From 0eeb68b390377148e31f000db1c533a9a49bf950 Mon Sep 17 00:00:00 2001 From: Christian König Date: Thu, 30 Mar 2017 14:49:50 +0200 Subject: drm/amdgpu: add VMHUB to ring association MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add the info which ring belonging to which VMHUB. Signed-off-by: Christian König Reviewed-by: Alex Deucher Reviewed-by: Andres Rodriguez Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c') diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 9bcf01469282..b5429223618f 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -1448,6 +1448,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = { .align_mask = 0xf, .nop = PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0), .support_64bit_ptrs = false, + .vmhub = AMDGPU_MMHUB, .get_rptr = uvd_v7_0_ring_get_rptr, .get_wptr = uvd_v7_0_ring_get_wptr, .set_wptr = uvd_v7_0_ring_set_wptr, @@ -1475,6 +1476,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = { .align_mask = 0x3f, .nop = HEVC_ENC_CMD_NO_OP, .support_64bit_ptrs = false, + .vmhub = AMDGPU_MMHUB, .get_rptr = uvd_v7_0_enc_ring_get_rptr, .get_wptr = uvd_v7_0_enc_ring_get_wptr, .set_wptr = uvd_v7_0_enc_ring_set_wptr, -- cgit v1.2.3 From 2e81984988adf8de92b7d3a14ba0fe8310d0bcf8 Mon Sep 17 00:00:00 2001 From: Christian König Date: Thu, 30 Mar 2017 16:50:47 +0200 Subject: drm/amdgpu: invalidate only the currently needed VMHUB v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Drop invalidating both hubs from each engine. v2: don't use hardcoded values Signed-off-by: Christian König Reviewed-by: Alex Deucher Reviewed-by: Andres Rodriguez Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 3 +- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 36 +++++------ drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 60 +++++++++--------- drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 111 +++++++++++++++------------------ drivers/gpu/drm/amd/amdgpu/vce_v4_0.c | 57 ++++++++--------- 5 files changed, 119 insertions(+), 148 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 3455b2c8652c..b38a8d7714c7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -406,8 +406,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, struct amdgpu_job *job) { struct amdgpu_device *adev = ring->adev; - /* Temporary use only the first VM manager */ - unsigned vmhub = 0; /*ring->funcs->vmhub;*/ + unsigned vmhub = ring->funcs->vmhub; struct amdgpu_vm_id_manager *id_mgr = &adev->vm_manager.id_mgr[vmhub]; uint64_t fence_context = adev->fence_context + ring->idx; struct dma_fence *updates = sync->last_vm_update; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 4a7fe9727e95..1badd294ff18 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -2956,35 +2956,29 @@ static void gfx_v9_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void gfx_v9_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); unsigned eng = ring->idx; - unsigned i; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->ctx0_ptb_addr_lo32 - + (2 * vm_id), - lower_32_bits(pd_addr)); + gfx_v9_0_write_data_to_reg(ring, usepfp, true, + hub->ctx0_ptb_addr_lo32 + (2 * vm_id), + lower_32_bits(pd_addr)); - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->ctx0_ptb_addr_hi32 - + (2 * vm_id), - upper_32_bits(pd_addr)); + gfx_v9_0_write_data_to_reg(ring, usepfp, true, + hub->ctx0_ptb_addr_hi32 + (2 * vm_id), + upper_32_bits(pd_addr)); - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->vm_inv_eng0_req + eng, req); + gfx_v9_0_write_data_to_reg(ring, usepfp, true, + hub->vm_inv_eng0_req + eng, req); - /* wait for the invalidate to complete */ - gfx_v9_0_wait_reg_mem(ring, 0, 0, 0, hub->vm_inv_eng0_ack + - eng, 0, 1 << vm_id, 1 << vm_id, 0x20); - } + /* wait for the invalidate to complete */ + gfx_v9_0_wait_reg_mem(ring, 0, 0, 0, hub->vm_inv_eng0_ack + + eng, 0, 1 << vm_id, 1 << vm_id, 0x20); /* compute doesn't have PFP */ if (usepfp) { @@ -3463,7 +3457,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = { .emit_frame_size = /* totally 242 maximum if 16 IBs */ 5 + /* COND_EXEC */ 7 + /* PIPELINE_SYNC */ - 46 + /* VM_FLUSH */ + 24 + /* VM_FLUSH */ 8 + /* FENCE for VM_FLUSH */ 20 + /* GDS switch */ 4 + /* double SWITCH_BUFFER, @@ -3510,7 +3504,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = { 7 + /* gfx_v9_0_ring_emit_hdp_flush */ 5 + /* gfx_v9_0_ring_emit_hdp_invalidate */ 7 + /* gfx_v9_0_ring_emit_pipeline_sync */ - 64 + /* gfx_v9_0_ring_emit_vm_flush */ + 24 + /* gfx_v9_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v9_0_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */ .emit_ib = gfx_v9_0_ring_emit_ib_compute, @@ -3540,7 +3534,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = { 7 + /* gfx_v9_0_ring_emit_hdp_flush */ 5 + /* gfx_v9_0_ring_emit_hdp_invalidate */ 7 + /* gfx_v9_0_ring_emit_pipeline_sync */ - 64 + /* gfx_v9_0_ring_emit_vm_flush */ + 24 + /* gfx_v9_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v9_0_ring_emit_fence_kiq x3 for user fence, vm fence */ .emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */ .emit_ib = gfx_v9_0_ring_emit_ib_compute, diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 7ba3e85a96c6..c3cf6bcadc60 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -1039,44 +1039,40 @@ static void sdma_v4_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void sdma_v4_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); unsigned eng = ring->idx; - unsigned i; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->ctx0_ptb_addr_lo32 + vm_id * 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->ctx0_ptb_addr_hi32 + vm_id * 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->vm_inv_eng0_req + eng); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | - SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) | - SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* equal */ - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 1 << vm_id); /* reference */ - amdgpu_ring_write(ring, 1 << vm_id); /* mask */ - amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | - SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); - } + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | + SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); + amdgpu_ring_write(ring, hub->ctx0_ptb_addr_lo32 + vm_id * 2); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | + SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); + amdgpu_ring_write(ring, hub->ctx0_ptb_addr_hi32 + vm_id * 2); + amdgpu_ring_write(ring, upper_32_bits(pd_addr)); + + /* flush TLB */ + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | + SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); + amdgpu_ring_write(ring, hub->vm_inv_eng0_req + eng); + amdgpu_ring_write(ring, req); + + /* wait for flush */ + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | + SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) | + SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* equal */ + amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 1 << vm_id); /* reference */ + amdgpu_ring_write(ring, 1 << vm_id); /* mask */ + amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | + SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); } static int sdma_v4_0_early_init(void *handle) @@ -1481,7 +1477,7 @@ static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs = { 6 + /* sdma_v4_0_ring_emit_hdp_flush */ 3 + /* sdma_v4_0_ring_emit_hdp_invalidate */ 6 + /* sdma_v4_0_ring_emit_pipeline_sync */ - 36 + /* sdma_v4_0_ring_emit_vm_flush */ + 18 + /* sdma_v4_0_ring_emit_vm_flush */ 10 + 10 + 10, /* sdma_v4_0_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 7 + 6, /* sdma_v4_0_ring_emit_ib */ .emit_ib = sdma_v4_0_ring_emit_ib, diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index b5429223618f..dc3a2145a452 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -1034,42 +1034,38 @@ static void uvd_v7_0_vm_reg_wait(struct amdgpu_ring *ring, static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); uint32_t data0, data1, mask; unsigned eng = ring->idx; - unsigned i; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - data0 = (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2; - data1 = upper_32_bits(pd_addr); - uvd_v7_0_vm_reg_write(ring, data0, data1); - - data0 = (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2; - data1 = lower_32_bits(pd_addr); - uvd_v7_0_vm_reg_write(ring, data0, data1); - - data0 = (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2; - data1 = lower_32_bits(pd_addr); - mask = 0xffffffff; - uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); - - /* flush