From de331f42ab6efd7739c9d02594bc796421779a5f Mon Sep 17 00:00:00 2001 From: Joshua Bakita Date: Wed, 18 Dec 2024 15:37:27 -0500 Subject: 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. --- libsmctrl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'libsmctrl.c') 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 } -- cgit v1.2.2