From: Mark Hounschell <markh@compro.net>
Subject: Re: BUG at drivers/iommu/amd_iommu.c:1436!
Reply-To: markh@compro.net
References: <838275de-6cac-93d8-ca04-54b06a24b3d4@compro.net>
 <20161117214141.GC2078@8bytes.org>
 <f4ababc2-64bf-fd0e-1d18-cb5963d23eb1@compro.net>
 <20161117220045.GD2078@8bytes.org>
 <b8260bfd-55c9-c029-1af1-e1ccc719dd3a@compro.net>
 <20161121153250.GE2078@8bytes.org>
To: Joerg Roedel <joro@8bytes.org>
Cc: iommu@lists.linux-foundation.org,
        Linux-kernel <linux-kernel@vger.kernel.org>
Organization: Compro Computer Svcs.
Message-ID: <4f6afa52-9634-79f4-01af-1ae047aa83f7@compro.net>
Date: Mon, 21 Nov 2016 16:47:59 -0500
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101
 Thunderbird/45.4.0
MIME-Version: 1.0
In-Reply-To: <20161121153250.GE2078@8bytes.org>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1622
Lines: 27

On 11/21/2016 10:32 AM, Joerg Roedel wrote:
> On Fri, Nov 18, 2016 at 09:04:05AM -0500, Mark Hounschell wrote:
>> OK. It's attached.
> 
>> [   45.033715] WARNING: CPU: 0 PID: 0 at lib/dma-debug.c:1158 check_unmap+0x4ef/0x990
>> [   45.037651] 3c59x 0000:04:04.0: DMA-API: device driver failed to check map error[device address=0x00000000ffedf840] [size=1536 bytes] [mapped as single]
> 
> Okay, there is a dma-debug warning from the 3c59x driver. Can you please
> blacklist this driver (or compile it out) and re-try? The dma-debug code
> will only show the first warning and stop then. So any possible warnings
> from your out-of-tree driver will not be shown as long as the 3c59x
> driver is loaded.
> 
> Alternativly you can write 1 to dma-api/all_errors in debug-fs so that
> it will show all errors. But I suggest to unload the 3com driver first
> and then load your driver.
> 

OK, I did get this message before the reported BUG message.

gpiohsd gpiohsd: DMA-API: device driver tries to free DMA memory it has not allocated [device address=0xffffffffffee8000] [size=8192 bytes]

But I've verified that the dma_addr_t that I get for the alloc, and also use for the free is 0x00000000ffee8000 in this case? Is device "address=0xffffffffffee8000" in that message a bug in the message or do we have a sign extended address problem? It seems strange to me, I've never seen a dma_addr_t given, when using the iommu, that high. In the past I've seen them as usually 0x00xxxxxx?

I have also verified that simply changing from pci_alloc/free_consistent to the newer DMA API fixes my issue and I get no such messages.

Mark