TLB */ - data0 = (hub->vm_inv_eng0_req + eng) << 2; - data1 = req; - uvd_v7_0_vm_reg_write(ring, data0, data1); - - /* wait for flush */ - data0 = (hub->vm_inv_eng0_ack + eng) << 2; - data1 = 1 << vm_id; - mask = 1 << vm_id; - uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); - } + data0 = (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2; + data1 = upper_32_bits(pd_addr); + uvd_v7_0_vm_reg_write(ring, data0, data1); + + data0 = (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2; + data1 = lower_32_bits(pd_addr); + uvd_v7_0_vm_reg_write(ring, data0, data1); + + data0 = (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2; + data1 = lower_32_bits(pd_addr); + mask = 0xffffffff; + uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); + + /* flush TLB */ + data0 = (hub->vm_inv_eng0_req + eng) << 2; + data1 = req; + uvd_v7_0_vm_reg_write(ring, data0, data1); + + /* wait for flush */ + data0 = (hub->vm_inv_eng0_ack + eng) << 2; + data1 = 1 << vm_id; + mask = 1 << vm_id; + uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); } static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring) @@ -1080,44 +1076,37 @@ static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring) static void uvd_v7_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned int vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); unsigned eng = ring->idx; - unsigned i; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, 0xffffffff); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 1 << vm_id); - amdgpu_ring_write(ring, 1 << vm_id); - } + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, upper_32_bits(pd_addr)); + + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, 0xffffffff); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + /* flush TLB */ + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); + amdgpu_ring_write(ring, req); + + /* wait for flush */ + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); + amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); + amdgpu_ring_write(ring, 1 << vm_id); + amdgpu_ring_write(ring, 1 << vm_id); } #if 0 @@ -1455,7 +1444,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = { .emit_frame_size = 2 + /* uvd_v7_0_ring_emit_hdp_flush */ 2 + /* uvd_v7_0_ring_emit_hdp_invalidate */ - 34 * AMDGPU_MAX_VMHUBS + /* uvd_v7_0_ring_emit_vm_flush */ + 34 + /* uvd_v7_0_ring_emit_vm_flush */ 14 + 14, /* uvd_v7_0_ring_emit_fence x2 vm fence */ .emit_ib_size = 8, /* uvd_v7_0_ring_emit_ib */ .emit_ib = uvd_v7_0_ring_emit_ib, @@ -1481,7 +1470,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = { .get_wptr = uvd_v7_0_enc_ring_get_wptr, .set_wptr = uvd_v7_0_enc_ring_set_wptr, .emit_frame_size = - 17 * AMDGPU_MAX_VMHUBS + /* uvd_v7_0_enc_ring_emit_vm_flush */ + 17 + /* uvd_v7_0_enc_ring_emit_vm_flush */ 5 + 5 + /* uvd_v7_0_enc_ring_emit_fence x2 vm fence */ 1, /* uvd_v7_0_enc_ring_insert_end */ .emit_ib_size = 5, /* uvd_v7_0_enc_ring_emit_ib */ diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c index f68ab3b24c31..2ffafbe6cd80 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c @@ -968,44 +968,37 @@ static void vce_v4_0_ring_insert_end(struct amdgpu_ring *ring) static void vce_v4_0_emit_vm_flush(struct amdgpu_ring *ring, unsigned int vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); unsigned eng = ring->idx; - unsigned i; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, 0xffffffff); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 1 << vm_id); - amdgpu_ring_write(ring, 1 << vm_id); - } + amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, upper_32_bits(pd_addr)); + + amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, 0xffffffff); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + /* flush TLB */ + amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); + amdgpu_ring_write(ring, req); + + /* wait for flush */ + amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); + amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); + amdgpu_ring_write(ring, 1 << vm_id); + amdgpu_ring_write(ring, 1 << vm_id); } static int vce_v4_0_set_interrupt_state(struct amdgpu_device *adev, @@ -1079,7 +1072,7 @@ static const struct amdgpu_ring_funcs vce_v4_0_ring_vm_funcs = { .set_wptr = vce_v4_0_ring_set_wptr, .parse_cs = amdgpu_vce_ring_parse_cs_vm, .emit_frame_size = - 17 * AMDGPU_MAX_VMHUBS + /* vce_v4_0_emit_vm_flush */ + 17 + /* vce_v4_0_emit_vm_flush */ 5 + 5 + /* amdgpu_vce_ring_emit_fence x2 vm fence */ 1, /* vce_v4_0_ring_insert_end */ .emit_ib_size = 5, /* vce_v4_0_ring_emit_ib */ -- cgit v1.2.3 From 4789c463cb04d725170406873caaec5208c99eb9 Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 31 Mar 2017 11:03:50 +0200 Subject: drm/amdgpu: assign VM invalidation engine manually v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For Vega10 we have 18 VM invalidation engines for each VMHUB. Start to assign them manually to the rings. v2: add a BUG_ON if we use to many engines Signed-off-by: Christian König Reviewed-by: Alex Deucher Reviewed-by: Andres Rodriguez Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h | 1 + drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 16 ++++++++++++++++ drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 4 ++-- drivers/gpu/drm/amd/amdgpu/vce_v4_0.c | 2 +- 6 files changed, 22 insertions(+), 5 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index de12f7549e6d..944443c5b90a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -179,6 +179,7 @@ struct amdgpu_ring { unsigned cond_exe_offs; u64 cond_exe_gpu_addr; volatile u32 *cond_exe_cpu_addr; + unsigned vm_inv_eng; #if defined(CONFIG_DEBUG_FS) struct dentry *ent; #endif diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 1badd294ff18..87596e48f65d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -2959,7 +2959,7 @@ static void gfx_v9_0_ring_emit_vm_flush(struct amdgpu_ring *ring, struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); - unsigned eng = ring->idx; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 7f5cc75f0da5..a71521e10763 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -386,6 +386,22 @@ static int gmc_v9_0_early_init(void *handle) static int gmc_v9_0_late_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + unsigned vm_inv_eng[AMDGPU_MAX_VMHUBS] = { 0 }; + unsigned i; + + for(i = 0; i < adev->num_rings; ++i) { + struct amdgpu_ring *ring = adev->rings[i]; + unsigned vmhub = ring->funcs->vmhub; + + ring->vm_inv_eng = vm_inv_eng[vmhub]++; + dev_info(adev->dev, "ring %u uses VM inv eng %u on hub %u\n", + ring->idx, ring->vm_inv_eng, ring->funcs->vmhub); + } + + /* Engine 17 is used for GART flushes */ + for(i = 0; i < AMDGPU_MAX_VMHUBS; ++i) + BUG_ON(vm_inv_eng[i] > 17); + return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0); } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index c3cf6bcadc60..88c5d8f6f414 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -1041,7 +1041,7 @@ static void sdma_v4_0_ring_emit_vm_flush(struct amdgpu_ring *ring, { struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); - unsigned eng = ring->idx; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index dc3a2145a452..c646570d1421 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -1037,7 +1037,7 @@ static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring, struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); uint32_t data0, data1, mask; - unsigned eng = ring->idx; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ @@ -1078,7 +1078,7 @@ static void uvd_v7_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring, { struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); - unsigned eng = ring->idx; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c index 2ffafbe6cd80..72d0edb9bc39 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c @@ -970,7 +970,7 @@ static void vce_v4_0_emit_vm_flush(struct amdgpu_ring *ring, { struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); - unsigned eng = ring->idx; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ -- cgit v1.2.3 From 247ac95141740157bd86d29819ff46e7ab396113 Mon Sep 17 00:00:00 2001 From: Frank Min Date: Mon, 17 Apr 2017 11:28:12 +0800 Subject: drm/amdgpu/uvd7: add sriov uvd initialization sequences Add UVD initialization for SRIOV. Signed-off-by: Frank Min Signed-off-by: Xiangliang Yu Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 246 ++++++++++++++++++++++++++++++++++ 1 file changed, 246 insertions(+) (limited to 'drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c') diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index c646570d1421..46c7bc490809 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -27,10 +27,14 @@ #include "amdgpu_uvd.h" #include "soc15d.h" #include "soc15_common.h" +#include "mmsch_v1_0.h" #include "vega10/soc15ip.h" #include "vega10/UVD/uvd_7_0_offset.h" #include "vega10/UVD/uvd_7_0_sh_mask.h" +#include "vega10/VCE/vce_4_0_offset.h" +#include "vega10/VCE/vce_4_0_default.h" +#include "vega10/VCE/vce_4_0_sh_mask.h" #include "vega10/NBIF/nbif_6_1_offset.h" #include "vega10/HDP/hdp_4_0_offset.h" #include "vega10/MMHUB/mmhub_1_0_offset.h" @@ -41,6 +45,7 @@ static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev); static void uvd_v7_0_set_irq_funcs(struct amdgpu_device *adev); static int uvd_v7_0_start(struct amdgpu_device *adev); static void uvd_v7_0_stop(struct amdgpu_device *adev); +static int uvd_v7_0_sriov_start(struct amdgpu_device *adev); /** * uvd_v7_0_ring_get_rptr - get read pointer @@ -618,6 +623,247 @@ static void uvd_v7_0_mc_resume(struct amdgpu_device *adev) WREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH4), adev->uvd.max_handles); } +static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev, + struct amdgpu_mm_table *table) +{ + uint32_t data = 0, loop; + uint64_t addr = table->gpu_addr; + struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)table->cpu_addr; + uint32_t size; + + size = header->header_size + header->vce_table_size + header->uvd_table_size; + + /* 1, write to vce_mmsch_vf_ctx_addr_lo/hi register with GPU mc addr of memory descriptor location */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_CTX_ADDR_LO), lower_32_bits(addr)); + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_CTX_ADDR_HI), upper_32_bits(addr)); + + /* 2, update vmid of descriptor */ + data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_VMID)); + data &= ~VCE_MMSCH_VF_VMID__VF_CTX_VMID_MASK; + data |= (0 << VCE_MMSCH_VF_VMID__VF_CTX_VMID__SHIFT); /* use domain0 for MM scheduler */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_VMID), data); + + /* 3, notify mmsch about the size of this descriptor */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_CTX_SIZE), size); + + /* 4, set resp to zero */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP), 0); + + /* 5, kick off the initialization and wait until VCE_MMSCH_VF_MAILBOX_RESP becomes non-zero */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_HOST), 0x10000001); + + data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP)); + loop = 1000; + while ((data & 0x10000002) != 0x10000002) { + udelay(10); + data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP)); + loop--; + if (!loop) + break; + } + + if (!loop) { + dev_err(adev->dev, "failed to init MMSCH, mmVCE_MMSCH_VF_MAILBOX_RESP = %x\n", data); + return -EBUSY; + } + + return 0; +} + +static int uvd_v7_0_sriov_start(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring; + uint32_t offset, size, tmp; + uint32_t table_size = 0; + struct mmsch_v1_0_cmd_direct_write direct_wt = { {0} }; + struct mmsch_v1_0_cmd_direct_read_modify_write direct_rd_mod_wt = { {0} }; + struct mmsch_v1_0_cmd_direct_polling direct_poll = { {0} }; + //struct mmsch_v1_0_cmd_indirect_write indirect_wt = {{0}}; + struct mmsch_v1_0_cmd_end end = { {0} }; + uint32_t *init_table = adev->virt.mm_table.cpu_addr; + struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)init_table; + + direct_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_WRITE; + direct_rd_mod_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_READ_MODIFY_WRITE; + direct_poll.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_POLLING; + end.cmd_header.command_type = MMSCH_COMMAND__END; + + if (header->uvd_table_offset == 0 && header->uvd_table_size == 0) { + header->version = MMSCH_VERSION; + header->header_size = sizeof(struct mmsch_v1_0_init_header) >> 2; + + if (header->vce_table_offset == 0 && header->vce_table_size == 0) + header->uvd_table_offset = header->header_size; + else + header->uvd_table_offset = header->vce_table_size + header->vce_table_offset; + + init_table += header->uvd_table_offset; + + ring = &adev->uvd.ring; + size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4); + + /* disable