From: Alex Deucher Date: Tue, 25 Jun 2019 13:42:25 +0000 (-0500) Subject: Merge branch 'drm-next' into drm-next-5.3 X-Git-Url: http://git.maquefel.me/?a=commitdiff_plain;h=d7929c1e13e3788e7cb741d75b5baec5e53eff21;p=linux.git Merge branch 'drm-next' into drm-next-5.3 Backmerge drm-next and fix up conflicts due to drmP.h removal. Signed-off-by: Alex Deucher --- d7929c1e13e3788e7cb741d75b5baec5e53eff21 diff --cc drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index c604f1504d3e7,9f32bf862d949..2e12eeb314a75 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@@ -339,10 -318,10 +340,10 @@@ int amdgpu_vcn_dec_ring_test_ring(struc amdgpu_ring_write(ring, 0xDEADBEEF); amdgpu_ring_commit(ring); for (i = 0; i < adev->usec_timeout; i++) { - tmp = RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_SCRATCH9)); + tmp = RREG32(adev->vcn.external.scratch9); if (tmp == 0xDEADBEEF) break; - DRM_UDELAY(1); + udelay(1); } if (i >= adev->usec_timeout) @@@ -660,10 -641,10 +661,10 @@@ int amdgpu_vcn_jpeg_ring_test_ring(stru amdgpu_ring_commit(ring); for (i = 0; i < adev->usec_timeout; i++) { - tmp = RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_SCRATCH9)); + tmp = RREG32(adev->vcn.external.jpeg_pitch); if (tmp == 0xDEADBEEF) break; - DRM_UDELAY(1); + udelay(1); } if (i >= adev->usec_timeout) @@@ -734,10 -715,10 +735,10 @@@ int amdgpu_vcn_jpeg_ring_test_ib(struc } for (i = 0; i < adev->usec_timeout; i++) { - tmp = RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_SCRATCH9)); + tmp = RREG32(adev->vcn.external.jpeg_pitch); if (tmp == 0xDEADBEEF) break; - DRM_UDELAY(1); + udelay(1); } if (i >= adev->usec_timeout) diff --cc drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c index acdd1868d010e,0000000000000..cec7c1fb14bf4 mode 100644,000000..100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c @@@ -1,916 -1,0 +1,917 @@@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include ++#include +#include "amdgpu.h" +#include "amdgpu_atomfirmware.h" +#include "gmc_v10_0.h" + +#include "hdp/hdp_5_0_0_offset.h" +#include "hdp/hdp_5_0_0_sh_mask.h" +#include "gc/gc_10_1_0_sh_mask.h" +#include "mmhub/mmhub_2_0_0_sh_mask.h" +#include "dcn/dcn_2_0_0_offset.h" +#include "dcn/dcn_2_0_0_sh_mask.h" +#include "oss/osssys_5_0_0_offset.h" +#include "ivsrcid/vmc/irqsrcs_vmc_1_0.h" +#include "navi10_enum.h" + +#include "soc15.h" +#include "soc15_common.h" + +#include "nbio_v2_3.h" + +#include "gfxhub_v2_0.h" +#include "mmhub_v2_0.h" +#include "athub_v2_0.h" +/* XXX Move this macro to navi10 header file, which is like vid.h for VI.*/ +#define AMDGPU_NUM_OF_VMIDS 8 + +#if 0 +static const struct soc15_reg_golden golden_settings_navi10_hdp[] = +{ + /* TODO add golden setting for hdp */ +}; +#endif + +static int +gmc_v10_0_vm_fault_interrupt_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, unsigned type, + enum amdgpu_interrupt_state state) +{ + struct amdgpu_vmhub *hub; + u32 tmp, reg, bits[AMDGPU_MAX_VMHUBS], i; + + bits[AMDGPU_GFXHUB] = GCVM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + GCVM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + GCVM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + GCVM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + GCVM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + GCVM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + GCVM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK; + + bits[AMDGPU_MMHUB] = MMVM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + MMVM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + MMVM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + MMVM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + MMVM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + MMVM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + MMVM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK; + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + /* MM HUB */ + hub = &adev->vmhub[AMDGPU_MMHUB]; + for (i = 0; i < 16; i++) { + reg = hub->vm_context0_cntl + i; + tmp = RREG32(reg); + tmp &= ~bits[AMDGPU_MMHUB]; + WREG32(reg, tmp); + } + + /* GFX HUB */ + hub = &adev->vmhub[AMDGPU_GFXHUB]; + for (i = 0; i < 16; i++) { + reg = hub->vm_context0_cntl + i; + tmp = RREG32(reg); + tmp &= ~bits[AMDGPU_GFXHUB]; + WREG32(reg, tmp); + } + break; + case AMDGPU_IRQ_STATE_ENABLE: + /* MM HUB */ + hub = &adev->vmhub[AMDGPU_MMHUB]; + for (i = 0; i < 16; i++) { + reg = hub->vm_context0_cntl + i; + tmp = RREG32(reg); + tmp |= bits[AMDGPU_MMHUB]; + WREG32(reg, tmp); + } + + /* GFX HUB */ + hub = &adev->vmhub[AMDGPU_GFXHUB]; + for (i = 0; i < 16; i++) { + reg = hub->vm_context0_cntl + i; + tmp = RREG32(reg); + tmp |= bits[AMDGPU_GFXHUB]; + WREG32(reg, tmp); + } + break; + default: + break; + } + + return 0; +} + +static int gmc_v10_0_process_interrupt(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + struct amdgpu_vmhub *hub = &adev->vmhub[entry->vmid_src]; + uint32_t status = 0; + u64 addr; + + addr = (u64)entry->src_data[0] << 12; + addr |= ((u64)entry->src_data[1] & 0xf) << 44; + + if (!amdgpu_sriov_vf(adev)) { + status = RREG32(hub->vm_l2_pro_fault_status); + WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1); + } + + if (printk_ratelimit()) { + dev_err(adev->dev, + "[%s] VMC page fault (src_id:%u ring:%u vmid:%u pasid:%u)\n", + entry->vmid_src ? "mmhub" : "gfxhub", + entry->src_id, entry->ring_id, entry->vmid, + entry->pasid); + dev_err(adev->dev, " at page 0x%016llx from %d\n", + addr, entry->client_id); + if (!amdgpu_sriov_vf(adev)) + dev_err(adev->dev, + "VM_L2_PROTECTION_FAULT_STATUS:0x%08X\n", + status); + } + + return 0; +} + +static const struct amdgpu_irq_src_funcs gmc_v10_0_irq_funcs = { + .set = gmc_v10_0_vm_fault_interrupt_state, + .process = gmc_v10_0_process_interrupt, +}; + +static void gmc_v10_0_set_irq_funcs(struct amdgpu_device *adev) +{ + adev->gmc.vm_fault.num_types = 1; + adev->gmc.vm_fault.funcs = &gmc_v10_0_irq_funcs; +} + +static uint32_t gmc_v10_0_get_invalidate_req(unsigned int vmid, + uint32_t flush_type) +{ + u32 req = 0; + + /* invalidate using legacy mode on vmid*/ + req = REG_SET_FIELD(req, GCVM_INVALIDATE_ENG0_REQ, + PER_VMID_INVALIDATE_REQ, 1 << vmid); + req = REG_SET_FIELD(req, GCVM_INVALIDATE_ENG0_REQ, FLUSH_TYPE, flush_type); + req = REG_SET_FIELD(req, GCVM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PTES, 1); + req = REG_SET_FIELD(req, GCVM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE0, 1); + req = REG_SET_FIELD(req, GCVM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE1, 1); + req = REG_SET_FIELD(req, GCVM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE2, 1); + req = REG_SET_FIELD(req, GCVM_INVALIDATE_ENG0_REQ, INVALIDATE_L1_PTES, 1); + req = REG_SET_FIELD(req, GCVM_INVALIDATE_ENG0_REQ, + CLEAR_PROTECTION_FAULT_STATUS_ADDR, 0); + + return req; +} + +/* + * GART + * VMID 0 is the physical GPU addresses as used by the kernel. + * VMIDs 1-15 are used for userspace clients and are handled + * by the amdgpu vm/hsa code. + */ + +static void gmc_v10_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid, + unsigned int vmhub, uint32_t flush_type) +{ + struct amdgpu_vmhub *hub = &adev->vmhub[vmhub]; + u32 tmp = gmc_v10_0_get_invalidate_req(vmid, flush_type); + /* Use register 17 for GART */ + const unsigned eng = 17; + unsigned int i; + + WREG32_NO_KIQ(hub->vm_inv_eng0_req + eng, tmp); + + /* Wait for ACK with a delay.*/ + for (i = 0; i < adev->usec_timeout; i++) { + tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_ack + eng); + tmp &= 1 << vmid; + if (tmp) + break; + + udelay(1); + } + + if (i < adev->usec_timeout) + return; + + DRM_ERROR("Timeout waiting for VM flush ACK!\n"); +} + +/** + * gmc_v10_0_flush_gpu_tlb - gart tlb flush callback + * + * @adev: amdgpu_device pointer + * @vmid: vm instance to flush + * + * Flush the TLB for the requested page table. + */ +static void gmc_v10_0_flush_gpu_tlb(struct amdgpu_device *adev, + uint32_t vmid, uint32_t flush_type) +{ + struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring; + struct dma_fence *fence; + struct amdgpu_job *job; + + int r; + + /* flush hdp cache */ + adev->nbio_funcs->hdp_flush(adev, NULL); + + mutex_lock(&adev->mman.gtt_window_lock); + + gmc_v10_0_flush_vm_hub(adev, vmid, AMDGPU_MMHUB, 0); + if (!adev->mman.buffer_funcs_enabled || !adev->ib_pool_ready || + adev->asic_type != CHIP_NAVI10) { + gmc_v10_0_flush_vm_hub(adev, vmid, AMDGPU_GFXHUB, 0); + mutex_unlock(&adev->mman.gtt_window_lock); + return; + } + + /* The SDMA on Navi has a bug which can theoretically result in memory + * corruption if an invalidation happens at the same time as an VA + * translation. Avoid this by doing the invalidation from the SDMA + * itself. + */ + r = amdgpu_job_alloc_with_ib(adev, 16 * 4, &job); + if (r) + goto error_alloc; + + job->vm_pd_addr = amdgpu_gmc_pd_addr(adev->gart.bo); + job->vm_needs_flush = true; + amdgpu_ring_pad_ib(ring, &job->ibs[0]); + r = amdgpu_job_submit(job, &adev->mman.entity, + AMDGPU_FENCE_OWNER_UNDEFINED, &fence); + if (r) + goto error_submit; + + mutex_unlock(&adev->mman.gtt_window_lock); + + dma_fence_wait(fence, false); + dma_fence_put(fence); + + return; + +error_submit: + amdgpu_job_free(job); + +error_alloc: + mutex_unlock(&adev->mman.gtt_window_lock); + DRM_ERROR("Error flushing GPU TLB using the SDMA (%d)!\n", r); +} + +static uint64_t gmc_v10_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, + unsigned vmid, uint64_t pd_addr) +{ + 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; + + amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_lo32 + (2 * vmid), + lower_32_bits(pd_addr)); + + amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_hi32 + (2 * vmid), + upper_32_bits(pd_addr)); + + amdgpu_ring_emit_wreg(ring, hub->vm_inv_eng0_req + eng, req); + + /* wait for the invalidate to complete */ + amdgpu_ring_emit_reg_wait(ring, hub->vm_inv_eng0_ack + eng, + 1 << vmid, 1 << vmid); + + return pd_addr; +} + +static void gmc_v10_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid, + unsigned pasid) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t reg; + + if (ring->funcs->vmhub == AMDGPU_GFXHUB) + reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_VMID_0_LUT) + vmid; + else + reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_VMID_0_LUT_MM) + vmid; + + amdgpu_ring_emit_wreg(ring, reg, pasid); +} + +/* + * PTE format on NAVI 10: + * 63:59 reserved + * 58:57 reserved + * 56 F + * 55 L + * 54 reserved + * 53:52 SW + * 51 T + * 50:48 mtype + * 47:12 4k physical page base address + * 11:7 fragment + * 6 write + * 5 read + * 4 exe + * 3 Z + * 2 snooped + * 1 system + * 0 valid + * + * PDE format on NAVI 10: + * 63:59 block fragment size + * 58:55 reserved + * 54 P + * 53:48 reserved + * 47:6 physical base address of PD or PTE + * 5:3 reserved + * 2 C + * 1 system + * 0 valid + */ +static uint64_t gmc_v10_0_get_vm_pte_flags(struct amdgpu_device *adev, + uint32_t flags) +{ + uint64_t pte_flag = 0; + + if (flags & AMDGPU_VM_PAGE_EXECUTABLE) + pte_flag |= AMDGPU_PTE_EXECUTABLE; + if (flags & AMDGPU_VM_PAGE_READABLE) + pte_flag |= AMDGPU_PTE_READABLE; + if (flags & AMDGPU_VM_PAGE_WRITEABLE) + pte_flag |= AMDGPU_PTE_WRITEABLE; + + switch (flags & AMDGPU_VM_MTYPE_MASK) { + case AMDGPU_VM_MTYPE_DEFAULT: + pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_NC); + break; + case AMDGPU_VM_MTYPE_NC: + pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_NC); + break; + case AMDGPU_VM_MTYPE_WC: + pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_WC); + break; + case AMDGPU_VM_MTYPE_CC: + pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_CC); + break; + case AMDGPU_VM_MTYPE_UC: + pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_UC); + break; + default: + pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_NC); + break; + } + + if (flags & AMDGPU_VM_PAGE_PRT) + pte_flag |= AMDGPU_PTE_PRT; + + return pte_flag; +} + +static void gmc_v10_0_get_vm_pde(struct amdgpu_device *adev, int level, + uint64_t *addr, uint64_t *flags) +{ + if (!(*flags & AMDGPU_PDE_PTE) && !(*flags & AMDGPU_PTE_SYSTEM)) + *addr = adev->vm_manager.vram_base_offset + *addr - + adev->gmc.vram_start; + BUG_ON(*addr & 0xFFFF00000000003FULL); + + if (!adev->gmc.translate_further) + return; + + if (level == AMDGPU_VM_PDB1) { + /* Set the block fragment size */ + if (!(*flags & AMDGPU_PDE_PTE)) + *flags |= AMDGPU_PDE_BFS(0x9); + + } else if (level == AMDGPU_VM_PDB0) { + if (*flags & AMDGPU_PDE_PTE) + *flags &= ~AMDGPU_PDE_PTE; + else + *flags |= AMDGPU_PTE_TF; + } +} + +static const struct amdgpu_gmc_funcs gmc_v10_0_gmc_funcs = { + .flush_gpu_tlb = gmc_v10_0_flush_gpu_tlb, + .emit_flush_gpu_tlb = gmc_v10_0_emit_flush_gpu_tlb, + .emit_pasid_mapping = gmc_v10_0_emit_pasid_mapping, + .get_vm_pte_flags = gmc_v10_0_get_vm_pte_flags, + .get_vm_pde = gmc_v10_0_get_vm_pde +}; + +static void gmc_v10_0_set_gmc_funcs(struct amdgpu_device *adev) +{ + if (adev->gmc.gmc_funcs == NULL) + adev->gmc.gmc_funcs = &gmc_v10_0_gmc_funcs; +} + +static int gmc_v10_0_early_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + gmc_v10_0_set_gmc_funcs(adev); + gmc_v10_0_set_irq_funcs(adev); + + adev->gmc.shared_aperture_start = 0x2000000000000000ULL; + adev->gmc.shared_aperture_end = + adev->gmc.shared_aperture_start + (4ULL << 30) - 1; + adev->gmc.private_aperture_start = 0x1000000000000000ULL; + adev->gmc.private_aperture_end = + adev->gmc.private_aperture_start + (4ULL << 30) - 1; + + return 0; +} + +static int gmc_v10_0_late_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + unsigned vm_inv_eng[AMDGPU_MAX_VMHUBS] = { 4, 4 }; + unsigned i; + + for(i = 0; i < adev->num_rings; ++i) { + struct amdgpu_ring *ring = adev->rings[i]; + unsigned vmhub = ring->funcs->vmhub; + + ring->vm_inv_eng = vm_inv_eng[vmhub]++; + dev_info(adev->dev, "ring %u(%s) uses VM inv eng %u on hub %u\n", + ring->idx, ring->name, ring->vm_inv_eng, + ring->funcs->vmhub); + } + + /* Engine 17 is used for GART flushes */ + for(i = 0; i < AMDGPU_MAX_VMHUBS; ++i) + BUG_ON(vm_inv_eng[i] > 17); + + return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); +} + +static void gmc_v10_0_vram_gtt_location(struct amdgpu_device *adev, + struct amdgpu_gmc *mc) +{ + u64 base = 0; + + if (!amdgpu_sriov_vf(adev)) + base = gfxhub_v2_0_get_fb_location(adev); + + amdgpu_gmc_vram_location(adev, &adev->gmc, base); + amdgpu_gmc_gart_location(adev, mc); + + /* base offset of vram pages */ + adev->vm_manager.vram_base_offset = gfxhub_v2_0_get_mc_fb_offset(adev); +} + +/** + * gmc_v10_0_mc_init - initialize the memory controller driver params + * + * @adev: amdgpu_device pointer + * + * Look up the amount of vram, vram width, and decide how to place + * vram and gart within the GPU's physical address space. + * Returns 0 for success. + */ +static int gmc_v10_0_mc_init(struct amdgpu_device *adev) +{ + int chansize, numchan; + + if (!amdgpu_emu_mode) + adev->gmc.vram_width = amdgpu_atomfirmware_get_vram_width(adev); + else { + /* hard code vram_width for emulation */ + chansize = 128; + numchan = 1; + adev->gmc.vram_width = numchan * chansize; + } + + /* Could aper size report 0 ? */ + adev->gmc.aper_base = pci_resource_start(adev->pdev, 0); + adev->gmc.aper_size = pci_resource_len(adev->pdev, 0); + + /* size in MB on si */ + adev->gmc.mc_vram_size = + adev->nbio_funcs->get_memsize(adev) * 1024ULL * 1024ULL; + adev->gmc.real_vram_size = adev->gmc.mc_vram_size; + adev->gmc.visible_vram_size = adev->gmc.aper_size; + + /* In case the PCI BAR is larger than the actual amount of vram */ + if (adev->gmc.visible_vram_size > adev->gmc.real_vram_size) + adev->gmc.visible_vram_size = adev->gmc.real_vram_size; + + /* set the gart size */ + if (amdgpu_gart_size == -1) { + switch (adev->asic_type) { + case CHIP_NAVI10: + default: + adev->gmc.gart_size = 512ULL << 20; + break; + } + } else + adev->gmc.gart_size = (u64)amdgpu_gart_size << 20; + + gmc_v10_0_vram_gtt_location(adev, &adev->gmc); + + return 0; +} + +static int gmc_v10_0_gart_init(struct amdgpu_device *adev) +{ + int r; + + if (adev->gart.bo) { + WARN(1, "NAVI10 PCIE GART already initialized\n"); + return 0; + } + + /* Initialize common gart structure */ + r = amdgpu_gart_init(adev); + if (r) + return r; + + adev->gart.table_size = adev->gart.num_gpu_pages * 8; + adev->gart.gart_pte_flags = AMDGPU_PTE_MTYPE_NV10(MTYPE_UC) | + AMDGPU_PTE_EXECUTABLE; + + return amdgpu_gart_table_vram_alloc(adev); +} + +static unsigned gmc_v10_0_get_vbios_fb_size(struct amdgpu_device *adev) +{ + u32 d1vga_control = RREG32_SOC15(DCE, 0, mmD1VGA_CONTROL); + unsigned size; + + if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) { + size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */ + } else { + u32 viewport; + u32 pitch; + + viewport = RREG32_SOC15(DCE, 0, mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION); + pitch = RREG32_SOC15(DCE, 0, mmHUBPREQ0_DCSURF_SURFACE_PITCH); + size = (REG_GET_FIELD(viewport, + HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION, PRI_VIEWPORT_HEIGHT) * + REG_GET_FIELD(pitch, HUBPREQ0_DCSURF_SURFACE_PITCH, PITCH) * + 4); + } + /* return 0 if the pre-OS buffer uses up most of vram */ + if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024)) { + DRM_ERROR("Warning: pre-OS buffer uses most of vram, \ + be aware of gart table overwrite\n"); + return 0; + } + + return size; +} + + + +static int gmc_v10_0_sw_init(void *handle) +{ + int r; + int dma_bits; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + gfxhub_v2_0_init(adev); + mmhub_v2_0_init(adev); + + spin_lock_init(&adev->gmc.invalidate_lock); + + adev->gmc.vram_type = amdgpu_atomfirmware_get_vram_type(adev); + switch (adev->asic_type) { + case CHIP_NAVI10: + /* + * To fulfill 4-level page support, + * vm size is 256TB (48bit), maximum size of Navi10, + * block size 512 (9bit) + */ + amdgpu_vm_adjust_size(adev, 256 * 1024, 9, 3, 48); + break; + default: + break; + } + + /* This interrupt is VMC page fault.*/ + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VMC, + VMC_1_0__SRCID__VM_FAULT, + &adev->gmc.vm_fault); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UTCL2, + UTCL2_1_0__SRCID__FAULT, + &adev->gmc.vm_fault); + if (r) + return r; + + /* + * Set the internal MC address mask This is the max address of the GPU's + * internal address space. + */ + adev->gmc.mc_mask = 0xffffffffffffULL; /* 48 bit MC */ + + /* + * Reserve 8M stolen memory for navi10 like vega10 + * TODO: will check if it's really needed on asic. + */ + if (amdgpu_emu_mode == 1) + adev->gmc.stolen_size = 0; + else + adev->gmc.stolen_size = 9 * 1024 *1024; + + /* + * Set DMA mask + need_dma32 flags. + * PCIE - can handle 44-bits. + * IGP - can handle 44-bits + * PCI - dma32 for legacy pci gart, 44 bits on navi10 + */ + adev->need_dma32 = false; + dma_bits = adev->need_dma32 ? 32 : 44; + + r = pci_set_dma_mask(adev->pdev, DMA_BIT_MASK(dma_bits)); + if (r) { + adev->need_dma32 = true; + dma_bits = 32; + printk(KERN_WARNING "amdgpu: No suitable DMA available.\n"); + } + + r = pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(dma_bits)); + if (r) { + pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32)); + printk(KERN_WARNING "amdgpu: No coherent DMA available.\n"); + } + + r = gmc_v10_0_mc_init(adev); + if (r) + return r; + + adev->gmc.stolen_size = gmc_v10_0_get_vbios_fb_size(adev); + + /* Memory manager */ + r = amdgpu_bo_init(adev); + if (r) + return r; + + r = gmc_v10_0_gart_init(adev); + if (r) + return r; + + /* + * number of VMs + * VMID 0 is reserved for System + * amdgpu graphics/compute will use VMIDs 1-7 + * amdkfd will use VMIDs 8-15 + */ + adev->vm_manager.id_mgr[AMDGPU_GFXHUB].num_ids = AMDGPU_NUM_OF_VMIDS; + adev->vm_manager.id_mgr[AMDGPU_MMHUB].num_ids = AMDGPU_NUM_OF_VMIDS; + + amdgpu_vm_manager_init(adev); + + return 0; +} + +/** + * gmc_v8_0_gart_fini - vm fini callback + * + * @adev: amdgpu_device pointer + * + * Tears down the driver GART/VM setup (CIK). + */ +static void gmc_v10_0_gart_fini(struct amdgpu_device *adev) +{ + amdgpu_gart_table_vram_free(adev); + amdgpu_gart_fini(adev); +} + +static int gmc_v10_0_sw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + amdgpu_vm_manager_fini(adev); + gmc_v10_0_gart_fini(adev); + amdgpu_gem_force_release(adev); + amdgpu_bo_fini(adev); + + return 0; +} + +static void gmc_v10_0_init_golden_registers(struct amdgpu_device *adev) +{ + switch (adev->asic_type) { + case CHIP_NAVI10: + break; + default: + break; + } +} + +/** + * gmc_v10_0_gart_enable - gart enable + * + * @adev: amdgpu_device pointer + */ +static int gmc_v10_0_gart_enable(struct amdgpu_device *adev) +{ + int r; + bool value; + u32 tmp; + + if (adev->gart.bo == NULL) { + dev_err(adev->dev, "No VRAM object for PCIE GART.\n"); + return -EINVAL; + } + + r = amdgpu_gart_table_vram_pin(adev); + if (r) + return r; + + r = gfxhub_v2_0_gart_enable(adev); + if (r) + return r; + + r = mmhub_v2_0_gart_enable(adev); + if (r) + return r; + + tmp = RREG32_SOC15(HDP, 0, mmHDP_MISC_CNTL); + tmp |= HDP_MISC_CNTL__FLUSH_INVALIDATE_CACHE_MASK; + WREG32_SOC15(HDP, 0, mmHDP_MISC_CNTL, tmp); + + tmp = RREG32_SOC15(HDP, 0, mmHDP_HOST_PATH_CNTL); + WREG32_SOC15(HDP, 0, mmHDP_HOST_PATH_CNTL, tmp); + + /* Flush HDP after it is initialized */ + adev->nbio_funcs->hdp_flush(adev, NULL); + + value = (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS) ? + false : true; + + gfxhub_v2_0_set_fault_enable_default(adev, value); + mmhub_v2_0_set_fault_enable_default(adev, value); + gmc_v10_0_flush_gpu_tlb(adev, 0, 0); + + DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n", + (unsigned)(adev->gmc.gart_size >> 20), + (unsigned long long)amdgpu_bo_gpu_offset(adev->gart.bo)); + + adev->gart.ready = true; + + return 0; +} + +static int gmc_v10_0_hw_init(void *handle) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + /* The sequence of these two function calls matters.*/ + gmc_v10_0_init_golden_registers(adev); + + r = gmc_v10_0_gart_enable(adev); + if (r) + return r; + + return 0; +} + +/** + * gmc_v10_0_gart_disable - gart disable + * + * @adev: amdgpu_device pointer + * + * This disables all VM page table. + */ +static void gmc_v10_0_gart_disable(struct amdgpu_device *adev) +{ + gfxhub_v2_0_gart_disable(adev); + mmhub_v2_0_gart_disable(adev); + amdgpu_gart_table_vram_unpin(adev); +} + +static int gmc_v10_0_hw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (amdgpu_sriov_vf(adev)) { + /* full access mode, so don't touch any GMC register */ + DRM_DEBUG("For SRIOV client, shouldn't do anything.\n"); + return 0; + } + + amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); + gmc_v10_0_gart_disable(adev); + + return 0; +} + +static int gmc_v10_0_suspend(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + gmc_v10_0_hw_fini(adev); + + return 0; +} + +static int gmc_v10_0_resume(void *handle) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + r = gmc_v10_0_hw_init(adev); + if (r) + return r; + + amdgpu_vmid_reset_all(adev); + + return 0; +} + +static bool gmc_v10_0_is_idle(void *handle) +{ + /* MC is always ready in GMC v10.*/ + return true; +} + +static int gmc_v10_0_wait_for_idle(void *handle) +{ + /* There is no need to wait for MC idle in GMC v10.*/ + return 0; +} + +static int gmc_v10_0_soft_reset(void *handle) +{ + return 0; +} + +static int gmc_v10_0_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + r = mmhub_v2_0_set_clockgating(adev, state); + if (r) + return r; + + return athub_v2_0_set_clockgating(adev, state); +} + +static void gmc_v10_0_get_clockgating_state(void *handle, u32 *flags) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + mmhub_v2_0_get_clockgating(adev, flags); + + athub_v2_0_get_clockgating(adev, flags); +} + +static int gmc_v10_0_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + return 0; +} + +const struct amd_ip_funcs gmc_v10_0_ip_funcs = { + .name = "gmc_v10_0", + .early_init = gmc_v10_0_early_init, + .late_init = gmc_v10_0_late_init, + .sw_init = gmc_v10_0_sw_init, + .sw_fini = gmc_v10_0_sw_fini, + .hw_init = gmc_v10_0_hw_init, + .hw_fini = gmc_v10_0_hw_fini, + .suspend = gmc_v10_0_suspend, + .resume = gmc_v10_0_resume, + .is_idle = gmc_v10_0_is_idle, + .wait_for_idle = gmc_v10_0_wait_for_idle, + .soft_reset = gmc_v10_0_soft_reset, + .set_clockgating_state = gmc_v10_0_set_clockgating_state, + .set_powergating_state = gmc_v10_0_set_powergating_state, + .get_clockgating_state = gmc_v10_0_get_clockgating_state, +}; + +const struct amdgpu_ip_block_version gmc_v10_0_ip_block = +{ + .type = AMD_IP_BLOCK_TYPE_GMC, + .major = 10, + .minor = 0, + .rev = 0, + .funcs = &gmc_v10_0_ip_funcs, +}; diff --cc drivers/gpu/drm/amd/amdgpu/mes_v10_1.c index 0d92b88a85b84,0000000000000..29fab7984855a mode 100644,000000..100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v10_1.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_v10_1.c @@@ -1,365 -1,0 +1,366 @@@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include ++#include +#include "amdgpu.h" +#include "soc15_common.h" +#include "nv.h" +#include "gc/gc_10_1_0_offset.h" +#include "gc/gc_10_1_0_sh_mask.h" + +MODULE_FIRMWARE("amdgpu/navi10_mes.bin"); + +static int mes_v10_1_add_hw_queue(struct amdgpu_mes *mes, + struct mes_add_queue_input *input) +{ + return 0; +} + +static int mes_v10_1_remove_hw_queue(struct amdgpu_mes *mes, + struct mes_remove_queue_input *input) +{ + return 0; +} + +static int mes_v10_1_suspend_gang(struct amdgpu_mes *mes, + struct mes_suspend_gang_input *input) +{ + return 0; +} + +static int mes_v10_1_resume_gang(struct amdgpu_mes *mes, + struct mes_resume_gang_input *input) +{ + return 0; +} + +static const struct amdgpu_mes_funcs mes_v10_1_funcs = { + .add_hw_queue = mes_v10_1_add_hw_queue, + .remove_hw_queue = mes_v10_1_remove_hw_queue, + .suspend_gang = mes_v10_1_suspend_gang, + .resume_gang = mes_v10_1_resume_gang, +}; + +static int mes_v10_1_init_microcode(struct amdgpu_device *adev) +{ + const char *chip_name; + char fw_name[30]; + int err; + const struct mes_firmware_header_v1_0 *mes_hdr; + + switch (adev->asic_type) { + case CHIP_NAVI10: + chip_name = "navi10"; + break; + default: + BUG(); + } + + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mes.bin", chip_name); + err = request_firmware(&adev->mes.fw, fw_name, adev->dev); + if (err) + return err; + + err = amdgpu_ucode_validate(adev->mes.fw); + if (err) { + release_firmware(adev->mes.fw); + adev->mes.fw = NULL; + return err; + } + + mes_hdr = (const struct mes_firmware_header_v1_0 *)adev->mes.fw->data; + adev->mes.ucode_fw_version = le32_to_cpu(mes_hdr->mes_ucode_version); + adev->mes.ucode_fw_version = + le32_to_cpu(mes_hdr->mes_ucode_data_version); + adev->mes.uc_start_addr = + le32_to_cpu(mes_hdr->mes_uc_start_addr_lo) | + ((uint64_t)(le32_to_cpu(mes_hdr->mes_uc_start_addr_hi)) << 32); + adev->mes.data_start_addr = + le32_to_cpu(mes_hdr->mes_data_start_addr_lo) | + ((uint64_t)(le32_to_cpu(mes_hdr->mes_data_start_addr_hi)) << 32); + + return 0; +} + +static void mes_v10_1_free_microcode(struct amdgpu_device *adev) +{ + release_firmware(adev->mes.fw); + adev->mes.fw = NULL; +} + +static int mes_v10_1_allocate_ucode_buffer(struct amdgpu_device *adev) +{ + int r; + const struct mes_firmware_header_v1_0 *mes_hdr; + const __le32 *fw_data; + unsigned fw_size; + + mes_hdr = (const struct mes_firmware_header_v1_0 *) + adev->mes.fw->data; + + fw_data = (const __le32 *)(adev->mes.fw->data + + le32_to_cpu(mes_hdr->mes_ucode_offset_bytes)); + fw_size = le32_to_cpu(mes_hdr->mes_ucode_size_bytes); + + r = amdgpu_bo_create_reserved(adev, fw_size, + PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT, + &adev->mes.ucode_fw_obj, + &adev->mes.ucode_fw_gpu_addr, + (void **)&adev->mes.ucode_fw_ptr); + if (r) { + dev_err(adev->dev, "(%d) failed to create mes fw bo\n", r); + return r; + } + + memcpy(adev->mes.ucode_fw_ptr, fw_data, fw_size); + + amdgpu_bo_kunmap(adev->mes.ucode_fw_obj); + amdgpu_bo_unreserve(adev->mes.ucode_fw_obj); + + return 0; +} + +static int mes_v10_1_allocate_ucode_data_buffer(struct amdgpu_device *adev) +{ + int r; + const struct mes_firmware_header_v1_0 *mes_hdr; + const __le32 *fw_data; + unsigned fw_size; + + mes_hdr = (const struct mes_firmware_header_v1_0 *) + adev->mes.fw->data; + + fw_data = (const __le32 *)(adev->mes.fw->data + + le32_to_cpu(mes_hdr->mes_ucode_data_offset_bytes)); + fw_size = le32_to_cpu(mes_hdr->mes_ucode_data_size_bytes); + + r = amdgpu_bo_create_reserved(adev, fw_size, + 64 * 1024, AMDGPU_GEM_DOMAIN_GTT, + &adev->mes.data_fw_obj, + &adev->mes.data_fw_gpu_addr, + (void **)&adev->mes.data_fw_ptr); + if (r) { + dev_err(adev->dev, "(%d) failed to create mes data fw bo\n", r); + return r; + } + + memcpy(adev->mes.data_fw_ptr, fw_data, fw_size); + + amdgpu_bo_kunmap(adev->mes.data_fw_obj); + amdgpu_bo_unreserve(adev->mes.data_fw_obj); + + return 0; +} + +static void mes_v10_1_free_ucode_buffers(struct amdgpu_device *adev) +{ + amdgpu_bo_free_kernel(&adev->mes.data_fw_obj, + &adev->mes.data_fw_gpu_addr, + (void **)&adev->mes.data_fw_ptr); + + amdgpu_bo_free_kernel(&adev->mes.ucode_fw_obj, + &adev->mes.ucode_fw_gpu_addr, + (void **)&adev->mes.ucode_fw_ptr); +} + +static void mes_v10_1_enable(struct amdgpu_device *adev, bool enable) +{ + uint32_t data = 0; + + if (enable) { + data = RREG32_SOC15(GC, 0, mmCP_MES_CNTL); + data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE0_RESET, 1); + WREG32_SOC15(GC, 0, mmCP_MES_CNTL, data); + + /* set ucode start address */ + WREG32_SOC15(GC, 0, mmCP_MES_PRGRM_CNTR_START, + (uint32_t)(adev->mes.uc_start_addr) >> 2); + + /* clear BYPASS_UNCACHED to avoid hangs after interrupt. */ + data = RREG32_SOC15(GC, 0, mmCP_MES_DC_OP_CNTL); + data = REG_SET_FIELD(data, CP_MES_DC_OP_CNTL, + BYPASS_UNCACHED, 0); + WREG32_SOC15(GC, 0, mmCP_MES_DC_OP_CNTL, data); + + /* unhalt MES and activate pipe0 */ + data = REG_SET_FIELD(0, CP_MES_CNTL, MES_PIPE0_ACTIVE, 1); + WREG32_SOC15(GC, 0, mmCP_MES_CNTL, data); + } else { + data = RREG32_SOC15(GC, 0, mmCP_MES_CNTL); + data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE0_ACTIVE, 0); + data = REG_SET_FIELD(data, CP_MES_CNTL, + MES_INVALIDATE_ICACHE, 1); + data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE0_RESET, 1); + data = REG_SET_FIELD(data, CP_MES_CNTL, MES_HALT, 1); + WREG32_SOC15(GC, 0, mmCP_MES_CNTL, data); + } +} + +/* This function is for backdoor MES firmware */ +static int mes_v10_1_load_microcode(struct amdgpu_device *adev) +{ + int r; + uint32_t data; + + if (!adev->mes.fw) + return -EINVAL; + + r = mes_v10_1_allocate_ucode_buffer(adev); + if (r) + return r; + + r = mes_v10_1_allocate_ucode_data_buffer(adev); + if (r) { + mes_v10_1_free_ucode_buffers(adev); + return r; + } + + mes_v10_1_enable(adev, false); + + WREG32_SOC15(GC, 0, mmCP_MES_IC_BASE_CNTL, 0); + + mutex_lock(&adev->srbm_mutex); + /* me=3, pipe=0, queue=0 */ + nv_grbm_select(adev, 3, 0, 0, 0); + + /* set ucode start address */ + WREG32_SOC15(GC, 0, mmCP_MES_PRGRM_CNTR_START, + (uint32_t)(adev->mes.uc_start_addr) >> 2); + + /* set ucode fimrware address */ + WREG32_SOC15(GC, 0, mmCP_MES_IC_BASE_LO, + lower_32_bits(adev->mes.ucode_fw_gpu_addr)); + WREG32_SOC15(GC, 0, mmCP_MES_IC_BASE_HI, + upper_32_bits(adev->mes.ucode_fw_gpu_addr)); + + /* set ucode instruction cache boundary to 2M-1 */ + WREG32_SOC15(GC, 0, mmCP_MES_MIBOUND_LO, 0x1FFFFF); + + /* set ucode data firmware address */ + WREG32_SOC15(GC, 0, mmCP_MES_MDBASE_LO, + lower_32_bits(adev->mes.data_fw_gpu_addr)); + WREG32_SOC15(GC, 0, mmCP_MES_MDBASE_HI, + upper_32_bits(adev->mes.data_fw_gpu_addr)); + + /* Set 0x3FFFF (256K-1) to CP_MES_MDBOUND_LO */ + WREG32_SOC15(GC, 0, mmCP_MES_MDBOUND_LO, 0x3FFFF); + + /* invalidate ICACHE */ + data = RREG32_SOC15(GC, 0, mmCP_MES_IC_OP_CNTL); + data = REG_SET_FIELD(data, CP_MES_IC_OP_CNTL, PRIME_ICACHE, 0); + data = REG_SET_FIELD(data, CP_MES_IC_OP_CNTL, INVALIDATE_CACHE, 1); + WREG32_SOC15(GC, 0, mmCP_MES_IC_OP_CNTL, data); + + /* prime the ICACHE. */ + data = RREG32_SOC15(GC, 0, mmCP_MES_IC_OP_CNTL); + data = REG_SET_FIELD(data, CP_MES_IC_OP_CNTL, PRIME_ICACHE, 1); + WREG32_SOC15(GC, 0, mmCP_MES_IC_OP_CNTL, data); + + nv_grbm_select(adev, 0, 0, 0, 0); + mutex_unlock(&adev->srbm_mutex); + + return 0; +} + +static int mes_v10_1_sw_init(void *handle) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + r = mes_v10_1_init_microcode(adev); + if (r) + return r; + + return 0; +} + +static int mes_v10_1_sw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + mes_v10_1_free_microcode(adev); + + return 0; +} + +static int mes_v10_1_hw_init(void *handle) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) { + r = mes_v10_1_load_microcode(adev); + if (r) { + DRM_ERROR("failed to MES fw, r=%d\n", r); + return r; + } + } else { + DRM_ERROR("only support direct fw loading on MES\n"); + return -EINVAL; + } + + mes_v10_1_enable(adev, true); + + return 0; +} + +static int mes_v10_1_hw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + mes_v10_1_enable(adev, false); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) + mes_v10_1_free_ucode_buffers(adev); + + return 0; +} + +static int mes_v10_1_suspend(void *handle) +{ + return 0; +} + +static int mes_v10_1_resume(void *handle) +{ + return 0; +} + +static const struct amd_ip_funcs mes_v10_1_ip_funcs = { + .name = "mes_v10_1", + .sw_init = mes_v10_1_sw_init, + .sw_fini = mes_v10_1_sw_fini, + .hw_init = mes_v10_1_hw_init, + .hw_fini = mes_v10_1_hw_fini, + .suspend = mes_v10_1_suspend, + .resume = mes_v10_1_resume, +}; + +const struct amdgpu_ip_block_version mes_v10_1_ip_block = { + .type = AMD_IP_BLOCK_TYPE_MES, + .major = 10, + .minor = 1, + .rev = 0, + .funcs = &mes_v10_1_ip_funcs, +}; diff --cc drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 58d7bbc5ada74,b16c658074d2a..4c70a0803b851 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@@ -62,10 -62,12 +62,12 @@@ #include #include #include + #include #include + #include #if defined(CONFIG_DRM_AMD_DC_DCN1_0) -#include "ivsrcid/irqsrcs_dcn_1_0.h" +#include "ivsrcid/dcn/irqsrcs_dcn_1_0.h" #include "dcn/dcn_1_0_offset.h" #include "dcn/dcn_1_0_sh_mask.h" diff --cc drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.c index 23362dd4b6d39,0000000000000..51a3dfe97f0ea mode 100644,000000..100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.c @@@ -1,157 -1,0 +1,159 @@@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + ++#include ++ +#include "reg_helper.h" +#include "core_types.h" +#include "dcn20_dccg.h" + +#define TO_DCN_DCCG(dccg)\ + container_of(dccg, struct dcn_dccg, base) + +#define REG(reg) \ + (dccg_dcn->regs->reg) + +#undef FN +#define FN(reg_name, field_name) \ + dccg_dcn->dccg_shift->field_name, dccg_dcn->dccg_mask->field_name + +#define CTX \ + dccg_dcn->base.ctx +#define DC_LOGGER \ + dccg->ctx->logger + +void dccg2_update_dpp_dto(struct dccg *dccg, int dpp_inst, int req_dppclk) +{ + struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + + if (dccg->ref_dppclk && req_dppclk) { + int ref_dppclk = dccg->ref_dppclk; + + ASSERT(req_dppclk <= ref_dppclk); + /* need to clamp to 8 bits */ + if (ref_dppclk > 0xff) { + int divider = (ref_dppclk + 0xfe) / 0xff; + + ref_dppclk /= divider; + req_dppclk = (req_dppclk + divider - 1) / divider; + if (req_dppclk > ref_dppclk) + req_dppclk = ref_dppclk; + } + REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0, + DPPCLK0_DTO_PHASE, req_dppclk, + DPPCLK0_DTO_MODULO, ref_dppclk); + REG_UPDATE(DPPCLK_DTO_CTRL, + DPPCLK_DTO_ENABLE[dpp_inst], 1); + } else { + REG_UPDATE(DPPCLK_DTO_CTRL, + DPPCLK_DTO_ENABLE[dpp_inst], 0); + } +} + +void dccg2_get_dccg_ref_freq(struct dccg *dccg, + unsigned int xtalin_freq_inKhz, + unsigned int *dccg_ref_freq_inKhz) +{ + struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + uint32_t clk_en = 0; + uint32_t clk_sel = 0; + + REG_GET_2(REFCLK_CNTL, REFCLK_CLOCK_EN, &clk_en, REFCLK_SRC_SEL, &clk_sel); + + if (clk_en != 0) { + // DCN20 has never been validated for non-xtalin as reference + // frequency. There's actually no way for DC to determine what + // frequency a non-xtalin source is. + ASSERT_CRITICAL(false); + } + + *dccg_ref_freq_inKhz = xtalin_freq_inKhz; + + return; +} + +void dccg2_init(struct dccg *dccg) +{ + struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + + // Fallthrough intentional to program all available dpp_dto's + switch (dccg_dcn->base.ctx->dc->res_pool->pipe_count) { + case 6: + REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[5], 1); + case 5: + REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[4], 1); + case 4: + REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[3], 1); + case 3: + REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[2], 1); + case 2: + REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[1], 1); + case 1: + REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[0], 1); + break; + default: + ASSERT(false); + break; + } +} + +static const struct dccg_funcs dccg2_funcs = { + .update_dpp_dto = dccg2_update_dpp_dto, + .get_dccg_ref_freq = dccg2_get_dccg_ref_freq, + .dccg_init = dccg2_init +}; + +struct dccg *dccg2_create( + struct dc_context *ctx, + const struct dccg_registers *regs, + const struct dccg_shift *dccg_shift, + const struct dccg_mask *dccg_mask) +{ + struct dcn_dccg *dccg_dcn = kzalloc(sizeof(*dccg_dcn), GFP_KERNEL); + struct dccg *base; + + if (dccg_dcn == NULL) { + BREAK_TO_DEBUGGER(); + return NULL; + } + + base = &dccg_dcn->base; + base->ctx = ctx; + base->funcs = &dccg2_funcs; + + dccg_dcn->regs = regs; + dccg_dcn->dccg_shift = dccg_shift; + dccg_dcn->dccg_mask = dccg_mask; + + return &dccg_dcn->base; +} + +void dcn_dccg_destroy(struct dccg **dccg) +{ + struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(*dccg); + + kfree(dccg_dcn); + *dccg = NULL; +} diff --cc drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c index c5ac25980f199,0000000000000..a8ba7d15abbb0 mode 100644,000000..100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c @@@ -1,3175 -1,0 +1,3177 @@@ +/* +* Copyright 2016 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + ++#include ++ +#include "dm_services.h" +#include "dc.h" + +#include "resource.h" +#include "include/irq_service_interface.h" +#include "dcn20/dcn20_resource.h" + +#include "dcn10/dcn10_hubp.h" +#include "dcn10/dcn10_ipp.h" +#include "dcn20_hubbub.h" +#include "dcn20_mpc.h" +#include "dcn20_hubp.h" +#include "irq/dcn20/irq_service_dcn20.h" +#include "dcn20_dpp.h" +#include "dcn20_optc.h" +#include "dcn20_hwseq.h" +#include "dce110/dce110_hw_sequencer.h" +#include "dcn10/dcn10_resource.h" +#include "dcn20_opp.h" + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT +#include "dcn20_dsc.h" +#endif + +#include "dcn20_link_encoder.h" +#include "dcn20_stream_encoder.h" +#include "dce/dce_clock_source.h" +#include "dce/dce_audio.h" +#include "dce/dce_hwseq.h" +#include "virtual/virtual_stream_encoder.h" +#include "dce110/dce110_resource.h" +#include "dml/display_mode_vba.h" +#include "dcn20_dccg.h" +#include "dcn20_vmid.h" + +#include "navi10_ip_offset.h" + +#include "dcn/dcn_2_0_0_offset.h" +#include "dcn/dcn_2_0_0_sh_mask.h" + +#include "nbio/nbio_2_3_offset.h" + +#include "dcn20/dcn20_dwb.h" +#include "dcn20/dcn20_mmhubbub.h" + +#include "mmhub/mmhub_2_0_0_offset.h" +#include "mmhub/mmhub_2_0_0_sh_mask.h" + +#include "reg_helper.h" +#include "dce/dce_abm.h" +#include "dce/dce_dmcu.h" +#include "dce/dce_aux.h" +#include "dce/dce_i2c.h" +#include "vm_helper.h" + +#include "amdgpu_socbb.h" + +#define SOC_BOUNDING_BOX_VALID false +#define DC_LOGGER_INIT(logger) + +struct _vcs_dpi_ip_params_st dcn2_0_ip = { + .odm_capable = 1, + .gpuvm_enable = 0, + .hostvm_enable = 0, + .gpuvm_max_page_table_levels = 4, + .hostvm_max_page_table_levels = 4, + .hostvm_cached_page_table_levels = 0, + .pte_group_size_bytes = 2048, +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + .num_dsc = 6, +#else + .num_dsc = 0, +#endif + .rob_buffer_size_kbytes = 168, + .det_buffer_size_kbytes = 164, + .dpte_buffer_size_in_pte_reqs_luma = 84, + .pde_proc_buffer_size_64k_reqs = 48, + .dpp_output_buffer_pixels = 2560, + .opp_output_buffer_lines = 1, + .pixel_chunk_size_kbytes = 8, + .pte_chunk_size_kbytes = 2, + .meta_chunk_size_kbytes = 2, + .writeback_chunk_size_kbytes = 2, + .line_buffer_size_bits = 789504, + .is_line_buffer_bpp_fixed = 0, + .line_buffer_fixed_bpp = 0, + .dcc_supported = true, + .max_line_buffer_lines = 12, + .writeback_luma_buffer_size_kbytes = 12, + .writeback_chroma_buffer_size_kbytes = 8, + .writeback_chroma_line_buffer_width_pixels = 4, + .writeback_max_hscl_ratio = 1, + .writeback_max_vscl_ratio = 1, + .writeback_min_hscl_ratio = 1, + .writeback_min_vscl_ratio = 1, + .writeback_max_hscl_taps = 12, + .writeback_max_vscl_taps = 12, + .writeback_line_buffer_luma_buffer_size = 0, + .writeback_line_buffer_chroma_buffer_size = 14643, + .cursor_buffer_size = 8, + .cursor_chunk_size = 2, + .max_num_otg = 6, + .max_num_dpp = 6, + .max_num_wb = 1, + .max_dchub_pscl_bw_pix_per_clk = 4, + .max_pscl_lb_bw_pix_per_clk = 2, + .max_lb_vscl_bw_pix_per_clk = 4, + .max_vscl_hscl_bw_pix_per_clk = 4, + .max_hscl_ratio = 8, + .max_vscl_ratio = 8, + .hscl_mults = 4, + .vscl_mults = 4, + .max_hscl_taps = 8, + .max_vscl_taps = 8, + .dispclk_ramp_margin_percent = 1, + .underscan_factor = 1.10, + .min_vblank_lines = 32, // + .dppclk_delay_subtotal = 77, // + .dppclk_delay_scl_lb_only = 16, + .dppclk_delay_scl = 50, + .dppclk_delay_cnvc_formatter = 8, + .dppclk_delay_cnvc_cursor = 6, + .dispclk_delay_subtotal = 87, // + .dcfclk_cstate_latency = 10, // SRExitTime + .max_inter_dcn_tile_repeaters = 8, + + .xfc_supported = true, + .xfc_fill_bw_overhead_percent = 10.0, + .xfc_fill_constant_bytes = 0, +}; + +struct _vcs_dpi_soc_bounding_box_st dcn2_0_soc = { 0 }; + + +#ifndef mmDP0_DP_DPHY_INTERNAL_CTRL + #define mmDP0_DP_DPHY_INTERNAL_CTRL 0x210f + #define mmDP0_DP_DPHY_INTERNAL_CTRL_BASE_IDX 2 + #define mmDP1_DP_DPHY_INTERNAL_CTRL 0x220f + #define mmDP1_DP_DPHY_INTERNAL_CTRL_BASE_IDX 2 + #define mmDP2_DP_DPHY_INTERNAL_CTRL 0x230f + #define mmDP2_DP_DPHY_INTERNAL_CTRL_BASE_IDX 2 + #define mmDP3_DP_DPHY_INTERNAL_CTRL 0x240f + #define mmDP3_DP_DPHY_INTERNAL_CTRL_BASE_IDX 2 + #define mmDP4_DP_DPHY_INTERNAL_CTRL 0x250f + #define mmDP4_DP_DPHY_INTERNAL_CTRL_BASE_IDX 2 + #define mmDP5_DP_DPHY_INTERNAL_CTRL 0x260f + #define mmDP5_DP_DPHY_INTERNAL_CTRL_BASE_IDX 2 + #define mmDP6_DP_DPHY_INTERNAL_CTRL 0x270f + #define mmDP6_DP_DPHY_INTERNAL_CTRL_BASE_IDX 2 +#endif + + +enum dcn20_clk_src_array_id { + DCN20_CLK_SRC_PLL0, + DCN20_CLK_SRC_PLL1, + DCN20_CLK_SRC_PLL2, + DCN20_CLK_SRC_PLL3, + DCN20_CLK_SRC_PLL4, + DCN20_CLK_SRC_PLL5, + DCN20_CLK_SRC_TOTAL +}; + +/* begin ********************* + * macros to expend register list macro defined in HW object header file */ + +/* DCN */ +/* TODO awful hack. fixup dcn20_dwb.h */ +#undef BASE_INNER +#define BASE_INNER(seg) DCN_BASE__INST0_SEG ## seg + +#define BASE(seg) BASE_INNER(seg) + +#define SR(reg_name)\ + .reg_name = BASE(mm ## reg_name ## _BASE_IDX) + \ + mm ## reg_name + +#define SRI(reg_name, block, id)\ + .reg_name = BASE(mm ## block ## id ## _ ## reg_name ## _BASE_IDX) + \ + mm ## block ## id ## _ ## reg_name + +#define SRIR(var_name, reg_name, block, id)\ + .var_name = BASE(mm ## block ## id ## _ ## reg_name ## _BASE_IDX) + \ + mm ## block ## id ## _ ## reg_name + +#define SRII(reg_name, block, id)\ + .reg_name[id] = BASE(mm ## block ## id ## _ ## reg_name ## _BASE_IDX) + \ + mm ## block ## id ## _ ## reg_name + +#define DCCG_SRII(reg_name, block, id)\ + .block ## _ ## reg_name[id] = BASE(mm ## block ## id ## _ ## reg_name ## _BASE_IDX) + \ + mm ## block ## id ## _ ## reg_name + +/* NBIO */ +#define NBIO_BASE_INNER(seg) \ + NBIO_BASE__INST0_SEG ## seg + +#define NBIO_BASE(seg) \ + NBIO_BASE_INNER(seg) + +#define NBIO_SR(reg_name)\ + .reg_name = NBIO_BASE(mm ## reg_name ## _BASE_IDX) + \ + mm ## reg_name + +/* MMHUB */ +#define MMHUB_BASE_INNER(seg) \ + MMHUB_BASE__INST0_SEG ## seg + +#define MMHUB_BASE(seg) \ + MMHUB_BASE_INNER(seg) + +#define MMHUB_SR(reg_name)\ + .reg_name = MMHUB_BASE(mmMM ## reg_name ## _BASE_IDX) + \ + mmMM ## reg_name + +static const struct bios_registers bios_regs = { + NBIO_SR(BIOS_SCRATCH_3), + NBIO_SR(BIOS_SCRATCH_6) +}; + +#define clk_src_regs(index, pllid)\ +[index] = {\ + CS_COMMON_REG_LIST_DCN2_0(index, pllid),\ +} + +static const struct dce110_clk_src_regs clk_src_regs[] = { + clk_src_regs(0, A), + clk_src_regs(1, B), + clk_src_regs(2, C), + clk_src_regs(3, D), + clk_src_regs(4, E), + clk_src_regs(5, F) +}; + +static const struct dce110_clk_src_shift cs_shift = { + CS_COMMON_MASK_SH_LIST_DCN2_0(__SHIFT) +}; + +static const struct dce110_clk_src_mask cs_mask = { + CS_COMMON_MASK_SH_LIST_DCN2_0(_MASK) +}; + +static const struct dce_dmcu_registers dmcu_regs = { + DMCU_DCN10_REG_LIST() +}; + +static const struct dce_dmcu_shift dmcu_shift = { + DMCU_MASK_SH_LIST_DCN10(__SHIFT) +}; + +static const struct dce_dmcu_mask dmcu_mask = { + DMCU_MASK_SH_LIST_DCN10(_MASK) +}; + +static const struct dce_abm_registers abm_regs = { + ABM_DCN20_REG_LIST() +}; + +static const struct dce_abm_shift abm_shift = { + ABM_MASK_SH_LIST_DCN20(__SHIFT) +}; + +static const struct dce_abm_mask abm_mask = { + ABM_MASK_SH_LIST_DCN20(_MASK) +}; + +#define audio_regs(id)\ +[id] = {\ + AUD_COMMON_REG_LIST(id)\ +} + +static const struct dce_audio_registers audio_regs[] = { + audio_regs(0), + audio_regs(1), + audio_regs(2), + audio_regs(3), + audio_regs(4), + audio_regs(5), + audio_regs(6), +}; + +#define DCE120_AUD_COMMON_MASK_SH_LIST(mask_sh)\ + SF(AZF0ENDPOINT0_AZALIA_F0_CODEC_ENDPOINT_INDEX, AZALIA_ENDPOINT_REG_INDEX, mask_sh),\ + SF(AZF0ENDPOINT0_AZALIA_F0_CODEC_ENDPOINT_DATA, AZALIA_ENDPOINT_REG_DATA, mask_sh),\ + AUD_COMMON_MASK_SH_LIST_BASE(mask_sh) + +static const struct dce_audio_shift audio_shift = { + DCE120_AUD_COMMON_MASK_SH_LIST(__SHIFT) +}; + +static const struct dce_aduio_mask audio_mask = { + DCE120_AUD_COMMON_MASK_SH_LIST(_MASK) +}; + +#define stream_enc_regs(id)\ +[id] = {\ + SE_DCN2_REG_LIST(id)\ +} + +static const struct dcn10_stream_enc_registers stream_enc_regs[] = { + stream_enc_regs(0), + stream_enc_regs(1), + stream_enc_regs(2), + stream_enc_regs(3), + stream_enc_regs(4), + stream_enc_regs(5), +}; + +static const struct dcn10_stream_encoder_shift se_shift = { + SE_COMMON_MASK_SH_LIST_DCN20(__SHIFT) +}; + +static const struct dcn10_stream_encoder_mask se_mask = { + SE_COMMON_MASK_SH_LIST_DCN20(_MASK) +}; + + +#define aux_regs(id)\ +[id] = {\ + DCN2_AUX_REG_LIST(id)\ +} + +static const struct dcn10_link_enc_aux_registers link_enc_aux_regs[] = { + aux_regs(0), + aux_regs(1), + aux_regs(2), + aux_regs(3), + aux_regs(4), + aux_regs(5) +}; + +#define hpd_regs(id)\ +[id] = {\ + HPD_REG_LIST(id)\ +} + +static const struct dcn10_link_enc_hpd_registers link_enc_hpd_regs[] = { + hpd_regs(0), + hpd_regs(1), + hpd_regs(2), + hpd_regs(3), + hpd_regs(4), + hpd_regs(5) +}; + +#define link_regs(id, phyid)\ +[id] = {\ + LE_DCN10_REG_LIST(id), \ + UNIPHY_DCN2_REG_LIST(phyid), \ + SRI(DP_DPHY_INTERNAL_CTRL, DP, id) \ +} + +static const struct dcn10_link_enc_registers link_enc_regs[] = { + link_regs(0, A), + link_regs(1, B), + link_regs(2, C), + link_regs(3, D), + link_regs(4, E), + link_regs(5, F) +}; + +static const struct dcn10_link_enc_shift le_shift = { + LINK_ENCODER_MASK_SH_LIST_DCN20(__SHIFT) +}; + +static const struct dcn10_link_enc_mask le_mask = { + LINK_ENCODER_MASK_SH_LIST_DCN20(_MASK) +}; + +#define ipp_regs(id)\ +[id] = {\ + IPP_REG_LIST_DCN20(id),\ +} + +static const struct dcn10_ipp_registers ipp_regs[] = { + ipp_regs(0), + ipp_regs(1), + ipp_regs(2), + ipp_regs(3), + ipp_regs(4), + ipp_regs(5), +}; + +static const struct dcn10_ipp_shift ipp_shift = { + IPP_MASK_SH_LIST_DCN20(__SHIFT) +}; + +static const struct dcn10_ipp_mask ipp_mask = { + IPP_MASK_SH_LIST_DCN20(_MASK), +}; + +#define opp_regs(id)\ +[id] = {\ + OPP_REG_LIST_DCN20(id),\ +} + +static const struct dcn20_opp_registers opp_regs[] = { + opp_regs(0), + opp_regs(1), + opp_regs(2), + opp_regs(3), + opp_regs(4), + opp_regs(5), +}; + +static const struct dcn20_opp_shift opp_shift = { + OPP_MASK_SH_LIST_DCN20(__SHIFT) +}; + +static const struct dcn20_opp_mask opp_mask = { + OPP_MASK_SH_LIST_DCN20(_MASK) +}; + +#define aux_engine_regs(id)\ +[id] = {\ + AUX_COMMON_REG_LIST0(id), \ + .AUXN_IMPCAL = 0, \ + .AUXP_IMPCAL = 0, \ + .AUX_RESET_MASK = DP_AUX0_AUX_CONTROL__AUX_RESET_MASK, \ +} + +static const struct dce110_aux_registers aux_engine_regs[] = { + aux_engine_regs(0), + aux_engine_regs(1), + aux_engine_regs(2), + aux_engine_regs(3), + aux_engine_regs(4), + aux_engine_regs(5) +}; + +#define tf_regs(id)\ +[id] = {\ + TF_REG_LIST_DCN20(id),\ +} + +static const struct dcn2_dpp_registers tf_regs[] = { + tf_regs(0), + tf_regs(1), + tf_regs(2), + tf_regs(3), + tf_regs(4), + tf_regs(5), +}; + +static const struct dcn2_dpp_shift tf_shift = { + TF_REG_LIST_SH_MASK_DCN20(__SHIFT) +}; + +static const struct dcn2_dpp_mask tf_mask = { + TF_REG_LIST_SH_MASK_DCN20(_MASK) +}; + +#define dwbc_regs_dcn2(id)\ +[id] = {\ + DWBC_COMMON_REG_LIST_DCN2_0(id),\ + } + +static const struct dcn20_dwbc_registers dwbc20_regs[] = { + dwbc_regs_dcn2(0), +}; + +static const struct dcn20_dwbc_shift dwbc20_shift = { + DWBC_COMMON_MASK_SH_LIST_DCN2_0(__SHIFT) +}; + +static const struct dcn20_dwbc_mask dwbc20_mask = { + DWBC_COMMON_MASK_SH_LIST_DCN2_0(_MASK) +}; + +#define mcif_wb_regs_dcn2(id)\ +[id] = {\ + MCIF_WB_COMMON_REG_LIST_DCN2_0(id),\ + } + +static const struct dcn20_mmhubbub_registers mcif_wb20_regs[] = { + mcif_wb_regs_dcn2(0), +}; + +static const struct dcn20_mmhubbub_shift mcif_wb20_shift = { + MCIF_WB_COMMON_MASK_SH_LIST_DCN2_0(__SHIFT) +}; + +static const struct dcn20_mmhubbub_mask mcif_wb20_mask = { + MCIF_WB_COMMON_MASK_SH_LIST_DCN2_0(_MASK) +}; + +static const struct dcn20_mpc_registers mpc_regs = { + MPC_REG_LIST_DCN2_0(0), + MPC_REG_LIST_DCN2_0(1), + MPC_REG_LIST_DCN2_0(2), + MPC_REG_LIST_DCN2_0(3), + MPC_REG_LIST_DCN2_0(4), + MPC_REG_LIST_DCN2_0(5), + MPC_OUT_MUX_REG_LIST_DCN2_0(0), + MPC_OUT_MUX_REG_LIST_DCN2_0(1), + MPC_OUT_MUX_REG_LIST_DCN2_0(2), + MPC_OUT_MUX_REG_LIST_DCN2_0(3), + MPC_OUT_MUX_REG_LIST_DCN2_0(4), + MPC_OUT_MUX_REG_LIST_DCN2_0(5), +}; + +static const struct dcn20_mpc_shift mpc_shift = { + MPC_COMMON_MASK_SH_LIST_DCN2_0(__SHIFT) +}; + +static const struct dcn20_mpc_mask mpc_mask = { + MPC_COMMON_MASK_SH_LIST_DCN2_0(_MASK) +}; + +#define tg_regs(id)\ +[id] = {TG_COMMON_REG_LIST_DCN2_0(id)} + + +static const struct dcn_optc_registers tg_regs[] = { + tg_regs(0), + tg_regs(1), + tg_regs(2), + tg_regs(3), + tg_regs(4), + tg_regs(5) +}; + +static const struct dcn_optc_shift tg_shift = { + TG_COMMON_MASK_SH_LIST_DCN2_0(__SHIFT) +}; + +static const struct dcn_optc_mask tg_mask = { + TG_COMMON_MASK_SH_LIST_DCN2_0(_MASK) +}; + +#define hubp_regs(id)\ +[id] = {\ + HUBP_REG_LIST_DCN20(id)\ +} + +static const struct dcn_hubp2_registers hubp_regs[] = { + hubp_regs(0), + hubp_regs(1), + hubp_regs(2), + hubp_regs(3), + hubp_regs(4), + hubp_regs(5) +}; + +static const struct dcn_hubp2_shift hubp_shift = { + HUBP_MASK_SH_LIST_DCN20(__SHIFT) +}; + +static const struct dcn_hubp2_mask hubp_mask = { + HUBP_MASK_SH_LIST_DCN20(_MASK) +}; + +static const struct dcn_hubbub_registers hubbub_reg = { + HUBBUB_REG_LIST_DCN20(0) +}; + +static const struct dcn_hubbub_shift hubbub_shift = { + HUBBUB_MASK_SH_LIST_DCN20(__SHIFT) +}; + +static const struct dcn_hubbub_mask hubbub_mask = { + HUBBUB_MASK_SH_LIST_DCN20(_MASK) +}; + +#define vmid_regs(id)\ +[id] = {\ + DCN20_VMID_REG_LIST(id)\ +} + +static const struct dcn_vmid_registers vmid_regs[] = { + vmid_regs(0), + vmid_regs(1), + vmid_regs(2), + vmid_regs(3), + vmid_regs(4), + vmid_regs(5), + vmid_regs(6), + vmid_regs(7), + vmid_regs(8), + vmid_regs(9), + vmid_regs(10), + vmid_regs(11), + vmid_regs(12), + vmid_regs(13), + vmid_regs(14), + vmid_regs(15) +}; + +static const struct dcn20_vmid_shift vmid_shifts = { + DCN20_VMID_MASK_SH_LIST(__SHIFT) +}; + +static const struct dcn20_vmid_mask vmid_masks = { + DCN20_VMID_MASK_SH_LIST(_MASK) +}; + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT +#define dsc_regsDCN20(id)\ +[id] = {\ + DSC_REG_LIST_DCN20(id)\ +} + +static const struct dcn20_dsc_registers dsc_regs[] = { + dsc_regsDCN20(0), + dsc_regsDCN20(1), + dsc_regsDCN20(2), + dsc_regsDCN20(3), + dsc_regsDCN20(4), + dsc_regsDCN20(5) +}; + +static const struct dcn20_dsc_shift dsc_shift = { + DSC_REG_LIST_SH_MASK_DCN20(__SHIFT) +}; + +static const struct dcn20_dsc_mask dsc_mask = { + DSC_REG_LIST_SH_MASK_DCN20(_MASK) +}; +#endif + +static const struct dccg_registers dccg_regs = { + DCCG_REG_LIST_DCN2() +}; + +static const struct dccg_shift dccg_shift = { + DCCG_MASK_SH_LIST_DCN2(__SHIFT) +}; + +static const struct dccg_mask dccg_mask = { + DCCG_MASK_SH_LIST_DCN2(_MASK) +}; + +static const struct resource_caps res_cap_nv10 = { + .num_timing_generator = 6, + .num_opp = 6, + .num_video_plane = 6, + .num_audio = 7, + .num_stream_encoder = 6, + .num_pll = 6, + .num_dwb = 1, + .num_ddc = 6, + .num_vmid = 16, +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + .num_dsc = 6, +#endif +}; + +static const struct dc_plane_cap plane_cap = { + .type = DC_PLANE_TYPE_DCN_UNIVERSAL, + .blends_with_above = true, + .blends_with_below = true, + .per_pixel_alpha = true, + + .pixel_format_support = { + .argb8888 = true, + .nv12 = true, + .fp16 = true + }, + + .max_upscale_factor = { + .argb8888 = 16000, + .nv12 = 16000, + .fp16 = 1 + }, + + .max_downscale_factor = { + .argb8888 = 250, + .nv12 = 250, + .fp16 = 1 + } +}; + +static const struct dc_debug_options debug_defaults_drv = { + .disable_dmcu = true, + .force_abm_enable = false, + .timing_trace = false, + .clock_trace = true, + .disable_pplib_clock_request = true, + .pipe_split_policy = MPC_SPLIT_DYNAMIC, + .force_single_disp_pipe_split = true, + .disable_dcc = DCC_ENABLE, + .vsr_support = true, + .performance_trace = false, + .max_downscale_src_width = 5120,/*upto 5K*/ + .disable_pplib_wm_range = false, + .scl_reset_length10 = true, + .sanity_checks = false, + .disable_tri_buf = true, + .underflow_assert_delay_us = 0xFFFFFFFF, +}; + +static const struct dc_debug_options debug_defaults_diags = { + .disable_dmcu = true, + .force_abm_enable = false, + .timing_trace = true, + .clock_trace = true, + .disable_dpp_power_gate = true, + .disable_hubp_power_gate = true, + .disable_clock_gate = true, + .disable_pplib_clock_request = true, + .disable_pplib_wm_range = true, + .disable_stutter = true, + .scl_reset_length10 = true, + .underflow_assert_delay_us = 0xFFFFFFFF, +}; + +void dcn20_dpp_destroy(struct dpp **dpp) +{ + kfree(TO_DCN20_DPP(*dpp)); + *dpp = NULL; +} + +struct dpp *dcn20_dpp_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct dcn20_dpp *dpp = + kzalloc(sizeof(struct dcn20_dpp), GFP_KERNEL); + + if (!dpp) + return NULL; + + if (dpp2_construct(dpp, ctx, inst, + &tf_regs[inst], &tf_shift, &tf_mask)) + return &dpp->base; + + BREAK_TO_DEBUGGER(); + kfree(dpp); + return NULL; +} + +struct input_pixel_processor *dcn20_ipp_create( + struct dc_context *ctx, uint32_t inst) +{ + struct dcn10_ipp *ipp = + kzalloc(sizeof(struct dcn10_ipp), GFP_KERNEL); + + if (!ipp) { + BREAK_TO_DEBUGGER(); + return NULL; + } + + dcn20_ipp_construct(ipp, ctx, inst, + &ipp_regs[inst], &ipp_shift, &ipp_mask); + return &ipp->base; +} + + +struct output_pixel_processor *dcn20_opp_create( + struct dc_context *ctx, uint32_t inst) +{ + struct dcn20_opp *opp = + kzalloc(sizeof(struct dcn20_opp), GFP_KERNEL); + + if (!opp) { + BREAK_TO_DEBUGGER(); + return NULL; + } + + dcn20_opp_construct(opp, ctx, inst, + &opp_regs[inst], &opp_shift, &opp_mask); + return &opp->base; +} + +struct dce_aux *dcn20_aux_engine_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct aux_engine_dce110 *aux_engine = + kzalloc(sizeof(struct aux_engine_dce110), GFP_KERNEL); + + if (!aux_engine) + return NULL; + + dce110_aux_engine_construct(aux_engine, ctx, inst, + SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD, + &aux_engine_regs[inst]); + + return &aux_engine->base; +} +#define i2c_inst_regs(id) { I2C_HW_ENGINE_COMMON_REG_LIST(id) } + +static const struct dce_i2c_registers i2c_hw_regs[] = { + i2c_inst_regs(1), + i2c_inst_regs(2), + i2c_inst_regs(3), + i2c_inst_regs(4), + i2c_inst_regs(5), + i2c_inst_regs(6), +}; + +static const struct dce_i2c_shift i2c_shifts = { + I2C_COMMON_MASK_SH_LIST_DCN2(__SHIFT) +}; + +static const struct dce_i2c_mask i2c_masks = { + I2C_COMMON_MASK_SH_LIST_DCN2(_MASK) +}; + +struct dce_i2c_hw *dcn20_i2c_hw_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct dce_i2c_hw *dce_i2c_hw = + kzalloc(sizeof(struct dce_i2c_hw), GFP_KERNEL); + + if (!dce_i2c_hw) + return NULL; + + dcn2_i2c_hw_construct(dce_i2c_hw, ctx, inst, + &i2c_hw_regs[inst], &i2c_shifts, &i2c_masks); + + return dce_i2c_hw; +} +struct mpc *dcn20_mpc_create(struct dc_context *ctx) +{ + struct dcn20_mpc *mpc20 = kzalloc(sizeof(struct dcn20_mpc), + GFP_KERNEL); + + if (!mpc20) + return NULL; + + dcn20_mpc_construct(mpc20, ctx, + &mpc_regs, + &mpc_shift, + &mpc_mask, + 6); + + return &mpc20->base; +} + +struct hubbub *dcn20_hubbub_create(struct dc_context *ctx) +{ + int i; + struct dcn20_hubbub *hubbub = kzalloc(sizeof(struct dcn20_hubbub), + GFP_KERNEL); + + if (!hubbub) + return NULL; + + hubbub2_construct(hubbub, ctx, + &hubbub_reg, + &hubbub_shift, + &hubbub_mask); + + for (i = 0; i < res_cap_nv10.num_vmid; i++) { + struct dcn20_vmid *vmid = &hubbub->vmid[i]; + + vmid->ctx = ctx; + + vmid->regs = &vmid_regs[i]; + vmid->shifts = &vmid_shifts; + vmid->masks = &vmid_masks; + } + + return &hubbub->base; +} + +struct timing_generator *dcn20_timing_generator_create( + struct dc_context *ctx, + uint32_t instance) +{ + struct optc *tgn10 = + kzalloc(sizeof(struct optc), GFP_KERNEL); + + if (!tgn10) + return NULL; + + tgn10->base.inst = instance; + tgn10->base.ctx = ctx; + + tgn10->tg_regs = &tg_regs[instance]; + tgn10->tg_shift = &tg_shift; + tgn10->tg_mask = &tg_mask; + + dcn20_timing_generator_init(tgn10); + + return &tgn10->base; +} + +static const struct encoder_feature_support link_enc_feature = { + .max_hdmi_deep_color = COLOR_DEPTH_121212, + .max_hdmi_pixel_clock = 600000, + .hdmi_ycbcr420_supported = true, + .dp_ycbcr420_supported = true, + .flags.bits.IS_HBR2_CAPABLE = true, + .flags.bits.IS_HBR3_CAPABLE = true, + .flags.bits.IS_TPS3_CAPABLE = true, + .flags.bits.IS_TPS4_CAPABLE = true +}; + +struct link_encoder *dcn20_link_encoder_create( + const struct encoder_init_data *enc_init_data) +{ + struct dcn20_link_encoder *enc20 = + kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); + + if (!enc20) + return NULL; + + dcn20_link_encoder_construct(enc20, + enc_init_data, + &link_enc_feature, + &link_enc_regs[enc_init_data->transmitter], + &link_enc_aux_regs[enc_init_data->channel - 1], + &link_enc_hpd_regs[enc_init_data->hpd_source], + &le_shift, + &le_mask); + + return &enc20->enc10.base; +} + +struct clock_source *dcn20_clock_source_create( + struct dc_context *ctx, + struct dc_bios *bios, + enum clock_source_id id, + const struct dce110_clk_src_regs *regs, + bool dp_clk_src) +{ + struct dce110_clk_src *clk_src = + kzalloc(sizeof(struct dce110_clk_src), GFP_KERNEL); + + if (!clk_src) + return NULL; + + if (dcn20_clk_src_construct(clk_src, ctx, bios, id, + regs, &cs_shift, &cs_mask)) { + clk_src->base.dp_clk_src = dp_clk_src; + return &clk_src->base; + } + + BREAK_TO_DEBUGGER(); + return NULL; +} + +static void read_dce_straps( + struct dc_context *ctx, + struct resource_straps *straps) +{ + generic_reg_get(ctx, mmDC_PINSTRAPS + BASE(mmDC_PINSTRAPS_BASE_IDX), + FN(DC_PINSTRAPS, DC_PINSTRAPS_AUDIO), &straps->dc_pinstraps_audio); +} + +static struct audio *dcn20_create_audio( + struct dc_context *ctx, unsigned int inst) +{ + return dce_audio_create(ctx, inst, + &audio_regs[inst], &audio_shift, &audio_mask); +} + +struct stream_encoder *dcn20_stream_encoder_create( + enum engine_id eng_id, + struct dc_context *ctx) +{ + struct dcn10_stream_encoder *enc1 = + kzalloc(sizeof(struct dcn10_stream_encoder), GFP_KERNEL); + + if (!enc1) + return NULL; + + dcn20_stream_encoder_construct(enc1, ctx, ctx->dc_bios, eng_id, + &stream_enc_regs[eng_id], + &se_shift, &se_mask); + + return &enc1->base; +} + +static const struct dce_hwseq_registers hwseq_reg = { + HWSEQ_DCN2_REG_LIST() +}; + +static const struct dce_hwseq_shift hwseq_shift = { + HWSEQ_DCN2_MASK_SH_LIST(__SHIFT) +}; + +static const struct dce_hwseq_mask hwseq_mask = { + HWSEQ_DCN2_MASK_SH_LIST(_MASK) +}; + +struct dce_hwseq *dcn20_hwseq_create( + struct dc_context *ctx) +{ + struct dce_hwseq *hws = kzalloc(sizeof(struct