diff options
52 files changed, 707 insertions, 440 deletions
diff --git a/MAINTAINERS b/MAINTAINERS index 02d5278a4c9a..b540ecc4e2a3 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4970,6 +4970,7 @@ F: include/linux/dma-buf* F: include/linux/reservation.h F: include/linux/*fence.h F: Documentation/driver-api/dma-buf.rst +K: dma_(buf|fence|resv) T: git git://anongit.freedesktop.org/drm/drm-misc DMA GENERIC OFFLOAD ENGINE SUBSYSTEM diff --git a/drivers/dma-buf/sync_file.c b/drivers/dma-buf/sync_file.c index 76fb072c22dc..5a5a1da01a00 100644 --- a/drivers/dma-buf/sync_file.c +++ b/drivers/dma-buf/sync_file.c @@ -221,7 +221,7 @@ static struct sync_file *sync_file_merge(const char *name, struct sync_file *a, a_fences = get_fences(a, &a_num_fences); b_fences = get_fences(b, &b_num_fences); if (a_num_fences > INT_MAX - b_num_fences) - return NULL; + goto err; num_fences = a_num_fences + b_num_fences; diff --git a/drivers/gpu/drm/amd/acp/Kconfig b/drivers/gpu/drm/amd/acp/Kconfig index d968c2471412..0d12ebf66174 100644 --- a/drivers/gpu/drm/amd/acp/Kconfig +++ b/drivers/gpu/drm/amd/acp/Kconfig @@ -1,4 +1,4 @@ -# SPDX-License-Identifier: GPL-2.0-only +# SPDX-License-Identifier: MIT menu "ACP (Audio CoProcessor) Configuration" config DRM_AMD_ACP diff --git a/drivers/gpu/drm/amd/amdgpu/Kconfig b/drivers/gpu/drm/amd/amdgpu/Kconfig index 2e98c016cb47..9375e7f12420 100644 --- a/drivers/gpu/drm/amd/amdgpu/Kconfig +++ b/drivers/gpu/drm/amd/amdgpu/Kconfig @@ -1,4 +1,4 @@ -# SPDX-License-Identifier: GPL-2.0-only +# SPDX-License-Identifier: MIT config DRM_AMDGPU_SI bool "Enable amdgpu support for SI parts" depends on DRM_AMDGPU diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c index 2cdaf3b2a721..6614d8a6f4c8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c @@ -604,11 +604,8 @@ void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr) continue; } - for (i = 0; i < num_entities; i++) { - mutex_lock(&ctx->adev->lock_reset); + for (i = 0; i < num_entities; i++) drm_sched_entity_fini(&ctx->entities[0][i].entity); - mutex_unlock(&ctx->adev->lock_reset); - } } } diff --git a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c index 16fbd2bc8ad1..4043ebcea5de 100644 --- a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c +++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c @@ -268,23 +268,29 @@ static void df_v3_6_update_medium_grain_clock_gating(struct amdgpu_device *adev, { u32 tmp; - /* Put DF on broadcast mode */ - adev->df_funcs->enable_broadcast_mode(adev, true); - - if (enable && (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG)) { - tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater); - tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK; - tmp |= DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY; - WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp); - } else { - tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater); - tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK; - tmp |= DF_V3_6_MGCG_DISABLE; - WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp); - } + if (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG) { + /* Put DF on broadcast mode */ + adev->df_funcs->enable_broadcast_mode(adev, true); + + if (enable) { + tmp = RREG32_SOC15(DF, 0, + mmDF_PIE_AON0_DfGlobalClkGater); + tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK; + tmp |= DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY; + WREG32_SOC15(DF, 0, + mmDF_PIE_AON0_DfGlobalClkGater, tmp); + } else { + tmp = RREG32_SOC15(DF, 0, + mmDF_PIE_AON0_DfGlobalClkGater); + tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK; + tmp |= DF_V3_6_MGCG_DISABLE; + WREG32_SOC15(DF, 0, + mmDF_PIE_AON0_DfGlobalClkGater, tmp); + } - /* Exit broadcast mode */ - adev->df_funcs->enable_broadcast_mode(adev, false); + /* Exit broadcast mode */ + adev->df_funcs->enable_broadcast_mode(adev, false); + } } static void df_v3_6_get_clockgating_state(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index f2c1b026397b..ba9e53a1abc3 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -117,10 +117,13 @@ static const struct soc15_reg_golden golden_settings_gc_10_1[] = SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CGTT_SCLK_CTRL, 0x10000000, 0x10000100), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CTRL2, 0xffffffff, 0x1402002f), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CTRL3, 0xffff9fff, 0x00001188), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_TIMEOUT_COUNTER, 0xffffffff, 0x00000800), SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE, 0x3fffffff, 0x08000009), SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0x00400000, 0x04440000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_2, 0x00000800, 0x00000820), SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000), SOC15_REG_GOLDEN_VALUE(GC, 0, mmRMI_SPARE, 0xffffffff, 0xffff3101), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL, 0x001f0000, 0x00070104), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_ALU_CLK_CTRL, 0xffffffff, 0xffffffff), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_ARB_CONFIG, 0x00000100, 0x00000130), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_LDS_CLK_CTRL, 0xffffffff, 0xffffffff), @@ -162,10 +165,13 @@ static const struct soc15_reg_golden golden_settings_gc_10_1_1[] = SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CGTT_SCLK_CTRL, 0xffff0fff, 0x10000100), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CTRL2, 0xffffffff, 0x1402002f), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CTRL3, 0xffffbfff, 0x00000188), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_TIMEOUT_COUNTER, 0xffffffff, 0x00000800), SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE, 0x3fffffff, 0x08000009), SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0x00400000, 0x04440000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_2, 0x00000800, 0x00000820), SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000), SOC15_REG_GOLDEN_VALUE(GC, 0, mmRMI_SPARE, 0xffffffff, 0xffff3101), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL, 0x001f0000, 0x00070105), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_ALU_CLK_CTRL, 0xffffffff, 0xffffffff), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_ARB_CONFIG, 0x00000133, 0x00000130), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_LDS_CLK_CTRL, 0xffffffff, 0xffffffff), diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 983db77999e7..52a647d7022d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -6146,7 +6146,23 @@ static void gfx_v8_0_ring_emit_fence_gfx(struct amdgpu_ring *ring, u64 addr, bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT; bool int_sel = flags & AMDGPU_FENCE_FLAG_INT; - /* EVENT_WRITE_EOP - flush caches, send int */ + /* Workaround for cache flush problems. First send a dummy EOP + * event down the pipe with seq one below. + */ + amdgpu_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4)); + amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN | + EOP_TC_ACTION_EN | + EOP_TC_WB_ACTION_EN | + EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) | + EVENT_INDEX(5))); + amdgpu_ring_write(ring, addr & 0xfffffffc); + amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xffff) | + DATA_SEL(1) | INT_SEL(0)); + amdgpu_ring_write(ring, lower_32_bits(seq - 1)); + amdgpu_ring_write(ring, upper_32_bits(seq - 1)); + + /* Then send the real EOP event down the pipe: + * EVENT_WRITE_EOP - flush caches, send int */ amdgpu_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4)); amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN | EOP_TC_ACTION_EN | @@ -6888,7 +6904,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { 5 + /* COND_EXEC */ 7 + /* PIPELINE_SYNC */ VI_FLUSH_GPU_TLB_NUM_WREG * 5 + 9 + /* VM_FLUSH */ - 8 + /* FENCE for VM_FLUSH */ + 12 + /* FENCE for VM_FLUSH */ 20 + /* GDS switch */ 4 + /* double SWITCH_BUFFER, the first COND_EXEC jump to the place just @@ -6900,7 +6916,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { 31 + /* DE_META */ 3 + /* CNTX_CTRL */ 5 + /* HDP_INVL */ - 8 + 8 + /* FENCE x2 */ + 12 + 12 + /* FENCE x2 */ 2, /* SWITCH_BUFFER */ .emit_ib_size = 4, /* gfx_v8_0_ring_emit_ib_gfx */ .emit_ib = gfx_v8_0_ring_emit_ib_gfx, diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c index 232469507446..f5725336a5f2 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c @@ -219,6 +219,21 @@ static uint32_t gmc_v10_0_get_invalidate_req(unsigned int vmid, return req; } +/** + * gmc_v10_0_use_invalidate_semaphore - judge whether to use semaphore + * + * @adev: amdgpu_device pointer + * @vmhub: vmhub type + * + */ +static bool gmc_v10_0_use_invalidate_semaphore(struct amdgpu_device *adev, + uint32_t vmhub) +{ + return ((vmhub == AMDGPU_MMHUB_0 || + vmhub == AMDGPU_MMHUB_1) && + (!amdgpu_sriov_vf(adev))); +} + /* * GART * VMID 0 is the physical GPU addresses as used by the kernel. @@ -229,6 +244,7 @@ static uint32_t gmc_v10_0_get_invalidate_req(unsigned int vmid, static void gmc_v10_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid, unsigned int vmhub, uint32_t flush_type) { + bool use_semaphore = gmc_v10_0_use_invalidate_semaphore(adev, vmhub); struct amdgpu_vmhub *hub = &adev->vmhub[vmhub]; u32 tmp = gmc_v10_0_get_invalidate_req(vmid, flush_type); /* Use register 17 for GART */ @@ -244,8 +260,7 @@ static void gmc_v10_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid, */ /* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */ - if (vmhub == AMDGPU_MMHUB_0 || - vmhub == AMDGPU_MMHUB_1) { + if (use_semaphore) { for (i = 0; i < adev->usec_timeout; i++) { /* a read return value of 1 means semaphore acuqire */ tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_sem + eng); @@ -278,8 +293,7 @@ static void gmc_v10_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid, } /* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */ - if (vmhub == AMDGPU_MMHUB_0 || - vmhub == AMDGPU_MMHUB_1) + if (use_semaphore) /* * add