Re: [PATCH 1/3] selftests/zram: Remove obsolete max_comp_streams interface

From: Petr Vorel <pvorel@suse.cz>
To: Shuah Khan <skhan@linuxfoundation.org>
Cc: "xuyang2018.jy@fujitsu.com" <xuyang2018.jy@fujitsu.com>,
	"linux-kselftest@vger.kernel.org"
	<linux-kselftest@vger.kernel.org>,
	"shuah@kernel.org" <shuah@kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"naresh.kamboju@linaro.org" <naresh.kamboju@linaro.org>,
	"aleksei.kodanev@bell-sw.com" <aleksei.kodanev@bell-sw.com>
Subject: Re: [PATCH 1/3] selftests/zram: Remove obsolete max_comp_streams interface
Date: Wed, 26 Jan 2022 19:24:55 +0100	[thread overview]
Message-ID: <YfGR9xZ2EuMXyAzb@pevik> (raw)
In-Reply-To: <590c1f1c-2da1-583a-d055-83c15969cf80@linuxfoundation.org>

> On 1/26/22 12:13 AM, Petr Vorel wrote:
> > Hi all,

> > > on 2022/1/26 4:33, Shuah Khan wrote :
> > > > On 12/15/21 2:56 AM, Yang Xu wrote:
> > > > > Since kernel commit 43209ea2d17a ("zram: remove max_comp_streams
> > > > > internals"), zram has
> > > > > switched to per-cpu streams. Even kernel still keep this interface for
> > > > > some reasons, but
> > > > > writing to max_comp_stream doesn't take any effect. So remove it.

> > > > I get that max_comp_stream doesn't do anything since this referenced
> > > > commit. Don't we need this test on older kernels since older kernels
> > > > still support max_comp_stream?

> > > I read the following info from kernel selftest documentation
> > > https://www.kernel.org/doc/html/latest/dev-tools/kselftest.html

> > > "The kernel contains a set of “self tests” under the
> > > tools/testing/selftests/ directory. These are intended to be small tests
> > > to exercise individual code paths in the kernel. Tests are intended to
> > > be run after building, installing and booting a kernel."

> > > So, we can build older kernel(use older kernel source) if we want to
> > > test older kernel.

> > > IMO, kernel selftest is different from other testsuit(ie ltp, this
> > > shuould think about api changes because ltp may test on different kernel).
> > Yes, that's how I understand the difference with approach of in kselftest - the
> > kernel tree testsuite and LTP - the out-of-tree testsuite.

> Removing max_comp_stream test appears to be motivated by the fact it isn't
> needed on newer kernels.

> Kselftest from mainline can be run on older stable kernels. This is a use-case
> for a lot test rings. The idea is that when a new test gets added for older
> code to regression test a bug, we should be able to run that test on an older
> kernel. This is the reason why we don't remove code that can still test an older
> kernel and make sure it skips gracefully.

Thanks for clarifying this approach. It might be worth of documenting it in
dev-tools/kselftest.rst.

Kind regards,
Petr

> Hence, I won't be taking this patch.

> thanks,
> -- Shuah