Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-12-13 22:53:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

drm/amdgpu: move JPEG2.5 out from VCN2.5

Browse Source 
And clean up the duplicated stuff

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Leo Liu 2019-11-11 09:56:32 -05:00 committed by Alex Deucher
parent 5be45a26c9
commit 14f43e8f88
9 changed files with 679 additions and 602 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
drivers/gpu/drm/amd/amdgpu/Makefile
View File

@ -155,7 +155,8 @@ amdgpu-y += \
vcn_v2_5.o \
amdgpu_jpeg.o \
jpeg_v1_0.o \
jpeg_v2_0.o
jpeg_v2_0.o \
jpeg_v2_5.o
# add ATHUB block
amdgpu-y += \

3
drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h
View File

@ -26,6 +26,9 @@
#define AMDGPU_MAX_JPEG_INSTANCES 2
#define AMDGPU_JPEG_HARVEST_JPEG0 (1 << 0)
#define AMDGPU_JPEG_HARVEST_JPEG1 (1 << 1)
struct amdgpu_jpeg_reg{
unsigned jpeg_pitch;
};

105
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
View File

@ -706,108 +706,3 @@ error:
amdgpu_bo_unref(&bo);
return r;
}
int amdgpu_vcn_jpeg_ring_test_ring(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
uint32_t tmp = 0;
unsigned i;
int r;
WREG32(adev->vcn.inst[ring->me].external.jpeg_pitch, 0xCAFEDEAD);
r = amdgpu_ring_alloc(ring, 3);
if (r)
return r;
amdgpu_ring_write(ring, PACKET0(adev->vcn.internal.jpeg_pitch, 0));
amdgpu_ring_write(ring, 0xDEADBEEF);
amdgpu_ring_commit(ring);
for (i = 0; i < adev->usec_timeout; i++) {
tmp = RREG32(adev->vcn.inst[ring->me].external.jpeg_pitch);
if (tmp == 0xDEADBEEF)
break;
udelay(1);
}
if (i >= adev->usec_timeout)
r = -ETIMEDOUT;
return r;
}
static int amdgpu_vcn_jpeg_set_reg(struct amdgpu_ring *ring, uint32_t handle,
struct dma_fence **fence)
{
struct amdgpu_device *adev = ring->adev;
struct amdgpu_job *job;
struct amdgpu_ib *ib;
struct dma_fence *f = NULL;
const unsigned ib_size_dw = 16;
int i, r;
r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
if (r)
return r;
ib = &job->ibs[0];
ib->ptr[0] = PACKETJ(adev->vcn.internal.jpeg_pitch, 0, 0, PACKETJ_TYPE0);
ib->ptr[1] = 0xDEADBEEF;
for (i = 2; i < 16; i += 2) {
ib->ptr[i] = PACKETJ(0, 0, 0, PACKETJ_TYPE6);
ib->ptr[i+1] = 0;
}
ib->length_dw = 16;
r = amdgpu_job_submit_direct(job, ring, &f);
if (r)
goto err;
if (fence)
*fence = dma_fence_get(f);
dma_fence_put(f);
return 0;
err:
amdgpu_job_free(job);
return r;
}
int amdgpu_vcn_jpeg_ring_test_ib(struct amdgpu_ring *ring, long timeout)
{
struct amdgpu_device *adev = ring->adev;
uint32_t tmp = 0;
unsigned i;
struct dma_fence *fence = NULL;
long r = 0;
r = amdgpu_vcn_jpeg_set_reg(ring, 1, &fence);
if (r)
goto error;
r = dma_fence_wait_timeout(fence, false, timeout);
if (r == 0) {
r = -ETIMEDOUT;
goto error;
} else if (r < 0) {
goto error;
} else {
r = 0;
}
for (i = 0; i < adev->usec_timeout; i++) {
tmp = RREG32(adev->vcn.inst[ring->me].external.jpeg_pitch);
if (tmp == 0xDEADBEEF)
break;
udelay(1);
}
if (i >= adev->usec_timeout)
r = -ETIMEDOUT;
dma_fence_put(fence);
error:
return r;
}

5
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
View File

@ -158,7 +158,6 @@ struct amdgpu_vcn_reg{
unsigned ib_size;
unsigned gp_scratch8;
unsigned scratch9;
unsigned jpeg_pitch;
};
struct amdgpu_vcn_inst {
@ -168,7 +167,6 @@ struct amdgpu_vcn_inst {
void *saved_bo;
struct amdgpu_ring ring_dec;
struct amdgpu_ring ring_enc[AMDGPU_VCN_MAX_ENC_RINGS];
struct amdgpu_ring ring_jpeg;
struct amdgpu_irq_src irq;
struct amdgpu_vcn_reg external;
};
@ -209,7 +207,4 @@ int amdgpu_vcn_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout);
int amdgpu_vcn_enc_ring_test_ring(struct amdgpu_ring *ring);
int amdgpu_vcn_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout);
int amdgpu_vcn_jpeg_ring_test_ring(struct amdgpu_ring *ring);
int amdgpu_vcn_jpeg_ring_test_ib(struct amdgpu_ring *ring, long timeout);
#endif

641
drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c Normal file
View File

@ -0,0 +1,641 @@
/*
* Copyright 2019 Advanced Micro Devices, Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*
*/
#include "amdgpu.h"
#include "amdgpu_jpeg.h"
#include "soc15.h"
#include "soc15d.h"
#include "jpeg_v2_0.h"
#include "vcn/vcn_2_5_offset.h"
#include "vcn/vcn_2_5_sh_mask.h"
#include "ivsrcid/vcn/irqsrcs_vcn_2_0.h"
#define mmUVD_JPEG_PITCH_INTERNAL_OFFSET 0x401f
#define JPEG25_MAX_HW_INSTANCES_ARCTURUS 2
static void jpeg_v2_5_set_dec_ring_funcs(struct amdgpu_device *adev);
static void jpeg_v2_5_set_irq_funcs(struct amdgpu_device *adev);
static int jpeg_v2_5_set_powergating_state(void *handle,
enum amd_powergating_state state);
static int amdgpu_ih_clientid_jpeg[] = {
SOC15_IH_CLIENTID_VCN,
SOC15_IH_CLIENTID_VCN1
};
/**
* jpeg_v2_5_early_init - set function pointers
*
* @handle: amdgpu_device pointer
*
* Set ring and irq function pointers
*/
static int jpeg_v2_5_early_init(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
if (adev->asic_type == CHIP_ARCTURUS) {
u32 harvest;
int i;
adev->jpeg.num_jpeg_inst = JPEG25_MAX_HW_INSTANCES_ARCTURUS;
for (i = 0; i < adev->jpeg.num_jpeg_inst; i++) {
harvest = RREG32_SOC15(JPEG, i, mmCC_UVD_HARVESTING);
if (harvest & CC_UVD_HARVESTING__UVD_DISABLE_MASK)
adev->jpeg.harvest_config |= 1 << i;
}
if (adev->jpeg.harvest_config == (AMDGPU_JPEG_HARVEST_JPEG0 |
AMDGPU_JPEG_HARVEST_JPEG1))
return -ENOENT;
} else
adev->jpeg.num_jpeg_inst = 1;
jpeg_v2_5_set_dec_ring_funcs(adev);
jpeg_v2_5_set_irq_funcs(adev);
return 0;
}
/**
* jpeg_v2_5_sw_init - sw init for JPEG block
*
* @handle: amdgpu_device pointer
*
* Load firmware and sw initialization
*/
static int jpeg_v2_5_sw_init(void *handle)
{
struct amdgpu_ring *ring;
int i, r;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
/* JPEG TRAP */
r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_jpeg[i],
VCN_2_0__SRCID__JPEG_DECODE, &adev->jpeg.inst[i].irq);
if (r)
return r;
}
r = amdgpu_jpeg_sw_init(adev);
if (r)
return r;
r = amdgpu_jpeg_resume(adev);
if (r)
return r;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
ring = &adev->jpeg.inst[i].ring_dec;
ring->use_doorbell = true;
ring->doorbell_index = (adev->doorbell_index.vcn.vcn_ring0_1 << 1) + 1 + 8 * i;
sprintf(ring->name, "jpeg_dec_%d", i);
r = amdgpu_ring_init(adev, ring, 512, &adev->jpeg.inst[i].irq, 0);
if (r)
return r;
adev->jpeg.internal.jpeg_pitch = mmUVD_JPEG_PITCH_INTERNAL_OFFSET;
adev->jpeg.inst[i].external.jpeg_pitch = SOC15_REG_OFFSET(JPEG, i, mmUVD_JPEG_PITCH);
}
return 0;
}
/**
* jpeg_v2_5_sw_fini - sw fini for JPEG block
*
* @handle: amdgpu_device pointer
*
* JPEG suspend and free up sw allocation
*/
static int jpeg_v2_5_sw_fini(void *handle)
{
int r;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
r = amdgpu_jpeg_suspend(adev);
if (r)
return r;
r = amdgpu_jpeg_sw_fini(adev);
return r;
}
/**
* jpeg_v2_5_hw_init - start and test JPEG block
*
* @handle: amdgpu_device pointer
*
*/
static int jpeg_v2_5_hw_init(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
struct amdgpu_ring *ring;
int i, r;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
ring = &adev->jpeg.inst[i].ring_dec;
adev->nbio.funcs->vcn_doorbell_range(adev, ring->use_doorbell,
(adev->doorbell_index.vcn.vcn_ring0_1 << 1) + 8 * i, i);
r = amdgpu_ring_test_helper(ring);
if (r)
return r;
}
DRM_INFO("JPEG decode initialized successfully.\n");
return 0;
}
/**
* jpeg_v2_5_hw_fini - stop the hardware block
*
* @handle: amdgpu_device pointer
*
* Stop the JPEG block, mark ring as not ready any more
*/
static int jpeg_v2_5_hw_fini(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
struct amdgpu_ring *ring;
int i;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
ring = &adev->jpeg.inst[i].ring_dec;
if (adev->jpeg.cur_state != AMD_PG_STATE_GATE &&
RREG32_SOC15(JPEG, i, mmUVD_JRBC_STATUS))
jpeg_v2_5_set_powergating_state(adev, AMD_PG_STATE_GATE);
ring->sched.ready = false;
}
return 0;
}
/**
* jpeg_v2_5_suspend - suspend JPEG block
*
* @handle: amdgpu_device pointer
*
* HW fini and suspend JPEG block
*/
static int jpeg_v2_5_suspend(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int r;
r = jpeg_v2_5_hw_fini(adev);
if (r)
return r;
r = amdgpu_jpeg_suspend(adev);
return r;
}
/**
* jpeg_v2_5_resume - resume JPEG block
*
* @handle: amdgpu_device pointer
*
* Resume firmware and hw init JPEG block
*/
static int jpeg_v2_5_resume(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int r;
r = amdgpu_jpeg_resume(adev);
if (r)
return r;
r = jpeg_v2_5_hw_init(adev);
return r;
}
static void jpeg_v2_5_disable_clock_gating(struct amdgpu_device* adev, int inst)
{
uint32_t data;
data = RREG32_SOC15(JPEG, inst, mmJPEG_CGC_CTRL);
