pcp
[Top] [All Lists]

Re: [PATCH v3 1/3] perfevent_pmda: Add extension for perf derived events

To: Nathan Scott <nathans@xxxxxxxxxx>
Subject: Re: [PATCH v3 1/3] perfevent_pmda: Add extension for perf derived events
From: Hemant Kumar <hemant@xxxxxxxxxxxxxxxxxx>
Date: Tue, 16 Feb 2016 20:43:29 +0530
Cc: pcp@xxxxxxxxxxx
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <484836703.20809064.1455498547101.JavaMail.zimbra@xxxxxxxxxx>
References: <1455060499-14670-1-git-send-email-hemant@xxxxxxxxxxxxxxxxxx> <1779243884.20121961.1455156620367.JavaMail.zimbra@xxxxxxxxxx> <56BE6E0D.6050101@xxxxxxxxxxxxxxxxxx> <484836703.20809064.1455498547101.JavaMail.zimbra@xxxxxxxxxx>
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.7.0


On 02/15/2016 06:39 AM, Nathan Scott wrote:
Hi Hemant,

----- Original Message -----
[...]
One solution might be to add a metric permanently below that point,
that is instantiated both with and without derived perfevent counts
e.g. perfevent.derived.active (mirror the existing perfevent.active)
- a count of #derived metrics - zero valued in the default case.
Right. Sent a patch to add a metric perfevent.derived.active and
that should fix the warning.
http://www.pcp.io/pipermail/pcp/2016-February/009610.html

Thanks for the suggestion!
No problem, it works nicely here and fixes the issue.  I think this new
metric may need some help text though?  Not sure how best to add that in
with the scheme used, so I'll leave that for you too, if thats OK?  Taa.

Sure, I will add a help text for this new metric too.

--
Thanks,
Hemant Kumar

<Prev in Thread] Current Thread [Next in Thread>