diff options
Diffstat (limited to 'drivers/gpu/drm/amd')
20 files changed, 265 insertions, 122 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c index 9b384a94d2f3..3e35a8f2c5e5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c @@ -574,6 +574,7 @@ static const struct amdgpu_px_quirk amdgpu_px_quirk_list[] = { { 0x1002, 0x6900, 0x1002, 0x0124, AMDGPU_PX_QUIRK_FORCE_ATPX }, { 0x1002, 0x6900, 0x1028, 0x0812, AMDGPU_PX_QUIRK_FORCE_ATPX }, { 0x1002, 0x6900, 0x1028, 0x0813, AMDGPU_PX_QUIRK_FORCE_ATPX }, + { 0x1002, 0x699f, 0x1028, 0x0814, AMDGPU_PX_QUIRK_FORCE_ATPX }, { 0x1002, 0x6900, 0x1025, 0x125A, AMDGPU_PX_QUIRK_FORCE_ATPX }, { 0x1002, 0x6900, 0x17AA, 0x3806, AMDGPU_PX_QUIRK_FORCE_ATPX }, { 0, 0, 0, 0, 0 }, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 4e4094f842e7..8b26c970a3cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -1143,6 +1143,9 @@ static int amdgpu_cs_process_syncobj_out_dep(struct amdgpu_cs_parser *p, num_deps = chunk->length_dw * 4 / sizeof(struct drm_amdgpu_cs_chunk_sem); + if (p->post_deps) + return -EINVAL; + p->post_deps = kmalloc_array(num_deps, sizeof(*p->post_deps), GFP_KERNEL); p->num_post_deps = 0; @@ -1166,8 +1169,7 @@ static int amdgpu_cs_process_syncobj_out_dep(struct amdgpu_cs_parser *p, static int amdgpu_cs_process_syncobj_timeline_out_dep(struct amdgpu_cs_parser *p, - struct amdgpu_cs_chunk - *chunk) + struct amdgpu_cs_chunk *chunk) { struct drm_amdgpu_cs_chunk_syncobj *syncobj_deps; unsigned num_deps; @@ -1177,6 +1179,9 @@ static int amdgpu_cs_process_syncobj_timeline_out_dep(struct amdgpu_cs_parser *p num_deps = chunk->length_dw * 4 / sizeof(struct drm_amdgpu_cs_chunk_syncobj); + if (p->post_deps) + return -EINVAL; + p->post_deps = kmalloc_array(num_deps, sizeof(*p->post_deps), GFP_KERNEL); p->num_post_deps = 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c index f539a2a92774..7398b4850649 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c @@ -534,21 +534,24 @@ int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, struct drm_sched_entity *entity) { struct amdgpu_ctx_entity *centity = to_amdgpu_ctx_entity(entity); - unsigned idx = centity->sequence & (amdgpu_sched_jobs - 1); - struct dma_fence *other = centity->fences[idx]; + struct dma_fence *other; + unsigned idx; + long r; - if (other) { - signed long r; - r = dma_fence_wait(other, true); - if (r < 0) { - if (r != -ERESTARTSYS) - DRM_ERROR("Error (%ld) waiting for fence!\n", r); + spin_lock(&ctx->ring_lock); + idx = centity->sequence & (amdgpu_sched_jobs - 1); + other = dma_fence_get(centity->fences[idx]); + spin_unlock(&ctx->ring_lock); - return r; - } - } + if (!other) + return 0; - return 0; + r = dma_fence_wait(other, true); + if (r < 0 && r != -ERESTARTSYS) + DRM_ERROR("Error (%ld) waiting for fence!\n", r); + + dma_fence_put(other); + return r; } void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h index df8a23554831..f6ac1e9548f2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h @@ -32,7 +32,6 @@ struct amdgpu_gds { uint32_t gws_size; uint32_t oa_size; uint32_t gds_compute_max_wave_id; - uint32_t vgt_gs_max_wave_id; }; struct amdgpu_gds_reg_offset { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h index 99f14fcc1460..19661c645703 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h @@ -30,6 +30,7 @@ #define AMDGPU_VCN_FIRMWARE_OFFSET 256 #define AMDGPU_VCN_MAX_ENC_RINGS 3 +#define VCN_DEC_KMD_CMD 0x80000000 #define VCN_DEC_CMD_FENCE 0x00000000 #define VCN_DEC_CMD_TRAP 0x00000001 #define