if (adev->cg_flags & AMD_CG_SUPPORT_JPEG_MGCG)
data |= 1 << JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
else
data &= ~JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
data |= 1 << JPEG_CGC_CTRL__CLK_GATE_DLY_TIMER__SHIFT;
data |= 4 << JPEG_CGC_CTRL__CLK_OFF_DELAY__SHIFT;
WREG32_SOC15(JPEG, inst, mmJPEG_CGC_CTRL, data);
data = RREG32_SOC15(JPEG, inst, mmJPEG_CGC_GATE);
data &= ~(JPEG_CGC_GATE__JPEG_DEC_MASK
| JPEG_CGC_GATE__JPEG2_DEC_MASK
| JPEG_CGC_GATE__JPEG_ENC_MASK
| JPEG_CGC_GATE__JMCIF_MASK
| JPEG_CGC_GATE__JRBBM_MASK);
WREG32_SOC15(JPEG, inst, mmJPEG_CGC_GATE, data);
data = RREG32_SOC15(JPEG, inst, mmJPEG_CGC_CTRL);
data &= ~(JPEG_CGC_CTRL__JPEG_DEC_MODE_MASK
| JPEG_CGC_CTRL__JPEG2_DEC_MODE_MASK
| JPEG_CGC_CTRL__JMCIF_MODE_MASK
| JPEG_CGC_CTRL__JRBBM_MODE_MASK);
WREG32_SOC15(JPEG, inst, mmJPEG_CGC_CTRL, data);
}
static void jpeg_v2_5_enable_clock_gating(struct amdgpu_device* adev, int inst)
{
uint32_t data;
data = RREG32_SOC15(JPEG, inst, mmJPEG_CGC_GATE);
data |= (JPEG_CGC_GATE__JPEG_DEC_MASK
|JPEG_CGC_GATE__JPEG2_DEC_MASK
|JPEG_CGC_GATE__JPEG_ENC_MASK
|JPEG_CGC_GATE__JMCIF_MASK
|JPEG_CGC_GATE__JRBBM_MASK);
WREG32_SOC15(JPEG, inst, mmJPEG_CGC_GATE, data);
}
/**
* jpeg_v2_5_start - start JPEG block
*
* @adev: amdgpu_device pointer
*
* Setup and start the JPEG block
*/
static int jpeg_v2_5_start(struct amdgpu_device *adev)
{
struct amdgpu_ring *ring;
int i;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
ring = &adev->jpeg.inst[i].ring_dec;
/* disable anti hang mechanism */
WREG32_P(SOC15_REG_OFFSET(JPEG, i, mmUVD_JPEG_POWER_STATUS), 0,
~UVD_JPEG_POWER_STATUS__JPEG_POWER_STATUS_MASK);
/* JPEG disable CGC */
jpeg_v2_5_disable_clock_gating(adev, i);
/* MJPEG global tiling registers */
WREG32_SOC15(JPEG, i, mmJPEG_DEC_GFX8_ADDR_CONFIG,
adev->gfx.config.gb_addr_config);
WREG32_SOC15(JPEG, i, mmJPEG_DEC_GFX10_ADDR_CONFIG,
adev->gfx.config.gb_addr_config);
/* enable JMI channel */
WREG32_P(SOC15_REG_OFFSET(JPEG, i, mmUVD_JMI_CNTL), 0,
~UVD_JMI_CNTL__SOFT_RESET_MASK);
/* enable System Interrupt for JRBC */
WREG32_P(SOC15_REG_OFFSET(JPEG, i, mmJPEG_SYS_INT_EN),
JPEG_SYS_INT_EN__DJRBC_MASK,
~JPEG_SYS_INT_EN__DJRBC_MASK);
WREG32_SOC15(JPEG, i, mmUVD_LMI_JRBC_RB_VMID, 0);
WREG32_SOC15(JPEG, i, mmUVD_JRBC_RB_CNTL, (0x00000001L | 0x00000002L));
WREG32_SOC15(JPEG, i, mmUVD_LMI_JRBC_RB_64BIT_BAR_LOW,
lower_32_bits(ring->gpu_addr));
WREG32_SOC15(JPEG, i, mmUVD_LMI_JRBC_RB_64BIT_BAR_HIGH,
upper_32_bits(ring->gpu_addr));
WREG32_SOC15(JPEG, i, mmUVD_JRBC_RB_RPTR, 0);
WREG32_SOC15(JPEG, i, mmUVD_JRBC_RB_WPTR, 0);
WREG32_SOC15(JPEG, i, mmUVD_JRBC_RB_CNTL, 0x00000002L);
WREG32_SOC15(JPEG, i, mmUVD_JRBC_RB_SIZE, ring->ring_size / 4);
ring->wptr = RREG32_SOC15(JPEG, i, mmUVD_JRBC_RB_WPTR);
}
return 0;
}
/**
* jpeg_v2_5_stop - stop JPEG block
*
* @adev: amdgpu_device pointer
*
* stop the JPEG block
*/
static int jpeg_v2_5_stop(struct amdgpu_device *adev)
{
int i;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
/* reset JMI */
WREG32_P(SOC15_REG_OFFSET(JPEG, i, mmUVD_JMI_CNTL),
UVD_JMI_CNTL__SOFT_RESET_MASK,
~UVD_JMI_CNTL__SOFT_RESET_MASK);
jpeg_v2_5_enable_clock_gating(adev, i);
/* enable anti hang mechanism */
WREG32_P(SOC15_REG_OFFSET(JPEG, i, mmUVD_JPEG_POWER_STATUS),
UVD_JPEG_POWER_STATUS__JPEG_POWER_STATUS_MASK,
~UVD_JPEG_POWER_STATUS__JPEG_POWER_STATUS_MASK);
}
return 0;
}
/**
* jpeg_v2_5_dec_ring_get_rptr - get read pointer
*
* @ring: amdgpu_ring pointer
*
* Returns the current hardware read pointer
*/
static uint64_t jpeg_v2_5_dec_ring_get_rptr(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
return RREG32_SOC15(JPEG, ring->me, mmUVD_JRBC_RB_RPTR);
}
/**
* jpeg_v2_5_dec_ring_get_wptr - get write pointer
*
* @ring: amdgpu_ring pointer
*
* Returns the current hardware write pointer
*/
static uint64_t jpeg_v2_5_dec_ring_get_wptr(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
if (ring->use_doorbell)
return adev->wb.wb[ring->wptr_offs];
else
return RREG32_SOC15(JPEG, ring->me, mmUVD_JRBC_RB_WPTR);
}
/**
* jpeg_v2_5_dec_ring_set_wptr - set write pointer
*
* @ring: amdgpu_ring pointer
*
* Commits the write pointer to the hardware
*/
static void jpeg_v2_5_dec_ring_set_wptr(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
if (ring->use_doorbell) {
adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr);
WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr));
} else {
WREG32_SOC15(JPEG, ring->me, mmUVD_JRBC_RB_WPTR, lower_32_bits(ring->wptr));
}
}
static bool jpeg_v2_5_is_idle(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int i, ret = 1;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