dce_hwseq), GFP_KERNEL); + + if (hws) { + hws->ctx = ctx; + hws->regs = &hwseq_reg; + hws->shifts = &hwseq_shift; + hws->masks = &hwseq_mask; + } + return hws; +} + +static const struct resource_create_funcs res_create_funcs = { + .read_dce_straps = read_dce_straps, + .create_audio = dcn20_create_audio, + .create_stream_encoder = dcn20_stream_encoder_create, + .create_hwseq = dcn20_hwseq_create, +}; + +static const struct resource_create_funcs res_create_maximus_funcs = { + .read_dce_straps = NULL, + .create_audio = NULL, + .create_stream_encoder = NULL, + .create_hwseq = dcn20_hwseq_create, +}; + +void dcn20_clock_source_destroy(struct clock_source **clk_src) +{ + kfree(TO_DCE110_CLK_SRC(*clk_src)); + *clk_src = NULL; +} + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + +struct display_stream_compressor *dcn20_dsc_create( + struct dc_context *ctx, uint32_t inst) +{ + struct dcn20_dsc *dsc = + kzalloc(sizeof(struct dcn20_dsc), GFP_KERNEL); + + if (!dsc) { + BREAK_TO_DEBUGGER(); + return NULL; + } + + dsc2_construct(dsc, ctx, inst, &dsc_regs[inst], &dsc_shift, &dsc_mask); + return &dsc->base; +} + +void dcn20_dsc_destroy(struct display_stream_compressor **dsc) +{ + kfree(container_of(*dsc, struct dcn20_dsc, base)); + *dsc = NULL; +} + +#endif + +static void destruct(struct dcn20_resource_pool *pool) +{ + unsigned int i; + + for (i = 0; i < pool->base.stream_enc_count; i++) { + if (pool->base.stream_enc[i] != NULL) { + kfree(DCN10STRENC_FROM_STRENC(pool->base.stream_enc[i])); + pool->base.stream_enc[i] = NULL; + } + } + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + for (i = 0; i < pool->base.res_cap->num_dsc; i++) { + if (pool->base.dscs[i] != NULL) + dcn20_dsc_destroy(&pool->base.dscs[i]); + } +#endif + + if (pool->base.mpc != NULL) { + kfree(TO_DCN20_MPC(pool->base.mpc)); + pool->base.mpc = NULL; + } + if (pool->base.hubbub != NULL) { + kfree(pool->base.hubbub); + pool->base.hubbub = NULL; + } + for (i = 0; i < pool->base.pipe_count; i++) { + if (pool->base.dpps[i] != NULL) + dcn20_dpp_destroy(&pool->base.dpps[i]); + + if (pool->base.ipps[i] != NULL) + pool->base.ipps[i]->funcs->ipp_destroy(&pool->base.ipps[i]); + + if (pool->base.hubps[i] != NULL) { + kfree(TO_DCN20_HUBP(pool->base.hubps[i])); + pool->base.hubps[i] = NULL; + } + + if (pool->base.irqs != NULL) { + dal_irq_service_destroy(&pool->base.irqs); + } + } + + for (i = 0; i < pool->base.res_cap->num_ddc; i++) { + if (pool->base.engines[i] != NULL) + dce110_engine_destroy(&pool->base.engines[i]); + if (pool->base.hw_i2cs[i] != NULL) { + kfree(pool->base.hw_i2cs[i]); + pool->base.hw_i2cs[i] = NULL; + } + if (pool->base.sw_i2cs[i] != NULL) { + kfree(pool->base.sw_i2cs[i]); + pool->base.sw_i2cs[i] = NULL; + } + } + + for (i = 0; i < pool->base.res_cap->num_opp; i++) { + if (pool->base.opps[i] != NULL) + pool->base.opps[i]->funcs->opp_destroy(&pool->base.opps[i]); + } + + for (i = 0; i < pool->base.res_cap->num_timing_generator; i++) { + if (pool->base.timing_generators[i] != NULL) { + kfree(DCN10TG_FROM_TG(pool->base.timing_generators[i])); + pool->base.timing_generators[i] = NULL; + } + } + + for (i = 0; i < pool->base.res_cap->num_dwb; i++) { + if (pool->base.dwbc[i] != NULL) { + kfree(TO_DCN20_DWBC(pool->base.dwbc[i])); + pool->base.dwbc[i] = NULL; + } + if (pool->base.mcif_wb[i] != NULL) { + kfree(TO_DCN20_MMHUBBUB(pool->base.mcif_wb[i])); + pool->base.mcif_wb[i] = NULL; + } + } + + for (i = 0; i < pool->base.audio_count; i++) { + if (pool->base.audios[i]) + dce_aud_destroy(&pool->base.audios[i]); + } + + for (i = 0; i < pool->base.clk_src_count; i++) { + if (pool->base.clock_sources[i] != NULL) { + dcn20_clock_source_destroy(&pool->base.clock_sources[i]); + pool->base.clock_sources[i] = NULL; + } + } + + if (pool->base.dp_clock_source != NULL) { + dcn20_clock_source_destroy(&pool->base.dp_clock_source); + pool->base.dp_clock_source = NULL; + } + + + if (pool->base.abm != NULL) + dce_abm_destroy(&pool->base.abm); + + if (pool->base.dmcu != NULL) + dce_dmcu_destroy(&pool->base.dmcu); + + if (pool->base.dccg != NULL) + dcn_dccg_destroy(&pool->base.dccg); + + if (pool->base.pp_smu != NULL) + dcn20_pp_smu_destroy(&pool->base.pp_smu); + +} + +struct hubp *dcn20_hubp_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct dcn20_hubp *hubp2 = + kzalloc(sizeof(struct dcn20_hubp), GFP_KERNEL); + + if (!hubp2) + return NULL; + + if (hubp2_construct(hubp2, ctx, inst, + &hubp_regs[inst], &hubp_shift, &hubp_mask)) + return &hubp2->base; + + BREAK_TO_DEBUGGER(); + kfree(hubp2); + return NULL; +} + +static void get_pixel_clock_parameters( + struct pipe_ctx *pipe_ctx, + struct pixel_clk_params *pixel_clk_params) +{ + const struct dc_stream_state *stream = pipe_ctx->stream; + bool odm_combine = dc_res_get_odm_bottom_pipe(pipe_ctx) != NULL; + + pixel_clk_params->requested_pix_clk_100hz = stream->timing.pix_clk_100hz; + pixel_clk_params->encoder_object_id = stream->link->link_enc->id; + pixel_clk_params->signal_type = pipe_ctx->stream->signal; + pixel_clk_params->controller_id = pipe_ctx->stream_res.tg->inst + 1; + /* TODO: un-hardcode*/ + pixel_clk_params->requested_sym_clk = LINK_RATE_LOW * + LINK_RATE_REF_FREQ_IN_KHZ; + pixel_clk_params->flags.ENABLE_SS = 0; + pixel_clk_params->color_depth = + stream->timing.display_color_depth; + pixel_clk_params->flags.DISPLAY_BLANKED = 1; + pixel_clk_params->pixel_encoding = stream->timing.pixel_encoding; + + if (stream->timing.pixel_encoding == PIXEL_ENCODING_YCBCR422) + pixel_clk_params->color_depth = COLOR_DEPTH_888; + + if (optc1_is_two_pixels_per_containter(&stream->timing) || odm_combine) + pixel_clk_params->requested_pix_clk_100hz /= 2; + + if (stream->timing.timing_3d_format == TIMING_3D_FORMAT_HW_FRAME_PACKING) + pixel_clk_params->requested_pix_clk_100hz *= 2; + +} + +static void build_clamping_params(struct dc_stream_state *stream) +{ + stream->clamping.clamping_level = CLAMPING_FULL_RANGE; + stream->clamping.c_depth = stream->timing.display_color_depth; + stream->clamping.pixel_encoding = stream->timing.pixel_encoding; +} + +static enum dc_status build_pipe_hw_param(struct pipe_ctx *pipe_ctx) +{ + + get_pixel_clock_parameters(pipe_ctx, &pipe_ctx->stream_res.pix_clk_params); + + pipe_ctx->clock_source->funcs->get_pix_clk_dividers( + pipe_ctx->clock_source, + &pipe_ctx->stream_res.pix_clk_params, + &pipe_ctx->pll_settings); + + pipe_ctx->stream->clamping.pixel_encoding = pipe_ctx->stream->timing.pixel_encoding; + + resource_build_bit_depth_reduction_params(pipe_ctx->stream, + &pipe_ctx->stream->bit_depth_params); + build_clamping_params(pipe_ctx->stream); + + return DC_OK; +} + +enum dc_status dcn20_build_mapped_resource(const struct dc *dc, struct dc_state *context, struct dc_stream_state *stream) +{ + enum dc_status status = DC_OK; + struct pipe_ctx *pipe_ctx = resource_get_head_pipe_for_stream(&context->res_ctx, stream); + + /*TODO Seems unneeded anymore */ + /* if (old_context && resource_is_stream_unchanged(old_context, stream)) { + if (stream != NULL && old_context->streams[i] != NULL) { + todo: shouldn't have to copy missing parameter here + resource_build_bit_depth_reduction_params(stream, + &stream->bit_depth_params); + stream->clamping.pixel_encoding = + stream->timing.pixel_encoding; + + resource_build_bit_depth_reduction_params(stream, + &stream->bit_depth_params); + build_clamping_params(stream); + + continue; + } + } + */ + + if (!pipe_ctx) + return DC_ERROR_UNEXPECTED; + + + status = build_pipe_hw_param(pipe_ctx); + + return status; +} + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + +static void acquire_dsc(struct resource_context *res_ctx, + const struct resource_pool *pool, + struct display_stream_compressor **dsc) +{ + int i; + + ASSERT(*dsc == NULL); + *dsc = NULL; + + /* Find first free DSC */ + for (i = 0; i < pool->res_cap->num_dsc; i++) + if (!res_ctx->is_dsc_acquired[i]) { + *dsc = pool->dscs[i]; + res_ctx->is_dsc_acquired[i] = true; + break; + } +} + +static void release_dsc(struct resource_context *res_ctx, + const struct resource_pool *pool, + struct display_stream_compressor **dsc) +{ + int i; + + for (i = 0; i < pool->res_cap->num_dsc; i++) + if (pool->dscs[i] == *dsc) { + res_ctx->is_dsc_acquired[i] = false; + *dsc = NULL; + break; + } +} + +#endif + + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT +static enum dc_status add_dsc_to_stream_resource(struct dc *dc, + struct dc_state *dc_ctx, + struct dc_stream_state *dc_stream) +{ + enum dc_status result = DC_OK; + int i; + const struct resource_pool *pool = dc->res_pool; + + /* Get a DSC if required and available */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe_ctx = &dc_ctx->res_ctx.pipe_ctx[i]; + + if (pipe_ctx->stream != dc_stream) + continue; + + acquire_dsc(&dc_ctx->res_ctx, pool, &pipe_ctx->stream_res.dsc); + + /* The number of DSCs can be less than the number of pipes */ + if (!pipe_ctx->stream_res.dsc) { + dm_output_to_console("No DSCs available\n"); + result = DC_NO_DSC_RESOURCE; + } + + break; + } + + return result; +} + + +static enum dc_status remove_dsc_from_stream_resource(struct dc *dc, + struct dc_state *new_ctx, + struct dc_stream_state *dc_stream) +{ + struct pipe_ctx *pipe_ctx = NULL; + int i; + + for (i = 0; i < MAX_PIPES; i++) { + if (new_ctx->res_ctx.pipe_ctx[i].stream == dc_stream && !new_ctx->res_ctx.pipe_ctx[i].top_pipe) { + pipe_ctx = &new_ctx->res_ctx.pipe_ctx[i]; + break; + } + } + + if (!pipe_ctx) + return DC_ERROR_UNEXPECTED; + + if (pipe_ctx->stream_res.dsc) { + struct pipe_ctx *odm_pipe = dc_res_get_odm_bottom_pipe(pipe_ctx); + + release_dsc(&new_ctx->res_ctx, dc->res_pool, &pipe_ctx->stream_res.dsc); + if (odm_pipe) + release_dsc(&new_ctx->res_ctx, dc->res_pool, &odm_pipe->stream_res.dsc); + } + + return DC_OK; +} +#endif + + +enum dc_status dcn20_add_stream_to_ctx(struct dc *dc, struct dc_state *new_ctx, struct dc_stream_state *dc_stream) +{ + enum dc_status result = DC_ERROR_UNEXPECTED; + + result = resource_map_pool_resources(dc, new_ctx, dc_stream); + + if (result == DC_OK) + result = resource_map_phy_clock_resources(dc, new_ctx, dc_stream); + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + /* Get a DSC if required and available */ + if (result == DC_OK && dc_stream->timing.flags.DSC) + result = add_dsc_to_stream_resource(dc, new_ctx, dc_stream); +#endif + + if (result == DC_OK) + result = dcn20_build_mapped_resource(dc, new_ctx, dc_stream); + + return result; +} + + +enum dc_status dcn20_remove_stream_from_ctx(struct dc *dc, struct dc_state *new_ctx, struct dc_stream_state *dc_stream) +{ + enum dc_status result = DC_OK; + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + result = remove_dsc_from_stream_resource(dc, new_ctx, dc_stream); +#endif + + return result; +} + + +static void swizzle_to_dml_params( + enum swizzle_mode_values swizzle, + unsigned int *sw_mode) +{ + switch (swizzle) { + case DC_SW_LINEAR: + *sw_mode = dm_sw_linear; + break; + case DC_SW_4KB_S: + *sw_mode = dm_sw_4kb_s; + break; + case DC_SW_4KB_S_X: + *sw_mode = dm_sw_4kb_s_x; + break; + case DC_SW_4KB_D: + *sw_mode = dm_sw_4kb_d; + break; + case DC_SW_4KB_D_X: + *sw_mode = dm_sw_4kb_d_x; + break; + case DC_SW_64KB_S: + *sw_mode = dm_sw_64kb_s; + break; + case DC_SW_64KB_S_X: + *sw_mode = dm_sw_64kb_s_x; + break; + case DC_SW_64KB_S_T: + *sw_mode = dm_sw_64kb_s_t; + break; + case DC_SW_64KB_D: + *sw_mode = dm_sw_64kb_d; + break; + case DC_SW_64KB_D_X: + *sw_mode = dm_sw_64kb_d_x; + break; + case DC_SW_64KB_D_T: + *sw_mode = dm_sw_64kb_d_t; + break; + case DC_SW_64KB_R_X: + *sw_mode = dm_sw_64kb_r_x; + break; + case DC_SW_VAR_S: + *sw_mode = dm_sw_var_s; + break; + case DC_SW_VAR_S_X: + *sw_mode = dm_sw_var_s_x; + break; + case DC_SW_VAR_D: + *sw_mode = dm_sw_var_d; + break; + case DC_SW_VAR_D_X: + *sw_mode = dm_sw_var_d_x; + break; + + default: + ASSERT(0); /* Not supported */ + break; + } +} + +static bool dcn20_split_stream_for_combine( + struct resource_context *res_ctx, + const struct resource_pool *pool, + struct pipe_ctx *primary_pipe, + struct pipe_ctx *secondary_pipe, + bool is_odm_combine) +{ + int pipe_idx = secondary_pipe->pipe_idx; + struct scaler_data *sd = &primary_pipe->plane_res.scl_data; + struct pipe_ctx *sec_bot_pipe = secondary_pipe->bottom_pipe; + int new_width; + + *secondary_pipe = *primary_pipe; + secondary_pipe->bottom_pipe = sec_bot_pipe; + + secondary_pipe->pipe_idx = pipe_idx; + secondary_pipe->plane_res.mi = pool->mis[secondary_pipe->pipe_idx]; + secondary_pipe->plane_res.hubp = pool->hubps[secondary_pipe->pipe_idx]; + secondary_pipe->plane_res.ipp = pool->ipps[secondary_pipe->pipe_idx]; + secondary_pipe->plane_res.xfm = pool->transforms[secondary_pipe->pipe_idx]; + secondary_pipe->plane_res.dpp = pool->dpps[secondary_pipe->pipe_idx]; + secondary_pipe->plane_res.mpcc_inst = pool->dpps[secondary_pipe->pipe_idx]->inst; +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + secondary_pipe->stream_res.dsc = NULL; +#endif + if (primary_pipe->bottom_pipe && primary_pipe->bottom_pipe != secondary_pipe) { + ASSERT(!secondary_pipe->bottom_pipe); + secondary_pipe->bottom_pipe = primary_pipe->bottom_pipe; + secondary_pipe->bottom_pipe->top_pipe = secondary_pipe; + } + primary_pipe->bottom_pipe = secondary_pipe; + secondary_pipe->top_pipe = primary_pipe; + + if (is_odm_combine) { + if (primary_pipe->plane_state) { + /* HACTIVE halved for odm combine */ + sd->h_active /= 2; + /* Copy scl_data to secondary pipe */ + secondary_pipe->plane_res.scl_data = *sd; + + /* Calculate new vp and recout for left pipe */ + /* Need at least 16 pixels width per side */ + if (sd->recout.x + 16 >= sd->h_active) + return false; + new_width = sd->h_active - sd->recout.x; + sd->viewport.width -= dc_fixpt_floor(dc_fixpt_mul_int( + sd->ratios.horz, sd->recout.width - new_width)); + sd->viewport_c.width -= dc_fixpt_floor(dc_fixpt_mul_int( + sd->ratios.horz_c, sd->recout.width - new_width)); + sd->recout.width = new_width; + + /* Calculate new vp and recout for right pipe */ + sd = &secondary_pipe->plane_res.scl_data; + new_width = sd->recout.width + sd->recout.x - sd->h_active; + /* Need at least 16 pixels width per side */ + if (new_width <= 16) + return false; + sd->viewport.width -= dc_fixpt_floor(dc_fixpt_mul_int( + sd->ratios.horz, sd->recout.width - new_width)); + sd->viewport_c.width -= dc_fixpt_floor(dc_fixpt_mul_int( + sd->ratios.horz_c, sd->recout.width - new_width)); + sd->recout.width = new_width; + sd->viewport.x += dc_fixpt_floor(dc_fixpt_mul_int( + sd->ratios.horz, sd->h_active - sd->recout.x)); + sd->viewport_c.x += dc_fixpt_floor(dc_fixpt_mul_int( + sd->ratios.horz_c, sd->h_active - sd->recout.x)); + sd->recout.x = 0; + } + secondary_pipe->stream_res.opp = pool->opps[secondary_pipe->pipe_idx]; +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + if (secondary_pipe->stream->timing.flags.DSC == 1) { + acquire_dsc(res_ctx, pool, &secondary_pipe->stream_res.dsc); + ASSERT(secondary_pipe->stream_res.dsc); + if (secondary_pipe->stream_res.dsc == NULL) + return false; + } +#endif + } else { + ASSERT(primary_pipe->plane_state); + resource_build_scaling_params(primary_pipe); + resource_build_scaling_params(secondary_pipe); + } + + return true; +} + +void dcn20_populate_dml_writeback_from_context( + struct dc *dc, struct resource_context *res_ctx, display_e2e_pipe_params_st *pipes) +{ + int pipe_cnt, i; + + for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) { + struct dc_writeback_info *wb_info = &res_ctx->pipe_ctx[i].stream->writeback_info[0]; + + if (!res_ctx->pipe_ctx[i].stream) + continue; + + /* Set writeback information */ + pipes[pipe_cnt].dout.wb_enable = (wb_info->wb_enabled == true) ? 1 : 0; + pipes[pipe_cnt].dout.num_active_wb++; + pipes[pipe_cnt].dout.wb.wb_src_height = wb_info->dwb_params.cnv_params.crop_height; + pipes[pipe_cnt].dout.wb.wb_src_width = wb_info->dwb_params.cnv_params.crop_width; + pipes[pipe_cnt].dout.wb.wb_dst_width = wb_info->dwb_params.dest_width; + pipes[pipe_cnt].dout.wb.wb_dst_height = wb_info->dwb_params.dest_height; + pipes[pipe_cnt].dout.wb.wb_htaps_luma = 1; + pipes[pipe_cnt].dout.wb.wb_vtaps_luma = 1; + pipes[pipe_cnt].dout.wb.wb_htaps_chroma = wb_info->dwb_params.scaler_taps.h_taps_c; + pipes[pipe_cnt].dout.wb.wb_vtaps_chroma = wb_info->dwb_params.scaler_taps.v_taps_c; + pipes[pipe_cnt].dout.wb.wb_hratio = 1.0; + pipes[pipe_cnt].dout.wb.wb_vratio = 1.0; + if (wb_info->dwb_params.out_format == dwb_scaler_mode_yuv420) { + if (wb_info->dwb_params.output_depth == DWB_OUTPUT_PIXEL_DEPTH_8BPC) + pipes[pipe_cnt].dout.wb.wb_pixel_format = dm_420_8; + else + pipes[pipe_cnt].dout.wb.wb_pixel_format = dm_420_10; + } else + pipes[pipe_cnt].dout.wb.wb_pixel_format = dm_444_32; + + pipe_cnt++; + } + +} + +int dcn20_populate_dml_pipes_from_context( + struct dc *dc, struct resource_context *res_ctx, display_e2e_pipe_params_st *pipes) +{ + int pipe_cnt, i; + bool synchronized_vblank = true; + + for (i = 0, pipe_cnt = -1; i < dc->res_pool->pipe_count; i++) { + if (!res_ctx->pipe_ctx[i].stream) + continue; + + if (pipe_cnt < 0) { + pipe_cnt = i; + continue; + } + if (!resource_are_streams_timing_synchronizable( + res_ctx->pipe_ctx[pipe_cnt].stream, + res_ctx->pipe_ctx[i].stream)) { + synchronized_vblank = false; + break; + } + } + + for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) { + struct dc_crtc_timing *timing = &res_ctx->pipe_ctx[i].stream->timing; + int output_bpc; + + if (!res_ctx->pipe_ctx[i].stream) + continue; + /* todo: + pipes[pipe_cnt].pipe.src.dynamic_metadata_enable = 0; + pipes[pipe_cnt].pipe.src.dcc = 0; + pipes[pipe_cnt].pipe.src.vm = 0;*/ + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + pipes[pipe_cnt].dout.dsc_enable = res_ctx->pipe_ctx[i].stream->timing.flags.DSC; + /* todo: rotation?*/ + pipes[pipe_cnt].dout.dsc_slices = res_ctx->pipe_ctx[i].stream->timing.dsc_cfg.num_slices_h; +#endif + if (res_ctx->pipe_ctx[i].stream->use_dynamic_meta) { + pipes[pipe_cnt].pipe.src.dynamic_metadata_enable = true; + /* 1/2 vblank */ + pipes[pipe_cnt].pipe.src.dynamic_metadata_lines_before_active = + (timing->v_total - timing->v_addressable + - timing->v_border_top - timing->v_border_bottom) / 2; + /* 36 bytes dp, 32 hdmi */ + pipes[pipe_cnt].pipe.src.dynamic_metadata_xmit_bytes = + dc_is_dp_signal(res_ctx->pipe_ctx[i].stream->signal) ? 