VCN_DEC_CMD_WRITE_REG 0x00000004 diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index 32773b7523d2..f41287f9000d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -4206,15 +4206,6 @@ static void gfx_v10_0_ring_emit_ib_gfx(struct amdgpu_ring *ring, unsigned vmid = AMDGPU_JOB_GET_VMID(job); u32 header, control = 0; - /* Prevent a hw deadlock due to a wave ID mismatch between ME and GDS. - * This resets the wave ID counters. (needed by transform feedback) - * TODO: This might only be needed on a VMID switch when we change - * the GDS OA mapping, not sure. - */ - amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); - amdgpu_ring_write(ring, mmVGT_GS_MAX_WAVE_ID); - amdgpu_ring_write(ring, ring->adev->gds.vgt_gs_max_wave_id); - if (ib->flags & AMDGPU_IB_FLAG_CE) header = PACKET3(PACKET3_INDIRECT_BUFFER_CNST, 2); else @@ -4961,7 +4952,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = { 5 + /* HDP_INVL */ 8 + 8 + /* FENCE x2 */ 2, /* SWITCH_BUFFER */ - .emit_ib_size = 7, /* gfx_v10_0_ring_emit_ib_gfx */ + .emit_ib_size = 4, /* gfx_v10_0_ring_emit_ib_gfx */ .emit_ib = gfx_v10_0_ring_emit_ib_gfx, .emit_fence = gfx_v10_0_ring_emit_fence, .emit_pipeline_sync = gfx_v10_0_ring_emit_pipeline_sync, @@ -5112,7 +5103,6 @@ static void gfx_v10_0_set_gds_init(struct amdgpu_device *adev) default: adev->gds.gds_size = 0x10000; adev->gds.gds_compute_max_wave_id = 0x4ff; - adev->gds.vgt_gs_max_wave_id = 0x3ff; break; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 751567f78567..ee1ccdcf2d30 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -1321,6 +1321,39 @@ static int gfx_v8_0_rlc_init(struct amdgpu_device *adev) return 0; } +static int gfx_v8_0_csb_vram_pin(struct amdgpu_device *adev) +{ + int r; + + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false); + if (unlikely(r != 0)) + return r; + + r = amdgpu_bo_pin(adev->gfx.rlc.clear_state_obj, + AMDGPU_GEM_DOMAIN_VRAM); + if (!r) + adev->gfx.rlc.clear_state_gpu_addr = + amdgpu_bo_gpu_offset(adev->gfx.rlc.clear_state_obj); + + amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); + + return r; +} + +static void gfx_v8_0_csb_vram_unpin(struct amdgpu_device *adev) +{ + int r; + + if (!adev->gfx.rlc.clear_state_obj) + return; + + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, true); + if (likely(r == 0)) { + amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); + } +} + static void gfx_v8_0_mec_fini(struct amdgpu_device *adev) { amdgpu_bo_free_kernel(&adev->gfx.mec.hpd_eop_obj, NULL, NULL); @@ -4785,6 +4818,10 @@ static int gfx_v8_0_hw_init(void *handle) gfx_v8_0_init_golden_registers(adev); gfx_v8_0_constants_init(adev); + r = gfx_v8_0_csb_vram_pin(adev); + if (r) + return r; + r = adev->gfx.rlc.funcs->resume(adev); if (r) return r; @@ -4901,6 +4938,9 @@ static int gfx_v8_0_hw_fini(void *handle) else pr_err("rlc is busy, skip halt rlc\n"); amdgpu_gfx_rlc_exit_safe_mode(adev); + + gfx_v8_0_csb_vram_unpin(adev); + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 1cf639a51178..c066e1d3f981 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -596,14 +596,18 @@ static void gfx_v9_0_check_if_need_gfxoff(struct amdgpu_device *adev) case CHIP_VEGA20: break; case CHIP_RAVEN: - if (adev->rev_id >= 0x8 || adev->pdev->device == 0x15d8) - break; - if ((adev->gfx.rlc_fw_version != 106 && - adev->gfx.rlc_fw_version < 531) || - (adev->gfx.rlc_fw_version == 53815) || - (adev->gfx.rlc_feature_version < 1) || - !adev->gfx.rlc.is_rlc_v2_1) + if (!