Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S966073AbcKKB1h (ORCPT ); Thu, 10 Nov 2016 20:27:37 -0500 Received: from galahad.ideasonboard.com ([185.26.127.97]:55454 "EHLO galahad.ideasonboard.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S965383AbcKKB1g (ORCPT ); Thu, 10 Nov 2016 20:27:36 -0500 From: Laurent Pinchart To: Robin Murphy Cc: Magnus Damm , iommu@lists.linux-foundation.org, laurent.pinchart+renesas@ideasonboard.com, geert+renesas@glider.be, joro@8bytes.org, linux-kernel@vger.kernel.org, linux-renesas-soc@vger.kernel.org, horms+renesas@verge.net.au, m.szyprowski@samsung.com Subject: Re: [PATCH v6 06/07] iommu/ipmmu-vmsa: ARM and ARM64 archdata access Date: Fri, 11 Nov 2016 03:27:38 +0200 Message-ID: <4483634.D6YpsPD8PO@avalon> User-Agent: KMail/4.14.10 (Linux/4.8.6-gentoo; KDE/4.14.24; x86_64; ; ) In-Reply-To: <792320ab-13ae-741e-7202-1e7983e0b7a4@arm.com> References: <20161019233533.10506.16810.sendpatchset@little-apple> <20161019233633.10506.3943.sendpatchset@little-apple> <792320ab-13ae-741e-7202-1e7983e0b7a4@arm.com> MIME-Version: 1.0 Content-Transfer-Encoding: 7Bit Content-Type: text/plain; charset="us-ascii" Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5527 Lines: 190 Hello, On Friday 21 Oct 2016 18:32:54 Robin Murphy wrote: > On 20/10/16 00:36, Magnus Damm wrote: > > From: Magnus Damm > > > > Not all architectures have an iommu member in their archdata, so > > use #ifdefs support build wit COMPILE_TEST on any architecture. > > As an alternative to this we could now use iommu_fwspec in place of the > custom ipmmu_vmsa_archdata - that's deliberately > architecture-independent. I converted the Mediatek drivers[1] as an > example to stand separately from the big SMMU rework, as those are the > ones I'm most familiar with, but it looks like the IPMMU is a similarly > perfect fit. > > Of course, that could always come afterwards as a cleanup - I wouldn't > consider it a blocker at this point - but it does look like it might > simplify some of the code being moved around in this series if it were > to be done beforehand. I like the suggestion, it looks much cleaner. It would also allow implementing .of_xlate() in a cleaner fashion. I don't think it needs to block this series, but if Magnus ends up sending a v7 to address all the other small comments, it would be worth a shot. > [1]:http://www.mail-archive.com/iommu@lists.linux-foundation.org/msg14577.ht > ml > > Signed-off-by: Magnus Damm > > Reviewed-by: Joerg Roedel > > --- > > > > Changes since V5: > > - None > > > > Changes since V4: > > - None > > > > Changes since V3: > > - New patch > > > > drivers/iommu/ipmmu-vmsa.c | 37 +++++++++++++++++++++++++++---------- > > 1 file changed, 27 insertions(+), 10 deletions(-) > > > > --- 0012/drivers/iommu/ipmmu-vmsa.c > > +++ work/drivers/iommu/ipmmu-vmsa.c 2016-09-20 21:59:21.690607110 +0900 > > @@ -70,6 +70,25 @@ static struct ipmmu_vmsa_domain *to_vmsa > > > > return container_of(dom, struct ipmmu_vmsa_domain, io_domain); > > > > } > > > > +#if defined(CONFIG_ARM) || defined(CONFIG_ARM64) > > +static struct ipmmu_vmsa_archdata *to_archdata(struct device *dev) > > +{ > > + return dev->archdata.iommu; > > +} > > +static void set_archdata(struct device *dev, struct ipmmu_vmsa_archdata > > *p) +{ > > + dev->archdata.iommu = p; > > +} > > +#else > > +static struct ipmmu_vmsa_archdata *to_archdata(struct device *dev) > > +{ > > + return NULL; > > +} > > +static void set_archdata(struct device *dev, struct ipmmu_vmsa_archdata > > *p) +{ > > +} > > +#endif > > + > > > > #define TLB_LOOP_TIMEOUT 100 /* 100us */ > > > > /* > > ------------------------------------------------------------------------ > > -----> > > @@ -539,7 +558,7 @@ static void ipmmu_domain_free(struct iom > > > > static int ipmmu_attach_device(struct iommu_domain *io_domain, > > > > struct device *dev) > > > > { > > > > - struct ipmmu_vmsa_archdata *archdata = dev->archdata.iommu; > > + struct ipmmu_vmsa_archdata *archdata = to_archdata(dev); > > > > struct ipmmu_vmsa_device *mmu = archdata->mmu; > > struct ipmmu_vmsa_domain *domain = to_vmsa_domain(io_domain); > > unsigned long flags; > > > > @@ -581,7 +600,7 @@ static int ipmmu_attach_device(struct io > > > > static void ipmmu_detach_device(struct iommu_domain *io_domain, > > > > struct device *dev) > > > > { > > > > - struct ipmmu_vmsa_archdata *archdata = dev->archdata.iommu; > > + struct ipmmu_vmsa_archdata *archdata = to_archdata(dev); > > > > struct ipmmu_vmsa_domain *domain = to_vmsa_domain(io_domain); > > unsigned int i; > > > > @@ -701,7 +720,7 @@ static int ipmmu_init_platform_device(st > > > > archdata->mmu = mmu; > > archdata->utlbs = utlbs; > > archdata->num_utlbs = num_utlbs; > > > > - dev->archdata.iommu = archdata; > > + set_archdata(dev, archdata); > > > > return 0; > > > > error: > > @@ -713,12 +732,11 @@ error: > > static int ipmmu_add_device(struct device *dev) > > { > > > > - struct ipmmu_vmsa_archdata *archdata; > > > > struct ipmmu_vmsa_device *mmu = NULL; > > struct iommu_group *group; > > int ret; > > > > - if (dev->archdata.iommu) { > > + if (to_archdata(dev)) { > > > > dev_warn(dev, "IOMMU driver already assigned to device %s\n", > > > > dev_name(dev)); > > > > return -EINVAL; > > > > @@ -754,8 +772,7 @@ static int ipmmu_add_device(struct devic > > > > * - Make the mapping size configurable ? We currently use a 2GB mapping > > * at a 1GB offset to ensure that NULL VAs will fault. > > */ > > > > - archdata = dev->archdata.iommu; > > - mmu = archdata->mmu; > > + mmu = to_archdata(dev)->mmu; > > > > if (!mmu->mapping) { > > > > struct dma_iommu_mapping *mapping; > > > > @@ -783,7 +800,7 @@ error: > > if (mmu) > > > > arm_iommu_release_mapping(mmu->mapping); > > > > - dev->archdata.iommu = NULL; > > + set_archdata(dev, NULL); > > > > if (!IS_ERR_OR_NULL(group)) > > > > iommu_group_remove_device(dev); > > > > @@ -793,7 +810,7 @@ error: > > static void ipmmu_remove_device(struct device *dev) > > { > > > > - struct ipmmu_vmsa_archdata *archdata = dev->archdata.iommu; > > + struct ipmmu_vmsa_archdata *archdata = to_archdata(dev); > > > > arm_iommu_detach_device(dev); > > iommu_group_remove_device(dev); > > > > @@ -801,7 +818,7 @@ static void ipmmu_remove_device(struct d > > > > kfree(archdata->utlbs); > > kfree(archdata); > > > > - dev->archdata.iommu = NULL; > > + set_archdata(dev, NULL); > > > > } > > > > static struct iommu_domain *ipmmu_domain_alloc(unsigned type) -- Regards, Laurent Pinchart