ret &= (((RREG32_SOC15(JPEG, i, mmUVD_JRBC_STATUS) &
UVD_JRBC_STATUS__RB_JOB_DONE_MASK) ==
UVD_JRBC_STATUS__RB_JOB_DONE_MASK));
}
return ret;
}
static int jpeg_v2_5_wait_for_idle(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int i, ret = 0;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
SOC15_WAIT_ON_RREG(JPEG, i, mmUVD_JRBC_STATUS,
UVD_JRBC_STATUS__RB_JOB_DONE_MASK,
UVD_JRBC_STATUS__RB_JOB_DONE_MASK, ret);
if (ret)
return ret;
}
return ret;
}
static int jpeg_v2_5_set_clockgating_state(void *handle,
enum amd_clockgating_state state)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
bool enable = (state == AMD_CG_STATE_GATE) ? true : false;
int i;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
if (enable) {
if (jpeg_v2_5_is_idle(handle))
return -EBUSY;
jpeg_v2_5_enable_clock_gating(adev, i);
} else {
jpeg_v2_5_disable_clock_gating(adev, i);
}
}
return 0;
}
static int jpeg_v2_5_set_powergating_state(void *handle,
enum amd_powergating_state state)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int ret;
if(state == adev->jpeg.cur_state)
return 0;
if (state == AMD_PG_STATE_GATE)
ret = jpeg_v2_5_stop(adev);
else
ret = jpeg_v2_5_start(adev);
if(!ret)
adev->jpeg.cur_state = state;
return ret;
}
static int jpeg_v2_5_set_interrupt_state(struct amdgpu_device *adev,
struct amdgpu_irq_src *source,
unsigned type,
enum amdgpu_interrupt_state state)
{
return 0;
}
static int jpeg_v2_5_process_interrupt(struct amdgpu_device *adev,
struct amdgpu_irq_src *source,
struct amdgpu_iv_entry *entry)
{
uint32_t ip_instance;
switch (entry->client_id) {
case SOC15_IH_CLIENTID_VCN:
ip_instance = 0;
break;
case SOC15_IH_CLIENTID_VCN1:
ip_instance = 1;
break;
default:
DRM_ERROR("Unhandled client id: %d\n", entry->client_id);
return 0;
}
DRM_DEBUG("IH: JPEG TRAP\n");
switch (entry->src_id) {
case VCN_2_0__SRCID__JPEG_DECODE:
amdgpu_fence_process(&adev->jpeg.inst[ip_instance].ring_dec);
break;
default:
DRM_ERROR("Unhandled interrupt: %d %d\n",
entry->src_id, entry->src_data[0]);
break;
}
return 0;
}
static const struct amd_ip_funcs jpeg_v2_5_ip_funcs = {
.name = "jpeg_v2_5",
.early_init = jpeg_v2_5_early_init,
.late_init = NULL,
.sw_init = jpeg_v2_5_sw_init,
.sw_fini = jpeg_v2_5_sw_fini,
.hw_init = jpeg_v2_5_hw_init,
.hw_fini = jpeg_v2_5_hw_fini,
.suspend = jpeg_v2_5_suspend,
.resume = jpeg_v2_5_resume,
.is_idle = jpeg_v2_5_is_idle,
.wait_for_idle = jpeg_v2_5_wait_for_idle,
.check_soft_reset = NULL,
.pre_soft_reset = NULL,
.soft_reset = NULL,
.post_soft_reset = NULL,
.set_clockgating_state = jpeg_v2_5_set_clockgating_state,
.set_powergating_state = jpeg_v2_5_set_powergating_state,
};
static const struct amdgpu_ring_funcs jpeg_v2_5_dec_ring_vm_funcs = {
.type = AMDGPU_RING_TYPE_VCN_JPEG,
.align_mask = 0xf,
.vmhub = AMDGPU_MMHUB_1,
.get_rptr = jpeg_v2_5_dec_ring_get_rptr,
.get_wptr = jpeg_v2_5_dec_ring_get_wptr,
.set_wptr = jpeg_v2_5_dec_ring_set_wptr,
.emit_frame_size =
SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
8 + /* jpeg_v2_5_dec_ring_emit_vm_flush */
18 + 18 + /* jpeg_v2_5_dec_ring_emit_fence x2 vm fence */
8 + 16,
.emit_ib_size = 22, /* jpeg_v2_5_dec_ring_emit_ib */
.emit_ib = jpeg_v2_0_dec_ring_emit_ib,
.emit_fence = jpeg_v2_0_dec_ring_emit_fence,
.emit_vm_flush = jpeg_v2_0_dec_ring_emit_vm_flush,
.test_ring = amdgpu_jpeg_dec_ring_test_ring,
.test_ib = amdgpu_jpeg_dec_ring_test_ib,
.insert_nop = jpeg_v2_0_dec_ring_nop,
.insert_start = jpeg_v2_0_dec_ring_insert_start,
.insert_end = jpeg_v2_0_dec_ring_insert_end,
.pad_ib = amdgpu_ring_generic_pad_ib,
.begin_use = amdgpu_jpeg_ring_begin_use,
.end_use = amdgpu_jpeg_ring_end_use,
.emit_wreg = jpeg_v2_0_dec_ring_emit_wreg,
.emit_reg_wait = jpeg_v2_0_dec_ring_emit_reg_wait,
.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
};
static void jpeg_v2_5_set_dec_ring_funcs(struct amdgpu_device *adev)
{
int i;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
adev->jpeg.inst[i].ring_dec.funcs = &jpeg_v2_5_dec_ring_vm_funcs;
adev->jpeg.inst[i].ring_dec.me = i;
DRM_INFO("JPEG(%d) JPEG decode is enabled in VM mode\n", i);
}
}
static const struct amdgpu_irq_src_funcs jpeg_v2_5_irq_funcs = {
.set = jpeg_v2_5_set_interrupt_state,
.process = jpeg_v2_5_process_interrupt,
};
static void jpeg_v2_5_set_irq_funcs(struct amdgpu_device *adev)
{
int i;
for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
if (adev->jpeg.harvest_config & (1 << i))
continue;
adev->jpeg.inst[i].irq.num_types = 1;
adev->jpeg.inst[i].irq.funcs = &jpeg_v2_5_irq_funcs;
}
}
const struct amdgpu_ip_block_version jpeg_v2_5_ip_block =
{
.type = AMD_IP_BLOCK_TYPE_JPEG,
.major = 2,
.minor = 5,
.rev = 0,
.funcs = &jpeg_v2_5_ip_funcs,
};

