Add support of all the CP GFX queues for gfx12 ipdump
to be used by devcoredump.
Signed-off-by: Sunil Khatri <sunil.khatri@xxxxxxx>
---
drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c | 94 ++++++++++++++++++++++++++
1 file changed, 94 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
index 21b8167f39bd..cf7209b93617 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
@@ -173,6 +173,35 @@ static const struct amdgpu_hwip_reg_entry gc_cp_reg_list_12[] = {
SOC15_REG_ENTRY_STR(GC, 0, regCP_HQD_DEQUEUE_STATUS)
};
+static const struct amdgpu_hwip_reg_entry gc_gfx_queue_reg_list_12[] = {
+ /* gfx queue registers */
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_ACTIVE),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_VMID),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_QUEUE_PRIORITY),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_QUANTUM),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_BASE),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_BASE_HI),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_OFFSET),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_CNTL),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_CSMD_RPTR),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_WPTR),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_WPTR_HI),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_DEQUEUE_REQUEST),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_MAPPED),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_QUE_MGR_CONTROL),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_HQ_CONTROL0),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_HQD_HQ_STATUS0),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_MQD_BASE_ADDR),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_GFX_MQD_BASE_ADDR_HI),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_RB_WPTR_POLL_ADDR_LO),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_RB_WPTR_POLL_ADDR_HI),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_RB_RPTR),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_IB1_BASE_LO),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_IB1_BASE_HI),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_IB1_CMD_BUFSZ),
+ SOC15_REG_ENTRY_STR(GC, 0, regCP_IB1_BUFSZ)
+};
+
#define DEFAULT_SH_MEM_CONFIG \
((SH_MEM_ADDRESS_MODE_64 << SH_MEM_CONFIG__ADDRESS_MODE__SHIFT) | \
(SH_MEM_ALIGNMENT_MODE_UNALIGNED << SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT) | \
@@ -1265,6 +1294,19 @@ static void gfx_v12_0_alloc_ip_dump(struct amdgpu_device *adev)
} else {
adev->gfx.ip_dump_compute_queues = ptr;
}
+
+ /* Allocate memory for gfx queue registers for all the instances */
+ reg_count = ARRAY_SIZE(gc_gfx_queue_reg_list_12);
+ inst = adev->gfx.me.num_me * adev->gfx.me.num_pipe_per_me *
+ adev->gfx.me.num_queue_per_pipe;
+
+ ptr = kcalloc(reg_count * inst, sizeof(uint32_t), GFP_KERNEL);
+ if (ptr == NULL) {
+ DRM_ERROR("Failed to allocate memory for GFX Queues IP Dump\n");
+ adev->gfx.ip_dump_gfx_queues = NULL;
+ } else {
+ adev->gfx.ip_dump_gfx_queues = ptr;
+ }
}
static int gfx_v12_0_sw_init(void *handle)
@@ -1462,6 +1504,7 @@ static int gfx_v12_0_sw_fini(void *handle)
kfree(adev->gfx.ip_dump_core);
kfree(adev->gfx.ip_dump_compute_queues);
+ kfree(adev->gfx.ip_dump_gfx_queues);
return 0;
}
@@ -4853,6 +4896,31 @@ static void gfx_v12_ip_print(void *handle, struct drm_printer *p)
}
}
}
+
+ /* print gfx queue registers for all instances */
+ if (!adev->gfx.ip_dump_gfx_queues)
+ return;
+
+ index = 0;
+ reg_count = ARRAY_SIZE(gc_gfx_queue_reg_list_12);
+ drm_printf(p, "\nnum_me: %d num_pipe: %d num_queue: %d\n",
+ adev->gfx.me.num_me,
+ adev->gfx.me.num_pipe_per_me,
+ adev->gfx.me.num_queue_per_pipe);
+
+ for (i = 0; i < adev->gfx.me.num_me; i++) {
+ for (j = 0; j < adev->gfx.me.num_pipe_per_me; j++) {
+ for (k = 0; k < adev->gfx.me.num_queue_per_pipe; k++) {
+ drm_printf(p, "\nme %d, pipe %d, queue %d\n", i, j, k);
+ for (reg = 0; reg < reg_count; reg++) {
+ drm_printf(p, "%-50s \t 0x%08x\n",
+ gc_gfx_queue_reg_list_12[reg].reg_name,
+ adev->gfx.ip_dump_gfx_queues[index + reg]);
+ }
+ index += reg_count;
+ }
+ }
+ }
}
static void gfx_v12_ip_dump(void *handle)
@@ -4893,6 +4961,32 @@ static void gfx_v12_ip_dump(void *handle)
soc24_grbm_select(adev, 0, 0, 0, 0);
mutex_unlock(&adev->srbm_mutex);
amdgpu_gfx_off_ctrl(adev, true);
+
+ /* dump gfx queue registers for all instances */
+ if (!adev->gfx.ip_dump_gfx_queues)
+ return;
+
+ index = 0;
+ reg_count = ARRAY_SIZE(gc_gfx_queue_reg_list_12);
+ amdgpu_gfx_off_ctrl(adev, false);
+ mutex_lock(&adev->srbm_mutex);
+ for (i = 0; i < adev->gfx.me.num_me; i++) {
+ for (j = 0; j < adev->gfx.me.num_pipe_per_me; j++) {
+ for (k = 0; k < adev->gfx.me.num_queue_per_pipe; k++) {
+ soc24_grbm_select(adev, i, j, k, 0);
+
+ for (reg = 0; reg < reg_count; reg++) {
+ adev->gfx.ip_dump_gfx_queues[index + reg] =
+ RREG32(SOC15_REG_ENTRY_OFFSET(
+ gc_gfx_queue_reg_list_12[reg]));
+ }
+ index += reg_count;
+ }
+ }
+ }
+ soc24_grbm_select(adev, 0, 0, 0, 0);
+ mutex_unlock(&adev->srbm_mutex);
+ amdgpu_gfx_off_ctrl(adev, true);
}
static const struct amd_ip_funcs gfx_v12_0_ip_funcs = {
--
2.34.1
