xiaomi-sm8450/android_kernel_xiaomi_sm8450
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge branch 'drm-next-3.12' of git://people.freedesktop.org/~agd5f/linux into drm-next

Browse Source 
Alex writes:
This is the radeon drm-next request.  Big changes include:
- support for dpm on CIK parts
- support for ASPM on CIK parts
- support for berlin GPUs
- major ring handling cleanup
- remove the old 3D blit code for bo moves in favor of CP DMA or sDMA
- lots of bug fixes

[airlied: fix up a bunch of conflicts from drm_order removal]

* 'drm-next-3.12' of git://people.freedesktop.org/~agd5f/linux: (898 commits)
  drm/radeon/dpm: make sure dc performance level limits are valid (CI)
  drm/radeon/dpm: make sure dc performance level limits are valid (BTC-SI) (v2)
  drm/radeon: gcc fixes for extended dpm tables
  drm/radeon: gcc fixes for kb/kv dpm
  drm/radeon: gcc fixes for ci dpm
  drm/radeon: gcc fixes for si dpm
  drm/radeon: gcc fixes for ni dpm
  drm/radeon: gcc fixes for trinity dpm
  drm/radeon: gcc fixes for sumo dpm
  drm/radeonn: gcc fixes for rv7xx/eg/btc dpm
  drm/radeon: gcc fixes for rv6xx dpm
  drm/radeon: gcc fixes for radeon_atombios.c
  drm/radeon: enable UVD interrupts on CIK
  drm/radeon: fix init ordering for r600+
  drm/radeon/dpm: only need to reprogram uvd if uvd pg is enabled
  drm/radeon: check the return value of uvd_v1_0_start in uvd_v1_0_init
  drm/radeon: split out radeon_uvd_resume from uvd_v4_2_resume
  radeon kms: fix uninitialised hotplug work usage in r100_irq_process()
  drm/radeon/audio: set up the sads on DCE3.2 asics
  drm/radeon: fix handling of variable sized arrays for router objects
  ...

Conflicts:
	drivers/gpu/drm/i915/i915_dma.c
	drivers/gpu/drm/i915/i915_gem_dmabuf.c
	drivers/gpu/drm/i915/intel_pm.c
	drivers/gpu/drm/radeon/cik.c
	drivers/gpu/drm/radeon/ni.c
	drivers/gpu/drm/radeon/r600.c
This commit is contained in:
Dave Airlie
2013-09-02 09:31:40 +10:00
parent efa27f9cec 679fe80fbe
commit 9c725e5bcd
882 changed files with 28521 additions and 10932 deletions
Download Patch File Download Diff File Expand all files Collapse all files

381
drivers/gpu/drm/radeon/ni.c
View File

@@ -35,7 +35,7 @@
#include "radeon_ucode.h"
#include "clearstate_cayman.h"
static u32 tn_rlc_save_restore_register_list[] =
static const u32 tn_rlc_save_restore_register_list[] =
{
0x98fc,
0x98f0,
@@ -160,7 +160,6 @@ static u32 tn_rlc_save_restore_register_list[] =
0x9830,
0x802c,
};
static u32 tn_rlc_save_restore_register_list_size = ARRAY_SIZE(tn_rlc_save_restore_register_list);
extern bool evergreen_is_display_hung(struct radeon_device *rdev);
extern void evergreen_print_gpu_status_regs(struct radeon_device *rdev);
@@ -175,6 +174,11 @@ extern void evergreen_pcie_gen2_enable(struct radeon_device *rdev);
extern void evergreen_program_aspm(struct radeon_device *rdev);
extern void sumo_rlc_fini(struct radeon_device *rdev);
extern int sumo_rlc_init(struct radeon_device *rdev);
extern void cayman_dma_vm_set_page(struct radeon_device *rdev,
struct radeon_ib *ib,
uint64_t pe,
uint64_t addr, unsigned count,
uint32_t incr, uint32_t flags);
/* Firmware Names */
MODULE_FIRMWARE("radeon/BARTS_pfp.bin");
@@ -794,9 +798,13 @@ int ni_init_microcode(struct radeon_device *rdev)
if ((rdev->family >= CHIP_BARTS) && (rdev->family <= CHIP_CAYMAN)) {
snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
if (err)
goto out;
if (rdev->smc_fw->size != smc_req_size) {
if (err) {
printk(KERN_ERR
