From: Tony Luck <tony.luck@intel.com>
To: Peter Newman <peternewman@google.com>
Cc: Reinette Chatre <reinette.chatre@intel.com>,
Fenghua Yu <fenghua.yu@intel.com>,
Jonathan Corbet <corbet@lwn.net>,
Shuah Khan <skhan@linuxfoundation.org>,
x86@kernel.org, James Morse <james.morse@arm.com>,
Jamie Iles <quic_jiles@quicinc.com>,
Babu Moger <babu.moger@amd.com>,
Randy Dunlap <rdunlap@infradead.org>,
"Shaopeng Tan (Fujitsu)" <tan.shaopeng@fujitsu.com>,
linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
patches@lists.linux.dev
Subject: Re: [PATCH v8 3/7] x86/resctrl: Refactor mbm_update()
Date: Thu, 14 Nov 2024 09:20:19 -0800 [thread overview]
Message-ID: <ZzYxUwwEEIOGdmT6@agluck-desk3> (raw)
In-Reply-To: <CALPaoCi94amaO4ALGhLPn=zWEJ3STJWyYid4L+kMjXYf9wKqAQ@mail.gmail.com>
On Thu, Nov 14, 2024 at 11:31:25AM +0100, Peter Newman wrote:
> At least for the purposes of reporting the mbps rate to userspace, my
> users will record from the files one per second, so it would be fine
> to just report Unavailable (or Unassigned when it's clear that the
> error is because counter wasn't unassigned) whenever either the
> current or previous reading was not successful. Then they can assume
> the value or error reported always refers to the most
> recently-completed one-second window.
First hack at keeping status for memory bandwidth values. Simple
state machine. Compile tested.
diff --git a/arch/x86/kernel/cpu/resctrl/internal.h b/arch/x86/kernel/cpu/resctrl/internal.h
index 6345ab3e0890..80de5c6b0754 100644
--- a/arch/x86/kernel/cpu/resctrl/internal.h
+++ b/arch/x86/kernel/cpu/resctrl/internal.h
@@ -359,14 +359,22 @@ struct rftype {
char *buf, size_t nbytes, loff_t off);
};
+enum mbm_state_status {
+ MBM_INVALID,
+ MBM_ONE_VALUE,
+ MBM_VALID
+};
+
/**
* struct mbm_state - status for each MBM counter in each domain
* @prev_bw_bytes: Previous bytes value read for bandwidth calculation
* @prev_bw: The most recent bandwidth in MBps
+ * @status: Validity of @prev_bw
*/
struct mbm_state {
- u64 prev_bw_bytes;
- u32 prev_bw;
+ u64 prev_bw_bytes;
+ u32 prev_bw;
+ enum mbm_state_status status;
};
/**
diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c b/arch/x86/kernel/cpu/resctrl/monitor.c
index da4ae21350c8..767a526af2f5 100644
--- a/arch/x86/kernel/cpu/resctrl/monitor.c
+++ b/arch/x86/kernel/cpu/resctrl/monitor.c
@@ -670,6 +670,17 @@ static void mbm_bw_count(u32 closid, u32 rmid, struct rmid_read *rr)
if (WARN_ON_ONCE(!m))
return;
+ if (rr->err || !rr->val) {
+ m->status = MBM_INVALID;
+ return;
+ }
+
+ if (m->status == MBM_INVALID) {
+ m->status = MBM_ONE_VALUE;
+ m->prev_bw_bytes = rr->val;
+ return;
+ }
+
cur_bytes = rr->val;
bytes = cur_bytes - m->prev_bw_bytes;
m->prev_bw_bytes = cur_bytes;
@@ -677,6 +688,7 @@ static void mbm_bw_count(u32 closid, u32 rmid, struct rmid_read *rr)
cur_bw = bytes / SZ_1M;
