Received-SPF: pass (google.com: domain of linux-kernel+bounces-62152-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Date: Mon, 12 Feb 2024 18:54:08 +0100
From: Borislav Petkov <bp@alien8.de>
To: "Luck, Tony" <tony.luck@intel.com>
Cc: "Naik, Avadhut" <avadnaik@amd.com>,
	"Mehta, Sohil" <sohil.mehta@intel.com>,
	"x86@kernel.org" <x86@kernel.org>,
	"linux-edac@vger.kernel.org" <linux-edac@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"yazen.ghannam@amd.com" <yazen.ghannam@amd.com>,
	Avadhut Naik <avadhut.naik@amd.com>
Subject: Re: [PATCH 2/2] x86/MCE: Add command line option to extend MCE
 Records pool
Message-ID: <20240212175408.GIZcpbQHVjEtwRKLS-@fat_crate.local>
References: <20240209200920.GFZcaGcOr757W9O3IG@fat_crate.local>
 <7a4945b0-322a-444e-a0ca-860a062a49c3@amd.com>
 <20240209205111.GGZcaQP1gb6C9m0WZB@fat_crate.local>
 <5DB0FF8D-C6DA-45DC-B287-201A9BF48BDA@alien8.de>
 <75ddf61d-8dda-47fa-9da0-24221feb22a2@amd.com>
 <20240211111455.GAZcisL09LeFPWa2EI@fat_crate.local>
 <774e7ca5-154d-4ca4-bc4c-2f945c20b938@amd.com>
 <20240212085801.GAZcndma4UTPtKm33e@fat_crate.local>
 <20240212093246.GBZcnlvkPKDC8C7rv5@fat_crate.local>
 <SJ1PR11MB6083B3511D18787BE823AB2CFC482@SJ1PR11MB6083.namprd11.prod.outlook.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <SJ1PR11MB6083B3511D18787BE823AB2CFC482@SJ1PR11MB6083.namprd11.prod.outlook.com>

On Mon, Feb 12, 2024 at 05:29:31PM +0000, Luck, Tony wrote:
> Walking the structures already allocated from the genpool in the #MC
> handler may be possible, but what is the criteria for "duplicates"?

for each i in pool:
	memcmp(mce[i], new_mce, sizeof(struct mce));

It'll probably need to mask out fields like ->time etc.

> Do we avoid entering duplicates into the pool altogether?

We could do

	mce[i].count++;

in the same loop.

Dunno yet if we even need to state how many duplicates are there...

> Or when the pool is full overwrite a duplicate?

Nah, not overwrite the duplicate but not add the new one. Cheaper.

> How about compile time allocation of extra space. Algorithm below for
> illustrative purposes only. May need some more thought about how
> to scale up.

Yeah, it is too static IMO. Especially if NR_CPUS is a build-time thing
- needs to be based on the actual number of CPUs on the machine.

BUT, we don't have an allocator yet.

So we end up allocating it there on the heap.

Unless we can do something with memblock...

But then this still needs code audit whether num_possible_cpus() is
final at that point. Because if it is, that would be the optimal thing
to do a memblock_alloc() there...

> [Diff pasted into Outlook, chances that it will automatically apply = 0%]

How you even do patches with outschmook is mindboggling :-)

At least use Thunderbird. That's what I do for the company mail but then
again I don't even try to do patches over company mail...

Thx.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette