linux-kernel - Re: [PATCH 0/6] mm: Remote LRU per-cpu pagevec cache/per-cpu page list drain support

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 21 Sep 2021 19:59:51 +0200
From:   Vlastimil Babka <vbabka@...e.cz>
To:     Nicolas Saenz Julienne <nsaenzju@...hat.com>,
        akpm@...ux-foundation.org, frederic@...nel.org
Cc:     linux-kernel@...r.kernel.org, linux-mm@...ck.org,
        tglx@...utronix.de, cl@...ux.com, peterz@...radead.org,
        juri.lelli@...hat.com, mingo@...hat.com, mtosatti@...hat.com,
        nilal@...hat.com, mgorman@...e.de, ppandit@...hat.com,
        williams@...hat.com, bigeasy@...utronix.de,
        anna-maria@...utronix.de, linux-rt-users@...r.kernel.org
Subject: Re: [PATCH 0/6] mm: Remote LRU per-cpu pagevec cache/per-cpu page
 list drain support

On 9/21/21 6:13 PM, Nicolas Saenz Julienne wrote:
> This series introduces an alternative locking scheme around mm/swap.c's per-cpu
> LRU pagevec caches and mm/page_alloc.c's per-cpu page lists which will allow
> for remote CPUs to drain them. Currently, only a local CPU is permitted to
> change its per-cpu lists, and it's expected to do so, on-demand, whenever a
> process demands it (by means of queueing an drain task on the local CPU). Most
> systems will handle this promptly, but it'll cause problems for NOHZ_FULL CPUs
> that can't take any sort of interruption without breaking their functional
> guarantees (latency, bandwidth, etc...). Having a way for these processes to
> remotely drain the lists themselves will make co-existing with isolated CPUs
> possible, at the cost of more constraining locks.
> 
> Fortunately for non-NOHZ_FULL users, the alternative locking scheme and remote
> drain code are conditional to a static key which is disabled by default. This
> guarantees minimal functional or performance regressions. The feature will only
> be enabled if NOHZ_FULL's initialization process was successful.
> 
> This work is based on a previous series by Thomas Gleixner, Anna-Maria
> Gleixner, and Sebastian Andrzej Siewior[1].

These days the pcplist protection is done by local_lock, which solved
the RT concerns. Probably a stupid/infeasible idea, but maybe what you
want to achieve could be more generally solved at the local_lock level?
That on NOHZ_FULL CPUs, local_locks could have this mode where they
could synchronize with remote cpus?

> [1] https://patchwork.kernel.org/project/linux-mm/patch/20190424111208.24459-3-bigeasy@linutronix.de/
> 
> Nicolas Saenz Julienne (6):
>   mm/swap: Introduce lru_cpu_needs_drain()
>   mm/swap: Introduce alternative per-cpu LRU cache locking
>   mm/swap: Allow remote LRU cache draining
>   mm/page_alloc: Introduce alternative per-cpu list locking
>   mm/page_alloc: Allow remote per-cpu page list draining
>   sched/isolation: Enable 'remote_pcpu_cache_access' on NOHZ_FULL
>     systems
> 
>  kernel/sched/isolation.c |   9 +-
>  mm/internal.h            |   2 +
>  mm/page_alloc.c          | 111 ++++++++++++++++-----
>  mm/swap.c                | 202 ++++++++++++++++++++++++++++++---------
>  4 files changed, 253 insertions(+), 71 deletions(-)
>