Merge drm/drm-next into drm-misc-next

Let's kickstart the v6.8 release cycle.

Signed-off-by: Maxime Ripard <mripard@kernel.org>

1 files changed, 84 insertions, 109 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
index 762d7a19f1be..0c6133cc5e57 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
@@ -60,6 +60,8 @@
 #define regCGTT_WD_CLK_CTRL_BASE_IDX	1
 #define regRLC_RLCS_BOOTLOAD_STATUS_gc_11_0_1	0x4e7e
 #define regRLC_RLCS_BOOTLOAD_STATUS_gc_11_0_1_BASE_IDX	1
+#define regPC_CONFIG_CNTL_1		0x194d
+#define regPC_CONFIG_CNTL_1_BASE_IDX	1
 
 MODULE_FIRMWARE("amdgpu/gc_11_0_0_pfp.bin");
 MODULE_FIRMWARE("amdgpu/gc_11_0_0_me.bin");
@@ -82,6 +84,10 @@ MODULE_FIRMWARE("amdgpu/gc_11_0_4_pfp.bin");
 MODULE_FIRMWARE("amdgpu/gc_11_0_4_me.bin");
 MODULE_FIRMWARE("amdgpu/gc_11_0_4_mec.bin");
 MODULE_FIRMWARE("amdgpu/gc_11_0_4_rlc.bin");
+MODULE_FIRMWARE("amdgpu/gc_11_5_0_pfp.bin");
+MODULE_FIRMWARE("amdgpu/gc_11_5_0_me.bin");
+MODULE_FIRMWARE("amdgpu/gc_11_5_0_mec.bin");
+MODULE_FIRMWARE("amdgpu/gc_11_5_0_rlc.bin");
 
 static const struct soc15_reg_golden golden_settings_gc_11_0_1[] =
 {
@@ -96,6 +102,23 @@ static const struct soc15_reg_golden golden_settings_gc_11_0_1[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, regTCP_CNTL2, 0xfcffffff, 0x0000000a)
 };
 
+static const struct soc15_reg_golden golden_settings_gc_11_5_0[] = {
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regDB_DEBUG5, 0xffffffff, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regGB_ADDR_CONFIG, 0x0c1807ff, 0x00000242),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regGCR_GENERAL_CNTL, 0x1ff1ffff, 0x00000500),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regGL2A_ADDR_MATCH_MASK, 0xffffffff, 0xfffffff3),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regGL2C_ADDR_MATCH_MASK, 0xffffffff, 0xfffffff3),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regGL2C_CTRL, 0xffffffff, 0xf37fff3f),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regGL2C_CTRL3, 0xfffffffb, 0x00f40188),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regGL2C_CTRL4, 0xf0ffffff, 0x8000b007),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regPA_CL_ENHANCE, 0xf1ffffff, 0x00880007),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regPC_CONFIG_CNTL_1, 0xffffffff, 0x00010000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regTA_CNTL_AUX, 0xf7f7ffff, 0x01030000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regTA_CNTL2, 0x007f0000, 0x00000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regTCP_CNTL2, 0xffcfffff, 0x0000200a),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, regUTCL1_CTRL_2, 0xffffffff, 0x0000048f)
+};
+
 #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) | \
@@ -132,6 +155,7 @@ static void gfx11_kiq_set_resources(struct amdgpu_ring *kiq_ring, uint64_t queue
 {
 	amdgpu_ring_write(kiq_ring, PACKET3(PACKET3_SET_RESOURCES, 6));
 	amdgpu_ring_write(kiq_ring, PACKET3_SET_RESOURCES_VMID_MASK(0) |
+			  PACKET3_SET_RESOURCES_UNMAP_LATENTY(0xa) | /* unmap_latency: 0xa (~ 1s) */
 			  PACKET3_SET_RESOURCES_QUEUE_TYPE(0));	/* vmid_mask:0 queue_type:0 (KIQ) */
 	amdgpu_ring_write(kiq_ring, lower_32_bits(queue_mask));	/* queue mask lo */
 	amdgpu_ring_write(kiq_ring, upper_32_bits(queue_mask));	/* queue mask hi */
@@ -265,13 +289,18 @@ static void gfx_v11_0_set_kiq_pm4_funcs(struct amdgpu_device *adev)
 
