From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-acpi-owner@vger.kernel.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII;
 format=flowed
Date: Thu, 02 Feb 2017 14:01:52 -0500
From: Nate Watterson <nwatters@codeaurora.org>
To: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Sinan Kaya <okaya@codeaurora.org>,
        Will Deacon <will.deacon@arm.com>,
        Robin Murphy <robin.murphy@arm.com>,
        Sricharan R <sricharan@codeaurora.org>,
        linux-arm-msm@vger.kernel.org, linux-acpi@vger.kernel.org,
        iommu@lists.linux-foundation.org, linux-pci@vger.kernel.org,
        bhelgaas@google.com, Tomasz Nowicki <tn@semihalf.com>,
        linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH V7 08/11] drivers: acpi: Handle IOMMU lookup failure with
 deferred probing or error
In-Reply-To: <20170201185230.GA1922@red-moon>
References: <f2370671-ece8-fbee-de19-350dd625088f@codeaurora.org>
 <93e79759-d614-9b36-d5ab-63e8eb725009@arm.com>
 <14751205-f034-7f0d-442a-854c3909425c@codeaurora.org>
 <5ba9f366d6e25397cdef8ad95b49e199@codeaurora.org>
 <175a3798-b824-ef1a-e112-9f6f472973ae@codeaurora.org>
 <20170130143851.GJ16461@arm.com>
 <b2e50c17de156341deab9fb8e91a8dec@codeaurora.org>
 <1e048aff-0d77-b9f2-ebf8-2ba315b90ca7@codeaurora.org>
 <20170130165124.GA11712@red-moon>
 <f401e27e-d619-0eec-2401-50d8e2fc2f0f@codeaurora.org>
 <20170201185230.GA1922@red-moon>
Message-ID: <3f4aa4c1e660c7e62256de9ca64fc822@codeaurora.org>
Sender: linux-acpi-owner@vger.kernel.org
List-ID: <linux-pci.vger.kernel.org>

On 2017-02-01 13:52, Lorenzo Pieralisi wrote:
> 
> I debugged the issue and Nate's fix is correct, the fact that you
> can't it hit it with mainline is just a matter of timing because it has
> to do with the CTX pointer value (we OR it with the existing value), so
> it may work or not depending on how the cdptr memory allocation
> pattern turns out to be (which explains why Nate and I can hit it with
> simple PCI device remove/add execution too).
> 
> So it is neither an ACPI nor an IOMMU probe deferral issue per-se,
> fix is already queued, so it is all good.
> 
> What about USB stalls ?
Our fault. The USB controller was getting 48-bit IOVAs, but the
bus it sits on only supports 44-bits so the controller's DMA
transactions ended up targeting addresses that were never mapped.

It started working once I applied the iort/dma-mapping patches I
sent out earlier this week that use the iort memory_address_limit
field to check if a dma_mask is sane.

Sorry for the false alarm.
Nate
> 
> Thanks !
> Lorenzo

-- 
Qualcomm Datacenter Technologies, Inc. on behalf of Qualcomm 
Technologies, Inc.
Qualcomm Technologies, Inc. is a member of the Code Aurora Forum, a 
Linux
Foundation Collaborative Project.