(adev->rev_id >= 0x8 || adev->pdev->device == 0x15d8) + &&((adev->gfx.rlc_fw_version != 106 && + adev->gfx.rlc_fw_version < 531) || + (adev->gfx.rlc_fw_version == 53815) || + (adev->gfx.rlc_feature_version < 1) || + !adev->gfx.rlc.is_rlc_v2_1)) adev->pm.pp_feature &= ~PP_GFXOFF_MASK; + + if (adev->pm.pp_feature & PP_GFXOFF_MASK) + adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG | + AMD_PG_SUPPORT_CP | + AMD_PG_SUPPORT_RLC_SMU_HS; break; default: break; @@ -4869,7 +4873,7 @@ static void gfx_v9_0_ring_soft_recovery(struct amdgpu_ring *ring, unsigned vmid) value = REG_SET_FIELD(value, SQ_CMD, MODE, 0x01); value = REG_SET_FIELD(value, SQ_CMD, CHECK_VMID, 1); value = REG_SET_FIELD(value, SQ_CMD, VM_ID, vmid); - WREG32(mmSQ_CMD, value); + WREG32_SOC15(GC, 0, mmSQ_CMD, value); } static void gfx_v9_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index 662612f89c70..9922bce3fd89 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -552,7 +552,6 @@ static int nv_common_early_init(void *handle) AMD_CG_SUPPORT_BIF_LS; adev->pg_flags = AMD_PG_SUPPORT_VCN | AMD_PG_SUPPORT_VCN_DPG | - AMD_PG_SUPPORT_MMHUB | AMD_PG_SUPPORT_ATHUB; adev->external_rev_id = adev->rev_id + 0x1; break; diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 23265414d448..04fbf05d7176 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -992,11 +992,6 @@ static int soc15_common_early_init(void *handle) adev->pg_flags = AMD_PG_SUPPORT_SDMA | AMD_PG_SUPPORT_VCN; } - - if (adev->pm.pp_feature & PP_GFXOFF_MASK) - adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG | - AMD_PG_SUPPORT_CP | - AMD_PG_SUPPORT_RLC_SMU_HS; break; default: /* FIXME: not supported yet */ diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c index 1cfc2620b2dd..dfde886cc6bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c @@ -1485,7 +1485,7 @@ static void vcn_v2_0_dec_ring_insert_start(struct amdgpu_ring *ring) amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0_INTERNAL_OFFSET, 0)); amdgpu_ring_write(ring, 0); amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0)); - amdgpu_ring_write(ring, VCN_DEC_CMD_PACKET_START << 1); + amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_PACKET_START << 1)); } /** @@ -1498,7 +1498,7 @@ static void vcn_v2_0_dec_ring_insert_start(struct amdgpu_ring *ring) static void vcn_v2_0_dec_ring_insert_end(struct amdgpu_ring *ring) { amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0)); - amdgpu_ring_write(ring, VCN_DEC_CMD_PACKET_END << 1); + amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_PACKET_END << 1)); } /** @@ -1543,7 +1543,7 @@ static void vcn_v2_0_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 amdgpu_ring_write(ring, upper_32_bits(addr) & 0xff); amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0)); - amdgpu_ring_write(ring, VCN_DEC_CMD_FENCE << 1); + amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_FENCE << 1)); amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0_INTERNAL_OFFSET, 0)); amdgpu_ring_write(ring, 0); @@ -1553,7 +1553,7 @@ static void vcn_v2_0_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0)); - amdgpu_ring_write(ring, VCN_DEC_CMD_TRAP << 1); + amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_TRAP << 1)); } /** @@ -1597,7 +1597,7 @@ static void vcn_v2_0_dec_ring_emit_reg_wait(struct amdgpu_ring *ring, amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0)); - amdgpu_ring_write(ring, VCN_DEC_CMD_REG_READ_COND_WAIT << 1); + amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_REG_READ_COND_WAIT << 