Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.32 as permitted sender) client-ip=23.128.96.32;
Date:   Wed, 25 Oct 2023 13:52:37 -0700
From:   Tony Luck <tony.luck@intel.com>
To:     "Moger, Babu" <babu.moger@amd.com>
Cc:     Peter Newman <peternewman@google.com>,
        Fenghua Yu <fenghua.yu@intel.com>,
        Reinette Chatre <reinette.chatre@intel.com>,
        Jonathan Corbet <corbet@lwn.net>,
        Shuah Khan <skhan@linuxfoundation.org>, x86@kernel.org,
        Shaopeng Tan <tan.shaopeng@fujitsu.com>,
        James Morse <james.morse@arm.com>,
        Jamie Iles <quic_jiles@quicinc.com>,
        Randy Dunlap <rdunlap@infradead.org>,
        linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
        patches@lists.linux.dev
Subject: Re: [PATCH] x86/resctrl: mba_MBps: Fall back to total b/w if local
 b/w unavailable
Message-ID: <ZTmAFVuYlMuCbQHz@agluck-desk3>
References: <20231024181600.8270-1-tony.luck@intel.com>
 <CALPaoChftF-H6GauKq4-c_qBJP1GJbR3-ByE5krsaQF4y4y9oQ@mail.gmail.com>
 <ZTluypa9bCWv4k2n@agluck-desk3>
 <e4994218-a7a2-4505-868e-a5c825ca4db0@amd.com>
 <b8ea0a74-347d-475f-a36d-8944ced16951@amd.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <b8ea0a74-347d-475f-a36d-8944ced16951@amd.com>
Precedence: bulk

On Wed, Oct 25, 2023 at 03:42:14PM -0500, Moger, Babu wrote:
> I meant, I was thinking bit different.
> 
> > 
> > You need these changes in only two functions, mbm_bw_count and
> > update_mba_bw. You decide which event you want to use based on availability,
> > 
> > Something like this. I updated mbm_bw_count.
> > 
> > diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c
> > b/arch/x86/kernel/cpu/resctrl/monitor.c
> > index 0ad23475fe16..302993e4fbc3 100644
> > --- a/arch/x86/kernel/cpu/resctrl/monitor.c
> > +++ b/arch/x86/kernel/cpu/resctrl/monitor.c
> > @@ -436,8 +436,16 @@ static int __mon_event_count(u32 rmid, struct
> > rmid_read *rr)
> >   */
> >  static void mbm_bw_count(u32 rmid, struct rmid_read *rr)
> >  {
> > -       struct mbm_state *m = &rr->d->mbm_local[rmid];
> >         u64 cur_bw, bytes, cur_bytes;
> > +       struct mbm_state *m;
> > +       int evtid;
> > +
> > +       if (is_mbm_local_enabled())
> > +               evtid = QOS_L3_MBM_LOCAL_EVENT_ID;
> > +       else
> > +               evtid = QOS_L3_MBM_TOTAL_EVENT_ID;
> > +
> > +       m = get_mbm_state(rr->d, rmid, evtid);

Ok. Yes. That seems simpler.

Maybe I should just set a global "mbm_evtid" at mount
time. No need to check every time to see if is_mbm_local_enabled()
somehow changed and local b/w measurements were suddenly
available!

-Tony