"smc: error loading firmware \"%s\"\n",
fw_name);
release_firmware(rdev->smc_fw);
rdev->smc_fw = NULL;
} else if (rdev->smc_fw->size != smc_req_size) {
printk(KERN_ERR
"ni_mc: Bogus length %zu in firmware \"%s\"\n",
rdev->mc_fw->size, fw_name);
@@ -1370,23 +1378,6 @@ void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
radeon_ring_write(ring, 10); /* poll interval */
}
void cayman_uvd_semaphore_emit(struct radeon_device *rdev,
struct radeon_ring *ring,
struct radeon_semaphore *semaphore,
bool emit_wait)
{
uint64_t addr = semaphore->gpu_addr;
radeon_ring_write(ring, PACKET0(UVD_SEMA_ADDR_LOW, 0));
radeon_ring_write(ring, (addr >> 3) & 0x000FFFFF);
radeon_ring_write(ring, PACKET0(UVD_SEMA_ADDR_HIGH, 0));
radeon_ring_write(ring, (addr >> 23) & 0x000FFFFF);
radeon_ring_write(ring, PACKET0(UVD_SEMA_CMD, 0));
radeon_ring_write(ring, 0x80 | (emit_wait ? 1 : 0));
}
static void cayman_cp_enable(struct radeon_device *rdev, bool enable)
{
if (enable)
@@ -1609,186 +1600,7 @@ static int cayman_cp_resume(struct radeon_device *rdev)
return 0;
}
/*
* DMA
* Starting with R600, the GPU has an asynchronous
* DMA engine. The programming model is very similar
* to the 3D engine (ring buffer, IBs, etc.), but the
* DMA controller has it's own packet format that is
* different form the PM4 format used by the 3D engine.
* It supports copying data, writing embedded data,
* solid fills, and a number of other things. It also
* has support for tiling/detiling of buffers.
* Cayman and newer support two asynchronous DMA engines.
*/
/**
* cayman_dma_ring_ib_execute - Schedule an IB on the DMA engine
*
* @rdev: radeon_device pointer
* @ib: IB object to schedule
*
* Schedule an IB in the DMA ring (cayman-SI).
*/
void cayman_dma_ring_ib_execute(struct radeon_device *rdev,
struct radeon_ib *ib)
{
struct radeon_ring *ring = &rdev->ring[ib->ring];
if (rdev->wb.enabled) {
u32 next_rptr = ring->wptr + 4;
while ((next_rptr & 7) != 5)
next_rptr++;
next_rptr += 3;
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 1));
radeon_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc);
radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xff);
radeon_ring_write(ring, next_rptr);
}
/* The indirect buffer packet must end on an 8 DW boundary in the DMA ring.
* Pad as necessary with NOPs.
*/
while ((ring->wptr & 7) != 5)
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
radeon_ring_write(ring, DMA_IB_PACKET(DMA_PACKET_INDIRECT_BUFFER, ib->vm ? ib->vm->id : 0, 0));
radeon_ring_write(ring, (ib->gpu_addr & 0xFFFFFFE0));
radeon_ring_write(ring, (ib->length_dw << 12) | (upper_32_bits(ib->gpu_addr) & 0xFF));
}
/**
* cayman_dma_stop - stop the async dma engines
*
* @rdev: radeon_device pointer
*
* Stop the async dma engines (cayman-SI).
*/
void cayman_dma_stop(struct radeon_device *rdev)
{
u32 rb_cntl;
radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
/* dma0 */
rb_cntl = RREG32(DMA_RB_CNTL + DMA0_REGISTER_OFFSET);
rb_cntl &= ~DMA_RB_ENABLE;
WREG32(DMA_RB_CNTL + DMA0_REGISTER_OFFSET, rb_cntl);
/* dma1 */
rb_cntl = RREG32(DMA_RB_CNTL + DMA1_REGISTER_OFFSET);
rb_cntl &= ~DMA_RB_ENABLE;
WREG32(DMA_RB_CNTL + DMA1_REGISTER_OFFSET, rb_cntl);
rdev->ring[R600_RING_TYPE_DMA_INDEX].ready = false;
rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX].ready = false;
}
/**
* cayman_dma_resume - setup and start the async dma engines
*
* @rdev: radeon_device pointer
*
* Set up the DMA ring buffers and enable them. (cayman-SI).