29
drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.h Normal file
View File

@ -0,0 +1,29 @@
/*
* Copyright 2019 Advanced Micro Devices, Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*
*/
#ifndef __JPEG_V2_5_H__
#define __JPEG_V2_5_H__
extern const struct amdgpu_ip_block_version jpeg_v2_5_ip_block;
#endif /* __JPEG_V2_5_H__ */

236
drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
View File

@ -47,26 +47,6 @@
#define mmUVD_LMI_RBC_IB_64BIT_BAR_LOW_INTERNAL_OFFSET 0x5a7
#define mmUVD_RBC_IB_SIZE_INTERNAL_OFFSET 0x1e2
#define mmUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET 0x1bfff
#define mmUVD_JPEG_GPCOM_CMD_INTERNAL_OFFSET 0x4029
#define mmUVD_JPEG_GPCOM_DATA0_INTERNAL_OFFSET 0x402a
#define mmUVD_JPEG_GPCOM_DATA1_INTERNAL_OFFSET 0x402b
#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_LOW_INTERNAL_OFFSET 0x40ea
#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_HIGH_INTERNAL_OFFSET 0x40eb
#define mmUVD_LMI_JRBC_IB_VMID_INTERNAL_OFFSET 0x40cf
#define mmUVD_LMI_JPEG_VMID_INTERNAL_OFFSET 0x40d1
#define mmUVD_LMI_JRBC_IB_64BIT_BAR_LOW_INTERNAL_OFFSET 0x40e8
#define mmUVD_LMI_JRBC_IB_64BIT_BAR_HIGH_INTERNAL_OFFSET 0x40e9
#define mmUVD_JRBC_IB_SIZE_INTERNAL_OFFSET 0x4082
#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW_INTERNAL_OFFSET 0x40ec
#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH_INTERNAL_OFFSET 0x40ed
#define mmUVD_JRBC_RB_COND_RD_TIMER_INTERNAL_OFFSET 0x4085
#define mmUVD_JRBC_RB_REF_DATA_INTERNAL_OFFSET 0x4084
#define mmUVD_JRBC_STATUS_INTERNAL_OFFSET 0x4089
#define mmUVD_JPEG_PITCH_INTERNAL_OFFSET 0x401f
#define JRBC_DEC_EXTERNAL_REG_WRITE_ADDR 0x18000
#define mmUVD_RBC_XX_IB_REG_CHECK 0x026b
#define mmUVD_RBC_XX_IB_REG_CHECK_BASE_IDX 1
#define mmUVD_REG_XX_MASK 0x026c
@ -1616,222 +1596,6 @@ void vcn_v2_0_enc_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_
amdgpu_ring_write(ring, val);
}
/**
* vcn_v2_0_jpeg_ring_insert_start - insert a start command
*
* @ring: amdgpu_ring pointer
*
* Write a start command to the ring.
*/
void vcn_v2_0_jpeg_ring_insert_start(struct amdgpu_ring *ring)
{
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x68e04);
amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x80010000);
}
/**
* vcn_v2_0_jpeg_ring_insert_end - insert a end command
*
* @ring: amdgpu_ring pointer
*
* Write a end command to the ring.
*/
void vcn_v2_0_jpeg_ring_insert_end(struct amdgpu_ring *ring)
{
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x68e04);
amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x00010000);
}
/**
* vcn_v2_0_jpeg_ring_emit_fence - emit an fence & trap command
*
* @ring: amdgpu_ring pointer
* @fence: fence to emit
*
* Write a fence and a trap command to the ring.
*/
void vcn_v2_0_jpeg_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
unsigned flags)
{
WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JPEG_GPCOM_DATA0_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, seq);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JPEG_GPCOM_DATA1_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, seq);
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_LOW_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, lower_32_bits(addr));
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_HIGH_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, upper_32_bits(addr));
amdgpu_ring_write(ring, PACKETJ(mmUVD_JPEG_GPCOM_CMD_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x8);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JPEG_GPCOM_CMD_INTERNAL_OFFSET,
0, PACKETJ_CONDITION_CHECK0, PACKETJ_TYPE4));
amdgpu_ring_write(ring, 0);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x3fbc);
amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x1);
amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE7));
amdgpu_ring_write(ring, 0);
}
/**
* vcn_v2_0_jpeg_ring_emit_ib - execute indirect buffer
*
* @ring: amdgpu_ring pointer
* @ib: indirect buffer to execute
*
* Write ring commands to execute the indirect buffer.
*/
void vcn_v2_0_jpeg_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_IB_VMID_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, (vmid | (vmid << 4)));
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JPEG_VMID_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, (vmid | (vmid << 4)));
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_IB_64BIT_BAR_LOW_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_IB_64BIT_BAR_HIGH_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr));
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_IB_SIZE_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, ib->length_dw);
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, lower_32_bits(ring->gpu_addr));
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, upper_32_bits(ring->gpu_addr));
amdgpu_ring_write(ring, PACKETJ(0, 0, PACKETJ_CONDITION_CHECK0, PACKETJ_TYPE2));
amdgpu_ring_write(ring, 0);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_RB_COND_RD_TIMER_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x01400200);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_RB_REF_DATA_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x2);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_STATUS_INTERNAL_OFFSET,
0, PACKETJ_CONDITION_CHECK3, PACKETJ_TYPE3));
amdgpu_ring_write(ring, 0x2);
}
void vcn_v2_0_jpeg_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
uint32_t val, uint32_t mask)
{
uint32_t reg_offset = (reg << 2);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_RB_COND_RD_TIMER_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, 0x01400200);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_RB_REF_DATA_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, val);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
if (reg_offset >= 0x10000 && reg_offset <= 0x105ff) {
amdgpu_ring_write(ring, 0);
amdgpu_ring_write(ring,
PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE3));
} else {
amdgpu_ring_write(ring, reg_offset);
amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR,
0, 0, PACKETJ_TYPE3));
}
amdgpu_ring_write(ring, mask);
}
void vcn_v2_0_jpeg_ring_emit_vm_flush(struct amdgpu_ring *ring,
unsigned vmid, uint64_t pd_addr)
{
struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
uint32_t data0, data1, mask;
pd_addr = amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr);
/* wait for register write */
data0 = hub->ctx0_ptb_addr_lo32 + vmid * 2;
data1 = lower_32_bits(pd_addr);
mask = 0xffffffff;
vcn_v2_0_jpeg_ring_emit_reg_wait(ring, data0, data1, mask);
}
void vcn_v2_0_jpeg_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val)
{
uint32_t reg_offset = (reg << 2);
amdgpu_ring_write(ring, PACKETJ(mmUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
if (reg_offset >= 0x10000 && reg_offset <= 0x105ff) {
amdgpu_ring_write(ring, 0);
amdgpu_ring_write(ring,
PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE0));
} else {
amdgpu_ring_write(ring, reg_offset);
amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR,
0, 0, PACKETJ_TYPE0));
}
amdgpu_ring_write(ring, val);
}
void vcn_v2_0_jpeg_ring_nop(struct amdgpu_ring *ring, uint32_t count)
{
int i;
WARN_ON(ring->wptr % 2 || count % 2);
for (i = 0; i < count / 2; i++) {
amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE6));
amdgpu_ring_write(ring, 0);
}
}
static int vcn_v2_0_set_interrupt_state(struct amdgpu_device *adev,
struct amdgpu_irq_src *source,
unsigned type,

