From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Date: Tue, 17 Apr 2018 13:54:07 +0000 Subject: Re: [PATCH] sh: mm: Fix unprotected access to struct device Message-Id: <20180417155407.49c981b3@windsurf.numericable.fr> List-Id: References: <1523972123-5700-1-git-send-email-jacopo+renesas@jmondi.org> In-Reply-To: <1523972123-5700-1-git-send-email-jacopo+renesas@jmondi.org> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Jacopo Mondi Cc: ysato@users.sourceforge.jp, dalias@libc.org, thomas.petazzoni@free-electrons.com, robin.murphy@arm.com, geert@linux-m68k.org, linux-renesas-soc@vger.kernel.org, linux-sh@vger.kernel.org, linux-kernel@vger.kernel.org Hello, On Tue, 17 Apr 2018 15:35:23 +0200, Jacopo Mondi wrote: > With commit ce88313069c36eef80f21fd7 ("arch/sh: make the DMA mapping > operations observe dev->dma_pfn_offset") the generic DMA allocation > function on which the SH 'dma_alloc_coherent()' function relies on, > access the 'dma_pfn_offset' field of struct device. > > Unfortunately the 'dma_generic_alloc_coherent()' function is called from > several places with a NULL struct device argument, halting the CPU > during the boot process. > > This patch fixes the issue protecting access to dev->dma_pfn_offset, > with a trivial check for validity. It also passes a valid 'struct device' > in the 'platform_resource_setup_memory' function which is the main user > of 'dma_alloc_coherent()', and inserting a WARN_ON() check to make future > (and existing) bogus users of this function they're should provide a valid > 'struct device' whenever possible. > > Fixes: ce88313069c36eef80f21fd7 ("arch/sh: make the DMA mapping operations observe dev->dma_pfn_offset") > Signed-off-by: Jacopo Mondi I would have done two commits here, one to fix: dma_alloc_coherent(&pdev->dev, memsize, &dma_handle, GFP_KERNEL); and one to switch to the WARN_ON + if(dev) model. But I don't really care either way, so: Reviewed-by: Thomas Petazzoni Note that even with the if (dev) check, you don't avoid all possible regressions. For example, some parts of the sh_eth driver were passing a non-NULL struct device, but it was the wrong struct device (the one inside struct net_device, and not the one part of struct platform_device). I fixed that for sh_eth, but there could be other drivers doing bogus things. Best regards, Thomas -- Thomas Petazzoni, CTO, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering https://bootlin.com