1)); } static void vcn_v2_0_dec_ring_emit_vm_flush(struct amdgpu_ring *ring, @@ -1626,7 +1626,7 @@ static void vcn_v2_0_dec_ring_emit_wreg(struct amdgpu_ring *ring, amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0)); - amdgpu_ring_write(ring, VCN_DEC_CMD_WRITE_REG << 1); + amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_WRITE_REG << 1)); } /** @@ -2079,6 +2079,36 @@ static int vcn_v2_0_process_interrupt(struct amdgpu_device *adev, return 0; } +static int vcn_v2_0_dec_ring_test_ring(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t tmp = 0; + unsigned i; + int r; + + WREG32(adev->vcn.external.scratch9, 0xCAFEDEAD); + r = amdgpu_ring_alloc(ring, 4); + if (r) + return r; + amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0)); + amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_PACKET_START << 1)); + amdgpu_ring_write(ring, PACKET0(adev->vcn.internal.scratch9, 0)); + amdgpu_ring_write(ring, 0xDEADBEEF); + amdgpu_ring_commit(ring); + for (i = 0; i < adev->usec_timeout; i++) { + tmp = RREG32(adev->vcn.external.scratch9); + if (tmp == 0xDEADBEEF) + break; + DRM_UDELAY(1); + } + + if (i >= adev->usec_timeout) + r = -ETIMEDOUT; + + return r; +} + + static int vcn_v2_0_set_powergating_state(void *handle, enum amd_powergating_state state) { @@ -2142,7 +2172,7 @@ static const struct amdgpu_ring_funcs vcn_v2_0_dec_ring_vm_funcs = { .emit_ib = vcn_v2_0_dec_ring_emit_ib, .emit_fence = vcn_v2_0_dec_ring_emit_fence, .emit_vm_flush = vcn_v2_0_dec_ring_emit_vm_flush, - .test_ring = amdgpu_vcn_dec_ring_test_ring, + .test_ring = vcn_v2_0_dec_ring_test_ring, .test_ib = amdgpu_vcn_dec_ring_test_ib, .insert_nop = vcn_v2_0_dec_ring_insert_nop, .insert_start = vcn_v2_0_dec_ring_insert_start, diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index 26b15cc56c31..1d3cd5c50d5f 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -1567,32 +1567,6 @@ copy_from_user_failed: return err; } -static int kfd_ioctl_alloc_queue_gws(struct file *filep, - struct kfd_process *p, void *data) -{ - int retval; - struct kfd_ioctl_alloc_queue_gws_args *args = data; - struct kfd_dev *dev; - - if (!hws_gws_support) - return -ENODEV; - - dev = kfd_device_by_id(args->gpu_id); - if (!dev) { - pr_debug("Could not find gpu id 0x%x\n", args->gpu_id); - return -ENODEV; - } - if (dev->dqm->sched_policy == KFD_SCHED_POLICY_NO_HWS) - return -ENODEV; - - mutex_lock(&p->mutex); - retval = pqm_set_gws(&p->pqm, args->queue_id, args->num_gws ? dev->gws : NULL); - mutex_unlock(&p->mutex); - - args->first_gws = 0; - return retval; -} - static int kfd_ioctl_get_dmabuf_info(struct file *filep, struct kfd_process *p, void *data) { @@ -1795,8 +1769,6 @@ static const struct amdkfd_ioctl_desc amdkfd_ioctls[] = { AMDKFD_IOCTL_DEF(AMDKFD_IOC_IMPORT_DMABUF, kfd_ioctl_import_dmabuf, 0), - AMDKFD_IOCTL_DEF(AMDKFD_IOC_ALLOC_QUEUE_GWS, - kfd_ioctl_alloc_queue_gws, 0), }; #define AMDKFD_CORE_IOCTL_COUNT ARRAY_SIZE(amdkfd_ioctls) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 4a29f72334d0..45be7a2132bb 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -3131,13 +3131,25 @@ static enum dc_color_depth convert_color_depth_from_display_info(const struct drm_connector *connector, const struct drm_connector_state *state) { - uint32_t bpc = connector->display_info.bpc; + uint8_t bpc = (uint8_t)connector->display_info.bpc; + + /* Assume 8 bpc by default if no bpc is specified. */ + bpc = bpc ? bpc : 8; if (!state) state = connector->state; if (state) { - bpc = state->max_bpc; + /* + * Cap