semaphore release after invalidation, * write with 0 means semaphore release @@ -369,6 +383,7 @@ error_alloc: static uint64_t gmc_v10_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { + bool use_semaphore = gmc_v10_0_use_invalidate_semaphore(ring->adev, ring->funcs->vmhub); struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = gmc_v10_0_get_invalidate_req(vmid, 0); unsigned eng = ring->vm_inv_eng; @@ -381,8 +396,7 @@ static uint64_t gmc_v10_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, */ /* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */ - if (ring->funcs->vmhub == AMDGPU_MMHUB_0 || - ring->funcs->vmhub == AMDGPU_MMHUB_1) + if (use_semaphore) /* a read return value of 1 means semaphore acuqire */ amdgpu_ring_emit_reg_wait(ring, hub->vm_inv_eng0_sem + eng, 0x1, 0x1); @@ -398,8 +412,7 @@ static uint64_t gmc_v10_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, req, 1 << vmid); /* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */ - if (ring->funcs->vmhub == AMDGPU_MMHUB_0 || - ring->funcs->vmhub == AMDGPU_MMHUB_1) + if (use_semaphore) /* * add semaphore release after invalidation, * write with 0 means semaphore release diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 3c355fb5d2b4..a5b68b5e452f 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -416,6 +416,24 @@ static uint32_t gmc_v9_0_get_invalidate_req(unsigned int vmid, return req; } +/** + * gmc_v9_0_use_invalidate_semaphore - judge whether to use semaphore + * + * @adev: amdgpu_device pointer + * @vmhub: vmhub type + * + */ +static bool gmc_v9_0_use_invalidate_semaphore(struct amdgpu_device *adev, + uint32_t vmhub) +{ + return ((vmhub == AMDGPU_MMHUB_0 || + vmhub == AMDGPU_MMHUB_1) && + (!amdgpu_sriov_vf(adev)) && + (!(adev->asic_type == CHIP_RAVEN && + adev->rev_id < 0x8 && + adev->pdev->device == 0x15d8))); +} + /* * GART * VMID 0 is the physical GPU addresses as used by the kernel. @@ -435,6 +453,7 @@ static uint32_t gmc_v9_0_get_invalidate_req(unsigned int vmid, static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid, uint32_t vmhub, uint32_t flush_type) { + bool use_semaphore = gmc_v9_0_use_invalidate_semaphore(adev, vmhub); const unsigned eng = 17; u32 j, tmp; struct amdgpu_vmhub *hub; @@ -468,8 +487,7 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid, */ /* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */ - if (vmhub == AMDGPU_MMHUB_0 || - vmhub == AMDGPU_MMHUB_1) { + if (use_semaphore) { for (j = 0; j < adev->usec_timeout; j++) { /* a read return value of 1 means semaphore acuqire */ tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_sem + eng); @@ -499,8 +517,7 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid, } /* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */ - if (vmhub == AMDGPU_MMHUB_0 || - vmhub == AMDGPU_MMHUB_1) + if (use_semaphore) /* * add semaphore release after invalidation, * write with 0 means semaphore release @@ -518,6 +535,7 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid, static uint64_t gmc_v9_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { + bool use_semaphore = gmc_v9_0_use_invalidate_semaphore(ring->adev, ring->funcs->vmhub); struct amdgpu_device *adev = ring->adev; struct amdgpu_vmhub *hub = &adev->vmhub[ring->funcs->vmhub]; uint32_t req = gmc_v9_0_get_invalidate_req(vmid, 0); @@ -531,8 +549,7 @@ static uint64_t gmc_v9_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, */ /* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */ - if (ring->funcs->vmhub == AMDGPU_MMHUB_0 || - ring->funcs->vmhub == AMDGPU_MMHUB_1) + if (use_semaphore) /* a read return value of 1 means semaphore acuqire */ amdgpu_ring_emit_reg_wait(ring, hub->vm_inv_eng0_sem + eng, 0x1, 0x1); @@ -548,8 +565,7 @@ static uint64_t gmc_v9_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, req, 1 << vmid); /* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */ - if (ring->funcs->vmhub == AMDGPU_MMHUB_0 || - ring->funcs->vmhub == AMDGPU_MMHUB_1) + if (use_semaphore) /* * add semaphore release after invalidation, * write with 0 means semaphore release diff --git a/drivers/gpu/drm/amd/amdkfd/Kconfig b/drivers/gpu/drm/amd/amdkfd/Kconfig index ba0e68057a89..b3672d10ea54 100644 --- a/drivers/gpu/drm/amd/amdkfd/Kconfig +++ b/drivers/gpu/drm/amd/amdkfd/Kconfig @@ -1,4 +1,4 @@ -# SPDX-License-Identifier: GPL-2.0-only +# SPDX-License-Identifier: MIT # # Heterogenous system architecture configuration # diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig index 313183b80032..ae161fe86ebb 100644 --- a/drivers/gpu/drm/amd/display/Kconfig +++ b/drivers/gpu/drm/amd/display/Kconfig @@ -1,4 +1,4 @@ -# SPDX-License-Identifier: GPL-2.0-only +# SPDX-License-Identifier: MIT menu "Display Engine Configuration" depends on DRM && DRM_AMDGPU diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c index 7873abea4112..5c3fcaa47410 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c @@ -1625,6 +1625,7 @@ static enum bp_result construct_integrated_info( /* Don't need to check major revision as they are all 1 */ switch (revision.minor) { case 11: + case 12: result = get_integrated_info_v11(bp, info); break; default: diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c index 790a2d211bd6..35c55e54eac0 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c @@ -471,12 +471,28 @@ static void rn_notify_wm_ranges(struct clk_mgr *clk_mgr_base) } +static bool rn_are_clock_states_equal(struct dc_clocks *a, + struct dc_clocks *b) +{ + if (a->dispclk_khz != b->dispclk_khz) + return false; + else if (a->dppclk_khz != b->dppclk_khz) + return false; + else if (a->dcfclk_khz != b->dcfclk_khz) + return false; + else if (a->dcfclk_deep_sleep_khz != b->dcfclk_deep_sleep_khz) + return false; + + return true; +} + + static struct clk_mgr_funcs dcn21_funcs = { .get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz, .update_clocks = rn_update_clocks, .init_clocks = rn_init_clocks, .enable_pme_wa = rn_enable_pme_wa, - /* .dump_clk_registers = rn_dump_clk_registers, */ + .are_clock_states_equal = rn_are_clock_states_equal, .notify_wm_ranges = rn_notify_wm_ranges }; @@ -518,36 +534,83 @@ struct clk_bw_params rn_bw_params = { .num_entries = 4, }, - .wm_table = { - .entries = { - { - .wm_inst = WM_A, - .wm_type = WM_TYPE_PSTATE_CHG, - .pstate_latency_us = 23.84, - .valid = true, - }, - { - .wm_inst = WM_B, - .wm_type = WM_TYPE_PSTATE_CHG, - .pstate_latency_us = 23.84, - .valid = true, - }, - { - .wm_inst = WM_C, - .wm_type = WM_TYPE_PSTATE_CHG, - .pstate_latency_us = 23.84, - .valid = true, - }, - { - .wm_inst = WM_D, - .wm_type = WM_TYPE_PSTATE_CHG, - .pstate_latency_us = 23.84, - .valid = true, - }, +}; + +struct wm_table ddr4_wm_table = { + .entries = { + { + .wm_inst = WM_A, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 6.09, + .sr_enter_plus_exit_time_us = 7.14, + .valid = true, + }, + { + .wm_inst = WM_B, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 10.12, + .sr_enter_plus_exit_time_us = 11.48, + .valid = true, + }, + { + .wm_inst = WM_C, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 10.12, + .sr_enter_plus_exit_time_us = 11.48, + .valid = true, + }, + { + .wm_inst = WM_D, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 10.12, + .sr_enter_plus_exit_time_us = 11.48, + .valid = true, }, } }; +struct wm_table lpddr4_wm_table = { + .entries = { + { + .wm_inst = WM_A, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 23.84, + .sr_exit_time_us = 12.5, + .sr_enter_plus_exit_time_us = 17.0, + .valid = true, + }, + { + .wm_inst = WM_B, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 23.84, + .sr_exit_time_us = 12.5, + .sr_enter_plus_exit_time_us = 17.0, + .valid = true, + }, + { + .wm_inst = WM_C, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 23.84, + .sr_exit_time_us = 12.5, + .sr_enter_plus_exit_time_us = 17.0, + .valid = true, + }, + { + .wm_inst = WM_D, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 23.84, + .sr_exit_time_us = 12.5, + .sr_enter_plus_exit_time_us = 17.0, + .valid = true, + }, + } +}; + + static unsigned int find_dcfclk_for_voltage(struct dpm_clocks *clock_table, unsigned int voltage) { int i; @@ -561,7 +624,7 @@ static unsigned int find_dcfclk_for_voltage(struct dpm_clocks *clock_table, unsi return 0; } -static void rn_clk_mgr_helper_populate_bw_params(struct clk_bw_params *bw_params, struct dpm_clocks *clock_table, struct hw_asic_id *asic_id) +static void rn_clk_mgr_helper_populate_bw_params(struct clk_bw_params *bw_params, struct dpm_clocks *clock_table, struct integrated_info *bios_info) { int i, j = 0; @@ -593,8 +656,8 @@ static void rn_clk_mgr_helper_populate_bw_params(struct clk_bw_params *bw_params bw_params->clk_table.entries[i].dcfclk_mhz = find_dcfclk_for_voltage(clock_table, clock_table->FClocks[j].Vol); } - bw_params->vram_type = asic_id->vram_type; - bw_params->num_channels = asic_id->vram_width / DDR4_DRAM_WIDTH; + bw_params->vram_type = bios_info->memory_type; + bw_params->num_channels = bios_info->ma_channel_number; for (i = 0; i < WM_SET_COUNT; i++) { bw_params->wm_table.entries[i].wm_inst = i; @@ -669,15 +732,24 @@ void rn_clk_mgr_construct( ASSERT(clk_mgr->base.dprefclk_khz == 600000); clk_mgr->base.dprefclk_khz = 600000; } + + if (ctx->dc_bios->integrated_info->memory_type == LpDdr4MemType) { + rn_bw_params.wm_table = lpddr4_wm_table; + } else { + rn_bw_params.wm_table = ddr4_wm_table; + } } dce_clock_read_ss_info(clk_mgr); + clk_mgr->base.bw_params = &rn_bw_params; if (pp_smu && pp_smu->rn_funcs.get_dpm_clock_table) { pp_smu->rn_funcs.get_dpm_clock_table(&pp_smu->rn_funcs.pp_smu, &clock_table); - rn_clk_mgr_helper_populate_bw_params(clk_mgr->base.bw_params, &clock_table, &ctx->asic_id); + if (ctx->dc_bios && ctx->dc_bios->integrated_info) { + rn_clk_mgr_helper_populate_bw_params (clk_mgr->base.bw_params, &clock_table, ctx->dc_bios->integrated_info); + } } if (!IS_FPGA_MAXIMUS_DC(ctx->dce_environment) && clk_mgr->smu_ver >= 0x00371500) { diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index 