Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 27 Mar 2023 18:49:22 -0700
From:   Christoph Hellwig <hch@infradead.org>
To:     "Darrick J. Wong" <djwong@kernel.org>
Cc:     Ye Bin <yebin@huaweicloud.com>, linux-xfs@vger.kernel.org,
        linux-kernel@vger.kernel.org, Ye Bin <yebin10@huawei.com>,
        Christoph Hellwig <hch@infradead.org>
Subject: Re: [PATCH] xfs: fix BUG_ON in xfs_getbmap()
Message-ID: <ZCJHotVOmvK+/P/k@infradead.org>
References: <20230327140218.4154709-1-yebin@huaweicloud.com>
 <20230327151524.GC16180@frogsfrogsfrogs>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230327151524.GC16180@frogsfrogsfrogs>
Precedence: bulk

On Mon, Mar 27, 2023 at 08:15:24AM -0700, Darrick J. Wong wrote:
> > There's issue as follows:
> > XFS: Assertion failed: (bmv->bmv_iflags & BMV_IF_DELALLOC) != 0, file: fs/xfs/xfs_bmap_util.c, line: 329
> 
> Why not get rid of the assertion?  It's not like it changes the course
> of the code flow -- userspace still gets told there's a delalloc extent.
> 
> Or, if the assert does serve some purpose, then do we need to take
> the mmaplock for cow fork reporting too?

Looking at the COW fork reporting I think it's actually even more
broken as it never tried to flush data to start with.  But COW
report is a DEBUG only feature, so maybe forcing or requiring
BMV_IF_DELALLOC there would make a whole lot of sense.