Re: [dm-devel] [PATCH v5-fix 1/5] block: Don't invalidate pagecache for invalid falloc modes

From: Mike Snitzer <snitzer@kernel.org>
To: Sarthak Kukreti <sarthakkukreti@chromium.org>
Cc: Jens Axboe <axboe@kernel.dk>,
	Christoph Hellwig <hch@infradead.org>,
	Theodore Ts'o <tytso@mit.edu>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	"Darrick J. Wong" <djwong@kernel.org>,
	Jason Wang <jasowang@redhat.com>,
	Bart Van Assche <bvanassche@google.com>,
	linux-kernel@vger.kernel.org, linux-block@vger.kernel.org,
	dm-devel@redhat.com, Andreas Dilger <adilger.kernel@dilger.ca>,
	Daniil Lunev <dlunev@google.com>,
	Stefan Hajnoczi <stefanha@redhat.com>,
	linux-fsdevel@vger.kernel.org, linux-ext4@vger.kernel.org,
	Brian Foster <bfoster@redhat.com>,
	Alasdair Kergon <agk@redhat.com>
Subject: Re: [dm-devel] [PATCH v5-fix 1/5] block: Don't invalidate pagecache for invalid falloc modes
Date: Thu, 20 Apr 2023 12:20:27 -0400	[thread overview]
Message-ID: <ZEFmS9h81Wwlv9+/@redhat.com> (raw)
In-Reply-To: <20230420014734.302304-1-sarthakkukreti@chromium.org>

On Wed, Apr 19 2023 at  9:47P -0400,
Sarthak Kukreti <sarthakkukreti@chromium.org> wrote:

> Only call truncate_bdev_range() if the fallocate mode is
> supported. This fixes a bug where data in the pagecache
> could be invalidated if the fallocate() was called on the
> block device with an invalid mode.
> 
> Fixes: 25f4c41415e5 ("block: implement (some of) fallocate for block devices")

You should add:

Cc: stable@vger.kernel.org
Reported-by: Darrick J. Wong <djwong@kernel.org>

> Signed-off-by: Sarthak Kukreti <sarthakkukreti@chromium.org>
> ---
>  block/fops.c | 37 ++++++++++++++++++++++++-------------
>  1 file changed, 24 insertions(+), 13 deletions(-)
> 
> diff --git a/block/fops.c b/block/fops.c
> index d2e6be4e3d1c..d359254c645d 100644
> --- a/block/fops.c
> +++ b/block/fops.c
> @@ -648,26 +648,37 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,
>  
>  	filemap_invalidate_lock(inode->i_mapping);
>  
> -	/* Invalidate the page cache, including dirty pages. */
> -	error = truncate_bdev_range(bdev, file->f_mode, start, end);
> -	if (error)
> -		goto fail;
> -

You remove the only user of the 'fail' label.  But I think it'd be
cleaner to keep using it below (reduces indentation churn too).

> +	/*
> +	 * Invalidate the page cache, including dirty pages, for valid
> +	 * de-allocate mode calls to fallocate().
> +	 */
>  	switch (mode) {
>  	case FALLOC_FL_ZERO_RANGE:
>  	case FALLOC_FL_ZERO_RANGE | FALLOC_FL_KEEP_SIZE:
> -		error = blkdev_issue_zeroout(bdev, start >> SECTOR_SHIFT,
> -					     len >> SECTOR_SHIFT, GFP_KERNEL,
> -					     BLKDEV_ZERO_NOUNMAP);
> +		error = truncate_bdev_range(bdev, file->f_mode, start, end);
> +		if (!error)
> +			error = blkdev_issue_zeroout(bdev,
> +						     start >> SECTOR_SHIFT,
> +						     len >> SECTOR_SHIFT,
> +						     GFP_KERNEL,
> +						     BLKDEV_ZERO_NOUNMAP);
>  		break;

So:

		error = truncate_bdev_range(bdev, file->f_mode, start, end);
		if (error)
		        goto fail;
		...

>  	case FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE:
> -		error = blkdev_issue_zeroout(bdev, start >> SECTOR_SHIFT,
> -					     len >> SECTOR_SHIFT, GFP_KERNEL,
> -					     BLKDEV_ZERO_NOFALLBACK);
> +		error = truncate_bdev_range(bdev, file->f_mode, start, end);
> +		if (!error)
> +			error = blkdev_issue_zeroout(bdev,
> +						     start >> SECTOR_SHIFT,
> +						     len >> SECTOR_SHIFT,
> +						     GFP_KERNEL,
> +						     BLKDEV_ZERO_NOFALLBACK);
>  		break;

Same.

>  	case FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE | FALLOC_FL_NO_HIDE_STALE:
> -		error = blkdev_issue_discard(bdev, start >> SECTOR_SHIFT,
> -					     len >> SECTOR_SHIFT, GFP_KERNEL);
> +		error = truncate_bdev_range(bdev, file->f_mode, start, end);
> +		if (!error)
> +			error = blkdev_issue_discard(bdev,
> +						     start >> SECTOR_SHIFT,
> +						     len >> SECTOR_SHIFT,
> +						     GFP_KERNEL);
>  		break;

Same.

--
dm-devel mailing list
dm-devel@redhat.com
https://listman.redhat.com/mailman/listinfo/dm-devel