From: Andrei-Edward Popa <andrei.popa105@yahoo.com>
To: m.szyprowski@samsung.com
Cc: robin.murphy@arm.com, iommu@lists.linux.dev,
linux-kernel@vger.kernel.org,
Andrei-Edward Popa <andrei.popa105@yahoo.com>
Subject: [PATCH] dma-remap: fix dma_common_find_pages() page lookup for offsets
Date: Fri, 12 Dec 2025 22:09:14 +0200 [thread overview]
Message-ID: <20251212200914.138310-1-andrei.popa105@yahoo.com> (raw)
In-Reply-To: <20251212200914.138310-1-andrei.popa105.ref@yahoo.com>
dma_common_find_pages() previously assumed that the CPU virtual address
always pointed to the start of a DMA-coherent allocation. This fails when
memory is allocated via dma_alloc_attrs() without DMA_ATTR_FORCE_CONTIGUOUS
and then subdivided into smaller blocks using a gen_pool, relevant only
when an IOMMU is enabled.
In such cases, userspace may request a mapping via dma_mmap_attrs()
for a CPU address that is offset inside the original allocation. The
previous code could return the wrong struct page pointer.
Example scenario:
- Allocate a large DMA buffer with dma_alloc_attrs() (non-contiguous)
- Create a gen_pool using this buffer
- Take sub-allocations from the gen_pool
- Map the sub-allocations to userspace via dma_mmap_attrs()
- dma_common_find_pages() must return the correct struct page for the offset
This patch computes the page index relative to the vm_struct backing
the DMA allocation:
page_index = (vaddr - area->addr) >> PAGE_SHIFT
Bounds checks ensure the CPU address is within the vm_struct and
page_index < area->nr_pages.
This ensures correct behavior for sub-allocated regions from gen_pool,
without affecting allocations starting at the base address or allocations
with DMA_ATTR_FORCE_CONTIGUOUS.
Signed-off-by: Andrei-Edward Popa <andrei.popa105@yahoo.com>
---
kernel/dma/remap.c | 15 ++++++++++++++-
1 file changed, 14 insertions(+), 1 deletion(-)
diff --git a/kernel/dma/remap.c b/kernel/dma/remap.c
index b7c1c0c92d0c..d27477e32ed8 100644
--- a/kernel/dma/remap.c
+++ b/kernel/dma/remap.c
@@ -9,12 +9,25 @@
struct page **dma_common_find_pages(void *cpu_addr)
{
struct vm_struct *area = find_vm_area(cpu_addr);
+ unsigned long vaddr, area_vaddr;
+ size_t page_index;
if (!area || !(area->flags & VM_DMA_COHERENT))
return NULL;
WARN(area->flags != VM_DMA_COHERENT,
"unexpected flags in area: %p\n", cpu_addr);
- return area->pages;
+
+ vaddr = (unsigned long)cpu_addr;
+ area_vaddr = (unsigned long)area->addr;
+ if (unlikely(vaddr < area_vaddr ||
+ vaddr >= area_vaddr + area->size))
+ return NULL;
+
+ page_index = (vaddr - area_vaddr) >> PAGE_SHIFT;
+ if (unlikely(page_index >= area->nr_pages))
+ return NULL;
+
+ return &area->pages[page_index];
}
/*
--
2.52.0
next parent reply other threads:[~2025-12-12 21:00 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20251212200914.138310-1-andrei.popa105.ref@yahoo.com>
2025-12-12 20:09 ` Andrei-Edward Popa [this message]
2025-12-15 5:57 ` Christoph Hellwig
2025-12-15 8:13 ` Marek Szyprowski
2025-12-15 11:31 ` Robin Murphy
2025-12-15 20:49 ` Andrei-Edward Popa
2025-12-15 10:52 ` Robin Murphy
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=20251212200914.138310-1-andrei.popa105@yahoo.com \
--to=andrei.popa105@yahoo.com \
--cc=iommu@lists.linux.dev \
--cc=linux-kernel@vger.kernel.org \
--cc=m.szyprowski@samsung.com \
--cc=robin.murphy@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
all inboxes | Powered by JetHome®