add support for GPUs using wavefront size of 32 (#285)

* add gfx1100 support

Add support for Radeon 7900 GPUs (RX and PRO), and 7800 PRO.

I was contemplating to add gfx1101 and gfx1102 GPUs as well, but those are the lower end models that are more unlikely to be used for compute intensive jobs. In addition, I do not have access to them to test the support.

* update WF_SIZe for different options

Radeon systems use a WarpSize of 32, unlike current Instinct systems,
which use a warp size of 64. For the device side, a gfx specific ifdef
is sufficient. For the host side, we need to query the device
properties.

* adjust functional tests to wf_size of 32

* update unit tests to handle wf_size of 32

* address reviewer comments
This commit is contained in:
Edgar Gabriel
2025-10-22 16:04:58 -05:00
committed by GitHub
parent 955c22aeed
commit d0c2845031
19 changed files with 192 additions and 56 deletions
+17
View File
@@ -44,6 +44,21 @@ typedef struct device_agent {
std::vector<device_agent_t> gpu_agents;
std::vector<device_agent_t> cpu_agents;
std::vector<device_prop_t> device_properties;
static void device_properties_init(void) {
int numDevices;
CHECK_HIP(hipGetDeviceCount(&numDevices));
device_prop_t prop;
hipDeviceProp_t hipprop;
for (int i=0; i<numDevices; i++) {
CHECK_HIP(hipGetDeviceProperties(&hipprop, i));
prop.warpSize = hipprop.warpSize;
prop.maxThreadsPerBlock = hipprop.maxThreadsPerBlock;
device_properties.push_back(prop);
}
}
hsa_status_t rocm_hsa_amd_memory_pool_callback(
hsa_amd_memory_pool_t memory_pool, void* data) {
hsa_amd_memory_pool_global_flag_t pool_flag{};
@@ -108,6 +123,8 @@ int rocm_init() {
return 1;
}
device_properties_init();
return 0;
}