Linux IOMMU Development
 help / color / mirror / Atom feed
From: Will Deacon <will.deacon-5wv7dgnIgG8@public.gmane.org>
To: iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org
Cc: jroedel-l3A5Bk7waGM@public.gmane.org,
	marc.zyngier-5wv7dgnIgG8@public.gmane.org,
	Will Deacon <will.deacon-5wv7dgnIgG8@public.gmane.org>,
	laurent.pinchart+renesas-ryLnwIuWjnjg/C1BVhZhaw@public.gmane.org,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org
Subject: [PATCH 04/13] iommu/arm-smmu: Sort out coherency
Date: Mon,  3 Aug 2015 14:25:46 +0100	[thread overview]
Message-ID: <1438608355-7335-5-git-send-email-will.deacon@arm.com> (raw)
In-Reply-To: <1438608355-7335-1-git-send-email-will.deacon-5wv7dgnIgG8@public.gmane.org>

From: Robin Murphy <Robin.Murphy-5wv7dgnIgG8@public.gmane.org>

Currently, we detect whether the SMMU has coherent page table walk
capability from the IDR0.CTTW field, and base our cache maintenance
decisions on that. In preparation for fixing the bogus DMA API usage,
however, we need to ensure that the DMA API agrees about this, which
necessitates deferring to the dma-coherent property in the device tree
for the final say.

As an added bonus, since systems exist where an external CTTW signal
has been tied off incorrectly at integration, allowing DT to override
it offers a neat workaround for coherency issues with such SMMUs.

Signed-off-by: Robin Murphy <robin.murphy-5wv7dgnIgG8@public.gmane.org>
Signed-off-by: Will Deacon <will.deacon-5wv7dgnIgG8@public.gmane.org>
---
 Documentation/devicetree/bindings/iommu/arm,smmu.txt |  6 ++++++
 drivers/iommu/arm-smmu.c                             | 20 +++++++++++++++++---
 2 files changed, 23 insertions(+), 3 deletions(-)

diff --git a/Documentation/devicetree/bindings/iommu/arm,smmu.txt b/Documentation/devicetree/bindings/iommu/arm,smmu.txt
index 06760503a819..718074501fcb 100644
--- a/Documentation/devicetree/bindings/iommu/arm,smmu.txt
+++ b/Documentation/devicetree/bindings/iommu/arm,smmu.txt
@@ -43,6 +43,12 @@ conditions.
 
 ** System MMU optional properties:
 
+- dma-coherent  : Present if page table walks made by the SMMU are
+                  cache coherent with the CPU.
+
+                  NOTE: this only applies to the SMMU itself, not
+                  masters connected upstream of the SMMU.
+
 - calxeda,smmu-secure-config-access : Enable proper handling of buggy
                   implementations that always use secure access to
                   SMMU configuration registers. In this case non-secure
diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c
index 4cd0c29cb585..0583ed2f33c0 100644
--- a/drivers/iommu/arm-smmu.c
+++ b/drivers/iommu/arm-smmu.c
@@ -37,6 +37,7 @@
 #include <linux/iopoll.h>
 #include <linux/module.h>
 #include <linux/of.h>
+#include <linux/of_address.h>
 #include <linux/pci.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
@@ -1532,6 +1533,7 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 	unsigned long size;
 	void __iomem *gr0_base = ARM_SMMU_GR0(smmu);
 	u32 id;
+	bool cttw_dt, cttw_reg;
 
 	dev_notice(smmu->dev, "probing hardware configuration...\n");
 	dev_notice(smmu->dev, "SMMUv%d with:\n", smmu->version);
@@ -1571,10 +1573,22 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 		dev_notice(smmu->dev, "\taddress translation ops\n");
 	}
 
