mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Lin Liu <lin.liu01@citrix.com>
To: jgross@suse.com, sstabellini@kernel.org
Cc: oleksandr_tyshchenko@epam.com, andrew.cooper3@citrix.com,
	ross.lagerwall@citrix.com, xen-devel@lists.xenproject.org,
	iommu@lists.linux.dev, linux-kernel@vger.kernel.org
Subject: [PATCH v2] xen/swiotlb: Bounce mappings that cross the segment boundary
Date: Thu, 19 Nov 2026 06:38:32 +0000	[thread overview]
Message-ID: <4a2d59f8d5a2b1399fc8bdf6e49202c9c2f394a7.1795069501.git.lin.liu01@citrix.com> (raw)
In-Reply-To: <807c3b040b02d7d89f1cac76eb137afbee91015d.1794573814.git.lin.liu01@citrix.com>

The block layer applies a device's DMA segment boundary to
pseudo-physical addresses, but in a Xen PV domain the bus address is the
machine address.  A segment that is contiguous in both spaces can still
cross the boundary in machine space and reach the device.  AHCI declares
a 4 GiB boundary, and controllers that do not carry across it (e.g.
Marvell 88SE9230) then DMA the tail of the transfer 4 GiB too low.

Bounce such streaming mappings.  Coherent allocations need no check:
they are naturally aligned in bus space (range_requires_alignment()), so
they cannot cross a boundary they fit in.

Signed-off-by: Lin Liu <lin.liu01@citrix.com>
---
v2:
- Only check streaming mappings; coherent allocations are naturally
  aligned in bus space and cannot cross the boundary (Ross, Jürgen).
- Leave range_straddles_page_boundary() unchanged.
- Comment why alloc_coherent needs no boundary check (Jürgen).

 drivers/xen/swiotlb-xen.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index e2538824e..5041be7d3 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -96,6 +96,15 @@ static inline int range_straddles_page_boundary(phys_addr_t p, size_t size)
 	return 0;
 }
 
+/* Bus-contiguous, but may still cross the device's segment boundary */
+static inline bool range_crosses_seg_boundary(struct device *dev,
+					      dma_addr_t dma, size_t size)
+{
+	unsigned long mask = dma_get_seg_boundary(dev);
+
+	return size && ((dma ^ (dma + size - 1)) & ~mask);
+}
+
 static struct io_tlb_pool *xen_swiotlb_find_pool(struct device *dev,
 						 dma_addr_t dma_addr)
 {
@@ -157,6 +166,7 @@ xen_swiotlb_alloc_coherent(struct device *dev, size_t size,
 	phys = virt_to_phys(ret);
 
 	*dma_handle = xen_phys_to_dma(dev, phys);
+	/* No seg boundary check: __get_free_pages() is naturally aligned */
 	if (*dma_handle + size - 1 > dma_mask ||
 	    range_straddles_page_boundary(phys, size) ||
 	    range_requires_alignment(phys, size)) {
@@ -234,6 +244,7 @@ static dma_addr_t xen_swiotlb_map_phys(struct device *dev, phys_addr_t phys,
 	if (dma_capable(dev, dev_addr, size, true, attrs) &&
 	    !dma_kmalloc_needs_bounce(dev, size, dir) &&
 	    !range_straddles_page_boundary(phys, size) &&
+	    !range_crosses_seg_boundary(dev, dev_addr, size) &&
 		!xen_arch_need_swiotlb(dev, phys, dev_addr) &&
 		!is_swiotlb_force_bounce(dev))
 		goto done;

base-commit: 7e1392a42ea288d70a266f6553fe06f57f8647ea
-- 
2.52.0


  parent reply	other threads:[~2026-11-19  6:38 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-13 12:47 [PATCH] xen/swiotlb: Check the segment boundary in range_straddles_page_boundary() Lin Liu
2026-11-18 11:25 ` Ross Lagerwall
2026-11-18 12:24   ` Jürgen Groß
2026-11-19  4:35     ` Lin Liu
2026-11-19  5:59       ` Jürgen Groß
2026-11-19  6:12         ` Lin Liu
2026-11-19  6:38 ` Lin Liu [this message]
2026-11-19  6:44   ` [PATCH v2] xen/swiotlb: Bounce mappings that cross the segment boundary Jürgen Groß

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=4a2d59f8d5a2b1399fc8bdf6e49202c9c2f394a7.1795069501.git.lin.liu01@citrix.com \
    --to=lin.liu01@citrix.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=iommu@lists.linux.dev \
    --cc=jgross@suse.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=oleksandr_tyshchenko@epam.com \
    --cc=ross.lagerwall@citrix.com \
    --cc=sstabellini@kernel.org \
    --cc=xen-devel@lists.xenproject.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

all inboxes | Powered by JetHome®