m->prev_bw = cur_bw;
+ m->status = MBM_VALID;
}
/*
@@ -781,6 +793,9 @@ static void update_mba_bw(struct rdtgroup *rgrp, struct rdt_mon_domain *dom_mbm)
if (WARN_ON_ONCE(!pmbm_data))
return;
+ if (pmbm_data->status != MBM_VALID)
+ return;
+
dom_mba = get_ctrl_domain_from_cpu(smp_processor_id(), r_mba);
if (!dom_mba) {
pr_warn_once("Failure to get domain for MBA update\n");
@@ -801,6 +816,9 @@ static void update_mba_bw(struct rdtgroup *rgrp, struct rdt_mon_domain *dom_mbm)
cmbm_data = get_mbm_state(dom_mbm, entry->closid, entry->mon.rmid, evt_id);
if (WARN_ON_ONCE(!cmbm_data))
return;
+ if (cmbm_data->status != MBM_VALID)
+ return;
+
cur_bw += cmbm_data->prev_bw;
}
next prev parent reply other threads:[~2024-11-14 17:20 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-29 17:28 [PATCH v8 0/7] x86/resctrl: mba_MBps enhancement Tony Luck
2024-10-29 17:28 ` [PATCH v8 1/7] x86/resctrl: Prepare for per-ctrl_mon group mba_MBps control Tony Luck
2024-11-01 22:03 ` Fenghua Yu
2024-11-01 22:40 ` Tony Luck
2024-11-12 19:24 ` Reinette Chatre
2024-10-29 17:28 ` [PATCH v8 2/7] x86/resctrl: Compute memory bandwidth for all supported events Tony Luck
2024-11-12 19:25 ` Reinette Chatre
2024-10-29 17:28 ` [PATCH v8 3/7] x86/resctrl: Refactor mbm_update() Tony Luck
2024-11-01 22:08 ` Fenghua Yu
2024-11-01 22:57 ` Tony Luck
2024-11-13 22:25 ` Reinette Chatre
2024-11-13 22:58 ` Tony Luck
2024-11-13 23:58 ` Reinette Chatre
2024-11-14 10:31 ` Peter Newman
2024-11-14 17:20 ` Tony Luck [this message]
2024-10-29 17:28 ` [PATCH v8 4/7] x86/resctrl: Relax checks for mba_MBps mount option Tony Luck
2024-10-29 17:28 ` [PATCH v8 5/7] x86/resctrl: Add "mba_MBps_event" file to ctrl_mon directories Tony Luck
2024-11-12 22:12 ` Reinette Chatre
2024-11-12 23:42 ` Luck, Tony
2024-11-13 0:20 ` Reinette Chatre
2024-11-13 0:53 ` Luck, Tony
2024-11-13 2:54 ` Reinette Chatre
2024-10-29 17:28 ` [PATCH v8 6/7] x86/resctrl: Add write option to "mba_MBps_event" file Tony Luck
2024-11-01 23:26 ` Fenghua Yu
2024-11-01 23:55 ` Luck, Tony
2024-11-02 0:57 ` Fenghua Yu
2024-11-12 22:00 ` Reinette Chatre
2024-11-12 23:57 ` Luck, Tony
2024-11-13 0:40 ` Reinette Chatre
2024-11-12 22:18 ` Reinette Chatre
2024-10-29 17:28 ` [PATCH v8 7/7] x86/resctrl: Document the new " Tony Luck
2024-11-12 22:25 ` Reinette Chatre
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZzYxUwwEEIOGdmT6@agluck-desk3 \
--to=tony.luck@intel.com \
--cc=babu.moger@amd.com \
--cc=corbet@lwn.net \
--cc=fenghua.yu@intel.com \
--cc=james.morse@arm.com \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=patches@lists.linux.dev \
--cc=peternewman@google.com \
--cc=quic_jiles@quicinc.com \
--cc=rdunlap@infradead.org \
--cc=reinette.chatre@intel.com \
--cc=skhan@linuxfoundation.org \
--cc=tan.shaopeng@fujitsu.com \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox