[dpdk-dev] [dpdk-stable] [PATCH v4] mbuf: fix reset on mbuf free

[Ali Alnubani]

> -----Original Message-----
> From: Olivier Matz <olivier.matz at 6wind.com>
> Sent: Thursday, September 30, 2021 12:39 AM
> To: Ali Alnubani <alialnu at nvidia.com>
> Cc: dev at dpdk.org; David Marchand <david.marchand at redhat.com>;
> Alexander Kozyrev <akozyrev at nvidia.com>; NBU-Contact-Thomas Monjalon
> <thomas at monjalon.net>; Ferruh Yigit <ferruh.yigit at intel.com>; Slava
> Ovsiienko <viacheslavo at nvidia.com>; zhaoyan.chen at intel.com; Morten
> Brørup <mb at smartsharesystems.com>; Andrew Rybchenko
> <andrew.rybchenko at oktetlabs.ru>; Ananyev, Konstantin
> <konstantin.ananyev at intel.com>; Ajit Khaparde
> <ajit.khaparde at broadcom.com>; jerinj at marvell.com
> Subject: Re: [dpdk-dev] [dpdk-stable] [PATCH v4] mbuf: fix reset on mbuf
> free
> 
> Hi Ali,
> 
> 
> On Wed, Sep 29, 2021 at 08:03:17AM +0000, Ali Alnubani wrote:
> > Hi Olivier,
> >
> > I wanted to retest the patch on latest main, but it no longer applies, could
> you please rebase it?
> 
> I rebased the patch:
> https://patchwork.dpdk.org/project/dpdk/patch/20210929213707.17727-1-
> olivier.matz at 6wind.com/
> 

Thanks for the update,
I tested v5 on main (84b3e4555a1f) and the regression we previously saw no longer reproduces.

Will the patch be backported to 20.11? I can still reproduce a 2% drop in single core packet forwarding performance with v4 applied on v20.11.3.

Testpmd: dpdk-testpmd -n 4  -a 0000:82:00.0  -a 0000:82:00.1 -c 0x9000  -- -i  --nb-cores=1  --txd=256 --rxd=256 --auto-start
Traffic generator is TREX with pattern (Ether / IP / UDP).
Drop is from ~50.5 mpps to ~49.5 mpps.

OS: 18.04.5 LTS
Device: ConnectX-5
MLNX_OFED: MLNX_OFED_LINUX-5.4-1.0.3.0
Firmware: 16.31.1014

Regards,
Ali