| Message ID | 1409788712-10741-3-git-send-email-s-anna@ti.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Suman, Thank you for the patch. On Wednesday 03 September 2014 18:58:32 Suman Anna wrote: > A device is tied to an iommu through its archdata field. The archdata > is allocated on the fly for DT-based devices automatically through the > .add_device iommu ops. The current logic incorrectly assigned the name > of the IOMMU user device, instead of the name of the IOMMU device as > required by the attach logic. Fix this issue so that DT-based devices > can attach successfully to an IOMMU domain. > > Signed-off-by: Suman Anna <s-anna@ti.com> > --- > drivers/iommu/omap-iommu.c | 10 +++++++++- > 1 file changed, 9 insertions(+), 1 deletion(-) > > diff --git a/drivers/iommu/omap-iommu.c b/drivers/iommu/omap-iommu.c > index f245d51..f47ac03 100644 > --- a/drivers/iommu/omap-iommu.c > +++ b/drivers/iommu/omap-iommu.c > @@ -26,6 +26,7 @@ > #include <linux/of.h> > #include <linux/of_iommu.h> > #include <linux/of_irq.h> > +#include <linux/of_platform.h> > > #include <asm/cacheflush.h> > > @@ -1244,6 +1245,7 @@ static int omap_iommu_add_device(struct device *dev) > { > struct omap_iommu_arch_data *arch_data; > struct device_node *np; > + struct platform_device *pdev; > > /* > * Allocate the archdata iommu structure for DT-based devices. > @@ -1258,13 +1260,19 @@ static int omap_iommu_add_device(struct device *dev) > if (!np) > return 0; > > + pdev = of_find_device_by_node(np); > + if (WARN_ON(!pdev)) { > + of_node_put(np); > + return -EINVAL; > + } I might be wrong, but I don't think there's a guarantee at this point that the IOMMU device is already instantiated :-/ Will Deacon has posted patches that rework the IOMMU core for better DT integration, have you seen them ? > + > arch_data = kzalloc(sizeof(*arch_data), GFP_KERNEL); > if (!arch_data) { > of_node_put(np); > return -ENOMEM; > } > > - arch_data->name = kstrdup(dev_name(dev), GFP_KERNEL); > + arch_data->name = kstrdup(dev_name(&pdev->dev), GFP_KERNEL); > dev->archdata.iommu = arch_data; > > of_node_put(np);
Hi Laurent, > > On Wednesday 03 September 2014 18:58:32 Suman Anna wrote: >> A device is tied to an iommu through its archdata field. The archdata >> is allocated on the fly for DT-based devices automatically through the >> .add_device iommu ops. The current logic incorrectly assigned the name >> of the IOMMU user device, instead of the name of the IOMMU device as >> required by the attach logic. Fix this issue so that DT-based devices >> can attach successfully to an IOMMU domain. >> >> Signed-off-by: Suman Anna <s-anna@ti.com> >> --- >> drivers/iommu/omap-iommu.c | 10 +++++++++- >> 1 file changed, 9 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/iommu/omap-iommu.c b/drivers/iommu/omap-iommu.c >> index f245d51..f47ac03 100644 >> --- a/drivers/iommu/omap-iommu.c >> +++ b/drivers/iommu/omap-iommu.c >> @@ -26,6 +26,7 @@ >> #include <linux/of.h> >> #include <linux/of_iommu.h> >> #include <linux/of_irq.h> >> +#include <linux/of_platform.h> >> >> #include <asm/cacheflush.h> >> >> @@ -1244,6 +1245,7 @@ static int omap_iommu_add_device(struct device *dev) >> { >> struct omap_iommu_arch_data *arch_data; >> struct device_node *np; >> + struct platform_device *pdev; >> >> /* >> * Allocate the archdata iommu structure for DT-based devices. >> @@ -1258,13 +1260,19 @@ static int omap_iommu_add_device(struct device *dev) >> if (!np) >> return 0; >> >> + pdev = of_find_device_by_node(np); >> + if (WARN_ON(!pdev)) { >> + of_node_put(np); >> + return -EINVAL; >> + } > > I might be wrong, but I don't think there's a guarantee at this point that the > IOMMU device is already instantiated :-/ The omap_iommu_init which registers the iommu_ops is a subsys_initcall, so the platform devices are guaranteed to be created by this point. My test on OMAP4 in fact has the dsp node created before the IOMMU node, and this is not an issue. I have added the WARN_ON in case some one has the IOMMU node disabled, but try to use it. > Will Deacon has posted patches that rework the IOMMU core for better DT > integration, have you seen them ? Can you point out the thread? Are you talking about http://marc.info/?l=linux-arm-kernel&m=140968072117851&w=2? regards Suman > >> + >> arch_data = kzalloc(sizeof(*arch_data), GFP_KERNEL); >> if (!arch_data) { >> of_node_put(np); >> return -ENOMEM; >> } >> >> - arch_data->name = kstrdup(dev_name(dev), GFP_KERNEL); >> + arch_data->name = kstrdup(dev_name(&pdev->dev), GFP_KERNEL); >> dev->archdata.iommu = arch_data; >> >> of_node_put(np); > -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hi Suman, On Thursday 04 September 2014 16:17:53 Suman Anna wrote: > Hi Laurent, > > > On Wednesday 03 September 2014 18:58:32 Suman Anna wrote: > >> A device is tied to an iommu through its archdata field. The archdata > >> is allocated on the fly for DT-based devices automatically through the > >> .add_device iommu ops. The current logic incorrectly assigned the name > >> of the IOMMU user device, instead of the name of the IOMMU device as > >> required by the attach logic. Fix this issue so that DT-based devices > >> can attach successfully to an IOMMU domain. > >> > >> Signed-off-by: Suman Anna <s-anna@ti.com> > >> --- > >> > >> drivers/iommu/omap-iommu.c | 10 +++++++++- > >> 1 file changed, 9 insertions(+), 1 deletion(-) > >> > >> diff --git a/drivers/iommu/omap-iommu.c b/drivers/iommu/omap-iommu.c > >> index f245d51..f47ac03 100644 > >> --- a/drivers/iommu/omap-iommu.c > >> +++ b/drivers/iommu/omap-iommu.c > >> @@ -26,6 +26,7 @@ > >> #include <linux/of.h> > >> #include <linux/of_iommu.h> > >> #include <linux/of_irq.h> > >> +#include <linux/of_platform.h> > >> > >> #include <asm/cacheflush.h> > >> > >> @@ -1244,6 +1245,7 @@ static int omap_iommu_add_device(struct device > >> *dev) > >> { > >> struct omap_iommu_arch_data *arch_data; > >> struct device_node *np; > >> + struct platform_device *pdev; > >> > >> /* > >> * Allocate the archdata iommu structure for DT-based devices. > >> @@ -1258,13 +1260,19 @@ static int omap_iommu_add_device(struct device > >> *dev) > >> if (!np) > >> return 0; > >> > >> + pdev = of_find_device_by_node(np); > >> + if (WARN_ON(!pdev)) { > >> + of_node_put(np); > >> + return -EINVAL; > >> + } > > > > I might be wrong, but I don't think there's a guarantee at this point that > > the IOMMU device is already instantiated :-/ > > The omap_iommu_init which registers the iommu_ops is a subsys_initcall, > so the platform devices are guaranteed to be created by this point. OK, no worries then. > My test on OMAP4 in fact has the dsp node created before the IOMMU node, > and this is not an issue. I have added the WARN_ON in case some one has > the IOMMU node disabled, but try to use it. > > > Will Deacon has posted patches that rework the IOMMU core for better DT > > integration, have you seen them ? > > Can you point out the thread? Are you talking about > http://marc.info/?l=linux-arm-kernel&m=140968072117851&w=2? Yes that's the one.
diff --git a/drivers/iommu/omap-iommu.c b/drivers/iommu/omap-iommu.c index f245d51..f47ac03 100644 --- a/drivers/iommu/omap-iommu.c +++ b/drivers/iommu/omap-iommu.c @@ -26,6 +26,7 @@ #include <linux/of.h> #include <linux/of_iommu.h> #include <linux/of_irq.h> +#include <linux/of_platform.h> #include <asm/cacheflush.h> @@ -1244,6 +1245,7 @@ static int omap_iommu_add_device(struct device *dev) { struct omap_iommu_arch_data *arch_data; struct device_node *np; + struct platform_device *pdev; /* * Allocate the archdata iommu structure for DT-based devices. @@ -1258,13 +1260,19 @@ static int omap_iommu_add_device(struct device *dev) if (!np) return 0; + pdev = of_find_device_by_node(np); + if (WARN_ON(!pdev)) { + of_node_put(np); + return -EINVAL; + } + arch_data = kzalloc(sizeof(*arch_data), GFP_KERNEL); if (!arch_data) { of_node_put(np); return -ENOMEM; } - arch_data->name = kstrdup(dev_name(dev), GFP_KERNEL); + arch_data->name = kstrdup(dev_name(&pdev->dev), GFP_KERNEL); dev->archdata.iommu = arch_data; of_node_put(np);
A device is tied to an iommu through its archdata field. The archdata is allocated on the fly for DT-based devices automatically through the .add_device iommu ops. The current logic incorrectly assigned the name of the IOMMU user device, instead of the name of the IOMMU device as required by the attach logic. Fix this issue so that DT-based devices can attach successfully to an IOMMU domain. Signed-off-by: Suman Anna <s-anna@ti.com> --- drivers/iommu/omap-iommu.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)