linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@lst.de>
To: Robin Murphy <robin.murphy@arm.com>
Cc: Joerg Roedel <joro@8bytes.org>,
	Catalin Marinas <catalin.marinas@arm.com>,
	Will Deacon <will.deacon@arm.com>,
	Tom Lendacky <thomas.lendacky@amd.com>,
	iommu@lists.linux-foundation.org,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org
Subject: [PATCH 07/19] dma-iommu: fix and refactor iommu_dma_get_sgtable
Date: Mon, 14 Jan 2019 10:41:47 +0100	[thread overview]
Message-ID: <20190114094159.27326-8-hch@lst.de> (raw)
In-Reply-To: <20190114094159.27326-1-hch@lst.de>

The current iommu_dma_get_sgtable code does not properly handle memory
from the page allocator that hasn't been remapped, which can happen in
the rare case of allocations for a coherent device that aren't allowed
to block.

Fix this by replacing iommu_dma_get_sgtable with a slightly tweaked copy
of dma_common_get_sgtable with special handling for the remapped array
of pages allocated from __iommu_dma_alloc.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/iommu/dma-iommu.c | 49 +++++++++++++++++++--------------------
 1 file changed, 24 insertions(+), 25 deletions(-)

diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
index 26f479d49103..8f3dc6ab3da1 100644
--- a/drivers/iommu/dma-iommu.c
+++ b/drivers/iommu/dma-iommu.c
@@ -620,6 +620,18 @@ static int iommu_dma_mmap_remap(void *cpu_addr, size_t size,
 	return ret;
 }
 
+static int iommu_dma_get_sgtable_remap(struct sg_table *sgt, void *cpu_addr,
+		size_t size)
+{
+	unsigned int count = PAGE_ALIGN(size) >> PAGE_SHIFT;
+	struct vm_struct *area = find_vm_area(cpu_addr);
+
+	if (WARN_ON(!area || !area->pages))
+		return -ENXIO;
+	return sg_alloc_table_from_pages(sgt, area->pages, count, 0, size,
+			GFP_KERNEL);
+}
+
 static void iommu_dma_sync_single_for_cpu(struct device *dev,
 		dma_addr_t dma_handle, size_t size, enum dma_data_direction dir)
 {
@@ -1080,37 +1092,24 @@ static int iommu_dma_mmap(struct device *dev, struct vm_area_struct *vma,
 			user_count << PAGE_SHIFT, vma->vm_page_prot);
 }
 
-static int __iommu_dma_get_sgtable_page(struct sg_table *sgt, struct page *page,
-		size_t size)
-{
-	int ret = sg_alloc_table(sgt, 1, GFP_KERNEL);
-
-	if (!ret)
-		sg_set_page(sgt->sgl, page, PAGE_ALIGN(size), 0);
-	return ret;
-}
-
 static int iommu_dma_get_sgtable(struct device *dev, struct sg_table *sgt,
 		void *cpu_addr, dma_addr_t dma_addr, size_t size,
 		unsigned long attrs)
 {
-	unsigned int count = PAGE_ALIGN(size) >> PAGE_SHIFT;
-	struct vm_struct *area = find_vm_area(cpu_addr);
-
-	if (attrs & DMA_ATTR_FORCE_CONTIGUOUS) {
-		/*
-		 * DMA_ATTR_FORCE_CONTIGUOUS allocations are always remapped,
-		 * hence in the vmalloc space.
-		 */
-		struct page *page = vmalloc_to_page(cpu_addr);
-		return __iommu_dma_get_sgtable_page(sgt, page, size);
-	}
+	struct page *page;
+	int ret;
 
-	if (WARN_ON(!area || !area->pages))
-		return -ENXIO;
+	if (is_vmalloc_addr(cpu_addr)) {
+		if (!(attrs & DMA_ATTR_FORCE_CONTIGUOUS))
+			return iommu_dma_get_sgtable_remap(sgt, cpu_addr, size);
+		page = vmalloc_to_page(cpu_addr);
+	} else
+		page = virt_to_page(cpu_addr);
 
