[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20211028074949.GB23476@willie-the-truck>
Date: Thu, 28 Oct 2021 08:49:49 +0100
From: Will Deacon <will@...nel.org>
To: Qian Cai <quic_qiancai@...cinc.com>
Cc: Catalin Marinas <catalin.marinas@....com>,
linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: Arm64 defconfig-debug
On Wed, Oct 27, 2021 at 03:45:32PM -0400, Qian Cai wrote:
> Hi there, some general debugging features like kmemleak, KASAN, lockdep,
> UBSAN etc help fix many bugs like a microscope. On the other hand, those
> features are scatter around and mixed up with more situational debugging
> options making them difficult to consume properly.
>
> Since I have been deal with those options on a daily basis for the last
> a few years and accumulated the knowledge of their pros and cons, I am
> thinking about to create an initial defconfig-debug available inside the
> tree, so that it could amplify the general debugging/testing efforts for
> arm64 across the broad.
>
> The downside is that this might open a gate for other people to push yet
> another subject-specific defconfig, but I don't see anything as
> important as general debugging/testing at the moment. Then, there are
> two defconfig need to be modified for some common changes, but I think
> that is quite trivial to do. We could even create a simple script to
> sync if that is not overkill.
I'm not in favour of adding a whole new debug-flavoured defconfig for arm64,
but you might be able to add some reduced config file containing just the
options you're interested in to kernel/configs/. That said, I'm not sure the
existing files under there are well-maintained at all. The Android ones,
especially, looks like they should just be removed.
Will
Powered by blists - more mailing lists