Received-SPF: pass (google.com: domain of linux-ext4-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 11 Jan 2023 10:30:19 -0800
From:   Eric Biggers <ebiggers@kernel.org>
To:     Jan Kara <jack@suse.cz>
Cc:     linux-ext4@vger.kernel.org, Ye Bin <yebin10@huawei.com>
Subject: Re: [PATCH] ext4: use ext4_fc_tl_mem in fast-commit replay path
Message-ID: <Y78AO7WX5Q6Zju3p@sol.localdomain>
References: <20221217050212.150665-1-ebiggers@kernel.org>
 <20230111144327.gb6p3foqj23mby7w@quack3>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230111144327.gb6p3foqj23mby7w@quack3>
Precedence: bulk

On Wed, Jan 11, 2023 at 03:43:27PM +0100, Jan Kara wrote:
> On Fri 16-12-22 21:02:12, Eric Biggers wrote:
> > From: Eric Biggers <ebiggers@google.com>
> > 
> > To avoid 'sparse' warnings about missing endianness conversions, don't
> > store native endianness values into struct ext4_fc_tl.  Instead, use a
> > separate struct type, ext4_fc_tl_mem.
> > 
> > Fixes: dcc5827484d6 ("ext4: factor out ext4_fc_get_tl()")
> > Cc: Ye Bin <yebin10@huawei.com>
> > Signed-off-by: Eric Biggers <ebiggers@google.com>
> 
> Looks good to me. Just one nit below:
> 
> > -static inline void ext4_fc_get_tl(struct ext4_fc_tl *tl, u8 *val)
> > +static inline void ext4_fc_get_tl(struct ext4_fc_tl_mem *tl, u8 *val)
> >  {
> > -	memcpy(tl, val, EXT4_FC_TAG_BASE_LEN);
> > -	tl->fc_len = le16_to_cpu(tl->fc_len);
> > -	tl->fc_tag = le16_to_cpu(tl->fc_tag);
> > +	struct ext4_fc_tl tl_disk;
> > +
> > +	memcpy(&tl_disk, val, EXT4_FC_TAG_BASE_LEN);
> > +	tl->fc_len = le16_to_cpu(tl_disk.fc_len);
> > +	tl->fc_tag = le16_to_cpu(tl_disk.fc_tag);
> >  }
> 
> So why not just:
> 
> 	struct ext4_fc_tl *tl_disk = (struct ext4_fc_tl *)val;
> 
> instead of memcpy?

That would result in unaligned memory accesses.

- Eric