static int vce_v2_0_soft_reset(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
WREG32_FIELD(SRBM_SOFT_RESET, SOFT_RESET_VCE, 1);
mdelay(5);
return vce_v2_0_start(adev);
}
/**
* vce_v2_0_start - start VCE block
*
* @adev: amdgpu_device pointer
*
* Setup and start the VCE block
*/
static int vce_v2_0_start(struct amdgpu_device *adev)
{
struct amdgpu_ring *ring;
int r;
vce_v2_0_mc_resume(adev);
/* set BUSY flag */
WREG32_P(mmVCE_STATUS, 1, ~1);
ring = &adev->vce.ring[0];
WREG32(mmVCE_RB_RPTR, ring->wptr);
WREG32(mmVCE_RB_WPTR, ring->wptr);
WREG32(mmVCE_RB_BASE_LO, ring->gpu_addr);
WREG32(mmVCE_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
WREG32(mmVCE_RB_SIZE, ring->ring_size / 4);
ring = &adev->vce.ring[1];
WREG32(mmVCE_RB_RPTR2, ring->wptr);
WREG32(mmVCE_RB_WPTR2, ring->wptr);
WREG32(mmVCE_RB_BASE_LO2, ring->gpu_addr);
WREG32(mmVCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
WREG32(mmVCE_RB_SIZE2, ring->ring_size / 4);
WREG32_FIELD(VCE_VCPU_CNTL, CLK_EN, 1);
WREG32_FIELD(VCE_SOFT_RESET, ECPU_SOFT_RESET, 1);
mdelay(100);
WREG32_FIELD(VCE_SOFT_RESET, ECPU_SOFT_RESET, 0);
r = vce_v2_0_firmware_loaded(adev);
/* clear BUSY flag */
WREG32_P(mmVCE_STATUS, 0, ~1);
if (r) {
DRM_ERROR("VCE not responding, giving up!!!\n");
return r;
}
return 0;
}
static void vce_v2_0_mc_resume(struct amdgpu_device *adev)
{
uint64_t addr = adev->vce.gpu_addr;
uint32_t size;
WREG32_P(mmVCE_CLOCK_GATING_A, 0, ~(1 << 16));
WREG32_P(mmVCE_UENC_CLOCK_GATING, 0x1FF000, ~0xFF9FF000);
WREG32_P(mmVCE_UENC_REG_CLOCK_GATING, 0x3F, ~0x3F);
WREG32(mmVCE_CLOCK_GATING_B, 0xf7);
WREG32(mmVCE_LMI_CTRL, 0x00398000);
WREG32_P(mmVCE_LMI_CACHE_CTRL, 0x0, ~0x1);
WREG32(mmVCE_LMI_SWAP_CNTL, 0);
WREG32(mmVCE_LMI_SWAP_CNTL1, 0);
WREG32(mmVCE_LMI_VM_CTRL, 0);
addr += AMDGPU_VCE_FIRMWARE_OFFSET;
size = VCE_V2_0_FW_SIZE;
WREG32(mmVCE_VCPU_CACHE_OFFSET0, addr & 0x7fffffff);
WREG32(mmVCE_VCPU_CACHE_SIZE0, size);
addr += size;
size = VCE_V2_0_STACK_SIZE;
WREG32(mmVCE_VCPU_CACHE_OFFSET1, addr & 0x7fffffff);
WREG32(mmVCE_VCPU_CACHE_SIZE1, size);
addr += size;
size = VCE_V2_0_DATA_SIZE;
WREG32(mmVCE_VCPU_CACHE_OFFSET2, addr & 0x7fffffff);
WREG32(mmVCE_VCPU_CACHE_SIZE2, size);
WREG32_P(mmVCE_LMI_CTRL2, 0x0, ~0x100);
WREG32_FIELD(VCE_SYS_INT_EN, VCE_SYS_INT_TRAP_INTERRUPT_EN, 1);
vce_v2_0_init_cg(adev);
}
/**
* uvd_v6_0_start - start UVD block
*
* @adev: amdgpu_device pointer
*
* Setup and start the UVD block
*/
static int uvd_v6_0_start(struct amdgpu_device *adev)
{
struct amdgpu_ring *ring = &adev->uvd.ring;
uint32_t rb_bufsz, tmp;
uint32_t lmi_swap_cntl;
uint32_t mp_swap_cntl;
int i, j, r;
/* disable DPG */
WREG32_P(mmUVD_POWER_STATUS, 0, ~UVD_POWER_STATUS__UVD_PG_MODE_MASK);
/* disable byte swapping */
lmi_swap_cntl = 0;
mp_swap_cntl = 0;