12ba6fdf89b7..62d8289abb4e 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -372,7 +372,7 @@ bool dc_link_is_dp_sink_present(struct dc_link *link) if (GPIO_RESULT_OK != dal_ddc_open( ddc, GPIO_MODE_INPUT, GPIO_DDC_CONFIG_TYPE_MODE_I2C)) { - dal_gpio_destroy_ddc(&ddc); + dal_ddc_close(ddc); return present; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c index 7f904d55c1bc..81789191d4ec 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c @@ -586,7 +586,7 @@ bool dal_ddc_service_query_ddc_data( bool dal_ddc_submit_aux_command(struct ddc_service *ddc, struct aux_payload *payload) { - uint8_t retrieved = 0; + uint32_t retrieved = 0; bool ret = 0; if (!ddc) diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index 0f59b68aa4c2..504055fc70e8 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c @@ -3522,7 +3522,14 @@ void dp_set_fec_enable(struct dc_link *link, bool enable) if (link_enc->funcs->fec_set_enable && link->dpcd_caps.fec_cap.bits.FEC_CAPABLE) { if (link->fec_state == dc_link_fec_ready && enable) { - msleep(1); + /* Accord to DP spec, FEC enable sequence can first + * be transmitted anytime after 1000 LL codes have + * been transmitted on the link after link training + * completion. Using 1 lane RBR should have the maximum + * time for transmitting 1000 LL codes which is 6.173 us. + * So use 7 microseconds delay instead. + */ + udelay(7); link_enc->funcs->fec_set_enable(link_enc, true); link->fec_state = dc_link_fec_enabled; } else if (link->fec_state == dc_link_fec_enabled && !enable) { diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c index e472608faf33..793c0cec407f 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c @@ -583,6 +583,8 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc, uint8_t reply; bool payload_reply = true; enum aux_channel_operation_result operation_result; + bool retry_on_defer = false; + int aux_ack_retries = 0, aux_defer_retries = 0, aux_i2c_defer_retries = 0, @@ -613,8 +615,10 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc, break; case AUX_TRANSACTION_REPLY_AUX_DEFER: - case AUX_TRANSACTION_REPLY_I2C_OVER_AUX_NACK: case AUX_TRANSACTION_REPLY_I2C_OVER_AUX_DEFER: + retry_on_defer = true; + /* fall through */ + case AUX_TRANSACTION_REPLY_I2C_OVER_AUX_NACK: if (++aux_defer_retries >= AUX_MAX_DEFER_RETRIES) { goto fail; } else { @@ -647,15 +651,24 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc, break; case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT: - if (++aux_timeout_retries >= AUX_MAX_TIMEOUT_RETRIES) - goto fail; - else { - /* - * DP 1.4, 2.8.2: AUX Transaction Response/Reply Timeouts - * According to the DP spec there should be 3 retries total - * with a 400us wait inbetween each. Hardware already waits - * for 550us therefore no wait is required here. - */ + // Check whether a DEFER had occurred before the timeout. + // If so, treat timeout as a DEFER. + if (retry_on_defer) { + if (++aux_defer_retries >= AUX_MAX_DEFER_RETRIES) + goto fail; + else if (payload->defer_delay > 0) + msleep(payload->defer_delay); + } else { + if (++aux_timeout_retries >= AUX_MAX_TIMEOUT_RETRIES) + goto fail; + else { + /* + * DP 1.4, 2.8.2: AUX Transaction Response/Reply Timeouts + * According to the DP spec there should be 3 retries total + * with a 400us wait inbetween each. Hardware already waits + * for 550us therefore no wait is required here. + */ + } } break; diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/Makefile b/drivers/gpu/drm/amd/display/dc/dcn20/Makefile index 63f3bddba7da..10b47986526b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn20/Makefile @@ -1,3 +1,4 @@ +# SPDX-License-Identifier: MIT # # Makefile for DCN. diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c index 09793336d84f..23ff2f1c75b5 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c @@ -923,7 +923,9 @@ static const struct resource_caps res_cap_nv14 = { .num_dwb = 1, .num_ddc = 5, .num_vmid = 16, +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT .num_dsc = 5, +#endif }; static const struct dc_debug_options debug_defaults_drv = { @@ -1536,13 +1538,20 @@ enum dc_status dcn20_build_mapped_resource(const struct dc *dc, struct dc_state static void acquire_dsc(struct resource_context *res_ctx, const struct resource_pool *pool, - struct display_stream_compressor **dsc) + struct display_stream_compressor **dsc, + int pipe_idx) { int i; ASSERT(*dsc == NULL); *dsc = NULL; + if (pool->res_cap->num_dsc == pool->res_cap->num_opp) { + *dsc = pool->dscs[pipe_idx]; + res_ctx->is_dsc_acquired[pipe_idx] = true; + return; + } + /* Find first free DSC */ for (i = 0; i < pool->res_cap->num_dsc; i++) if (!res_ctx->is_dsc_acquired[i]) { @@ -1585,7 +1594,7 @@ static enum dc_status add_dsc_to_stream_resource(struct dc *dc, if (pipe_ctx->stream != dc_stream) continue; - acquire_dsc(&dc_ctx->res_ctx, pool, &pipe_ctx->stream_res.dsc); + acquire_dsc(&dc_ctx->res_ctx, pool, &pipe_ctx->stream_res.dsc, i); /* The number of DSCs can be less than the number of pipes */ if (!pipe_ctx->stream_res.dsc) { @@ -1785,7 +1794,7 @@ bool dcn20_split_stream_for_odm( next_odm_pipe->stream_res.opp = pool->opps[next_odm_pipe->pipe_idx]; #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT if (next_odm_pipe->stream->timing.flags.DSC == 1) { - acquire_dsc(res_ctx, pool, &next_odm_pipe->stream_res.dsc); + acquire_dsc(res_ctx, pool, &next_odm_pipe->stream_res.dsc, next_odm_pipe->pipe_idx); ASSERT(next_odm_pipe->stream_res.dsc); if (next_odm_pipe->stream_res.dsc == NULL) return false; diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c index 4b3401616434..fcb3877b4fcb 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c @@ -492,15 +492,23 @@ void enc2_stream_encoder_dp_unblank( DP_VID_N_MUL, n_multiply); } - /* set DIG_START to 0x1 to reset FIFO */ + /* make sure stream is disabled before resetting steer fifo */ + REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, false); + REG_WAIT(DP_VID_STREAM_CNTL, DP_VID_STREAM_STATUS, 0, 10, 5000); + /* set DIG_START to 0x1 to reset FIFO */ REG_UPDATE(DIG_FE_CNTL, DIG_START, 1); + udelay(1); /* write 0 to take the FIFO out of reset */ REG_UPDATE(DIG_FE_CNTL, DIG_START, 0); - /* switch DP encoder to CRTC data */ + /* switch DP encoder to CRTC data, but reset it the fifo first. It may happen + * that it overflows during mode transition, and sometimes doesn't recover. + */ + REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 1); + udelay(10); REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 0); diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/Makefile b/drivers/gpu/drm/amd/display/dc/dcn21/Makefile index 14113ccf498d..5b8c17564bc1 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn21/Makefile @@ -1,3 +1,4 @@ +# SPDX-License-Identifier: MIT # # Makefile for DCN21. diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c index 459bd9a5caed..b29b2c99a564 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c @@ -23,6 +23,8 @@ * */ +#include <linux/slab.h> + #include "dm_services.h" #include "dc.h" @@ -257,7 +259,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn2_1_soc = { .vmm_page_size_bytes = 4096, .dram_clock_change_latency_us = 23.84, .return_bus_width_bytes = 64, - .dispclk_dppclk_vco_speed_mhz = 3550, + .dispclk_dppclk_vco_speed_mhz = 3600, .xfc_bus_transport_time_us = 4, .xfc_xbuf_latency_tolerance_us = 4, .use_urgent_burst_bw = 1, @@ -1000,6 +1002,8 @@ static void calculate_wm_set_for_vlevel( pipes[0].clks_cfg.socclk_mhz = dml->soc.clock_limits[vlevel].socclk_mhz; dml->soc.dram_clock_change_latency_us = table_entry->pstate_latency_us; + dml->soc.sr_exit_time_us = table_entry->sr_exit_time_us; + dml->soc.sr_enter_plus_exit_time_us = table_entry->sr_enter_plus_exit_time_us; wm_set->urgent_ns = get_wm_urgent(dml, pipes, pipe_cnt) * 1000; wm_set->cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(dml, pipes, pipe_cnt) * 1000; @@ -1017,14 +1021,21 @@ static void calculate_wm_set_for_vlevel( static void patch_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb) { + int i; + kernel_fpu_begin(); if (dc->bb_overrides.sr_exit_time_ns) { - bb->sr_exit_time_us = dc->bb_overrides.sr_exit_time_ns / 1000.0; + for (i = 0; i < WM_SET_COUNT; i++) { + dc->clk_mgr->bw_params->wm_table.entries[i].sr_exit_time_us = + dc->bb_overrides.sr_exit_time_ns / 1000.0; + } } if (dc->bb_overrides.sr_enter_plus_exit_time_ns) { - bb->sr_enter_plus_exit_time_us = - dc->bb_overrides.sr_enter_plus_exit_time_ns / 1000.0; + for (i = 0; i < WM_SET_COUNT; i++) { + dc->clk_mgr->bw_params->wm_table.entries[i].sr_enter_plus_exit_time_us = + dc->bb_overrides.sr_enter_plus_exit_time_ns / 1000.0; + } } if (dc->bb_overrides.urgent_latency_ns) { @@ -1032,9 +1043,12 @@ static void patch_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_s } if (dc->bb_overrides.dram_clock_change_latency_ns) { - bb->dram_clock_change_latency_us = + for (i = 0; i < WM_SET_COUNT; i++) { + dc->clk_mgr->bw_params->wm_table.entries[i].pstate_latency_us = dc->bb_overrides.dram_clock_change_latency_ns / 1000.0; + } } + kernel_fpu_end(); } diff --git a/drivers/gpu/drm/amd/display/dc/dsc/Makefile b/drivers/gpu/drm/amd/display/dc/dsc/Makefile index 970737217e53..641ffb7cfaed 100644 --- a/drivers/gpu/drm/amd/display/dc/dsc/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dsc/Makefile @@ -1,3 +1,4 @@ +# SPDX-License-Identifier: MIT # # Makefile for the 'dsc' sub-component of DAL. diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h index 4e18e77dcf42..026e6a2a2c44 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h @@ -69,6 +69,8 @@ struct wm_range_table_entry { unsigned int wm_inst; unsigned int wm_type; double pstate_latency_us; + double sr_exit_time_us; + double sr_enter_plus_exit_time_us; bool valid; }; diff --git a/drivers/gpu/drm/amd/display/include/i2caux_interface.h b/drivers/gpu/drm/amd/display/include/i2caux_interface.h index bb012cb1a9f5..c7fbb9c3ad6b 100644 --- a/drivers/gpu/drm/amd/display/include/i2caux_interface.h +++ b/drivers/gpu/drm/amd/display/include/i2caux_interface.h @@ -42,7 +42,7 @@ struct aux_payload { bool write; bool mot; uint32_t address; - uint8_t length; + uint32_t length; uint8_t *data; /* * used to return the reply type of the transaction diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index 16e69bbc69aa..5437b50e9f90 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c @@ -37,8 +37,8 @@ #define STATIC_SCREEN_RAMP_DELTA_REFRESH_RATE_PER_FRAME ((1000 / 60) * 65) /* Number of elements in the render times cache array */ #define RENDER_TIMES_MAX_COUNT 10 -/* Threshold to exit/exit BTR (to avoid frequent enter-exits at the lower limit) */ -#define BTR_MAX_MARGIN 2500 +/* Threshold to exit BTR (to avoid frequent enter-exits at the lower limit) */ +#define BTR_EXIT_MARGIN 2000 /* Threshold to change BTR multiplier (to avoid frequent changes) */ #define BTR_DRIFT_MARGIN 2000 /*Threshold to exit fixed refresh rate*/ @@ -254,22 +254,24 @@ static void apply_below_the_range(struct core_freesync *core_freesync, unsigned int delta_from_mid_point_in_us_1 = 0xFFFFFFFF; unsigned int delta_from_mid_point_in_us_2 = 0xFFFFFFFF; unsigned int frames_to_insert = 0; + unsigned int min_frame_duration_in_ns = 0; + unsigned int max_render_time_in_us = in_out_vrr->max_duration_in_us; unsigned int delta_from_mid_point_delta_in_us; - unsigned int max_render_time_in_us = - in_out_vrr->max_duration_in_us - in_out_vrr->btr.margin_in_us; + + min_frame_duration_in_ns = ((unsigned int) (div64_u64( + (1000000000ULL * 1000000), + in_out_vrr->max_refresh_in_uhz))); /* Program BTR */ - if ((last_render_time_in_us + in_out_vrr->btr.margin_in_us / 2) < max_render_time_in_us) { + if (last_render_time_in_us + BTR_EXIT_MARGIN < max_render_time_in_us) { /* Exit Below the Range */ if (in_out_vrr->btr.btr_active) { in_out_vrr->btr.frame_counter = 0; in_out_vrr->btr.btr_active = false; } - } else if (last_render_time_in_us > (max_render_time_in_us + in_out_vrr->btr.margin_in_us / 2)) { + } else if (last_render_time_in_us > max_render_time_in_us) { /* Enter Below the Range */ - if (!in_out_vrr->btr.btr_active) { - in_out_vrr->btr.btr_active = true; - } + in_out_vrr->btr.btr_active = true; } /* BTR set to "not active" so disengage */ @@ -325,9 +327,7 @@ static void apply_below_the_range(struct core_freesync *core_freesync, /* Choose number of frames to insert based on how close it * can get to the mid point of the variable range. */ - if ((frame_time_in_us / mid_point_frames_ceil) > in_out_vrr->min_duration_in_us && - (delta_from_mid_point_in_us_1 < delta_from_mid_point_in_us_2 || - mid_point_frames_floor < 2)) { + if (delta_from_mid_point_in_us_1 < delta_from_mid_point_in_us_2) { frames_to_insert = mid_point_frames_ceil; delta_from_mid_point_delta_in_us = delta_from_mid_point_in_us_2 - delta_from_mid_point_in_us_1; @@ -343,7 +343,7 @@ static void apply_below_the_range(struct core_freesync *core_freesync, if (in_out_vrr->btr.frames_to_insert != 0 && delta_from_mid_point_delta_in_us < BTR_DRIFT_MARGIN) { if (((last_render_time_in_us / in_out_vrr->btr.frames_to_insert) < - max_render_time_in_us) && + in_out_vrr->max_duration_in_us) && ((last_render_time_in_us / in_out_vrr->btr.frames_to_insert) > in_out_vrr->min_duration_in_us)) frames_to_insert = in_out_vrr->btr.frames_to_insert; @@ -796,11 +796,6 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync, refresh_range = in_out_vrr->max_refresh_in_uhz - in_out_vrr->min_refresh_in_uhz; - in_out_vrr->btr.margin_in_us = in_out_vrr->max_duration_in_us - - 2 * in_out_vrr->min_duration_in_us; - if (in_out_vrr->btr.margin_in_us > BTR_MAX_MARGIN) - in_out_vrr->btr.margin_in_us = BTR_MAX_MARGIN; - in_out_vrr->supported = true; } @@ -816,7 +811,6 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync, in_out_vrr->btr.inserted_duration_in_us = 0; in_out_vrr->btr.frames_to_insert = 0; in_out_vrr->btr.frame_counter = 0; - in_out_vrr->btr.mid_point_in_us = (in_out_vrr->min_duration_in_us + in_out_vrr->max_duration_in_us) / 2; diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h index dbe7835aabcf..dc187844d10b 100644 --- a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h +++ b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h @@ -92,7 +92,6 @@ struct mod_vrr_params_btr { uint32_t inserted_duration_in_us; uint32_t frames_to_insert; uint32_t frame_counter; - uint32_t margin_in_us; }; struct mod_vrr_params_fixed_refresh { diff --git a/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c b/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c index ce3566ca3e24..cc71a1078a7a 100644 --- a/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c +++ b/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c @@ -1313,12 +1313,17 @@ static int arcturus_get_power_profile_mode(struct smu_context *smu, "VR", "COMPUTE", "CUSTOM"}; + static const char *title[] = { + "PROFILE_INDEX(NAME)"}; uint32_t i, size = 0; int16_t workload_type = 0; if (!smu->pm_enabled || !buf) return -EINVAL; + size += sprintf(buf + size, "%16s\n", + title[0]); + for (i = 0; i <= PP_SMC_POWER_PROFILE_CUSTOM; i++) { /* * Conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c index c7c2b349858d..2a27fb5d7dc6 100644 --- a/drivers/gpu/drm/i915/display/intel_ddi.c +++ b/drivers/gpu/drm/i915/display/intel_ddi.c @@ -3986,6 +3986,7 @@ static void intel_enable_ddi(struct intel_encoder *encoder, if (conn_state->content_protection == DRM_MODE_CONTENT_PROTECTION_DESIRED) intel_hdcp_enable(to_intel_connector(conn_state->connector), + crtc_state->cpu_transcoder, (u8)conn_state->hdcp_content_type); } @@ -4089,7 +4090,9 @@ static void intel_ddi_update_pipe(struct intel_encoder *encoder, if (conn_state->content_protection == DRM_MODE_CONTENT_PROTECTION_DESIRED || content_protection_type_changed) - intel_hdcp_enable(connector, (u8)conn_state->hdcp_content_type); + intel_hdcp_enable(connector, + crtc_state->cpu_transcoder, + (u8)conn_state->hdcp_content_type); } static void diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c index 050655a1a3d8..b05b2191b919 100644 --- a/drivers/gpu/drm/i915/display/intel_dp.c +++ b/drivers/gpu/drm/i915/display/intel_dp.c @@ -2414,9 +2414,6 @@ intel_dp_compute_config(struct intel_encoder *encoder, intel_psr_compute_config(intel_dp, pipe_config); - intel_hdcp_transcoder_config(intel_connector, - pipe_config->cpu_transcoder); - return 0; } diff --git a/drivers/gpu/drm/i915/display/intel_fbc.c b/drivers/gpu/drm/i915/display/intel_fbc.c index 3111ecaeabd0..20616639b8ab 100644 --- a/drivers/gpu/drm/i915/display/intel_fbc.c +++ b/drivers/gpu/drm/i915/display/intel_fbc.c @@ -1284,7 +1284,7 @@ static int intel_sanitize_fbc_option(struct drm_i915_private *dev_priv) return 0; /* https://bugs.freedesktop.org/show_bug.cgi?id=108085 */ - if (IS_GEMINILAKE(dev_priv)) + if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) return 0; if (IS_BROADWELL(dev_priv) || INTEL_GEN(dev_priv) >= 9) diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.c b/drivers/gpu/drm/i915/display/intel_hdcp.c index f1f41ca8402b..a448815d8fc2 100644 --- a/drivers/gpu/drm/i915/display/intel_hdcp.c +++ b/drivers/gpu/drm/i915/display/intel_hdcp.c @@ -1821,23 +1821,6 @@ enum mei_fw_tc intel_get_mei_fw_tc(enum transcoder cpu_transcoder) } } -void intel_hdcp_transcoder_config(struct intel_connector *connector, - enum transcoder cpu_transcoder) -{ - struct drm_i915_private *dev_priv = to_i915(connector->base.dev); - struct intel_hdcp *hdcp = &connector->hdcp; - - if (!hdcp->shim) - return; - - if (INTEL_GEN(dev_priv) >= 12) { - mutex_lock(&hdcp->mutex); - hdcp->cpu_transcoder = cpu_transcoder; - hdcp->port_data.fw_tc = intel_get_mei_fw_tc(cpu_transcoder); - mutex_unlock(&hdcp->mutex); - } -} - static inline int initialize_hdcp_port_data(struct intel_connector *connector, const struct intel_hdcp_shim *shim) { @@ -1959,8 +1942,10 @@ int intel_hdcp_init(struct intel_connector *connector, return 0; } -int intel_hdcp_enable(struct intel_connector *connector, u8 content_type) +int intel_hdcp_enable(struct intel_connector *connector, + enum transcoder cpu_transcoder, u8 content_type) { + struct drm_i915_private *dev_priv = to_i915(connector->base.dev); struct intel_hdcp *hdcp = &connector->hdcp; unsigned long check_link_interval = DRM_HDCP_CHECK_PERIOD_MS; int ret = -EINVAL; @@ -1972,6 +1957,11 @@ int intel_hdcp_enable(struct intel_connector *connector, u8 content_type) WARN_ON(hdcp->value == DRM_MODE_CONTENT_PROTECTION_ENABLED); hdcp->content_type = content_type; + if (INTEL_GEN(dev_priv) >= 12) { + hdcp->cpu_transcoder = cpu_transcoder; + hdcp->port_data.fw_tc = intel_get_mei_fw_tc(cpu_transcoder); + } + /* * Considering that HDCP2.2 is more secure than HDCP1.4, If the setup * is capable of HDCP2.2, it is preferred to use HDCP2.2. diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.h b/drivers/gpu/drm/i915/display/intel_hdcp.h index 41c1053d9e38..f3c3272e712a 100644 --- a/drivers/gpu/drm/i915/display/intel_hdcp.h +++ b/drivers/gpu/drm/i915/display/intel_hdcp.h @@ -21,11 +21,10 @@ enum transcoder; void intel_hdcp_atomic_check(struct drm_connector *connector, struct drm_connector_state *old_state, struct drm_connector_state *new_state); -void intel_hdcp_transcoder_config(struct intel_connector *connector, - enum transcoder cpu_transcoder); int intel_hdcp_init(struct intel_connector *connector, const struct intel_hdcp_shim *hdcp_shim); -int intel_hdcp_enable(struct intel_connector *connector, u8 content_type); +int intel_hdcp_enable(struct intel_connector *connector, + enum transcoder cpu_transcoder, u8 content_type); int intel_hdcp_disable(struct intel_connector *connector); bool is_hdcp_supported(struct drm_i915_private *dev_priv, enum port port); bool intel_hdcp_capable(struct intel_connector *connector); diff --git a/drivers/gpu/drm/i915/display/intel_hdmi.c b/drivers/gpu/drm/i915/display/intel_hdmi.c index f6f5312205c4..f56fffc474fa 100644 --- a/drivers/gpu/drm/i915/display/intel_hdmi.c +++ b/drivers/gpu/drm/i915/display/intel_hdmi.c @@ -2489,9 +2489,6 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder, return -EINVAL; } - intel_hdcp_transcoder_config(intel_hdmi->attached_connector, - pipe_config->cpu_transcoder); - return 0; } diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 9fdefbdc3546..75dd0e0367b7 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -845,12 +845,6 @@ static const u8 *reg_offsets(const struct intel_engine_cs *engine) } } -static void unwind_wa_tail(struct i915_request *rq) -{ - rq->tail = intel_ring_wrap(rq->ring, rq->wa_tail - WA_TAIL_BYTES); - assert_ring_tail_valid(rq->ring, rq->tail); -} - static struct i915_request * __unwind_incomplete_requests(struct intel_engine_cs *engine) { @@ -863,12 +857,10 @@ __unwind_incomplete_requests(struct intel_engine_cs *engine) list_for_each_entry_safe_reverse(rq, rn, &engine->active.requests, sched.link) { - if (i915_request_completed(rq)) continue; /* XXX */ __i915_request_unsubmit(rq); - unwind_wa_tail(rq); /* * Push the request back into the queue for later resubmission. @@ -1161,13 +1153,29 @@ execlists_schedule_out(struct i915_request *rq) i915_request_put(rq); } -static u64 execlists_update_context(const struct i915_request *rq) +static u64 execlists_update_context(struct i915_request *rq) { struct intel_context *ce = rq->hw_context; - u64 desc; + u64 desc = ce->lrc_desc; + u32 tail; - ce->lrc_reg_state[CTX_RING_TAIL] = - intel_ring_set_tail(rq->ring, rq->tail); + /* + * WaIdleLiteRestore:bdw,skl + * + * We should never submit the context with the same RING_TAIL twice + * just in case we submit an empty ring, which confuses the HW. + * + * We append a couple of NOOPs (gen8_emit_wa_tail) after the end of + * the normal request to be able to always advance the RING_TAIL on + * subsequent resubmissions (for lite restore). Should that fail us, + * and we try and submit the same tail again, force the context + * reload. + */ + tail = intel_ring_set_tail(rq->ring, rq->tail); + if (unlikely(ce->lrc_reg_state[CTX_RING_TAIL] == tail)) + desc |= CTX_DESC_FORCE_RESTORE; + ce->lrc_reg_state[CTX_RING_TAIL] = tail; + rq->tail = rq->wa_tail; /* * Make sure the context image is complete before we submit it to HW. @@ -1186,13 +1194,11 @@ static u64 execlists_update_context(const struct i915_request *rq) */ mb(); - desc = ce->lrc_desc; - ce->lrc_desc &= ~CTX_DESC_FORCE_RESTORE; - /* Wa_1607138340:tgl */ if (IS_TGL_REVID(rq->i915, TGL_REVID_A0, TGL_REVID_A0)) desc |= CTX_DESC_FORCE_RESTORE; + ce->lrc_desc &= ~CTX_DESC_FORCE_RESTORE; return desc; } @@ -1703,16 +1709,6 @@ static void execlists_dequeue(struct intel_engine_cs *engine) return; } - - /* - * WaIdleLiteRestore:bdw,skl - * Apply the wa NOOPs to prevent - * ring:HEAD == rq:TAIL as we resubmit the - * request. See gen8_emit_fini_breadcrumb() for - * where we prepare the padding after the - * end of the request. - */ - last->tail = last->wa_tail; } } @@ -4120,17 +4116,18 @@ static void virtual_context_destroy(struct kref *kref) for (n = 0; n < ve->num_siblings; n++) { struct intel_engine_cs *sibling = ve->siblings[n]; struct rb_node *node = &ve->nodes[sibling->id].rb; + unsigned long flags; if (RB_EMPTY_NODE(node)) continue; - spin_lock_irq(&sibling->active.lock); + spin_lock_irqsave(&sibling->active.lock, flags); /* Detachment is lazily performed in the execlists tasklet */ if (!RB_EMPTY_NODE(node)) rb_erase_cached(node, &sibling->execlists.virtual); - spin_unlock_irq(&sibling->active.lock); + spin_unlock_irqrestore(&sibling->active.lock, flags); } GEM_BUG_ON(__tasklet_is_scheduled(&ve->base.execlists.tasklet)); diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index b9eb6b3149b7..d034fa413164 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -45,6 +45,7 @@ #include "gem/i915_gem_context.h" #include "gem/i915_gem_ioctls.h" #include "gem/i915_gem_pm.h" +#include "gt/intel_context.h" #include "gt/intel_engine_user.h" #include "gt/intel_gt.h" #include "gt/intel_gt_pm.h" @@ -1053,6 +1054,18 @@ out: return err; } +static int __intel_context_flush_retire(struct intel_context *ce) +{ + struct intel_timeline *tl; + + tl = intel_context_timeline_lock(ce); + if (IS_ERR(tl)) + return PTR_ERR(tl); + + intel_context_timeline_unlock(tl); + return 0; +} + static int __intel_engines_record_defaults(struct intel_gt *gt) { struct i915_request *requests[I915_NUM_ENGINES] = {}; @@ -1121,13 +1134,20 @@ err_rq: if (!rq) continue; - /* We want to be able to unbind the state from the GGTT */ - GEM_BUG_ON(intel_context_is_pinned(rq->hw_context)); - + GEM_BUG_ON(!test_bit(CONTEXT_ALLOC_BIT, + &rq->hw_context->flags)); state = rq->hw_context->state; if (!state) continue; + /* Serialise with retirement on another CPU */ + err = __intel_context_flush_retire(rq->hw_context); + if (err) + goto out; + + /* We want to be able to unbind the state from the GGTT */ + GEM_BUG_ON(intel_context_is_pinned(rq->hw_context)); + /* * As we will hold a reference to the logical state, it will * not be torn down with the context, and importantly the diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index 65d7c2e599de..2ae14bc14931 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -2078,20 +2078,12 @@ gen8_update_reg_state_unlocked(const struct intel_context *ce, u32 *reg_state = ce->lrc_reg_state; int i; - if (IS_GEN(stream->perf->i915, 12)) { - u32 format = stream->oa_buffer.format; + reg_state[ctx_oactxctrl + 1] = + (stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | + (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) | + GEN8_OA_COUNTER_RESUME; - reg_state[ctx_oactxctrl + 1] = - (format << GEN12_OAR_OACONTROL_COUNTER_FORMAT_SHIFT) | - (stream->oa_config ? GEN12_OAR_OACONTROL_COUNTER_ENABLE : 0); - } else { - reg_state[ctx_oactxctrl + 1] = - (stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | - (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) | - GEN8_OA_COUNTER_RESUME; - } - - for (i = 0; !!ctx_flexeu0 && i < ARRAY_SIZE(flex_regs); i++) + for (i = 0; i < ARRAY_SIZE(flex_regs); i++) reg_state[ctx_flexeu0 + i * 2 + 1] = oa_config_flex_reg(stream->oa_config, flex_regs[i]); @@ -2224,34 +2216,51 @@ static int gen8_configure_context(struct i915_gem_context *ctx, return err; } -static int gen12_emit_oar_config(struct intel_context *ce, bool enable) +static int gen12_configure_oar_context(struct i915_perf_stream *stream, bool enable) { - struct i915_request *rq; - u32 *cs; - int err = 0; - - rq = i915_request_create(ce); - if (IS_ERR(rq)) - return PTR_ERR(rq); - - cs = intel_ring_begin(rq, 4); - if (IS_ERR(cs)) { - err = PTR_ERR(cs); - goto out; - } - - *cs++ = MI_LOAD_REGISTER_IMM(1); - *cs++ = i915_mmio_reg_offset(RING_CONTEXT_CONTROL(ce->engine->mmio_base)); - *cs++ = _MASKED_FIELD(GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE, - enable ? GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE : 0); - *cs++ = MI_NOOP; + int err; + struct intel_context *ce = stream->pinned_ctx; + u32 format = stream->oa_buffer.format; + struct flex regs_context[] = { + { + GEN8_OACTXCONTROL, + stream->perf->ctx_oactxctrl_offset + 1, + enable ? GEN8_OA_COUNTER_RESUME : 0, + }, + }; + /* Offsets in regs_lri are not used since this configuration is only + * applied using LRI. Initialize the correct offsets for posterity. + */ +#define GEN12_OAR_OACONTROL_OFFSET 0x5B0 + struct flex regs_lri[] = { + { + GEN12_OAR_OACONTROL, + GEN12_OAR_OACONTROL_OFFSET + 1, + (format << GEN12_OAR_OACONTROL_COUNTER_FORMAT_SHIFT) | + (enable ? GEN12_OAR_OACONTROL_COUNTER_ENABLE : 0) + }, + { + RING_CONTEXT_CONTROL(ce->engine->mmio_base), + CTX_CONTEXT_CONTROL, + _MASKED_FIELD(GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE, + enable ? + GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE : + 0) + }, + }; - intel_ring_advance(rq, cs); + /* Modify the context image of pinned context with regs_context*/ + err = intel_context_lock_pinned(ce); + if (err) + return err; -out: - i915_request_add(rq); + err = gen8_modify_context(ce, regs_context, ARRAY_SIZE(regs_context)); + intel_context_unlock_pinned(ce); + if (err) + return err; - return err; + /* Apply regs_lri using LRI with pinned context */ + return gen8_modify_self(ce, regs_lri, ARRAY_SIZE(regs_lri)); } /* @@ -2277,53 +2286,16 @@ out: * per-context OA state. * * Note: it's only the RCS/Render context that has any OA state. + * Note: the first flex register passed must always be R_PWR_CLK_STATE */ -static int lrc_configure_all_contexts(struct i915_perf_stream *stream, - const struct i915_oa_config *oa_config) +static int oa_configure_all_contexts(struct i915_perf_stream *stream, + struct flex *regs, + size_t num_regs) { struct drm_i915_private *i915 = stream->perf->i915; - /* The MMIO offsets for Flex EU registers aren't contiguous */ - const u32 ctx_flexeu0 = stream->perf->ctx_flexeu0_offset; -#define ctx_flexeuN(N) (ctx_flexeu0 + 2 * (N) + 1) - struct flex regs[] = { - { - GEN8_R_PWR_CLK_STATE, - CTX_R_PWR_CLK_STATE, - }, - { - IS_GEN(i915, 12) ? - GEN12_OAR_OACONTROL : GEN8_OACTXCONTROL, - stream->perf->ctx_oactxctrl_offset + 1, - }, - { EU_PERF_CNTL0, ctx_flexeuN(0) }, - { EU_PERF_CNTL1, ctx_flexeuN(1) }, - { EU_PERF_CNTL2, ctx_flexeuN(2) }, - { EU_PERF_CNTL3, ctx_flexeuN(3) }, - { EU_PERF_CNTL4, ctx_flexeuN(4) }, - { EU_PERF_CNTL5, ctx_flexeuN(5) }, - { EU_PERF_CNTL6, ctx_flexeuN(6) }, - }; -#undef ctx_flexeuN struct intel_engine_cs *engine; struct i915_gem_context *ctx, *cn; - size_t array_size = IS_GEN(i915, 12) ? 