-----Original Message-----
From: John Garry <redacted>
Sent: 2021年1月20日 17:26
To: Joakim Zhang <redacted>; acme@kernel.org;
jolsa@redhat.com
Cc: will@kernel.org; linux-perf-users@vger.kernel.org; dl-linux-imx
[off-list ref]; mark.rutland@arm.com;
linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH V1 3/4] perf vendor events: Add JSON metrics for imx8mq
DDR Perf

On 20/01/2021 05:46, Joakim Zhang wrote:

quoted

quoted

-----Original Message-----
From: John Garry<redacted>
Sent: 2021年1月19日 20:27
To: Joakim Zhang<redacted>;acme@kernel.org;
jolsa@redhat.com
Cc:will@kernel.org;linux-perf-users@vger.kernel.org; dl-linux-imx
[off-list ref];mark.rutland@arm.com;
linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH V1 3/4] perf vendor events: Add JSON metrics for
imx8mq DDR Perf

On 19/01/2021 11:00, Joakim Zhang wrote:

quoted

quoted

-----Original Message-----
From: John Garry<redacted>
Sent: 2021年1月19日 18:40
To: Joakim Zhang<redacted>;acme@kernel.org;
jolsa@redhat.com
Cc:will@kernel.org;linux-perf-users@vger.kernel.org; dl-linux-imx
[off-list ref];mark.rutland@arm.com;
linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH V1 3/4] perf vendor events: Add JSON metrics
for imx8mq DDR Perf

On 19/01/2021 02:56, Joakim Zhang wrote:

quoted

index 000000000000..c3e9f89a4a4d

--- /dev/null
+++

b/tools/perf/pmu-events/arch/arm64/freescale/imx8mq/sys/metrics.

quoted

quoted

quoted

quoted

quoted

+++ js
+++ on

@@ -0,0 +1,26 @@
+[
+   {
+	    "BriefDescription": "bandwidth usage for lpddr4 evk board",

Why mention the evk board specifically? Is this metric only
specific to that board, and not all i.MX8MQ platforms?

Yes, John, we have lpddr4 evk board and ddr4 evk borad......

How is this supposed to work?

So perf matches on the identifier file for system PMUs; however the
contents of this file are based on the HW implementation for imx DDR PMU,

not the board.

quoted

Yes, I know this feature is for system PMU, not board. You know DDR

bandwidth calculation is based on DDR clock frequency, which is vary form
boards.

quoted

e.g.
lpddr4 board, DDR clock frequency is 750MHZ:
"MetricExpr": "(( imx8_ddr0@read\\-cycles@ + imx8_ddr0@write\\-cycles@ )

* 4 * 4 / duration_time) / (750 * 1000000 * 4 * 4)"

quoted

ddr4 board, DDR clock frequency is 600MHZ.
"MetricExpr": "(( imx8_ddr0@read\\-cycles@ + imx8_ddr0@write\\-cycles@ )

* 4 * 4 / duration_time) / (600 * 1000000 * 4 * 4)"

quoted

I give different metric names,
"MetricName": "imx8mm_bandwidth_usage.lpddr4"
"MetricName": "imx8mm_bandwidth_usage.ddr4"

But that is just not how things are done. We - or I - would expect that perf tool
would have the intelligence to know the appropriate metric for that specific
platform. Or, more specifically, we should not present the user with many
metrics and expect him/her to know the appropriate metric for the specific
board.

Are there other cases for other architectures where things are done like this?

I don't know if exist other cases. If you can’t accept it, I can only push part of it first, and let bandwidth metric locally.


Best Regards,
Joakim Zhang

Thanks,
John

quoted

So users can select correct bandwidth metric according to their own board

types. To a certain extent, it can also support some board-level metrics.

quoted

If it is okay for you, I will send a V2 to replace ( imx8_ddr0@read\\-cycles@ +

imx8_ddr0@write\\-cycles@ ) with ( imx8mm_ddr.read_cycles +
imx8mm_ddr.write_cycles ).

quoted

quoted

quoted

John, change to your new framework, I found it can't support all
features

which is supported in 5.10 kernel.

Yes, so let me know about all the problems.

It should have worked as 5.10 kernel at my side. Thanks.