* Returns 0 for success, error for failure.
*/
int cayman_dma_resume(struct radeon_device *rdev)
{
struct radeon_ring *ring;
u32 rb_cntl, dma_cntl, ib_cntl;
u32 rb_bufsz;
u32 reg_offset, wb_offset;
int i, r;
/* Reset dma */
WREG32(SRBM_SOFT_RESET, SOFT_RESET_DMA | SOFT_RESET_DMA1);
RREG32(SRBM_SOFT_RESET);
udelay(50);
WREG32(SRBM_SOFT_RESET, 0);
for (i = 0; i < 2; i++) {
if (i == 0) {
ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
reg_offset = DMA0_REGISTER_OFFSET;
wb_offset = R600_WB_DMA_RPTR_OFFSET;
} else {
ring = &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX];
reg_offset = DMA1_REGISTER_OFFSET;
wb_offset = CAYMAN_WB_DMA1_RPTR_OFFSET;
}
WREG32(DMA_SEM_INCOMPLETE_TIMER_CNTL + reg_offset, 0);
WREG32(DMA_SEM_WAIT_FAIL_TIMER_CNTL + reg_offset, 0);
/* Set ring buffer size in dwords */
rb_bufsz = order_base_2(ring->ring_size / 4);
rb_cntl = rb_bufsz << 1;
#ifdef __BIG_ENDIAN
rb_cntl |= DMA_RB_SWAP_ENABLE | DMA_RPTR_WRITEBACK_SWAP_ENABLE;
#endif
WREG32(DMA_RB_CNTL + reg_offset, rb_cntl);
/* Initialize the ring buffer's read and write pointers */
WREG32(DMA_RB_RPTR + reg_offset, 0);
WREG32(DMA_RB_WPTR + reg_offset, 0);
/* set the wb address whether it's enabled or not */
WREG32(DMA_RB_RPTR_ADDR_HI + reg_offset,
upper_32_bits(rdev->wb.gpu_addr + wb_offset) & 0xFF);
WREG32(DMA_RB_RPTR_ADDR_LO + reg_offset,
((rdev->wb.gpu_addr + wb_offset) & 0xFFFFFFFC));
if (rdev->wb.enabled)
rb_cntl |= DMA_RPTR_WRITEBACK_ENABLE;
WREG32(DMA_RB_BASE + reg_offset, ring->gpu_addr >> 8);
/* enable DMA IBs */
ib_cntl = DMA_IB_ENABLE | CMD_VMID_FORCE;
#ifdef __BIG_ENDIAN
ib_cntl |= DMA_IB_SWAP_ENABLE;
#endif
WREG32(DMA_IB_CNTL + reg_offset, ib_cntl);
dma_cntl = RREG32(DMA_CNTL + reg_offset);
dma_cntl &= ~CTXEMPTY_INT_ENABLE;
WREG32(DMA_CNTL + reg_offset, dma_cntl);
ring->wptr = 0;
WREG32(DMA_RB_WPTR + reg_offset, ring->wptr << 2);
ring->rptr = RREG32(DMA_RB_RPTR + reg_offset) >> 2;
WREG32(DMA_RB_CNTL + reg_offset, rb_cntl | DMA_RB_ENABLE);
ring->ready = true;
r = radeon_ring_test(rdev, ring->idx, ring);
if (r) {
ring->ready = false;
return r;
}
}
radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
return 0;
}
/**
* cayman_dma_fini - tear down the async dma engines
*
* @rdev: radeon_device pointer
*
* Stop the async dma engines and free the rings (cayman-SI).
*/
void cayman_dma_fini(struct radeon_device *rdev)
{
cayman_dma_stop(rdev);
radeon_ring_fini(rdev, &rdev->ring[R600_RING_TYPE_DMA_INDEX]);
radeon_ring_fini(rdev, &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX]);
}
static u32 cayman_gpu_check_soft_reset(struct radeon_device *rdev)
u32 cayman_gpu_check_soft_reset(struct radeon_device *rdev)
{
u32 reset_mask = 0;
u32 tmp;
@@ -2041,34 +1853,6 @@ bool cayman_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
return radeon_ring_test_lockup(rdev, ring);
}
/**
* cayman_dma_is_lockup - Check if the DMA engine is locked up
*
* @rdev: radeon_device pointer
* @ring: radeon_ring structure holding ring information
*
* Check if the async DMA engine is locked up.
