summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c')
-rw-r--r--drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c272
1 files changed, 133 insertions, 139 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
index 86a4865b1ae5..da0534ff1271 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
@@ -55,6 +55,9 @@
#define GFX7_NUM_GFX_RINGS 1
#define GFX7_MEC_HPD_SIZE 2048
+#define BONAIRE_GB_ADDR_CONFIG_GOLDEN 0x12010001
+#define HAWAII_GB_ADDR_CONFIG_GOLDEN 0x12011003
+
static void gfx_v7_0_set_ring_funcs(struct amdgpu_device *adev);
static void gfx_v7_0_set_irq_funcs(struct amdgpu_device *adev);
static void gfx_v7_0_set_gds_init(struct amdgpu_device *adev);
@@ -909,7 +912,6 @@ static void gfx_v7_0_free_microcode(struct amdgpu_device *adev)
static int gfx_v7_0_init_microcode(struct amdgpu_device *adev)
{
const char *chip_name;
- char fw_name[30];
int err;
DRM_DEBUG("\n");
@@ -934,40 +936,44 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev)
BUG();
}
- snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_pfp.bin", chip_name);
- err = amdgpu_ucode_request(adev, &adev->gfx.pfp_fw, fw_name);
+ err = amdgpu_ucode_request(adev, &adev->gfx.pfp_fw,
+ AMDGPU_UCODE_REQUIRED,
+ "amdgpu/%s_pfp.bin", chip_name);
if (err)
goto out;
- snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_me.bin", chip_name);
- err = amdgpu_ucode_request(adev, &adev->gfx.me_fw, fw_name);
+ err = amdgpu_ucode_request(adev, &adev->gfx.me_fw,
+ AMDGPU_UCODE_REQUIRED,
+ "amdgpu/%s_me.bin", chip_name);
if (err)
goto out;
- snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ce.bin", chip_name);
- err = amdgpu_ucode_request(adev, &adev->gfx.ce_fw, fw_name);
+ err = amdgpu_ucode_request(adev, &adev->gfx.ce_fw,
+ AMDGPU_UCODE_REQUIRED,
+ "amdgpu/%s_ce.bin", chip_name);
if (err)
goto out;
- snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec.bin", chip_name);
- err = amdgpu_ucode_request(adev, &adev->gfx.mec_fw, fw_name);
+ err = amdgpu_ucode_request(adev, &adev->gfx.mec_fw,
+ AMDGPU_UCODE_REQUIRED,
+ "amdgpu/%s_mec.bin", chip_name);
if (err)
goto out;
if (adev->asic_type == CHIP_KAVERI) {
- snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec2.bin", chip_name);
- err = amdgpu_ucode_request(adev, &adev->gfx.mec2_fw, fw_name);
+ err = amdgpu_ucode_request(adev, &adev->gfx.mec2_fw,
+ AMDGPU_UCODE_REQUIRED,
+ "amdgpu/%s_mec2.bin", chip_name);
if (err)
goto out;
}
- snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_rlc.bin", chip_name);
- err = amdgpu_ucode_request(adev, &adev->gfx.rlc_fw, fw_name);
- if (err)
- goto out;
+ err = amdgpu_ucode_request(adev, &adev->gfx.rlc_fw,
+ AMDGPU_UCODE_REQUIRED,
+ "amdgpu/%s_rlc.bin", chip_name);
out:
if (err) {
- pr_err("gfx7: Failed to load firmware \"%s\"\n", fw_name);
+ pr_err("gfx7: Failed to load firmware %s gfx firmware\n", chip_name);
gfx_v7_0_free_microcode(adev);
}
return err;
@@ -2117,6 +2123,8 @@ static void gfx_v7_0_ring_emit_fence_gfx(struct amdgpu_ring *ring, u64 addr,
{
bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT;
bool int_sel = flags & AMDGPU_FENCE_FLAG_INT;
+ bool exec = flags & AMDGPU_FENCE_FLAG_EXEC;
+
/* Workaround for cache flush problems. First send a dummy EOP
* event down the pipe with seq one below.