uvd_v6_0_mc_resume(adev);
/* disable clock gating */
WREG32_FIELD(UVD_CGC_CTRL, DYN_CLOCK_MODE, 0);
/* disable interupt */
WREG32_FIELD(UVD_MASTINT_EN, VCPU_EN, 0);
/* stall UMC and register bus before resetting VCPU */
WREG32_FIELD(UVD_LMI_CTRL2, STALL_ARB_UMC, 1);
mdelay(1);
/* put LMI, VCPU, RBC etc... into reset */
WREG32(mmUVD_SOFT_RESET,
UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK);
mdelay(5);
/* take UVD block out of reset */
WREG32_FIELD(SRBM_SOFT_RESET, SOFT_RESET_UVD, 0);
mdelay(5);
/* initialize UVD memory controller */
WREG32(mmUVD_LMI_CTRL,
(0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
UVD_LMI_CTRL__REQ_MODE_MASK |
UVD_LMI_CTRL__DISABLE_ON_FWV_FAIL_MASK);
#ifdef __BIG_ENDIAN
/* swap (8 in 32) RB and IB */
lmi_swap_cntl = 0xa;
mp_swap_cntl = 0;
#endif
WREG32(mmUVD_LMI_SWAP_CNTL, lmi_swap_cntl);
WREG32(mmUVD_MP_SWAP_CNTL, mp_swap_cntl);
WREG32(mmUVD_MPC_SET_MUXA0, 0x40c2040);
WREG32(mmUVD_MPC_SET_MUXA1, 0x0);
WREG32(mmUVD_MPC_SET_MUXB0, 0x40c2040);
WREG32(mmUVD_MPC_SET_MUXB1, 0x0);
WREG32(mmUVD_MPC_SET_ALU, 0);
WREG32(mmUVD_MPC_SET_MUX, 0x88);
/* take all subblocks out of reset, except VCPU */
WREG32(mmUVD_SOFT_RESET, UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
mdelay(5);
/* enable VCPU clock */
WREG32(mmUVD_VCPU_CNTL, UVD_VCPU_CNTL__CLK_EN_MASK);
/* enable UMC */
WREG32_FIELD(UVD_LMI_CTRL2, STALL_ARB_UMC, 0);
/* boot up the VCPU */
WREG32(mmUVD_SOFT_RESET, 0);
mdelay(10);
for (i = 0; i < 10; ++i) {
uint32_t status;
for (j = 0; j < 100; ++j) {
status = RREG32(mmUVD_STATUS);
if (status & 2)
break;
mdelay(10);
}
r = 0;
if (status & 2)
break;
DRM_ERROR("UVD not responding, trying to reset the VCPU!!!\n");
WREG32_FIELD(UVD_SOFT_RESET, VCPU_SOFT_RESET, 1);
mdelay(10);
WREG32_FIELD(UVD_SOFT_RESET, VCPU_SOFT_RESET, 0);
mdelay(10);
r = -1;
}
if (r) {
DRM_ERROR("UVD not responding, giving up!!!\n");
return r;
}
/* enable master interrupt */
WREG32_P(mmUVD_MASTINT_EN,
(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
/* clear the bit 4 of UVD_STATUS */
WREG32_P(mmUVD_STATUS, 0, ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT));
/* force RBC into idle state */
rb_bufsz = order_base_2(ring->ring_size);
tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, rb_bufsz);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BLKSZ, 1);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_WPTR_POLL_EN, 0);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1);
WREG32(mmUVD_RBC_RB_CNTL, tmp);
/* set the write pointer delay */
WREG32(mmUVD_RBC_RB_WPTR_CNTL, 0);
/* set the wb address */
WREG32(mmUVD_RBC_RB_RPTR_ADDR, (upper_32_bits(ring->gpu_addr) >> 2));
/* programm the RB_BASE for ring buffer */
WREG32(mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
