linux-kernel - Re: [PATCH] Fix Oops with Atmel SPI

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] [day] [month] [year] [list]

Date:	Fri, 28 May 2010 11:27:44 +0200
From:	Haavard Skinnemoen <haavard.skinnemoen@...el.com>
To:	Andrew Morton <akpm@...ux-foundation.org>
Cc:	Anders Larsen <al@...rsen.net>,
	Iwo Mergler <iwo@...l-direct.com.au>,
	linux-mtd@...ts.infradead.org,
	Artem Bityutskiy <Artem.Bityutskiy@...ia.com>,
	Ian McDonnell <ian@...ghtstareng.com>,
	Nicolas Pitre <nico@...xnic.net>, linux-kernel@...r.kernel.org,
	Matthias Kaehlcke <matthias@...hlcke.net>,
	David Woodhouse <dwmw2@...radead.org>,
	Haavard Skinnemoen <hskinnemoen@...el.com>
Subject: Re: [PATCH] Fix Oops with Atmel SPI

Andrew Morton <akpm@...ux-foundation.org> wrote:
> On Wed, 19 May 2010 13:05:00 +0200
> Anders Larsen <al@...rsen.net> wrote:
> 
> > On 2010-04-22 00:24:10, Andrew Morton wrote:
> > > Finally..  Wouldn't it be better to just fix the atmel SPI driver so
> > > that it doesn't barf when handed vmalloc'ed memory?  Who do we ridicule
> > > about that?  <checks, adds cc>
> > 
> > You mean something like this instead?
> 
> That looks simple enough.  How do we get it tested, changelogged and
> merged up?  Haavard, can you please take a look?

Sure. Sorry for the late response; I've been traveling for the last two
weeks.

Did anyone check what other drivers do to handle this case? Surely this
isn't the only driver which supports DMA?

> > diff --git a/drivers/spi/atmel_spi.c b/drivers/spi/atmel_spi.c
> > index c4e0442..a9ad5e8 100644
> > --- a/drivers/spi/atmel_spi.c
> > +++ b/drivers/spi/atmel_spi.c
> > @@ -352,16 +352,30 @@ atmel_spi_dma_map_xfer(struct atmel_spi *as, struct spi_transfer *xfer)
> >  
> >  	xfer->tx_dma = xfer->rx_dma = INVALID_DMA_ADDRESS;
> >  	if (xfer->tx_buf) {
> > -		xfer->tx_dma = dma_map_single(dev,
> > -				(void *) xfer->tx_buf, xfer->len,
> > -				DMA_TO_DEVICE);
> > +		if (is_vmalloc_addr(xfer->tx_buf))
> > +			xfer->tx_dma = dma_map_page(dev,
> > +					vmalloc_to_page(xfer->tx_buf),
> > +					(unsigned long)xfer->tx_buf & (PAGE_SIZE-1),
> > +					xfer->len,
> > +					DMA_TO_DEVICE);

Ok, this should be fine for small transfers, but what happens if the
transfer crosses a page boundary? Are there any guarantees that this
will never happen? What callers are passing vmalloc'ed memory in the
first place?

Ditto for the rx path.

Haavard
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/