36 : 32; + } + pipes[pipe_cnt].pipe.src.dcc = false; + pipes[pipe_cnt].pipe.src.dcc_rate = 1; + pipes[pipe_cnt].pipe.dest.synchronized_vblank_all_planes = synchronized_vblank; + pipes[pipe_cnt].pipe.dest.hblank_start = timing->h_total - timing->h_front_porch; + pipes[pipe_cnt].pipe.dest.hblank_end = pipes[pipe_cnt].pipe.dest.hblank_start + - timing->h_addressable + - timing->h_border_left + - timing->h_border_right; + pipes[pipe_cnt].pipe.dest.vblank_start = timing->v_total - timing->v_front_porch; + pipes[pipe_cnt].pipe.dest.vblank_end = pipes[pipe_cnt].pipe.dest.vblank_start + - timing->v_addressable + - timing->v_border_top + - timing->v_border_bottom; + pipes[pipe_cnt].pipe.dest.htotal = timing->h_total; + pipes[pipe_cnt].pipe.dest.vtotal = timing->v_total; + pipes[pipe_cnt].pipe.dest.hactive = timing->h_addressable; + pipes[pipe_cnt].pipe.dest.vactive = timing->v_addressable; + pipes[pipe_cnt].pipe.dest.interlaced = timing->flags.INTERLACE; + pipes[pipe_cnt].pipe.dest.pixel_rate_mhz = timing->pix_clk_100hz/10000.0; + if (timing->timing_3d_format == TIMING_3D_FORMAT_HW_FRAME_PACKING) + pipes[pipe_cnt].pipe.dest.pixel_rate_mhz *= 2; + pipes[pipe_cnt].pipe.dest.otg_inst = res_ctx->pipe_ctx[i].stream_res.tg->inst; + pipes[pipe_cnt].dout.dp_lanes = 4; + pipes[pipe_cnt].pipe.dest.vtotal_min = res_ctx->pipe_ctx[i].stream->adjust.v_total_min; + pipes[pipe_cnt].pipe.dest.vtotal_max = res_ctx->pipe_ctx[i].stream->adjust.v_total_max; + + switch (res_ctx->pipe_ctx[i].stream->signal) { + case SIGNAL_TYPE_DISPLAY_PORT_MST: + case SIGNAL_TYPE_DISPLAY_PORT: + pipes[pipe_cnt].dout.output_type = dm_dp; + break; + case SIGNAL_TYPE_EDP: + pipes[pipe_cnt].dout.output_type = dm_edp; + break; + case SIGNAL_TYPE_HDMI_TYPE_A: + case SIGNAL_TYPE_DVI_SINGLE_LINK: + case SIGNAL_TYPE_DVI_DUAL_LINK: + pipes[pipe_cnt].dout.output_type = dm_hdmi; + break; + default: + /* In case there is no signal, set dp with 4 lanes to allow max config */ + pipes[pipe_cnt].dout.output_type = dm_dp; + pipes[pipe_cnt].dout.dp_lanes = 4; + } + + switch (res_ctx->pipe_ctx[i].stream->timing.display_color_depth) { + case COLOR_DEPTH_666: + output_bpc = 6; + break; + case COLOR_DEPTH_888: + output_bpc = 8; + break; + case COLOR_DEPTH_101010: + output_bpc = 10; + break; + case COLOR_DEPTH_121212: + output_bpc = 12; + break; + case COLOR_DEPTH_141414: + output_bpc = 14; + break; + case COLOR_DEPTH_161616: + output_bpc = 16; + break; +#ifdef CONFIG_DRM_AMD_DC_DCN2_0 + case COLOR_DEPTH_999: + output_bpc = 9; + break; + case COLOR_DEPTH_111111: + output_bpc = 11; + break; +#endif + default: + output_bpc = 8; + break; + } + + + switch (res_ctx->pipe_ctx[i].stream->timing.pixel_encoding) { + case PIXEL_ENCODING_RGB: + case PIXEL_ENCODING_YCBCR444: + pipes[pipe_cnt].dout.output_format = dm_444; + pipes[pipe_cnt].dout.output_bpp = output_bpc * 3; + break; + case PIXEL_ENCODING_YCBCR420: + pipes[pipe_cnt].dout.output_format = dm_420; + pipes[pipe_cnt].dout.output_bpp = (output_bpc * 3) / 2; + break; + case PIXEL_ENCODING_YCBCR422: + if (true) /* todo */ + pipes[pipe_cnt].dout.output_format = dm_s422; + else + pipes[pipe_cnt].dout.output_format = dm_n422; + pipes[pipe_cnt].dout.output_bpp = output_bpc * 2; + break; + default: + pipes[pipe_cnt].dout.output_format = dm_444; + pipes[pipe_cnt].dout.output_bpp = output_bpc * 3; + } + pipes[pipe_cnt].pipe.src.hsplit_grp = res_ctx->pipe_ctx[i].pipe_idx; + if (res_ctx->pipe_ctx[i].top_pipe && res_ctx->pipe_ctx[i].top_pipe->plane_state + == res_ctx->pipe_ctx[i].plane_state) + pipes[pipe_cnt].pipe.src.hsplit_grp = res_ctx->pipe_ctx[i].top_pipe->pipe_idx; + + /* todo: default max for now, until there is logic reflecting this in dc*/ + pipes[pipe_cnt].dout.output_bpc = 12; + /* + * Use max cursor settings for calculations to minimize + * bw calculations due to cursor on/off + */ + pipes[pipe_cnt].pipe.src.num_cursors = 2; + pipes[pipe_cnt].pipe.src.cur0_src_width = 256; + pipes[pipe_cnt].pipe.src.cur0_bpp = dm_cur_32bit; + pipes[pipe_cnt].pipe.src.cur1_src_width = 256; + pipes[pipe_cnt].pipe.src.cur1_bpp = dm_cur_32bit; + + if (!res_ctx->pipe_ctx[i].plane_state) { + pipes[pipe_cnt].pipe.src.source_scan = dm_horz; + pipes[pipe_cnt].pipe.src.sw_mode = dm_sw_linear; + pipes[pipe_cnt].pipe.src.macro_tile_size = dm_64k_tile; + pipes[pipe_cnt].pipe.src.viewport_width = timing->h_addressable; + if (pipes[pipe_cnt].pipe.src.viewport_width > 1920) + pipes[pipe_cnt].pipe.src.viewport_width = 1920; + pipes[pipe_cnt].pipe.src.viewport_height = timing->v_addressable; + if (pipes[pipe_cnt].pipe.src.viewport_height > 1080) + pipes[pipe_cnt].pipe.src.viewport_height = 1080; + pipes[pipe_cnt].pipe.src.data_pitch = ((pipes[pipe_cnt].pipe.src.viewport_width + 63) / 64) * 64; /* linear sw only */ + pipes[pipe_cnt].pipe.src.source_format = dm_444_32; + pipes[pipe_cnt].pipe.dest.recout_width = pipes[pipe_cnt].pipe.src.viewport_width; /*vp_width/hratio*/ + pipes[pipe_cnt].pipe.dest.recout_height = pipes[pipe_cnt].pipe.src.viewport_height; /*vp_height/vratio*/ + pipes[pipe_cnt].pipe.dest.full_recout_width = pipes[pipe_cnt].pipe.dest.recout_width; /*when is_hsplit != 1*/ + pipes[pipe_cnt].pipe.dest.full_recout_height = pipes[pipe_cnt].pipe.dest.recout_height; /*when is_hsplit != 1*/ + pipes[pipe_cnt].pipe.scale_ratio_depth.lb_depth = dm_lb_16; + pipes[pipe_cnt].pipe.scale_ratio_depth.hscl_ratio = 1.0; + pipes[pipe_cnt].pipe.scale_ratio_depth.vscl_ratio = 1.0; + pipes[pipe_cnt].pipe.scale_ratio_depth.scl_enable = 0; /*Lb only or Full scl*/ + pipes[pipe_cnt].pipe.scale_taps.htaps = 1; + pipes[pipe_cnt].pipe.scale_taps.vtaps = 1; + pipes[pipe_cnt].pipe.src.is_hsplit = 0; + pipes[pipe_cnt].pipe.dest.odm_combine = 0; + pipes[pipe_cnt].pipe.dest.vtotal_min = timing->v_total; + pipes[pipe_cnt].pipe.dest.vtotal_max = timing->v_total; + } else { + struct dc_plane_state *pln = res_ctx->pipe_ctx[i].plane_state; + struct scaler_data *scl = &res_ctx->pipe_ctx[i].plane_res.scl_data; + + pipes[pipe_cnt].pipe.src.immediate_flip = pln->flip_immediate; + pipes[pipe_cnt].pipe.src.is_hsplit = (res_ctx->pipe_ctx[i].bottom_pipe + && res_ctx->pipe_ctx[i].bottom_pipe->plane_state == pln) + || (res_ctx->pipe_ctx[i].top_pipe + && res_ctx->pipe_ctx[i].top_pipe->plane_state == pln); + pipes[pipe_cnt].pipe.dest.odm_combine = (res_ctx->pipe_ctx[i].bottom_pipe + && res_ctx->pipe_ctx[i].bottom_pipe->plane_state == pln + && res_ctx->pipe_ctx[i].bottom_pipe->stream_res.opp + != res_ctx->pipe_ctx[i].stream_res.opp) + || (res_ctx->pipe_ctx[i].top_pipe + && res_ctx->pipe_ctx[i].top_pipe->plane_state == pln + && res_ctx->pipe_ctx[i].top_pipe->stream_res.opp + != res_ctx->pipe_ctx[i].stream_res.opp); + pipes[pipe_cnt].pipe.src.source_scan = pln->rotation == ROTATION_ANGLE_90 + || pln->rotation == ROTATION_ANGLE_270 ? dm_vert : dm_horz; + pipes[pipe_cnt].pipe.src.viewport_y_y = scl->viewport.y; + pipes[pipe_cnt].pipe.src.viewport_y_c = scl->viewport_c.y; + pipes[pipe_cnt].pipe.src.viewport_width = scl->viewport.width; + pipes[pipe_cnt].pipe.src.viewport_width_c = scl->viewport_c.width; + pipes[pipe_cnt].pipe.src.viewport_height = scl->viewport.height; + pipes[pipe_cnt].pipe.src.viewport_height_c = scl->viewport_c.height; + if (pln->format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) { + pipes[pipe_cnt].pipe.src.data_pitch = pln->plane_size.video.luma_pitch; + pipes[pipe_cnt].pipe.src.data_pitch_c = pln->plane_size.video.chroma_pitch; + pipes[pipe_cnt].pipe.src.meta_pitch = pln->dcc.video.meta_pitch_l; + pipes[pipe_cnt].pipe.src.meta_pitch_c = pln->dcc.video.meta_pitch_c; + } else { + pipes[pipe_cnt].pipe.src.data_pitch = pln->plane_size.grph.surface_pitch; + pipes[pipe_cnt].pipe.src.meta_pitch = pln->dcc.grph.meta_pitch; + } + pipes[pipe_cnt].pipe.src.dcc = pln->dcc.enable; + pipes[pipe_cnt].pipe.dest.recout_width = scl->recout.width; + pipes[pipe_cnt].pipe.dest.recout_height = scl->recout.height; + pipes[pipe_cnt].pipe.dest.full_recout_width = scl->recout.width; + pipes[pipe_cnt].pipe.dest.full_recout_height = scl->recout.height; + if (res_ctx->pipe_ctx[i].bottom_pipe && res_ctx->pipe_ctx[i].bottom_pipe->plane_state == pln) { + pipes[pipe_cnt].pipe.dest.full_recout_width += + res_ctx->pipe_ctx[i].bottom_pipe->plane_res.scl_data.recout.width; + pipes[pipe_cnt].pipe.dest.full_recout_height += + res_ctx->pipe_ctx[i].bottom_pipe->plane_res.scl_data.recout.height; + } else if (res_ctx->pipe_ctx[i].top_pipe && res_ctx->pipe_ctx[i].top_pipe->plane_state == pln) { + pipes[pipe_cnt].pipe.dest.full_recout_width += + res_ctx->pipe_ctx[i].top_pipe->plane_res.scl_data.recout.width; + pipes[pipe_cnt].pipe.dest.full_recout_height += + res_ctx->pipe_ctx[i].top_pipe->plane_res.scl_data.recout.height; + } + + pipes[pipe_cnt].pipe.scale_ratio_depth.lb_depth = dm_lb_16; + pipes[pipe_cnt].pipe.scale_ratio_depth.hscl_ratio = (double) scl->ratios.horz.value / (1ULL<<32); + pipes[pipe_cnt].pipe.scale_ratio_depth.hscl_ratio_c = (double) scl->ratios.horz_c.value / (1ULL<<32); + pipes[pipe_cnt].pipe.scale_ratio_depth.vscl_ratio = (double) scl->ratios.vert.value / (1ULL<<32); + pipes[pipe_cnt].pipe.scale_ratio_depth.vscl_ratio_c = (double) scl->ratios.vert_c.value / (1ULL<<32); + pipes[pipe_cnt].pipe.scale_ratio_depth.scl_enable = + scl->ratios.vert.value != dc_fixpt_one.value + || scl->ratios.horz.value != dc_fixpt_one.value + || scl->ratios.vert_c.value != dc_fixpt_one.value + || scl->ratios.horz_c.value != dc_fixpt_one.value /*Lb only or Full scl*/ + || dc->debug.always_scale; /*support always scale*/ + pipes[pipe_cnt].pipe.scale_taps.htaps = scl->taps.h_taps; + pipes[pipe_cnt].pipe.scale_taps.htaps_c = scl->taps.h_taps_c; + pipes[pipe_cnt].pipe.scale_taps.vtaps = scl->taps.v_taps; + pipes[pipe_cnt].pipe.scale_taps.vtaps_c = scl->taps.v_taps_c; + + pipes[pipe_cnt].pipe.src.macro_tile_size = + swizzle_mode_to_macro_tile_size(pln->tiling_info.gfx9.swizzle); + swizzle_to_dml_params(pln->tiling_info.gfx9.swizzle, + &pipes[pipe_cnt].pipe.src.sw_mode); + + switch (pln->format) { + case SURFACE_PIXEL_FORMAT_VIDEO_420_YCbCr: + case SURFACE_PIXEL_FORMAT_VIDEO_420_YCrCb: + pipes[pipe_cnt].pipe.src.source_format = dm_420_8; + break; + case SURFACE_PIXEL_FORMAT_VIDEO_420_10bpc_YCbCr: + case SURFACE_PIXEL_FORMAT_VIDEO_420_10bpc_YCrCb: + pipes[pipe_cnt].pipe.src.source_format = dm_420_10; + break; + case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616: + case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616F: + case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616F: + pipes[pipe_cnt].pipe.src.source_format = dm_444_64; + break; + case SURFACE_PIXEL_FORMAT_GRPH_ARGB1555: + case SURFACE_PIXEL_FORMAT_GRPH_RGB565: + pipes[pipe_cnt].pipe.src.source_format = dm_444_16; + break; + case SURFACE_PIXEL_FORMAT_GRPH_PALETA_256_COLORS: + pipes[pipe_cnt].pipe.src.source_format = dm_444_8; + break; + default: + pipes[pipe_cnt].pipe.src.source_format = dm_444_32; + break; + } + } + + pipe_cnt++; + } + + /* populate writeback information */ + dc->res_pool->funcs->populate_dml_writeback_from_context(dc, res_ctx, pipes); + + return pipe_cnt; +} + +unsigned int dcn20_calc_max_scaled_time( + unsigned int time_per_pixel, + enum mmhubbub_wbif_mode mode, + unsigned int urgent_watermark) +{ + unsigned int time_per_byte = 0; + unsigned int total_y_free_entry = 0x200; /* two memory piece for luma */ + unsigned int total_c_free_entry = 0x140; /* two memory piece for chroma */ + unsigned int small_free_entry, max_free_entry; + unsigned int buf_lh_capability; + unsigned int max_scaled_time; + + if (mode == PACKED_444) /* packed mode */ + time_per_byte = time_per_pixel/4; + else if (mode == PLANAR_420_8BPC) + time_per_byte = time_per_pixel; + else if (mode == PLANAR_420_10BPC) /* p010 */ + time_per_byte = time_per_pixel * 819/1024; + + if (time_per_byte == 0) + time_per_byte = 1; + + small_free_entry = (total_y_free_entry > total_c_free_entry) ? total_c_free_entry : total_y_free_entry; + max_free_entry = (mode == PACKED_444) ? total_y_free_entry + total_c_free_entry : small_free_entry; + buf_lh_capability = max_free_entry*time_per_byte*32/16; /* there is 4bit fraction */ + max_scaled_time = buf_lh_capability - urgent_watermark; + return max_scaled_time; +} + +void dcn20_set_mcif_arb_params( + struct dc *dc, + struct dc_state *context, + display_e2e_pipe_params_st *pipes, + int pipe_cnt) +{ + enum mmhubbub_wbif_mode wbif_mode; + struct mcif_arb_params *wb_arb_params; + int i, j, k, dwb_pipe; + + /* Writeback MCIF_WB arbitration parameters */ + dwb_pipe = 0; + for (i = 0; i < dc->res_pool->pipe_count; i++) { + + if (!context->res_ctx.pipe_ctx[i].stream) + continue; + + for (j = 0; j < MAX_DWB_PIPES; j++) { + if (context->res_ctx.pipe_ctx[i].stream->writeback_info[j].wb_enabled == false) + continue; + + //wb_arb_params = &context->res_ctx.pipe_ctx[i].stream->writeback_info[j].mcif_arb_params; + wb_arb_params = &context->bw_ctx.bw.dcn.bw_writeback.mcif_wb_arb[dwb_pipe]; + + if (context->res_ctx.pipe_ctx[i].stream->writeback_info[j].dwb_params.out_format == dwb_scaler_mode_yuv420) { + if (context->res_ctx.pipe_ctx[i].stream->writeback_info[j].dwb_params.output_depth == DWB_OUTPUT_PIXEL_DEPTH_8BPC) + wbif_mode = PLANAR_420_8BPC; + else + wbif_mode = PLANAR_420_10BPC; + } else + wbif_mode = PACKED_444; + + for (k = 0; k < sizeof(wb_arb_params->cli_watermark)/sizeof(wb_arb_params->cli_watermark[0]); k++) { + wb_arb_params->cli_watermark[k] = get_wm_writeback_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + wb_arb_params->pstate_watermark[k] = get_wm_writeback_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + } + wb_arb_params->time_per_pixel = 16.0 / context->res_ctx.pipe_ctx[i].stream->phy_pix_clk; /* 4 bit fraction, ms */ + wb_arb_params->slice_lines = 32; + wb_arb_params->arbitration_slice = 2; + wb_arb_params->max_scaled_time = dcn20_calc_max_scaled_time(wb_arb_params->time_per_pixel, + wbif_mode, + wb_arb_params->cli_watermark[0]); /* assume 4 watermark sets have the same value */ + + dwb_pipe++; + + if (dwb_pipe >= MAX_DWB_PIPES) + return; + } + if (dwb_pipe >= MAX_DWB_PIPES) + return; + } +} + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT +static bool dcn20_validate_dsc(struct dc *dc, struct dc_state *new_ctx) +{ + int i; + + /* Validate DSC config, dsc count validation is already done */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe_ctx = &new_ctx->res_ctx.pipe_ctx[i]; + struct dc_stream_state *stream = pipe_ctx->stream; + struct dsc_config dsc_cfg; + + /* Only need to validate top pipe */ + if (pipe_ctx->top_pipe || !stream || !stream->timing.flags.DSC) + continue; + + dsc_cfg.pic_width = stream->timing.h_addressable + stream->timing.h_border_left + + stream->timing.h_border_right; + dsc_cfg.pic_height = stream->timing.v_addressable + stream->timing.v_border_top + + stream->timing.v_border_bottom; + if (dc_res_get_odm_bottom_pipe(pipe_ctx)) + dsc_cfg.pic_width /= 2; + dsc_cfg.pixel_encoding = stream->timing.pixel_encoding; + dsc_cfg.color_depth = stream->timing.display_color_depth; + dsc_cfg.dc_dsc_cfg = stream->timing.dsc_cfg; + + if (!pipe_ctx->stream_res.dsc->funcs->dsc_validate_stream(pipe_ctx->stream_res.dsc, &dsc_cfg)) + return false; + } + return true; +} +#endif + +bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context, + bool fast_validate) +{ + bool out = false; + + BW_VAL_TRACE_SETUP(); + + int pipe_cnt, i, pipe_idx, vlevel, vlevel_unsplit; + int pipe_split_from[MAX_PIPES]; + bool odm_capable = context->bw_ctx.dml.ip.odm_capable; + bool force_split = false; +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + bool failed_non_odm_dsc = false; +#endif + int split_threshold = dc->res_pool->pipe_count / 2; + bool avoid_split = dc->debug.pipe_split_policy != MPC_SPLIT_DYNAMIC; + display_e2e_pipe_params_st *pipes = kzalloc(dc->res_pool->pipe_count * sizeof(display_e2e_pipe_params_st), GFP_KERNEL); + DC_LOGGER_INIT(dc->ctx->logger); + + BW_VAL_TRACE_COUNT(); + + ASSERT(pipes); + if (!pipes) + return false; + + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; + struct pipe_ctx *hsplit_pipe = pipe->bottom_pipe; + + if (!hsplit_pipe || hsplit_pipe->plane_state != pipe->plane_state) + continue; + + /* merge previously split pipe since mode support needs to make the decision */ + pipe->bottom_pipe = hsplit_pipe->bottom_pipe; + if (hsplit_pipe->bottom_pipe) + hsplit_pipe->bottom_pipe->top_pipe = pipe; + hsplit_pipe->plane_state = NULL; + hsplit_pipe->stream = NULL; + hsplit_pipe->top_pipe = NULL; + hsplit_pipe->bottom_pipe = NULL; +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + if (hsplit_pipe->stream_res.dsc && hsplit_pipe->stream_res.dsc != pipe->stream_res.dsc) + release_dsc(&context->res_ctx, dc->res_pool, &hsplit_pipe->stream_res.dsc); +#endif + /* Clear plane_res and stream_res */ + memset(&hsplit_pipe->plane_res, 0, sizeof(hsplit_pipe->plane_res)); + memset(&hsplit_pipe->stream_res, 0, sizeof(hsplit_pipe->stream_res)); + if (pipe->plane_state) + resource_build_scaling_params(pipe); + } + + if (dc->res_pool->funcs->populate_dml_pipes) + pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, + &context->res_ctx, pipes); + else + pipe_cnt = dcn20_populate_dml_pipes_from_context(dc, + &context->res_ctx, pipes); + + if (!pipe_cnt) { + BW_VAL_TRACE_SKIP(pass); + out = true; + goto validate_out; + } + + context->bw_ctx.dml.ip.odm_capable = 0; + + vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt); + + context->bw_ctx.dml.ip.odm_capable = odm_capable; + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + /* 1 dsc per stream dsc validation */ + if (vlevel <= context->bw_ctx.dml.soc.num_states) + if (!dcn20_validate_dsc(dc, context)) { + failed_non_odm_dsc = true; + vlevel = context->bw_ctx.dml.soc.num_states + 1; + } +#endif + + if (vlevel > context->bw_ctx.dml.soc.num_states && odm_capable) + vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt); + + if (vlevel > context->bw_ctx.dml.soc.num_states) + goto validate_fail; + + if ((context->stream_count > split_threshold && dc->current_state->stream_count <= split_threshold) + || (context->stream_count <= split_threshold && dc->current_state->stream_count > split_threshold)) + context->commit_hints.full_update_needed = true; + + /*initialize pipe_just_split_from to invalid idx*/ + for (i = 0; i < MAX_PIPES; i++) + pipe_split_from[i] = -1; + + /* Single display only conditionals get set here */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; + bool exit_loop = false; + + if (!pipe->stream || pipe->top_pipe) + continue; + + if (dc->debug.force_single_disp_pipe_split) { + if (!force_split) + force_split = true; + else { + force_split = false; + exit_loop = true; + } + } + if (dc->debug.pipe_split_policy == MPC_SPLIT_AVOID_MULT_DISP) { + if (avoid_split) + avoid_split = false; + else { + avoid_split = true; + exit_loop = true; + } + } + if (exit_loop) + break; + } + + if (context->stream_count > split_threshold) + avoid_split = true; + + vlevel_unsplit = vlevel; + for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { + if (!context->res_ctx.pipe_ctx[i].stream) + continue; + for (; vlevel_unsplit <= context->bw_ctx.dml.soc.num_states; vlevel_unsplit++) + if (context->bw_ctx.dml.vba.NoOfDPP[vlevel_unsplit][0][pipe_idx] == 1) + break; + pipe_idx++; + } + + for (i = 0, pipe_idx = -1; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; + struct pipe_ctx *hsplit_pipe = pipe->bottom_pipe; + bool need_split = true; + bool need_split3d; + + if (!pipe->stream || pipe_split_from[i] >= 0) + continue; + + pipe_idx++; + + if (dc->debug.force_odm_combine & (1 << pipe->stream_res.tg->inst)) { + force_split = true; + context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx] = true; + context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_idx] = true; + } + if (force_split && context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] == 1) + context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] /= 2; + if (dc->config.forced_clocks == true) { + context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] = + context->bw_ctx.dml.soc.clock_limits[0].dppclk_mhz; + } + if (!pipe->top_pipe && !pipe->plane_state && context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx]) { + hsplit_pipe = find_idle_secondary_pipe(&context->res_ctx, dc->res_pool, pipe); + ASSERT(hsplit_pipe); + if (!dcn20_split_stream_for_combine( + &context->res_ctx, dc->res_pool, + pipe, hsplit_pipe, + true)) + goto validate_fail; + pipe_split_from[hsplit_pipe->pipe_idx] = pipe_idx; + dcn20_build_mapped_resource(dc, context, pipe->stream); + } + + if (!pipe->plane_state) + continue; + /* Skip 2nd half of already split pipe */ + if (pipe->top_pipe && pipe->plane_state == pipe->top_pipe->plane_state) + continue; + + need_split3d = ((pipe->stream->view_format == + VIEW_3D_FORMAT_SIDE_BY_SIDE || + pipe->stream->view_format == + VIEW_3D_FORMAT_TOP_AND_BOTTOM) && + (pipe->stream->timing.timing_3d_format == + TIMING_3D_FORMAT_TOP_AND_BOTTOM || + pipe->stream->timing.timing_3d_format == + TIMING_3D_FORMAT_SIDE_BY_SIDE)); + + if (avoid_split && vlevel_unsplit <= context->bw_ctx.dml.soc.num_states && !force_split && !need_split3d) { + need_split = false; + vlevel = vlevel_unsplit; + context->bw_ctx.dml.vba.maxMpcComb = 0; + } else + need_split = context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] == 2; + + /* We do not support mpo + odm at the moment */ + if (hsplit_pipe && hsplit_pipe->plane_state != pipe->plane_state + && context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx]) + goto validate_fail; + + if (need_split3d || need_split || force_split) { + if (!hsplit_pipe || hsplit_pipe->plane_state != pipe->plane_state) { + /* pipe not split previously needs split */ + hsplit_pipe = find_idle_secondary_pipe(&context->res_ctx, dc->res_pool, pipe); + ASSERT(hsplit_pipe || force_split); + if (!hsplit_pipe) + continue; + + if (!dcn20_split_stream_for_combine( + &context->res_ctx, dc->res_pool, + pipe, hsplit_pipe, + context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx])) + goto validate_fail; + pipe_split_from[hsplit_pipe->pipe_idx] = pipe_idx; + } + } else if (hsplit_pipe && hsplit_pipe->plane_state == pipe->plane_state) { + /* merge should already have been done */ + ASSERT(0); + } + } +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + /* Actual dsc count per stream dsc validation*/ + if (failed_non_odm_dsc && !dcn20_validate_dsc(dc, context)) { + context->bw_ctx.dml.vba.ValidationStatus[context->bw_ctx.dml.vba.soc.num_states] = + DML_FAIL_DSC_VALIDATION_FAILURE; + goto validate_fail; + } +#endif + + BW_VAL_TRACE_END_VOLTAGE_LEVEL(); + + if (fast_validate) { + BW_VAL_TRACE_SKIP(fast); + out = true; + goto validate_out; + } + + for (i = 0, pipe_idx = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) { + if (!context->res_ctx.pipe_ctx[i].stream) + continue; + + pipes[pipe_cnt].clks_cfg.refclk_mhz = dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000.0; + pipes[pipe_cnt].clks_cfg.dispclk_mhz = context->bw_ctx.dml.vba.RequiredDISPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; + + if (pipe_split_from[i] < 0) { + pipes[pipe_cnt].clks_cfg.dppclk_mhz = + context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx]; + if (context->bw_ctx.dml.vba.BlendingAndTiming[pipe_idx] == pipe_idx) + pipes[pipe_cnt].pipe.dest.odm_combine = + context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_idx]; + else + pipes[pipe_cnt].pipe.dest.odm_combine = 0; + pipe_idx++; + } else { + pipes[pipe_cnt].clks_cfg.dppclk_mhz = + context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_split_from[i]]; + if (context->bw_ctx.dml.vba.BlendingAndTiming[pipe_split_from[i]] == pipe_split_from[i]) + pipes[pipe_cnt].pipe.dest.odm_combine = + context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_split_from[i]]; + else + pipes[pipe_cnt].pipe.dest.odm_combine = 0; + } + if (dc->config.forced_clocks) { + pipes[pipe_cnt].clks_cfg.dispclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dispclk_mhz; + pipes[pipe_cnt].clks_cfg.dppclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dppclk_mhz; + } + pipe_cnt++; + } + + if (pipe_cnt != pipe_idx) { + if (dc->res_pool->funcs->populate_dml_pipes) + pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, + &context->res_ctx, pipes); + else + pipe_cnt = dcn20_populate_dml_pipes_from_context(dc, + &context->res_ctx, pipes); + } + + pipes[0].clks_cfg.voltage = vlevel; + pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].dcfclk_mhz; + pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; + + /* only pipe 0 is read for voltage and dcf/soc clocks */ + if (vlevel < 1) { + pipes[0].clks_cfg.voltage = 1; + pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[1].dcfclk_mhz; + pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[1].socclk_mhz; + } + context->bw_ctx.bw.dcn.watermarks.b.