/* * Copyright 2020 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 #include #include "amdgpu.h" #include "amdgpu_ucode.h" #include "amdgpu_trace.h" #include "gc/gc_11_0_0_offset.h" #include "gc/gc_11_0_0_sh_mask.h" #include "gc/gc_11_0_0_default.h" #include "hdp/hdp_6_0_0_offset.h" #include "ivsrcid/gfx/irqsrcs_gfx_11_0_0.h" #include "soc15_common.h" #include "soc15.h" #include "sdma_v6_0_0_pkt_open.h" #include "nbio_v4_3.h" #include "sdma_common.h" #include "sdma_v6_0.h" #include "v11_structs.h" MODULE_FIRMWARE("amdgpu/sdma_6_0_0.bin"); MODULE_FIRMWARE("amdgpu/sdma_6_0_1.bin"); MODULE_FIRMWARE("amdgpu/sdma_6_0_2.bin"); #define SDMA1_REG_OFFSET 0x600 #define SDMA0_HYP_DEC_REG_START 0x5880 #define SDMA0_HYP_DEC_REG_END 0x589a #define SDMA1_HYP_DEC_REG_OFFSET 0x20 static void sdma_v6_0_set_ring_funcs(struct amdgpu_device *adev); static void sdma_v6_0_set_buffer_funcs(struct amdgpu_device *adev); static void sdma_v6_0_set_vm_pte_funcs(struct amdgpu_device *adev); static void sdma_v6_0_set_irq_funcs(struct amdgpu_device *adev); static int sdma_v6_0_start(struct amdgpu_device *adev); static u32 sdma_v6_0_get_reg_offset(struct amdgpu_device *adev, u32 instance, u32 internal_offset) { u32 base; if (internal_offset >= SDMA0_HYP_DEC_REG_START && internal_offset <= SDMA0_HYP_DEC_REG_END) { base = adev->reg_offset[GC_HWIP][0][1]; if (instance != 0) internal_offset += SDMA1_HYP_DEC_REG_OFFSET * instance; } else { base = adev->reg_offset[GC_HWIP][0][0]; if (instance == 1) internal_offset += SDMA1_REG_OFFSET; } return base + internal_offset; } static int sdma_v6_0_init_inst_ctx(struct amdgpu_sdma_instance *sdma_inst) { int err = 0; const struct sdma_firmware_header_v2_0 *hdr; err = amdgpu_ucode_validate(sdma_inst->fw); if (err) return err; hdr = (const struct sdma_firmware_header_v2_0 *)sdma_inst->fw->data; sdma_inst->fw_version = le32_to_cpu(hdr->header.ucode_version); sdma_inst->feature_version = le32_to_cpu(hdr->ucode_feature_version); if (sdma_inst->feature_version >= 20) sdma_inst->burst_nop = true; return 0; } static void sdma_v6_0_destroy_inst_ctx(struct amdgpu_device *adev) { release_firmware(adev->sdma.instance[0].fw); memset((void*)adev->sdma.instance, 0, sizeof(struct amdgpu_sdma_instance) * AMDGPU_MAX_SDMA_INSTANCES); } /** * sdma_v6_0_init_microcode - load ucode images from disk * * @adev: amdgpu_device pointer * * Use the firmware interface to load the ucode images into * the driver (not loaded into hw). * Returns 0 on success, error on failure. */ // emulation only, won't work on real chip // sdma 6.0.0 real chip need to use PSP to load firmware static int sdma_v6_0_init_microcode(struct amdgpu_device *adev) { char fw_name[30]; char ucode_prefix[30]; int err = 0, i; struct amdgpu_firmware_info *info = NULL; const struct sdma_firmware_header_v2_0 *sdma_hdr; DRM_DEBUG("\n"); amdgpu_ucode_ip_version_decode(adev, SDMA0_HWIP, ucode_prefix, sizeof(ucode_prefix)); snprintf(fw_name, sizeof(fw_name), "amdgpu/%s.bin", ucode_prefix); err = request_firmware(&adev->sdma.instance[0].fw, fw_name, adev->dev); if (err) goto out; err = sdma_v6_0_init_inst_ctx(&adev->sdma.instance[0]); if (err) goto out; for (i = 1; i < adev->sdma.num_instances; i++) { memcpy((void*)&adev->sdma.instance[i], (void*)&adev->sdma.instance[0], sizeof(struct amdgpu_sdma_instance)); } DRM_DEBUG("psp_load == '%s'\n", adev->firmware.load_type == AMDGPU_FW_LOAD_PSP ? "true" : "false"); if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { sdma_hdr = (const struct sdma_firmware_header_v2_0 *)adev->sdma.instance[0].fw->data; info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA_UCODE_TH0]; info->ucode_id = AMDGPU_UCODE_ID_SDMA_UCODE_TH0; info->fw = adev->sdma.instance[0].fw; adev->firmware.fw_size += ALIGN(le32_to_cpu(sdma_hdr->ctx_ucode_size_bytes), PAGE_SIZE); info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA_UCODE_TH1]; info->ucode_id = AMDGPU_UCODE_ID_SDMA_UCODE_TH1; info->fw = adev->sdma.instance[0].fw; adev->firmware.fw_size += ALIGN(le32_to_cpu(sdma_hdr->ctl_ucode_size_bytes), PAGE_SIZE); } out: if (err) { DRM_ERROR("sdma_v6_0: Failed to load firmware \"%s\"\n", fw_name); sdma_v6_0_destroy_inst_ctx(adev); } return err; } static unsigned sdma_v6_0_ring_init_cond_exec(struct amdgpu_ring *ring) { unsigned ret; amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_COND_EXE)); amdgpu_ring_write(ring, lower_32_bits(ring->cond_exe_gpu_addr)); amdgpu_ring_write(ring, upper_32_bits(ring->cond_exe_gpu_addr)); amdgpu_ring_write(ring, 1); ret = ring->wptr & ring->buf_mask;/* this is the offset we need patch later */ amdgpu_ring_write(ring, 0x55aa55aa);/* insert dummy here and patch it later */ return ret; } static void sdma_v6_0_ring_patch_cond_exec(struct amdgpu_ring *ring, unsigned offset) { unsigned cur; BUG_ON(offset > ring->buf_mask); BUG_ON(ring->ring[offset] != 0x55aa55aa); cur = (ring->wptr - 1) & ring->buf_mask; if (cur > offset) ring->ring[offset] = cur - offset; else ring->ring[offset] = (ring->buf_mask + 1) - offset + cur; } /** * sdma_v6_0_ring_get_rptr - get the current read pointer * * @ring: amdgpu ring pointer * * Get the current rptr from the hardware. */ static uint64_t sdma_v6_0_ring_get_rptr(struct amdgpu_ring *ring) { u64 *rptr; /* XXX check if swapping is necessary on BE */ rptr = (u64 *)ring->rptr_cpu_addr; DRM_DEBUG("rptr before shift == 0x%016llx\n", *rptr); return ((*rptr) >> 2); } /** * sdma_v6_0_ring_get_wptr - get the current write pointer * * @ring: amdgpu ring pointer * * Get the current wptr from the hardware. */ static uint64_t sdma_v6_0_ring_get_wptr(struct amdgpu_ring *ring) { u64 wptr = 0; if (ring->use_doorbell) { /* XXX check if swapping is necessary on BE */ wptr = READ_ONCE(*((u64 *)ring->wptr_cpu_addr)); DRM_DEBUG("wptr/doorbell before shift == 0x%016llx\n", wptr); } return wptr >> 2; } /** * sdma_v6_0_ring_set_wptr - commit the write pointer * * @ring: amdgpu ring pointer * * Write the wptr back to the hardware. */ static void sdma_v6_0_ring_set_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; uint32_t *wptr_saved; uint32_t *is_queue_unmap; uint64_t aggregated_db_index; uint32_t mqd_size = adev->mqds[AMDGPU_HW_IP_DMA].mqd_size; DRM_DEBUG("Setting write pointer\n"); if (ring->is_mes_queue) { wptr_saved = (uint32_t *)(ring->mqd_ptr + mqd_size); is_queue_unmap = (uint32_t *)(ring->mqd_ptr + mqd_size + sizeof(uint32_t)); aggregated_db_index = amdgpu_mes_get_aggregated_doorbell_index(adev, ring->hw_prio); atomic64_set((atomic64_t *)ring->wptr_cpu_addr, ring->wptr << 2); *wptr_saved = ring->wptr << 2; if (*is_queue_unmap) { WDOORBELL64(aggregated_db_index, ring->wptr << 2); DRM_DEBUG("calling WDOORBELL64(0x%08x, 0x%016llx)\n", ring->doorbell_index, ring->wptr << 2); WDOORBELL64(ring->doorbell_index, ring->wptr << 2); } else { DRM_DEBUG("calling WDOORBELL64(0x%08x, 0x%016llx)\n", ring->doorbell_index, ring->wptr << 2); WDOORBELL64(ring->doorbell_index, ring->wptr << 2); if (*is_queue_unmap) WDOORBELL64(aggregated_db_index, ring->wptr << 2); } } else { if (ring->use_doorbell) { DRM_DEBUG("Using doorbell -- " "wptr_offs == 0x%08x " "lower_32_bits(ring->wptr) << 2 == 0x%08x " "upper_32_bits(ring->wptr) << 2 == 0x%08x\n", ring->wptr_offs, lower_32_bits(ring->wptr << 2), upper_32_bits(ring->wptr << 2)); /* XXX check if swapping is necessary on BE */ atomic64_set((atomic64_t *)ring->wptr_cpu_addr, ring->wptr << 2); DRM_DEBUG("calling WDOORBELL64(0x%08x, 0x%016llx)\n", ring->doorbell_index, ring->wptr << 2); WDOORBELL64(ring->doorbell_index, ring->wptr << 2); } else { DRM_DEBUG("Not using doorbell -- " "regSDMA%i_GFX_RB_WPTR == 0x%08x " "regSDMA%i_GFX_RB_WPTR_HI == 0x%08x\n", ring->me, lower_32_bits(ring->wptr << 2), ring->me, upper_32_bits(ring->wptr << 2)); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, ring->me, regSDMA0_QUEUE0_RB_WPTR), lower_32_bits(ring->wptr << 2)); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, ring->me, regSDMA0_QUEUE0_RB_WPTR_HI), upper_32_bits(ring->wptr << 2)); } } } static void sdma_v6_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count) { struct amdgpu_sdma_instance *sdma = amdgpu_sdma_get_instance_from_ring(ring); int i; for (i = 0; i < count; i++) if (sdma && sdma->burst_nop && (i == 0)) amdgpu_ring_write(ring, ring->funcs->nop | SDMA_PKT_NOP_HEADER_COUNT(count - 1)); else amdgpu_ring_write(ring, ring->funcs->nop); } /** * sdma_v6_0_ring_emit_ib - Schedule an IB on the DMA engine * * @ring: amdgpu ring pointer * @ib: IB object to schedule * * Schedule an IB in the DMA ring. */ static void sdma_v6_0_ring_emit_ib(struct amdgpu_ring *ring, struct amdgpu_job *job, struct amdgpu_ib *ib, uint32_t flags) { unsigned vmid = AMDGPU_JOB_GET_VMID(job); uint64_t csa_mc_addr = amdgpu_sdma_get_csa_mc_addr(ring, vmid); /* An IB packet must end on a 8 DW boundary--the next dword * must be on a 8-dword boundary. Our IB packet below is 6 * dwords long, thus add x number of NOPs, such that, in * modular arithmetic, * wptr + 6 + x = 8k, k >= 0, which in C is, * (wptr + 6 + x) % 8 = 0. * The expression below, is a solution of x. */ sdma_v6_0_ring_insert_nop(ring, (2 - lower_32_bits(ring->wptr)) & 7); amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_INDIRECT) | SDMA_PKT_INDIRECT_HEADER_VMID(vmid & 0xf)); /* base must be 32 byte aligned */ amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr) & 0xffffffe0); amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr)); amdgpu_ring_write(ring, ib->length_dw); amdgpu_ring_write(ring, lower_32_bits(csa_mc_addr)); amdgpu_ring_write(ring, upper_32_bits(csa_mc_addr)); } /** * sdma_v6_0_ring_emit_mem_sync - flush the IB by graphics cache rinse * * @ring: amdgpu ring pointer * @job: job to retrieve vmid from * @ib: IB object to schedule * * flush the IB by graphics cache rinse. */ static void sdma_v6_0_ring_emit_mem_sync(struct amdgpu_ring *ring) { uint32_t gcr_cntl = SDMA_GCR_GL2_INV | SDMA_GCR_GL2_WB | SDMA_GCR_GLM_INV | SDMA_GCR_GL1_INV | SDMA_GCR_GLV_INV | SDMA_GCR_GLK_INV | SDMA_GCR_GLI_INV(1); /* flush entire cache L0/L1/L2, this can be optimized by performance requirement */ amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_GCR_REQ)); amdgpu_ring_write(ring, SDMA_PKT_GCR_REQ_PAYLOAD1_BASE_VA_31_7(0)); amdgpu_ring_write(ring, SDMA_PKT_GCR_REQ_PAYLOAD2_GCR_CONTROL_15_0(gcr_cntl) | SDMA_PKT_GCR_REQ_PAYLOAD2_BASE_VA_47_32(0)); amdgpu_ring_write(ring, SDMA_PKT_GCR_REQ_PAYLOAD3_LIMIT_VA_31_7(0) | SDMA_PKT_GCR_REQ_PAYLOAD3_GCR_CONTROL_18_16(gcr_cntl >> 16)); amdgpu_ring_write(ring, SDMA_PKT_GCR_REQ_PAYLOAD4_LIMIT_VA_47_32(0) | SDMA_PKT_GCR_REQ_PAYLOAD4_VMID(0)); } /** * sdma_v6_0_ring_emit_hdp_flush - emit an hdp flush on the DMA ring * * @ring: amdgpu ring pointer * * Emit an hdp flush packet on the requested DMA ring. */ static void sdma_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; u32 ref_and_mask = 0; const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio.hdp_flush_reg; ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0 << ring->me; amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_POLL_REGMEM) | SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(1) | SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */ amdgpu_ring_write(ring, (adev->nbio.funcs->get_hdp_flush_done_offset(adev)) << 2); amdgpu_ring_write(ring, (adev->nbio.funcs->get_hdp_flush_req_offset(adev)) << 2); amdgpu_ring_write(ring, ref_and_mask); /* reference */ amdgpu_ring_write(ring, ref_and_mask); /* mask */ amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ } /** * sdma_v6_0_ring_emit_fence - emit a fence on the DMA ring * * @ring: amdgpu ring pointer * @fence: amdgpu fence object * * Add a DMA fence packet to the ring to write * the fence seq number and DMA trap packet to generate * an interrupt if needed. */ static void sdma_v6_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq, unsigned flags) { bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT; /* write the fence */ amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_FENCE) | SDMA_PKT_FENCE_HEADER_MTYPE(0x3)); /* Ucached(UC) */ /* zero in first two bits */ BUG_ON(addr & 0x3); amdgpu_ring_write(ring, lower_32_bits(addr)); amdgpu_ring_write(ring, upper_32_bits(addr)); amdgpu_ring_write(ring, lower_32_bits(seq)); /* optionally write high bits as well */ if (write64bit) { addr += 4; amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_FENCE) | SDMA_PKT_FENCE_HEADER_MTYPE(0x3)); /* zero in first two bits */ BUG_ON(addr & 0x3); amdgpu_ring_write(ring, lower_32_bits(addr)); amdgpu_ring_write(ring, upper_32_bits(addr)); amdgpu_ring_write(ring, upper_32_bits(seq)); } if (flags & AMDGPU_FENCE_FLAG_INT) { uint32_t ctx = ring->is_mes_queue ? (ring->hw_queue_id | AMDGPU_FENCE_MES_QUEUE_FLAG) : 0; /* generate an interrupt */ amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_TRAP)); amdgpu_ring_write(ring, SDMA_PKT_TRAP_INT_CONTEXT_INT_CONTEXT(ctx)); } } /** * sdma_v6_0_gfx_stop - stop the gfx async dma engines * * @adev: amdgpu_device pointer * * Stop the gfx async dma ring buffers. */ static void sdma_v6_0_gfx_stop(struct amdgpu_device *adev) { struct amdgpu_ring *sdma0 = &adev->sdma.instance[0].ring; struct amdgpu_ring *sdma1 = &adev->sdma.instance[1].ring; u32 rb_cntl, ib_cntl; int i; if ((adev->mman.buffer_funcs_ring == sdma0) || (adev->mman.buffer_funcs_ring == sdma1)) amdgpu_ttm_set_buffer_funcs_status(adev, false); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL)); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_ENABLE, 0); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL), rb_cntl); ib_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL)); ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_QUEUE0_IB_CNTL, IB_ENABLE, 0); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL), ib_cntl); } sdma0->sched.ready = false; sdma1->sched.ready = false; } /** * sdma_v6_0_rlc_stop - stop the compute async dma engines * * @adev: amdgpu_device pointer * * Stop the compute async dma queues. */ static void sdma_v6_0_rlc_stop(struct amdgpu_device *adev) { /* XXX todo */ } /** * sdma_v6_0_ctx_switch_enable - stop the async dma engines context switch * * @adev: amdgpu_device pointer * @enable: enable/disable the DMA MEs context switch. * * Halt or unhalt the async dma engines context switch. */ static void sdma_v6_0_ctx_switch_enable(struct amdgpu_device *adev, bool enable) { } /** * sdma_v6_0_enable - stop the async dma engines * * @adev: amdgpu_device pointer * @enable: enable/disable the DMA MEs. * * Halt or unhalt the async dma engines. */ static void sdma_v6_0_enable(struct amdgpu_device *adev, bool enable) { u32 f32_cntl; int i; if (!enable) { sdma_v6_0_gfx_stop(adev); sdma_v6_0_rlc_stop(adev); } for (i = 0; i < adev->sdma.num_instances; i++) { f32_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL)); f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_F32_CNTL, HALT, enable ? 0 : 1); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL), f32_cntl); } } /** * sdma_v6_0_gfx_resume - setup and start the async dma engines * * @adev: amdgpu_device pointer * * Set up the gfx DMA ring buffers and enable them. * Returns 0 for success, error for failure. */ static int sdma_v6_0_gfx_resume(struct amdgpu_device *adev) { struct amdgpu_ring *ring; u32 rb_cntl, ib_cntl; u32 rb_bufsz; u32 doorbell; u32 doorbell_offset; u32 temp; u64 wptr_gpu_addr; int i, r; for (i = 0; i < adev->sdma.num_instances; i++) { ring = &adev->sdma.instance[i].ring; WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); /* Set ring buffer size in dwords */ rb_bufsz = order_base_2(ring->ring_size / 4); rb_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL)); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_SIZE, rb_bufsz); #ifdef __BIG_ENDIAN rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_SWAP_ENABLE, 1); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RPTR_WRITEBACK_SWAP_ENABLE, 1); #endif rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_PRIV, 1); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL), rb_cntl); /* Initialize the ring buffer's read and write pointers */ WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR), 0); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_HI), 0); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR), 0); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_HI), 0); /* setup the wptr shadow polling */ wptr_gpu_addr = ring->wptr_gpu_addr; WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_POLL_ADDR_LO), lower_32_bits(wptr_gpu_addr)); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_POLL_ADDR_HI), upper_32_bits(wptr_gpu_addr)); /* set the wb address whether it's enabled or not */ WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_ADDR_HI), upper_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFF); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_ADDR_LO), lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, WPTR_POLL_ENABLE, 0); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, F32_WPTR_POLL_ENABLE, 1); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_BASE), ring->gpu_addr >> 8); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_BASE_HI), ring->gpu_addr >> 40); ring->wptr = 0; /* before programing wptr to a less value, need set minor_ptr_update first */ WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_MINOR_PTR_UPDATE), 1); if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */ WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR), lower_32_bits(ring->wptr) << 2); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_HI), upper_32_bits(ring->wptr) << 2); } doorbell = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL)); doorbell_offset = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL_OFFSET)); if (ring->use_doorbell) { doorbell = REG_SET_FIELD(doorbell, SDMA0_QUEUE0_DOORBELL, ENABLE, 1); doorbell_offset = REG_SET_FIELD(doorbell_offset, SDMA0_QUEUE0_DOORBELL_OFFSET, OFFSET, ring->doorbell_index); } else { doorbell = REG_SET_FIELD(doorbell, SDMA0_QUEUE0_DOORBELL, ENABLE, 0); } WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL), doorbell); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL_OFFSET), doorbell_offset); if (i == 0) adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell, ring->doorbell_index, adev->doorbell_index.sdma_doorbell_range * adev->sdma.num_instances); if (amdgpu_sriov_vf(adev)) sdma_v6_0_ring_set_wptr(ring); /* set minor_ptr_update to 0 after wptr programed */ WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_MINOR_PTR_UPDATE), 0); /* Set up RESP_MODE to non-copy addresses */ temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_CNTL)); temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, REDO_DELAY, 9); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_CNTL), temp); /* program default cache read and write policy */ temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_PAGE)); /* clean read policy and write policy bits */ temp &= 0xFF0FFF; temp |= ((CACHE_READ_POLICY_L2__DEFAULT << 12) | (CACHE_WRITE_POLICY_L2__DEFAULT << 14) | SDMA0_UTCL1_PAGE__LLC_NOALLOC_MASK); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_PAGE), temp); if (!amdgpu_sriov_vf(adev)) { /* unhalt engine */ temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL)); temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0); temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, TH1_RESET, 0); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL), temp); } /* enable DMA RB */ rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_ENABLE, 1); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL), rb_cntl); ib_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL)); ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_QUEUE0_IB_CNTL, IB_ENABLE, 1); #ifdef __BIG_ENDIAN ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_QUEUE0_IB_CNTL, IB_SWAP_ENABLE, 1); #endif /* enable DMA IBs */ WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL), ib_cntl); ring->sched.ready = true; if (amdgpu_sriov_vf(adev)) { /* bare-metal sequence doesn't need below to lines */ sdma_v6_0_ctx_switch_enable(adev, true); sdma_v6_0_enable(adev, true); } r = amdgpu_ring_test_helper(ring); if (r) { ring->sched.ready = false; return r; } if (adev->mman.buffer_funcs_ring == ring) amdgpu_ttm_set_buffer_funcs_status(adev, true); } return 0; } /** * sdma_v6_0_rlc_resume - setup and start the async dma engines * * @adev: amdgpu_device pointer * * Set up the compute DMA queues and enable them. * Returns 0 for success, error for failure. */ static int sdma_v6_0_rlc_resume(struct amdgpu_device *adev) { return 0; } /** * sdma_v6_0_load_microcode - load the sDMA ME ucode * * @adev: amdgpu_device pointer * * Loads the sDMA0/1 ucode. * Returns 0 for success, -EINVAL if the ucode is not available. */ static int sdma_v6_0_load_microcode(struct amdgpu_device *adev) { const struct sdma_firmware_header_v2_0 *hdr; const __le32 *fw_data; u32 fw_size; int i, j; bool use_broadcast; /* halt the MEs */ sdma_v6_0_enable(adev, false); if (!adev->sdma.instance[0].fw) return -EINVAL; /* use broadcast mode to load SDMA microcode by default */ use_broadcast = true; if (use_broadcast) { dev_info(adev->dev, "Use broadcast method to load SDMA firmware\n"); /* load Control Thread microcode */ hdr = (const struct sdma_firmware_header_v2_0 *)adev->sdma.instance[0].fw->data; amdgpu_ucode_print_sdma_hdr(&hdr->header); fw_size = le32_to_cpu(hdr->ctx_jt_offset + hdr->ctx_jt_size) / 4; fw_data = (const __le32 *) (adev->sdma.instance[0].fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); WREG32(sdma_v6_0_get_reg_offset(adev, 0, regSDMA0_BROADCAST_UCODE_ADDR), 0); for (j = 0; j < fw_size; j++) { if (amdgpu_emu_mode == 1 && j % 500 == 0) msleep(1); WREG32(sdma_v6_0_get_reg_offset(adev, 0, regSDMA0_BROADCAST_UCODE_DATA), le32_to_cpup(fw_data++)); } /* load Context Switch microcode */ fw_size = le32_to_cpu(hdr->ctl_jt_offset + hdr->ctl_jt_size) / 4; fw_data = (const __le32 *) (adev->sdma.instance[0].fw->data + le32_to_cpu(hdr->ctl_ucode_offset)); WREG32(sdma_v6_0_get_reg_offset(adev, 0, regSDMA0_BROADCAST_UCODE_ADDR), 0x8000); for (j = 0; j < fw_size; j++) { if (amdgpu_emu_mode == 1 && j % 500 == 0) msleep(1); WREG32(sdma_v6_0_get_reg_offset(adev, 0, regSDMA0_BROADCAST_UCODE_DATA), le32_to_cpup(fw_data++)); } } else { dev_info(adev->dev, "Use legacy method to load SDMA firmware\n"); for (i = 0; i < adev->sdma.num_instances; i++) { /* load Control Thread microcode */ hdr = (const struct sdma_firmware_header_v2_0 *)adev->sdma.instance[0].fw->data; amdgpu_ucode_print_sdma_hdr(&hdr->header); fw_size = le32_to_cpu(hdr->ctx_jt_offset + hdr->ctx_jt_size) / 4; fw_data = (const __le32 *) (adev->sdma.instance[0].fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); WREG32(sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UCODE_ADDR), 0); for (j = 0; j < fw_size; j++) { if (amdgpu_emu_mode == 1 && j % 500 == 0) msleep(1); WREG32(sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UCODE_DATA), le32_to_cpup(fw_data++)); } WREG32(sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UCODE_ADDR), adev->sdma.instance[0].fw_version); /* load Context Switch microcode */ fw_size = le32_to_cpu(hdr->ctl_jt_offset + hdr->ctl_jt_size) / 4; fw_data = (const __le32 *) (adev->sdma.instance[0].fw->data + le32_to_cpu(hdr->ctl_ucode_offset)); WREG32(sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UCODE_ADDR), 0x8000); for (j = 0; j < fw_size; j++) { if (amdgpu_emu_mode == 1 && j % 500 == 0) msleep(1); WREG32(sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UCODE_DATA), le32_to_cpup(fw_data++)); } WREG32(sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UCODE_ADDR), adev->sdma.instance[0].fw_version); } } return 0; } static int sdma_v6_0_soft_reset(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; u32 tmp; int i; sdma_v6_0_gfx_stop(adev); for (i = 0; i < adev->sdma.num_instances; i++) { tmp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_FREEZE)); tmp |= SDMA0_FREEZE__FREEZE_MASK; WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_FREEZE), tmp); tmp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL)); tmp |= SDMA0_F32_CNTL__HALT_MASK; tmp |= SDMA0_F32_CNTL__TH1_RESET_MASK; WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL), tmp); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_PREEMPT), 0); udelay(100); tmp = GRBM_SOFT_RESET__SOFT_RESET_SDMA0_MASK << i; WREG32_SOC15(GC, 0, regGRBM_SOFT_RESET, tmp); tmp = RREG32_SOC15(GC, 0, regGRBM_SOFT_RESET); udelay(100); WREG32_SOC15(GC, 0, regGRBM_SOFT_RESET, 0); tmp = RREG32_SOC15(GC, 0, regGRBM_SOFT_RESET); udelay(100); } return sdma_v6_0_start(adev); } static bool sdma_v6_0_check_soft_reset(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; struct amdgpu_ring *ring; int i, r; long tmo = msecs_to_jiffies(1000); for (i = 0; i < adev->sdma.num_instances; i++) { ring = &adev->sdma.instance[i].ring; r = amdgpu_ring_test_ib(ring, tmo); if (r) return true; } return false; } /** * sdma_v6_0_start - setup and start the async dma engines * * @adev: amdgpu_device pointer * * Set up the DMA engines and enable them. * Returns 0 for success, error for failure. */ static int sdma_v6_0_start(struct amdgpu_device *adev) { int r = 0; if (amdgpu_sriov_vf(adev)) { sdma_v6_0_ctx_switch_enable(adev, false); sdma_v6_0_enable(adev, false); /* set RB registers */ r = sdma_v6_0_gfx_resume(adev); return r; } if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) { r = sdma_v6_0_load_microcode(adev); if (r) return r; /* The value of regSDMA_F32_CNTL is invalid the moment after loading fw */ if (amdgpu_emu_mode == 1) msleep(1000); } /* unhalt the MEs */ sdma_v6_0_enable(adev, true); /* enable sdma ring preemption */ sdma_v6_0_ctx_switch_enable(adev, true); /* start the gfx rings and rlc compute queues */ r = sdma_v6_0_gfx_resume(adev); if (r) return r; r = sdma_v6_0_rlc_resume(adev); return r; } static int sdma_v6_0_mqd_init(struct amdgpu_device *adev, void *mqd, struct amdgpu_mqd_prop *prop) { struct v11_sdma_mqd *m = mqd; uint64_t wb_gpu_addr; m->sdmax_rlcx_rb_cntl = order_base_2(prop->queue_size / 4) << SDMA0_QUEUE0_RB_CNTL__RB_SIZE__SHIFT | 1 << SDMA0_QUEUE0_RB_CNTL__RPTR_WRITEBACK_ENABLE__SHIFT | 4 << SDMA0_QUEUE0_RB_CNTL__RPTR_WRITEBACK_TIMER__SHIFT; m->sdmax_rlcx_rb_base = lower_32_bits(prop->hqd_base_gpu_addr >> 8); m->sdmax_rlcx_rb_base_hi = upper_32_bits(prop->hqd_base_gpu_addr >> 8); wb_gpu_addr = prop->wptr_gpu_addr; m->sdmax_rlcx_rb_wptr_poll_addr_lo = lower_32_bits(wb_gpu_addr); m->sdmax_rlcx_rb_wptr_poll_addr_hi = upper_32_bits(wb_gpu_addr); wb_gpu_addr = prop->rptr_gpu_addr; m->sdmax_rlcx_rb_rptr_addr_lo = lower_32_bits(wb_gpu_addr); m->sdmax_rlcx_rb_rptr_addr_hi = upper_32_bits(wb_gpu_addr); m->sdmax_rlcx_ib_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, 0, regSDMA0_QUEUE0_IB_CNTL)); m->sdmax_rlcx_doorbell_offset = prop->doorbell_index << SDMA0_QUEUE0_DOORBELL_OFFSET__OFFSET__SHIFT; m->sdmax_rlcx_doorbell = REG_SET_FIELD(0, SDMA0_QUEUE0_DOORBELL, ENABLE, 1); m->sdmax_rlcx_skip_cntl = 0; m->sdmax_rlcx_context_status = 0; m->sdmax_rlcx_doorbell_log = 0; m->sdmax_rlcx_rb_aql_cntl = regSDMA0_QUEUE0_RB_AQL_CNTL_DEFAULT; m->sdmax_rlcx_dummy_reg = regSDMA0_QUEUE0_DUMMY_REG_DEFAULT; return 0; } static void sdma_v6_0_set_mqd_funcs(struct amdgpu_device *adev) { adev->mqds[AMDGPU_HW_IP_DMA].mqd_size = sizeof(struct v11_sdma_mqd); adev->mqds[AMDGPU_HW_IP_DMA].init_mqd = sdma_v6_0_mqd_init; } /** * sdma_v6_0_ring_test_ring - simple async dma engine test * * @ring: amdgpu_ring structure holding ring information * * Test the DMA engine by writing using it to write an * value to memory. * Returns 0 for success, error for failure. */ static int sdma_v6_0_ring_test_ring(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; unsigned i; unsigned index; int r; u32 tmp; u64 gpu_addr; volatile uint32_t *cpu_ptr = NULL; tmp = 0xCAFEDEAD; if (ring->is_mes_queue) { uint32_t offset = 0; offset = amdgpu_mes_ctx_get_offs(ring, AMDGPU_MES_CTX_PADDING_OFFS); gpu_addr = amdgpu_mes_ctx_get_offs_gpu_addr(ring, offset); cpu_ptr = amdgpu_mes_ctx_get_offs_cpu_addr(ring, offset); *cpu_ptr = tmp; } else { r = amdgpu_device_wb_get(adev, &index); if (r) { dev_err(adev->dev, "(%d) failed to allocate wb slot\n", r); return r; } gpu_addr = adev->wb.gpu_addr + (index * 4); adev->wb.wb[index] = cpu_to_le32(tmp); } r = amdgpu_ring_alloc(ring, 5); if (r) { DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r); amdgpu_device_wb_free(adev, index); return r; } amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_WRITE) | SDMA_PKT_COPY_LINEAR_HEADER_SUB_OP(SDMA_SUBOP_WRITE_LINEAR)); amdgpu_ring_write(ring, lower_32_bits(gpu_addr)); amdgpu_ring_write(ring, upper_32_bits(gpu_addr)); amdgpu_ring_write(ring, SDMA_PKT_WRITE_UNTILED_DW_3_COUNT(0)); amdgpu_ring_write(ring, 0xDEADBEEF); amdgpu_ring_commit(ring); for (i = 0; i < adev->usec_timeout; i++) { if (ring->is_mes_queue) tmp = le32_to_cpu(*cpu_ptr); else tmp = le32_to_cpu(adev->wb.wb[index]); if (tmp == 0xDEADBEEF) break; if (amdgpu_emu_mode == 1) msleep(1); else udelay(1); } if (i >= adev->usec_timeout) r = -ETIMEDOUT; if (!ring->is_mes_queue) amdgpu_device_wb_free(adev, index); return r; } /** * sdma_v6_0_ring_test_ib - test an IB on the DMA engine * * @ring: amdgpu_ring structure holding ring information * * Test a simple IB in the DMA ring. * Returns 0 on success, error on failure. */ static int sdma_v6_0_ring_test_ib(struct amdgpu_ring *ring, long timeout) { struct amdgpu_device *adev = ring->adev; struct amdgpu_ib ib; struct dma_fence *f = NULL; unsigned index; long r; u32 tmp = 0; u64 gpu_addr; volatile uint32_t *cpu_ptr = NULL; tmp = 0xCAFEDEAD; memset(&ib, 0, sizeof(ib)); if (ring->is_mes_queue) { uint32_t offset = 0; offset = amdgpu_mes_ctx_get_offs(ring, AMDGPU_MES_CTX_IB_OFFS); ib.gpu_addr = amdgpu_mes_ctx_get_offs_gpu_addr(ring, offset); ib.ptr = (void *)amdgpu_mes_ctx_get_offs_cpu_addr(ring, offset); offset = amdgpu_mes_ctx_get_offs(ring, AMDGPU_MES_CTX_PADDING_OFFS); gpu_addr = amdgpu_mes_ctx_get_offs_gpu_addr(ring, offset); cpu_ptr = amdgpu_mes_ctx_get_offs_cpu_addr(ring, offset); *cpu_ptr = tmp; } else { r = amdgpu_device_wb_get(adev, &index); if (r) { dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r); return r; } gpu_addr = adev->wb.gpu_addr + (index * 4); adev->wb.wb[index] = cpu_to_le32(tmp); r = amdgpu_ib_get(adev, NULL, 256, AMDGPU_IB_POOL_DIRECT, &ib); if (r) { DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r); goto err0; } } ib.ptr[0] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_WRITE) | SDMA_PKT_COPY_LINEAR_HEADER_SUB_OP(SDMA_SUBOP_WRITE_LINEAR); ib.ptr[1] = lower_32_bits(gpu_addr); ib.ptr[2] = upper_32_bits(gpu_addr); ib.ptr[3] = SDMA_PKT_WRITE_UNTILED_DW_3_COUNT(0); ib.ptr[4] = 0xDEADBEEF; ib.ptr[5] = SDMA_PKT_NOP_HEADER_OP(SDMA_OP_NOP); ib.ptr[6] = SDMA_PKT_NOP_HEADER_OP(SDMA_OP_NOP); ib.ptr[7] = SDMA_PKT_NOP_HEADER_OP(SDMA_OP_NOP); ib.length_dw = 8; r = amdgpu_ib_schedule(ring, 1, &ib, NULL, &f); if (r) goto err1; r = dma_fence_wait_timeout(f, false, timeout); if (r == 0) { DRM_ERROR("amdgpu: IB test timed out\n"); r = -ETIMEDOUT; goto err1; } else if (r < 0) { DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); goto err1; } if (ring->is_mes_queue) tmp = le32_to_cpu(*cpu_ptr); else tmp = le32_to_cpu(adev->wb.wb[index]); if (tmp == 0xDEADBEEF) r = 0; else r = -EINVAL; err1: amdgpu_ib_free(adev, &ib, NULL); dma_fence_put(f); err0: if (!ring->is_mes_queue) amdgpu_device_wb_free(adev, index); return r; } /** * sdma_v6_0_vm_copy_pte - update PTEs by copying them from the GART * * @ib: indirect buffer to fill with commands * @pe: addr of the page entry * @src: src addr to copy from * @count: number of page entries to update * * Update PTEs by copying them from the GART using sDMA. */ static void sdma_v6_0_vm_copy_pte(struct amdgpu_ib *ib, uint64_t pe, uint64_t src, unsigned count) { unsigned bytes = count * 8; ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_COPY_LINEAR_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR); ib->ptr[ib->length_dw++] = bytes - 1; ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */ ib->ptr[ib->length_dw++] = lower_32_bits(src); ib->ptr[ib->length_dw++] = upper_32_bits(src); ib->ptr[ib->length_dw++] = lower_32_bits(pe); ib->ptr[ib->length_dw++] = upper_32_bits(pe); } /** * sdma_v6_0_vm_write_pte - update PTEs by writing them manually * * @ib: indirect buffer to fill with commands * @pe: addr of the page entry * @addr: dst addr to write into pe * @count: number of page entries to update * @incr: increase next addr by incr bytes * @flags: access flags * * Update PTEs by writing them manually using sDMA. */ static void sdma_v6_0_vm_write_pte(struct amdgpu_ib *ib, uint64_t pe, uint64_t value, unsigned count, uint32_t incr) { unsigned ndw = count * 2; ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_WRITE) | SDMA_PKT_COPY_LINEAR_HEADER_SUB_OP(SDMA_SUBOP_WRITE_LINEAR); ib->ptr[ib->length_dw++] = lower_32_bits(pe); ib->ptr[ib->length_dw++] = upper_32_bits(pe); ib->ptr[ib->length_dw++] = ndw - 1; for (; ndw > 0; ndw -= 2) { ib->ptr[ib->length_dw++] = lower_32_bits(value); ib->ptr[ib->length_dw++] = upper_32_bits(value); value += incr; } } /** * sdma_v6_0_vm_set_pte_pde - update the page tables using sDMA * * @ib: indirect buffer to fill with commands * @pe: addr of the page entry * @addr: dst addr to write into pe * @count: number of page entries to update * @incr: increase next addr by incr bytes * @flags: access flags * * Update the page tables using sDMA. */ static void sdma_v6_0_vm_set_pte_pde(struct amdgpu_ib *ib, uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, uint64_t flags) { /* for physically contiguous pages (vram) */ ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_PTEPDE); ib->ptr[ib->length_dw++] = lower_32_bits(pe); /* dst addr */ ib->ptr[ib->length_dw++] = upper_32_bits(pe); ib->ptr[ib->length_dw++] = lower_32_bits(flags); /* mask */ ib->ptr[ib->length_dw++] = upper_32_bits(flags); ib->ptr[ib->length_dw++] = lower_32_bits(addr); /* value */ ib->ptr[ib->length_dw++] = upper_32_bits(addr); ib->ptr[ib->length_dw++] = incr; /* increment size */ ib->ptr[ib->length_dw++] = 0; ib->ptr[ib->length_dw++] = count - 1; /* number of entries */ } /** * sdma_v6_0_ring_pad_ib - pad the IB * @ib: indirect buffer to fill with padding * * Pad the IB with NOPs to a boundary multiple of 8. */ static void sdma_v6_0_ring_pad_ib(struct amdgpu_ring *ring, struct amdgpu_ib *ib) { struct amdgpu_sdma_instance *sdma = amdgpu_sdma_get_instance_from_ring(ring); u32 pad_count; int i; pad_count = (-ib->length_dw) & 0x7; for (i = 0; i < pad_count; i++) if (sdma && sdma->burst_nop && (i == 0)) ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_NOP) | SDMA_PKT_NOP_HEADER_COUNT(pad_count - 1); else ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_NOP); } /** * sdma_v6_0_ring_emit_pipeline_sync - sync the pipeline * * @ring: amdgpu_ring pointer * * Make sure all previous operations are completed (CIK). */ static void sdma_v6_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) { uint32_t seq = ring->fence_drv.sync_seq; uint64_t addr = ring->fence_drv.gpu_addr; /* wait for idle */ amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_POLL_REGMEM) | SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) | SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3) | /* equal */ SDMA_PKT_POLL_REGMEM_HEADER_MEM_POLL(1)); amdgpu_ring_write(ring, addr & 0xfffffffc); amdgpu_ring_write(ring, upper_32_bits(addr) & 0xffffffff); amdgpu_ring_write(ring, seq); /* reference */ amdgpu_ring_write(ring, 0xffffffff); /* mask */ amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(4)); /* retry count, poll interval */ } /** * sdma_v6_0_ring_emit_vm_flush - vm flush using sDMA * * @ring: amdgpu_ring pointer * @vm: amdgpu_vm pointer * * Update the page table base and flush the VM TLB * using sDMA. */ static void sdma_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); } static void