Fix a potential bug with stream masking on CUDA 12.6 on aarch64 Jetson

Commit 3f9bda39 made an error by using the pre-CUDA-12 mask
structure layout on CUDA 12.6 on aarch64 Jetson. Switch to the
CUDA 12+ layout (as used on x86_64).

Tests work either way on the Jetson Orin, so this change is not
strictly required, but seems advisable to support potenital large
(PCIe-attached?) GPUs on Jetson/DRIVE platforms.

1 files changed, 1 insertions, 1 deletions

diff --git a/libsmctrl.c b/libsmctrl.c
index 7202572..7debe3a 100644
--- a/libsmctrl.c
+++ b/libsmctrl.c
@@ -421,7 +421,7 @@ void libsmctrl_set_stream_mask_ext(void* stream, uint128_t mask) {
                 hw_mask = (struct stream_sm_mask*)(stream_struct_base + CU_11_4_MASK_OFF_JETSON);
                 break;
         case 12060:
-                hw_mask = (struct stream_sm_mask*)(stream_struct_base + CU_12_6_MASK_OFF_JETSON);
+                hw_mask_v2 = (void*)(stream_struct_base + CU_12_6_MASK_OFF_JETSON);
                 break;
 #endif
         }