linux-kernel - Re: [PATCH 1/2] mtd: mtdconcat: Judge callback function existence getting from master for each partition

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20210807123243.7661e4e3@xps13>
Date:   Sat, 7 Aug 2021 12:32:43 +0200
From:   Miquel Raynal <miquel.raynal@...tlin.com>
To:     Zhihao Cheng <chengzhihao1@...wei.com>
Cc:     <richard@....at>, <vigneshr@...com>, <bbrezillon@...nel.org>,
        <linux-mtd@...ts.infradead.org>, <linux-kernel@...r.kernel.org>,
        <yukuai3@...wei.com>
Subject: Re: [PATCH 1/2] mtd: mtdconcat: Judge callback function existence
 getting from master for each partition

Hi Zhihao,

Zhihao Cheng <chengzhihao1@...wei.com> wrote on Sat, 7 Aug 2021
10:15:46 +0800:

> 在 2021/8/7 3:28, Miquel Raynal 写道:
> Hi Miquel,
> > Hi Zhihao,
> >
> > Zhihao Cheng <chengzhihao1@...wei.com> wrote on Sat, 31 Jul 2021
> > 10:32:42 +0800:
> > @@ -721,14 +724,15 @@ struct mtd_info *mtd_concat_create(struct mtd_info *subdev[],	/* subdevices to c
> >   				    subdev[i]->flags & MTD_WRITEABLE;
> >   		}  
> >   > +		subdev_master = mtd_get_master(subdev[i]);  
> >   		concat->mtd.size += subdev[i]->size;
> >   		concat->mtd.ecc_stats.badblocks +=
> >   			subdev[i]->ecc_stats.badblocks;
> >   		if (concat->mtd.writesize   !=  subdev[i]->writesize ||
> >   		    concat->mtd.subpage_sft != subdev[i]->subpage_sft ||
> >   		    concat->mtd.oobsize    !=  subdev[i]->oobsize ||
> > -		    !concat->mtd._read_oob  != !subdev[i]->_read_oob ||
> > -		    !concat->mtd._write_oob != !subdev[i]->_write_oob) {
> > +		    !concat->mtd._read_oob  != !subdev_master->_read_oob ||
> > +		    !concat->mtd._write_oob != !subdev_master->_write_oob) {
> > Do you really need this change?  
> 
> I think both "!concat->mtd._read_oob != !subdev[i]->_read_oob" and "!concat->mtd._write_oob != !subdev[i]->_write_oob" need to be modified otherwise concatenating goes failure.
> 
> I thought there exists two problems:
> 
>    1. Wrong callback fetching in mtd partition device
> 
>    2. Warning for existence of _read and _read_oob at the same time
> 
> so I solved them in two steps to make history commit logs a bit clear.
> 
> Though these two patches can be combined to one.

No please keep the split.

What I mean here is that I don't think your fix is valid. Maybe we
should propagate these callbacks as well instead of trying to hack into
this condition. I don't see why you should check against subdev[i] for
half of the callbacks and check for subdev_master for the last two.


Thanks,
Miquèl