display bpc based on the user requested value. + * + * The value for state->max_bpc may not correctly updated + * depending on when the connector gets added to the state + * or if this was called outside of atomic check, so it + * can't be used directly. + */ + bpc = min(bpc, state->max_requested_bpc); + /* Round down to the nearest even number. */ bpc = bpc - (bpc & 1); } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index fa20201eef3a..cbc480a33376 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -23,6 +23,7 @@ */ #include <linux/slab.h> +#include <linux/mm.h> #include "dm_services.h" @@ -1171,8 +1172,8 @@ bool dc_post_update_surfaces_to_stream(struct dc *dc) struct dc_state *dc_create_state(struct dc *dc) { - struct dc_state *context = kzalloc(sizeof(struct dc_state), - GFP_KERNEL); + struct dc_state *context = kvzalloc(sizeof(struct dc_state), + GFP_KERNEL); if (!context) return NULL; @@ -1192,11 +1193,11 @@ struct dc_state *dc_create_state(struct dc *dc) struct dc_state *dc_copy_state(struct dc_state *src_ctx) { int i, j; - struct dc_state *new_ctx = kmemdup(src_ctx, - sizeof(struct dc_state), GFP_KERNEL); + struct dc_state *new_ctx = kvmalloc(sizeof(struct dc_state), GFP_KERNEL); if (!new_ctx) return NULL; + memcpy(new_ctx, src_ctx, sizeof(struct dc_state)); for (i = 0; i < MAX_PIPES; i++) { struct pipe_ctx *cur_pipe = &new_ctx->res_ctx.pipe_ctx[i]; @@ -1230,7 +1231,7 @@ static void dc_state_free(struct kref *kref) { struct dc_state *context = container_of(kref, struct dc_state, refcount); dc_resource_state_destruct(context); - kfree(context); + kvfree(context); } void dc_release_state(struct dc_state *context) diff --git a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c index 0685a3388e38..8a3eadeebdcb 100644 --- a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c @@ -315,6 +315,8 @@ int smu_get_power_num_states(struct smu_context *smu, int smu_common_read_sensor(struct smu_context *smu, enum amd_pp_sensors sensor, void *data, uint32_t *size) { + struct smu_power_context *smu_power = &smu->smu_power; + struct smu_power_gate *power_gate = &smu_power->power_gate; int ret = 0; switch (sensor) { @@ -339,7 +341,7 @@ int smu_common_read_sensor(struct smu_context *smu, enum amd_pp_sensors sensor, *size = 4; break; case AMDGPU_PP_SENSOR_VCN_POWER_STATE: - *(uint32_t *)data = smu_feature_is_enabled(smu, SMU_FEATURE_VCN_PG_BIT) ? 1 : 0; + *(uint32_t *)data = power_gate->vcn_gated ? 0 : 1; *size = 4; break; default: diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c index f27c6fbb192e..90c4e87ac5ad 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c @@ -2101,7 +2101,11 @@ static int vega20_get_gpu_power(struct pp_hwmgr *hwmgr, if (ret) return ret; - *query = metrics_table.CurrSocketPower << 8; + /* For the 40.46 release, they changed the value name */ + if (hwmgr->smu_version == 0x282e00) + *query = metrics_table.AverageSocketPower << 8; + else + *query = metrics_table.CurrSocketPower << 8; return ret; } @@ -2349,12 +2353,16 @@ static int vega20_force_dpm_highest(struct pp_hwmgr *hwmgr) data->dpm_table.soc_table.dpm_state.soft_max_level = data->dpm_table.soc_table.dpm_levels[soft_level].value; - ret = vega20_upload_dpm_min_level(hwmgr, 0xFFFFFFFF); + ret = vega20_upload_dpm_min_level(hwmgr, FEATURE_DPM_GFXCLK_MASK | + FEATURE_DPM_UCLK_MASK | + FEATURE_DPM_SOCCLK_MASK); PP_ASSERT_WITH_CODE(!ret, "Failed to upload boot level to highest!", return ret); - ret = vega20_upload_dpm_max_level(hwmgr, 