*/
@@ -2136,7 +2144,8 @@ static void gfx_v7_0_ring_emit_fence_gfx(struct amdgpu_ring *ring, u64 addr,
amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN |
EOP_TC_ACTION_EN |
EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) |
- EVENT_INDEX(5)));
+ EVENT_INDEX(5) |
+ (exec ? EOP_EXEC : 0)));
amdgpu_ring_write(ring, addr & 0xfffffffc);
amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xffff) |
DATA_SEL(write64bit ? 2 : 1) | INT_SEL(int_sel ? 2 : 0));
@@ -2324,7 +2333,7 @@ static int gfx_v7_0_ring_test_ib(struct amdgpu_ring *ring, long timeout)
r = -EINVAL;
error:
- amdgpu_ib_free(adev, &ib, NULL);
+ amdgpu_ib_free(&ib, NULL);
dma_fence_put(f);
return r;
}
@@ -2559,7 +2568,7 @@ static int gfx_v7_0_cp_gfx_resume(struct amdgpu_device *adev)
ring->wptr = 0;
WREG32(mmCP_RB0_WPTR, lower_32_bits(ring->wptr));
- /* set the wb address wether it's enabled or not */
+ /* set the wb address whether it's enabled or not */
rptr_addr = ring->rptr_gpu_addr;
WREG32(mmCP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr));
WREG32(mmCP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF);
@@ -2757,44 +2766,6 @@ static int gfx_v7_0_mec_init(struct amdgpu_device *adev)
return 0;
}
-struct hqd_registers {
- u32 cp_mqd_base_addr;
- u32 cp_mqd_base_addr_hi;
- u32 cp_hqd_active;
- u32 cp_hqd_vmid;
- u32 cp_hqd_persistent_state;
- u32 cp_hqd_pipe_priority;
- u32 cp_hqd_queue_priority;
- u32 cp_hqd_quantum;
- u32 cp_hqd_pq_base;
- u32 cp_hqd_pq_base_hi;
- u32 cp_hqd_pq_rptr;
- u32 cp_hqd_pq_rptr_report_addr;
- u32 cp_hqd_pq_rptr_report_addr_hi;
- u32 cp_hqd_pq_wptr_poll_addr;
- u32 cp_hqd_pq_wptr_poll_addr_hi;
- u32 cp_hqd_pq_doorbell_control;
- u32 cp_hqd_pq_wptr;
- u32 cp_hqd_pq_control;
- u32 cp_hqd_ib_base_addr;
- u32 cp_hqd_ib_base_addr_hi;
- u32 cp_hqd_ib_rptr;
- u32 cp_hqd_ib_control;
- u32 cp_hqd_iq_timer;
- u32 cp_hqd_iq_rptr;
- u32 cp_hqd_dequeue_request;
- u32 cp_hqd_dma_offload;
- u32 cp_hqd_sema_cmd;
- u32 cp_hqd_msg_type;
- u32 cp_hqd_atomic0_preop_lo;
- u32 cp_hqd_atomic0_preop_hi;
- u32 cp_hqd_atomic1_preop_lo;
- u32 cp_hqd_atomic1_preop_hi;
- u32 cp_hqd_hq_scheduler0;
- u32 cp_hqd_hq_scheduler1;
- u32 cp_mqd_control;
-};
-
static void gfx_v7_0_compute_pipe_init(struct amdgpu_device *adev,
int mec, int pipe)
{
@@ -2914,7 +2885,7 @@ static void gfx_v7_0_mqd_init(struct amdgpu_device *adev,
mqd->cp_hqd_pq_wptr_poll_addr_lo = wb_gpu_addr & 0xfffffffc;
mqd->cp_hqd_pq_wptr_poll_addr_hi = upper_32_bits(wb_gpu_addr) & 0xffff;
- /* set the wb address wether it's enabled or not */
+ /* set the wb address whether it's enabled or not */
wb_gpu_addr = ring->rptr_gpu_addr;
mqd->cp_hqd_pq_rptr_report_addr_lo = wb_gpu_addr & 0xfffffffc;
mqd->cp_hqd_pq_rptr_report_addr_hi =
@@ -3914,67 +3885,22 @@ static u32 gfx_v7_0_get_csb_size(struct amdgpu_device *adev)
static void gfx_v7_0_get_csb_buffer(struct amdgpu_device *adev,
volatile u32 *buffer)
{
- u32 count = 0, i;
- const struct cs_section_def *sect = NULL;
- const struct cs_extent_def *ext = NULL;
+ u32 count = 0;
if (adev->gfx.rlc.cs_data == NULL)
return;
if (buffer == NULL)
return;
- buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0));
- buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_BEGIN_CLEAR_STATE);
-
- buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CONTEXT_CONTROL, 1));
- buffer[count++] = cpu_to_le32(0x80000000);
- buffer[count++] = cpu_to_le32(0x80000000);
-
- for (sect = adev->gfx.rlc.cs_data; sect->section != NULL; ++sect) {
- for (ext = sect->section; ext->extent != NULL; ++ext) {
- if (sect->id == SECT_CONTEXT) {
- buffer[count++] =
- cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count));
- buffer[count++] = cpu_to_le32(ext->reg_index - PACKET3_SET_CONTEXT_REG_START);
- for (i = 0; i < ext->reg_count; i++)
- buffer[count++] = cpu_to_le32(ext->extent[i]);
- } else {
- return;
- }
- }
- }
+ count = amdgpu_gfx_csb_preamble_start(buffer);
+ count = amdgpu_gfx_csb_data_parser(adev, buffer, count);
buffer[count++] = cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, 2));
buffer[count++] = cpu_to_le32(mmPA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START);
- switch (adev->asic_type) {
- case CHIP_BONAIRE:
- buffer[count++] = cpu_to_le32(0x16000012);
- buffer[count++] = cpu_to_le32(0x00000000);
- break;
- case CHIP_KAVERI:
- buffer[count++] = cpu_to_le32(0x00000000); /* XXX */
- buffer[count++] = cpu_to_le32(0x00000000);
- break;
- case CHIP_KABINI:
- case CHIP_MULLINS:
- buffer[count++] = cpu_to_le32(0x00000000); /* XXX */
- buffer[count++] = cpu_to_le32(0x00000000);
- break;
- case CHIP_HAWAII:
- buffer[count++] = cpu_to_le32(0x3a00161a);
- buffer[count++] = cpu_to_le32(0x0000002e);
- break;
- default:
- buffer[count++] = cpu_to_le32(0x00000000);
- buffer[count++] = cpu_to_le32(0x00000000);
- break;
- }
+ buffer[count++] = cpu_to_le32(adev->gfx.config.rb_config[0][0].raster_config);
+ buffer[count++] = cpu_to_le32(adev->gfx.config.rb_config[0][0].raster_config_1);
- buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0));
- buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_END_CLEAR_STATE);
-
- buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CLEAR_STATE, 0));
- buffer[count++] = cpu_to_le32(0);
+ amdgpu_gfx_csb_preamble_end(buffer, count);
}
static void gfx_v7_0_init_pg(struct amdgpu_device *adev)
@@ -4172,9 +4098,9 @@ static const struct amdgpu_rlc_funcs gfx_v7_0_rlc_funcs = {
.update_spm_vmid = gfx_v7_0_update_spm_vmid
};
-static int gfx_v7_0_early_init(void *handle)
+static int gfx_v7_0_early_init(struct amdgpu_ip_block *ip_block)
{
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
adev->gfx.xcc_mask = 1;
adev->gfx.num_gfx_rings = GFX7_NUM_GFX_RINGS;
@@ -4189,9 +4115,9 @@ static int gfx_v7_0_early_init(void *handle)
return 0;
}
-static int gfx_v7_0_late_init(void *handle)
+static int gfx_v7_0_late_init(struct amdgpu_ip_block *ip_block)
{
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
int r;
r = amdgpu_irq_get(adev, &adev->gfx.priv_reg_irq, 0);
@@ -4381,10 +4307,10 @@ static int gfx_v7_0_compute_ring_init(struct amdgpu_device *adev, int ring_id,
return 0;
}
-static int gfx_v7_0_sw_init(void *handle)
+static int gfx_v7_0_sw_init(struct amdgpu_ip_block *ip_block)