 static void gfx_v11_0_init_golden_registers(struct amdgpu_device *adev)
 {
-	switch (adev->ip_versions[GC_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(11, 0, 1):
 	case IP_VERSION(11, 0, 4):
 		soc15_program_register_sequence(adev,
 						golden_settings_gc_11_0_1,
 						(const u32)ARRAY_SIZE(golden_settings_gc_11_0_1));
 		break;
+	case IP_VERSION(11, 5, 0):
+		soc15_program_register_sequence(adev,
+						golden_settings_gc_11_5_0,
+						(const u32)ARRAY_SIZE(golden_settings_gc_11_5_0));
+		break;
 	default:
 		break;
 	}
@@ -465,7 +494,7 @@ out:
 
 static void gfx_v11_0_check_fw_cp_gfx_shadow(struct amdgpu_device *adev)
 {
-	switch (adev->ip_versions[GC_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(11, 0, 0):
 	case IP_VERSION(11, 0, 2):
 	case IP_VERSION(11, 0, 3):
@@ -561,6 +590,14 @@ static int gfx_v11_0_init_microcode(struct amdgpu_device *adev)
 	adev->gfx.mec2_fw = NULL;
 
 	gfx_v11_0_check_fw_cp_gfx_shadow(adev);
+
+	if (adev->gfx.imu.funcs && adev->gfx.imu.funcs->init_microcode) {
+		err = adev->gfx.imu.funcs->init_microcode(adev);
+		if (err)
+			DRM_ERROR("Failed to init imu firmware!\n");
+		return err;
+	}
+
 out:
 	if (err) {
 		amdgpu_ucode_release(&adev->gfx.pfp_fw);
@@ -856,8 +893,7 @@ static const struct amdgpu_gfx_funcs gfx_v11_0_gfx_funcs = {
 
 static int gfx_v11_0_gpu_early_init(struct amdgpu_device *adev)
 {
-
-	switch (adev->ip_versions[GC_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(11, 0, 0):
 	case IP_VERSION(11, 0, 2):
 		adev->gfx.config.max_hw_contexts = 8;
@@ -876,6 +912,7 @@ static int gfx_v11_0_gpu_early_init(struct amdgpu_device *adev)
 		break;
 	case IP_VERSION(11, 0, 1):
 	case IP_VERSION(11, 0, 4):
+	case IP_VERSION(11, 5, 0):
 		adev->gfx.config.max_hw_contexts = 8;
 		adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
 		adev->gfx.config.sc_prim_fifo_size_backend = 0x100;
@@ -1301,9 +1338,7 @@ static int gfx_v11_0_sw_init(void *handle)
 	struct amdgpu_kiq *kiq;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	adev->gfxhub.funcs->init(adev);
-
-	switch (adev->ip_versions[GC_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(11, 0, 0):
 	case IP_VERSION(11, 0, 2):
 	case IP_VERSION(11, 0, 3):
@@ -1316,6 +1351,7 @@ static int gfx_v11_0_sw_init(void *handle)
 		break;
 	case IP_VERSION(11, 0, 1):
 	case IP_VERSION(11, 0, 4):
+	case IP_VERSION(11, 5, 0):
 		adev->gfx.me.num_me = 1;
 		adev->gfx.me.num_pipe_per_me = 1;
 		adev->gfx.me.num_queue_per_pipe = 1;
@@ -1334,8 +1370,8 @@ static int gfx_v11_0_sw_init(void *handle)
 	}
 
 	/* Enable CG flag in one VF mode for enabling RLC safe mode enter/exit */
-	if (adev->ip_versions[GC_HWIP][0] == IP_VERSION(11, 0, 3) &&
-		amdgpu_sriov_is_pp_one_vf(adev))
+	if (amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(11, 0, 3) &&
+	    amdgpu_sriov_is_pp_one_vf(adev))
 		adev->cg_flags = AMD_CG_SUPPORT_GFX_CGCG;
 
 	/* EOP Event */
@@ -1368,14 +1404,6 @@ static int gfx_v11_0_sw_init(void *handle)
 
 	adev->gfx.gfx_current_status = AMDGPU_GFX_NORMAL_MODE;
 
