[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20160223152353.GA22447@lnxrabinv.se.axis.com>
Date: Tue, 23 Feb 2016 16:23:53 +0100
From: Rabin Vincent <rabin@....in>
To: Russell King - ARM Linux <linux@....linux.org.uk>
Cc: Rabin Vincent <rabin.vincent@...s.com>, mina86@...a86.com,
akpm@...ux-foundation.org, linux-arm-kernel@...ts.infradead.org,
linux-mm@...ck.org, linux-kernel@...r.kernel.org,
Rabin Vincent <rabinv@...s.com>
Subject: Re: [PATCH 2/2] ARM: dma-mapping: fix alloc/free for coherent + CMA
+ gfp=0
On Fri, Feb 19, 2016 at 02:06:00PM +0000, Russell King - ARM Linux wrote:
> On Fri, Feb 19, 2016 at 09:12:04AM +0100, Rabin Vincent wrote:
> > Given a device which uses arm_coherent_dma_ops and on which
> > dev_get_cma_area(dev) returns non-NULL, the following usage of the DMA
> > API with gfp=0 results in a memory leak and memory corruption.
> >
> > p = dma_alloc_coherent(dev, sz, &dma, 0);
> > if (p)
> > dma_free_coherent(dev, sz, p, dma);
> >
> > The memory leak is because the alloc allocates using
> > __alloc_simple_buffer() but the free attempts
> > dma_release_from_contiguous(), which does not do free anything since the
> > page is not in the CMA area.
>
> I'd really like to see a better solution to this problem: over the course
> of the years, I've seen a number of patches that rearrange the test order
> at allocation time because of some problem or the other.
>
> What we need is a better way to ensure that we use the correct release
> functionality - having two independent set of tests where the order
> matters is really not very good.
I've sent a v2 of this series which refactors the code so that we no
longer have two independent sets of tests.
Powered by blists - more mailing lists