{
struct amdgpu_ring *ring;
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
int i, j, k, r, ring_id;
switch (adev->asic_type) {
@@ -4477,9 +4403,9 @@ static int gfx_v7_0_sw_init(void *handle)
return r;
}
-static int gfx_v7_0_sw_fini(void *handle)
+static int gfx_v7_0_sw_fini(struct amdgpu_ip_block *ip_block)
{
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
int i;
for (i = 0; i < adev->gfx.num_gfx_rings; i++)
@@ -4503,10 +4429,10 @@ static int gfx_v7_0_sw_fini(void *handle)
return 0;
}
-static int gfx_v7_0_hw_init(void *handle)
+static int gfx_v7_0_hw_init(struct amdgpu_ip_block *ip_block)
{
int r;
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
gfx_v7_0_constants_init(adev);
@@ -4524,9 +4450,9 @@ static int gfx_v7_0_hw_init(void *handle)
return r;
}
-static int gfx_v7_0_hw_fini(void *handle)
+static int gfx_v7_0_hw_fini(struct amdgpu_ip_block *ip_block)
{
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0);
@@ -4537,23 +4463,19 @@ static int gfx_v7_0_hw_fini(void *handle)
return 0;
}
-static int gfx_v7_0_suspend(void *handle)
+static int gfx_v7_0_suspend(struct amdgpu_ip_block *ip_block)
{
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
- return gfx_v7_0_hw_fini(adev);
+ return gfx_v7_0_hw_fini(ip_block);
}
-static int gfx_v7_0_resume(void *handle)
+static int gfx_v7_0_resume(struct amdgpu_ip_block *ip_block)
{
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
- return gfx_v7_0_hw_init(adev);
+ return gfx_v7_0_hw_init(ip_block);
}
-static bool gfx_v7_0_is_idle(void *handle)
+static bool gfx_v7_0_is_idle(struct amdgpu_ip_block *ip_block)
{
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
if (RREG32(mmGRBM_STATUS) & GRBM_STATUS__GUI_ACTIVE_MASK)
return false;
@@ -4561,11 +4483,11 @@ static bool gfx_v7_0_is_idle(void *handle)
return true;
}
-static int gfx_v7_0_wait_for_idle(void *handle)
+static int gfx_v7_0_wait_for_idle(struct amdgpu_ip_block *ip_block)
{
unsigned i;
u32 tmp;
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
for (i = 0; i < adev->usec_timeout; i++) {
/* read MC_STATUS */
@@ -4578,11 +4500,11 @@ static int gfx_v7_0_wait_for_idle(void *handle)
return -ETIMEDOUT;
}
-static int gfx_v7_0_soft_reset(void *handle)
+static int gfx_v7_0_soft_reset(struct amdgpu_ip_block *ip_block)
{
u32 grbm_soft_reset = 0, srbm_soft_reset = 0;
u32 tmp;
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
/* GRBM_STATUS */
tmp = RREG32(mmGRBM_STATUS);
@@ -4888,11 +4810,11 @@ static int gfx_v7_0_priv_inst_irq(struct amdgpu_device *adev,
return 0;
}
-static int gfx_v7_0_set_clockgating_state(void *handle,
+static int gfx_v7_0_set_clockgating_state(struct amdgpu_ip_block *ip_block,
enum amd_clockgating_state state)
{
bool gate = false;
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
if (state == AMD_CG_STATE_GATE)
gate = true;
@@ -4911,11 +4833,11 @@ static int gfx_v7_0_set_clockgating_state(void *handle,
return 0;
}
-static int gfx_v7_0_set_powergating_state(void *handle,
+static int gfx_v7_0_set_powergating_state(struct amdgpu_ip_block *ip_block,
enum amd_powergating_state state)
{
bool gate = false;