* Returns true if the engine appears to be locked up, false if not.
*/
bool cayman_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
{
u32 reset_mask = cayman_gpu_check_soft_reset(rdev);
u32 mask;
if (ring->idx == R600_RING_TYPE_DMA_INDEX)
mask = RADEON_RESET_DMA;
else
mask = RADEON_RESET_DMA1;
if (!(reset_mask & mask)) {
radeon_ring_lockup_update(ring);
return false;
}
/* force ring activities */
radeon_ring_force_activity(rdev, ring);
return radeon_ring_test_lockup(rdev, ring);
}
static int cayman_startup(struct radeon_device *rdev)
{
struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
@@ -2079,6 +1863,13 @@ static int cayman_startup(struct radeon_device *rdev)
/* enable aspm */
evergreen_program_aspm(rdev);
/* scratch needs to be initialized before MC */
r = r600_vram_scratch_init(rdev);
if (r)
return r;
evergreen_mc_program(rdev);
if (rdev->flags & RADEON_IS_IGP) {
if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
r = ni_init_microcode(rdev);
@@ -2103,27 +1894,16 @@ static int cayman_startup(struct radeon_device *rdev)
}
}
r = r600_vram_scratch_init(rdev);
if (r)
return r;
evergreen_mc_program(rdev);
r = cayman_pcie_gart_enable(rdev);
if (r)
return r;
cayman_gpu_init(rdev);
r = evergreen_blit_init(rdev);
if (r) {
r600_blit_fini(rdev);
rdev->asic->copy.copy = NULL;
dev_warn(rdev->dev, "failed blitter (%d) falling back to memcpy\n", r);
}
/* allocate rlc buffers */
if (rdev->flags & RADEON_IS_IGP) {
rdev->rlc.reg_list = tn_rlc_save_restore_register_list;
rdev->rlc.reg_list_size = tn_rlc_save_restore_register_list_size;
rdev->rlc.reg_list_size =
(u32)ARRAY_SIZE(tn_rlc_save_restore_register_list);
rdev->rlc.cs_data = cayman_cs_data;
r = sumo_rlc_init(rdev);
if (r) {
@@ -2143,7 +1923,7 @@ static int cayman_startup(struct radeon_device *rdev)
return r;
}
r = rv770_uvd_resume(rdev);
r = uvd_v2_2_resume(rdev);
if (!r) {
r = radeon_fence_driver_start_ring(rdev,
R600_RING_TYPE_UVD_INDEX);
@@ -2194,7 +1974,7 @@ static int cayman_startup(struct radeon_device *rdev)
r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
CP_RB0_RPTR, CP_RB0_WPTR,
0, 0xfffff, RADEON_CP_PACKET2);
RADEON_CP_PACKET2);
if (r)
return r;
@@ -2202,7 +1982,7 @@ static int cayman_startup(struct radeon_device *rdev)
r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
DMA_RB_RPTR + DMA0_REGISTER_OFFSET,
DMA_RB_WPTR + DMA0_REGISTER_OFFSET,
2, 0x3fffc, DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
if (r)
return r;
@@ -2210,7 +1990,7 @@ static int cayman_startup(struct radeon_device *rdev)
r = radeon_ring_init(rdev, ring, ring->ring_size, CAYMAN_WB_DMA1_RPTR_OFFSET,
DMA_RB_RPTR + DMA1_REGISTER_OFFSET,
DMA_RB_WPTR + DMA1_REGISTER_OFFSET,
2, 0x3fffc, DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
if (r)
return r;
@@ -2227,12 +2007,11 @@ static int cayman_startup(struct radeon_device *rdev)
ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
if (ring->ring_size) {
r = radeon_ring_init(rdev, ring, ring->ring_size,
R600_WB_UVD_RPTR_OFFSET,
r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
0, 0xfffff, RADEON_CP_PACKET2);
RADEON_CP_PACKET2);
if (!r)
r = r600_uvd_init(rdev);
r = uvd_v1_0_init(rdev);
if (r)
DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
}
@@ -2249,9 +2028,15 @@ static int cayman_startup(struct radeon_device *rdev)
return r;
}
r = r600_audio_init(rdev);
if (r)
return r;
if (ASIC_IS_DCE6(rdev)) {
r = dce6_audio_init(rdev);
if (r)
return r;
} else {
r = r600_audio_init(rdev);
if (r)
return r;
