From mboxrd@z Thu Jan 1 00:00:00 1970 From: Benjamin Herrenschmidt Date: Thu, 02 Apr 2015 22:19:36 +0000 Subject: Re: [PATCH v8 RFC 1/3] sparc: Break up monolithic iommu table/lock into finer graularity pools and l Message-Id: <1428013176.20500.285.camel@kernel.crashing.org> List-Id: References: <8406d119fb885255387a400551de994cb4a4c331.1427761300.git.sowmini.varadhan@oracle.com> <1428008044.20500.272.camel@kernel.crashing.org> <20150402214343.GA15680@oracle.com> <1428011820.20500.277.camel@kernel.crashing.org> <20150402221505.GD15680@oracle.com> In-Reply-To: <20150402221505.GD15680@oracle.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Sowmini Varadhan Cc: aik@au1.ibm.com, anton@au1.ibm.com, paulus@samba.org, sparclinux@vger.kernel.org, linuxppc-dev@lists.ozlabs.org, davem@davemloft.net On Thu, 2015-04-02 at 18:15 -0400, Sowmini Varadhan wrote: > On (04/03/15 08:57), Benjamin Herrenschmidt wrote: > > > > > > I only just noticed too, you completely dropped the code to honor > > > > the dma mask. Why that ? Some devices rely on this. > > /* Sowmini's comment about this coming from sparc origins.. */ > > > Probably, not that many devices have limits on DMA mask but they do > > exist. It becomes more important if we decide to create a very large > > IOMMU window that spans beyond 4G in order to support devices with > > 32-bit DMA masks. Otherwise it's older devices mostly with <32-bit > > masks. > > > > In any case, for a generic piece of code, this should be supported. > > Basically, assume that if we have something in the powerpc code, we need > > it, if you remove it, we won't be able to use your code generically. > > I see. > > is the mask something that can be stored in the iommu_map_table as > part of the init? > No, the mask is per device and has to be retrieved from the device. Additionally the mask used for dma_map_* can be different from the consistent mask used for alloc_coherent (we have a bug there on powerpc which I'm trying to fix btw). So it should be passed as an argument by the caller. > I can see that the align_order has to be an additional arg to > iommu_tbl_range_alloc, not sure if mask falls in that category > as well. It does. Cheers, Ben.