netdev - Re: [net-next PATCH 06/10] octeontx2-af: Add NIX1 interfaces to NPC

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] [thread-next>] [day] [month] [year] [list]

Message-ID: <20201016104851.01ac62f4@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
Date:   Fri, 16 Oct 2020 10:48:51 -0700
From:   Jakub Kicinski <kuba@...nel.org>
To:     sundeep subbaraya <sundeep.lkml@...il.com>
Cc:     David Miller <davem@...emloft.net>, netdev@...r.kernel.org,
        rsaladi2@...vell.com, Sunil Kovvuri Goutham <sgoutham@...vell.com>,
        Subbaraya Sundeep <sbhatta@...vell.com>
Subject: Re: [net-next PATCH 06/10] octeontx2-af: Add NIX1 interfaces to NPC

On Fri, 16 Oct 2020 08:59:43 +0530 sundeep subbaraya wrote:
> On Thu, Oct 15, 2020 at 9:02 PM Jakub Kicinski <kuba@...nel.org> wrote:
> > On Thu, 15 Oct 2020 17:53:07 +0530 sundeep subbaraya wrote:  
> > > Hi Jakub,
> > >
> > > On Thu, Oct 15, 2020 at 8:18 AM Jakub Kicinski <kuba@...nel.org> wrote:  
> > > >
> > > > On Tue, 13 Oct 2020 15:56:28 +0530 sundeep.lkml@...il.com wrote:  
> > > > > -static const struct npc_mcam_kex npc_mkex_default = {
> > > > > +static struct npc_mcam_kex npc_mkex_default = {
> > > > >       .mkex_sign = MKEX_SIGN,
> > > > >       .name = "default",
> > > > >       .kpu_version = NPC_KPU_PROFILE_VER,  
> > > >
> > > > Why is this no longer constant? Are you modifying global data based
> > > > on the HW discovered in the system?  
> > >
> > > Yes. Due to an errata present on earlier silicons
> > > npc_mkex_default.keyx_cfg[NIX_INTF_TX]
> > > and npc_mkex_default.keyx_cfg[NIX_INTF_RX] needs to be identical.  
> >
> > Does this run on the SoC? Is there no possibility that the same kernel
> > will have to drive two different pieces of hardware?  
> 
> If kernel runs on SoC with errata present then
> npc_mkex_default.keyx_cfg[NIX_INTF_TX]
> is modified to be same as npc_mkex_default.keyx_cfg[NIX_INTF_RX]. And if errata
> is not applicable to SoC then npc_mkex_default.keyx_cfg[NIX_INTF_TX]
> is unchanged and the values present in TX and RX are programmed to TX and RX
> interface registers.

Let me rephrase the question - can the AF driver only run on the SoC 
or are there configurations in which host can control the AF?

I see that Kconfig depends on ARM64 but is that what you choose 
to support or a HW limitation?