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_device *adev = ip_block->adev;
if (state == AMD_PG_STATE_GATE)
gate = true;
@@ -4962,6 +4884,76 @@ static void gfx_v7_0_emit_mem_sync_compute(struct amdgpu_ring *ring)
amdgpu_ring_write(ring, 0x0000000A); /* poll interval */
}
+static void gfx_v7_0_wait_reg_mem(struct amdgpu_ring *ring, int eng_sel,
+ int mem_space, int opt, uint32_t addr0,
+ uint32_t addr1, uint32_t ref, uint32_t mask,
+ uint32_t inv)
+{
+ amdgpu_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5));
+ amdgpu_ring_write(ring,
+ /* memory (1) or register (0) */
+ (WAIT_REG_MEM_MEM_SPACE(mem_space) |
+ WAIT_REG_MEM_OPERATION(opt) | /* wait */
+ WAIT_REG_MEM_FUNCTION(3) | /* equal */
+ WAIT_REG_MEM_ENGINE(eng_sel)));
+
+ if (mem_space)
+ BUG_ON(addr0 & 0x3); /* Dword align */
+ amdgpu_ring_write(ring, addr0);
+ amdgpu_ring_write(ring, addr1);
+ amdgpu_ring_write(ring, ref);
+ amdgpu_ring_write(ring, mask);
+ amdgpu_ring_write(ring, inv); /* poll interval */
+}
+
+static void gfx_v7_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
+ uint32_t val, uint32_t mask)
+{
+ gfx_v7_0_wait_reg_mem(ring, 0, 0, 0, reg, 0, val, mask, 0x20);
+}
+
+static int gfx_v7_0_reset_kgq(struct amdgpu_ring *ring, unsigned int vmid)
+{
+ struct amdgpu_device *adev = ring->adev;
+ struct amdgpu_kiq *kiq = &adev->gfx.kiq[0];
+ struct amdgpu_ring *kiq_ring = &kiq->ring;
+ unsigned long flags;
+ u32 tmp;
+ int r;
+
+ if (amdgpu_sriov_vf(adev))
+ return -EINVAL;
+
+ if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues)
+ return -EINVAL;
+
+ spin_lock_irqsave(&kiq->ring_lock, flags);
+
+ if (amdgpu_ring_alloc(kiq_ring, 5)) {
+ spin_unlock_irqrestore(&kiq->ring_lock, flags);
+ return -ENOMEM;
+ }
+
+ tmp = REG_SET_FIELD(0, CP_VMID_RESET, RESET_REQUEST, 1 << vmid);
+ gfx_v7_0_ring_emit_wreg(kiq_ring, mmCP_VMID_RESET, tmp);
+ amdgpu_ring_commit(kiq_ring);
+
+ spin_unlock_irqrestore(&kiq->ring_lock, flags);
+
+ r = amdgpu_ring_test_ring(kiq_ring);
+ if (r)
+ return r;
+
+ if (amdgpu_ring_alloc(ring, 7 + 12 + 5))
+ return -ENOMEM;
+ gfx_v7_0_ring_emit_fence_gfx(ring, ring->fence_drv.gpu_addr,
+ ring->fence_drv.sync_seq, AMDGPU_FENCE_FLAG_EXEC);
+ gfx_v7_0_ring_emit_reg_wait(ring, mmCP_VMID_RESET, 0, 0xffff);
+ gfx_v7_0_ring_emit_wreg(ring, mmCP_VMID_RESET, 0);
+
+ return amdgpu_ring_test_ring(ring);
+}
+
static const struct amd_ip_funcs gfx_v7_0_ip_funcs = {
.name = "gfx_v7_0",
.early_init = gfx_v7_0_early_init,
@@ -5011,6 +5003,7 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = {
.emit_wreg = gfx_v7_0_ring_emit_wreg,
.soft_recovery = gfx_v7_0_ring_soft_recovery,
.emit_mem_sync = gfx_v7_0_emit_mem_sync,
+ .reset = gfx_v7_0_reset_kgq,
};
static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = {
@@ -5041,6 +5034,7 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = {
.insert_nop = amdgpu_ring_insert_nop,
.pad_ib = amdgpu_ring_generic_pad_ib,
.emit_wreg = gfx_v7_0_ring_emit_wreg,
+ .soft_recovery = gfx_v7_0_ring_soft_recovery,
.emit_mem_sync = gfx_v7_0_emit_mem_sync_compute,
};
generated by cgit (git 2.34.1) at 2025-07-23 10:46:46 +0000