mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH] iommu/amd: Skip non-IVMD entries when parsing IVRS memory definitions
@ 2026-10-18  7:45 Akshay Jindal
  0 siblings, 0 replies; only message in thread
From: Akshay Jindal @ 2026-10-18  7:45 UTC (permalink / raw)
  To: iommu
  Cc: Suravee Suthikulpanit, Vasant Hegde, Joerg Roedel, Will Deacon,
	Robin Murphy, linux-kernel, Akshay Jindal, stable

init_memory_definitions() walks every top-level subtable including the
IVHD blocks (type 0x10/0x11/0x40) in the IVRS ACPI table. It interprets
each one as a 'struct ivmd_header', solely on m->flags, without first
checking that the subtable's type actually is one of the IVMD types
(ACPI_IVMD_TYPE, ACPI_IVMD_TYPE_ALL, ACPI_IVMD_TYPE_RANGE).

    if (m->flags & (IVMD_FLAG_UNITY_MAP | IVMD_FLAG_EXCL_RANGE))
            init_unity_map_range(m, table);

IVMD_FLAG_UNITY_MAP (0x01) occupies the same bit position as the IVHD
flags field's HtTunEn bit. Unlike real hardware, QEMU's AMD IOMMU
emulation sets HtTunEn unconditionally (flags=0xD1 for type 0x10,
0x11 for type 0x11), so every QEMU-emitted IVHD block is misidentified
as a unity-map/exclusion-range IVMD entry, which subsequently results
in the creation of a bogus pci segment. Eg:

    AMD-Vi: PCI segment : 0x4000, last bdf : 0x0000

Fix by checking the subtable's type before treating it as IVMD.

Fixes: be2a022c0dd0 ("x86, AMD IOMMU: add functions to parse IOMMU memory mapping requirements for devices")
Cc: stable@vger.kernel.org
Signed-off-by: Akshay Jindal <Akshay.Jindal@amd.com>
---

Testing:
========
- Reproduced via amd_iommu_dump=1 on a QEMU multi-IOMMU (pxb-pcie)
  topology; confirmed "PCI segment : 0x4000, last bdf : 0x0000"
  no longer appears after the fix.
- Verified IVHD device-entry parsing (DEV_SELECT/RANGE/SPECIAL) is
  unaffected by diffing amd_iommu_dump=1 output before/after.
- Booted real hardware (IVHD flags=0xB0, HtTunEn clear) to confirm
  no regression; this bug was not observable there to begin with.

 drivers/iommu/amd/init.c | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/drivers/iommu/amd/init.c b/drivers/iommu/amd/init.c
index 8ec8a6fccaaf..00e1d0be749f 100644
--- a/drivers/iommu/amd/init.c
+++ b/drivers/iommu/amd/init.c
@@ -2716,6 +2716,15 @@ static int __init init_unity_map_range(struct ivmd_header *m,
 	return 0;
 }
 
+static bool ivrs_block_type_ivmd(u8 type)
+{
+	if (type == ACPI_IVMD_TYPE ||
+	    type == ACPI_IVMD_TYPE_ALL ||
+	    type == ACPI_IVMD_TYPE_RANGE)
+		return true;
+	return false;
+}
+
 /* iterates over all memory definitions we find in the ACPI table */
 static int __init init_memory_definitions(struct acpi_table_header *table)
 {
@@ -2727,10 +2736,11 @@ static int __init init_memory_definitions(struct acpi_table_header *table)
 
 	while (p < end) {
 		m = (struct ivmd_header *)p;
+		p += m->length;
+		if (!ivrs_block_type_ivmd(m->type))
+			continue;
 		if (m->flags & (IVMD_FLAG_UNITY_MAP | IVMD_FLAG_EXCL_RANGE))
 			init_unity_map_range(m, table);
-
-		p += m->length;
 	}
 
 	return 0;
-- 
2.43.0


^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2026-10-18  7:46 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-10-18  7:45 [PATCH] iommu/amd: Skip non-IVMD entries when parsing IVRS memory definitions Akshay Jindal

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®