[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20241121045714.GA20680@lst.de>
Date: Thu, 21 Nov 2024 05:57:14 +0100
From: Christoph Hellwig <hch@....de>
To: Jens Axboe <axboe@...nel.dk>
Cc: Chaitanya Kulkarni <chaitanyak@...dia.com>,
Saeed Mirzamohammadi <saeed.mirzamohammadi@...cle.com>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"linux-nvme@...ts.infradead.org" <linux-nvme@...ts.infradead.org>,
Ramanan Govindarajan <ramanan.govindarajan@...cle.com>,
Sagi Grimberg <sagi@...mberg.me>,
Paul Webb <paul.x.webb@...cle.com>, Christoph Hellwig <hch@....de>,
Keith Busch <kbusch@...nel.org>
Subject: Re: [bug-report] 5-9% FIO randomwrite ext4 perf regression on
6.12.y kernel
On Wed, Nov 20, 2024 at 06:20:12PM -0700, Jens Axboe wrote:
> There's no way that commit is involved, the test as quoted doesn't even
> touch write zeroes. Hence if there really is a regression here, then
> it's either not easily bisectable, some error was injected while
> bisecting, or the test itself is bimodal.
ext4 actually has some weird lazy init code using write zeroes. So
if the test actually wasn't a steady state one but only run for a short
time after init, and the mentioned commit dropped the intel hack for
deallocate as write zeroes it might actually make a difference.
To check for that do a :
/sys/block/nvmeXn1/queue/write_zeroes_max_bytes
with and without that commit.
Powered by blists - more mailing lists