[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180906122956.GF10768@zn.tnic>
Date: Thu, 6 Sep 2018 14:29:56 +0200
From: Borislav Petkov <bp@...en8.de>
To: Pu Wen <puwen@...on.cn>
Cc: tglx@...utronix.de, mingo@...hat.com, hpa@...or.com,
x86@...nel.org, thomas.lendacky@....com, pbonzini@...hat.com,
tony.luck@...el.com, linux-kernel@...r.kernel.org,
linux-arch@...r.kernel.org, linux-edac@...r.kernel.org
Subject: Re: [PATCH v5 10/16] x86/mce: enable Hygon support to MCE
infrastructure
On Thu, Sep 06, 2018 at 07:40:52PM +0800, Pu Wen wrote:
> The first is the compilation test. Test the kernel with both MCE selected
> and unselected to see if is the compiling process successful or not.
This is a kernel compile test - it has nothing to do with actual machine
testing.
> The other test is functionality test. For example, during the MCE BIOS
> development, to see if the MCE functions OK, it may need to select and
> unselect MCE in kernel for double checking.
Say what now?! What testing do you do if you boot a kernel which doesn't
even use the functionality you're testing?!
> Actually in normal use scenario and in real product the MCE should be
> always selected.
It better be!
> To ensure this, I think there are two ways as below:
> - Select X86_MCE_AMD in CPU_SUP_HYGON config entry, it also cater to the
> first test scenario, but meanwhile lacks flexibility.
> - The linux distros(Ubuntu, CentOS, etc) ensure that X86_MCE_AMD is
> selected to the default config file, and indeed they do.
> Which way is better?
The first way but change that dependency to CPU_SUP_AMD because there
are people building their own kernels and don't run distro configs. And
you need CPU_SUP_AMD because you're using their code.
--
Regards/Gruss,
Boris.
Good mailing practices for 400: avoid top-posting and trim the reply.
Powered by blists - more mailing lists