linux-kernel - Re: [PATCH net] macmace: Set platform device coherent_dma

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: <alpine.LNX.2.21.1805091804290.72@nippy.intranet>
Date:   Thu, 10 May 2018 11:25:09 +1000 (AEST)
From:   Finn Thain <fthain@...egraphics.com.au>
To:     Christoph Hellwig <hch@....de>
cc:     Geert Uytterhoeven <geert@...ux-m68k.org>,
        "David S. Miller" <davem@...emloft.net>,
        linux-m68k <linux-m68k@...ts.linux-m68k.org>,
        netdev <netdev@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH net] macmace: Set platform device coherent_dma_mask

On Thu, 3 May 2018, Christoph Hellwig wrote:

> On Thu, May 03, 2018 at 10:46:56AM +0200, Geert Uytterhoeven wrote:
> > Perhaps you can add a new helper 
> > (platform_device_register_simple_dma()?) that takes the DMA mask, too? 
> > With people setting the mask to kill the WARNING splat, this may 
> > become more common.
> > 
> > struct platform_device_info already has a dma_mask field, but 
> > platform_device_register_resndata() explicitly sets it to zero.
> 
> Yes, that would be useful.  The other assumption could be that platform 
> devices always allow an all-0xff dma mask.

Could that have unintended side effects? The mask is presently unset by 
default, and my worry would be that changing it may cause some drivers to 
behave differently.

A quick grep turns up this in drivers/spi/spi-au1550.c for example,

	if (pdev->dev.dma_mask == NULL)
		dev_warn(&pdev->dev, "no dma mask\n");
	else
		hw->usedma = 1;

Also, if pdev.dev.dma_mask is to get a default value, shouldn't it use the 
same default as dma_get_mask, to avoid unintended side effects?

static inline u64 dma_get_mask(struct device *dev)
{
        if (dev && dev->dma_mask && *dev->dma_mask)
                return *dev->dma_mask;
        return DMA_BIT_MASK(32);
}

--