-	if (id & ID0_CTTW) {
+	/*
+	 * In order for DMA API calls to work properly, we must defer to what
+	 * the DT says about coherency, regardless of what the hardware claims.
+	 * Fortunately, this also opens up a workaround for systems where the
+	 * ID register value has ended up configured incorrectly.
+	 */
+	cttw_dt = of_dma_is_coherent(smmu->dev->of_node);
+	cttw_reg = !!(id & ID0_CTTW);
+	if (cttw_dt)
 		smmu->features |= ARM_SMMU_FEAT_COHERENT_WALK;
-		dev_notice(smmu->dev, "\tcoherent table walk\n");
-	}
+	if (cttw_dt || cttw_reg)
+		dev_notice(smmu->dev, "\t%scoherent table walk\n",
+			   cttw_dt ? "" : "non-");
+	if (cttw_dt != cttw_reg)
+		dev_notice(smmu->dev,
+			   "\t(IDR0.CTTW overridden by dma-coherent property)\n");
 
 	if (id & ID0_SMS) {
 		u32 smr, sid, mask;
-- 
2.1.4

  parent reply	other threads:[~2015-08-03 13:25 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-08-03 13:25 [PATCH 00/13] iommu/arm-smmu: Updates for 4.3 Will Deacon
     [not found] ` <1438608355-7335-1-git-send-email-will.deacon-5wv7dgnIgG8@public.gmane.org>
2015-08-03 13:25   ` [PATCH 01/13] iommu/arm-smmu: Fix enabling of PRIQ interrupt Will Deacon
2015-08-03 13:25   ` [PATCH 02/13] iommu/arm-smmu: Fix MSI memory attributes to match specification Will Deacon
2015-08-03 13:25   ` [PATCH 03/13] iommu/arm-smmu: Limit 2-level strtab allocation for small SID sizes Will Deacon
2015-08-03 13:25   ` Will Deacon [this message]
2015-08-03 13:25   ` [PATCH 05/13] iommu/io-pgtable-arm: Allow appropriate DMA API use Will Deacon
     [not found]     ` <1438608355-7335-6-git-send-email-will.deacon-5wv7dgnIgG8@public.gmane.org>
2015-08-04 13:16       ` Laurent Pinchart
2015-08-04 14:47         ` Robin Murphy
     [not found]           ` <55C0D071.1040104-5wv7dgnIgG8@public.gmane.org>
2015-08-04 14:56             ` Russell King - ARM Linux
     [not found]               ` <20150804145642.GQ7557-l+eeeJia6m9vn6HldHNs0ANdhmdF6hFW@public.gmane.org>
2015-08-04 20:54                 ` Laurent Pinchart
2015-08-05 16:24                   ` Will Deacon
     [not found]                     ` <20150805162452.GH6092-5wv7dgnIgG8@public.gmane.org>
2015-08-06 19:10                       ` Laurent Pinchart
2015-08-03 13:25   ` [PATCH 06/13] iommu/arm-smmu: Clean up DMA API usage Will Deacon
2015-08-03 13:25   ` [PATCH 07/13] " Will Deacon
2015-08-03 13:25   ` [PATCH 08/13] iommu/ipmmu-vmsa: " Will Deacon
2015-08-03 13:25   ` [PATCH 09/13] iommu/io-pgtable-arm: Centralise sync points Will Deacon
2015-08-03 13:25   ` [PATCH 10/13] iommu/arm-smmu: Remove arm_smmu_flush_pgtable() Will Deacon
2015-08-03 13:25   ` [PATCH 11/13] " Will Deacon
2015-08-03 13:25   ` [PATCH 12/13] iommu/io-pgtable: Remove flush_pgtable callback Will Deacon
2015-08-03 13:25   ` [PATCH 13/13] iommu/arm-smmu: Treat unknown OAS as 48-bit Will Deacon
     [not found]     ` <1438608355-7335-14-git-send-email-will.deacon-5wv7dgnIgG8@public.gmane.org>
2015-08-03 18:23       ` Sergei Shtylyov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1438608355-7335-5-git-send-email-will.deacon@arm.com \
    --to=will.deacon-5wv7dgnigg8@public.gmane.org \
    --cc=iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org \
    --cc=jroedel-l3A5Bk7waGM@public.gmane.org \
    --cc=laurent.pinchart+renesas-ryLnwIuWjnjg/C1BVhZhaw@public.gmane.org \
    --cc=linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org \
    --cc=marc.zyngier-5wv7dgnIgG8@public.gmane.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox