[PATCH v3 4/6] drm/amdgpu: add ring mode support for lsdma
Yifan Zhang
yifan1.zhang at amd.com
Tue Nov 21 07:30:12 UTC 2023
This patch is to add ring mode support for lsdma.
Signed-off-by: Yifan Zhang <yifan1.zhang at amd.com>
Reviewed-by: Tim Huang <Tim.Huang at amd.com>
Reviewed-by: Lang Yu <lang.yu at amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 20 +
drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.c | 40 +
drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.h | 33 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h | 1 +
drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.c | 1293 +++++++++++++++++
drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.h | 3 +-
6 files changed, 1384 insertions(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
index f96811bbe40e..3e7ad86d7fa9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
@@ -2044,6 +2044,22 @@ static int amdgpu_discovery_set_sdma_ip_blocks(struct amdgpu_device *adev)
return 0;
}
+static int amdgpu_discovery_set_lsdma_ip_blocks(struct amdgpu_device *adev)
+{
+ switch (amdgpu_ip_version(adev, LSDMA_HWIP, 0)) {
+ case IP_VERSION(6, 0, 0):
+ case IP_VERSION(6, 0, 1):
+ case IP_VERSION(6, 0, 2):
+ case IP_VERSION(6, 0, 3):
+ case IP_VERSION(6, 1, 0):
+ amdgpu_device_ip_block_add(adev, &lsdma_v6_0_ip_block);
+ break;
+ default:
+ break;
+ }
+ return 0;
+}
+
static int amdgpu_discovery_set_mm_ip_blocks(struct amdgpu_device *adev)
{
if (amdgpu_ip_version(adev, VCE_HWIP, 0)) {
@@ -2699,6 +2715,10 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
if (r)
return r;
+ r = amdgpu_discovery_set_lsdma_ip_blocks(adev);
+ if (r)
+ return r;
+
r = amdgpu_discovery_set_mes_ip_blocks(adev);
if (r)
return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.c
index 4d1d4994ea3f..44428c4de958 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.c
@@ -25,6 +25,8 @@
#include "amdgpu_lsdma.h"
#define AMDGPU_LSDMA_MAX_SIZE 0x2000000ULL
+/* LSDMA CSA resides in the 5th page of CSA */
+#define AMDGPU_CSA_LSDMA_OFFSET (4096 * 4)
int amdgpu_lsdma_wait_for(struct amdgpu_device *adev,
uint32_t reg_index, uint32_t reg_val,
@@ -89,3 +91,41 @@ int amdgpu_lsdma_fill_mem(struct amdgpu_device *adev,
return 0;
}
+
+uint64_t amdgpu_lsdma_get_csa_mc_addr(struct amdgpu_ring *ring,
+ unsigned int vmid)
+{
+ struct amdgpu_device *adev = ring->adev;
+ uint64_t csa_mc_addr;
+
+ /* don't enable OS preemption on SDMA under SRIOV */
+ if (amdgpu_sriov_vf(adev) || vmid == 0 || !adev->gfx.mcbp)
+ return 0;
+
+ if (ring->is_mes_queue) {
+ uint32_t offset = 0;
+
+ offset = offsetof(struct amdgpu_mes_ctx_meta_data,
+ sdma[ring->idx].sdma_meta_data);
+ csa_mc_addr = amdgpu_mes_ctx_get_offs_gpu_addr(ring, offset);
+ } else {
+ csa_mc_addr = amdgpu_csa_vaddr(adev) + AMDGPU_CSA_LSDMA_OFFSET;
+ }
+
+ return csa_mc_addr;
+}
+
+void amdgpu_lsdma_destroy_inst_ctx(struct amdgpu_device *adev)
+{
+ amdgpu_ucode_release(&adev->lsdma.fw);
+}
+
+void amdgpu_lsdma_unset_buffer_funcs_helper(struct amdgpu_device *adev)
+{
+ struct amdgpu_ring *lsdma;
+
+ lsdma = &adev->lsdma.ring;
+
+ if (adev->mman.buffer_funcs_ring == lsdma)
+ amdgpu_ttm_set_buffer_funcs_status(adev, false);
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.h
index c61ba58c5ee0..933c59ded652 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_lsdma.h
@@ -24,10 +24,6 @@
#ifndef __AMDGPU_LSDMA_H__
#define __AMDGPU_LSDMA_H__
-struct amdgpu_lsdma {
- const struct amdgpu_lsdma_funcs *funcs;
-};
-
struct amdgpu_lsdma_funcs {
int (*copy_mem)(struct amdgpu_device *adev, uint64_t src_addr,
uint64_t dst_addr, uint64_t size);
@@ -43,4 +39,33 @@ int amdgpu_lsdma_fill_mem(struct amdgpu_device *adev, uint64_t dst_addr,
int amdgpu_lsdma_wait_for(struct amdgpu_device *adev, uint32_t reg_index,
uint32_t reg_val, uint32_t mask);
+struct amdgpu_lsdma {
+ struct amdgpu_irq_src trap_irq;
+ struct amdgpu_irq_src illegal_inst_irq;
+ struct amdgpu_irq_src ecc_irq;
+ struct amdgpu_irq_src vm_hole_irq;
+ struct amdgpu_irq_src doorbell_invalid_irq;
+ struct amdgpu_irq_src pool_timeout_irq;
+ struct amdgpu_irq_src srbm_write_irq;
+
+ uint32_t lsdma_mask;
+
+ /* LSDMA firmware */
+ const struct firmware *fw;
+ uint32_t fw_version;
+ uint32_t feature_version;
+
+ struct amdgpu_ring ring;
+ struct amdgpu_ring page;
+ bool burst_nop;
+ uint32_t aid_id;
+
+ const struct amdgpu_lsdma_funcs *funcs;
+
+};
+
+uint64_t amdgpu_lsdma_get_csa_mc_addr(struct amdgpu_ring *ring, unsigned int vmid);
+void amdgpu_lsdma_unset_buffer_funcs_helper(struct amdgpu_device *adev);
+void amdgpu_lsdma_destroy_inst_ctx(struct amdgpu_device *adev);
+
#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
index 7f6a14e2282e..95004c2de81d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
@@ -79,6 +79,7 @@ enum amdgpu_ring_type {
AMDGPU_RING_TYPE_VCN_ENC = AMDGPU_HW_IP_VCN_ENC,
AMDGPU_RING_TYPE_VCN_JPEG = AMDGPU_HW_IP_VCN_JPEG,
AMDGPU_RING_TYPE_VPE = AMDGPU_HW_IP_VPE,
+ AMDGPU_RING_TYPE_LSDMA = AMDGPU_HW_IP_LSDMA,
AMDGPU_RING_TYPE_KIQ,
AMDGPU_RING_TYPE_MES,
AMDGPU_RING_TYPE_UMSCH_MM,
diff --git a/drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.c
index 1a285b531881..26577525df0f 100644
--- a/drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.c
@@ -22,12 +22,21 @@
*/
#include <linux/delay.h>
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+
#include "amdgpu.h"
#include "lsdma_v6_0.h"
#include "amdgpu_lsdma.h"
#include "lsdma/lsdma_6_0_0_offset.h"
#include "lsdma/lsdma_6_0_0_sh_mask.h"
+#include "ivsrcid/gfx/irqsrcs_gfx_11_0_0.h"
+#include "sdma_v6_0_0_pkt_open.h"
+#include "v11_structs.h"
+
+static void lsdma_v6_0_set_buffer_funcs(struct amdgpu_device *adev);
static int lsdma_v6_0_wait_pio_status(struct amdgpu_device *adev)
{
@@ -119,3 +128,1287 @@ const struct amdgpu_lsdma_funcs lsdma_v6_0_funcs = {
.fill_mem = lsdma_v6_0_fill_mem,
.update_memory_power_gating = lsdma_v6_0_update_memory_power_gating
};
+
+/**
+ * lsdma_v6_0_ctxempty_int_enable - enable or disable context empty interrupts
+ *
+ * @adev: amdgpu_device pointer
+ * @enable: enable/disable context switching due to queue empty conditions
+ *
+ * Enable or disable the async dma engines queue empty context switch.
+ */
+static void lsdma_v6_0_ctxempty_int_enable(struct amdgpu_device *adev, bool enable)
+{
+ uint32_t f32_cntl;
+
+
+ if (!amdgpu_sriov_vf(adev)) {
+ f32_cntl = RREG32_SOC15(LSDMA, 0, regLSDMA_INT_CNTL);
+ f32_cntl = REG_SET_FIELD(f32_cntl, LSDMA_INT_CNTL,
+ CTX_EMPTY_INT_ENABLE, enable ? 1 : 0);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_INT_CNTL, f32_cntl);
+ }
+}
+
+/**
+ * lsdma_v6_0_gfx_stop - stop the gfx async dma engines
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Stop the gfx async dma ring buffers.
+ */
+static void lsdma_v6_0_gfx_stop(struct amdgpu_device *adev)
+{
+ u32 rb_cntl, ib_cntl;
+
+ amdgpu_lsdma_unset_buffer_funcs_helper(adev);
+
+ rb_cntl = RREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_CNTL);
+ rb_cntl = REG_SET_FIELD(rb_cntl, LSDMA_QUEUE0_RB_CNTL, RB_ENABLE, 0);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_CNTL, rb_cntl);
+
+ ib_cntl = RREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_IB_CNTL);
+ ib_cntl = REG_SET_FIELD(rb_cntl, LSDMA_QUEUE0_IB_CNTL, IB_ENABLE, 0);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_CNTL, ib_cntl);
+
+}
+
+/**
+ * lsdma_v6_0_rlc_stop - stop the compute async dma engines
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Stop the compute async dma queues.
+ */
+static void lsdma_v6_0_rlc_stop(struct amdgpu_device *adev)
+{
+ /* XXX todo */
+}
+
+
+/**
+ * lsdma_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 lsdma_v6_0_enable(struct amdgpu_device *adev, bool enable)
+{
+ u32 f32_cntl;
+
+ if (!enable) {
+ lsdma_v6_0_gfx_stop(adev);
+ lsdma_v6_0_rlc_stop(adev);
+ }
+
+ if (amdgpu_sriov_vf(adev))
+ return;
+
+ f32_cntl = RREG32_SOC15(LSDMA, 0, regLSDMA_F32_CNTL);
+ f32_cntl = REG_SET_FIELD(f32_cntl, LSDMA_F32_CNTL, HALT, enable ? 0 : 1);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_F32_CNTL, f32_cntl);
+}
+
+
+/**
+ * lsdma_v6_0_emit_copy_buffer - copy buffer using the sDMA engine
+ *
+ * @ib: indirect buffer to fill with commands
+ * @src_offset: src GPU address
+ * @dst_offset: dst GPU address
+ * @byte_count: number of bytes to xfer
+ * @tmz: if a secure copy should be used
+ *
+ * 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 lsdma_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);
+}
+
+/**
+ * lsdma_v6_0_emit_fill_buffer - fill buffer using the sDMA engine
+ *
+ * @ib: indirect buffer to fill
+ * @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 lsdma_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 lsdma_v6_0_buffer_funcs = {
+ .copy_max_bytes = 0x400000,
+ .copy_num_dw = 7,
+ .emit_copy_buffer = lsdma_v6_0_emit_copy_buffer,
+
+ .fill_max_bytes = 0x400000,
+ .fill_num_dw = 5,
+ .emit_fill_buffer = lsdma_v6_0_emit_fill_buffer,
+};
+
+
+static void lsdma_v6_0_set_buffer_funcs(struct amdgpu_device *adev)
+{
+ adev->mman.buffer_funcs = &lsdma_v6_0_buffer_funcs;
+ adev->mman.buffer_funcs_ring = &adev->lsdma.ring;
+}
+
+
+/**
+ * lsdma_v6_0_ring_get_rptr - get the current read pointer
+ *
+ * @ring: amdgpu ring pointer
+ *
+ * Get the current rptr from the hardware.
+ */
+static uint64_t lsdma_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);
+}
+
+
+/**
+ * lsdma_v6_0_ring_get_wptr - get the current write pointer
+ *
+ * @ring: amdgpu ring pointer
+ *
+ * Get the current wptr from the hardware.
+ */
+static uint64_t lsdma_v6_0_ring_get_wptr(struct amdgpu_ring *ring)
+{
+ struct amdgpu_device *adev = ring->adev;
+ 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);
+ } else {
+ wptr = (u64)RREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_WPTR) |
+ (u64)RREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_WPTR_HI) << 32;
+ }
+
+ 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 lsdma_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;
+
+ 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 -- "
+ "ring->wptr << 2 0x%016llx\n"
+ "*(ring->rptr_cpu_addr) 0x%016llx\n",
+ ring->wptr << 2,
+ (u64)(*ring->rptr_cpu_addr));
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_WPTR, lower_32_bits(ring->wptr << 2));
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_WPTR_HI, upper_32_bits(ring->wptr << 2));
+ }
+ }
+}
+
+static void lsdma_v6_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+ struct amdgpu_lsdma *lsdma = &ring->adev->lsdma;
+ int i;
+
+ for (i = 0; i < count; i++)
+ if (lsdma && lsdma->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);
+}
+
+
+/*
+ * lsdma_v6_0_ring_emit_ib - Schedule an IB on the DMA engine
+ *
+ * @ring: amdgpu ring pointer
+ * @ib: IB object to schedule
+ * @flags: unused
+ * @job: job to retrieve vmid from
+ *
+ * Schedule an IB in the DMA ring.
+ */
+static void lsdma_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_lsdma_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.
+ */
+ lsdma_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));
+}
+
+/**
+ * lsdma_v6_0_ring_emit_fence - emit a fence on the DMA ring
+ *
+ * @ring: amdgpu ring pointer
+ * @addr: address
+ * @seq: fence seq number
+ * @flags: fence flags
+ *
+ * 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 lsdma_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));
+ }
+}
+/**
+ * lsdma_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 lsdma_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 */
+}
+
+/*
+ * lsdma_v6_0_ring_emit_vm_flush - vm flush using sDMA
+ *
+ * @ring: amdgpu_ring pointer
+ * @vmid: vmid number to use
+ * @pd_addr: address
+ *
+ * Update the page table base and flush the VM TLB
+ * using sDMA.
+ */
+static void lsdma_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
+ unsigned vmid, uint64_t pd_addr)
+{
+ struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->vm_hub];
+ uint32_t req = hub->vmhub_funcs->get_invalidate_req(vmid, 0);
+
+ /* Update the PD address for this VMID. */
+ amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_lo32 +
+ (hub->ctx_addr_distance * vmid),
+ lower_32_bits(pd_addr));
+ amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_hi32 +
+ (hub->ctx_addr_distance * vmid),
+ upper_32_bits(pd_addr));
+
+ /* Trigger invalidation. */
+ amdgpu_ring_write(ring,
+ SDMA_PKT_VM_INVALIDATION_HEADER_OP(SDMA_OP_POLL_REGMEM) |
+ SDMA_PKT_VM_INVALIDATION_HEADER_SUB_OP(SDMA_SUBOP_VM_INVALIDATION) |
+ SDMA_PKT_VM_INVALIDATION_HEADER_GFX_ENG_ID(ring->vm_inv_eng) |
+ SDMA_PKT_VM_INVALIDATION_HEADER_MM_ENG_ID(0x1f));
+ amdgpu_ring_write(ring, req);
+ amdgpu_ring_write(ring, 0xFFFFFFFF);
+ amdgpu_ring_write(ring,
+ SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_INVALIDATEACK(1 << vmid) |
+ SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_ADDRESSRANGEHI(0x1F));
+}
+
+static int lsdma_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, 0xabcdef12);
+ 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 == 0xabcdef12)
+ 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;
+}
+
+
+/*
+ * lsdma_v6_0_ring_test_ib - test an IB on the DMA engine
+ *
+ * @ring: amdgpu_ring structure holding ring information
+ * @timeout: timeout value in jiffies, or MAX_SCHEDULE_TIMEOUT
+ *
+ * Test a simple IB in the DMA ring.
+ * Returns 0 on success, error on failure.
+ */
+static int lsdma_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] = 0xfeffe123;
+ 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 == 0xfeffe123) {
+ 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;
+}
+
+static void lsdma_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 lsdma_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 lsdma_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 unsigned lsdma_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 lsdma_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;
+}
+
+static int lsdma_v6_0_ring_preempt_ib(struct amdgpu_ring *ring)
+{
+ int i, r = 0;
+ struct amdgpu_device *adev = ring->adev;
+
+ /* assert preemption condition */
+ amdgpu_ring_set_preempt_cond_exec(ring, false);
+
+ /* emit the trailing fence */
+ ring->trail_seq += 1;
+ amdgpu_ring_alloc(ring, 10);
+ lsdma_v6_0_ring_emit_fence(ring, ring->trail_fence_gpu_addr,
+ ring->trail_seq, 0);
+ amdgpu_ring_commit(ring);
+
+ /* assert IB preemption */
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_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_SOC15(LSDMA, 0, regLSDMA_QUEUE0_PREEMPT, 0);
+
+ /* deassert the preemption condition */
+ amdgpu_ring_set_preempt_cond_exec(ring, true);
+ return r;
+}
+
+/*
+ * lsdma_v6_0_ring_pad_ib - pad the IB
+ * @ib: indirect buffer to fill with padding
+ * @ring: amdgpu ring pointer
+ *
+ * Pad the IB with NOPs to a boundary multiple of 8.
+ */
+static void lsdma_v6_0_ring_pad_ib(struct amdgpu_ring *ring, struct amdgpu_ib *ib)
+{
+ struct amdgpu_lsdma *lsdma = &ring->adev->lsdma;
+ u32 pad_count;
+ int i;
+
+ pad_count = (-ib->length_dw) & 0x7;
+ for (i = 0; i < pad_count; i++)
+ if (lsdma->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);
+}
+
+/**
+ * lsdma_v6_0_ring_emit_pipeline_sync - sync the pipeline
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Make sure all previous operations are completed (CIK).
+ */
+static void lsdma_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 */
+}
+
+
+
+
+static const struct amdgpu_ring_funcs lsdma_v6_0_ring_funcs = {
+ .type = AMDGPU_RING_TYPE_LSDMA,
+ .align_mask = 0xf,
+ .nop = SDMA_PKT_NOP_HEADER_OP(SDMA_OP_NOP),
+ .support_64bit_ptrs = true,
+ .secure_submission_supported = true,
+ .get_rptr = lsdma_v6_0_ring_get_rptr,
+ .get_wptr = lsdma_v6_0_ring_get_wptr,
+ .set_wptr = lsdma_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 = lsdma_v6_0_ring_emit_ib,
+ .emit_fence = lsdma_v6_0_ring_emit_fence,
+ .emit_pipeline_sync = lsdma_v6_0_ring_emit_pipeline_sync,
+ .emit_vm_flush = lsdma_v6_0_ring_emit_vm_flush,
+ .emit_hdp_flush = lsdma_v6_0_ring_emit_hdp_flush,
+ .test_ring = lsdma_v6_0_ring_test_ring,
+ .test_ib = lsdma_v6_0_ring_test_ib,
+ .insert_nop = lsdma_v6_0_ring_insert_nop,
+ .pad_ib = lsdma_v6_0_ring_pad_ib,
+ .emit_wreg = lsdma_v6_0_ring_emit_wreg,
+ .emit_reg_wait = lsdma_v6_0_ring_emit_reg_wait,
+ .emit_reg_write_reg_wait = lsdma_v6_0_ring_emit_reg_write_reg_wait,
+ .init_cond_exec = lsdma_v6_0_ring_init_cond_exec,
+ .patch_cond_exec = lsdma_v6_0_ring_patch_cond_exec,
+ .preempt_ib = lsdma_v6_0_ring_preempt_ib,
+};
+
+static void lsdma_v6_0_set_ring_funcs(struct amdgpu_device *adev)
+{
+ adev->lsdma.ring.funcs = &lsdma_v6_0_ring_funcs;
+}
+
+static int lsdma_v6_0_set_trap_irq_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ u32 lsdma_int_cntl;
+
+
+ if (!amdgpu_sriov_vf(adev)) {
+ lsdma_int_cntl = RREG32_SOC15(LSDMA, 0, regLSDMA_INT_CNTL);
+ lsdma_int_cntl = REG_SET_FIELD(lsdma_int_cntl, LSDMA_INT_CNTL, TRAP_INT_ENABLE,
+ state == AMDGPU_IRQ_STATE_ENABLE ? 1 : 0);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_INT_CNTL, lsdma_int_cntl);
+ }
+
+ return 0;
+}
+
+static int lsdma_v6_0_process_trap_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ int queue;
+ uint32_t mes_queue_id = entry->src_data[0];
+
+ DRM_DEBUG("IH: LSDMA trap, client_id %d, src_id %d, ring_id %d,\
+ vmid %d, mes_queue_id %d\n", entry->client_id, entry->src_id,
+ entry->ring_id, entry->vmid, mes_queue_id);
+
+ 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 lsmda 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;
+
+ switch (entry->client_id) {
+ case SOC21_IH_CLIENTID_LSDMA:
+ switch (queue) {
+ case 0:
+ amdgpu_fence_process(&adev->lsdma.ring);
+ break;
+ default:
+ break;
+ }
+ break;
+ }
+ return 0;
+}
+
+
+static const struct amdgpu_irq_src_funcs lsdma_v6_0_trap_irq_funcs = {
+ .set = lsdma_v6_0_set_trap_irq_state,
+ .process = lsdma_v6_0_process_trap_irq,
+};
+
+static void lsdma_v6_0_set_irq_funcs(struct amdgpu_device *adev)
+{
+ adev->lsdma.trap_irq.num_types = 1;
+ adev->lsdma.trap_irq.funcs = &lsdma_v6_0_trap_irq_funcs;
+}
+
+/**
+ * lsdma_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 lsdma_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);
+
+}
+
+/**
+ * lsdma_v6_0_vm_write_pte - update PTEs by writing them manually
+ *
+ * @ib: indirect buffer to fill with commands
+ * @pe: addr of the page entry
+ * @value: dst addr to write into pe
+ * @count: number of page entries to update
+ * @incr: increase next addr by incr bytes
+ *
+ * Update PTEs by writing them manually using sDMA.
+ */
+static void lsdma_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 lsdma_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 */
+}
+
+static const struct amdgpu_vm_pte_funcs lsdma_v6_0_vm_pte_funcs = {
+ .copy_pte_num_dw = 7,
+ .copy_pte = lsdma_v6_0_vm_copy_pte,
+ .write_pte = lsdma_v6_0_vm_write_pte,
+ .set_pte_pde = lsdma_v6_0_vm_set_pte_pde,
+};
+
+
+static void lsdma_v6_0_set_vm_pte_funcs(struct amdgpu_device *adev)
+{
+ adev->vm_manager.vm_pte_funcs = &lsdma_v6_0_vm_pte_funcs;
+ adev->vm_manager.vm_pte_scheds[0] = &adev->lsdma.ring.sched;
+ adev->vm_manager.vm_pte_num_scheds = 1;
+}
+
+static int lsdma_v6_0_early_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+
+ lsdma_v6_0_set_ring_funcs(adev);
+ lsdma_v6_0_set_buffer_funcs(adev);
+ lsdma_v6_0_set_vm_pte_funcs(adev);
+ lsdma_v6_0_set_irq_funcs(adev);
+
+ return 0;
+}
+
+static int lsdma_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_LSDMA,
+ GFX_11_0_0__SRCID__LSDMA_TRAP,
+ &adev->lsdma.trap_irq);
+ if (r)
+ return r;
+
+ ring = &adev->lsdma.ring;
+ ring->ring_obj = NULL;
+ ring->use_doorbell = false;
+
+ DRM_INFO("LSDMA %d use_doorbell being set to: [%s]\n", i,
+ ring->use_doorbell?"true":"false");
+
+ ring->vm_hub = AMDGPU_MMHUB0(0);
+ sprintf(ring->name, "lsdma");
+ r = amdgpu_ring_init(adev, ring, 1024, &adev->lsdma.trap_irq,
+ 0, AMDGPU_RING_PRIO_DEFAULT, NULL);
+ if (r)
+ return r;
+
+ return r;
+}
+
+static int lsdma_v6_0_sw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+
+ amdgpu_ring_fini(&adev->lsdma.ring);
+
+ amdgpu_lsdma_destroy_inst_ctx(adev);
+
+ return 0;
+}
+
+
+/**
+ * lsdma_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 lsdma_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;
+ int r;
+
+ ring = &adev->lsdma.ring;
+
+ if (!amdgpu_sriov_vf(adev))
+ WREG32_SOC15(LSDMA, 0, regLSDMA_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(LSDMA, 0, regLSDMA_QUEUE0_RB_CNTL);
+ rb_cntl = REG_SET_FIELD(rb_cntl, LSDMA_QUEUE0_RB_CNTL, RB_SIZE, rb_bufsz);
+#ifdef __BIG_ENDIAN
+ rb_cntl = REG_SET_FIELD(rb_cntl, LSDMA_QUEUE0_RB_CNTL, RB_SWAP_ENABLE, 1);
+ rb_cntl = REG_SET_FIELD(rb_cntl, LSDMA_QUEUE0_RB_CNTL,
+ RPTR_WRITEBACK_SWAP_ENABLE, 1);
+#endif
+ rb_cntl = REG_SET_FIELD(rb_cntl, LSDMA_QUEUE0_RB_CNTL, RB_PRIV, 1);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_CNTL, rb_cntl);
+
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_WPTR_POLL_ADDR_LO, lower_32_bits(ring->wptr_gpu_addr));
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_WPTR_POLL_ADDR_HI, upper_32_bits(ring->wptr_gpu_addr));
+
+ /* set the wb address whether it's enabled or not */
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_RPTR_ADDR_HI, upper_32_bits(ring->rptr_gpu_addr));
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_RPTR_ADDR_LO, lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC);
+
+ rb_cntl = REG_SET_FIELD(rb_cntl, LSDMA_QUEUE0_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1);
+
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_BASE, ring->gpu_addr >> 8);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_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(LSDMA, 0, regLSDMA_QUEUE0_MINOR_PTR_UPDATE, 1);
+
+ if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_WPTR, lower_32_bits(ring->wptr) << 2);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_WPTR_HI, upper_32_bits(ring->wptr) << 2);
+ }
+
+ doorbell = RREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_DOORBELL);
+ doorbell_offset = RREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_DOORBELL_OFFSET);
+
+ if (ring->use_doorbell) {
+ doorbell = REG_SET_FIELD(doorbell, LSDMA_QUEUE0_DOORBELL, ENABLE, 1);
+ doorbell_offset = REG_SET_FIELD(doorbell_offset, LSDMA_QUEUE0_DOORBELL_OFFSET,
+ OFFSET, ring->doorbell_index);
+ } else {
+ doorbell = REG_SET_FIELD(doorbell, LSDMA_QUEUE0_DOORBELL, ENABLE, 0);
+ }
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_DOORBELL, doorbell);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_DOORBELL_OFFSET, doorbell_offset);
+
+
+ if (amdgpu_sriov_vf(adev))
+ lsdma_v6_0_ring_set_wptr(ring);
+
+ /* set minor_ptr_update to 0 after wptr programed */
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_MINOR_PTR_UPDATE, 0);
+
+ /* enable DMA RB */
+ rb_cntl = REG_SET_FIELD(rb_cntl, LSDMA_QUEUE0_RB_CNTL, RB_ENABLE, 1);
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_RB_CNTL, rb_cntl);
+
+ ib_cntl = RREG32_SOC15_IP(LSDMA, regLSDMA_QUEUE0_IB_CNTL);
+ ib_cntl = REG_SET_FIELD(ib_cntl, LSDMA_QUEUE0_IB_CNTL, IB_ENABLE, 1);
+#ifdef __BIG_ENDIAN
+ ib_cntl = REG_SET_FIELD(ib_cntl, LSDMA_QUEUE0_IB_CNTL, IB_SWAP_ENABLE, 1);
+#endif
+ /* enable DMA IBs */
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_IB_CNTL, ib_cntl);
+
+ if (amdgpu_sriov_vf(adev))
+ lsdma_v6_0_enable(adev, true);
+
+ r = amdgpu_ring_test_helper(ring);
+ if (r)
+ return r;
+
+ if (adev->mman.buffer_funcs_ring == ring)
+ amdgpu_ttm_set_buffer_funcs_status(adev, true);
+
+ adev->lsdma.burst_nop = true;
+
+ return 0;
+}
+
+/**
+ * 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 lsdma_v6_0_start(struct amdgpu_device *adev)
+{
+ int r;
+
+ lsdma_v6_0_ctxempty_int_enable(adev, true);
+
+ /* start the gfx rings and rlc compute queues */
+ r = lsdma_v6_0_gfx_resume(adev);
+ if (r)
+ return r;
+
+ return 0;
+}
+
+static int lsdma_v6_0_hw_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return lsdma_v6_0_start(adev);
+}
+
+static int lsdma_v6_0_hw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (amdgpu_sriov_vf(adev)) {
+ /* disable the scheduler for SDMA */
+ amdgpu_lsdma_unset_buffer_funcs_helper(adev);
+ return 0;
+ }
+
+ lsdma_v6_0_ctxempty_int_enable(adev, false);
+ lsdma_v6_0_enable(adev, false);
+
+ return 0;
+}
+
+static int lsdma_v6_0_suspend(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return lsdma_v6_0_hw_fini(adev);
+}
+
+static int lsdma_v6_0_resume(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return lsdma_v6_0_hw_init(adev);
+}
+
+static bool lsdma_v6_0_is_idle(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 tmp;
+
+ tmp = RREG32_SOC15(LSDMA, 0, regLSDMA_STATUS_REG);
+ if (!(tmp & LSDMA_STATUS_REG__IDLE_MASK))
+ return false;
+
+ return true;
+}
+
+static int lsdma_v6_0_wait_for_idle(void *handle)
+{
+ unsigned i;
+ u32 tmp;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ tmp = RREG32_SOC15(LSDMA, 0, regLSDMA_STATUS_REG);
+
+ if (tmp & LSDMA_STATUS_REG__IDLE_MASK)
+ return 0;
+ udelay(1);
+ }
+ return -ETIMEDOUT;
+}
+
+static int lsdma_v6_0_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 tmp;
+
+ lsdma_v6_0_gfx_stop(adev);
+
+ tmp = RREG32_SOC15(LSDMA, 0, regLSDMA_FREEZE);
+ tmp |= LSDMA_FREEZE__FREEZE_MASK;
+ WREG32_SOC15(LSDMA, 0, regLSDMA_FREEZE, tmp);
+
+ tmp = RREG32_SOC15(LSDMA, 0, regLSDMA_F32_CNTL);
+ tmp |= LSDMA_F32_CNTL__HALT_MASK;
+ tmp |= LSDMA_F32_CNTL__RESET_MASK;
+ WREG32_SOC15(LSDMA, 0, regLSDMA_F32_CNTL, tmp);
+
+ WREG32_SOC15(LSDMA, 0, regLSDMA_QUEUE0_PREEMPT, 0);
+
+ udelay(100);
+
+ return lsdma_v6_0_start(adev);
+}
+
+static bool lsdma_v6_0_check_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_ring *ring;
+ int r;
+ long tmo = msecs_to_jiffies(1000);
+
+ ring = &adev->lsdma.ring;
+ r = amdgpu_ring_test_ib(ring, tmo);
+ if (r)
+ return true;
+
+ return false;
+}
+
+static int lsdma_v6_0_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ return 0;
+}
+
+static int lsdma_v6_0_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ return 0;
+}
+
+static void lsdma_v6_0_get_clockgating_state(void *handle, u64 *flags)
+{
+}
+
+
+const struct amd_ip_funcs lsdma_v6_0_ip_funcs = {
+ .name = "lsdma_v6_0",
+ .early_init = lsdma_v6_0_early_init,
+ .late_init = NULL,
+ .sw_init = lsdma_v6_0_sw_init,
+ .sw_fini = lsdma_v6_0_sw_fini,
+ .hw_init = lsdma_v6_0_hw_init,
+ .hw_fini = lsdma_v6_0_hw_fini,
+ .suspend = lsdma_v6_0_suspend,
+ .resume = lsdma_v6_0_resume,
+ .is_idle = lsdma_v6_0_is_idle,
+ .wait_for_idle = lsdma_v6_0_wait_for_idle,
+ .soft_reset = lsdma_v6_0_soft_reset,
+ .check_soft_reset = lsdma_v6_0_check_soft_reset,
+ .set_clockgating_state = lsdma_v6_0_set_clockgating_state,
+ .set_powergating_state = lsdma_v6_0_set_powergating_state,
+ .get_clockgating_state = lsdma_v6_0_get_clockgating_state,
+};
+
+const struct amdgpu_ip_block_version lsdma_v6_0_ip_block = {
+ .type = AMD_IP_BLOCK_TYPE_LSDMA,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &lsdma_v6_0_ip_funcs,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.h b/drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.h
index 3ef79be1a9bf..c6940bf0914e 100644
--- a/drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.h
+++ b/drivers/gpu/drm/amd/amdgpu/lsdma_v6_0.h
@@ -24,8 +24,7 @@
#ifndef __LSDMA_V6_0_H__
#define __LSDMA_V6_0_H__
-#include "soc15_common.h"
-
extern const struct amdgpu_lsdma_funcs lsdma_v6_0_funcs;
+extern const struct amdgpu_ip_block_version lsdma_v6_0_ip_block;
#endif /* __LSDMA_V6_0_H__ */
--
2.37.3
More information about the amd-gfx
mailing list