Re: [REGRESSION] 6.7.1: md: raid5 hang and unresponsive system; successfully bisected

From: Roman Mamedov <rm@romanrm.net>
To: Carlos Carvalho <carlos@fisica.ufpr.br>
Cc: Dan Moulding <dan@danm.net>,
	junxiao.bi@oracle.com, gregkh@linuxfoundation.org,
	linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org,
	regressions@lists.linux.dev, song@kernel.org,
	stable@vger.kernel.org, yukuai1@huaweicloud.com
Subject: Re: [REGRESSION] 6.7.1: md: raid5 hang and unresponsive system; successfully bisected
Date: Fri, 26 Jan 2024 21:21:50 +0500	[thread overview]
Message-ID: <20240126212150.4b56ae05@nvm> (raw)
In-Reply-To: <ZbMnZnvyIyoWeIro@fisica.ufpr.br>

On Fri, 26 Jan 2024 00:30:46 -0300
Carlos Carvalho <carlos@fisica.ufpr.br> wrote:

> Dan Moulding (dan@danm.net) wrote on Thu, Jan 25, 2024 at 05:31:30PM -03:
> > I then created an ext4 file system on the "data" volume, mounted it, and used
> > "dd" to copy 1MiB blocks from /dev/urandom to a file on the "data" file
> > system, and just let it run. Eventually "dd" hangs and top shows that
> > md0_raid5 is using 100% CPU.
> 
> It's known that ext4 has these symptoms with parity raid. To make sure it's a
> raid problem you should try another filesystem or remount it with stripe=0.

If Ext4 wouldn't work properly on parity RAID, then it is a bug that should be
tracked down and fixed, not worked around by using a different FS. I am in
disbelief you are seriously suggesting that, and to be honest really doubt
there is any such high-profile "known" issue that stays unfixed and is just
commonly worked around.

-- 
With respect,
Roman