13
drivers/gpu/drm/amd/amdgpu/vcn_v2_0.h
View File

@ -49,19 +49,6 @@ extern void vcn_v2_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring,
unsigned int vmid, uint64_t pd_addr);
extern void vcn_v2_0_enc_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val);
extern void vcn_v2_0_jpeg_ring_insert_start(struct amdgpu_ring *ring);
extern void vcn_v2_0_jpeg_ring_insert_end(struct amdgpu_ring *ring);
extern void vcn_v2_0_jpeg_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
unsigned flags);
extern void vcn_v2_0_jpeg_ring_emit_ib(struct amdgpu_ring *ring, struct amdgpu_job *job,
struct amdgpu_ib *ib, uint32_t flags);
extern void vcn_v2_0_jpeg_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
uint32_t val, uint32_t mask);
extern void vcn_v2_0_jpeg_ring_emit_vm_flush(struct amdgpu_ring *ring,
unsigned vmid, uint64_t pd_addr);
extern void vcn_v2_0_jpeg_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val);
extern void vcn_v2_0_jpeg_ring_nop(struct amdgpu_ring *ring, uint32_t count);
extern const struct amdgpu_ip_block_version vcn_v2_0_ip_block;
#endif /* __VCN_V2_0_H__ */

244
drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
View File

@ -47,13 +47,10 @@
#define mmUVD_LMI_RBC_IB_64BIT_BAR_HIGH_INTERNAL_OFFSET 0x3b5
#define mmUVD_RBC_IB_SIZE_INTERNAL_OFFSET 0x25c
#define mmUVD_JPEG_PITCH_INTERNAL_OFFSET 0x401f
#define VCN25_MAX_HW_INSTANCES_ARCTURUS 2
static void vcn_v2_5_set_dec_ring_funcs(struct amdgpu_device *adev);
static void vcn_v2_5_set_enc_ring_funcs(struct amdgpu_device *adev);
static void vcn_v2_5_set_jpeg_ring_funcs(struct amdgpu_device *adev);
static void vcn_v2_5_set_irq_funcs(struct amdgpu_device *adev);
static int vcn_v2_5_set_powergating_state(void *handle,
enum amd_powergating_state state);
@ -95,7 +92,6 @@ static int vcn_v2_5_early_init(void *handle)
vcn_v2_5_set_dec_ring_funcs(adev);
vcn_v2_5_set_enc_ring_funcs(adev);
vcn_v2_5_set_jpeg_ring_funcs(adev);
vcn_v2_5_set_irq_funcs(adev);
return 0;
@ -130,12 +126,6 @@ static int vcn_v2_5_sw_init(void *handle)
if (r)
return r;
}
/* VCN JPEG TRAP */
r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_vcns[j],
VCN_2_0__SRCID__JPEG_DECODE, &adev->vcn.inst[j].irq);
if (r)
return r;
}
r = amdgpu_vcn_sw_init(adev);
@ -184,9 +174,6 @@ static int vcn_v2_5_sw_init(void *handle)
adev->vcn.internal.nop = mmUVD_NO_OP_INTERNAL_OFFSET;
adev->vcn.inst[j].external.nop = SOC15_REG_OFFSET(UVD, j, mmUVD_NO_OP);
adev->vcn.internal.jpeg_pitch = mmUVD_JPEG_PITCH_INTERNAL_OFFSET;
adev->vcn.inst[j].external.jpeg_pitch = SOC15_REG_OFFSET(UVD, j, mmUVD_JPEG_PITCH);
ring = &adev->vcn.inst[j].ring_dec;
ring->use_doorbell = true;
ring->doorbell_index = (adev->doorbell_index.vcn.vcn_ring0_1 << 1) + 8*j;
@ -204,14 +191,6 @@ static int vcn_v2_5_sw_init(void *handle)
if (r)
return r;
}
ring = &adev->vcn.inst[j].ring_jpeg;
ring->use_doorbell = true;
ring->doorbell_index = (adev->doorbell_index.vcn.vcn_ring0_1 << 1) + 1 + 8*j;
sprintf(ring->name, "vcn_jpeg_%d", j);
r = amdgpu_ring_init(adev, ring, 512, &adev->vcn.inst[j].irq, 0);
if (r)
return r;
}
return 0;
@ -269,12 +248,8 @@ static int vcn_v2_5_hw_init(void *handle)
if (r)
goto done;
}
ring = &adev->vcn.inst[j].ring_jpeg;
r = amdgpu_ring_test_helper(ring);
if (r)
goto done;
}
done:
if (!r)
DRM_INFO("VCN decode and encode initialized successfully.\n");
@ -309,9 +284,6 @@ static int vcn_v2_5_hw_fini(void *handle)
ring = &adev->vcn.inst[i].ring_enc[i];
ring->sched.ready = false;
}
ring = &adev->vcn.inst[i].ring_jpeg;
ring->sched.ready = false;
}
return 0;
@ -592,115 +564,6 @@ static void vcn_v2_5_enable_clock_gating(struct amdgpu_device *adev)