-	return sg_alloc_table_from_pages(sgt, area->pages, count, 0, size,
-					 GFP_KERNEL);
+	ret = sg_alloc_table(sgt, 1, GFP_KERNEL);
+	if (!ret)
+		sg_set_page(sgt->sgl, page, PAGE_ALIGN(size), 0);
+	return ret;
 }
 
 static const struct dma_map_ops iommu_dma_ops = {
-- 
2.20.1


  parent reply	other threads:[~2019-01-14  9:42 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-01-14  9:41 implement generic dma_map_ops for IOMMUs Christoph Hellwig
2019-01-14  9:41 ` [PATCH 01/19] dma-mapping: add a Kconfig symbol to indicated arch_dma_prep_coherent presence Christoph Hellwig
2019-02-01 14:22   ` Robin Murphy
2019-02-01 16:12     ` Christoph Hellwig
2019-01-14  9:41 ` [PATCH 02/19] dma-iommu: cleanup dma-iommu.h Christoph Hellwig
2019-02-01 14:47   ` Robin Murphy
2019-02-01 16:13     ` Christoph Hellwig
2019-02-06 15:08       ` Robin Murphy
2019-02-11 15:59         ` Christoph Hellwig
2019-01-14  9:41 ` [PATCH 03/19] dma-iommu: don't use a scatterlist in iommu_dma_alloc Christoph Hellwig
2019-02-01 15:24   ` Robin Murphy
2019-02-01 16:16     ` Christoph Hellwig
2019-02-06 15:28       ` Robin Murphy
2019-02-11 16:00         ` Christoph Hellwig
2019-01-14  9:41 ` [PATCH 04/19] dma-iommu: remove the flush_page callback Christoph Hellwig
2019-02-01 15:28   ` Robin Murphy
2019-01-14  9:41 ` [PATCH 05/19] dma-iommu: move the arm64 wrappers to common code Christoph Hellwig
2019-01-14  9:41 ` [PATCH 06/19] dma-iommu: fix and refactor iommu_dma_mmap Christoph Hellwig
2019-02-05 15:02   ` Robin Murphy
2019-02-11 16:03     ` Christoph Hellwig
2019-01-14  9:41 ` Christoph Hellwig [this message]
2019-01-14  9:41 ` [PATCH 08/19] dma-iommu: move __iommu_dma_map Christoph Hellwig
2019-01-14  9:41 ` [PATCH 09/19] dma-iommu: refactor page array remap helpers Christoph Hellwig
2019-01-14  9:41 ` [PATCH 10/19] dma-iommu: factor atomic pool allocations into helpers Christoph Hellwig
2019-01-14  9:41 ` [PATCH 11/19] dma-iommu: factor contiguous " Christoph Hellwig
2019-01-14  9:41 ` [PATCH 12/19] dma-iommu: refactor iommu_dma_free Christoph Hellwig
2019-01-14  9:41 ` [PATCH 13/19] dma-iommu: don't remap contiguous allocations for coherent devices Christoph Hellwig
2019-01-14  9:41 ` [PATCH 14/19] dma-iommu: factor contiguous remapped allocations into helpers Christoph Hellwig
2019-01-14  9:41 ` [PATCH 15/19] dma-iommu: refactor iommu_dma_alloc Christoph Hellwig
2019-01-14  9:41 ` [PATCH 16/19] dma-iommu: don't depend on CONFIG_DMA_DIRECT_REMAP Christoph Hellwig
2019-02-06 11:55   ` Robin Murphy
2019-02-11 16:39     ` Christoph Hellwig
2019-01-14  9:41 ` [PATCH 17/19] dma-iommu: switch copyright boilerplace to SPDX Christoph Hellwig
2019-02-06 11:57   ` Robin Murphy
2019-01-14  9:41 ` [PATCH 18/19] arm64: switch copyright boilerplace to SPDX in dma-mapping.c Christoph Hellwig
2019-02-06 12:19   ` Robin Murphy
2019-01-14  9:41 ` [PATCH 19/19] arm64: trim includes " Christoph Hellwig
2019-01-28  7:53 ` implement generic dma_map_ops for IOMMUs Christoph Hellwig

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=20190114094159.27326-8-hch@lst.de \
    --to=hch@lst.de \
    --cc=catalin.marinas@arm.com \
    --cc=iommu@lists.linux-foundation.org \
    --cc=joro@8bytes.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=robin.murphy@arm.com \
    --cc=thomas.lendacky@amd.com \
    --cc=will.deacon@arm.com \
    /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;
as well as URLs for NNTP newsgroup(s).