[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20161005132658.GC30363@kernel.org>
Date: Wed, 5 Oct 2016 10:26:58 -0300
From: Arnaldo Carvalho de Melo <acme@...nel.org>
To: Jiri Olsa <jolsa@...hat.com>
Cc: Jiri Olsa <jolsa@...nel.org>, lkml <linux-kernel@...r.kernel.org>,
Don Zickus <dzickus@...hat.com>, Joe Mario <jmario@...hat.com>,
Ingo Molnar <mingo@...nel.org>,
Peter Zijlstra <a.p.zijlstra@...llo.nl>,
Namhyung Kim <namhyung@...nel.org>,
David Ahern <dsahern@...il.com>,
Andi Kleen <andi@...stfloor.org>
Subject: Re: [PATCH 20/57] perf c2c report: Add dcacheline dimension key
Em Wed, Oct 05, 2016 at 03:09:29PM +0200, Jiri Olsa escreveu:
> On Wed, Oct 05, 2016 at 02:45:37PM +0200, Jiri Olsa wrote:
>
> SNIP
>
> > > > +
> > > > + if (he->mem_info)
> > > > + addr = cl_address(he->mem_info->daddr.addr);
> > > > +
> > > > + return snprintf(hpp->buf, hpp->size, "%*s", width, hex_str(addr));
> > >
> > > So here you get that static buffer and then truncate it? Wouldn't the
> > > perf_hpp stuff take care of this? Can't we stop using that static buffer
> > > and this truncation at such a level?
> >
> > I think we need to cut it on this level, but I actualy might recall some
> > change you did for perf_hpp to cut this on column width later on?
> >
> > I'll check on that..
>
> ok, so it's cut later on, but it allows only for left-side alignment
> while we use the right-side one
>
> if I leave it on perf_hpp to deal with it I end up with following output:
> (check the Cacheline column)
Which is not _that_ bad, I guess it gets like that because we expect
kernel addresses as well (longer)?
[root@...et ~]# grep icmp_rcv /proc/kallsyms | cut -d' ' -f 1 | wc -c
17
[root@...et ~]# echo -n .................. | wc -c
18
[root@...et ~]#
How to indicate to the hpp code that we want right alignment? Namhyung?
>
> # Total Rmt ----- LLC Load Hitm ----- ---- Store Reference ---- --- Load Dram ---- LLC Total ----- Core Load Hit
> # Index Cacheline records Hitm Total Lcl Rmt Total L1Hit L1Miss Lcl Rmt Ld Miss Loads FB L1
> # ..... .................. ....... ....... ....... ....... ....... ....... ....... ....... ........ ........ ....... ....... ....... ....... ..
> #
> 0 0x3d2e300 273 0.53% 44 22 22 40 40 0 0 0 22 233 107 78
> 1 0x3d001c0 68 0.51% 22 1 21 2 2 0 0 2 25 66 30 7
> 2 0x3d00200 165 0.48% 22 2 20 20 20 0 0 0 20 145 89 34
> 3 0x3d5ca80 22 0.41% 19 2 17 3 3 0 0 0 17 19 0 0
>
>
> while current code does:
>
> # Total Rmt ----- LLC Load Hitm ----- ---- Store Reference ---- --- Load Dram ---- LLC Total ----- Core Load Hit
> # Index Cacheline records Hitm Total Lcl Rmt Total L1Hit L1Miss Lcl Rmt Ld Miss Loads FB L1
> # ..... .................. ....... ....... ....... ....... ....... ....... ....... ....... ........ ........ ....... ....... ....... ....... ..
> #
> 0 0x3d2e300 273 0.53% 44 22 22 40 40 0 0 0 22 233 107 78
> 1 0x3d001c0 68 0.51% 22 1 21 2 2 0 0 2 25 66 30 7
> 2 0x3d00200 165 0.48% 22 2 20 20 20 0 0 0 20 145 89 34
> 3 0x3d5ca80 22 0.41% 19 2 17 3 3 0 0 0 17 19 0 0
>
>
> I'll make the snprintf/scnprintf replacement
> based on your acme/tmp.perf/c2c.2
Thanks,
> thanks,
> jirka
Powered by blists - more mailing lists