on different elements of the various arrays. Or do you
> believe that the srcu_drive_gp() use needs adjusting?
My bad, I missed that they are using different elements of array.
Please ignore this comment.
Thanks
Neeraj
>
> Either way, the overhead of READ_ONCE() is absolutely not at all
> a problem. Would you like to put together a patch so that I can see
> exactly what you are suggesting?
>
> Thanx, Paul
>
>> Thanks
>> Neeraj
>>
>>>>> Thanks
>>>>> Neeraj
>>>>>
>>>>>> diff --git a/kernel/rcu/srcutiny.c b/kernel/rcu/srcutiny.c
>>>>>> index 6208c1d..5598cf6 100644
>>>>>> --- a/kernel/rcu/srcutiny.c
>>>>>> +++ b/kernel/rcu/srcutiny.c
>>>>>> @@ -124,11 +124,12 @@ void srcu_drive_gp(struct work_struct *wp)
>>>>>> ssp->srcu_cb_head = NULL;
>>>>>> ssp->srcu_cb_tail = &ssp->srcu_cb_head;
>>>>>> local_irq_enable();
>>>>>> - idx = ssp->srcu_idx;
>>>>>> - WRITE_ONCE(ssp->srcu_idx, !ssp->srcu_idx);
>>>>>> + idx = (ssp->srcu_idx & 0x2) / 2;
>>>>>> + WRITE_ONCE(ssp->srcu_idx, ssp->srcu_idx + 1);
>>>>>> WRITE_ONCE(ssp->srcu_gp_waiting, true); /*
>>>>>> srcu_read_unlock() wakes! */
>>>>>> swait_event_exclusive(ssp->srcu_wq,
>>>>>> !READ_ONCE(ssp->srcu_lock_nesting[idx]));
>>>>>> WRITE_ONCE(ssp->srcu_gp_waiting, false); /*
>>>>>> srcu_read_unlock() cheap. */
>>>>>> + WRITE_ONCE(ssp->srcu_idx, ssp->srcu_idx + 1);
>>>>>> /* Invoke the callbacks we removed above. */
>>>>>> while (lh) {
>>>>>>
>>>>>
>>>>> --
>>>>> QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is
>>>>> a member of
>>>>> the Code Aurora Forum, hosted by The Linux Foundation
>>>
>>
>> --
>> QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of
>> the Code Aurora Forum, hosted by The Linux Foundation
--
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a
member of the Code Aurora Forum, hosted by The Linux Foundation
Powered by blists - more mailing lists