0xFFFFFFFF); + ret = vega20_upload_dpm_max_level(hwmgr, FEATURE_DPM_GFXCLK_MASK | + FEATURE_DPM_UCLK_MASK | + FEATURE_DPM_SOCCLK_MASK); PP_ASSERT_WITH_CODE(!ret, "Failed to upload dpm max level to highest!", return ret); @@ -2387,12 +2395,16 @@ static int vega20_force_dpm_lowest(struct pp_hwmgr *hwmgr) data->dpm_table.soc_table.dpm_state.soft_max_level = data->dpm_table.soc_table.dpm_levels[soft_level].value; - ret = vega20_upload_dpm_min_level(hwmgr, 0xFFFFFFFF); + ret = vega20_upload_dpm_min_level(hwmgr, FEATURE_DPM_GFXCLK_MASK | + FEATURE_DPM_UCLK_MASK | + FEATURE_DPM_SOCCLK_MASK); PP_ASSERT_WITH_CODE(!ret, "Failed to upload boot level to highest!", return ret); - ret = vega20_upload_dpm_max_level(hwmgr, 0xFFFFFFFF); + ret = vega20_upload_dpm_max_level(hwmgr, FEATURE_DPM_GFXCLK_MASK | + FEATURE_DPM_UCLK_MASK | + FEATURE_DPM_SOCCLK_MASK); PP_ASSERT_WITH_CODE(!ret, "Failed to upload dpm max level to highest!", return ret); @@ -2403,14 +2415,54 @@ static int vega20_force_dpm_lowest(struct pp_hwmgr *hwmgr) static int vega20_unforce_dpm_levels(struct pp_hwmgr *hwmgr) { + struct vega20_hwmgr *data = + (struct vega20_hwmgr *)(hwmgr->backend); + uint32_t soft_min_level, soft_max_level; int ret = 0; - ret = vega20_upload_dpm_min_level(hwmgr, 0xFFFFFFFF); + /* gfxclk soft min/max settings */ + soft_min_level = + vega20_find_lowest_dpm_level(&(data->dpm_table.gfx_table)); + soft_max_level = + vega20_find_highest_dpm_level(&(data->dpm_table.gfx_table)); + + data->dpm_table.gfx_table.dpm_state.soft_min_level = + data->dpm_table.gfx_table.dpm_levels[soft_min_level].value; + data->dpm_table.gfx_table.dpm_state.soft_max_level = + data->dpm_table.gfx_table.dpm_levels[soft_max_level].value; + + /* uclk soft min/max settings */ + soft_min_level = + vega20_find_lowest_dpm_level(&(data->dpm_table.mem_table)); + soft_max_level = + vega20_find_highest_dpm_level(&(data->dpm_table.mem_table)); + + data->dpm_table.mem_table.dpm_state.soft_min_level = + data->dpm_table.mem_table.dpm_levels[soft_min_level].value; + data->dpm_table.mem_table.dpm_state.soft_max_level = + data->dpm_table.mem_table.dpm_levels[soft_max_level].value; + + /* socclk soft min/max settings */ + soft_min_level = + vega20_find_lowest_dpm_level(&(data->dpm_table.soc_table)); + soft_max_level = + vega20_find_highest_dpm_level(&(data->dpm_table.soc_table)); + + data->dpm_table.soc_table.dpm_state.soft_min_level = + data->dpm_table.soc_table.dpm_levels[soft_min_level].value; + data->dpm_table.soc_table.dpm_state.soft_max_level = + data->dpm_table.soc_table.dpm_levels[soft_max_level].value; + + ret = vega20_upload_dpm_min_level(hwmgr, FEATURE_DPM_GFXCLK_MASK | + FEATURE_DPM_UCLK_MASK | + FEATURE_DPM_SOCCLK_MASK); PP_ASSERT_WITH_CODE(!ret, "Failed to upload DPM Bootup Levels!", return ret); - ret = vega20_upload_dpm_max_level(hwmgr, 0xFFFFFFFF); + ret = vega20_upload_dpm_max_level(hwmgr, FEATURE_DPM_GFXCLK_MASK | + FEATURE_DPM_UCLK_MASK | + FEATURE_DPM_SOCCLK_MASK); PP_ASSERT_WITH_CODE(!ret, "Failed to upload DPM Max Levels!", return ret); diff --git a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h index 208e6711d506..a78b2e295895 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h @@ -451,6 +451,7 @@ struct smu_dpm_context { struct smu_power_gate { bool uvd_gated; bool vce_gated; + bool vcn_gated; }; struct smu_power_context { @@ -906,8 +907,6 @@ struct smu_funcs ((smu)->funcs->register_irq_handler ? (smu)->funcs->register_irq_handler(smu) : 0) #define smu_set_azalia_d3_pme(smu) \ ((smu)->funcs->set_azalia_d3_pme ? (smu)->funcs->set_azalia_d3_pme((smu)) : 0) -#define smu_get_uclk_dpm_states(smu, clocks_in_khz, num_states) \ - ((smu)->ppt_funcs->get_uclk_dpm_states ? (smu)->ppt_funcs->get_uclk_dpm_states((smu), (clocks_in_khz), (num_states)) : 0) #define smu_get_max_sustainable_clocks_by_dc(smu, max_clocks) \ ((smu)->funcs->get_max_sustainable_clocks_by_dc ? (smu)->funcs->get_max_sustainable_clocks_by_dc((smu), (max_clocks)) : 0) #define smu_get_uclk_dpm_states(smu, clocks_in_khz, num_states) \ diff --git a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c index cc0a3b2256af..b81c7e715dc9 100644 --- a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c +++ b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c @@ -502,6 +502,8 @@ static int navi10_store_powerplay_table(struct smu_context *smu) static int navi10_tables_init(struct smu_context *smu, struct smu_table *tables) { + struct smu_table_context *smu_table = &smu->smu_table; + SMU_TABLE_INIT(tables, SMU_TABLE_PPTABLE, sizeof(PPTable_t), PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); SMU_TABLE_INIT(tables, SMU_TABLE_WATERMARKS, sizeof(Watermarks_t), @@ -516,9 +518,35 @@ static int navi10_tables_init(struct smu_context *smu, struct smu_table *tables) sizeof(DpmActivityMonitorCoeffInt_t), PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + smu_table->metrics_table = kzalloc(sizeof(SmuMetrics_t), GFP_KERNEL); + if (!smu_table->metrics_table) + return -ENOMEM; + smu_table->metrics_time = 0; + return 0; } +static int navi10_get_metrics_table(struct smu_context *smu, + SmuMetrics_t *metrics_table) +{ + struct smu_table_context *smu_table= &smu->smu_table; + int ret = 0; + + if (!smu_table->metrics_time || time_after(jiffies, smu_table->metrics_time + HZ / 1000)) { + ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, + (void *)smu_table->metrics_table, false); + if (ret) { + pr_info("Failed to export SMU metrics table!\n"); + return ret; + } + smu_table->metrics_time = jiffies; + } + + memcpy(metrics_table, smu_table->metrics_table, sizeof(SmuMetrics_t)); + + return ret; +} + static int navi10_allocate_dpm_context(struct smu_context *smu) { struct smu_dpm_context *smu_dpm = &smu->smu_dpm; @@ -577,20 +605,27 @@ static int navi10_set_default_dpm_table(struct smu_context *smu) static int navi10_dpm_set_uvd_enable(struct smu_context *smu, bool enable) { + struct smu_power_context *smu_power = &smu->smu_power; + struct smu_power_gate *power_gate = &smu_power->power_gate; int ret = 0; if (enable) { - ret = smu_send_smc_msg_with_param(smu, SMU_MSG_PowerUpVcn, 1); - if (ret) - return ret; + /* vcn dpm on is a prerequisite for vcn power gate messages */ + if (smu_feature_is_enabled(smu, SMU_FEATURE_VCN_PG_BIT)) { + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_PowerUpVcn, 1); + if (ret) + return ret; + } + power_gate->vcn_gated = false; } else { - ret = smu_send_smc_msg(smu, SMU_MSG_PowerDownVcn); - if (ret) - return ret; + if (smu_feature_is_enabled(smu, SMU_FEATURE_VCN_PG_BIT)) { + ret = smu_send_smc_msg(smu, SMU_MSG_PowerDownVcn); + if (ret) + return ret; + } + power_gate->vcn_gated = true; } - ret = smu_feature_set_enabled(smu, SMU_FEATURE_VCN_PG_BIT, enable); - return ret; } @@ -598,15 +633,10 @@ static int navi10_get_current_clk_freq_by_table(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t *value) { - static SmuMetrics_t metrics; int ret = 0, clk_id = 0; + SmuMetrics_t metrics; - if (!value) - return -EINVAL; - - memset(&metrics, 