-	if (adev->gfx.imu.funcs) {
-		if (adev->gfx.imu.funcs->init_microcode) {
-			r = adev->gfx.imu.funcs->init_microcode(adev);
-			if (r)
-				DRM_ERROR("Failed to load imu firmware!\n");
-		}
-	}
-
 	gfx_v11_0_me_init(adev);
 
 	r = gfx_v11_0_rlc_init(adev);
@@ -2562,8 +2590,11 @@ static int gfx_v11_0_wait_for_rlc_autoload_complete(struct amdgpu_device *adev)
 	for (i = 0; i < adev->usec_timeout; i++) {
 		cp_status = RREG32_SOC15(GC, 0, regCP_STAT);
 
-		if (adev->ip_versions[GC_HWIP][0] == IP_VERSION(11, 0, 1) ||
-				adev->ip_versions[GC_HWIP][0] == IP_VERSION(11, 0, 4))
+		if (amdgpu_ip_version(adev, GC_HWIP, 0) ==
+			    IP_VERSION(11, 0, 1) ||
+		    amdgpu_ip_version(adev, GC_HWIP, 0) ==
+			    IP_VERSION(11, 0, 4) ||
+		    amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(11, 5, 0))
 			bootload_status = RREG32_SOC15(GC, 0,
 					regRLC_RLCS_BOOTLOAD_STATUS_gc_11_0_1);
 		else
@@ -3684,11 +3715,11 @@ static int gfx_v11_0_gfx_init_queue(struct amdgpu_ring *ring)
 		soc21_grbm_select(adev, 0, 0, 0, 0);
 		mutex_unlock(&adev->srbm_mutex);
 		if (adev->gfx.me.mqd_backup[mqd_idx])
-			memcpy(adev->gfx.me.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
+			memcpy_fromio(adev->gfx.me.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
 	} else {
 		/* restore mqd with the backup copy */
 		if (adev->gfx.me.mqd_backup[mqd_idx])
-			memcpy(mqd, adev->gfx.me.mqd_backup[mqd_idx], sizeof(*mqd));
+			memcpy_toio(mqd, adev->gfx.me.mqd_backup[mqd_idx], sizeof(*mqd));
 		/* reset the ring */
 		ring->wptr = 0;
 		*ring->wptr_cpu_addr = 0;
@@ -3977,7 +4008,7 @@ static int gfx_v11_0_kiq_init_queue(struct amdgpu_ring *ring)
 	if (amdgpu_in_reset(adev)) { /* for GPU_RESET case */
 		/* reset MQD to a clean status */
 		if (adev->gfx.kiq[0].mqd_backup)
-			memcpy(mqd, adev->gfx.kiq[0].mqd_backup, sizeof(*mqd));
+			memcpy_toio(mqd, adev->gfx.kiq[0].mqd_backup, sizeof(*mqd));
 
 		/* reset ring buffer */
 		ring->wptr = 0;
@@ -4000,7 +4031,7 @@ static int gfx_v11_0_kiq_init_queue(struct amdgpu_ring *ring)
 		mutex_unlock(&adev->srbm_mutex);
 
 		if (adev->gfx.kiq[0].mqd_backup)
-			memcpy(adev->gfx.kiq[0].mqd_backup, mqd, sizeof(*mqd));
+			memcpy_fromio(adev->gfx.kiq[0].mqd_backup, mqd, sizeof(*mqd));
 	}
 
 	return 0;
@@ -4021,11 +4052,11 @@ static int gfx_v11_0_kcq_init_queue(struct amdgpu_ring *ring)
 		mutex_unlock(&adev->srbm_mutex);
 
 		if (adev->gfx.mec.mqd_backup[mqd_idx])
-			memcpy(adev->gfx.mec.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
+			memcpy_fromio(adev->gfx.mec.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
 	} else {
 		/* restore MQD to a clean status */
 		if (adev->gfx.mec.mqd_backup[mqd_idx])
-			memcpy(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(*mqd));
+			memcpy_toio(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(*mqd));
 		/* reset ring buffer */
 		ring->wptr = 0;
 		atomic64_set((atomic64_t *)ring->wptr_cpu_addr, 0);
@@ -4343,6 +4374,10 @@ static int gfx_v11_0_hw_init(void *handle)
 	if (r)
 		return r;
 
+	/* get IMU version from HW if it's not set */
+	if (!adev->gfx.imu_fw_version)
+		adev->gfx.imu_fw_version = RREG32_SOC15(GC, 0, regGFX_IMU_SCRATCH_0);
+
 	return r;
 }
 