clock gating */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_POWER_STATUS), + ~UVD_POWER_STATUS__UVD_PG_MODE_MASK, 0); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), + 0xFFFFFFFF, 0x00000004); + /* mc resume*/ + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), + lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), + upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr)); + offset = 0; + } else { + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), + lower_32_bits(adev->uvd.gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), + upper_32_bits(adev->uvd.gpu_addr)); + offset = size; + } + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0), + AMDGPU_UVD_FIRMWARE_OFFSET >> 3); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE0), size); + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), + lower_32_bits(adev->uvd.gpu_addr + offset)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), + upper_32_bits(adev->uvd.gpu_addr + offset)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), (1 << 21)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_UVD_HEAP_SIZE); + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW), + lower_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH), + upper_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2), (2 << 21)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE2), + AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40)); + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_UDEC_ADDR_CONFIG), + adev->gfx.config.gb_addr_config); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_UDEC_DB_ADDR_CONFIG), + adev->gfx.config.gb_addr_config); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_UDEC_DBW_ADDR_CONFIG), + adev->gfx.config.gb_addr_config); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH4), adev->uvd.max_handles); + /* mc resume end*/ + + /* disable clock gating */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_CGC_CTRL), + ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK, 0); + + /* disable interupt */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN), + ~UVD_MASTINT_EN__VCPU_EN_MASK, 0); + + /* stall UMC and register bus before resetting VCPU */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), + ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, + UVD_LMI_CTRL2__STALL_ARB_UMC_MASK); + + /* put LMI, VCPU, RBC etc... into reset */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), + (uint32_t)(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)); + + /* initialize UVD memory controller */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL), + (uint32_t)((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 | + 0x00100000L)); + + /* disable byte swapping */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_SWAP_CNTL), 0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MP_SWAP_CNTL), 0); + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUXA0), 0x40c2040); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUXA1), 0x0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUXB0), 0x40c2040); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUXB1), 0x0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_ALU), 0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUX), 0x88); + + /* take all subblocks out of reset, except VCPU */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), + UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK); + + /* enable VCPU clock */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CNTL), + UVD_VCPU_CNTL__CLK_EN_MASK); + + /* enable UMC */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), + ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, 0); + + /* boot up the VCPU */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), 0); + + MMSCH_V1_0_INSERT_DIRECT_POLL(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), 0x02, 0x02); + + /* enable master interrupt */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, 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 */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), + ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT), 0); + + /* force RBC into idle state */ + size = order_base_2(ring->ring_size); + tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, size); + 