}
return 0;
}
@@ -2282,11 +2067,14 @@ int cayman_resume(struct radeon_device *rdev)
int cayman_suspend(struct radeon_device *rdev)
{
r600_audio_fini(rdev);
if (ASIC_IS_DCE6(rdev))
dce6_audio_fini(rdev);
else
r600_audio_fini(rdev);
radeon_vm_manager_fini(rdev);
cayman_cp_enable(rdev, false);
cayman_dma_stop(rdev);
r600_uvd_rbc_stop(rdev);
uvd_v1_0_fini(rdev);
radeon_uvd_suspend(rdev);
evergreen_irq_suspend(rdev);
radeon_wb_disable(rdev);
@@ -2408,7 +2196,6 @@ int cayman_init(struct radeon_device *rdev)
void cayman_fini(struct radeon_device *rdev)
{
r600_blit_fini(rdev);
cayman_cp_fini(rdev);
cayman_dma_fini(rdev);
r600_irq_fini(rdev);
@@ -2418,6 +2205,7 @@ void cayman_fini(struct radeon_device *rdev)
radeon_vm_manager_fini(rdev);
radeon_ib_pool_fini(rdev);
radeon_irq_kms_fini(rdev);
uvd_v1_0_fini(rdev);
radeon_uvd_fini(rdev);
cayman_pcie_gart_fini(rdev);
r600_vram_scratch_fini(rdev);
@@ -2678,61 +2466,7 @@ void cayman_vm_set_page(struct radeon_device *rdev,
}
}
} else {
if ((flags & RADEON_VM_PAGE_SYSTEM) ||
(count == 1)) {
while (count) {
ndw = count * 2;
if (ndw > 0xFFFFE)
ndw = 0xFFFFE;
/* for non-physically contiguous pages (system) */
ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_WRITE, 0, 0, ndw);
ib->ptr[ib->length_dw++] = pe;
ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff;
for (; ndw > 0; ndw -= 2, --count, pe += 8) {
if (flags & RADEON_VM_PAGE_SYSTEM) {
value = radeon_vm_map_gart(rdev, addr);
value &= 0xFFFFFFFFFFFFF000ULL;
} else if (flags & RADEON_VM_PAGE_VALID) {
value = addr;
} else {
value = 0;
}
addr += incr;
value |= r600_flags;
ib->ptr[ib->length_dw++] = value;
ib->ptr[ib->length_dw++] = upper_32_bits(value);
}
}
while (ib->length_dw & 0x7)
ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0);
} else {
while (count) {
ndw = count * 2;
if (ndw > 0xFFFFE)
ndw = 0xFFFFE;
if (flags & RADEON_VM_PAGE_VALID)
value = addr;
else
value = 0;
/* for physically contiguous pages (vram) */
ib->ptr[ib->length_dw++] = DMA_PTE_PDE_PACKET(ndw);
ib->ptr[ib->length_dw++] = pe; /* dst addr */
ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff;
ib->ptr[ib->length_dw++] = r600_flags; /* mask */
ib->ptr[ib->length_dw++] = 0;
ib->ptr[ib->length_dw++] = value; /* value */
ib->ptr[ib->length_dw++] = upper_32_bits(value);
ib->ptr[ib->length_dw++] = incr; /* increment size */
ib->ptr[ib->length_dw++] = 0;
pe += ndw * 4;
addr += (ndw / 2) * incr;
count -= ndw / 2;
}
}
while (ib->length_dw & 0x7)
ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0);
cayman_dma_vm_set_page(rdev, ib, pe, addr, count, incr, flags);
}
}
@@ -2766,26 +2500,3 @@ void cayman_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
radeon_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0));
radeon_ring_write(ring, 0x0);
}
void cayman_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
{
struct radeon_ring *ring = &rdev->ring[ridx];
if (vm == NULL)
return;
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0));
radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm->id << 2)) >> 2));
radeon_ring_write(ring, vm->pd_gpu_addr >> 12);
/* flush hdp cache */
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0));
radeon_ring_write(ring, (0xf << 16) | (HDP_MEM_COHERENCY_FLUSH_CNTL >> 2));
radeon_ring_write(ring, 1);
/* bits 0-7 are the VM contexts0-7 */
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0));
radeon_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST >> 2));
radeon_ring_write(ring, 1 << vm->id);
}