sdma_v6_0_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val) { amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_SRBM_WRITE) | SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); amdgpu_ring_write(ring, reg); amdgpu_ring_write(ring, val); } static void sdma_v6_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg, uint32_t val, uint32_t mask) { amdgpu_ring_write(ring, SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_POLL_REGMEM) | SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) | SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* equal */ amdgpu_ring_write(ring, reg << 2); amdgpu_ring_write(ring, 0); amdgpu_ring_write(ring, val); /* reference */ amdgpu_ring_write(ring, mask); /* mask */ amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); } static void sdma_v6_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring, uint32_t reg0, uint32_t reg1, uint32_t ref, uint32_t mask) { amdgpu_ring_emit_wreg(ring, reg0, ref); /* wait for a cycle to reset vm_inv_eng*_ack */ amdgpu_ring_emit_reg_wait(ring, reg0, 0, 0); amdgpu_ring_emit_reg_wait(ring, reg1, mask, mask); } static int sdma_v6_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; sdma_v6_0_set_ring_funcs(adev); sdma_v6_0_set_buffer_funcs(adev); sdma_v6_0_set_vm_pte_funcs(adev); sdma_v6_0_set_irq_funcs(adev); sdma_v6_0_set_mqd_funcs(adev); return 0; } static int sdma_v6_0_sw_init(void *handle) { struct amdgpu_ring *ring; int r, i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* SDMA trap event */ r = amdgpu_irq_add_id(adev, SOC21_IH_CLIENTID_GFX, GFX_11_0_0__SRCID__SDMA_TRAP, &adev->sdma.trap_irq); if (r) return r; r = sdma_v6_0_init_microcode(adev); if (r) { DRM_ERROR("Failed to load sdma firmware!\n"); return r; } for (i = 0; i < adev->sdma.num_instances; i++) { ring = &adev->sdma.instance[i].ring; ring->ring_obj = NULL; ring->use_doorbell = true; ring->me = i; DRM_DEBUG("SDMA %d use_doorbell being set to: [%s]\n", i, ring->use_doorbell?"true":"false"); ring->doorbell_index = (adev->doorbell_index.sdma_engine[i] << 1); // get DWORD offset sprintf(ring->name, "sdma%d", i); r = amdgpu_ring_init(adev, ring, 1024, &adev->sdma.trap_irq, AMDGPU_SDMA_IRQ_INSTANCE0 + i, AMDGPU_RING_PRIO_DEFAULT, NULL); if (r) return r; } return r; } static int sdma_v6_0_sw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int i; for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); sdma_v6_0_destroy_inst_ctx(adev); return 0; } static int sdma_v6_0_hw_init(void *handle) { int r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; r = sdma_v6_0_start(adev); return r; } static int sdma_v6_0_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (amdgpu_sriov_vf(adev)) return 0; sdma_v6_0_ctx_switch_enable(adev, false); sdma_v6_0_enable(adev, false); return 0; } static int sdma_v6_0_suspend(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; return sdma_v6_0_hw_fini(adev); } static int sdma_v6_0_resume(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; return sdma_v6_0_hw_init(adev); } static bool sdma_v6_0_is_idle(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; u32 i; for (i = 0; i < adev->sdma.num_instances; i++) { u32 tmp = RREG32(sdma_v6_0_get_reg_offset(adev, i, regSDMA0_STATUS_REG)); if (!(tmp & SDMA0_STATUS_REG__IDLE_MASK)) return false; } return true; } static int sdma_v6_0_wait_for_idle(void *handle) { unsigned i; u32 sdma0, sdma1; struct amdgpu_device *adev = (struct amdgpu_device *)handle; for (i = 0; i < adev->usec_timeout; i++) { sdma0 = RREG32(sdma_v6_0_get_reg_offset(adev, 0, regSDMA0_STATUS_REG)); sdma1 = RREG32(sdma_v6_0_get_reg_offset(adev, 1, regSDMA0_STATUS_REG)); if (sdma0 & sdma1 & SDMA0_STATUS_REG__IDLE_MASK) return 0; udelay(1); } return -ETIMEDOUT; } static int sdma_v6_0_ring_preempt_ib(struct amdgpu_ring *ring) { int i, r = 0; struct amdgpu_device *adev = ring->adev; u32 index = 0; u64 sdma_gfx_preempt; amdgpu_sdma_get_index_from_ring(ring, &index); sdma_gfx_preempt = sdma_v6_0_get_reg_offset(adev, index, regSDMA0_QUEUE0_PREEMPT); /* assert preemption condition */ amdgpu_ring_set_preempt_cond_exec(ring, false); /* emit the trailing fence */ ring->trail_seq += 1; amdgpu_ring_alloc(ring, 10); sdma_v6_0_ring_emit_fence(ring, ring->trail_fence_gpu_addr, ring->trail_seq, 0); amdgpu_ring_commit(ring); /* assert IB preemption */ WREG32(sdma_gfx_preempt, 1); /* poll the trailing fence */ for (i = 0; i < adev->usec_timeout; i++) { if (ring->trail_seq == le32_to_cpu(*(ring->trail_fence_cpu_addr))) break; udelay(1); } if (i >= adev->usec_timeout) { r = -EINVAL; DRM_ERROR("ring %d failed to be preempted\n", ring->idx); } /* deassert IB preemption */ WREG32(sdma_gfx_preempt, 0); /* deassert the preemption condition */ amdgpu_ring_set_preempt_cond_exec(ring, true); return r; } static int sdma_v6_0_set_trap_irq_state(struct amdgpu_device *adev, struct amdgpu_irq_src *source, unsigned type, enum amdgpu_interrupt_state state) { u32 sdma_cntl; u32 reg_offset = sdma_v6_0_get_reg_offset(adev, type, regSDMA0_CNTL); sdma_cntl = RREG32(reg_offset); sdma_cntl = REG_SET_FIELD(sdma_cntl, SDMA0_CNTL, TRAP_ENABLE, state == AMDGPU_IRQ_STATE_ENABLE ? 1 : 0); WREG32(reg_offset, sdma_cntl); return 0; } static int sdma_v6_0_process_trap_irq(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { int instances, queue; uint32_t mes_queue_id = entry->src_data[0]; DRM_DEBUG("IH: SDMA trap\n"); if (adev->enable_mes && (mes_queue_id & AMDGPU_FENCE_MES_QUEUE_FLAG)) { struct amdgpu_mes_queue *queue; mes_queue_id &= AMDGPU_FENCE_MES_QUEUE_ID_MASK; spin_lock(&adev->mes.queue_id_lock); queue = idr_find(&adev->mes.queue_id_idr, mes_queue_id); if (queue) { DRM_DEBUG("process smda queue id = %d\n", mes_queue_id); amdgpu_fence_process(queue->ring); } spin_unlock(&adev->mes.queue_id_lock); return 0; } queue = entry->ring_id & 0xf; instances = (entry->ring_id & 0xf0) >> 4; if (instances > 1) { DRM_ERROR("IH: wrong ring_ID detected, as wrong sdma instance\n"); return -EINVAL; } switch (entry->client_id) { case SOC21_IH_CLIENTID_GFX: switch (queue) { case 0: amdgpu_fence_process(&adev->sdma.instance[instances].ring); break; default: break; } break; } return 0; } static int sdma_v6_0_process_illegal_inst_irq(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { return 0; } static int sdma_v6_0_set_clockgating_state(void *handle, enum amd_clockgating_state state) { return 0; } static int sdma_v6_0_set_powergating_state(void *handle, enum amd_powergating_state state) { return 0; } static void sdma_v6_0_get_clockgating_state(void *handle, u64 *flags) { } const struct amd_ip_funcs sdma_v6_0_ip_funcs = { .name = "sdma_v6_0", .early_init = sdma_v6_0_early_init, .late_init = NULL, .sw_init = sdma_v6_0_sw_init, .sw_fini = sdma_v6_0_sw_fini, .hw_init = sdma_v6_0_hw_init, .hw_fini = sdma_v6_0_hw_fini, .suspend = sdma_v6_0_suspend, .resume = sdma_v6_0_resume, .is_idle = sdma_v6_0_is_idle, .wait_for_idle = sdma_v6_0_wait_for_idle, .soft_reset = sdma_v6_0_soft_reset, .check_soft_reset = sdma_v6_0_check_soft_reset, .set_clockgating_state = sdma_v6_0_set_clockgating_state, .set_powergating_state = sdma_v6_0_set_powergating_state, .get_clockgating_state = sdma_v6_0_get_clockgating_state, }; static const struct amdgpu_ring_funcs sdma_v6_0_ring_funcs = { .type = AMDGPU_RING_TYPE_SDMA, .align_mask = 0xf, .nop = SDMA_PKT_NOP_HEADER_OP(SDMA_OP_NOP), .support_64bit_ptrs = true, .vmhub = AMDGPU_GFXHUB_0, .get_rptr = sdma_v6_0_ring_get_rptr, .get_wptr = sdma_v6_0_ring_get_wptr, .set_wptr = sdma_v6_0_ring_set_wptr, .emit_frame_size = 5 + /* sdma_v6_0_ring_init_cond_exec */ 6 + /* sdma_v6_0_ring_emit_hdp_flush */ 6 + /* sdma_v6_0_ring_emit_pipeline_sync */ /* sdma_v6_0_ring_emit_vm_flush */ SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 6 + 10 + 10 + 10, /* sdma_v6_0_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 5 + 7 + 6, /* sdma_v6_0_ring_emit_ib */ .emit_ib = sdma_v6_0_ring_emit_ib, .emit_mem_sync = sdma_v6_0_ring_emit_mem_sync, .emit_fence = sdma_v6_0_ring_emit_fence, .emit_pipeline_sync = sdma_v6_0_ring_emit_pipeline_sync, .emit_vm_flush = sdma_v6_0_ring_emit_vm_flush, .emit_hdp_flush = sdma_v6_0_ring_emit_hdp_flush, .test_ring = sdma_v6_0_ring_test_ring, .test_ib = sdma_v6_0_ring_test_ib, .insert_nop = sdma_v6_0_ring_insert_nop, .pad_ib = sdma_v6_0_ring_pad_ib, .emit_wreg = sdma_v6_0_ring_emit_wreg, .emit_reg_wait = sdma_v6_0_ring_emit_reg_wait, .emit_reg_write_reg_wait = sdma_v6_0_ring_emit_reg_write_reg_wait, .init_cond_exec = sdma_v6_0_ring_init_cond_exec, .patch_cond_exec = sdma_v6_0_ring_patch_cond_exec, .preempt_ib = sdma_v6_0_ring_preempt_ib, }; static void sdma_v6_0_set_ring_funcs(struct amdgpu_device *adev) { int i; for (i = 0; i < adev->sdma.num_instances; i++) { adev->sdma.instance[i].ring.funcs = &sdma_v6_0_ring_funcs; adev->sdma.instance[i].ring.me = i; } } static const struct amdgpu_irq_src_funcs sdma_v6_0_trap_irq_funcs = { .set = sdma_v6_0_set_trap_irq_state, .process = sdma_v6_0_process_trap_irq, }; static const struct amdgpu_irq_src_funcs sdma_v6_0_illegal_inst_irq_funcs = { .process = sdma_v6_0_process_illegal_inst_irq, }; static void sdma_v6_0_set_irq_funcs(struct amdgpu_device *adev) { adev->sdma.trap_irq.num_types = AMDGPU_SDMA_IRQ_INSTANCE0 + adev->sdma.num_instances; adev->sdma.trap_irq.funcs = &sdma_v6_0_trap_irq_funcs; adev->sdma.illegal_inst_irq.funcs = &sdma_v6_0_illegal_inst_irq_funcs; } /** * sdma_v6_0_emit_copy_buffer - copy buffer using the sDMA engine * * @ring: amdgpu_ring structure holding ring information * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer * * Copy GPU buffers using the DMA engine. * Used by the amdgpu ttm implementation to move pages if * registered as the asic copy callback. */ static void sdma_v6_0_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, bool tmz) { ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_COPY_LINEAR_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR) | SDMA_PKT_COPY_LINEAR_HEADER_TMZ(tmz ? 1 : 0); ib->ptr[ib->length_dw++] = byte_count - 1; ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */ ib->ptr[ib->length_dw++] = lower_32_bits(src_offset); ib->ptr[ib->length_dw++] = upper_32_bits(src_offset); ib->ptr[ib->length_dw++] = lower_32_bits(dst_offset); ib->ptr[ib->length_dw++] = upper_32_bits(dst_offset); } /** * sdma_v6_0_emit_fill_buffer - fill buffer using the sDMA engine * * @ring: amdgpu_ring structure holding ring information * @src_data: value to write to buffer * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer * * Fill GPU buffers using the DMA engine. */ static void sdma_v6_0_emit_fill_buffer(struct amdgpu_ib *ib, uint32_t src_data, uint64_t dst_offset, uint32_t byte_count) { ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_CONST_FILL); ib->ptr[ib->length_dw++] = lower_32_bits(dst_offset); ib->ptr[ib->length_dw++] = upper_32_bits(dst_offset); ib->ptr[ib->length_dw++] = src_data; ib->ptr[ib->length_dw++] = byte_count - 1; } static const struct amdgpu_buffer_funcs sdma_v6_0_buffer_funcs = { .copy_max_bytes = 0x400000, .copy_num_dw = 7, .emit_copy_buffer = sdma_v6_0_emit_copy_buffer, .fill_max_bytes = 0x400000, .fill_num_dw = 5, .emit_fill_buffer = sdma_v6_0_emit_fill_buffer, }; static void sdma_v6_0_set_buffer_funcs(struct amdgpu_device *adev) { adev->mman.buffer_funcs = &sdma_v6_0_buffer_funcs; adev->mman.buffer_funcs_ring = &adev->sdma.instance[0].ring; } static const struct amdgpu_vm_pte_funcs sdma_v6_0_vm_pte_funcs = { .copy_pte_num_dw = 7, .copy_pte = sdma_v6_0_vm_copy_pte, .write_pte = sdma_v6_0_vm_write_pte, .set_pte_pde = sdma_v6_0_vm_set_pte_pde, }; static void sdma_v6_0_set_vm_pte_funcs(struct amdgpu_device *adev) { unsigned i; adev->vm_manager.vm_pte_funcs = &sdma_v6_0_vm_pte_funcs; for (i = 0; i < adev->sdma.num_instances; i++) { adev->vm_manager.vm_pte_scheds[i] = &adev->sdma.instance[i].ring.sched; } adev->vm_manager.vm_pte_num_scheds = adev->sdma.num_instances; } const struct amdgpu_ip_block_version sdma_v6_0_ip_block = { .type = AMD_IP_BLOCK_TYPE_SDMA, .major = 6, .minor = 0, .rev = 0, .funcs = &sdma_v6_0_ip_funcs, };