2 : ARRAY_SIZE(regs); - int i, err; - - if (IS_GEN(i915, 12)) { - u32 format = stream->oa_buffer.format; - - regs[1].value = - (format << GEN12_OAR_OACONTROL_COUNTER_FORMAT_SHIFT) | - (oa_config ? GEN12_OAR_OACONTROL_COUNTER_ENABLE : 0); - } else { - regs[1].value = - (stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | - (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) | - GEN8_OA_COUNTER_RESUME; - } - - for (i = 2; !!ctx_flexeu0 && i < array_size; i++) - regs[i].value = oa_config_flex_reg(oa_config, regs[i].reg); + int err; lockdep_assert_held(&stream->perf->lock); @@ -2353,7 +2325,7 @@ static int lrc_configure_all_contexts(struct i915_perf_stream *stream, spin_unlock(&i915->gem.contexts.lock); - err = gen8_configure_context(ctx, regs, array_size); + err = gen8_configure_context(ctx, regs, num_regs); if (err) { i915_gem_context_put(ctx); return err; @@ -2378,7 +2350,7 @@ static int lrc_configure_all_contexts(struct i915_perf_stream *stream, regs[0].value = intel_sseu_make_rpcs(i915, &ce->sseu); - err = gen8_modify_self(ce, regs, array_size); + err = gen8_modify_self(ce, regs, num_regs); if (err) return err; } @@ -2386,6 +2358,56 @@ static int lrc_configure_all_contexts(struct i915_perf_stream *stream, return 0; } +static int gen12_configure_all_contexts(struct i915_perf_stream *stream, + const struct i915_oa_config *oa_config) +{ + struct flex regs[] = { + { + GEN8_R_PWR_CLK_STATE, + CTX_R_PWR_CLK_STATE, + }, + }; + + return oa_configure_all_contexts(stream, regs, ARRAY_SIZE(regs)); +} + +static int lrc_configure_all_contexts(struct i915_perf_stream *stream, + const struct i915_oa_config *oa_config) +{ + /* The MMIO offsets for Flex EU registers aren't contiguous */ + const u32 ctx_flexeu0 = stream->perf->ctx_flexeu0_offset; +#define ctx_flexeuN(N) (ctx_flexeu0 + 2 * (N) + 1) + struct flex regs[] = { + { + GEN8_R_PWR_CLK_STATE, + CTX_R_PWR_CLK_STATE, + }, + { + GEN8_OACTXCONTROL, + stream->perf->ctx_oactxctrl_offset + 1, + }, + { EU_PERF_CNTL0, ctx_flexeuN(0) }, + { EU_PERF_CNTL1, ctx_flexeuN(1) }, + { EU_PERF_CNTL2, ctx_flexeuN(2) }, + { EU_PERF_CNTL3, ctx_flexeuN(3) }, + { EU_PERF_CNTL4, ctx_flexeuN(4) }, + { EU_PERF_CNTL5, ctx_flexeuN(5) }, + { EU_PERF_CNTL6, ctx_flexeuN(6) }, + }; +#undef ctx_flexeuN + int i; + + regs[1].value = + (stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | + (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) | + GEN8_OA_COUNTER_RESUME; + + for (i = 2; i < ARRAY_SIZE(regs); i++) + regs[i].value = oa_config_flex_reg(oa_config, regs[i].reg); + + return oa_configure_all_contexts(stream, regs, ARRAY_SIZE(regs)); +} + static int gen8_enable_metric_set(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; @@ -2464,7 +2486,7 @@ static int gen12_enable_metric_set(struct i915_perf_stream *stream) * to make sure all slices/subslices are ON before writing to NOA * registers. */ - ret = lrc_configure_all_contexts(stream, oa_config); + ret = gen12_configure_all_contexts(stream, oa_config); if (ret) return ret; @@ -2474,8 +2496,7 @@ static int gen12_enable_metric_set(struct i915_perf_stream *stream) * requested this. */ if (stream->ctx) { - ret = gen12_emit_oar_config(stream->pinned_ctx, - oa_config != NULL); + ret = gen12_configure_oar_context(stream, true); if (ret) return ret; } @@ -2509,11 +2530,11 @@ static void gen12_disable_metric_set(struct i915_perf_stream *stream) struct intel_uncore *uncore = stream->uncore; /* Reset all contexts' slices/subslices configurations. */ - lrc_configure_all_contexts(stream, NULL); + gen12_configure_all_contexts(stream, NULL); /* disable the context save/restore or OAR counters */ if (stream->ctx) - gen12_emit_oar_config(stream->pinned_ctx, false); + gen12_configure_oar_context(stream, false); /* Make sure we disable noa to save power. */ intel_uncore_rmw(uncore, RPM_CONFIG1, GEN10_GT_NOA_ENABLE, 0); @@ -2713,7 +2734,8 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, return -EINVAL; } - if (!(props->sample_flags & SAMPLE_OA_REPORT)) { + if (!(props->sample_flags & SAMPLE_OA_REPORT) && + (INTEL_GEN(perf->i915) < 12 || !stream->ctx)) { DRM_DEBUG("Only OA report sampling supported\n"); return -EINVAL; } @@ -2745,7 +2767,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, format_size = perf->oa_formats[props->oa_format].size; - stream->sample_flags |= SAMPLE_OA_REPORT; + stream->sample_flags = props->sample_flags; stream->sample_size += format_size; stream->oa_buffer.format_size = format_size; @@ -2854,7 +2876,11 @@ void i915_oa_init_reg_state(const struct intel_context *ce, return; stream = engine->i915->perf.exclusive_stream; - if (stream) + /* + * For gen12, only CTX_R_PWR_CLK_STATE needs update, but the caller + * is already doing that, so nothing to be done for gen12 here. + */ + if (stream && INTEL_GEN(stream->perf->i915) < 12) gen8_update_reg_state_unlocked(ce, stream); } diff --git a/drivers/gpu/drm/mcde/mcde_dsi.c b/drivers/gpu/drm/mcde/mcde_dsi.c index d6214d3c8b33..ef4c630afe3f 100644 --- a/drivers/gpu/drm/mcde/mcde_dsi.c +++ b/drivers/gpu/drm/mcde/mcde_dsi.c @@ -935,11 +935,13 @@ static int mcde_dsi_bind(struct device *dev, struct device *master, for_each_available_child_of_node(dev->of_node, child) { panel = of_drm_find_panel(child); if (IS_ERR(panel)) { - dev_err(dev, "failed to find panel try bridge (%lu)\n", + dev_err(dev, "failed to find panel try bridge (%ld)\n", PTR_ERR(panel)); + panel = NULL; + bridge = of_drm_find_bridge(child); if (IS_ERR(bridge)) { - dev_err(dev, "failed to find bridge (%lu)\n", + dev_err(dev, "failed to find bridge (%ld)\n", PTR_ERR(bridge)); return PTR_ERR(bridge); } diff --git a/drivers/gpu/drm/meson/meson_venc_cvbs.c b/drivers/gpu/drm/meson/meson_venc_cvbs.c index 9ab27aecfcf3..1bd6b6d15ffb 100644 --- a/drivers/gpu/drm/meson/meson_venc_cvbs.c +++ b/drivers/gpu/drm/meson/meson_venc_cvbs.c @@ -64,6 +64,25 @@ struct meson_cvbs_mode meson_cvbs_modes[MESON_CVBS_MODES_COUNT] = { }, }; +static const struct meson_cvbs_mode * +meson_cvbs_get_mode(const struct drm_display_mode *req_mode) +{ + int i; + + for (i = 0; i < MESON_CVBS_MODES_COUNT; ++i) { + struct meson_cvbs_mode *meson_mode = &meson_cvbs_modes[i]; + + if (drm_mode_match(req_mode, &meson_mode->mode, + DRM_MODE_MATCH_TIMINGS | + DRM_MODE_MATCH_CLOCK | + DRM_MODE_MATCH_FLAGS | + DRM_MODE_MATCH_3D_FLAGS)) + return meson_mode; + } + + return NULL; +} + /* Connector */ static void meson_cvbs_connector_destroy(struct drm_connector *connector) @@ -136,14 +155,8 @@ static int meson_venc_cvbs_encoder_atomic_check(struct drm_encoder *encoder, struct drm_crtc_state *crtc_state, struct drm_connector_state *conn_state) { - int i; - - for (i = 0; i < MESON_CVBS_MODES_COUNT; ++i) { - struct meson_cvbs_mode *meson_mode = &meson_cvbs_modes[i]; - - if (drm_mode_equal(&crtc_state->mode, &meson_mode->mode)) - return 0; - } + if (meson_cvbs_get_mode(&crtc_state->mode)) + return 0; return -EINVAL; } @@ -191,24 +204,17 @@ static void meson_venc_cvbs_encoder_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) { + const struct meson_cvbs_mode *meson_mode = meson_cvbs_get_mode(mode); struct meson_venc_cvbs *meson_venc_cvbs = encoder_to_meson_venc_cvbs(encoder); struct meson_drm *priv = meson_venc_cvbs->priv; - int i; - for (i = 0; i < MESON_CVBS_MODES_COUNT; ++i) { - struct meson_cvbs_mode *meson_mode = &meson_cvbs_modes[i]; + if (meson_mode) { + meson_venci_cvbs_mode_set(priv, meson_mode->enci); - if (drm_mode_equal(mode, &meson_mode->mode)) { - meson_venci_cvbs_mode_set(priv, - meson_mode->enci); - - /* Setup 27MHz vclk2 for ENCI and VDAC */ - meson_vclk_setup(priv, MESON_VCLK_TARGET_CVBS, - MESON_VCLK_CVBS, MESON_VCLK_CVBS, - MESON_VCLK_CVBS, true); - break; - } + /* Setup 27MHz vclk2 for ENCI and VDAC */ + meson_vclk_setup(priv, MESON_VCLK_TARGET_CVBS, MESON_VCLK_CVBS, + MESON_VCLK_CVBS, MESON_VCLK_CVBS, true); } } diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.c b/drivers/gpu/drm/mgag200/mgag200_drv.c index d43951caeea0..b113876c2428 100644 --- a/drivers/gpu/drm/mgag200/mgag200_drv.c +++ b/drivers/gpu/drm/mgag200/mgag200_drv.c @@ -30,9 +30,8 @@ module_param_named(modeset, mgag200_modeset, int, 0400); static struct drm_driver driver; static const struct pci_device_id pciidlist[] = { - { PCI_VENDOR_ID_MATROX, 0x522, PCI_VENDOR_ID_SUN, 0x4852, 0, 0, + { PCI_VENDOR_ID_MATROX, 0x522, PCI_ANY_ID, PCI_ANY_ID, 0, 0, G200_SE_A | MGAG200_FLAG_HW_BUG_NO_STARTADD}, - { PCI_VENDOR_ID_MATROX, 0x522, PCI_ANY_ID, PCI_ANY_ID, 0, 0, G200_SE_A }, { PCI_VENDOR_ID_MATROX, 0x524, PCI_ANY_ID, PCI_ANY_ID, 0, 0, G200_SE_B }, { PCI_VENDOR_ID_MATROX, 0x530, PCI_ANY_ID, PCI_ANY_ID, 0, 0, G200_EV }, { PCI_VENDOR_ID_MATROX, 0x532, PCI_ANY_ID, PCI_ANY_ID, 0, 0, G200_WB }, diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h index 43df86c38f58..24f7700768da 100644 --- a/drivers/gpu/drm/nouveau/dispnv50/atom.h +++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h @@ -114,6 +114,7 @@ struct nv50_head_atom { u8 nhsync:1; u8 nvsync:1; u8 depth:4; + u8 bpc; } or; /* Currently only used for MST */ diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c index 549486f1d937..63425e246018 100644 --- a/drivers/gpu/drm/nouveau/dispnv50/disp.c +++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c @@ -326,9 +326,9 @@ nv50_outp_atomic_check_view(struct drm_encoder *encoder, * same size as the native one (e.g. different * refresh rate) */ - if (adjusted_mode->hdisplay == native_mode->hdisplay && - adjusted_mode->vdisplay == native_mode->vdisplay && - adjusted_mode->type & DRM_MODE_TYPE_DRIVER) + if (mode->hdisplay == native_mode->hdisplay && + mode->vdisplay == native_mode->vdisplay && + mode->type & DRM_MODE_TYPE_DRIVER) break; mode = native_mode; asyc->scaler.full = true; @@ -353,10 +353,20 @@ nv50_outp_atomic_check(struct drm_encoder *encoder, struct drm_crtc_state *crtc_state, struct drm_connector_state *conn_state) { - struct nouveau_connector *nv_connector = - nouveau_connector(conn_state->connector); - return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state, - nv_connector->native_mode); + struct drm_connector *connector = conn_state->connector; + struct nouveau_connector *nv_connector = nouveau_connector(connector); + struct nv50_head_atom *asyh = nv50_head_atom(crtc_state); + int ret; + + ret = nv50_outp_atomic_check_view(encoder, crtc_state, conn_state, + nv_connector->native_mode); + if (ret) + return ret; + + if (crtc_state->mode_changed || crtc_state->connectors_changed) + asyh->or.bpc = connector->display_info.bpc; + + return 