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + + if (vlevel < 2) { + pipes[0].clks_cfg.voltage = 2; + pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[2].dcfclk_mhz; + pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[2].socclk_mhz; + } + context->bw_ctx.bw.dcn.watermarks.c.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + + if (vlevel < 3) { + pipes[0].clks_cfg.voltage = 3; + pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[2].dcfclk_mhz; + pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[2].socclk_mhz; + } + context->bw_ctx.bw.dcn.watermarks.d.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + + pipes[0].clks_cfg.voltage = vlevel; + pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].dcfclk_mhz; + pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; + context->bw_ctx.bw.dcn.watermarks.a.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + /* Writeback MCIF_WB arbitration parameters */ + dc->res_pool->funcs->set_mcif_arb_params(dc, context, pipes, pipe_cnt); + + context->bw_ctx.bw.dcn.clk.dispclk_khz = context->bw_ctx.dml.vba.DISPCLK * 1000; + context->bw_ctx.bw.dcn.clk.dcfclk_khz = context->bw_ctx.dml.vba.DCFCLK * 1000; + context->bw_ctx.bw.dcn.clk.socclk_khz = context->bw_ctx.dml.vba.SOCCLK * 1000; + context->bw_ctx.bw.dcn.clk.dramclk_khz = context->bw_ctx.dml.vba.DRAMSpeed * 1000 / 16; + context->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz = context->bw_ctx.dml.vba.DCFCLKDeepSleep * 1000; + context->bw_ctx.bw.dcn.clk.fclk_khz = context->bw_ctx.dml.vba.FabricClock * 1000; + context->bw_ctx.bw.dcn.clk.p_state_change_support = + context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb] + != dm_dram_clock_change_unsupported; + context->bw_ctx.bw.dcn.clk.dppclk_khz = 0; + + BW_VAL_TRACE_END_WATERMARKS(); + + for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { + if (!context->res_ctx.pipe_ctx[i].stream) + continue; + pipes[pipe_idx].pipe.dest.vstartup_start = context->bw_ctx.dml.vba.VStartup[pipe_idx]; + pipes[pipe_idx].pipe.dest.vupdate_offset = context->bw_ctx.dml.vba.VUpdateOffsetPix[pipe_idx]; + pipes[pipe_idx].pipe.dest.vupdate_width = context->bw_ctx.dml.vba.VUpdateWidthPix[pipe_idx]; + pipes[pipe_idx].pipe.dest.vready_offset = context->bw_ctx.dml.vba.VReadyOffsetPix[pipe_idx]; + if (context->bw_ctx.bw.dcn.clk.dppclk_khz < pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000) + context->bw_ctx.bw.dcn.clk.dppclk_khz = pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000; + context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz = + pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000; +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + context->res_ctx.pipe_ctx[i].stream_res.dscclk_khz = + context->bw_ctx.dml.vba.DSCCLK_calculated[pipe_idx] * 1000; +#endif + context->res_ctx.pipe_ctx[i].pipe_dlg_param = pipes[pipe_idx].pipe.dest; + pipe_idx++; + } + + for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { + bool cstate_en = context->bw_ctx.dml.vba.PrefetchMode[vlevel][context->bw_ctx.dml.vba.maxMpcComb] != 2; + + if (!context->res_ctx.pipe_ctx[i].stream) + continue; + + context->bw_ctx.dml.funcs.rq_dlg_get_dlg_reg(&context->bw_ctx.dml, + &context->res_ctx.pipe_ctx[i].dlg_regs, + &context->res_ctx.pipe_ctx[i].ttu_regs, + pipes, + pipe_cnt, + pipe_idx, + cstate_en, + context->bw_ctx.bw.dcn.clk.p_state_change_support, + false, false, false); + + context->bw_ctx.dml.funcs.rq_dlg_get_rq_reg(&context->bw_ctx.dml, + &context->res_ctx.pipe_ctx[i].rq_regs, + pipes[pipe_idx].pipe); + pipe_idx++; + } + + out = true; + goto validate_out; + +validate_fail: + DC_LOG_WARNING("Mode Validation Warning: %s failed validation.\n", + dml_get_status_message(context->bw_ctx.dml.vba.ValidationStatus[context->bw_ctx.dml.vba.soc.num_states])); + + BW_VAL_TRACE_SKIP(fail); + out = false; + +validate_out: + kfree(pipes); + + BW_VAL_TRACE_FINISH(); + + return out; +} + +struct pipe_ctx *dcn20_acquire_idle_pipe_for_layer( + struct dc_state *state, + const struct resource_pool *pool, + struct dc_stream_state *stream) +{ + struct resource_context *res_ctx = &state->res_ctx; + struct pipe_ctx *head_pipe = resource_get_head_pipe_for_stream(res_ctx, stream); + struct pipe_ctx *idle_pipe = find_idle_secondary_pipe(res_ctx, pool, head_pipe); + + if (!head_pipe) + ASSERT(0); + + if (!idle_pipe) + return false; + + idle_pipe->stream = head_pipe->stream; + idle_pipe->stream_res.tg = head_pipe->stream_res.tg; + idle_pipe->stream_res.opp = head_pipe->stream_res.opp; + + idle_pipe->plane_res.hubp = pool->hubps[idle_pipe->pipe_idx]; + idle_pipe->plane_res.ipp = pool->ipps[idle_pipe->pipe_idx]; + idle_pipe->plane_res.dpp = pool->dpps[idle_pipe->pipe_idx]; + idle_pipe->plane_res.mpcc_inst = pool->dpps[idle_pipe->pipe_idx]->inst; + + return idle_pipe; +} + +bool dcn20_get_dcc_compression_cap(const struct dc *dc, + const struct dc_dcc_surface_param *input, + struct dc_surface_dcc_cap *output) +{ + return dc->res_pool->hubbub->funcs->get_dcc_compression_cap( + dc->res_pool->hubbub, + input, + output); +} + +static void dcn20_destroy_resource_pool(struct resource_pool **pool) +{ + struct dcn20_resource_pool *dcn20_pool = TO_DCN20_RES_POOL(*pool); + + destruct(dcn20_pool); + kfree(dcn20_pool); + *pool = NULL; +} + + +static struct dc_cap_funcs cap_funcs = { + .get_dcc_compression_cap = dcn20_get_dcc_compression_cap +}; + + +enum dc_status dcn20_get_default_swizzle_mode(struct dc_plane_state *plane_state) +{ + enum dc_status result = DC_OK; + + enum surface_pixel_format surf_pix_format = plane_state->format; + unsigned int bpp = resource_pixel_format_to_bpp(surf_pix_format); + + enum swizzle_mode_values swizzle = DC_SW_LINEAR; + + if (bpp == 64) + swizzle = DC_SW_64KB_D; + else + swizzle = DC_SW_64KB_S; + + plane_state->tiling_info.gfx9.swizzle = swizzle; + return result; +} + +static struct resource_funcs dcn20_res_pool_funcs = { + .destroy = dcn20_destroy_resource_pool, + .link_enc_create = dcn20_link_encoder_create, + .validate_bandwidth = dcn20_validate_bandwidth, + .acquire_idle_pipe_for_layer = dcn20_acquire_idle_pipe_for_layer, + .add_stream_to_ctx = dcn20_add_stream_to_ctx, + .remove_stream_from_ctx = dcn20_remove_stream_from_ctx, + .populate_dml_writeback_from_context = dcn20_populate_dml_writeback_from_context, + .get_default_swizzle_mode = dcn20_get_default_swizzle_mode, + .set_mcif_arb_params = dcn20_set_mcif_arb_params, + .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link +}; + +bool dcn20_dwbc_create(struct dc_context *ctx, struct resource_pool *pool) +{ + int i; + uint32_t pipe_count = pool->res_cap->num_dwb; + + ASSERT(pipe_count > 0); + + for (i = 0; i < pipe_count; i++) { + struct dcn20_dwbc *dwbc20 = kzalloc(sizeof(struct dcn20_dwbc), + GFP_KERNEL); + + if (!dwbc20) { + dm_error("DC: failed to create dwbc20!\n"); + return false; + } + dcn20_dwbc_construct(dwbc20, ctx, + &dwbc20_regs[i], + &dwbc20_shift, + &dwbc20_mask, + i); + pool->dwbc[i] = &dwbc20->base; + } + return true; +} + +bool dcn20_mmhubbub_create(struct dc_context *ctx, struct resource_pool *pool) +{ + int i; + uint32_t pipe_count = pool->res_cap->num_dwb; + + ASSERT(pipe_count > 0); + + for (i = 0; i < pipe_count; i++) { + struct dcn20_mmhubbub *mcif_wb20 = kzalloc(sizeof(struct dcn20_mmhubbub), + GFP_KERNEL); + + if (!mcif_wb20) { + dm_error("DC: failed to create mcif_wb20!\n"); + return false; + } + + dcn20_mmhubbub_construct(mcif_wb20, ctx, + &mcif_wb20_regs[i], + &mcif_wb20_shift, + &mcif_wb20_mask, + i); + + pool->mcif_wb[i] = &mcif_wb20->base; + } + return true; +} + +struct pp_smu_funcs *dcn20_pp_smu_create(struct dc_context *ctx) +{ + struct pp_smu_funcs *pp_smu = kzalloc(sizeof(*pp_smu), GFP_KERNEL); + + if (!pp_smu) + return pp_smu; + + dm_pp_get_funcs(ctx, pp_smu); + + if (pp_smu->ctx.ver != PP_SMU_VER_NV) + pp_smu = memset(pp_smu, 0, sizeof(struct pp_smu_funcs)); + + return pp_smu; +} + +void dcn20_pp_smu_destroy(struct pp_smu_funcs **pp_smu) +{ + if (pp_smu && *pp_smu) { + kfree(*pp_smu); + *pp_smu = NULL; + } +} + +static void cap_soc_clocks( + struct _vcs_dpi_soc_bounding_box_st *bb, + struct pp_smu_nv_clock_table max_clocks) +{ + int i; + + // First pass - cap all clocks higher than the reported max + for (i = 0; i < bb->num_states; i++) { + if ((bb->clock_limits[i].dcfclk_mhz > (max_clocks.dcfClockInKhz / 1000)) + && max_clocks.dcfClockInKhz != 0) + bb->clock_limits[i].dcfclk_mhz = (max_clocks.dcfClockInKhz / 1000); + + if ((bb->clock_limits[i].dram_speed_mts > (max_clocks.uClockInKhz / 1000) * 16) + && max_clocks.uClockInKhz != 0) + bb->clock_limits[i].dram_speed_mts = (max_clocks.uClockInKhz / 1000) * 16; + + if ((bb->clock_limits[i].fabricclk_mhz > (max_clocks.fabricClockInKhz / 1000)) + && max_clocks.fabricClockInKhz != 0) + bb->clock_limits[i].fabricclk_mhz = (max_clocks.fabricClockInKhz / 1000); + + if ((bb->clock_limits[i].dispclk_mhz > (max_clocks.displayClockInKhz / 1000)) + && max_clocks.displayClockInKhz != 0) + bb->clock_limits[i].dispclk_mhz = (max_clocks.displayClockInKhz / 1000); + + if ((bb->clock_limits[i].dppclk_mhz > (max_clocks.dppClockInKhz / 1000)) + && max_clocks.dppClockInKhz != 0) + bb->clock_limits[i].dppclk_mhz = (max_clocks.dppClockInKhz / 1000); + + if ((bb->clock_limits[i].phyclk_mhz > (max_clocks.phyClockInKhz / 1000)) + && max_clocks.phyClockInKhz != 0) + bb->clock_limits[i].phyclk_mhz = (max_clocks.phyClockInKhz / 1000); + + if ((bb->clock_limits[i].socclk_mhz > (max_clocks.socClockInKhz / 1000)) + && max_clocks.socClockInKhz != 0) + bb->clock_limits[i].socclk_mhz = (max_clocks.socClockInKhz / 1000); + + if ((bb->clock_limits[i].dscclk_mhz > (max_clocks.dscClockInKhz / 1000)) + && max_clocks.dscClockInKhz != 0) + bb->clock_limits[i].dscclk_mhz = (max_clocks.dscClockInKhz / 1000); + } + + // Second pass - remove all duplicate clock states + for (i = bb->num_states - 1; i > 1; i--) { + bool duplicate = true; + + if (bb->clock_limits[i-1].dcfclk_mhz != bb->clock_limits[i].dcfclk_mhz) + duplicate = false; + if (bb->clock_limits[i-1].dispclk_mhz != bb->clock_limits[i].dispclk_mhz) + duplicate = false; + if (bb->clock_limits[i-1].dppclk_mhz != bb->clock_limits[i].dppclk_mhz) + duplicate = false; + if (bb->clock_limits[i-1].dram_speed_mts != bb->clock_limits[i].dram_speed_mts) + duplicate = false; + if (bb->clock_limits[i-1].dscclk_mhz != bb->clock_limits[i].dscclk_mhz) + duplicate = false; + if (bb->clock_limits[i-1].fabricclk_mhz != bb->clock_limits[i].fabricclk_mhz) + duplicate = false; + if (bb->clock_limits[i-1].phyclk_mhz != bb->clock_limits[i].phyclk_mhz) + duplicate = false; + if (bb->clock_limits[i-1].socclk_mhz != bb->clock_limits[i].socclk_mhz) + duplicate = false; + + if (duplicate) + bb->num_states--; + } +} + +static void update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb, + struct pp_smu_nv_clock_table *max_clocks, unsigned int *uclk_states, unsigned int num_states) +{ + struct _vcs_dpi_voltage_scaling_st calculated_states[MAX_CLOCK_LIMIT_STATES] = {0}; + int i; + int num_calculated_states = 0; + int min_dcfclk = 0; + + if (num_states == 0) + return; + + if (dc->bb_overrides.min_dcfclk_mhz > 0) + min_dcfclk = dc->bb_overrides.min_dcfclk_mhz; + + for (i = 0; i < num_states; i++) { + int min_fclk_required_by_uclk; + calculated_states[i].state = i; + calculated_states[i].dram_speed_mts = uclk_states[i] * 16 / 1000; + + // FCLK:UCLK ratio is 1.08 + min_fclk_required_by_uclk = ((unsigned long long)uclk_states[i]) * 1080 / 1000000; + + calculated_states[i].fabricclk_mhz = (min_fclk_required_by_uclk < min_dcfclk) ? + min_dcfclk : min_fclk_required_by_uclk; + + calculated_states[i].socclk_mhz = (calculated_states[i].fabricclk_mhz > max_clocks->socClockInKhz / 1000) ? + max_clocks->socClockInKhz / 1000 : calculated_states[i].fabricclk_mhz; + + calculated_states[i].dcfclk_mhz = (calculated_states[i].fabricclk_mhz > max_clocks->dcfClockInKhz / 1000) ? + max_clocks->dcfClockInKhz / 1000 : calculated_states[i].fabricclk_mhz; + + calculated_states[i].dispclk_mhz = max_clocks->displayClockInKhz / 1000; + calculated_states[i].dppclk_mhz = max_clocks->displayClockInKhz / 1000; + calculated_states[i].dscclk_mhz = max_clocks->displayClockInKhz / (1000 * 3); + + calculated_states[i].phyclk_mhz = max_clocks->phyClockInKhz / 1000; + + num_calculated_states++; + } + + memcpy(bb->clock_limits, calculated_states, sizeof(bb->clock_limits)); + bb->num_states = num_calculated_states; + + // Duplicate the last state, DML always an extra state identical to max state to work + memcpy(&bb->clock_limits[num_calculated_states], &bb->clock_limits[num_calculated_states - 1], sizeof(struct _vcs_dpi_voltage_scaling_st)); + bb->clock_limits[num_calculated_states].state = bb->num_states; +} + +static void patch_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb) +{ + kernel_fpu_begin(); + if ((int)(bb->sr_exit_time_us * 1000) != dc->bb_overrides.sr_exit_time_ns + && dc->bb_overrides.sr_exit_time_ns) { + bb->sr_exit_time_us = dc->bb_overrides.sr_exit_time_ns / 1000.0; + } + + if ((int)(bb->sr_enter_plus_exit_time_us * 1000) + != dc->bb_overrides.sr_enter_plus_exit_time_ns + && 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; + } + + if ((int)(bb->urgent_latency_us * 1000) != dc->bb_overrides.urgent_latency_ns + && dc->bb_overrides.urgent_latency_ns) { + bb->urgent_latency_us = dc->bb_overrides.urgent_latency_ns / 1000.0; + } + + if ((int)(bb->dram_clock_change_latency_us * 1000) + != dc->bb_overrides.dram_clock_change_latency_ns + && dc->bb_overrides.dram_clock_change_latency_ns) { + bb->dram_clock_change_latency_us = + dc->bb_overrides.dram_clock_change_latency_ns / 1000.0; + } + kernel_fpu_end(); +} + +#define fixed16_to_double(x) (((double) x) / ((double) (1 << 16))) +#define fixed16_to_double_to_cpu(x) fixed16_to_double(le32_to_cpu(x)) + +static bool init_soc_bounding_box(struct dc *dc, + struct dcn20_resource_pool *pool) +{ + const struct gpu_info_soc_bounding_box_v1_0 *bb = dc->soc_bounding_box; + DC_LOGGER_INIT(dc->ctx->logger); + + if (!bb && !SOC_BOUNDING_BOX_VALID) { + DC_LOG_ERROR("%s: not valid soc bounding box/n", __func__); + return false; + } + + if (bb && !SOC_BOUNDING_BOX_VALID) { + int i; + + dcn2_0_soc.sr_exit_time_us = + fixed16_to_double_to_cpu(bb->sr_exit_time_us); + dcn2_0_soc.sr_enter_plus_exit_time_us = + fixed16_to_double_to_cpu(bb->sr_enter_plus_exit_time_us); + dcn2_0_soc.urgent_latency_us = + fixed16_to_double_to_cpu(bb->urgent_latency_us); + dcn2_0_soc.urgent_latency_pixel_data_only_us = + fixed16_to_double_to_cpu(bb->urgent_latency_pixel_data_only_us); + dcn2_0_soc.urgent_latency_pixel_mixed_with_vm_data_us = + fixed16_to_double_to_cpu(bb->urgent_latency_pixel_mixed_with_vm_data_us); + dcn2_0_soc.urgent_latency_vm_data_only_us = + fixed16_to_double_to_cpu(bb->urgent_latency_vm_data_only_us); + dcn2_0_soc.urgent_out_of_order_return_per_channel_pixel_only_bytes = + le32_to_cpu(bb->urgent_out_of_order_return_per_channel_pixel_only_bytes); + dcn2_0_soc.urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = + le32_to_cpu(bb->urgent_out_of_order_return_per_channel_pixel_and_vm_bytes); + dcn2_0_soc.urgent_out_of_order_return_per_channel_vm_only_bytes = + le32_to_cpu(bb->urgent_out_of_order_return_per_channel_vm_only_bytes); + dcn2_0_soc.pct_ideal_dram_sdp_bw_after_urgent_pixel_only = + fixed16_to_double_to_cpu(bb->pct_ideal_dram_sdp_bw_after_urgent_pixel_only); + dcn2_0_soc.pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = + fixed16_to_double_to_cpu(bb->pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm); + dcn2_0_soc.pct_ideal_dram_sdp_bw_after_urgent_vm_only = + fixed16_to_double_to_cpu(bb->pct_ideal_dram_sdp_bw_after_urgent_vm_only); + dcn2_0_soc.max_avg_sdp_bw_use_normal_percent = + fixed16_to_double_to_cpu(bb->max_avg_sdp_bw_use_normal_percent); + dcn2_0_soc.max_avg_dram_bw_use_normal_percent = + fixed16_to_double_to_cpu(bb->max_avg_dram_bw_use_normal_percent); + dcn2_0_soc.writeback_latency_us = + fixed16_to_double_to_cpu(bb->writeback_latency_us); + dcn2_0_soc.ideal_dram_bw_after_urgent_percent = + fixed16_to_double_to_cpu(bb->ideal_dram_bw_after_urgent_percent); + dcn2_0_soc.max_request_size_bytes = + le32_to_cpu(bb->max_request_size_bytes); + dcn2_0_soc.dram_channel_width_bytes = + le32_to_cpu(bb->dram_channel_width_bytes); + dcn2_0_soc.fabric_datapath_to_dcn_data_return_bytes = + le32_to_cpu(bb->fabric_datapath_to_dcn_data_return_bytes); + dcn2_0_soc.dcn_downspread_percent = + fixed16_to_double_to_cpu(bb->dcn_downspread_percent); + dcn2_0_soc.downspread_percent = + fixed16_to_double_to_cpu(bb->downspread_percent); + dcn2_0_soc.dram_page_open_time_ns = + fixed16_to_double_to_cpu(bb->dram_page_open_time_ns); + dcn2_0_soc.dram_rw_turnaround_time_ns = + fixed16_to_double_to_cpu(bb->dram_rw_turnaround_time_ns); + dcn2_0_soc.dram_return_buffer_per_channel_bytes = + le32_to_cpu(bb->dram_return_buffer_per_channel_bytes); + dcn2_0_soc.round_trip_ping_latency_dcfclk_cycles = + le32_to_cpu(bb->round_trip_ping_latency_dcfclk_cycles); + dcn2_0_soc.urgent_out_of_order_return_per_channel_bytes = + le32_to_cpu(bb->urgent_out_of_order_return_per_channel_bytes); + dcn2_0_soc.channel_interleave_bytes = + le32_to_cpu(bb->channel_interleave_bytes); + dcn2_0_soc.num_banks = + le32_to_cpu(bb->num_banks); + dcn2_0_soc.num_chans = + le32_to_cpu(bb->num_chans); + dcn2_0_soc.vmm_page_size_bytes = + le32_to_cpu(bb->vmm_page_size_bytes); + dcn2_0_soc.dram_clock_change_latency_us = + fixed16_to_double_to_cpu(bb->dram_clock_change_latency_us); + dcn2_0_soc.writeback_dram_clock_change_latency_us = + fixed16_to_double_to_cpu(bb->writeback_dram_clock_change_latency_us); + dcn2_0_soc.return_bus_width_bytes = + le32_to_cpu(bb->return_bus_width_bytes); + dcn2_0_soc.dispclk_dppclk_vco_speed_mhz = + le32_to_cpu(bb->dispclk_dppclk_vco_speed_mhz); + dcn2_0_soc.xfc_bus_transport_time_us = + le32_to_cpu(bb->xfc_bus_transport_time_us); + dcn2_0_soc.xfc_xbuf_latency_tolerance_us = + le32_to_cpu(bb->xfc_xbuf_latency_tolerance_us); + dcn2_0_soc.use_urgent_burst_bw = + le32_to_cpu(bb->use_urgent_burst_bw); + dcn2_0_soc.num_states = + le32_to_cpu(bb->num_states); + + for (i = 0; i < dcn2_0_soc.num_states; i++) { + dcn2_0_soc.clock_limits[i].state = + le32_to_cpu(bb->clock_limits[i].state); + dcn2_0_soc.clock_limits[i].dcfclk_mhz = + fixed16_to_double_to_cpu(bb->clock_limits[i].dcfclk_mhz); + dcn2_0_soc.clock_limits[i].fabricclk_mhz = + fixed16_to_double_to_cpu(bb->clock_limits[i].fabricclk_mhz); + dcn2_0_soc.clock_limits[i].dispclk_mhz = + fixed16_to_double_to_cpu(bb->clock_limits[i].dispclk_mhz); + dcn2_0_soc.clock_limits[i].dppclk_mhz = + fixed16_to_double_to_cpu(bb->clock_limits[i].dppclk_mhz); + dcn2_0_soc.clock_limits[i].phyclk_mhz = + fixed16_to_double_to_cpu(bb->clock_limits[i].phyclk_mhz); + dcn2_0_soc.clock_limits[i].socclk_mhz = + fixed16_to_double_to_cpu(bb->clock_limits[i].socclk_mhz); + dcn2_0_soc.clock_limits[i].dscclk_mhz = + fixed16_to_double_to_cpu(bb->clock_limits[i].dscclk_mhz); + dcn2_0_soc.clock_limits[i].dram_speed_mts = + fixed16_to_double_to_cpu(bb->clock_limits[i].dram_speed_mts); + } + } + + if (pool->base.pp_smu) { + struct pp_smu_nv_clock_table max_clocks = {0}; + unsigned int uclk_states[8] = {0}; + unsigned int num_states = 0; + enum pp_smu_status status; + bool clock_limits_available = false; + bool uclk_states_available = false; + + if (pool->base.pp_smu->nv_funcs.get_uclk_dpm_states) { + status = (pool->base.pp_smu->nv_funcs.get_uclk_dpm_states) + (&pool->base.pp_smu->nv_funcs.pp_smu, uclk_states, &num_states); + + uclk_states_available = (status == PP_SMU_RESULT_OK); + } + + if (pool->base.pp_smu->nv_funcs.get_maximum_sustainable_clocks) { + status = (*pool->base.pp_smu->nv_funcs.get_maximum_sustainable_clocks) + (&pool->base.pp_smu->nv_funcs.pp_smu, &max_clocks); + /* SMU cannot set DCF clock to anything equal to or higher than SOC clock + */ + if (max_clocks.dcfClockInKhz >= max_clocks.socClockInKhz) + max_clocks.dcfClockInKhz = max_clocks.socClockInKhz - 1000; + clock_limits_available = (status == PP_SMU_RESULT_OK); + } + + if (clock_limits_available && uclk_states_available && num_states) + update_bounding_box(dc, &dcn2_0_soc, &max_clocks, uclk_states, num_states); + else if (clock_limits_available) + cap_soc_clocks(&dcn2_0_soc, max_clocks); + } + + dcn2_0_ip.max_num_otg = pool->base.res_cap->num_timing_generator; + dcn2_0_ip.max_num_dpp = pool->base.pipe_count; + patch_bounding_box(dc, &dcn2_0_soc); + + return true; +} + +static bool construct( + uint8_t num_virtual_links, + struct dc *dc, + struct dcn20_resource_pool *pool) +{ + int i; + struct dc_context *ctx = dc->ctx; + struct irq_service_init_data init_data; + + ctx->dc_bios->regs = &bios_regs; + + pool->base.res_cap = &res_cap_nv10; + pool->base.funcs = &dcn20_res_pool_funcs; + + /************************************************* + * Resource + asic cap harcoding * + *************************************************/ + pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE; + + pool->base.pipe_count = 6; + pool->base.mpcc_count = 6; + dc->caps.max_downscale_ratio = 200; + dc->caps.i2c_speed_in_khz = 100; + dc->caps.max_cursor_size = 256; + dc->caps.dmdata_alloc_size = 2048; + + dc->caps.max_slave_planes = 1; + dc->caps.post_blend_color_processing = true; + dc->caps.force_dp_tps4_for_cp2520 = true; + dc->caps.hw_3d_lut = true; + + if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV) + dc->debug = debug_defaults_drv; + else if (dc->ctx->dce_environment == DCE_ENV_FPGA_MAXIMUS) { + pool->base.pipe_count = 4; + + pool->base.mpcc_count = pool->base.pipe_count; + dc->debug = debug_defaults_diags; + } else + dc->debug = debug_defaults_diags; + //dcn2.0x + dc->work_arounds.dedcn20_305_wa = true; + + // Init the vm_helper + if (dc->vm_helper) + vm_helper_init(dc->vm_helper, 16); + + /************************************************* + * Create resources * + *************************************************/ + + pool->base.clock_sources[DCN20_CLK_SRC_PLL0] = + dcn20_clock_source_create(ctx, ctx->dc_bios, + CLOCK_SOURCE_COMBO_PHY_PLL0, + &clk_src_regs[0], false); + pool->base.clock_sources[DCN20_CLK_SRC_PLL1] = + dcn20_clock_source_create(ctx, ctx->dc_bios, + CLOCK_SOURCE_COMBO_PHY_PLL1, + &clk_src_regs[1], false); + pool->base.clock_sources[DCN20_CLK_SRC_PLL2] = + dcn20_clock_source_create(ctx, ctx->dc_bios, + CLOCK_SOURCE_COMBO_PHY_PLL2, + &clk_src_regs[2], false); + pool->base.clock_sources[DCN20_CLK_SRC_PLL3] = + dcn20_clock_source_create(ctx, ctx->dc_bios, + CLOCK_SOURCE_COMBO_PHY_PLL3, + &clk_src_regs[3], false); + pool->base.clock_sources[DCN20_CLK_SRC_PLL4] = + dcn20_clock_source_create(ctx, ctx->dc_bios, + CLOCK_SOURCE_COMBO_PHY_PLL4, + &clk_src_regs[4], false); + pool->base.clock_sources[DCN20_CLK_SRC_PLL5] = + dcn20_clock_source_create(ctx, ctx->dc_bios, + CLOCK_SOURCE_COMBO_PHY_PLL5, + &clk_src_regs[5], false); + pool->base.clk_src_count = DCN20_CLK_SRC_TOTAL; + /* todo: not reuse phy_pll registers */ + pool->base.dp_clock_source = + dcn20_clock_source_create(ctx, ctx->dc_bios, + CLOCK_SOURCE_ID_DP_DTO, + &clk_src_regs[0], true); + + for (i = 0; i < pool->base.clk_src_count; i++) { + if (pool->base.clock_sources[i] == NULL) { + dm_error("DC: failed to create clock sources!\n"); + BREAK_TO_DEBUGGER(); + goto create_fail; + } + } + + pool->base.dccg = dccg2_create(ctx, &dccg_regs, &dccg_shift, &dccg_mask); + if (pool->base.dccg == NULL) { + dm_error("DC: failed to create dccg!\n"); + BREAK_TO_DEBUGGER(); + goto create_fail; + } + + pool->base.dmcu = dcn20_dmcu_create(ctx, + &dmcu_regs, + &dmcu_shift, + &dmcu_mask); + if (pool->base.dmcu == NULL) { + dm_error("DC: failed to create dmcu!\n"); + BREAK_TO_DEBUGGER(); + goto create_fail; + } + + pool->base.abm = dce_abm_create(ctx, + &abm_regs, + &abm_shift, + &abm_mask); + if (pool->base.abm == NULL) { + dm_error("DC: failed to create abm!\n"); + BREAK_TO_DEBUGGER(); + goto create_fail; + } + + pool->base.pp_smu = dcn20_pp_smu_create(ctx); + + + if (!init_soc_bounding_box(dc, pool)) { + dm_error("DC: failed to initialize soc bounding box!\n"); + BREAK_TO_DEBUGGER(); + goto create_fail; + } + + dml_init_instance(&dc->dml, &dcn2_0_soc, &dcn2_0_ip, DML_PROJECT_NAVI10); + + if (!dc->debug.disable_pplib_wm_range) { + struct pp_smu_wm_range_sets ranges = {0}; + int i = 0; + + ranges.num_reader_wm_sets = 0; + + if (dcn2_0_soc.num_states == 1) { + ranges.reader_wm_sets[0].wm_inst = i; + ranges.reader_wm_sets[0].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN; + ranges.reader_wm_sets[0].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX; + ranges.reader_wm_sets[0].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN; + ranges.reader_wm_sets[0].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX; + + ranges.num_reader_wm_sets = 1; + } else if (dcn2_0_soc.num_states > 1) { + for (i = 0; i < 4 && i < dcn2_0_soc.num_states; i++) { + ranges.reader_wm_sets[i].wm_inst = i; + ranges.reader_wm_sets[i].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN; + ranges.reader_wm_sets[i].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX; + ranges.reader_wm_sets[i].min_fill_clk_mhz = (i > 0) ? (dcn2_0_soc.clock_limits[i - 1].dram_speed_mts / 16) + 1 : 0; + ranges.reader_wm_sets[i].max_fill_clk_mhz = dcn2_0_soc.clock_limits[i].dram_speed_mts / 16; + + ranges.num_reader_wm_sets = i + 1; + } + + ranges.reader_wm_sets[0].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN; + ranges.reader_wm_sets[ranges.num_reader_wm_sets - 1].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX; + } + + ranges.num_writer_wm_sets = 1; + + ranges.writer_wm_sets[0].wm_inst = 0; + ranges.writer_wm_sets[0].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN; + ranges.writer_wm_sets[0].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX; + ranges.writer_wm_sets[0].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN; + ranges.writer_wm_sets[0].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX; + + /* Notify PP Lib/SMU which Watermarks to use for which clock ranges */ + if (pool->base.pp_smu->nv_funcs.set_wm_ranges) + pool->base.pp_smu->nv_funcs.set_wm_ranges(&pool->base.pp_smu->nv_funcs.pp_smu, &ranges); + } + + init_data.ctx = dc->ctx; + pool->base.irqs = dal_irq_service_dcn20_create(&init_data); + if (!pool->base.irqs) + goto create_fail; + + /* mem input -> ipp -> dpp -> opp -> TG */ + for (i = 0; i < pool->base.pipe_count; i++) { + pool->base.hubps[i] = dcn20_hubp_create(ctx, i); + if (pool->base.hubps[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC: failed to create memory input!\n"); + goto create_fail; + } + + pool->base.ipps[i] = dcn20_ipp_create(ctx, i); + if (pool->base.ipps[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC: failed to create input pixel processor!\n"); + goto create_fail; + } + + pool->base.dpps[i] = dcn20_dpp_create(ctx, i); + if (pool->base.dpps[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC: failed to create dpps!\n"); + goto create_fail; + } + } + for (i = 0; i < pool->base.res_cap->num_ddc; i++) { + pool->base.engines[i] = dcn20_aux_engine_create(ctx, i); + if (pool->base.engines[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC:failed to create aux engine!!\n"); + goto create_fail; + } + pool->base.hw_i2cs[i] = dcn20_i2c_hw_create(ctx, i); + if (pool->base.hw_i2cs[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC:failed to create hw i2c!!\n"); + goto create_fail; + } + pool->base.sw_i2cs[i] = NULL; + } + + for (i = 0; i < pool->base.res_cap->num_opp; i++) { + pool->base.opps[i] = dcn20_opp_create(ctx, i); + if (pool->base.opps[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC: failed to create output pixel processor!\n"); + goto create_fail; + } + } + + for (i = 0; i < pool->base.res_cap->num_timing_generator; i++) { + pool->base.timing_generators[i] = dcn20_timing_generator_create( + ctx, i); + if (pool->base.timing_generators[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error("DC: failed to create tg!\n"); + goto create_fail; + } + } + + pool->base.timing_generator_count = i; + + pool->base.mpc = dcn20_mpc_create(ctx); + if (pool->base.mpc == NULL) { + BREAK_TO_DEBUGGER(); + dm_error("DC: failed to create mpc!\n"); + goto create_fail; + } + + pool->base.hubbub = dcn20_hubbub_create(ctx); + if (pool->base.hubbub == NULL) { + BREAK_TO_DEBUGGER(); + dm_error("DC: failed to create hubbub!\n"); + goto create_fail; + } + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + for (i = 0; i < pool->base.res_cap->num_dsc; i++) { + pool->base.dscs[i] = dcn20_dsc_create(ctx, i); + if (pool->base.dscs[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error("DC: failed to create display stream compressor %d!\n", i); + goto create_fail; + } + } +#endif + + if (!dcn20_dwbc_create(ctx, &pool->base)) { + BREAK_TO_DEBUGGER(); + dm_error("DC: failed to create dwbc!\n"); + goto create_fail; + } + if (!dcn20_mmhubbub_create(ctx, &pool->base)) { + BREAK_TO_DEBUGGER(); + dm_error("DC: failed to create mcif_wb!\n"); + goto create_fail; + } + + if (!resource_construct(num_virtual_links, dc, &pool->base, + (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment) ? + &res_create_funcs : &res_create_maximus_funcs))) + goto create_fail; + + dcn20_hw_sequencer_construct(dc); + + dc->caps.max_planes = pool->base.pipe_count; + + for (i = 0; i < dc->caps.max_planes; ++i) + dc->caps.planes[i] = plane_cap; + + dc->cap_funcs = cap_funcs; + + return true; + +create_fail: + + destruct(pool); + + return false; +} + +struct resource_pool *dcn20_create_resource_pool( + const struct dc_init_data *init_data, + struct dc *dc) +{ + struct dcn20_resource_pool *pool = + kzalloc(sizeof(struct dcn20_resource_pool), GFP_KERNEL); + + if (!pool) + return NULL; + + if (construct(init_data->num_virtual_links, dc, pool)) + return &pool->base; + + BREAK_TO_DEBUGGER(); + kfree(pool); + return NULL; +} diff --cc drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c index 791aa745efd22,0000000000000..f5bcffc426b84 mode 100644,000000..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 @@@ -1,608 -1,0 +1,610 @@@ +/* + * Copyright 2012-15 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + ++#include ++ +#include "dc_bios_types.h" +#include "dcn20_stream_encoder.h" +#include "reg_helper.h" +#include "hw_shared.h" + +#define DC_LOGGER \ + enc1->base.ctx->logger + + +#define REG(reg)\ + (enc1->regs->reg) + +#undef FN +#define FN(reg_name, field_name) \ + enc1->se_shift->field_name, enc1->se_mask->field_name + + +#define CTX \ + enc1->base.ctx + + +static void enc2_update_hdmi_info_packet( + struct dcn10_stream_encoder *enc1, + uint32_t packet_index, + const struct dc_info_packet *info_packet) +{ + uint32_t cont, send, line; + + if (info_packet->valid) { + enc1_update_generic_info_packet( + enc1, + packet_index, + info_packet); + + /* enable transmission of packet(s) - + * packet transmission begins on the next frame */ + cont = 1; + /* send packet(s) every frame */ + send = 1; + /* select line number to send packets on */ + line = 2; + } else { + cont = 0; + send = 0; + line = 0; + } + + /* DP_SEC_GSP[x]_LINE_REFERENCE - keep default value REFER_TO_DP_SOF */ + + /* choose which generic packet control to use */ + switch (packet_index) { + case 0: + REG_UPDATE_2(HDMI_GENERIC_PACKET_CONTROL0, + HDMI_GENERIC0_CONT, cont, + HDMI_GENERIC0_SEND, send); + REG_UPDATE(HDMI_GENERIC_PACKET_CONTROL1, + HDMI_GENERIC0_LINE, line); + break; + case 1: + REG_UPDATE_2(HDMI_GENERIC_PACKET_CONTROL0, + HDMI_GENERIC1_CONT, cont, + HDMI_GENERIC1_SEND, send); + REG_UPDATE(HDMI_GENERIC_PACKET_CONTROL1, + HDMI_GENERIC1_LINE, line); + break; + case 2: + REG_UPDATE_2(HDMI_GENERIC_PACKET_CONTROL0, + HDMI_GENERIC2_CONT, cont, + HDMI_GENERIC2_SEND, send); + REG_UPDATE(HDMI_GENERIC_PACKET_CONTROL2, + HDMI_GENERIC2_LINE, line); + break; + case 3: + REG_UPDATE_2(HDMI_GENERIC_PACKET_CONTROL0, + HDMI_GENERIC3_CONT, cont, + HDMI_GENERIC3_SEND, send); + REG_UPDATE(HDMI_GENERIC_PACKET_CONTROL2, + HDMI_GENERIC3_LINE, line); + break; + case 4: + REG_UPDATE_2(HDMI_GENERIC_PACKET_CONTROL0, + HDMI_GENERIC4_CONT, cont, + HDMI_GENERIC4_SEND, send); + REG_UPDATE(HDMI_GENERIC_PACKET_CONTROL3, + HDMI_GENERIC4_LINE, line); + break; + case 5: + REG_UPDATE_2(HDMI_GENERIC_PACKET_CONTROL0, + HDMI_GENERIC5_CONT, cont, + HDMI_GENERIC5_SEND, send); + REG_UPDATE(HDMI_GENERIC_PACKET_CONTROL3, + HDMI_GENERIC5_LINE, line); + break; + case 6: + REG_UPDATE_2(HDMI_GENERIC_PACKET_CONTROL0, + HDMI_GENERIC6_CONT, cont, + HDMI_GENERIC6_SEND, send); + REG_UPDATE(HDMI_GENERIC_PACKET_CONTROL4, + HDMI_GENERIC6_LINE, line); + break; + case 7: + REG_UPDATE_2(HDMI_GENERIC_PACKET_CONTROL0, + HDMI_GENERIC7_CONT, cont, + HDMI_GENERIC7_SEND, send); + REG_UPDATE(HDMI_GENERIC_PACKET_CONTROL4, + HDMI_GENERIC7_LINE, line); + break; + default: + /* invalid HW packet index */ + DC_LOG_WARNING( + "Invalid HW packet index: %s()\n", + __func__); + return; + } +} + +static void enc2_stream_encoder_update_hdmi_info_packets( + struct stream_encoder *enc, + const struct encoder_info_frame *info_frame) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + + /* for bring up, disable dp double TODO */ + REG_UPDATE(HDMI_DB_CONTROL, HDMI_DB_DISABLE, 1); + + /*Always add mandatory packets first followed by optional ones*/ + enc2_update_hdmi_info_packet(enc1, 0, &info_frame->avi); + enc2_update_hdmi_info_packet(enc1, 5, &info_frame->hfvsif); + enc2_update_hdmi_info_packet(enc1, 2, &info_frame->gamut); + enc2_update_hdmi_info_packet(enc1, 1, &info_frame->vendor); + enc2_update_hdmi_info_packet(enc1, 3, &info_frame->spd); + enc2_update_hdmi_info_packet(enc1, 4, &info_frame->hdrsmd); +} + +static void enc2_stream_encoder_stop_hdmi_info_packets( + struct stream_encoder *enc) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + + /* stop generic packets 0,1 on HDMI */ + REG_SET_4(HDMI_GENERIC_PACKET_CONTROL0, 0, + HDMI_GENERIC0_CONT, 0, + HDMI_GENERIC0_SEND, 0, + HDMI_GENERIC1_CONT, 0, + HDMI_GENERIC1_SEND, 0); + REG_SET_2(HDMI_GENERIC_PACKET_CONTROL1, 0, + HDMI_GENERIC0_LINE, 0, + HDMI_GENERIC1_LINE, 0); + + /* stop generic packets 2,3 on HDMI */ + REG_SET_4(HDMI_GENERIC_PACKET_CONTROL0, 0, + HDMI_GENERIC2_CONT, 0, + HDMI_GENERIC2_SEND, 0, + HDMI_GENERIC3_CONT, 0, + HDMI_GENERIC3_SEND, 0); + REG_SET_2(HDMI_GENERIC_PACKET_CONTROL2, 0, + HDMI_GENERIC2_LINE, 0, + HDMI_GENERIC3_LINE, 0); + + /* stop generic packets 4,5 on HDMI */ + REG_SET_4(HDMI_GENERIC_PACKET_CONTROL0, 0, + HDMI_GENERIC4_CONT, 0, + HDMI_GENERIC4_SEND, 0, + HDMI_GENERIC5_CONT, 0, + HDMI_GENERIC5_SEND, 0); + REG_SET_2(HDMI_GENERIC_PACKET_CONTROL3, 0, + HDMI_GENERIC4_LINE, 0, + HDMI_GENERIC5_LINE, 0); + + /* stop generic packets 6,7 on HDMI */ + REG_SET_4(HDMI_GENERIC_PACKET_CONTROL0, 0, + HDMI_GENERIC6_CONT, 0, + HDMI_GENERIC6_SEND, 0, + HDMI_GENERIC7_CONT, 0, + HDMI_GENERIC7_SEND, 0); + REG_SET_2(HDMI_GENERIC_PACKET_CONTROL4, 0, + HDMI_GENERIC6_LINE, 0, + HDMI_GENERIC7_LINE, 0); +} + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + + +/* Update GSP7 SDP 128 byte long */ +static void enc2_send_gsp7_128_info_packet( + struct dcn10_stream_encoder *enc1, + const struct dc_info_packet_128 *info_packet) +{ + uint32_t i; + + /* TODOFPGA Figure out a proper number for max_retries polling for lock + * use 50 for now. + */ + uint32_t max_retries = 50; + const uint32_t *content = (const uint32_t *) &info_packet->sb[0]; + + ASSERT(info_packet->hb1 == DC_DP_INFOFRAME_TYPE_PPS); + + /* Configure for PPS packet size (128 bytes) */ + REG_UPDATE(DP_SEC_CNTL2, DP_SEC_GSP7_PPS, 1); + + /* We need turn on clock before programming AFMT block*/ + REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1); + + /* Poll dig_update_lock is not locked -> asic internal signal + * assumes otg master lock will unlock it + */ + /*REG_WAIT(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_LOCK_STATUS, 0, 10, max_retries);*/ + + /* Wait for HW/SW GSP memory access conflict to go away */ + REG_WAIT(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT, + 0, 10, max_retries); + + /* Clear HW/SW memory access conflict flag */ + REG_UPDATE(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT_CLR, 1); + + /* write generic packet header */ + REG_UPDATE(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_INDEX, 7); + REG_SET_4(AFMT_GENERIC_HDR, 0, + AFMT_GENERIC_HB0, info_packet->hb0, + AFMT_GENERIC_HB1, info_packet->hb1, + AFMT_GENERIC_HB2, info_packet->hb2, + AFMT_GENERIC_HB3, info_packet->hb3); + + /* Write generic packet content 128 bytes long. Four sets are used (indexes 7 + * through 10) to fit 128 bytes. + */ + for (i = 0; i < 4; i++) { + uint32_t packet_index = 7 + i; + REG_UPDATE(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_INDEX, packet_index); + + REG_WRITE(AFMT_GENERIC_0, *content++); + REG_WRITE(AFMT_GENERIC_1, *content++); + REG_WRITE(AFMT_GENERIC_2, *content++); + REG_WRITE(AFMT_GENERIC_3, *content++); + REG_WRITE(AFMT_GENERIC_4, *content++); + REG_WRITE(AFMT_GENERIC_5, *content++); + REG_WRITE(AFMT_GENERIC_6, *content++); + REG_WRITE(AFMT_GENERIC_7, *content++); + } + + REG_UPDATE(AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC7_FRAME_UPDATE, 1); +} + +/* Set DSC-related configuration. + * dsc_mode: 0 disables DSC, other values enable DSC in specified format + * sc_bytes_per_pixel: Bytes per pixel in u3.28 format + * dsc_slice_width: Slice width in pixels + */ +static void enc2_dp_set_dsc_config(struct stream_encoder *enc, + enum optc_dsc_mode dsc_mode, + uint32_t dsc_bytes_per_pixel, + uint32_t dsc_slice_width, + uint8_t *dsc_packed_pps) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + uint32_t dsc_value = 0; + + dsc_value = REG_READ(DP_DSC_CNTL); + + /* dsc disable skip */ + if ((dsc_value & 0x3) == 0x0) + return; + + + REG_UPDATE_2(DP_DSC_CNTL, + DP_DSC_MODE, dsc_mode, + DP_DSC_SLICE_WIDTH, dsc_slice_width); + + REG_SET(DP_DSC_BYTES_PER_PIXEL, 0, + DP_DSC_BYTES_PER_PIXEL, dsc_bytes_per_pixel); + + if (dsc_mode != OPTC_DSC_DISABLED) { + struct dc_info_packet_128 pps_sdp; + + ASSERT(dsc_packed_pps); + + /* Load PPS into infoframe (SDP) registers */ + pps_sdp.valid = true; + pps_sdp.hb0 = 0; + pps_sdp.hb1 = DC_DP_INFOFRAME_TYPE_PPS; + pps_sdp.hb2 = 127; + pps_sdp.hb3 = 0; + memcpy(&pps_sdp.sb[0], dsc_packed_pps, sizeof(pps_sdp.sb)); + enc2_send_gsp7_128_info_packet(enc1, &pps_sdp); + + /* Enable Generic Stream Packet 7 (GSP) transmission */ + //REG_UPDATE(DP_SEC_CNTL, + // DP_SEC_GSP7_ENABLE, 1); + + /* SW should make sure VBID[6] update line number is bigger + * than PPS transmit line number + */ + REG_UPDATE(DP_SEC_CNTL6, + DP_SEC_GSP7_LINE_NUM, 2); + REG_UPDATE_2(DP_MSA_VBID_MISC, + DP_VBID6_LINE_REFERENCE, 0, + DP_VBID6_LINE_NUM, 3); + + /* Send PPS data at the line number specified above. + * DP spec requires PPS to be sent only when it changes, however since + * decoder has to be able to handle its change on every frame, we're + * sending it always (i.e. on every frame) to reduce the chance it'd be + * missed by decoder. If it turns out required to send PPS only when it + * changes, we can use DP_SEC_GSP7_SEND register. + */ + REG_UPDATE_2(DP_SEC_CNTL, + DP_SEC_GSP7_ENABLE, 1, + DP_SEC_STREAM_ENABLE, 1); + } else { + /* Disable Generic Stream Packet 7 (GSP) transmission */ + REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP7_ENABLE, 0); + REG_UPDATE(DP_SEC_CNTL2, DP_SEC_GSP7_PPS, 0); + } +} +#endif + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT +/* this function read dsc related register fields to be logged later in dcn10_log_hw_state + * into a dcn_dsc_state struct. + */ +static void enc2_read_state(struct stream_encoder *enc, struct enc_state *s) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + + //if dsc is enabled, continue to read + REG_GET(DP_DSC_CNTL, DP_DSC_MODE, &s->dsc_mode); + if (s->dsc_mode) { + REG_GET(DP_DSC_CNTL, DP_DSC_SLICE_WIDTH, &s->dsc_slice_width); + REG_GET(DP_SEC_CNTL6, DP_SEC_GSP7_LINE_NUM, &s->sec_gsp_pps_line_num); + + REG_GET(DP_MSA_VBID_MISC, DP_VBID6_LINE_REFERENCE, &s->vbid6_line_reference); + REG_GET(DP_MSA_VBID_MISC, DP_VBID6_LINE_NUM, &s->vbid6_line_num); + + REG_GET(DP_SEC_CNTL, DP_SEC_GSP7_ENABLE, &s->sec_gsp_pps_enable); + REG_GET(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, &s->sec_stream_enable); + } +} +#endif + +/* Set Dynamic Metadata-configuration. + * enable_dme: TRUE: enables Dynamic Metadata Enfine, FALSE: disables DME + * hubp_requestor_id: HUBP physical instance that is the source of dynamic metadata + * only needs to be set when enable_dme is TRUE + * dmdata_mode: dynamic metadata packet type: DP, HDMI, or Dolby Vision + * + * Ensure the OTG master update lock is set when changing DME configuration. + */ +static void enc2_set_dynamic_metadata(struct stream_encoder *enc, + bool enable_dme, + uint32_t hubp_requestor_id, + enum dynamic_metadata_mode dmdata_mode) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + + if (enable_dme) { + REG_UPDATE_2(DME_CONTROL, + METADATA_HUBP_REQUESTOR_ID, hubp_requestor_id, + METADATA_STREAM_TYPE, (dmdata_mode == dmdata_dolby_vision) ? 1 : 0); + + /* Use default line reference DP_SOF for bringup. + * Should use OTG_SOF for DRR cases + */ + if (dmdata_mode == dmdata_dp) + REG_UPDATE_3(DP_SEC_METADATA_TRANSMISSION, + DP_SEC_METADATA_PACKET_ENABLE, 1, + DP_SEC_METADATA_PACKET_LINE_REFERENCE, 0, + DP_SEC_METADATA_PACKET_LINE, 20); + else { + REG_UPDATE_3(HDMI_METADATA_PACKET_CONTROL, + HDMI_METADATA_PACKET_ENABLE, 1, + HDMI_METADATA_PACKET_LINE_REFERENCE, 0, + HDMI_METADATA_PACKET_LINE, 2); + + if (dmdata_mode == dmdata_dolby_vision) + REG_UPDATE(DIG_FE_CNTL, + DOLBY_VISION_EN, 1); + } + + REG_UPDATE(DME_CONTROL, + METADATA_ENGINE_EN, 1); + } else { + REG_UPDATE(DME_CONTROL, + METADATA_ENGINE_EN, 0); + + if (dmdata_mode == dmdata_dp) + REG_UPDATE(DP_SEC_METADATA_TRANSMISSION, + DP_SEC_METADATA_PACKET_ENABLE, 0); + else { + REG_UPDATE(HDMI_METADATA_PACKET_CONTROL, + HDMI_METADATA_PACKET_ENABLE, 0); + REG_UPDATE(DIG_FE_CNTL, + DOLBY_VISION_EN, 0); + } + } +} + +static void enc2_stream_encoder_update_dp_info_packets( + struct stream_encoder *enc, + const struct encoder_info_frame *info_frame) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + uint32_t dmdata_packet_enabled = 0; + + enc1_stream_encoder_update_dp_info_packets(enc, info_frame); + + /* check if dynamic metadata packet transmission is enabled */ + REG_GET(DP_SEC_METADATA_TRANSMISSION, + DP_SEC_METADATA_PACKET_ENABLE, &dmdata_packet_enabled); + + if (dmdata_packet_enabled) + REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1); +} + +static bool is_two_pixels_per_containter(const struct dc_crtc_timing *timing) +{ + bool two_pix = timing->pixel_encoding == PIXEL_ENCODING_YCBCR420; + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + two_pix = two_pix || (timing->flags.DSC && timing->pixel_encoding == PIXEL_ENCODING_YCBCR422 + && !timing->dsc_cfg.ycbcr422_simple); +#endif + return two_pix; +} + +void enc2_stream_encoder_dp_unblank( + struct stream_encoder *enc, + const struct encoder_unblank_param *param) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + + if (param->link_settings.link_rate != LINK_RATE_UNKNOWN) { + uint32_t n_vid = 0x8000; + uint32_t m_vid; + uint32_t n_multiply = 0; + uint64_t m_vid_l = n_vid; + + /* YCbCr 4:2:0 : Computed VID_M will be 2X the input rate */ + if (is_two_pixels_per_containter(¶m->timing) || param->odm) { + /*this logic should be the same in get_pixel_clock_parameters() */ + n_multiply = 1; + } + /* M / N = Fstream / Flink + * m_vid / n_vid = pixel rate / link rate + */ + + m_vid_l *= param->timing.pix_clk_100hz / 10; + m_vid_l = div_u64(m_vid_l, + param->link_settings.link_rate + * LINK_RATE_REF_FREQ_IN_KHZ); + + m_vid = (uint32_t) m_vid_l; + + /* enable auto measurement */ + + REG_UPDATE(DP_VID_TIMING, DP_VID_M_N_GEN_EN, 0); + + /* auto measurement need 1 full 0x8000 symbol cycle to kick in, + * therefore program initial value for Mvid and Nvid + */ + + REG_UPDATE(DP_VID_N, DP_VID_N, n_vid); + + REG_UPDATE(DP_VID_M, DP_VID_M, m_vid); + + REG_UPDATE_2(DP_VID_TIMING, + DP_VID_M_N_GEN_EN, 1, + DP_VID_N_MUL, n_multiply); + } + + /* set DIG_START to 0x1 to reset FIFO */ + + REG_UPDATE(DIG_FE_CNTL, DIG_START, 1); + + /* write 0 to take the FIFO out of reset */ + + REG_UPDATE(DIG_FE_CNTL, DIG_START, 0); + + /* switch DP encoder to CRTC data */ + + REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 0); + + /* wait 100us for DIG/DP logic to prime + * (i.e. a few video lines) + */ + udelay(100); + + /* the hardware would start sending video at the start of the next DP + * frame (i.e. rising edge of the vblank). + * NOTE: We used to program DP_VID_STREAM_DIS_DEFER = 2 here, but this + * register has no effect on enable transition! HW always guarantees + * VID_STREAM enable at start of next frame, and this is not + * programmable + */ + + REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, true); +} + +static void enc2_dp_set_odm_combine( + struct stream_encoder *enc, + bool odm_combine) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + + REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_COMBINE, odm_combine); +} + +void enc2_stream_encoder_dp_set_stream_attribute( + struct stream_encoder *enc, + struct dc_crtc_timing *crtc_timing, + enum dc_color_space output_color_space, + uint32_t enable_sdp_splitting) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + + enc1_stream_encoder_dp_set_stream_attribute(enc, crtc_timing, output_color_space, enable_sdp_splitting); + + REG_UPDATE(DP_SEC_FRAMING4, + DP_SST_SDP_SPLITTING, enable_sdp_splitting); +} + +static const struct stream_encoder_funcs dcn20_str_enc_funcs = { + .dp_set_odm_combine = + enc2_dp_set_odm_combine, + .dp_set_stream_attribute = + enc2_stream_encoder_dp_set_stream_attribute, + .hdmi_set_stream_attribute = + enc1_stream_encoder_hdmi_set_stream_attribute, + .dvi_set_stream_attribute = + enc1_stream_encoder_dvi_set_stream_attribute, + .set_mst_bandwidth = + enc1_stream_encoder_set_mst_bandwidth, + .update_hdmi_info_packets = + enc2_stream_encoder_update_hdmi_info_packets, + .stop_hdmi_info_packets = + enc2_stream_encoder_stop_hdmi_info_packets, + .update_dp_info_packets = + enc2_stream_encoder_update_dp_info_packets, + .stop_dp_info_packets = + enc1_stream_encoder_stop_dp_info_packets, + .dp_blank = + enc1_stream_encoder_dp_blank, + .dp_unblank = + enc2_stream_encoder_dp_unblank, + .audio_mute_control = enc1_se_audio_mute_control, + + .dp_audio_setup = enc1_se_dp_audio_setup, + .dp_audio_enable = enc1_se_dp_audio_enable, + .dp_audio_disable = enc1_se_dp_audio_disable, + + .hdmi_audio_setup = enc1_se_hdmi_audio_setup, + .hdmi_audio_disable = enc1_se_hdmi_audio_disable, + .setup_stereo_sync = enc1_setup_stereo_sync, + .set_avmute = enc1_stream_encoder_set_avmute, + .dig_connect_to_otg = enc1_dig_connect_to_otg, +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + .enc_read_state = enc2_read_state, +#endif + +#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT + .dp_set_dsc_config = enc2_dp_set_dsc_config, +#endif + .set_dynamic_metadata = enc2_set_dynamic_metadata, +}; + +void dcn20_stream_encoder_construct( + struct dcn10_stream_encoder *enc1, + struct dc_context *ctx, + struct dc_bios *bp, + enum engine_id eng_id, + const struct dcn10_stream_enc_registers *regs, + const struct dcn10_stream_encoder_shift *se_shift, + const struct dcn10_stream_encoder_mask *se_mask) +{ + enc1->base.funcs = &dcn20_str_enc_funcs; + enc1->base.ctx = ctx; + enc1->base.id = eng_id; + enc1->base.bp = bp; + enc1->regs = regs; + enc1->se_shift = se_shift; + enc1->se_mask = se_mask; +} + diff --cc drivers/gpu/drm/amd/display/dc/irq/dcn20/irq_service_dcn20.c index 65866d620759e,0000000000000..3cc0f2a1f77cc mode 100644,000000..100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dcn20/irq_service_dcn20.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dcn20/irq_service_dcn20.c @@@ -1,373 -1,0 +1,375 @@@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + ++#include ++ +#include "dm_services.h" + +#include "include/logger_interface.h" + +#include "../dce110/irq_service_dce110.h" + +#include "dcn/dcn_2_0_0_offset.h" +#include "dcn/dcn_2_0_0_sh_mask.h" +#include "navi10_ip_offset.h" + + +#include "irq_service_dcn20.h" + +#include "ivsrcid/dcn/irqsrcs_dcn_1_0.h" + +enum dc_irq_source to_dal_irq_source_dcn20( + struct irq_service *irq_service, + uint32_t src_id, + uint32_t ext_id) +{ + switch (src_id) { + case DCN_1_0__SRCID__DC_D1_OTG_VSTARTUP: + return DC_IRQ_SOURCE_VBLANK1; + case DCN_1_0__SRCID__DC_D2_OTG_VSTARTUP: + return DC_IRQ_SOURCE_VBLANK2; + case DCN_1_0__SRCID__DC_D3_OTG_VSTARTUP: + return DC_IRQ_SOURCE_VBLANK3; + case DCN_1_0__SRCID__DC_D4_OTG_VSTARTUP: + return DC_IRQ_SOURCE_VBLANK4; + case DCN_1_0__SRCID__DC_D5_OTG_VSTARTUP: + return DC_IRQ_SOURCE_VBLANK5; + case DCN_1_0__SRCID__DC_D6_OTG_VSTARTUP: + return DC_IRQ_SOURCE_VBLANK6; + case DCN_1_0__SRCID__HUBP0_FLIP_INTERRUPT: + return DC_IRQ_SOURCE_PFLIP1; + case DCN_1_0__SRCID__HUBP1_FLIP_INTERRUPT: + return DC_IRQ_SOURCE_PFLIP2; + case DCN_1_0__SRCID__HUBP2_FLIP_INTERRUPT: + return DC_IRQ_SOURCE_PFLIP3; + case DCN_1_0__SRCID__HUBP3_FLIP_INTERRUPT: + return DC_IRQ_SOURCE_PFLIP4; + case DCN_1_0__SRCID__HUBP4_FLIP_INTERRUPT: + return DC_IRQ_SOURCE_PFLIP5; + case DCN_1_0__SRCID__HUBP5_FLIP_INTERRUPT: + return DC_IRQ_SOURCE_PFLIP6; + case DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE1; + case DCN_1_0__SRCID__OTG1_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE2; + case DCN_1_0__SRCID__OTG2_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE3; + case DCN_1_0__SRCID__OTG3_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE4; + case DCN_1_0__SRCID__OTG4_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE5; + case DCN_1_0__SRCID__OTG5_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE6; + + case DCN_1_0__SRCID__DC_HPD1_INT: + /* generic src_id for all HPD and HPDRX interrupts */ + switch (ext_id) { + case DCN_1_0__CTXID__DC_HPD1_INT: + return DC_IRQ_SOURCE_HPD1; + case DCN_1_0__CTXID__DC_HPD2_INT: + return DC_IRQ_SOURCE_HPD2; + case DCN_1_0__CTXID__DC_HPD3_INT: + return DC_IRQ_SOURCE_HPD3; + case DCN_1_0__CTXID__DC_HPD4_INT: + return DC_IRQ_SOURCE_HPD4; + case DCN_1_0__CTXID__DC_HPD5_INT: + return DC_IRQ_SOURCE_HPD5; + case DCN_1_0__CTXID__DC_HPD6_INT: + return DC_IRQ_SOURCE_HPD6; + case DCN_1_0__CTXID__DC_HPD1_RX_INT: + return DC_IRQ_SOURCE_HPD1RX; + case DCN_1_0__CTXID__DC_HPD2_RX_INT: + return DC_IRQ_SOURCE_HPD2RX; + case DCN_1_0__CTXID__DC_HPD3_RX_INT: + return DC_IRQ_SOURCE_HPD3RX; + case DCN_1_0__CTXID__DC_HPD4_RX_INT: + return DC_IRQ_SOURCE_HPD4RX; + case DCN_1_0__CTXID__DC_HPD5_RX_INT: + return DC_IRQ_SOURCE_HPD5RX; + case DCN_1_0__CTXID__DC_HPD6_RX_INT: + return DC_IRQ_SOURCE_HPD6RX; + default: + return DC_IRQ_SOURCE_INVALID; + } + break; + + default: + return DC_IRQ_SOURCE_INVALID; + } +} + +static bool hpd_ack( + struct irq_service *irq_service, + const struct irq_source_info *info) +{ + uint32_t addr = info->status_reg; + uint32_t value = dm_read_reg(irq_service->ctx, addr); + uint32_t current_status = + get_reg_field_value( + value, + HPD0_DC_HPD_INT_STATUS, + DC_HPD_SENSE_DELAYED); + + dal_irq_service_ack_generic(irq_service, info); + + value = dm_read_reg(irq_service->ctx, info->enable_reg); + + set_reg_field_value( + value, + current_status ? 0 : 1, + HPD0_DC_HPD_INT_CONTROL, + DC_HPD_INT_POLARITY); + + dm_write_reg(irq_service->ctx, info->enable_reg, value); + + return true; +} + +static const struct irq_source_info_funcs hpd_irq_info_funcs = { + .set = NULL, + .ack = hpd_ack +}; + +static const struct irq_source_info_funcs hpd_rx_irq_info_funcs = { + .set = NULL, + .ack = NULL +}; + +static const struct irq_source_info_funcs pflip_irq_info_funcs = { + .set = NULL, + .ack = NULL +}; + +static const struct irq_source_info_funcs vblank_irq_info_funcs = { + .set = NULL, + .ack = NULL +}; + +#undef BASE_INNER +#define BASE_INNER(seg) DCN_BASE__INST0_SEG ## seg + +/* compile time expand base address. */ +#define BASE(seg) \ + BASE_INNER(seg) + + +#define SRI(reg_name, block, id)\ + BASE(mm ## block ## id ## _ ## reg_name ## _BASE_IDX) + \ + mm ## block ## id ## _ ## reg_name + + +#define IRQ_REG_ENTRY(block, reg_num, reg1, mask1, reg2, mask2)\ + .enable_reg = SRI(reg1, block, reg_num),\ + .enable_mask = \ + block ## reg_num ## _ ## reg1 ## __ ## mask1 ## _MASK,\ + .enable_value = {\ + block ## reg_num ## _ ## reg1 ## __ ## mask1 ## _MASK,\ + ~block ## reg_num ## _ ## reg1 ## __ ## mask1 ## _MASK \ + },\ + .ack_reg = SRI(reg2, block, reg_num),\ + .ack_mask = \ + block ## reg_num ## _ ## reg2 ## __ ## mask2 ## _MASK,\ + .ack_value = \ + block ## reg_num ## _ ## reg2 ## __ ## mask2 ## _MASK \ + + + +#define hpd_int_entry(reg_num)\ + [DC_IRQ_SOURCE_HPD1 + reg_num] = {\ + IRQ_REG_ENTRY(HPD, reg_num,\ + DC_HPD_INT_CONTROL, DC_HPD_INT_EN,\ + DC_HPD_INT_CONTROL, DC_HPD_INT_ACK),\ + .status_reg = SRI(DC_HPD_INT_STATUS, HPD, reg_num),\ + .funcs = &hpd_irq_info_funcs\ + } + +#define hpd_rx_int_entry(reg_num)\ + [DC_IRQ_SOURCE_HPD1RX + reg_num] = {\ + IRQ_REG_ENTRY(HPD, reg_num,\ + DC_HPD_INT_CONTROL, DC_HPD_RX_INT_EN,\ + DC_HPD_INT_CONTROL, DC_HPD_RX_INT_ACK),\ + .status_reg = SRI(DC_HPD_INT_STATUS, HPD, reg_num),\ + .funcs = &hpd_rx_irq_info_funcs\ + } +#define pflip_int_entry(reg_num)\ + [DC_IRQ_SOURCE_PFLIP1 + reg_num] = {\ + IRQ_REG_ENTRY(HUBPREQ, reg_num,\ + DCSURF_SURFACE_FLIP_INTERRUPT, SURFACE_FLIP_INT_MASK,\ + DCSURF_SURFACE_FLIP_INTERRUPT, SURFACE_FLIP_CLEAR),\ + .funcs = &pflip_irq_info_funcs\ + } + +#define vupdate_int_entry(reg_num)\ + [DC_IRQ_SOURCE_VUPDATE1 + reg_num] = {\ + IRQ_REG_ENTRY(OTG, reg_num,\ + OTG_GLOBAL_SYNC_STATUS, VUPDATE_INT_EN,\ + OTG_GLOBAL_SYNC_STATUS, VUPDATE_EVENT_CLEAR),\ + .funcs = &vblank_irq_info_funcs\ + } + +#define vblank_int_entry(reg_num)\ + [DC_IRQ_SOURCE_VBLANK1 + reg_num] = {\ + IRQ_REG_ENTRY(OTG, reg_num,\ + OTG_GLOBAL_SYNC_STATUS, VSTARTUP_INT_EN,\ + OTG_GLOBAL_SYNC_STATUS, VSTARTUP_EVENT_CLEAR),\ + .funcs = &vblank_irq_info_funcs\ + } + +#define dummy_irq_entry() \ + {\ + .funcs = &dummy_irq_info_funcs\ + } + +#define i2c_int_entry(reg_num) \ + [DC_IRQ_SOURCE_I2C_DDC ## reg_num] = dummy_irq_entry() + +#define dp_sink_int_entry(reg_num) \ + [DC_IRQ_SOURCE_DPSINK ## reg_num] = dummy_irq_entry() + +#define gpio_pad_int_entry(reg_num) \ + [DC_IRQ_SOURCE_GPIOPAD ## reg_num] = dummy_irq_entry() + +#define dc_underflow_int_entry(reg_num) \ + [DC_IRQ_SOURCE_DC ## reg_num ## UNDERFLOW] = dummy_irq_entry() + +static const struct irq_source_info_funcs dummy_irq_info_funcs = { + .set = dal_irq_service_dummy_set, + .ack = dal_irq_service_dummy_ack +}; + +static const struct irq_source_info +irq_source_info_dcn20[DAL_IRQ_SOURCES_NUMBER] = { + [DC_IRQ_SOURCE_INVALID] = dummy_irq_entry(), + hpd_int_entry(0), + hpd_int_entry(1), + hpd_int_entry(2), + hpd_int_entry(3), + hpd_int_entry(4), + hpd_int_entry(5), + hpd_rx_int_entry(0), + hpd_rx_int_entry(1), + hpd_rx_int_entry(2), + hpd_rx_int_entry(3), + hpd_rx_int_entry(4), + hpd_rx_int_entry(5), + i2c_int_entry(1), + i2c_int_entry(2), + i2c_int_entry(3), + i2c_int_entry(4), + i2c_int_entry(5), + i2c_int_entry(6), + dp_sink_int_entry(1), + dp_sink_int_entry(2), + dp_sink_int_entry(3), + dp_sink_int_entry(4), + dp_sink_int_entry(5), + dp_sink_int_entry(6), + [DC_IRQ_SOURCE_TIMER] = dummy_irq_entry(), + pflip_int_entry(0), + pflip_int_entry(1), + pflip_int_entry(2), + pflip_int_entry(3), + [DC_IRQ_SOURCE_PFLIP5] = dummy_irq_entry(), + [DC_IRQ_SOURCE_PFLIP6] = dummy_irq_entry(), + [DC_IRQ_SOURCE_PFLIP_UNDERLAY0] = dummy_irq_entry(), + gpio_pad_int_entry(0), + gpio_pad_int_entry(1), + gpio_pad_int_entry(2), + gpio_pad_int_entry(3), + gpio_pad_int_entry(4), + gpio_pad_int_entry(5), + gpio_pad_int_entry(6), + gpio_pad_int_entry(7), + gpio_pad_int_entry(8), + gpio_pad_int_entry(9), + gpio_pad_int_entry(10), + gpio_pad_int_entry(11), + gpio_pad_int_entry(12), + gpio_pad_int_entry(13), + gpio_pad_int_entry(14), + gpio_pad_int_entry(15), + gpio_pad_int_entry(16), + gpio_pad_int_entry(17), + gpio_pad_int_entry(18), + gpio_pad_int_entry(19), + gpio_pad_int_entry(20), + gpio_pad_int_entry(21), + gpio_pad_int_entry(22), + gpio_pad_int_entry(23), + gpio_pad_int_entry(24), + gpio_pad_int_entry(25), + gpio_pad_int_entry(26), + gpio_pad_int_entry(27), + gpio_pad_int_entry(28), + gpio_pad_int_entry(29), + gpio_pad_int_entry(30), + dc_underflow_int_entry(1), + dc_underflow_int_entry(2), + dc_underflow_int_entry(3), + dc_underflow_int_entry(4), + dc_underflow_int_entry(5), + dc_underflow_int_entry(6), + [DC_IRQ_SOURCE_DMCU_SCP] = dummy_irq_entry(), + [DC_IRQ_SOURCE_VBIOS_SW] = dummy_irq_entry(), + vupdate_int_entry(0), + vupdate_int_entry(1), + vupdate_int_entry(2), + vupdate_int_entry(3), + vupdate_int_entry(4), + vupdate_int_entry(5), + vblank_int_entry(0), + vblank_int_entry(1), + vblank_int_entry(2), + vblank_int_entry(3), + vblank_int_entry(4), + vblank_int_entry(5), +}; + +static const struct irq_service_funcs irq_service_funcs_dcn20 = { + .to_dal_irq_source = to_dal_irq_source_dcn20 +}; + +static void construct( + struct irq_service *irq_service, + struct irq_service_init_data *init_data) +{ + dal_irq_service_construct(irq_service, init_data); + + irq_service->info = irq_source_info_dcn20; + irq_service->funcs = &irq_service_funcs_dcn20; +} + +struct irq_service *dal_irq_service_dcn20_create( + struct irq_service_init_data *init_data) +{ + struct irq_service *irq_service = kzalloc(sizeof(*irq_service), + GFP_KERNEL); + + if (!irq_service) + return NULL; + + construct(irq_service, init_data); + return irq_service; +} diff --cc drivers/gpu/drm/amd/powerplay/smu_v11_0.c index 5237246f89f75,463275f88e89e..f63a2becb7770 --- a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c +++ b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c @@@ -20,8 -20,10 +20,11 @@@ * OTHER DEALINGS IN THE SOFTWARE. */ - #include "pp_debug.h" #include + #include ++#include + + #include "pp_debug.h" #include "amdgpu.h" #include "amdgpu_smu.h" #include "atomfirmware.h"