@@ -4953,23 +4988,16 @@ static int gfx_v11_0_update_gfx_clock_gating(struct amdgpu_device *adev,
 
 static void gfx_v11_0_update_spm_vmid(struct amdgpu_device *adev, unsigned vmid)
 {
-	u32 reg, data;
+	u32 data;
 
 	amdgpu_gfx_off_ctrl(adev, false);
 
-	reg = SOC15_REG_OFFSET(GC, 0, regRLC_SPM_MC_CNTL);
-	if (amdgpu_sriov_is_pp_one_vf(adev))
-		data = RREG32_NO_KIQ(reg);
-	else
-		data = RREG32(reg);
+	data = RREG32_SOC15_NO_KIQ(GC, 0, regRLC_SPM_MC_CNTL);
 
 	data &= ~RLC_SPM_MC_CNTL__RLC_SPM_VMID_MASK;
 	data |= (vmid & RLC_SPM_MC_CNTL__RLC_SPM_VMID_MASK) << RLC_SPM_MC_CNTL__RLC_SPM_VMID__SHIFT;
 
-	if (amdgpu_sriov_is_pp_one_vf(adev))
-		WREG32_SOC15_NO_KIQ(GC, 0, regRLC_SPM_MC_CNTL, data);
-	else
-		WREG32_SOC15(GC, 0, regRLC_SPM_MC_CNTL, data);
+	WREG32_SOC15_NO_KIQ(GC, 0, regRLC_SPM_MC_CNTL, data);
 
 	amdgpu_gfx_off_ctrl(adev, true);
 }
@@ -5001,9 +5029,10 @@ static void gfx_v11_cntl_power_gating(struct amdgpu_device *adev, bool enable)
 
 	// Program RLC_PG_DELAY3 for CGPG hysteresis
 	if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)) {
-		switch (adev->ip_versions[GC_HWIP][0]) {
+		switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 		case IP_VERSION(11, 0, 1):
 		case IP_VERSION(11, 0, 4):
+		case IP_VERSION(11, 5, 0):
 			WREG32_SOC15(GC, 0, regRLC_PG_DELAY_3, RLC_PG_DELAY_3_DEFAULT_GC_11_0_1);
 			break;
 		default:
@@ -5030,7 +5059,7 @@ static int gfx_v11_0_set_powergating_state(void *handle,
 	if (amdgpu_sriov_vf(adev))
 		return 0;
 
-	switch (adev->ip_versions[GC_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(11, 0, 0):
 	case IP_VERSION(11, 0, 2):
 	case IP_VERSION(11, 0, 3):
@@ -5038,6 +5067,7 @@ static int gfx_v11_0_set_powergating_state(void *handle,
 		break;
 	case IP_VERSION(11, 0, 1):
 	case IP_VERSION(11, 0, 4):
+	case IP_VERSION(11, 5, 0):
 		if (!enable)
 			amdgpu_gfx_off_ctrl(adev, false);
 
@@ -5062,12 +5092,13 @@ static int gfx_v11_0_set_clockgating_state(void *handle,
 	if (amdgpu_sriov_vf(adev))
 	        return 0;
 
-	switch (adev->ip_versions[GC_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(11, 0, 0):
 	case IP_VERSION(11, 0, 1):
 	case IP_VERSION(11, 0, 2):
 	case IP_VERSION(11, 0, 3):
 	case IP_VERSION(11, 0, 4):
+	case IP_VERSION(11, 5, 0):
 	        gfx_v11_0_update_gfx_clock_gating(adev,
 	                        state ==  AMD_CG_STATE_GATE);
 	        break;
@@ -5144,45 +5175,17 @@ static u64 gfx_v11_0_ring_get_wptr_gfx(struct amdgpu_ring *ring)
 static void gfx_v11_0_ring_set_wptr_gfx(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
-	uint32_t *wptr_saved;
-	uint32_t *is_queue_unmap;
-	uint64_t aggregated_db_index;
-	uint32_t mqd_size = adev->mqds[AMDGPU_HW_IP_GFX].mqd_size;
-	uint64_t wptr_tmp;
 
