drm/amdkfd: fix and enable ttmp setup for gfx11

The MES cached process context must be cleared on adding any queue for
the first time.

For proper debug support, the MES will clear it's cached process context
on the first call to SET_SHADER_DEBUGGER.

This allows TTMPs to be pesistently enabled in a safe manner.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Eric Huang <jinhuieric@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Jonathan Kim
2023-06-12 11:31:07 -04:00
committed by Alex Deucher
parent f9acfafc34
commit fc7f1d9697
6 changed files with 39 additions and 20 deletions

View File

@@ -38,6 +38,7 @@
#include "kfd_device_queue_manager.h"
#include "kfd_iommu.h"
#include "kfd_svm.h"
#include "kfd_debug.h"
#include "amdgpu_amdkfd.h"
#include "amdgpu_ras.h"
#include "amdgpu.h"
@@ -1931,6 +1932,9 @@ static void kfd_topology_set_capabilities(struct kfd_topology_device *dev)
HSA_CAP_TRAP_DEBUG_WAVE_LAUNCH_TRAP_OVERRIDE_SUPPORTED |
HSA_CAP_TRAP_DEBUG_WAVE_LAUNCH_MODE_SUPPORTED;
if (kfd_dbg_has_ttmps_always_setup(dev->gpu))
dev->node_props.debug_prop |= HSA_DBG_DISPATCH_INFO_ALWAYS_VALID;
if (KFD_GC_VERSION(dev->gpu) < IP_VERSION(10, 0, 0)) {
if (KFD_GC_VERSION(dev->gpu) == IP_VERSION(9, 4, 3))
dev->node_props.debug_prop |=
@@ -1941,10 +1945,6 @@ static void kfd_topology_set_capabilities(struct kfd_topology_device *dev)
HSA_DBG_WATCH_ADDR_MASK_LO_BIT_GFX9 |
HSA_DBG_WATCH_ADDR_MASK_HI_BIT;
if (KFD_GC_VERSION(dev->gpu) != IP_VERSION(9, 4, 2))
dev->node_props.debug_prop |=
HSA_DBG_DISPATCH_INFO_ALWAYS_VALID;
if (KFD_GC_VERSION(dev->gpu) >= IP_VERSION(9, 4, 2))
dev->node_props.capability |=
HSA_CAP_TRAP_DEBUG_PRECISE_MEMORY_OPERATIONS_SUPPORTED;
@@ -1952,9 +1952,7 @@ static void kfd_topology_set_capabilities(struct kfd_topology_device *dev)
dev->node_props.debug_prop |= HSA_DBG_WATCH_ADDR_MASK_LO_BIT_GFX10 |
HSA_DBG_WATCH_ADDR_MASK_HI_BIT;
if (KFD_GC_VERSION(dev->gpu) < IP_VERSION(11, 0, 0))
dev->node_props.debug_prop |= HSA_DBG_DISPATCH_INFO_ALWAYS_VALID;
else
if (KFD_GC_VERSION(dev->gpu) >= IP_VERSION(11, 0, 0))
dev->node_props.capability |=
HSA_CAP_TRAP_DEBUG_PRECISE_MEMORY_OPERATIONS_SUPPORTED;
}