[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <PH0PR12MB5481A6D78F0304C5BDED25C8DC229@PH0PR12MB5481.namprd12.prod.outlook.com>
Date: Fri, 28 May 2021 03:42:57 +0000
From: Parav Pandit <parav@...dia.com>
To: Saeed Mahameed <saeed@...nel.org>,
"David S. Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>
CC: "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
Tariq Toukan <tariqt@...dia.com>, Eli Cohen <elic@...dia.com>
Subject: RE: [net-next 17/17] net/mlx5: Improve performance in SF allocation
> From: Saeed Mahameed <saeed@...nel.org>
> Sent: Friday, May 28, 2021 12:22 AM
>
> On Thu, 2021-05-27 at 05:36 +0000, Parav Pandit wrote:
> >
> > > From: Saeed Mahameed <saeed@...nel.org>
> > > Sent: Thursday, May 27, 2021 10:06 AM
> > > From: Eli Cohen <elic@...dia.com>
> > >
> > > + continue;
> > > }
> > > +
> > > + if (hwc->sfs[i].usr_sfnum == usr_sfnum)
> > > + return -EEXIST;
> > > }
> > > - return -ENOSPC;
> > > +
> > > + if (free_idx == -1)
> > > + return -ENOSPC;
> > > +
> > > + hwc->sfs[free_idx].usr_sfnum = usr_sfnum;
> > > + hwc->sfs[free_idx].allocated = true;
> > > + return 0;
> > This is incorrect. It must return the free_index and not zero.
>
> Thanks Parav, I will drop the patches,
> Please post those comments internally so Eli will handle, I waited for your
> comments internally for weeks.
My bad.
When regression tests started failing it caught my attention and triggered the review.
I will review them timely going forward.
Powered by blists - more mailing lists