0; } /****************************************************************************** @@ -770,32 +780,54 @@ nv50_msto_atomic_check(struct drm_encoder *encoder, struct nv50_mstm *mstm = mstc->mstm; struct nv50_head_atom *asyh = nv50_head_atom(crtc_state); int slots; + int ret; + + ret = nv50_outp_atomic_check_view(encoder, crtc_state, conn_state, + mstc->native); + if (ret) + return ret; + + if (!crtc_state->mode_changed && !crtc_state->connectors_changed) + return 0; + + /* + * When restoring duplicated states, we need to make sure that the bw + * remains the same and avoid recalculating it, as the connector's bpc + * may have changed after the state was duplicated + */ + if (!state->duplicated) { + const int clock = crtc_state->adjusted_mode.clock; - if (crtc_state->mode_changed || crtc_state->connectors_changed) { /* - * When restoring duplicated states, we need to make sure that - * the bw remains the same and avoid recalculating it, as the - * connector's bpc may have changed after the state was - * duplicated + * XXX: Since we don't use HDR in userspace quite yet, limit + * the bpc to 8 to save bandwidth on the topology. In the + * future, we'll want to properly fix this by dynamically + * selecting the highest possible bpc that would fit in the + * topology */ - if (!state->duplicated) { - const int bpp = connector->display_info.bpc * 3; - const int clock = crtc_state->adjusted_mode.clock; + asyh->or.bpc = min(connector->display_info.bpc, 8U); + asyh->dp.pbn = drm_dp_calc_pbn_mode(clock, asyh->or.bpc * 3); + } - asyh->dp.pbn = drm_dp_calc_pbn_mode(clock, bpp); - } + slots = drm_dp_atomic_find_vcpi_slots(state, &mstm->mgr, mstc->port, + asyh->dp.pbn); + if (slots < 0) + return slots; - slots = drm_dp_atomic_find_vcpi_slots(state, &mstm->mgr, - mstc->port, - asyh->dp.pbn); - if (slots < 0) - return slots; + asyh->dp.tu = slots; - asyh->dp.tu = slots; - } + return 0; +} - return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state, - mstc->native); +static u8 +nv50_dp_bpc_to_depth(unsigned int bpc) +{ + switch (bpc) { + case 6: return 0x2; + case 8: return 0x5; + case 10: /* fall-through */ + default: return 0x6; + } } static void @@ -808,7 +840,7 @@ nv50_msto_enable(struct drm_encoder *encoder) struct nv50_mstm *mstm = NULL; struct drm_connector *connector; struct drm_connector_list_iter conn_iter; - u8 proto, depth; + u8 proto; bool r; drm_connector_list_iter_begin(encoder->dev, &conn_iter); @@ -837,14 +869,8 @@ nv50_msto_enable(struct drm_encoder *encoder) else proto = 0x9; - switch (mstc->connector.display_info.bpc) { - case 6: depth = 0x2; break; - case 8: depth = 0x5; break; - case 10: - default: depth = 0x6; break; - } - - mstm->outp->update(mstm->outp, head->base.index, armh, proto, depth); + mstm->outp->update(mstm->outp, head->base.index, armh, proto, + nv50_dp_bpc_to_depth(armh->or.bpc)); msto->head = head; msto->mstc = mstc; @@ -1498,20 +1524,14 @@ nv50_sor_enable(struct drm_encoder *encoder) lvds.lvds.script |= 0x0200; } - if (nv_connector->base.display_info.bpc == 8) + if (asyh->or.bpc == 8) lvds.lvds.script |= 0x0200; } nvif_mthd(&disp->disp->object, 0, &lvds, sizeof(lvds)); break; case DCB_OUTPUT_DP: - if (nv_connector->base.display_info.bpc == 6) - depth = 0x2; - else - if (nv_connector->base.display_info.bpc == 8) - depth = 0x5; - else - depth = 0x6; + depth = nv50_dp_bpc_to_depth(asyh->or.bpc); if (nv_encoder->link & 1) proto = 0x8; @@ -1662,7 +1682,7 @@ nv50_pior_enable(struct drm_encoder *encoder) nv50_outp_acquire(nv_encoder); nv_connector = nouveau_encoder_connector_get(nv_encoder); - switch (nv_connector->base.display_info.bpc) { + switch (asyh->or.bpc) { case 10: asyh->or.depth = 0x6; break; case 8: asyh->or.depth = 0x5; break; case 6: asyh->or.depth = 0x2; break; diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.c b/drivers/gpu/drm/nouveau/dispnv50/head.c index 71c23bf1fe25..c9692df2b76c 100644 --- a/drivers/gpu/drm/nouveau/dispnv50/head.c +++ b/drivers/gpu/drm/nouveau/dispnv50/head.c @@ -81,18 +81,17 @@ nv50_head_atomic_check_dither(struct nv50_head_atom *armh, struct nv50_head_atom *asyh, struct nouveau_conn_atom *asyc) { - struct drm_connector *connector = asyc->state.connector; u32 mode = 0x00; if (asyc->dither.mode == DITHERING_MODE_AUTO) { - if (asyh->base.depth > connector->display_info.bpc * 3) + if (asyh->base.depth > asyh->or.bpc * 3) mode = DITHERING_MODE_DYNAMIC2X2; } else { mode = asyc->dither.mode; } if (asyc->dither.depth == DITHERING_DEPTH_AUTO) { - if (connector->display_info.bpc >= 8) + if (asyh->or.bpc >= 8) mode |= DITHERING_DEPTH_8BPC; } else { mode |= asyc->dither.depth; diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c index 5b413588b823..9a9a7f5003d3 100644 --- a/drivers/gpu/drm/nouveau/nouveau_connector.c +++ b/drivers/gpu/drm/nouveau/nouveau_connector.c @@ -245,14 +245,22 @@ nouveau_conn_atomic_duplicate_state(struct drm_connector *connector) void nouveau_conn_reset(struct drm_connector *connector) { + struct nouveau_connector *nv_connector = nouveau_connector(connector); struct nouveau_conn_atom *asyc; - if (WARN_ON(!(asyc = kzalloc(sizeof(*asyc), GFP_KERNEL)))) - return; + if (drm_drv_uses_atomic_modeset(connector->dev)) { + if (WARN_ON(!(asyc = kzalloc(sizeof(*asyc), GFP_KERNEL)))) + return; + + if (connector->state) + nouveau_conn_atomic_destroy_state(connector, + connector->state); + + __drm_atomic_helper_connector_reset(connector, &asyc->state); + } else { + asyc = &nv_connector->properties_state; + } - if (connector->state) - nouveau_conn_atomic_destroy_state(connector, connector->state); - __drm_atomic_helper_connector_reset(connector, &asyc->state); asyc->dither.mode = DITHERING_MODE_AUTO; asyc->dither.depth = DITHERING_DEPTH_AUTO; asyc->scaler.mode = DRM_MODE_SCALE_NONE; @@ -276,8 +284,14 @@ void nouveau_conn_attach_properties(struct drm_connector *connector) { struct drm_device *dev = connector->dev; - struct nouveau_conn_atom *armc = nouveau_conn_atom(connector->state); struct nouveau_display *disp = nouveau_display(dev); + struct nouveau_connector *nv_connector = nouveau_connector(connector); + struct nouveau_conn_atom *armc; + + if (drm_drv_uses_atomic_modeset(connector->dev)) + armc = nouveau_conn_atom(connector->state); + else + armc = &nv_connector->properties_state; /* Init DVI-I specific properties. */ if (connector->connector_type == DRM_MODE_CONNECTOR_DVII) @@ -748,9 +762,9 @@ static int nouveau_connector_set_property(struct drm_connector *connector, struct drm_property *property, uint64_t value) { - struct nouveau_conn_atom *asyc = nouveau_conn_atom(connector->state); struct nouveau_connector *nv_connector = nouveau_connector(connector); struct nouveau_encoder *nv_encoder = nv_connector->detected_encoder; + struct nouveau_conn_atom *asyc = &nv_connector->properties_state; struct drm_encoder *encoder = to_drm_encoder(nv_encoder); int ret; diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.h b/drivers/gpu/drm/nouveau/nouveau_connector.h index f43a8d63aef8..de84fb4708c7 100644 --- a/drivers/gpu/drm/nouveau/nouveau_connector.h +++ b/drivers/gpu/drm/nouveau/nouveau_connector.h @@ -29,6 +29,7 @@ #include <nvif/notify.h> +#include <drm/drm_crtc.h> #include <drm/drm_edid.h> #include <drm/drm_encoder.h> #include <drm/drm_dp_helper.h> @@ -44,6 +45,60 @@ struct dcb_output; struct nouveau_backlight; #endif +#define nouveau_conn_atom(p) \ + container_of((p), struct nouveau_conn_atom, state) + +struct nouveau_conn_atom { + struct drm_connector_state state; + + struct { + /* The enum values specifically defined here match nv50/gf119 + * hw values, and the code relies on this. + */ + enum { + DITHERING_MODE_OFF = 0x00, + DITHERING_MODE_ON = 0x01, + DITHERING_MODE_DYNAMIC2X2 = 0x10 | DITHERING_MODE_ON, + DITHERING_MODE_STATIC2X2 = 0x18 | DITHERING_MODE_ON, + DITHERING_MODE_TEMPORAL = 0x20 | DITHERING_MODE_ON, + DITHERING_MODE_AUTO + } mode; + enum { + DITHERING_DEPTH_6BPC = 0x00, + DITHERING_DEPTH_8BPC = 0x02, + DITHERING_DEPTH_AUTO + } depth; + } dither; + + struct { + int mode; /* DRM_MODE_SCALE_* */ + struct { + enum { + UNDERSCAN_OFF, + UNDERSCAN_ON, + UNDERSCAN_AUTO, + } mode; + u32 hborder; + u32 vborder; + } underscan; + bool full; + } scaler; + + struct { + int color_vibrance; + int vibrant_hue; + } procamp; + + union { + struct { + bool dither:1; + bool scaler:1; + bool procamp:1; + }; + u8 mask; + } set; +}; + struct nouveau_connector { struct drm_connector base; enum dcb_connector_type type; @@ -63,6 +118,12 @@ struct nouveau_connector { #ifdef CONFIG_DRM_NOUVEAU_BACKLIGHT struct nouveau_backlight *backlight; #endif + /* + * Our connector property code expects a nouveau_conn_atom struct + * even on pre-nv50 where we do not support atomic. This embedded + * version gets used in the non atomic modeset case. + */ + struct nouveau_conn_atom properties_state; }; static inline struct nouveau_connector *nouveau_connector( @@ -121,61 +182,6 @@ extern int nouveau_ignorelid; extern int nouveau_duallink; extern int nouveau_hdmimhz; -#include <drm/drm_crtc.h> -#define nouveau_conn_atom(p) \ - container_of((p), struct nouveau_conn_atom, state) - -struct nouveau_conn_atom { - struct drm_connector_state state; - - struct { - /* The enum values specifically defined here match nv50/gf119 - * hw values, and the code relies on this. - */ - enum { - DITHERING_MODE_OFF = 0x00, - DITHERING_MODE_ON = 0x01, - DITHERING_MODE_DYNAMIC2X2 = 0x10 | DITHERING_MODE_ON, - DITHERING_MODE_STATIC2X2 = 0x18 | DITHERING_MODE_ON, - DITHERING_MODE_TEMPORAL = 0x20 | DITHERING_MODE_ON, - DITHERING_MODE_AUTO - } mode; - enum { - DITHERING_DEPTH_6BPC = 0x00, - DITHERING_DEPTH_8BPC = 0x02, - DITHERING_DEPTH_AUTO - } depth; - } dither; - - struct { - int mode; /* DRM_MODE_SCALE_* */ - struct { - enum { - UNDERSCAN_OFF, - UNDERSCAN_ON, - UNDERSCAN_AUTO, - } mode; - u32 hborder; - u32 vborder; - } underscan; - bool full; - } scaler; - - struct { - int color_vibrance; - int vibrant_hue; - } procamp; - - union { - struct { - bool dither:1; - bool scaler:1; - bool procamp:1; - }; - u8 