LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Lu Baolu <baolu.lu@linux.intel.com>
To: David Woodhouse <dwmw2@infradead.org>, Joerg Roedel <joro@8bytes.org>
Cc: ashok.raj@intel.com, jacob.jun.pan@intel.com,
	kevin.tian@intel.com, jamessewart@arista.com, tmurphy@arista.com,
	dima@arista.com, sai.praneeth.prakhya@intel.com,
	iommu@lists.linux-foundation.org, linux-kernel@vger.kernel.org,
	Lu Baolu <baolu.lu@linux.intel.com>
Subject: [PATCH v4 09/15] iommu/vt-d: Handle 32bit device with identity default domain
Date: Sat, 25 May 2019 13:41:30 +0800	[thread overview]
Message-ID: <20190525054136.27810-10-baolu.lu@linux.intel.com> (raw)
In-Reply-To: <20190525054136.27810-1-baolu.lu@linux.intel.com>

The iommu driver doesn't know whether the bit width of a PCI
device is sufficient for access to the whole system memory.
Hence, the driver checks this when the driver calls into the
dma APIs. If a device is using an identity domain, but the
bit width is less than the system requirement, we need to use
a dma domain instead. This also applies after we delegated
the domain life cycle management to the upper layer.

Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
---
 drivers/iommu/intel-iommu.c | 55 +++++++++++++++----------------------
 1 file changed, 22 insertions(+), 33 deletions(-)

diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c
index 08da484e01d6..b7f5a6390be6 100644
--- a/drivers/iommu/intel-iommu.c
+++ b/drivers/iommu/intel-iommu.c
@@ -3012,25 +3012,6 @@ static int device_def_domain_type(struct device *dev, int startup)
 			return IOMMU_DOMAIN_DMA;
 	}
 
-	/*
-	 * At boot time, we don't yet know if devices will be 64-bit capable.
-	 * Assume that they will — if they turn out not to be, then we can
-	 * take them out of the 1:1 domain later.
-	 */
-	if (!startup) {
-		/*
-		 * If the device's dma_mask is less than the system's memory
-		 * size then this is not a candidate for identity mapping.
-		 */
-		u64 dma_mask = *dev->dma_mask;
-
-		if (dev->coherent_dma_mask &&
-		    dev->coherent_dma_mask < dma_mask)
-			dma_mask = dev->coherent_dma_mask;
-
-		return dma_mask >= dma_get_required_mask(dev);
-	}
-
 	return (iommu_identity_mapping & IDENTMAP_ALL) ?
 			IOMMU_DOMAIN_IDENTITY : 0;
 }
@@ -3642,14 +3623,19 @@ struct dmar_domain *get_valid_domain_for_dev(struct device *dev)
 /* Check if the dev needs to go through non-identity map and unmap process.*/
 static bool iommu_need_mapping(struct device *dev)
 {
-	int found;
+	int ret;
 
 	if (iommu_dummy(dev))
 		return false;
 
-	found = identity_mapping(dev);
-	if (found) {
-		if (iommu_should_identity_map(dev, 0))
+	ret = identity_mapping(dev);
+	if (ret) {
+		u64 dma_mask = *dev->dma_mask;
+
+		if (dev->coherent_dma_mask && dev->coherent_dma_mask < dma_mask)
+			dma_mask = dev->coherent_dma_mask;
+
+		if (dma_mask >= dma_get_required_mask(dev))
 			return false;
 
 		/*
@@ -3657,17 +3643,20 @@ static bool iommu_need_mapping(struct device *dev)
 		 * non-identity mapping.
 		 */
 		dmar_remove_one_dev_info(dev);
-		dev_info(dev, "32bit DMA uses non-identity mapping\n");
-	} else {
-		/*
-		 * In case of a detached 64 bit DMA device from vm, the device
-		 * is put into si_domain for identity mapping.
-		 */
-		if (iommu_should_identity_map(dev, 0) &&
-		    !domain_add_dev_info(si_domain, dev)) {
-			dev_info(dev, "64bit DMA uses identity mapping\n");
-			return false;
+		ret = iommu_request_dma_domain_for_dev(dev);
+		if (ret) {
+			struct iommu_domain *domain;
+			struct dmar_domain *dmar_domain;
+
+			domain = iommu_get_domain_for_dev(dev);
+			if (domain) {
+				dmar_domain = to_dmar_domain(domain);
+				dmar_domain->flags |= DOMAIN_FLAG_LOSE_CHILDREN;
+			}
+			get_valid_domain_for_dev(dev);
 		}
+
+		dev_info(dev, "32bit DMA uses non-identity mapping\n");
 	}
 
