* [PATCH V3] accel/amdxdna: Set HMM_PFN_REQ_WRITE for writable BO faults
@ 2026-09-17 17:19 Lizhi Hou
2026-09-17 17:47 ` Max Zhen
0 siblings, 1 reply; 2+ messages in thread
From: Lizhi Hou @ 2026-09-17 17:19 UTC (permalink / raw)
To: ogabbay, quic_jhugo, mario.limonciello, karol.wachowski,
dri-devel, max.zhen
Cc: Lizhi Hou, linux-kernel, sonal.santan
When aie2_populate_range() faults in pages, it only sets
HMM_PFN_REQ_FAULT. Without HMM_PFN_REQ_WRITE, handle_mm_fault() does not
request write permissions and may populate the CPU page tables with
read-only PTEs.
Check the VMA flags and set HMM_PFN_REQ_WRITE when the VMA has VM_WRITE.
Fixes: e486147c912f ("accel/amdxdna: Add BO import and export")
Signed-off-by: Lizhi Hou <lizhi.hou@amd.com>
---
V3:
Revert the V2 change. Instead, add comment to describe why it is not
an issue.
V2:
Fix sashiko comment.
drivers/accel/amdxdna/amdxdna_gem.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/drivers/accel/amdxdna/amdxdna_gem.c b/drivers/accel/amdxdna/amdxdna_gem.c
index e861db6f9369..a5e30855c8a1 100644
--- a/drivers/accel/amdxdna/amdxdna_gem.c
+++ b/drivers/accel/amdxdna/amdxdna_gem.c
@@ -418,7 +418,15 @@ static int amdxdna_hmm_register(struct amdxdna_gem_obj *abo,
mapp->range.notifier = &mapp->notifier;
mapp->range.start = vma->vm_start;
mapp->range.end = vma->vm_end;
+ /*
+ * Access permissions are fixed at mmap() time. Changing them later
+ * with mprotect() is not supported: the range keeps requesting the
+ * original permissions, so the application may see a fault failure
+ * or an IOMMU fault.
+ */
mapp->range.default_flags = HMM_PFN_REQ_FAULT;
+ if (vma->vm_flags & VM_WRITE)
+ mapp->range.default_flags |= HMM_PFN_REQ_WRITE;
mapp->abo = abo;
kref_init(&mapp->refcnt);
--
2.34.1
^ permalink raw reply [flat|nested] 2+ messages in thread* Re: [PATCH V3] accel/amdxdna: Set HMM_PFN_REQ_WRITE for writable BO faults
2026-09-17 17:19 [PATCH V3] accel/amdxdna: Set HMM_PFN_REQ_WRITE for writable BO faults Lizhi Hou
@ 2026-09-17 17:47 ` Max Zhen
0 siblings, 0 replies; 2+ messages in thread
From: Max Zhen @ 2026-09-17 17:47 UTC (permalink / raw)
To: Lizhi Hou, ogabbay, quic_jhugo, mario.limonciello,
karol.wachowski, dri-devel
Cc: linux-kernel, sonal.santan
On 9/17/2026 Thu 10:19, Lizhi Hou wrote:
> When aie2_populate_range() faults in pages, it only sets
> HMM_PFN_REQ_FAULT. Without HMM_PFN_REQ_WRITE, handle_mm_fault() does not
> request write permissions and may populate the CPU page tables with
> read-only PTEs.
>
> Check the VMA flags and set HMM_PFN_REQ_WRITE when the VMA has VM_WRITE.
>
> Fixes: e486147c912f ("accel/amdxdna: Add BO import and export")
> Signed-off-by: Lizhi Hou <lizhi.hou@amd.com>
Reviewed-by: Max Zhen <max.zhen@amd.com>
> ---
> V3:
> Revert the V2 change. Instead, add comment to describe why it is not
> an issue.
> V2:
> Fix sashiko comment.
>
> drivers/accel/amdxdna/amdxdna_gem.c | 8 ++++++++
> 1 file changed, 8 insertions(+)
>
> diff --git a/drivers/accel/amdxdna/amdxdna_gem.c b/drivers/accel/amdxdna/amdxdna_gem.c
> index e861db6f9369..a5e30855c8a1 100644
> --- a/drivers/accel/amdxdna/amdxdna_gem.c
> +++ b/drivers/accel/amdxdna/amdxdna_gem.c
> @@ -418,7 +418,15 @@ static int amdxdna_hmm_register(struct amdxdna_gem_obj *abo,
> mapp->range.notifier = &mapp->notifier;
> mapp->range.start = vma->vm_start;
> mapp->range.end = vma->vm_end;
> + /*
> + * Access permissions are fixed at mmap() time. Changing them later
> + * with mprotect() is not supported: the range keeps requesting the
> + * original permissions, so the application may see a fault failure
> + * or an IOMMU fault.
> + */
> mapp->range.default_flags = HMM_PFN_REQ_FAULT;
> + if (vma->vm_flags & VM_WRITE)
> + mapp->range.default_flags |= HMM_PFN_REQ_WRITE;
> mapp->abo = abo;
> kref_init(&mapp->refcnt);
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-09-17 17:48 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-09-17 17:19 [PATCH V3] accel/amdxdna: Set HMM_PFN_REQ_WRITE for writable BO faults Lizhi Hou
2026-09-17 17:47 ` Max Zhen
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®