0, sizeof(metrics)); - - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)&metrics, false); + ret = navi10_get_metrics_table(smu, &metrics); if (ret) return ret; @@ -894,8 +924,9 @@ static int navi10_get_gpu_power(struct smu_context *smu, uint32_t *value) if (!value) return -EINVAL; - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)&metrics, - false); + ret = navi10_get_metrics_table(smu, &metrics); + if (ret) + return ret; if (ret) return ret; @@ -914,10 +945,7 @@ static int navi10_get_current_activity_percent(struct smu_context *smu, if (!value) return -EINVAL; - msleep(1); - - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, - (void *)&metrics, false); + ret = navi10_get_metrics_table(smu, &metrics); if (ret) return ret; @@ -956,10 +984,9 @@ static int navi10_get_fan_speed_rpm(struct smu_context *smu, if (!speed) return -EINVAL; - memset(&metrics, 0, sizeof(metrics)); - - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, - (void *)&metrics, false); + ret = navi10_get_metrics_table(smu, &metrics); + if (ret) + return ret; if (ret) return ret; @@ -1307,7 +1334,7 @@ static int navi10_thermal_get_temperature(struct smu_context *smu, if (!value) return -EINVAL; - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)&metrics, false); + ret = navi10_get_metrics_table(smu, &metrics); if (ret) return ret; diff --git a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c index ac5b26228e75..53097961bf2b 100644 --- a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c +++ b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c @@ -326,7 +326,8 @@ static int smu_v11_0_setup_pptable(struct smu_context *smu) struct amdgpu_device *adev = smu->adev; const struct smc_firmware_header_v1_0 *hdr; int ret, index; - uint32_t size; + uint32_t size = 0; + uint16_t atom_table_size; uint8_t frev, crev; void *table; uint16_t version_major, version_minor; @@ -354,10 +355,11 @@ static int smu_v11_0_setup_pptable(struct smu_context *smu) index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, powerplayinfo); - ret = smu_get_atom_data_table(smu, index, (uint16_t *)&size, &frev, &crev, + ret = smu_get_atom_data_table(smu, index, &atom_table_size, &frev, &crev, (uint8_t **)&table); if (ret) return ret; + size = atom_table_size; } if (!smu->smu_table.power_play_table) @@ -1391,7 +1393,7 @@ smu_v11_0_smc_fan_control(struct smu_context *smu, bool start) { int ret = 0; - if (smu_feature_is_supported(smu, SMU_FEATURE_FAN_CONTROL_BIT)) + if (!smu_feature_is_supported(smu, SMU_FEATURE_FAN_CONTROL_BIT)) return 0; ret = smu_feature_set_enabled(smu, SMU_FEATURE_FAN_CONTROL_BIT, start); diff --git a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c index dd6fd1c8bf24..6a14497257e4 100644 --- a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c +++ b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c @@ -3050,6 +3050,7 @@ static int vega20_get_fan_speed_percent(struct smu_context *smu, static int vega20_get_gpu_power(struct smu_context *smu, uint32_t *value) { + uint32_t smu_version; int ret = 0; SmuMetrics_t metrics; @@ -3060,7 +3061,15 @@ static int vega20_get_gpu_power(struct smu_context *smu, uint32_t *value) if (ret) return ret; - *value = metrics.CurrSocketPower << 8; + ret = smu_get_smc_version(smu, NULL, &smu_version); + if (ret) + return ret; + + /* For the 40.46 release, they changed the value name */ + if (smu_version == 0x282e00) + *value = metrics.AverageSocketPower << 8; + else + *value = metrics.CurrSocketPower << 8; return 0; } |