linux-kernel - Re: [PATCH v2] selftests/intel

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CABq9Dx5DDJXO3Xc+F_2o86xxfGJD35u39o1nmBkhbBqDo=HS5g@mail.gmail.com>
Date: Mon, 21 Oct 2024 23:33:43 +0200
From: Alessandro Zanni <alessandro.zanni87@...il.com>
To: Shuah Khan <skhan@...uxfoundation.org>
Cc: shuah@...nel.org, linux-kselftest@...r.kernel.org, 
	linux-kernel@...r.kernel.org, anupnewsmail@...il.com
Subject: Re: [PATCH v2] selftests/intel_pstate: fix operand expected

> On 10/21/24 09:04, Alessandro Zanni wrote:
> >> On 24/10/14 06:05, Shuah Khan wrote:
> >>> On 10/14/24 11:21, Alessandro Zanni wrote:
> >>>> This fix solves theses errors, when calling kselftest with
> >>>> targets "intel_pstate":
> >>>>
> >>>> ./run.sh: line 90: / 1000: syntax error: operand expected (error token is "/ 1000")
> >>>>
> >>>> ./run.sh: line 92: / 1000: syntax error: operand expected (error token is "/ 1000")
> >>>>
> >>>> To error was found by running tests manually with the command:
> >>>> make kselftest TARGETS=intel_pstate
> >>>>
> >>>> Signed-off-by: Alessandro Zanni <alessandro.zanni87@...il.com>
> >>>> ---
> >>>>
> >>>> Notes:
> >>>>       v2: removed debug echos
> >>>
> >>> See my comments on your v1. It would help to wait a bit
> >>> to send v2.
> >>
> >> Ok and thanks for the comments.
> >>
> >>> I can't reproduce this problem on Linux 6.12-rc3.
> >>> What's you environment like?
> >>
> >> My kernel version is 6.12.0-rc3 from "make kernelversion".
> >>
> >> I think the errors are related to the bash type and version, rather than the kernel version.
> >> My bash version is: GNU bash, version 5.2.21(1)-release (x86_64-pc-linux-gnu)
> >>
> >> In fact, some shell do not complete expressions in variables and $var and command substitutions
> >> are done before the arithmetic expression itself is parsed.
> >> That expansion happens without regard for the arithmetic syntax, so with $var you can mess
> >> with that.
> >> So, I suggest to avoid to use $var inside a arithmetic expansion in order to be cross-platform.
> >
> > Hello,
> > any thoughts about this patch?
> >
> > Were you able to replicate the error?
> >
>
> Yes I was able to reproduce what you are seeing.
>
> >>>>
> >>>>    tools/testing/selftests/intel_pstate/run.sh | 4 ++--
> >>>>    1 file changed, 2 insertions(+), 2 deletions(-)
> >>>>
> >>>> diff --git a/tools/testing/selftests/intel_pstate/run.sh b/tools/testing/selftests/intel_pstate/run.sh
> >>>> index e7008f614ad7..0c1b6c1308a4 100755
> >>>> --- a/tools/testing/selftests/intel_pstate/run.sh
> >>>> +++ b/tools/testing/selftests/intel_pstate/run.sh
> >>>> @@ -87,9 +87,9 @@ mkt_freq=${_mkt_freq}0
> >>>>    # Get the ranges from cpupower
> >>>>    _min_freq=$(cpupower frequency-info -l | tail -1 | awk ' { print $1 } ')
> >>>> -min_freq=$(($_min_freq / 1000))
> >>>> +min_freq=$((_min_freq / 1000))
> >>>>    _max_freq=$(cpupower frequency-info -l | tail -1 | awk ' { print $2 } ')
> >>>> -max_freq=$(($_max_freq / 1000))
> >>>> +max_freq=$((_max_freq / 1000))
> >>>>    [ $EVALUATE_ONLY -eq 0 ] && for freq in `seq $max_freq -100 $min_freq`
> >>>
>
> The patch is fine. I applied and run it. I found another problem
> when cpupower command doesn't run
>
> # ./run.sh: line 89: cpupower: command not found
> # ./run.sh: line 91: cpupower: command not found
>
> So you would have to check if min_freq and max_freq are valid
> and don't continue if cpupower isn't found. This test depends
> on cpupower.

Thanks for the feedback.
I'm going to work also on the cpupower issue.

> You can do that as a separate patch and send it as a series with
> commit log changes I suggested on v1 of this patch.

Ok.

> thanks,
> -- Shuah

Thank again,
Alessandro