}
}
/**
* jpeg_v2_5_start - start JPEG block
*
* @adev: amdgpu_device pointer
*
* Setup and start the JPEG block
*/
static int jpeg_v2_5_start(struct amdgpu_device *adev)
{
struct amdgpu_ring *ring;
uint32_t tmp;
int i;
for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
if (adev->vcn.harvest_config & (1 << i))
continue;
ring = &adev->vcn.inst[i].ring_jpeg;
/* disable anti hang mechanism */
WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_JPEG_POWER_STATUS), 0,
~UVD_JPEG_POWER_STATUS__JPEG_POWER_STATUS_MASK);
/* JPEG disable CGC */
tmp = RREG32_SOC15(VCN, i, mmJPEG_CGC_CTRL);
tmp |= 1 << JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
tmp |= 1 << JPEG_CGC_CTRL__CLK_GATE_DLY_TIMER__SHIFT;
tmp |= 4 << JPEG_CGC_CTRL__CLK_OFF_DELAY__SHIFT;
WREG32_SOC15(VCN, i, mmJPEG_CGC_CTRL, tmp);
tmp = RREG32_SOC15(VCN, i, mmJPEG_CGC_GATE);
tmp &= ~(JPEG_CGC_GATE__JPEG_DEC_MASK
| JPEG_CGC_GATE__JPEG2_DEC_MASK
| JPEG_CGC_GATE__JMCIF_MASK
| JPEG_CGC_GATE__JRBBM_MASK);
WREG32_SOC15(VCN, i, mmJPEG_CGC_GATE, tmp);
tmp = RREG32_SOC15(VCN, i, mmJPEG_CGC_CTRL);
tmp &= ~(JPEG_CGC_CTRL__JPEG_DEC_MODE_MASK
| JPEG_CGC_CTRL__JPEG2_DEC_MODE_MASK
| JPEG_CGC_CTRL__JMCIF_MODE_MASK
| JPEG_CGC_CTRL__JRBBM_MODE_MASK);
WREG32_SOC15(VCN, i, mmJPEG_CGC_CTRL, tmp);
/* MJPEG global tiling registers */
WREG32_SOC15(UVD, i, mmJPEG_DEC_GFX8_ADDR_CONFIG,
adev->gfx.config.gb_addr_config);
WREG32_SOC15(UVD, i, mmJPEG_DEC_GFX10_ADDR_CONFIG,
adev->gfx.config.gb_addr_config);
/* enable JMI channel */
WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_JMI_CNTL), 0,
~UVD_JMI_CNTL__SOFT_RESET_MASK);
/* enable System Interrupt for JRBC */
WREG32_P(SOC15_REG_OFFSET(VCN, i, mmJPEG_SYS_INT_EN),
JPEG_SYS_INT_EN__DJRBC_MASK,
~JPEG_SYS_INT_EN__DJRBC_MASK);
WREG32_SOC15(UVD, i, mmUVD_LMI_JRBC_RB_VMID, 0);
WREG32_SOC15(UVD, i, mmUVD_JRBC_RB_CNTL, (0x00000001L | 0x00000002L));
WREG32_SOC15(UVD, i, mmUVD_LMI_JRBC_RB_64BIT_BAR_LOW,
lower_32_bits(ring->gpu_addr));
WREG32_SOC15(UVD, i, mmUVD_LMI_JRBC_RB_64BIT_BAR_HIGH,
upper_32_bits(ring->gpu_addr));
WREG32_SOC15(UVD, i, mmUVD_JRBC_RB_RPTR, 0);
WREG32_SOC15(UVD, i, mmUVD_JRBC_RB_WPTR, 0);
WREG32_SOC15(UVD, i, mmUVD_JRBC_RB_CNTL, 0x00000002L);
WREG32_SOC15(UVD, i, mmUVD_JRBC_RB_SIZE, ring->ring_size / 4);
ring->wptr = RREG32_SOC15(UVD, i, mmUVD_JRBC_RB_WPTR);
}
return 0;
}
/**
* jpeg_v2_5_stop - stop JPEG block
*
* @adev: amdgpu_device pointer
*
* stop the JPEG block
*/
static int jpeg_v2_5_stop(struct amdgpu_device *adev)
{
uint32_t tmp;
int i;
for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
if (adev->vcn.harvest_config & (1 << i))
continue;
/* reset JMI */
WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_JMI_CNTL),
UVD_JMI_CNTL__SOFT_RESET_MASK,
~UVD_JMI_CNTL__SOFT_RESET_MASK);
tmp = RREG32_SOC15(VCN, i, mmJPEG_CGC_GATE);
tmp |= (JPEG_CGC_GATE__JPEG_DEC_MASK
|JPEG_CGC_GATE__JPEG2_DEC_MASK
|JPEG_CGC_GATE__JMCIF_MASK
|JPEG_CGC_GATE__JRBBM_MASK);
WREG32_SOC15(VCN, i, mmJPEG_CGC_GATE, tmp);
/* enable anti hang mechanism */
WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_JPEG_POWER_STATUS),
UVD_JPEG_POWER_STATUS__JPEG_POWER_STATUS_MASK,
~UVD_JPEG_POWER_STATUS__JPEG_POWER_STATUS_MASK);
}
return 0;
}
static int vcn_v2_5_start(struct amdgpu_device *adev)
{
struct amdgpu_ring *ring;
@ -874,19 +737,14 @@ static int vcn_v2_5_start(struct amdgpu_device *adev)
WREG32_SOC15(UVD, i, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
WREG32_SOC15(UVD, i, mmUVD_RB_SIZE2, ring->ring_size / 4);
}
r = jpeg_v2_5_start(adev);
return r;
return 0;
}
static int vcn_v2_5_stop(struct amdgpu_device *adev)
