[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <36ce6959-a4ed-bf33-9e7c-f35a63bf9434@huawei.com>
Date: Wed, 21 Sep 2022 18:54:01 +0800
From: "wangjie (L)" <wangjie125@...wei.com>
To: Peng Zhang <zhangpeng.00@...edance.com>, <joro@...tes.org>,
<will@...nel.org>
CC: <iommu@...ts.linux.dev>, <linux-kernel@...r.kernel.org>,
<robin.murphy@....com>, <haifeng.zhao@...ux.intel.com>,
<john.garry@...wei.com>
Subject: Re: [PATCH v2] iommu/iova: Optimize alloc_iova with rbtree_augmented
Tested-by: Jie Wang <wangjie125@...wei.com>
On 2022/9/21 11:55, Peng Zhang wrote:
>> This patch seems to solve the performance issues i have. Currently my
>> nic's rx performance is unstable in large-capacity scenarios. I
>> applied this patch to 5.19 rc4 and tested 8 times rx performance in
>> these scenes. Here are test results, "before" row is the result of
>> 5.19 rc4. "after" row means 5.19 rc4 with this patch, the unit is Mbits/s
>>
>> 1 2 3 4 5 6 7 8
>> before 55430 76701 84194 77560 88292 90106
>> 87770 77273
>> after 92770 92767 92792 92764 92742 92696
>> 92781 92756
>>
>> Obviously, after using this patch, the performance is stable.
>
> Thank you for your test. Can you add a "Tested-by" token?
> I plan to update the commit message in the next version.
> .
>
Powered by blists - more mailing lists