 	return true;
-- 
2.17.1


  parent reply	other threads:[~2019-05-25  5:49 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-05-25  5:41 [PATCH v4 00/15] iommu/vt-d: Delegate DMA domain to generic iommu Lu Baolu
2019-05-25  5:41 ` [PATCH v4 01/15] iommu: Add API to request DMA domain for device Lu Baolu
2019-05-25  5:41 ` [PATCH v4 02/15] iommu/vt-d: Implement apply_resv_region iommu ops entry Lu Baolu
2019-05-25  5:41 ` [PATCH v4 03/15] iommu/vt-d: Expose ISA direct mapping region via iommu_get_resv_regions Lu Baolu
2019-05-25  5:41 ` [PATCH v4 04/15] iommu/vt-d: Enable DMA remapping after rmrr mapped Lu Baolu
2019-05-25  5:41 ` [PATCH v4 05/15] iommu/vt-d: Add device_def_domain_type() helper Lu Baolu
2019-05-25  5:41 ` [PATCH v4 06/15] iommu/vt-d: Delegate the identity domain to upper layer Lu Baolu
2019-05-25  5:41 ` [PATCH v4 07/15] iommu/vt-d: Delegate the dma " Lu Baolu
2020-08-21 18:33   ` Chris Wilson
2020-08-24  6:31     ` Lu Baolu
2020-08-24  8:35       ` Chris Wilson
2020-08-25  3:13         ` Lu Baolu
2019-05-25  5:41 ` [PATCH v4 08/15] iommu/vt-d: Identify default domains replaced with private Lu Baolu
2019-05-25  5:41 ` Lu Baolu [this message]
2019-05-25  5:41 ` [PATCH v4 10/15] iommu/vt-d: Probe DMA-capable ACPI name space devices Lu Baolu
2019-05-29  6:16   ` Christoph Hellwig
2019-06-03  0:35     ` Lu Baolu
2019-05-25  5:41 ` [PATCH v4 11/15] iommu/vt-d: Implement is_attach_deferred iommu ops entry Lu Baolu
2019-05-25  5:41 ` [PATCH v4 12/15] iommu/vt-d: Cleanup get_valid_domain_for_dev() Lu Baolu
2019-07-18  3:12   ` Alex Williamson
2019-07-19  9:04     ` Lu Baolu
2019-07-19 15:23       ` Alex Williamson
2019-08-02  1:30         ` Alex Williamson
2019-08-02  7:17           ` Lu Baolu
2019-08-02 16:54             ` Alex Williamson
2019-08-04  3:16               ` Lu Baolu
2019-08-06  0:06               ` Lu Baolu
2019-05-25  5:41 ` [PATCH v4 13/15] iommu/vt-d: Remove startup parameter from device_def_domain_type() Lu Baolu
2019-05-25  5:41 ` [PATCH v4 14/15] iommu/vt-d: Remove duplicated code for device hotplug Lu Baolu
2019-05-25  5:41 ` [PATCH v4 15/15] iommu/vt-d: Remove static identity map code Lu Baolu
2019-05-27 15:00 ` [PATCH v4 00/15] iommu/vt-d: Delegate DMA domain to generic iommu Joerg Roedel

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=20190525054136.27810-10-baolu.lu@linux.intel.com \
    --to=baolu.lu@linux.intel.com \
    --cc=ashok.raj@intel.com \
    --cc=dima@arista.com \
    --cc=dwmw2@infradead.org \
    --cc=iommu@lists.linux-foundation.org \
    --cc=jacob.jun.pan@intel.com \
    --cc=jamessewart@arista.com \
    --cc=joro@8bytes.org \
    --cc=kevin.tian@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=sai.praneeth.prakhya@intel.com \
    --cc=tmurphy@arista.com \
    --subject='Re: [PATCH v4 09/15] iommu/vt-d: Handle 32bit device with identity default domain' \
    /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

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).