Hi Mostafa, Eric,
On 2026/9/1 22:01, Mostafa Saleh wrote:
On Fri, Aug 14, 2026 at 12:24:43AM +0800, Tao Tang wrote:
Pass sec_sid into decode_ste() and treat a secure STE that enables stage-2
translation as an illegal STE.
The current model does not implement Secure stage-2 translation, so reject
Secure STE configurations that enable stage 2 and report BAD_STE. This
restriction can be relaxed when SEL2 and Secure stage-2 support are
implemented.
Signed-off-by: Tao Tang <[email protected]>
---
hw/arm/smmuv3.c | 12 ++++++++++--
1 file changed, 10 insertions(+), 2 deletions(-)
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index b5dc67dcb01..3ff2f498a3b 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -597,7 +597,7 @@ static void decode_ste_config(SMMUTransCfg *cfg, uint32_t
config)
/* Returns < 0 in case of invalid STE, 0 otherwise */
static int decode_ste(SMMUv3State *s, SMMUTransCfg *cfg,
- STE *ste, SMMUEventInfo *event)
+ STE *ste, SMMUEventInfo *event, SMMUSecSID sec_sid)
{
uint32_t config;
/* OAS field only presents on NS-IDR5 so we use hardcoded SMMU_SEC_SID_NS
*/
@@ -651,6 +651,14 @@ static int decode_ste(SMMUv3State *s, SMMUTransCfg *cfg,
if (ret) {
goto bad_ste;
}
+
+ /*
+ * It is ILLEGAL to set STE.Config == 0b11x according to (IHI 0070G.b)
+ * 5.2 STE, Stream Table Entry, Page 218.
+ */
This is not the latest spec, and there is no need to mention the page.
It might be more clear to mention that Qemu doesn’t support secure
stage-2 rather than say it is ILLEGAL.
Agreed. I'll change the comment to say that the model does not support
Secure stage 2 translation, and remove the spec version/page reference.
I'll also clarify SMMU_S_IDR1.SEL2 in the commit message.
@Eric, I've replied to your questions about the S2 plumbing and the
temporary SEL2 override in the threads for patches 6 [1] and 28 [2].
[1]
https://lore.kernel.org/qemu-devel/[email protected]/
[2]
https://lore.kernel.org/qemu-devel/[email protected]/
Best regards,
Tao