mask; - } set; -}; - void nouveau_conn_attach_properties(struct drm_connector *); void nouveau_conn_reset(struct drm_connector *); struct drm_connector_state * diff --git a/drivers/gpu/drm/panfrost/panfrost_devfreq.c b/drivers/gpu/drm/panfrost/panfrost_devfreq.c index 4c4e8a30a1ac..536ba93b0f46 100644 --- a/drivers/gpu/drm/panfrost/panfrost_devfreq.c +++ b/drivers/gpu/drm/panfrost/panfrost_devfreq.c @@ -18,15 +18,18 @@ static void panfrost_devfreq_update_utilization(struct panfrost_device *pfdev); static int panfrost_devfreq_target(struct device *dev, unsigned long *freq, u32 flags) { - struct panfrost_device *pfdev = dev_get_drvdata(dev); + struct dev_pm_opp *opp; int err; + opp = devfreq_recommended_opp(dev, freq, flags); + if (IS_ERR(opp)) + return PTR_ERR(opp); + dev_pm_opp_put(opp); + err = dev_pm_opp_set_rate(dev, *freq); if (err) return err; - *freq = clk_get_rate(pfdev->clock); - return 0; } @@ -60,20 +63,10 @@ static int panfrost_devfreq_get_dev_status(struct device *dev, return 0; } -static int panfrost_devfreq_get_cur_freq(struct device *dev, unsigned long *freq) -{ - struct panfrost_device *pfdev = platform_get_drvdata(to_platform_device(dev)); - - *freq = clk_get_rate(pfdev->clock); - - return 0; -} - static struct devfreq_dev_profile panfrost_devfreq_profile = { .polling_ms = 50, /* ~3 frames */ .target = panfrost_devfreq_target, .get_dev_status = panfrost_devfreq_get_dev_status, - .get_cur_freq = panfrost_devfreq_get_cur_freq, }; int panfrost_devfreq_init(struct panfrost_device *pfdev) diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c b/drivers/gpu/drm/panfrost/panfrost_drv.c index 9458dc6c750c..f61364f7c471 100644 --- a/drivers/gpu/drm/panfrost/panfrost_drv.c +++ b/drivers/gpu/drm/panfrost/panfrost_drv.c @@ -303,14 +303,17 @@ static int panfrost_ioctl_mmap_bo(struct drm_device *dev, void *data, } /* Don't allow mmapping of heap objects as pages are not pinned. */ - if (to_panfrost_bo(gem_obj)->is_heap) - return -EINVAL; + if (to_panfrost_bo(gem_obj)->is_heap) { + ret = -EINVAL; + goto out; + } ret = drm_gem_create_mmap_offset(gem_obj); if (ret == 0) args->offset = drm_vma_node_offset_addr(&gem_obj->vma_node); - drm_gem_object_put_unlocked(gem_obj); +out: + drm_gem_object_put_unlocked(gem_obj); return ret; } @@ -347,20 +350,19 @@ static int panfrost_ioctl_madvise(struct drm_device *dev, void *data, return -ENOENT; } + mutex_lock(&pfdev->shrinker_lock); args->retained = drm_gem_shmem_madvise(gem_obj, args->madv); if (args->retained) { struct panfrost_gem_object *bo = to_panfrost_bo(gem_obj); - mutex_lock(&pfdev->shrinker_lock); - if (args->madv == PANFROST_MADV_DONTNEED) - list_add_tail(&bo->base.madv_list, &pfdev->shrinker_list); + list_add_tail(&bo->base.madv_list, + &pfdev->shrinker_list); else if (args->madv == PANFROST_MADV_WILLNEED) list_del_init(&bo->base.madv_list); - - mutex_unlock(&pfdev->shrinker_lock); } + mutex_unlock(&pfdev->shrinker_lock); drm_gem_object_put_unlocked(gem_obj); return 0; @@ -443,7 +445,7 @@ panfrost_postclose(struct drm_device *dev, struct drm_file *file) { struct panfrost_file_priv *panfrost_priv = file->driver_priv; - panfrost_perfcnt_close(panfrost_priv); + panfrost_perfcnt_close(file); panfrost_job_close(panfrost_priv); panfrost_mmu_pgtable_free(panfrost_priv); diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.c b/drivers/gpu/drm/panfrost/panfrost_gem.c index deca0c30bbd4..fd766b1395fb 100644 --- a/drivers/gpu/drm/panfrost/panfrost_gem.c +++ b/drivers/gpu/drm/panfrost/panfrost_gem.c @@ -19,6 +19,16 @@ static void panfrost_gem_free_object(struct drm_gem_object *obj) struct panfrost_gem_object *bo = to_panfrost_bo(obj); struct panfrost_device *pfdev = obj->dev->dev_private; + /* + * Make sure the BO is no longer inserted in the shrinker list before + * taking care of the destruction itself. If we don't do that we have a + * race condition between this function and what's done in + * panfrost_gem_shrinker_scan(). + */ + mutex_lock(&pfdev->shrinker_lock); + list_del_init(&bo->base.madv_list); + mutex_unlock(&pfdev->shrinker_lock); + if (bo->sgts) { int i; int n_sgt = bo->base.base.size / SZ_2M; @@ -33,15 +43,10 @@ static void panfrost_gem_free_object(struct drm_gem_object *obj) kfree(bo->sgts); } - mutex_lock(&pfdev->shrinker_lock); - if (!list_empty(&bo->base.madv_list)) - list_del(&bo->base.madv_list); - mutex_unlock(&pfdev->shrinker_lock); - drm_gem_shmem_free_object(obj); } -static int panfrost_gem_open(struct drm_gem_object *obj, struct drm_file *file_priv) +int panfrost_gem_open(struct drm_gem_object *obj, struct drm_file *file_priv) { int ret; size_t size = obj->size; @@ -80,7 +85,7 @@ static int panfrost_gem_open(struct drm_gem_object *obj, struct drm_file *file_p return ret; } -static void panfrost_gem_close(struct drm_gem_object *obj, struct drm_file *file_priv) +void panfrost_gem_close(struct drm_gem_object *obj, struct drm_file *file_priv) { struct panfrost_gem_object *bo = to_panfrost_bo(obj); struct panfrost_file_priv *priv = file_priv->driver_priv; diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.h b/drivers/gpu/drm/panfrost/panfrost_gem.h index 50920819cc16..4b17e7308764 100644 --- a/drivers/gpu/drm/panfrost/panfrost_gem.h +++ b/drivers/gpu/drm/panfrost/panfrost_gem.h @@ -45,6 +45,10 @@ panfrost_gem_create_with_handle(struct drm_file *file_priv, u32 flags, uint32_t *handle); +int panfrost_gem_open(struct drm_gem_object *obj, struct drm_file *file_priv); +void panfrost_gem_close(struct drm_gem_object *obj, + struct drm_file *file_priv); + void panfrost_gem_shrinker_init(struct drm_device *dev); void panfrost_gem_shrinker_cleanup(struct drm_device *dev); diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c index 2dba192bf198..2c04e858c50a 100644 --- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c +++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c @@ -67,9 +67,10 @@ static int panfrost_perfcnt_dump_locked(struct panfrost_device *pfdev) } static int panfrost_perfcnt_enable_locked(struct panfrost_device *pfdev, - struct panfrost_file_priv *user, + struct drm_file *file_priv, unsigned int counterset) { + struct panfrost_file_priv *user = file_priv->driver_priv; struct panfrost_perfcnt *perfcnt = pfdev->perfcnt; struct drm_gem_shmem_object *bo; u32 cfg; @@ -91,14 +92,14 @@ static int panfrost_perfcnt_enable_locked(struct panfrost_device *pfdev, perfcnt->bo = to_panfrost_bo(&bo->base); /* Map the perfcnt buf in the address space attached to file_priv. */ - ret = panfrost_mmu_map(perfcnt->bo); + ret = panfrost_gem_open(&perfcnt->bo->base.base, file_priv); if (ret) goto err_put_bo; perfcnt->buf = drm_gem_shmem_vmap(&bo->base); if (IS_ERR(perfcnt->buf)) { ret = PTR_ERR(perfcnt->buf); - goto err_put_bo; + goto err_close_bo; } /* @@ -157,14 +158,17 @@ static int panfrost_perfcnt_enable_locked(struct panfrost_device *pfdev, err_vunmap: drm_gem_shmem_vunmap(&perfcnt->bo->base.base, perfcnt->buf); +err_close_bo: + panfrost_gem_close(&perfcnt->bo->base.base, file_priv); err_put_bo: drm_gem_object_put_unlocked(&bo->base); return ret; } static int panfrost_perfcnt_disable_locked(struct panfrost_device *pfdev, - struct panfrost_file_priv *user) + struct drm_file *file_priv) { + struct panfrost_file_priv *user = file_priv->driver_priv; struct panfrost_perfcnt *perfcnt = pfdev->perfcnt; if (user != perfcnt->user) @@ -180,6 +184,7 @@ static int panfrost_perfcnt_disable_locked(struct panfrost_device *pfdev, perfcnt->user = NULL; drm_gem_shmem_vunmap(&perfcnt->bo->base.base, perfcnt->buf); perfcnt->buf = NULL; + panfrost_gem_close(&perfcnt->bo->base.base, file_priv); drm_gem_object_put_unlocked(&perfcnt->bo->base.base); perfcnt->bo = NULL; pm_runtime_mark_last_busy(pfdev->dev); @@ -191,7 +196,6 @@ static int panfrost_perfcnt_disable_locked(struct panfrost_device *pfdev, int panfrost_ioctl_perfcnt_enable(struct drm_device *dev, void *data, struct drm_file *file_priv) { - struct panfrost_file_priv *pfile = file_priv->driver_priv; struct panfrost_device *pfdev = dev->dev_private; struct panfrost_perfcnt *perfcnt = pfdev->perfcnt; struct drm_panfrost_perfcnt_enable *req = data; @@ -207,10 +211,10 @@ int panfrost_ioctl_perfcnt_enable(struct drm_device *dev, void *data, mutex_lock(&perfcnt->lock); if (req->enable) - ret = panfrost_perfcnt_enable_locked(pfdev, pfile, + ret = panfrost_perfcnt_enable_locked(pfdev, file_priv, req->counterset); else - ret = panfrost_perfcnt_disable_locked(pfdev, pfile); + ret = panfrost_perfcnt_disable_locked(pfdev, file_priv); mutex_unlock(&perfcnt->lock); return ret; @@ -248,15 +252,16 @@ out: return ret; } -void panfrost_perfcnt_close(struct panfrost_file_priv *pfile) +void panfrost_perfcnt_close(struct drm_file *file_priv) { + struct panfrost_file_priv *pfile = file_priv->driver_priv; struct panfrost_device *pfdev = pfile->pfdev; struct panfrost_perfcnt *perfcnt = pfdev->perfcnt; pm_runtime_get_sync(pfdev->dev); mutex_lock(&perfcnt->lock); if (perfcnt->user == pfile) - panfrost_perfcnt_disable_locked(pfdev, pfile); + panfrost_perfcnt_disable_locked(pfdev, file_priv); mutex_unlock(&perfcnt->lock); pm_runtime_mark_last_busy(pfdev->dev); pm_runtime_put_autosuspend(pfdev->dev); diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.h b/drivers/gpu/drm/panfrost/panfrost_perfcnt.h index 13b8fdaa1b43..8bbcf5f5fb33 100644 --- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.h +++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.h @@ -9,7 +9,7 @@ void panfrost_perfcnt_sample_done(struct panfrost_device *pfdev); void panfrost_perfcnt_clean_cache_done(struct panfrost_device *pfdev); int panfrost_perfcnt_init(struct panfrost_device *pfdev); void panfrost_perfcnt_fini(struct panfrost_device *pfdev); -void panfrost_perfcnt_close(struct panfrost_file_priv *pfile); +void panfrost_perfcnt_close(struct drm_file *file_priv); int panfrost_ioctl_perfcnt_enable(struct drm_device *dev, void *data, struct drm_file *file_priv); int panfrost_ioctl_perfcnt_dump(struct drm_device *dev, void *data, |