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); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_CNTL), tmp); + + /* set the write pointer delay */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_WPTR_CNTL), 0); + + /* set the wb address */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_RPTR_ADDR), + (upper_32_bits(ring->gpu_addr) >> 2)); + + /* programm the RB_BASE for ring buffer */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW), + lower_32_bits(ring->gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH), + upper_32_bits(ring->gpu_addr)); + + ring->wptr = 0; + ring = &adev->uvd.ring_enc[0]; + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_LO), ring->gpu_addr); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_SIZE), ring->ring_size / 4); + + ring = &adev->uvd.ring_enc[1]; + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_LO2), ring->gpu_addr); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_HI2), upper_32_bits(ring->gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_SIZE2), ring->ring_size / 4); + + /* add end packet */ + memcpy((void *)init_table, &end, sizeof(struct mmsch_v1_0_cmd_end)); + table_size += sizeof(struct mmsch_v1_0_cmd_end) / 4; + header->uvd_table_size = table_size; + + return uvd_v7_0_mmsch_start(adev, &adev->virt.mm_table); + } + return -EINVAL; /* already initializaed ? */ +} + /** * uvd_v7_0_start - start UVD block * -- cgit v1.2.3 From beb2ced51b70a6cbccd8676b450e282fecf65565 Mon Sep 17 00:00:00 2001 From: Frank Min Date: Mon, 17 Apr 2017 11:45:35 +0800 Subject: drm/amdgpu/uvd7: add uvd doorbell initialization for sriov Add UVD doorbell for SRIOV. Signed-off-by: Frank Min Signed-off-by: Xiangliang Yu Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) (limited to 'drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c') diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 46c7bc490809..552bfcdd3236 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -103,6 +103,9 @@ static uint64_t uvd_v7_0_enc_ring_get_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; + if (ring->use_doorbell) + return adev->wb.wb[ring->wptr_offs]; + if (ring == &adev->uvd.ring_enc[0]) return RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_WPTR)); else @@ -134,6 +137,13 @@ static void uvd_v7_0_enc_ring_set_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; + if (ring->use_doorbell) { + /* XXX check if swapping is necessary on BE */ + adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr); + WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr)); + return; + } + if (ring == &adev->uvd.ring_enc[0]) WREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_WPTR), lower_32_bits(ring->wptr)); @@ -421,6 +431,10 @@ static int uvd_v7_0_sw_init(void *handle) for (i = 0; i < adev->uvd.num_enc_rings; ++i) { ring = &adev->uvd.ring_enc[i]; sprintf(ring->name, "uvd_enc%d", i); + if (amdgpu_sriov_vf(adev)) { + ring->use_doorbell = true; + ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING0_1 * 2; + } r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0); if (r) return r; -- cgit v1.2.3 From 6fa336a777712fac9da763d3df348fcfc96633f8 Mon Sep 17 00:00:00 2001 From: Frank Min Date: Mon, 17 Apr 2017 11:51:44 +0800 Subject: drm/amdgpu/uvd7: add UVD hw init sequences for sriov Add UVD hw init. Signed-off-by: Frank Min Signed-off-by: Xiangliang Yu Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 101 +++++++++++++++++++--------------- 1 file changed, 56 insertions(+), 45 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c') diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 552bfcdd3236..eca8f6e01e97 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -368,7 +368,10 @@ static int uvd_v7_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - adev->uvd.num_enc_rings = 2; + if (amdgpu_sriov_vf(adev)) + adev->uvd.num_enc_rings = 1; + else + adev->uvd.num_enc_rings = 2; uvd_v7_0_set_ring_funcs(adev); uvd_v7_0_set_enc_ring_funcs(adev); uvd_v7_0_set_irq_funcs(adev); @@ -421,12 +424,14 @@ static int uvd_v7_0_sw_init(void *handle) r = amdgpu_uvd_resume(adev); if (r) return r; + if (!amdgpu_sriov_vf(adev)) { + ring = &adev->uvd.ring; + sprintf(ring->name, "uvd"); + r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0); + if (r) + return r; + } - ring = &adev->uvd.ring; - sprintf(ring->name, "uvd"); - r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0); - if (r) - return r; for (i = 0; i < adev->uvd.num_enc_rings; ++i) { ring = &adev->uvd.ring_enc[i]; @@ -440,6 +445,10 @@ static int uvd_v7_0_sw_init(void *handle) return r; } + r = amdgpu_virt_alloc_mm_table(adev); + if (r) + return r; + return r; } @@ -448,6 +457,8 @@ static int uvd_v7_0_sw_fini(void *handle) int i, r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + amdgpu_virt_free_mm_table(adev); + r = amdgpu_uvd_suspend(adev); if (r) return r; @@ -474,48 +485,53 @@ static int uvd_v7_0_hw_init(void *handle) uint32_t tmp; int i, r; - r = uvd_v7_0_start(adev); + if (amdgpu_sriov_vf(adev)) + r = uvd_v7_0_sriov_start(adev); + else + r = uvd_v7_0_start(adev); if (r) goto done; - ring->ready = true; - r = amdgpu_ring_test_ring(ring); - if (r) { - ring->ready = false; - goto done; - } + if (!amdgpu_sriov_vf(adev)) { + ring->ready = true; + r = amdgpu_ring_test_ring(ring); + if (r) { + ring->ready = false; + goto done; + } - r = amdgpu_ring_alloc(ring, 10); - if (r) { - DRM_ERROR("amdgpu: ring failed to lock UVD ring (%d).\n", r); - goto done; - } + r = amdgpu_ring_alloc(ring, 10); + if (r) { + DRM_ERROR("amdgpu: ring failed to lock UVD ring (%d).\n", r); + goto done; + } - tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL), 0); - amdgpu_ring_write(ring, tmp); - amdgpu_ring_write(ring, 0xFFFFF); + tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL), 0); + amdgpu_ring_write(ring, tmp); + amdgpu_ring_write(ring, 0xFFFFF); - tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL), 0); - amdgpu_ring_write(ring, tmp); - amdgpu_ring_write(ring, 0xFFFFF); + tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL), 0); + amdgpu_ring_write(ring, tmp); + amdgpu_ring_write(ring, 0xFFFFF); - tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL), 0); - amdgpu_ring_write(ring, tmp); - amdgpu_ring_write(ring, 0xFFFFF); + tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL), 0); + amdgpu_ring_write(ring, tmp); + amdgpu_ring_write(ring, 0xFFFFF); - /* Clear timeout status bits */ - amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_TIMEOUT_STATUS), 0)); - amdgpu_ring_write(ring, 0x8); + /* Clear timeout status bits */ + amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_TIMEOUT_STATUS), 0)); + amdgpu_ring_write(ring, 0x8); - amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_CNTL), 0)); - amdgpu_ring_write(ring, 3); + amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_CNTL), 0)); + amdgpu_ring_write(ring, 3); - amdgpu_ring_commit(ring); + amdgpu_ring_commit(ring); + } for (i = 0; i < adev->uvd.num_enc_rings; ++i) { ring = &adev->uvd.ring_enc[i]; @@ -692,7 +708,6 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev) struct mmsch_v1_0_cmd_direct_write direct_wt = { {0} }; struct mmsch_v1_0_cmd_direct_read_modify_write direct_rd_mod_wt = { {0} }; struct mmsch_v1_0_cmd_direct_polling direct_poll = { {0} }; - //struct mmsch_v1_0_cmd_indirect_write indirect_wt = {{0}}; struct mmsch_v1_0_cmd_end end = { {0} }; uint32_t *init_table = adev->virt.mm_table.cpu_addr; struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)init_table; @@ -863,11 +878,6 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev) MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr)); MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_SIZE), ring->ring_size / 4); - ring = &adev->uvd.ring_enc[1]; - MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_LO2), ring->gpu_addr); - MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_HI2), upper_32_bits(ring->gpu_addr)); - MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_SIZE2), ring->ring_size / 4); - /* add end packet */ memcpy((void *)init_table, &end, sizeof(struct mmsch_v1_0_cmd_end)); table_size += sizeof(struct mmsch_v1_0_cmd_end) / 4; @@ -1489,7 +1499,8 @@ static int uvd_v7_0_process_interrupt(struct amdgpu_device *adev, amdgpu_fence_process(&adev->uvd.ring_enc[0]); break; case 120: - amdgpu_fence_process(&adev->uvd.ring_enc[1]); + if (!amdgpu_sriov_vf(adev)) + amdgpu_fence_process(&adev->uvd.ring_enc[1]); break; default: DRM_ERROR("Unhandled interrupt: %d %d\n", -- cgit v1.2.3