linux-kernel - Re: [PATCH 2/2] mtd: mtdconcat: Remove concat_{read|write}

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: <20210807122814.109ec082@xps13>
Date:   Sat, 7 Aug 2021 12:28:14 +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 2/2] mtd: mtdconcat: Remove concat_{read|write}_oob

Hi Zhihao,

Zhihao Cheng <chengzhihao1@...wei.com> wrote on Sat, 7 Aug 2021
10:59:32 +0800:

> 在 2021/8/7 3:26, Miquel Raynal 写道:
> Hi Miquel,
> >>   static int concat_erase(struct mtd_info *mtd, struct erase_info *instr)
> >>   {
> >>   	struct mtd_concat *concat = CONCAT(mtd);
> >> @@ -684,10 +580,6 @@ struct mtd_info *mtd_concat_create(struct mtd_info *subdev[],	/* subdevices to c
> >>   	subdev_master = mtd_get_master(subdev[0]);
> >>   	if (subdev_master->_writev)
> >>   		concat->mtd._writev = concat_writev;
> >> -	if (subdev_master->_read_oob)
> >> -		concat->mtd._read_oob = concat_read_oob;
> >> -	if (subdev_master->_write_oob)
> >> -		concat->mtd._write_oob = concat_write_oob;  
> > Actually I am not sure _read|write_oob() is the right callback to
> > remove.
> >
> > Richard, what is your input on this? Shall we remove _read|write()
> > instead? I don't remember the exact rationale behind these two helpers.  
> 
> Oh, I guess I made a mistake. It looks like that reserving _{read|write}_oob is a better method in my limited knowledge to nand driver. For example, nand_do_read_oob() behaves different from nand_do_read_ops(), and calling which function is decided by mtd_oob_ops.databuf.
> Callback _read_oobs() can support both functions, but callback _read() don't support nand_do_read_oob(). So mtd_read_oobs() covers mtd_read()?
> Is my understand right?
> 

Yes please let's drop _read|write() instead.

Thanks,
Miquèl