From: Ben Goz ben.goz@amd.com
This commit is a bug fix for 32b hsa processes support
Signed-off-by: Ben Goz ben.goz@amd.com Signed-off-by: Oded Gabbay oded.gabbay@amd.com --- drivers/gpu/hsa/radeon/cik_regs.h | 1 + drivers/gpu/hsa/radeon/kfd_device_queue_manager.c | 8 +++++--- 2 files changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/hsa/radeon/cik_regs.h b/drivers/gpu/hsa/radeon/cik_regs.h index fa5ec01..a6404e3 100644 --- a/drivers/gpu/hsa/radeon/cik_regs.h +++ b/drivers/gpu/hsa/radeon/cik_regs.h @@ -45,6 +45,7 @@ /* if PTR32, this is the upper limit of GPUVM */ #define SH_MEM_CONFIG 0x8C34 #define PTR32 (1 << 0) +#define PRIVATE_ATC (1 << 1) #define ALIGNMENT_MODE(x) ((x) << 2) #define SH_MEM_ALIGNMENT_MODE_DWORD 0 #define SH_MEM_ALIGNMENT_MODE_DWORD_STRICT 1 diff --git a/drivers/gpu/hsa/radeon/kfd_device_queue_manager.c b/drivers/gpu/hsa/radeon/kfd_device_queue_manager.c index 01573b1..3e1def1 100644 --- a/drivers/gpu/hsa/radeon/kfd_device_queue_manager.c +++ b/drivers/gpu/hsa/radeon/kfd_device_queue_manager.c @@ -90,15 +90,17 @@ static void init_process_memory(struct device_queue_manager *dqm, struct qcm_pro if (qpd->pqm->process->is_32bit_user_mode) { temp = get_sh_mem_bases_32(qpd->pqm->process, dqm->dev); qpd->sh_mem_bases = SHARED_BASE(temp); + qpd->sh_mem_config = PTR32; } else { temp = get_sh_mem_bases_nybble_64(qpd->pqm->process, dqm->dev); qpd->sh_mem_bases = compute_sh_mem_bases_64bit(temp); + qpd->sh_mem_config = 0; }
- qpd->sh_mem_config = ALIGNMENT_MODE(SH_MEM_ALIGNMENT_MODE_UNALIGNED); + qpd->sh_mem_config |= ALIGNMENT_MODE(SH_MEM_ALIGNMENT_MODE_UNALIGNED); qpd->sh_mem_config |= DEFAULT_MTYPE(MTYPE_NONCACHED); qpd->sh_mem_ape1_limit = 0; - qpd->sh_mem_ape1_base = 1; + qpd->sh_mem_ape1_base = 0;
pr_debug("kfd: is32bit process: %d sh_mem_bases nybble: 0x%X and register 0x%X\n", qpd->pqm->process->is_32bit_user_mode, temp, qpd->sh_mem_bases); @@ -854,7 +856,7 @@ static int execute_queues_cpsch(struct device_queue_manager *dqm) }
if (dqm->queue_count <= 0 || dqm->processes_count <= 0) - return 0; + return 0;
mutex_lock(&dqm->lock); if (dqm->active_runlist) {