lower_32_bits(ring->gpu_addr));
WREG32(mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
upper_32_bits(ring->gpu_addr));
/* Initialize the ring buffer's read and write pointers */
WREG32(mmUVD_RBC_RB_RPTR, 0);
ring->wptr = RREG32(mmUVD_RBC_RB_RPTR);
WREG32(mmUVD_RBC_RB_WPTR, ring->wptr);
WREG32_FIELD(UVD_RBC_RB_CNTL, RB_NO_FETCH, 0);
return 0;
}
static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd,
enum kfd_preempt_type reset_type,
unsigned int utimeout, uint32_t pipe_id,
uint32_t queue_id)
{
struct amdgpu_device *adev = get_amdgpu_device(kgd);
uint32_t temp;
enum hqd_dequeue_request_type type;
unsigned long flags, end_jiffies;
int retry;
struct vi_mqd *m = get_mqd(mqd);
if (adev->in_gpu_reset)
return -EIO;
acquire_queue(kgd, pipe_id, queue_id);
if (m->cp_hqd_vmid == 0)
WREG32_FIELD(RLC_CP_SCHEDULERS, scheduler1, 0);
switch (reset_type) {
case KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN:
type = DRAIN_PIPE;
break;
case KFD_PREEMPT_TYPE_WAVEFRONT_RESET:
type = RESET_WAVES;
break;
default:
type = DRAIN_PIPE;
break;
}
/* Workaround: If IQ timer is active and the wait time is close to or
* equal to 0, dequeueing is not safe. Wait until either the wait time
* is larger or timer is cleared. Also, ensure that IQ_REQ_PEND is
* cleared before continuing. Also, ensure wait times are set to at
* least 0x3.
*/
local_irq_save(flags);
preempt_disable();
retry = 5000; /* wait for 500 usecs at maximum */
while (true) {
temp = RREG32(mmCP_HQD_IQ_TIMER);
if (REG_GET_FIELD(temp, CP_HQD_IQ_TIMER, PROCESSING_IQ)) {
pr_debug("HW is processing IQ\n");
goto loop;
}
if (REG_GET_FIELD(temp, CP_HQD_IQ_TIMER, ACTIVE)) {
if (REG_GET_FIELD(temp, CP_HQD_IQ_TIMER, RETRY_TYPE)
== 3) /* SEM-rearm is safe */
break;
/* Wait time 3 is safe for CP, but our MMIO read/write
* time is close to 1 microsecond, so check for 10 to
* leave more buffer room
*/
if (REG_GET_FIELD(temp, CP_HQD_IQ_TIMER, WAIT_TIME)
>= 10)
break;
pr_debug("IQ timer is active\n");
} else
break;
loop:
if (!retry) {
pr_err("CP HQD IQ timer status time out\n");
break;
}
ndelay(100);
--retry;
}
retry = 1000;
while (true) {
temp = RREG32(mmCP_HQD_DEQUEUE_REQUEST);
if (!(temp & CP_HQD_DEQUEUE_REQUEST__IQ_REQ_PEND_MASK))
break;
pr_debug("Dequeue request is pending\n");
if (!retry) {
pr_err("CP HQD dequeue request time out\n");
break;
}
ndelay(100);
--retry;
}
local_irq_restore(flags);
preempt_enable();
WREG32(mmCP_HQD_DEQUEUE_REQUEST, type);
end_jiffies = (utimeout * HZ / 1000) + jiffies;
while (true) {
temp = RREG32(mmCP_HQD_ACTIVE);
if (!(temp & CP_HQD_ACTIVE__ACTIVE_MASK))
break;
if (time_after(jiffies, end_jiffies)) {
pr_err("cp queue preemption time out.\n");
release_queue(kgd);
return -ETIME;
}
usleep_range(500, 1000);
}
release_queue(kgd);
return 0;
}