linux-kernel - Re: [PATCH v5] ata/pata_buddha: Probe via modalias instead of initcall

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]

Date:   Tue, 20 Aug 2019 17:59:47 +0200
From:   Max Staudt <max@...as.org>
To:     Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>
Cc:     axboe@...nel.dk, linux-ide@...r.kernel.org,
        linux-m68k@...r.kernel.org, linux-kernel@...r.kernel.org,
        glaubitz@...sik.fu-berlin.de, schmitzmic@...il.com,
        geert@...ux-m68k.org
Subject: Re: [PATCH v5] ata/pata_buddha: Probe via modalias instead of
 initcall

Hi Bartlomiej,

Thank you very much for your review!

Question below.

On 08/20/2019 02:06 PM, Bartlomiej Zolnierkiewicz wrote:
>> +	/* Workaround for X-Surf: Save drvdata in case zorro8390 has set it */
>> +	old_drvdata = dev_get_drvdata(&z->dev);
> 
> This should be done only for type == BOARD_XSURF.

Agreed, as I want to keep unloading functional for Buddha/Catweasel - see below.

>> +static struct zorro_driver pata_buddha_driver = {
>> +	.name           = "pata_buddha",
>> +	.id_table       = pata_buddha_zorro_tbl,
>> +	.probe          = pata_buddha_probe,
>> +	.remove         = pata_buddha_remove,
> 
> I think that we should also add:
> 
> 	.driver  = {
> 		.suppress_bind_attrs = true,
> 	},
> 
> to prevent the device from being unbinded (and thus ->remove called)
> from the driver using sysfs interface.

Interesting idea - here's my question now:

My intention is to allow remove() for boards where we support IDE only (Buddha, Catweasel) - these are autoprobed via zorro_register_driver().
This shouldn't affect the X-Surf case, as it's not autoprobed in this way anyway - and thus pata_buddha_driver isn't even used.

Am I missing something? We want to inhibit module unloading (hence no module_exit()), but driver unbinding for Buddha/Catweasel should be fine to remain, right?

> Please also always check your patches with scripts/checkpatch.pl and
> fix the reported issues:

Apologies, must've been something in my coffee. I will.

Thanks for the review, I'll send a new patch once my question above is resolved.

Max