{
uint32_t tmp;
int i, r;
r = jpeg_v2_5_stop(adev);
if (r)
return r;
int i, r = 0;
for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
if (adev->vcn.harvest_config & (1 << i))
@ -1125,86 +983,6 @@ static const struct amdgpu_ring_funcs vcn_v2_5_enc_ring_vm_funcs = {
.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
};
/**
* vcn_v2_5_jpeg_ring_get_rptr - get read pointer
*
* @ring: amdgpu_ring pointer
*
* Returns the current hardware read pointer
*/
static uint64_t vcn_v2_5_jpeg_ring_get_rptr(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
return RREG32_SOC15(UVD, ring->me, mmUVD_JRBC_RB_RPTR);
}
/**
* vcn_v2_5_jpeg_ring_get_wptr - get write pointer
*
* @ring: amdgpu_ring pointer
*
* Returns the current hardware write pointer
*/
static uint64_t vcn_v2_5_jpeg_ring_get_wptr(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
if (ring->use_doorbell)
return adev->wb.wb[ring->wptr_offs];
else
return RREG32_SOC15(UVD, ring->me, mmUVD_JRBC_RB_WPTR);
}
/**
* vcn_v2_5_jpeg_ring_set_wptr - set write pointer
*
* @ring: amdgpu_ring pointer
*
* Commits the write pointer to the hardware
*/
static void vcn_v2_5_jpeg_ring_set_wptr(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
if (ring->use_doorbell) {
adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr);
WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr));
} else {
WREG32_SOC15(UVD, ring->me, mmUVD_JRBC_RB_WPTR, lower_32_bits(ring->wptr));
}
}
static const struct amdgpu_ring_funcs vcn_v2_5_jpeg_ring_vm_funcs = {
.type = AMDGPU_RING_TYPE_VCN_JPEG,
.align_mask = 0xf,
.vmhub = AMDGPU_MMHUB_1,
.get_rptr = vcn_v2_5_jpeg_ring_get_rptr,
.get_wptr = vcn_v2_5_jpeg_ring_get_wptr,
.set_wptr = vcn_v2_5_jpeg_ring_set_wptr,
.emit_frame_size =
SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
8 + /* vcn_v2_0_jpeg_ring_emit_vm_flush */
18 + 18 + /* vcn_v2_0_jpeg_ring_emit_fence x2 vm fence */
8 + 16,
.emit_ib_size = 22, /* vcn_v2_0_jpeg_ring_emit_ib */
.emit_ib = vcn_v2_0_jpeg_ring_emit_ib,
.emit_fence = vcn_v2_0_jpeg_ring_emit_fence,
.emit_vm_flush = vcn_v2_0_jpeg_ring_emit_vm_flush,
.test_ring = amdgpu_vcn_jpeg_ring_test_ring,
.test_ib = amdgpu_vcn_jpeg_ring_test_ib,
.insert_nop = vcn_v2_0_jpeg_ring_nop,
.insert_start = vcn_v2_0_jpeg_ring_insert_start,
.insert_end = vcn_v2_0_jpeg_ring_insert_end,
.pad_ib = amdgpu_ring_generic_pad_ib,
.begin_use = amdgpu_vcn_ring_begin_use,
.end_use = amdgpu_vcn_ring_end_use,
.emit_wreg = vcn_v2_0_jpeg_ring_emit_wreg,
.emit_reg_wait = vcn_v2_0_jpeg_ring_emit_reg_wait,
.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
};
static void vcn_v2_5_set_dec_ring_funcs(struct amdgpu_device *adev)
{
int i;
@ -1233,19 +1011,6 @@ static void vcn_v2_5_set_enc_ring_funcs(struct amdgpu_device *adev)
}
}
static void vcn_v2_5_set_jpeg_ring_funcs(struct amdgpu_device *adev)
{
int i;
for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
if (adev->vcn.harvest_config & (1 << i))
continue;
adev->vcn.inst[i].ring_jpeg.funcs = &vcn_v2_5_jpeg_ring_vm_funcs;
adev->vcn.inst[i].ring_jpeg.me = i;
DRM_INFO("VCN(%d) jpeg decode is enabled in VM mode\n", i);
}
}
static bool vcn_v2_5_is_idle(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
@ -1352,9 +1117,6 @@ static int vcn_v2_5_process_interrupt(struct amdgpu_device *adev,
case VCN_2_0__SRCID__UVD_ENC_LOW_LATENCY:
amdgpu_fence_process(&adev->vcn.inst[ip_instance].ring_enc[1]);
break;
case VCN_2_0__SRCID__JPEG_DECODE:
amdgpu_fence_process(&adev->vcn.inst[ip_instance].ring_jpeg);
break;
default:
DRM_ERROR("Unhandled interrupt: %d %d\n",
entry->src_id, entry->src_data[0]);