-	if (ring->is_mes_queue) {
-		wptr_saved = (uint32_t *)(ring->mqd_ptr + mqd_size);
-		is_queue_unmap = (uint32_t *)(ring->mqd_ptr + mqd_size +
-					      sizeof(uint32_t));
-		aggregated_db_index =
-			amdgpu_mes_get_aggregated_doorbell_index(adev,
-								 ring->hw_prio);
-
-		wptr_tmp = ring->wptr & ring->buf_mask;
-		atomic64_set((atomic64_t *)ring->wptr_cpu_addr, wptr_tmp);
-		*wptr_saved = wptr_tmp;
-		/* assume doorbell always being used by mes mapped queue */
-		if (*is_queue_unmap) {
-			WDOORBELL64(aggregated_db_index, wptr_tmp);
-			WDOORBELL64(ring->doorbell_index, wptr_tmp);
-		} else {
-			WDOORBELL64(ring->doorbell_index, wptr_tmp);
-
-			if (*is_queue_unmap)
-				WDOORBELL64(aggregated_db_index, wptr_tmp);
-		}
+	if (ring->use_doorbell) {
+		/* XXX check if swapping is necessary on BE */
+		atomic64_set((atomic64_t *)ring->wptr_cpu_addr,
+			     ring->wptr);
+		WDOORBELL64(ring->doorbell_index, ring->wptr);
 	} else {
-		if (ring->use_doorbell) {
-			/* XXX check if swapping is necessary on BE */
-			atomic64_set((atomic64_t *)ring->wptr_cpu_addr,
-				     ring->wptr);
-			WDOORBELL64(ring->doorbell_index, ring->wptr);
-		} else {
-			WREG32_SOC15(GC, 0, regCP_RB0_WPTR,
-				     lower_32_bits(ring->wptr));
-			WREG32_SOC15(GC, 0, regCP_RB0_WPTR_HI,
-				     upper_32_bits(ring->wptr));
-		}
+		WREG32_SOC15(GC, 0, regCP_RB0_WPTR,
+			     lower_32_bits(ring->wptr));
+		WREG32_SOC15(GC, 0, regCP_RB0_WPTR_HI,
+			     upper_32_bits(ring->wptr));
 	}
 }
 
@@ -5207,42 +5210,14 @@ static u64 gfx_v11_0_ring_get_wptr_compute(struct amdgpu_ring *ring)
 static void gfx_v11_0_ring_set_wptr_compute(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
-	uint32_t *wptr_saved;
-	uint32_t *is_queue_unmap;
-	uint64_t aggregated_db_index;
-	uint32_t mqd_size = adev->mqds[AMDGPU_HW_IP_COMPUTE].mqd_size;
-	uint64_t wptr_tmp;
 
-	if (ring->is_mes_queue) {
-		wptr_saved = (uint32_t *)(ring->mqd_ptr + mqd_size);
-		is_queue_unmap = (uint32_t *)(ring->mqd_ptr + mqd_size +
-					      sizeof(uint32_t));
-		aggregated_db_index =
-			amdgpu_mes_get_aggregated_doorbell_index(adev,
-								 ring->hw_prio);
-
-		wptr_tmp = ring->wptr & ring->buf_mask;
-		atomic64_set((atomic64_t *)ring->wptr_cpu_addr, wptr_tmp);
-		*wptr_saved = wptr_tmp;
-		/* assume doorbell always used by mes mapped queue */
-		if (*is_queue_unmap) {
-			WDOORBELL64(aggregated_db_index, wptr_tmp);
-			WDOORBELL64(ring->doorbell_index, wptr_tmp);
-		} else {
-			WDOORBELL64(ring->doorbell_index, wptr_tmp);
-
-			if (*is_queue_unmap)
-				WDOORBELL64(aggregated_db_index, wptr_tmp);
-		}
+	/* XXX check if swapping is necessary on BE */
+	if (ring->use_doorbell) {
+		atomic64_set((atomic64_t *)ring->wptr_cpu_addr,
+			     ring->wptr);
+		WDOORBELL64(ring->doorbell_index, ring->wptr);
 	} else {
-		/* XXX check if swapping is necessary on BE */
-		if (ring->use_doorbell) {
-			atomic64_set((atomic64_t *)ring->wptr_cpu_addr,
-				     ring->wptr);
-			WDOORBELL64(ring->doorbell_index, ring->wptr);
-		} else {
-			BUG(); /* only DOORBELL method supported on gfx11 now */
-		}
+		BUG(); /* only DOORBELL method supported on gfx11 now */
 	}
 }