From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ferruh Yigit Subject: Re: [PATCH 00/24] Refactor mlx5 to improve performance Date: Mon, 20 Jun 2016 16:03:00 +0100 Message-ID: <576805A4.1030204@intel.com> References: <1465379291-25310-1-git-send-email-nelio.laranjeiro@6wind.com> <576420C7.6010301@intel.com> <20160620073854.GE14221@autoinstall.dev.6wind.com> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: quoted-printable Cc: dev@dpdk.org To: =?UTF-8?Q?N=c3=a9lio_Laranjeiro?= Return-path: Received: from mga03.intel.com (mga03.intel.com [134.134.136.65]) by dpdk.org (Postfix) with ESMTP id 5878295DE for ; Mon, 20 Jun 2016 17:03:03 +0200 (CEST) In-Reply-To: <20160620073854.GE14221@autoinstall.dev.6wind.com> List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" On 6/20/2016 8:38 AM, N=E9lio Laranjeiro wrote: > On Fri, Jun 17, 2016 at 05:09:43PM +0100, Ferruh Yigit wrote: >> On 6/8/2016 10:47 AM, Nelio Laranjeiro wrote: >>> Enhance mlx5 with a data path that bypasses Verbs. >>> >>> The first half of this patchset removes support for functionality com= pletely >>> rewritten in the second half (scatter/gather, inline send), while the= data >>> path is refactored without Verbs. >>> >>> The PMD remains usable during the transition. >>> >>> This patchset must be applied after "Miscellaneous fixes for mlx4 and= mlx5". >>> >>> Adrien Mazarguil (8): >>> mlx5: replace countdown with threshold for TX completions >>> mlx5: add debugging information about TX queues capabilities >>> mlx5: check remaining space while processing TX burst >>> mlx5: resurrect TX gather support >>> mlx5: work around spurious compilation errors >>> mlx5: remove redundant RX queue initialization code >>> mlx5: make RX queue reinitialization safer >>> mlx5: resurrect RX scatter support >>> >>> Nelio Laranjeiro (15): >>> mlx5: split memory registration function for better performance >>> mlx5: remove TX gather support >>> mlx5: remove RX scatter support >>> mlx5: remove configuration variable for maximum number of segments >>> mlx5: remove inline TX support >>> mlx5: split TX queue structure >>> mlx5: split RX queue structure >>> mlx5: update prerequisites for upcoming enhancements >>> mlx5: add definitions for data path without Verbs >>> mlx5: add support for configuration through kvargs >>> mlx5: add TX/RX burst function selection wrapper >>> mlx5: refactor RX data path >>> mlx5: refactor TX data path >>> mlx5: handle RX CQE compression >>> mlx5: add support for multi-packet send >>> >>> Yaacov Hazan (1): >>> mlx5: add support for inline send >>> >> >> I run basic checks to the patchset: >> >> There are various checkpatch warnings, all are warning or check level >> >> Patch 8 and 13 failed to apply with via git, -looks line line numbers >> shifted a little, this is not a problem since eventually it applies bu= t >> just for your information. >> >> check-git-log is giving following errors, it is mainly case issue in R= x/Tx: >> Wrong headline lowercase: >> mlx5: resurrect RX scatter support >> mlx5: make RX queue reinitialization safer >> mlx5: remove redundant RX queue initialization code >> mlx5: resurrect TX gather support >> mlx5: check remaining space while processing TX burst >> mlx5: add debugging information about TX queues capabilities >> mlx5: replace countdown with threshold for TX completions >> mlx5: handle RX CQE compression >> mlx5: refactor RX data path >> mlx5: add TX/RX burst function selection wrapper >> mlx5: split RX queue structure >> mlx5: split TX queue structure >> mlx5: remove inline TX support >> mlx5: remove RX scatter support >> mlx5: remove TX gather support >> Headline too long: >> mlx5: remove configuration variable for maximum number of segm= ents >> mlx5: split memory registration function for better performanc= e >> >> >> It compiles fine. >> >> Regards, >> ferruh >=20 > Hi ferruh, >=20 > In fact, It does not apply well on top the current DPDK master branch. >=20 I did able to apply on top of rel_16_07 branch using "patch" binary. but if you think it doesn't apply well, any plan to send a new version? Thanks, ferruh