Lets make wifi fast again!
 help / color / mirror / Atom feed
From: Jesper Dangaard Brouer <brouer@redhat.com>
To: Dave Taht <dave.taht@gmail.com>
Cc: make-wifi-fast@lists.bufferbloat.net, brouer@redhat.com
Subject: Re: [Make-wifi-fast] a bit of profiling on the archer
Date: Mon, 21 Nov 2016 17:24:18 +0100	[thread overview]
Message-ID: <20161121172418.04d17aad@redhat.com> (raw)
In-Reply-To: <CAA93jw6-fHLqLbq6fg7Q98kRMo6QqprRK0Kke96RrM8323A1Ow@mail.gmail.com>

On Sat, 19 Nov 2016 10:30:10 -0800
Dave Taht <dave.taht@gmail.com> wrote:

> Jesper/eric:
> 
> If you have any satisfying network oneliners and stats to monitor via
> perf along the lines of
> 
> http://www.brendangregg.com/perf.html#OneLiners
> 
> it would be helpful longer term.

Good point... I'll think where we can easily publish such oneliners.

> The mips platforms have sprouted more events than it ever had before
> (well, until last year, perf didn't work at all):
> 
> Of these besides the cpu-cycles thing, the only stuff I've looked at
> are various wake_tx_queue related events, and I have a scar involved
> in unaligned_accesses (but we're not doing any, soo)
> 
>   branch-instructions OR branches                    [Hardware event]
>   branch-misses                                      [Hardware event]
>   cpu-cycles OR cycles                               [Hardware event]
>   instructions                                       [Hardware event]

I'm very happy to see this is HW events in this platform! :-)
Guess, I should buy one of these ;-)

This is the Archer? which is what kind of CPU?


>   alignment-faults                                   [Software event]
>   bpf-output                                         [Software event]
>   context-switches OR cs                             [Software event]
>   cpu-clock                                          [Software event]
>   cpu-migrations OR migrations                       [Software event]
>   dummy                                              [Software event]
>   emulation-faults                                   [Software event]
>   major-faults                                       [Software event]
>   minor-faults                                       [Software event]
>   page-faults OR faults                              [Software event]
>   task-clock                                         [Software event]

>   L1-dcache-load-misses                              [Hardware cache event]
>   L1-dcache-loads                                    [Hardware cache event]
>   L1-dcache-store-misses                             [Hardware cache event]
>   L1-dcache-stores                                   [Hardware cache event]

>   L1-icache-load-misses                              [Hardware cache event]
>   L1-icache-loads                                    [Hardware cache event]
>   L1-icache-prefetches                               [Hardware cache event]

The icache HW events are actually quite interesting.  And three of
them, on my Intel Skylake CPU I only have "L1-icache-load-misses".


>   LLC-load-misses                                    [Hardware cache event]
>   LLC-loads                                          [Hardware cache event]
>   LLC-store-misses                                   [Hardware cache event]
>   LLC-stores                                         [Hardware cache event]

LLC = Last Level Cache.

Do you have any info on what the cache layout and sizes of this CPU is?
LLC indicate it might have a L2 cache?

>   branch-load-misses                                 [Hardware cache event]
>   branch-loads                                       [Hardware cache event]
>   iTLB-load-misses                                   [Hardware cache event]
>   iTLB-loads                                         [Hardware cache event]
>   rNNN                                               [Raw hardware event descriptor]
>   cpu/t1=v1[,t2=v2,t3 ...]/modifier                  [Raw hardware event descriptor]
>   mem:<addr>[/len][:access]                          [Hardware breakpoint]
>   ath10k:ath10k_htt_pktlog                           [Tracepoint event]
>   ath10k:ath10k_htt_rx_desc                          [Tracepoint event]

You can use tracepoints if you want to find/analyse very specific
issue, but not so useful for general perf monitoring.

-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Principal Kernel Engineer at Red Hat
  Author of http://www.iptv-analyzer.org
  LinkedIn: http://www.linkedin.com/in/brouer

  reply	other threads:[~2016-11-21 16:24 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-11-18  4:14 Dave Taht
2016-11-18  7:55 ` Jesper Dangaard Brouer
2016-11-18 10:26   ` Toke Høiland-Jørgensen
2016-11-18 14:13   ` Eric Dumazet
2016-11-18 14:30     ` Toke Høiland-Jørgensen
2016-11-19 17:33   ` Dave Taht
2016-11-21 16:49     ` Jesper Dangaard Brouer
2016-11-19 18:30   ` Dave Taht
2016-11-21 16:24     ` Jesper Dangaard Brouer [this message]
2016-11-21 16:36       ` Jonathan Morton

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

  List information: https://lists.bufferbloat.net/postorius/lists/make-wifi-fast.lists.bufferbloat.net/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20161121172418.04d17aad@redhat.com \
    --to=brouer@redhat.com \
    --cc=dave.taht@gmail.com \
    --cc=make-wifi-fast@lists.bufferbloat.net \
    /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