netdev - Re: [PATCH][net-next] openvswitch: eliminate cpu_used_mask from sw

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] [day] [month] [year] [list]

Date:   Sat, 28 Jul 2018 23:12:43 -0700 (PDT)
From:   David Miller <davem@...emloft.net>
To:     lirongqing@...du.com
Cc:     netdev@...r.kernel.org, pshelar@....org
Subject: Re: [PATCH][net-next] openvswitch: eliminate cpu_used_mask from
 sw_flow

From: Li RongQing <lirongqing@...du.com>
Date: Fri, 27 Jul 2018 16:03:57 +0800

> The size of struct cpumask varies with CONFIG_NR_CPUS, some config
> CONFIG_NR_CPUS is very larger, like 5120, struct cpumask will take
> 640 bytes, if there is thousands of flows, it will take lots of
> memory
> 
> cpu_used_mask has two purposes
> 1: Assume first cpu as cpu0 which maybe not true; now use
>    cpumask_first(cpu_possible_mask)
> 2: when get/clear statistic, reduce the iteratation; but it
>    is not hot path, so use for_each_possible_cpu
> 
> Signed-off-by: Zhang Yu <zhangyu31@...du.com>
> Signed-off-by: Li RongQing <lirongqing@...du.com>

This seems to completely undo the optimization done by:

commit c4b2bf6b4a35348fe6d1eb06928eb68d7b9d99a9
Author: Tonghao Zhang <xiangxia.m.yue@...il.com>
Date:   Mon Jul 17 23:28:06 2017 -0700

    openvswitch: Optimize operations for OvS flow_stats.

And in that commit message it states clearly that flow_free()
performance matters, and that the iteration over cpu_possible_mask in
the for() loop is the problem.

At a minimum, we can't apply this unless you explain why the
above performance issue won't be reintroudced by your change.

Thank you.