From: "Doug Smythies" <dsmythies@telus.net>
To: "'Borislav Petkov'" <bp@suse.de>,
"'Artem Bityutskiy'" <dedekind1@gmail.com>
Cc: "'Rafael J. Wysocki'" <rafael@kernel.org>,
"'Len Brown'" <lenb@kernel.org>,
"'Linux PM Mailing List'" <linux-pm@vger.kernel.org>
Subject: RE: [PATCH] tools/power/turbostat: fix compatibility with older kernels
Date: Wed, 27 Jan 2021 12:15:10 -0800 [thread overview]
Message-ID: <002201d6f4e9$1e9f6c10$5bde4430$@net> (raw)
In-Reply-To: <20210127185957.GD8115@zn.tnic>
On 2021.01.27 Borislav Petkov wrote:
> On Wed, Jan 27, 2021 at 03:24:44PM +0200, Artem Bityutskiy wrote:
> > From: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
> >
> > Commit
> >
> > 6d6501d912a9 tools/power/turbostat: Read energy_perf_bias from sysfs
> >
> > added a useful EPB print by reading EPB from sysfs's 'energy_perf_bias' file.
> > However, older kernels, which do not necessarily have that sysfs file (e.g.,
> > Centos 7's stock kernel does not have it). As a result, turbostat fails with
> > older kernels.
> >
> > This patch fixes the problem by ignoring the sysfs file read errors.
> >
> > Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
> > ---
> > tools/power/x86/turbostat/turbostat.c | 4 +++-
> > 1 file changed, 3 insertions(+), 1 deletion(-)
> >
> > diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c
> > index 389ea5209a83..12e014f2c24b 100644
> > --- a/tools/power/x86/turbostat/turbostat.c
> > +++ b/tools/power/x86/turbostat/turbostat.c
> > @@ -1839,7 +1839,9 @@ int get_epb(int cpu)
> >
> > sprintf(path, "/sys/devices/system/cpu/cpu%d/power/energy_perf_bias", cpu);
> >
> > - fp = fopen_or_die(path, "r");
> > + fp = fopen(path, "r");
> > + if (!fp)
> > + return -1;
> >
> > ret = fscanf(fp, "%d", &epb);
> > if (ret != 1)
>
> So I was under the impression that things in tools/ are tied to the
> kernel version they're shipped with. Which means, you should probably
> get the one from the centos kernel.
>
> However, if this is supposed to work on older kernels too, then
It is supposed to work with old kernels. Here is a quote from Len from [1] (2019.09.05) :
> The latest turbostat and x86_energy_perf_policy utilities
> in the upstream kernel tree should always be backward
> compatible with all old kernels. If that is EVER not the
> case, I want to know about it.
>
> Yes, I know that some distros ship old versions of these
> utilities built out of their matching kernel tree snapshots.
> Yes, applying upstream fixes to .stable for such distros is a good thing.
>
> However, the better solution for these particular utilities, is that
> they simply always use upstream utilities -- even with old kernels.
>
> When somebody reports a problem and I need them to run these tools,
> 100% of the time, I start by sending them the latest upstream version
> to replace the old version shipped by the distro.
Which is also what I do. I was also trying, so far without success, to
get a distro to relax its stringent, unnecessary, turbostat dependency
checking [2].
>
> 6d6501d912a9 ("tools/power/turbostat: Read energy_perf_bias from sysfs")
>
> would need to be extended to first test the sysfs file's existence and
> if not there, fall back to the MSR reading...
Yes agree, the information is still required.
[1] https://marc.info/?l=linux-pm&m=156770359620861&w=2
[2] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1844201
next prev parent reply other threads:[~2021-01-27 20:15 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-27 13:24 [PATCH] tools/power/turbostat: fix compatibility with older kernels Artem Bityutskiy
2021-01-27 18:59 ` Borislav Petkov
2021-01-27 20:15 ` Doug Smythies [this message]
2021-01-27 20:33 ` Borislav Petkov
2021-01-28 17:10 ` Borislav Petkov
2021-02-03 11:06 ` [tip: x86/urgent] tools/power/turbostat: Fallback to an MSR read for EPB tip-bot2 for Borislav Petkov
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='002201d6f4e9$1e9f6c10$5bde4430$@net' \
--to=dsmythies@telus.net \
--cc=bp@suse.de \
--cc=dedekind1@gmail.com \
--cc=lenb@kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=rafael@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.