* [PATCH] iommu/amd: Fix ineffective error check in nested domain allocation
@ 2026-08-25 10:05 Hemanth Selam
2026-08-25 11:14 ` Vasant Hegde
0 siblings, 1 reply; 2+ messages in thread
From: Hemanth Selam @ 2026-08-25 10:05 UTC (permalink / raw)
To: joro
Cc: suravee.suthikulpanit, vasant.hegde, will, robin.murphy, jgg,
iommu, linux-kernel
amd_iommu_pdom_id_alloc() returns an int: a domain ID on success, or the
negative errno from ida_alloc_range() when the ID space is exhausted or
memory is short. amd_iommu_alloc_domain_nested() stores that return value
in gdom_info->hdom_id, which is a u32, and only then tests it:
gdom_info->hdom_id = amd_iommu_pdom_id_alloc();
if (gdom_info->hdom_id <= 0) {
The assignment discards the sign, so -ENOSPC becomes 0xffffffe4 and the
test never fires. The nested domain is then set up with a host domain ID
that was never allocated, instead of the allocation failing with -ENOSPC.
Keep the value in an int, test it there, and store it only once it is
known to be valid, which is what the other amd_iommu_pdom_id_alloc()
callers already do.
Fixes: 757d2b1fdf5b ("iommu/amd: Introduce gDomID-to-hDomID Mapping and handle parent domain invalidation")
Signed-off-by: Hemanth Selam <hemanth.selam@gmail.com>
---
Verified by driving amd_iommu_alloc_domain_nested() directly from a debugfs
hook, since the normal entry point needs iommufd to request a nested HWPT on
real hardware. No error injection was used: the harness allocates domain IDs
until ida_alloc_range() genuinely runs out, then calls the function with a
zeroed guest DTE, which is what validate_gdte_nested() accepts.
Without this patch:
NESTEDTEST: first alloc, IDs available -> success, hdom_id=1 (0x00000001)
NESTEDTEST: second alloc, same gDomID (reuse) -> success, hdom_id=1 (0x00000001)
NESTEDTEST: exhausted the ID space after 65534 IDs, allocator returns -28
NESTEDTEST: third alloc, IDs exhausted -> success, hdom_id=4294967268 (0xffffffe4)
With this patch:
NESTEDTEST: first alloc, IDs available -> success, hdom_id=1 (0x00000001)
NESTEDTEST: second alloc, same gDomID (reuse) -> success, hdom_id=1 (0x00000001)
NESTEDTEST: exhausted the ID space after 65534 IDs, allocator returns -28
NESTEDTEST: third alloc, IDs exhausted -> error -28
The successful and the reuse paths are unchanged; only the exhausted case
differs, and the result was the same in three runs of each kernel. Built with
CONFIG_AMD_IOMMU_IOMMUFD=y; sparse and checkpatch are clean.
drivers/iommu/amd/nested.c | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/drivers/iommu/amd/nested.c b/drivers/iommu/amd/nested.c
index 63b53b29e029..f1c7987fc585 100644
--- a/drivers/iommu/amd/nested.c
+++ b/drivers/iommu/amd/nested.c
@@ -96,7 +96,7 @@ struct iommu_domain *
amd_iommu_alloc_domain_nested(struct iommufd_viommu *viommu, u32 flags,
const struct iommu_user_data *user_data)
{
- int ret;
+ int ret, hdom_id;
unsigned long irqflags;
struct nested_domain *ndom;
struct guest_domain_mapping_info *gdom_info;
@@ -161,8 +161,8 @@ amd_iommu_alloc_domain_nested(struct iommufd_viommu *viommu, u32 flags,
}
/* The gDomID does not exist. We allocate new hdom_id */
- gdom_info->hdom_id = amd_iommu_pdom_id_alloc();
- if (gdom_info->hdom_id <= 0) {
+ hdom_id = amd_iommu_pdom_id_alloc();
+ if (hdom_id <= 0) {
__xa_cmpxchg(&aviommu->gdomid_array,
ndom->gdom_id, gdom_info, NULL, GFP_ATOMIC);
xa_unlock_irqrestore(&aviommu->gdomid_array, irqflags);
@@ -170,6 +170,7 @@ amd_iommu_alloc_domain_nested(struct iommufd_viommu *viommu, u32 flags,
goto out_err_gdom_info;
}
+ gdom_info->hdom_id = hdom_id;
ndom->gdom_info = gdom_info;
refcount_set(&gdom_info->users, 1);
--
2.43.7
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH] iommu/amd: Fix ineffective error check in nested domain allocation
2026-08-25 10:05 [PATCH] iommu/amd: Fix ineffective error check in nested domain allocation Hemanth Selam
@ 2026-08-25 11:14 ` Vasant Hegde
0 siblings, 0 replies; 2+ messages in thread
From: Vasant Hegde @ 2026-08-25 11:14 UTC (permalink / raw)
To: Hemanth Selam, joro
Cc: suravee.suthikulpanit, will, robin.murphy, jgg, iommu,
linux-kernel
On 8/25/2026 3:35 PM, Hemanth Selam wrote:
>
> amd_iommu_pdom_id_alloc() returns an int: a domain ID on success, or the
> negative errno from ida_alloc_range() when the ID space is exhausted or
> memory is short. amd_iommu_alloc_domain_nested() stores that return value
> in gdom_info->hdom_id, which is a u32, and only then tests it:
>
> gdom_info->hdom_id = amd_iommu_pdom_id_alloc();
> if (gdom_info->hdom_id <= 0) {
>
> The assignment discards the sign, so -ENOSPC becomes 0xffffffe4 and the
> test never fires. The nested domain is then set up with a host domain ID
> that was never allocated, instead of the allocation failing with -ENOSPC.
>
> Keep the value in an int, test it there, and store it only once it is
> known to be valid, which is what the other amd_iommu_pdom_id_alloc()
> callers already do.
>
> Fixes: 757d2b1fdf5b ("iommu/amd: Introduce gDomID-to-hDomID Mapping and handle parent domain invalidation")
> Signed-off-by: Hemanth Selam <hemanth.selam@gmail.com>
Reviewed-by: Vasant Hegde <vasant.hegde@amd.com>
-Vasant
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-08-25 11:14 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-08-25 10:05 [PATCH] iommu/amd: Fix ineffective error check in nested domain allocation Hemanth Selam
2026-08-25 11:14 ` Vasant Hegde
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox