1
0
Fork 0
alistair23-linux/drivers/misc/habanalabs/goya/goya_coresight.c

695 lines
21 KiB
C
Raw Normal View History

habanalabs: add new IOCTL for debug, tracing and profiling Habanalabs ASICs use the ARM coresight infrastructure to support debug, tracing and profiling of neural networks topologies. Because the coresight is configured using register writes and reads, and some of the registers hold sensitive information (e.g. the address in the device's DRAM where the trace data is written to), the user must go through the kernel driver to configure this mechanism. This patch implements the common code of the IOCTL and calls the ASIC-specific function for the actual H/W configuration. The IOCTL supports configuration of seven coresight components: ETR, ETF, STM, FUNNEL, BMON, SPMU and TIMESTAMP The user specifies which component he wishes to configure and provides a pointer to a structure (located in its process space) that contains the relevant configuration. The common code copies the relevant data from the user-space to kernel space and then calls the ASIC-specific function to do the H/W configuration. After the configuration is done, which is usually composed of several IOCTL calls depending on what the user wanted to trace, the user can start executing the topology. The trace data will be written to the user's area in the device's DRAM. After the tracing operation is complete, and user will call the IOCTL again to disable the tracing operation. The user also need to read values from registers for some of the components (e.g. the size of the trace data in the device's DRAM). In that case, the user will provide a pointer to an "output" structure in user-space, which the IOCTL code will fill according the to selected component. Signed-off-by: Omer Shpigelman <oshpigelman@habana.ai> Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
2019-04-01 13:31:22 -06:00
// SPDX-License-Identifier: GPL-2.0
/*
* Copyright 2016-2019 HabanaLabs, Ltd.
* All Rights Reserved.
*/
#include "goyaP.h"
#include "include/goya/goya_coresight.h"
#include "include/goya/asic_reg/goya_regs.h"
#include <uapi/misc/habanalabs.h>
#include <linux/coresight.h>
#define GOYA_PLDM_CORESIGHT_TIMEOUT_USEC (CORESIGHT_TIMEOUT_USEC * 100)
#define SPMU_SECTION_SIZE DMA_CH_0_CS_SPMU_MAX_OFFSET
#define SPMU_EVENT_TYPES_OFFSET 0x400
#define SPMU_MAX_COUNTERS 6
static u64 debug_stm_regs[GOYA_STM_LAST + 1] = {
[GOYA_STM_CPU] = mmCPU_STM_BASE,
[GOYA_STM_DMA_CH_0_CS] = mmDMA_CH_0_CS_STM_BASE,
[GOYA_STM_DMA_CH_1_CS] = mmDMA_CH_1_CS_STM_BASE,
[GOYA_STM_DMA_CH_2_CS] = mmDMA_CH_2_CS_STM_BASE,
[GOYA_STM_DMA_CH_3_CS] = mmDMA_CH_3_CS_STM_BASE,
[GOYA_STM_DMA_CH_4_CS] = mmDMA_CH_4_CS_STM_BASE,
[GOYA_STM_DMA_MACRO_CS] = mmDMA_MACRO_CS_STM_BASE,
[GOYA_STM_MME1_SBA] = mmMME1_SBA_STM_BASE,
[GOYA_STM_MME3_SBB] = mmMME3_SBB_STM_BASE,
[GOYA_STM_MME4_WACS2] = mmMME4_WACS2_STM_BASE,
[GOYA_STM_MME4_WACS] = mmMME4_WACS_STM_BASE,
[GOYA_STM_MMU_CS] = mmMMU_CS_STM_BASE,
[GOYA_STM_PCIE] = mmPCIE_STM_BASE,
[GOYA_STM_PSOC] = mmPSOC_STM_BASE,
[GOYA_STM_TPC0_EML] = mmTPC0_EML_STM_BASE,
[GOYA_STM_TPC1_EML] = mmTPC1_EML_STM_BASE,
[GOYA_STM_TPC2_EML] = mmTPC2_EML_STM_BASE,
[GOYA_STM_TPC3_EML] = mmTPC3_EML_STM_BASE,
[GOYA_STM_TPC4_EML] = mmTPC4_EML_STM_BASE,
[GOYA_STM_TPC5_EML] = mmTPC5_EML_STM_BASE,
[GOYA_STM_TPC6_EML] = mmTPC6_EML_STM_BASE,
[GOYA_STM_TPC7_EML] = mmTPC7_EML_STM_BASE
};
static u64 debug_etf_regs[GOYA_ETF_LAST + 1] = {
[GOYA_ETF_CPU_0] = mmCPU_ETF_0_BASE,
[GOYA_ETF_CPU_1] = mmCPU_ETF_1_BASE,
[GOYA_ETF_CPU_TRACE] = mmCPU_ETF_TRACE_BASE,
[GOYA_ETF_DMA_CH_0_CS] = mmDMA_CH_0_CS_ETF_BASE,
[GOYA_ETF_DMA_CH_1_CS] = mmDMA_CH_1_CS_ETF_BASE,
[GOYA_ETF_DMA_CH_2_CS] = mmDMA_CH_2_CS_ETF_BASE,
[GOYA_ETF_DMA_CH_3_CS] = mmDMA_CH_3_CS_ETF_BASE,
[GOYA_ETF_DMA_CH_4_CS] = mmDMA_CH_4_CS_ETF_BASE,
[GOYA_ETF_DMA_MACRO_CS] = mmDMA_MACRO_CS_ETF_BASE,
[GOYA_ETF_MME1_SBA] = mmMME1_SBA_ETF_BASE,
[GOYA_ETF_MME3_SBB] = mmMME3_SBB_ETF_BASE,
[GOYA_ETF_MME4_WACS2] = mmMME4_WACS2_ETF_BASE,
[GOYA_ETF_MME4_WACS] = mmMME4_WACS_ETF_BASE,
[GOYA_ETF_MMU_CS] = mmMMU_CS_ETF_BASE,
[GOYA_ETF_PCIE] = mmPCIE_ETF_BASE,
[GOYA_ETF_PSOC] = mmPSOC_ETF_BASE,
[GOYA_ETF_TPC0_EML] = mmTPC0_EML_ETF_BASE,
[GOYA_ETF_TPC1_EML] = mmTPC1_EML_ETF_BASE,
[GOYA_ETF_TPC2_EML] = mmTPC2_EML_ETF_BASE,
[GOYA_ETF_TPC3_EML] = mmTPC3_EML_ETF_BASE,
[GOYA_ETF_TPC4_EML] = mmTPC4_EML_ETF_BASE,
[GOYA_ETF_TPC5_EML] = mmTPC5_EML_ETF_BASE,
[GOYA_ETF_TPC6_EML] = mmTPC6_EML_ETF_BASE,
[GOYA_ETF_TPC7_EML] = mmTPC7_EML_ETF_BASE
};
static u64 debug_funnel_regs[GOYA_FUNNEL_LAST + 1] = {
[GOYA_FUNNEL_CPU] = mmCPU_FUNNEL_BASE,
[GOYA_FUNNEL_DMA_CH_6_1] = mmDMA_CH_FUNNEL_6_1_BASE,
[GOYA_FUNNEL_DMA_MACRO_3_1] = mmDMA_MACRO_FUNNEL_3_1_BASE,
[GOYA_FUNNEL_MME0_RTR] = mmMME0_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_MME1_RTR] = mmMME1_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_MME2_RTR] = mmMME2_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_MME3_RTR] = mmMME3_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_MME4_RTR] = mmMME4_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_MME5_RTR] = mmMME5_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_PCIE] = mmPCIE_FUNNEL_BASE,
[GOYA_FUNNEL_PSOC] = mmPSOC_FUNNEL_BASE,
[GOYA_FUNNEL_TPC0_EML] = mmTPC0_EML_FUNNEL_BASE,
[GOYA_FUNNEL_TPC1_EML] = mmTPC1_EML_FUNNEL_BASE,
[GOYA_FUNNEL_TPC1_RTR] = mmTPC1_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_TPC2_EML] = mmTPC2_EML_FUNNEL_BASE,
[GOYA_FUNNEL_TPC2_RTR] = mmTPC2_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_TPC3_EML] = mmTPC3_EML_FUNNEL_BASE,
[GOYA_FUNNEL_TPC3_RTR] = mmTPC3_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_TPC4_EML] = mmTPC4_EML_FUNNEL_BASE,
[GOYA_FUNNEL_TPC4_RTR] = mmTPC4_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_TPC5_EML] = mmTPC5_EML_FUNNEL_BASE,
[GOYA_FUNNEL_TPC5_RTR] = mmTPC5_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_TPC6_EML] = mmTPC6_EML_FUNNEL_BASE,
[GOYA_FUNNEL_TPC6_RTR] = mmTPC6_RTR_FUNNEL_BASE,
[GOYA_FUNNEL_TPC7_EML] = mmTPC7_EML_FUNNEL_BASE
};
static u64 debug_bmon_regs[GOYA_BMON_LAST + 1] = {
[GOYA_BMON_CPU_RD] = mmCPU_RD_BMON_BASE,
[GOYA_BMON_CPU_WR] = mmCPU_WR_BMON_BASE,
[GOYA_BMON_DMA_CH_0_0] = mmDMA_CH_0_BMON_0_BASE,
[GOYA_BMON_DMA_CH_0_1] = mmDMA_CH_0_BMON_1_BASE,
[GOYA_BMON_DMA_CH_1_0] = mmDMA_CH_1_BMON_0_BASE,
[GOYA_BMON_DMA_CH_1_1] = mmDMA_CH_1_BMON_1_BASE,
[GOYA_BMON_DMA_CH_2_0] = mmDMA_CH_2_BMON_0_BASE,
[GOYA_BMON_DMA_CH_2_1] = mmDMA_CH_2_BMON_1_BASE,
[GOYA_BMON_DMA_CH_3_0] = mmDMA_CH_3_BMON_0_BASE,
[GOYA_BMON_DMA_CH_3_1] = mmDMA_CH_3_BMON_1_BASE,
[GOYA_BMON_DMA_CH_4_0] = mmDMA_CH_4_BMON_0_BASE,
[GOYA_BMON_DMA_CH_4_1] = mmDMA_CH_4_BMON_1_BASE,
[GOYA_BMON_DMA_MACRO_0] = mmDMA_MACRO_BMON_0_BASE,
[GOYA_BMON_DMA_MACRO_1] = mmDMA_MACRO_BMON_1_BASE,
[GOYA_BMON_DMA_MACRO_2] = mmDMA_MACRO_BMON_2_BASE,
[GOYA_BMON_DMA_MACRO_3] = mmDMA_MACRO_BMON_3_BASE,
[GOYA_BMON_DMA_MACRO_4] = mmDMA_MACRO_BMON_4_BASE,
[GOYA_BMON_DMA_MACRO_5] = mmDMA_MACRO_BMON_5_BASE,
[GOYA_BMON_DMA_MACRO_6] = mmDMA_MACRO_BMON_6_BASE,
[GOYA_BMON_DMA_MACRO_7] = mmDMA_MACRO_BMON_7_BASE,
[GOYA_BMON_MME1_SBA_0] = mmMME1_SBA_BMON0_BASE,
[GOYA_BMON_MME1_SBA_1] = mmMME1_SBA_BMON1_BASE,
[GOYA_BMON_MME3_SBB_0] = mmMME3_SBB_BMON0_BASE,
[GOYA_BMON_MME3_SBB_1] = mmMME3_SBB_BMON1_BASE,
[GOYA_BMON_MME4_WACS2_0] = mmMME4_WACS2_BMON0_BASE,
[GOYA_BMON_MME4_WACS2_1] = mmMME4_WACS2_BMON1_BASE,
[GOYA_BMON_MME4_WACS2_2] = mmMME4_WACS2_BMON2_BASE,
[GOYA_BMON_MME4_WACS_0] = mmMME4_WACS_BMON0_BASE,
[GOYA_BMON_MME4_WACS_1] = mmMME4_WACS_BMON1_BASE,
[GOYA_BMON_MME4_WACS_2] = mmMME4_WACS_BMON2_BASE,
[GOYA_BMON_MME4_WACS_3] = mmMME4_WACS_BMON3_BASE,
[GOYA_BMON_MME4_WACS_4] = mmMME4_WACS_BMON4_BASE,
[GOYA_BMON_MME4_WACS_5] = mmMME4_WACS_BMON5_BASE,
[GOYA_BMON_MME4_WACS_6] = mmMME4_WACS_BMON6_BASE,
[GOYA_BMON_MMU_0] = mmMMU_BMON_0_BASE,
[GOYA_BMON_MMU_1] = mmMMU_BMON_1_BASE,
[GOYA_BMON_PCIE_MSTR_RD] = mmPCIE_BMON_MSTR_RD_BASE,
[GOYA_BMON_PCIE_MSTR_WR] = mmPCIE_BMON_MSTR_WR_BASE,
[GOYA_BMON_PCIE_SLV_RD] = mmPCIE_BMON_SLV_RD_BASE,
[GOYA_BMON_PCIE_SLV_WR] = mmPCIE_BMON_SLV_WR_BASE,
[GOYA_BMON_TPC0_EML_0] = mmTPC0_EML_BUSMON_0_BASE,
[GOYA_BMON_TPC0_EML_1] = mmTPC0_EML_BUSMON_1_BASE,
[GOYA_BMON_TPC0_EML_2] = mmTPC0_EML_BUSMON_2_BASE,
[GOYA_BMON_TPC0_EML_3] = mmTPC0_EML_BUSMON_3_BASE,
[GOYA_BMON_TPC1_EML_0] = mmTPC1_EML_BUSMON_0_BASE,
[GOYA_BMON_TPC1_EML_1] = mmTPC1_EML_BUSMON_1_BASE,
[GOYA_BMON_TPC1_EML_2] = mmTPC1_EML_BUSMON_2_BASE,
[GOYA_BMON_TPC1_EML_3] = mmTPC1_EML_BUSMON_3_BASE,
[GOYA_BMON_TPC2_EML_0] = mmTPC2_EML_BUSMON_0_BASE,
[GOYA_BMON_TPC2_EML_1] = mmTPC2_EML_BUSMON_1_BASE,
[GOYA_BMON_TPC2_EML_2] = mmTPC2_EML_BUSMON_2_BASE,
[GOYA_BMON_TPC2_EML_3] = mmTPC2_EML_BUSMON_3_BASE,
[GOYA_BMON_TPC3_EML_0] = mmTPC3_EML_BUSMON_0_BASE,
[GOYA_BMON_TPC3_EML_1] = mmTPC3_EML_BUSMON_1_BASE,
[GOYA_BMON_TPC3_EML_2] = mmTPC3_EML_BUSMON_2_BASE,
[GOYA_BMON_TPC3_EML_3] = mmTPC3_EML_BUSMON_3_BASE,
[GOYA_BMON_TPC4_EML_0] = mmTPC4_EML_BUSMON_0_BASE,
[GOYA_BMON_TPC4_EML_1] = mmTPC4_EML_BUSMON_1_BASE,
[GOYA_BMON_TPC4_EML_2] = mmTPC4_EML_BUSMON_2_BASE,
[GOYA_BMON_TPC4_EML_3] = mmTPC4_EML_BUSMON_3_BASE,
[GOYA_BMON_TPC5_EML_0] = mmTPC5_EML_BUSMON_0_BASE,
[GOYA_BMON_TPC5_EML_1] = mmTPC5_EML_BUSMON_1_BASE,
[GOYA_BMON_TPC5_EML_2] = mmTPC5_EML_BUSMON_2_BASE,
[GOYA_BMON_TPC5_EML_3] = mmTPC5_EML_BUSMON_3_BASE,
[GOYA_BMON_TPC6_EML_0] = mmTPC6_EML_BUSMON_0_BASE,
[GOYA_BMON_TPC6_EML_1] = mmTPC6_EML_BUSMON_1_BASE,
[GOYA_BMON_TPC6_EML_2] = mmTPC6_EML_BUSMON_2_BASE,
[GOYA_BMON_TPC6_EML_3] = mmTPC6_EML_BUSMON_3_BASE,
[GOYA_BMON_TPC7_EML_0] = mmTPC7_EML_BUSMON_0_BASE,
[GOYA_BMON_TPC7_EML_1] = mmTPC7_EML_BUSMON_1_BASE,
[GOYA_BMON_TPC7_EML_2] = mmTPC7_EML_BUSMON_2_BASE,
[GOYA_BMON_TPC7_EML_3] = mmTPC7_EML_BUSMON_3_BASE
};
static u64 debug_spmu_regs[GOYA_SPMU_LAST + 1] = {
[GOYA_SPMU_DMA_CH_0_CS] = mmDMA_CH_0_CS_SPMU_BASE,
[GOYA_SPMU_DMA_CH_1_CS] = mmDMA_CH_1_CS_SPMU_BASE,
[GOYA_SPMU_DMA_CH_2_CS] = mmDMA_CH_2_CS_SPMU_BASE,
[GOYA_SPMU_DMA_CH_3_CS] = mmDMA_CH_3_CS_SPMU_BASE,
[GOYA_SPMU_DMA_CH_4_CS] = mmDMA_CH_4_CS_SPMU_BASE,
[GOYA_SPMU_DMA_MACRO_CS] = mmDMA_MACRO_CS_SPMU_BASE,
[GOYA_SPMU_MME1_SBA] = mmMME1_SBA_SPMU_BASE,
[GOYA_SPMU_MME3_SBB] = mmMME3_SBB_SPMU_BASE,
[GOYA_SPMU_MME4_WACS2] = mmMME4_WACS2_SPMU_BASE,
[GOYA_SPMU_MME4_WACS] = mmMME4_WACS_SPMU_BASE,
[GOYA_SPMU_MMU_CS] = mmMMU_CS_SPMU_BASE,
[GOYA_SPMU_PCIE] = mmPCIE_SPMU_BASE,
[GOYA_SPMU_TPC0_EML] = mmTPC0_EML_SPMU_BASE,
[GOYA_SPMU_TPC1_EML] = mmTPC1_EML_SPMU_BASE,
[GOYA_SPMU_TPC2_EML] = mmTPC2_EML_SPMU_BASE,
[GOYA_SPMU_TPC3_EML] = mmTPC3_EML_SPMU_BASE,
[GOYA_SPMU_TPC4_EML] = mmTPC4_EML_SPMU_BASE,
[GOYA_SPMU_TPC5_EML] = mmTPC5_EML_SPMU_BASE,
[GOYA_SPMU_TPC6_EML] = mmTPC6_EML_SPMU_BASE,
[GOYA_SPMU_TPC7_EML] = mmTPC7_EML_SPMU_BASE
};
static int goya_coresight_timeout(struct hl_device *hdev, u64 addr,
int position, bool up)
{
int rc;
u32 val, timeout_usec;
if (hdev->pldm)
timeout_usec = GOYA_PLDM_CORESIGHT_TIMEOUT_USEC;
else
timeout_usec = CORESIGHT_TIMEOUT_USEC;
rc = hl_poll_timeout(
hdev,
addr,
val,
up ? val & BIT(position) : !(val & BIT(position)),
1000,
timeout_usec);
if (rc) {
dev_err(hdev->dev,
"Timeout while waiting for coresight, addr: 0x%llx, position: %d, up: %d\n",
addr, position, up);
return -EFAULT;
}
return 0;
}
static int goya_config_stm(struct hl_device *hdev,
struct hl_debug_params *params)
{
struct hl_debug_params_stm *input;
u64 base_reg;
int rc;
if (params->reg_idx >= ARRAY_SIZE(debug_stm_regs)) {
dev_err(hdev->dev, "Invalid register index in STM\n");
return -EINVAL;
}
base_reg = debug_stm_regs[params->reg_idx] - CFG_BASE;
WREG32(base_reg + 0xFB0, CORESIGHT_UNLOCK);
if (params->enable) {
input = params->input;
if (!input)
return -EINVAL;
WREG32(base_reg + 0xE80, 0x80004);
WREG32(base_reg + 0xD64, 7);
WREG32(base_reg + 0xD60, 0);
WREG32(base_reg + 0xD00, lower_32_bits(input->he_mask));
WREG32(base_reg + 0xD20, lower_32_bits(input->sp_mask));
WREG32(base_reg + 0xD60, 1);
WREG32(base_reg + 0xD00, upper_32_bits(input->he_mask));
WREG32(base_reg + 0xD20, upper_32_bits(input->sp_mask));
WREG32(base_reg + 0xE70, 0x10);
WREG32(base_reg + 0xE60, 0);
WREG32(base_reg + 0xE64, 0x420000);
WREG32(base_reg + 0xE00, 0xFFFFFFFF);
WREG32(base_reg + 0xE20, 0xFFFFFFFF);
WREG32(base_reg + 0xEF4, input->id);
WREG32(base_reg + 0xDF4, 0x80);
WREG32(base_reg + 0xE8C, input->frequency);
WREG32(base_reg + 0xE90, 0x7FF);
WREG32(base_reg + 0xE80, 0x7 | (input->id << 16));
} else {
WREG32(base_reg + 0xE80, 4);
WREG32(base_reg + 0xD64, 0);
WREG32(base_reg + 0xD60, 1);
WREG32(base_reg + 0xD00, 0);
WREG32(base_reg + 0xD20, 0);
WREG32(base_reg + 0xD60, 0);
WREG32(base_reg + 0xE20, 0);
WREG32(base_reg + 0xE00, 0);
WREG32(base_reg + 0xDF4, 0x80);
WREG32(base_reg + 0xE70, 0);
WREG32(base_reg + 0xE60, 0);
WREG32(base_reg + 0xE64, 0);
WREG32(base_reg + 0xE8C, 0);
rc = goya_coresight_timeout(hdev, base_reg + 0xE80, 23, false);
if (rc) {
dev_err(hdev->dev,
"Failed to disable STM on timeout, error %d\n",
rc);
return rc;
}
WREG32(base_reg + 0xE80, 4);
}
return 0;
}
static int goya_config_etf(struct hl_device *hdev,
struct hl_debug_params *params)
{
struct hl_debug_params_etf *input;
u64 base_reg;
u32 val;
int rc;
if (params->reg_idx >= ARRAY_SIZE(debug_etf_regs)) {
dev_err(hdev->dev, "Invalid register index in ETF\n");
return -EINVAL;
}
base_reg = debug_etf_regs[params->reg_idx] - CFG_BASE;
WREG32(base_reg + 0xFB0, CORESIGHT_UNLOCK);
val = RREG32(base_reg + 0x304);
val |= 0x1000;
WREG32(base_reg + 0x304, val);
val |= 0x40;
WREG32(base_reg + 0x304, val);
rc = goya_coresight_timeout(hdev, base_reg + 0x304, 6, false);
if (rc) {
dev_err(hdev->dev,
"Failed to %s ETF on timeout, error %d\n",
params->enable ? "enable" : "disable", rc);
return rc;
}
rc = goya_coresight_timeout(hdev, base_reg + 0xC, 2, true);
if (rc) {
dev_err(hdev->dev,
"Failed to %s ETF on timeout, error %d\n",
params->enable ? "enable" : "disable", rc);
return rc;
}
WREG32(base_reg + 0x20, 0);
if (params->enable) {
input = params->input;
if (!input)
return -EINVAL;
WREG32(base_reg + 0x34, 0x3FFC);
WREG32(base_reg + 0x28, input->sink_mode);
WREG32(base_reg + 0x304, 0x4001);
WREG32(base_reg + 0x308, 0xA);
WREG32(base_reg + 0x20, 1);
} else {
WREG32(base_reg + 0x34, 0);
WREG32(base_reg + 0x28, 0);
WREG32(base_reg + 0x304, 0);
}
return 0;
}
static int goya_etr_validate_address(struct hl_device *hdev, u64 addr,
u32 size)
{
struct asic_fixed_properties *prop = &hdev->asic_prop;
u64 range_start, range_end;
if (hdev->mmu_enable) {
range_start = prop->va_space_dram_start_address;
range_end = prop->va_space_dram_end_address;
} else {
range_start = prop->dram_user_base_address;
range_end = prop->dram_end_address;
}
return hl_mem_area_inside_range(addr, size, range_start, range_end);
}
static int goya_config_etr(struct hl_device *hdev,
struct hl_debug_params *params)
{
struct hl_debug_params_etr *input;
u64 base_reg = mmPSOC_ETR_BASE - CFG_BASE;
u32 val;
int rc;
WREG32(base_reg + 0xFB0, CORESIGHT_UNLOCK);
val = RREG32(base_reg + 0x304);
val |= 0x1000;
WREG32(base_reg + 0x304, val);
val |= 0x40;
WREG32(base_reg + 0x304, val);
rc = goya_coresight_timeout(hdev, base_reg + 0x304, 6, false);
if (rc) {
dev_err(hdev->dev, "Failed to %s ETR on timeout, error %d\n",
params->enable ? "enable" : "disable", rc);
return rc;
}
rc = goya_coresight_timeout(hdev, base_reg + 0xC, 2, true);
if (rc) {
dev_err(hdev->dev, "Failed to %s ETR on timeout, error %d\n",
params->enable ? "enable" : "disable", rc);
return rc;
}
WREG32(base_reg + 0x20, 0);
if (params->enable) {
input = params->input;
if (!input)
return -EINVAL;
if (input->buffer_size == 0) {
dev_err(hdev->dev,
"ETR buffer size should be bigger than 0\n");
return -EINVAL;
}
if (!goya_etr_validate_address(hdev,
input->buffer_address, input->buffer_size)) {
dev_err(hdev->dev, "buffer address is not valid\n");
return -EINVAL;
}
WREG32(base_reg + 0x34, 0x3FFC);
WREG32(base_reg + 0x4, input->buffer_size);
WREG32(base_reg + 0x28, input->sink_mode);
WREG32(base_reg + 0x110, 0x700);
WREG32(base_reg + 0x118,
lower_32_bits(input->buffer_address));
WREG32(base_reg + 0x11C,
upper_32_bits(input->buffer_address));
WREG32(base_reg + 0x304, 3);
WREG32(base_reg + 0x308, 0xA);
WREG32(base_reg + 0x20, 1);
} else {
WREG32(base_reg + 0x34, 0);
WREG32(base_reg + 0x4, 0x400);
WREG32(base_reg + 0x118, 0);
WREG32(base_reg + 0x11C, 0);
WREG32(base_reg + 0x308, 0);
WREG32(base_reg + 0x28, 0);
WREG32(base_reg + 0x304, 0);
if (params->output_size >= sizeof(u64)) {
u32 rwp, rwphi;
/*
* The trace buffer address is 40 bits wide. The end of
* the buffer is set in the RWP register (lower 32
* bits), and in the RWPHI register (upper 8 bits).
*/
rwp = RREG32(base_reg + 0x18);
rwphi = RREG32(base_reg + 0x3c) & 0xff;
*(u64 *) params->output = ((u64) rwphi << 32) | rwp;
}
}
return 0;
}
static int goya_config_funnel(struct hl_device *hdev,
struct hl_debug_params *params)
{
u64 base_reg;
if (params->reg_idx >= ARRAY_SIZE(debug_funnel_regs)) {
dev_err(hdev->dev, "Invalid register index in FUNNEL\n");
return -EINVAL;
}
base_reg = debug_funnel_regs[params->reg_idx] - CFG_BASE;
WREG32(base_reg + 0xFB0, CORESIGHT_UNLOCK);
WREG32(base_reg, params->enable ? 0x33F : 0);
return 0;
}
static int goya_config_bmon(struct hl_device *hdev,
struct hl_debug_params *params)
{
struct hl_debug_params_bmon *input;
u64 base_reg;
u32 pcie_base = 0;
if (params->reg_idx >= ARRAY_SIZE(debug_bmon_regs)) {
dev_err(hdev->dev, "Invalid register index in BMON\n");
return -EINVAL;
}
base_reg = debug_bmon_regs[params->reg_idx] - CFG_BASE;
WREG32(base_reg + 0x104, 1);
if (params->enable) {
input = params->input;
if (!input)
return -EINVAL;
WREG32(base_reg + 0x200, lower_32_bits(input->start_addr0));
WREG32(base_reg + 0x204, upper_32_bits(input->start_addr0));
WREG32(base_reg + 0x208, lower_32_bits(input->addr_mask0));
WREG32(base_reg + 0x20C, upper_32_bits(input->addr_mask0));
WREG32(base_reg + 0x240, lower_32_bits(input->start_addr1));
WREG32(base_reg + 0x244, upper_32_bits(input->start_addr1));
WREG32(base_reg + 0x248, lower_32_bits(input->addr_mask1));
WREG32(base_reg + 0x24C, upper_32_bits(input->addr_mask1));
WREG32(base_reg + 0x224, 0);
WREG32(base_reg + 0x234, 0);
WREG32(base_reg + 0x30C, input->bw_win);
WREG32(base_reg + 0x308, input->win_capture);
/* PCIE IF BMON bug WA */
if (params->reg_idx != GOYA_BMON_PCIE_MSTR_RD &&
params->reg_idx != GOYA_BMON_PCIE_MSTR_WR &&
params->reg_idx != GOYA_BMON_PCIE_SLV_RD &&
params->reg_idx != GOYA_BMON_PCIE_SLV_WR)
pcie_base = 0xA000000;
WREG32(base_reg + 0x700, pcie_base | 0xB00 | (input->id << 12));
WREG32(base_reg + 0x708, pcie_base | 0xA00 | (input->id << 12));
WREG32(base_reg + 0x70C, pcie_base | 0xC00 | (input->id << 12));
WREG32(base_reg + 0x100, 0x11);
WREG32(base_reg + 0x304, 0x1);
} else {
WREG32(base_reg + 0x200, 0);
WREG32(base_reg + 0x204, 0);
WREG32(base_reg + 0x208, 0xFFFFFFFF);
WREG32(base_reg + 0x20C, 0xFFFFFFFF);
WREG32(base_reg + 0x240, 0);
WREG32(base_reg + 0x244, 0);
WREG32(base_reg + 0x248, 0xFFFFFFFF);
WREG32(base_reg + 0x24C, 0xFFFFFFFF);
WREG32(base_reg + 0x224, 0xFFFFFFFF);
WREG32(base_reg + 0x234, 0x1070F);
WREG32(base_reg + 0x30C, 0);
WREG32(base_reg + 0x308, 0xFFFF);
WREG32(base_reg + 0x700, 0xA000B00);
WREG32(base_reg + 0x708, 0xA000A00);
WREG32(base_reg + 0x70C, 0xA000C00);
WREG32(base_reg + 0x100, 1);
WREG32(base_reg + 0x304, 0);
WREG32(base_reg + 0x104, 0);
}
return 0;
}
static int goya_config_spmu(struct hl_device *hdev,
struct hl_debug_params *params)
{
u64 base_reg;
struct hl_debug_params_spmu *input = params->input;
u64 *output;
u32 output_arr_len;
u32 events_num;
u32 overflow_idx;
u32 cycle_cnt_idx;
int i;
if (params->reg_idx >= ARRAY_SIZE(debug_spmu_regs)) {
dev_err(hdev->dev, "Invalid register index in SPMU\n");
return -EINVAL;
}
base_reg = debug_spmu_regs[params->reg_idx] - CFG_BASE;
if (params->enable) {
input = params->input;
if (!input)
return -EINVAL;
if (input->event_types_num < 3) {
dev_err(hdev->dev,
"not enough event types values for SPMU enable\n");
return -EINVAL;
}
if (input->event_types_num > SPMU_MAX_COUNTERS) {
dev_err(hdev->dev,
"too many event types values for SPMU enable\n");
return -EINVAL;
}
WREG32(base_reg + 0xE04, 0x41013046);
WREG32(base_reg + 0xE04, 0x41013040);
for (i = 0 ; i < input->event_types_num ; i++)
WREG32(base_reg + SPMU_EVENT_TYPES_OFFSET + i * 4,
input->event_types[i]);
WREG32(base_reg + 0xE04, 0x41013041);
WREG32(base_reg + 0xC00, 0x8000003F);
} else {
output = params->output;
output_arr_len = params->output_size / 8;
events_num = output_arr_len - 2;
overflow_idx = output_arr_len - 2;
cycle_cnt_idx = output_arr_len - 1;
if (!output)
return -EINVAL;
if (output_arr_len < 3) {
dev_err(hdev->dev,
"not enough values for SPMU disable\n");
return -EINVAL;
}
if (events_num > SPMU_MAX_COUNTERS) {
dev_err(hdev->dev,
"too many events values for SPMU disable\n");
return -EINVAL;
}
WREG32(base_reg + 0xE04, 0x41013040);
for (i = 0 ; i < events_num ; i++)
output[i] = RREG32(base_reg + i * 8);
output[overflow_idx] = RREG32(base_reg + 0xCC0);
output[cycle_cnt_idx] = RREG32(base_reg + 0xFC);
output[cycle_cnt_idx] <<= 32;
output[cycle_cnt_idx] |= RREG32(base_reg + 0xF8);
WREG32(base_reg + 0xCC0, 0);
}
return 0;
}
habanalabs: add new IOCTL for debug, tracing and profiling Habanalabs ASICs use the ARM coresight infrastructure to support debug, tracing and profiling of neural networks topologies. Because the coresight is configured using register writes and reads, and some of the registers hold sensitive information (e.g. the address in the device's DRAM where the trace data is written to), the user must go through the kernel driver to configure this mechanism. This patch implements the common code of the IOCTL and calls the ASIC-specific function for the actual H/W configuration. The IOCTL supports configuration of seven coresight components: ETR, ETF, STM, FUNNEL, BMON, SPMU and TIMESTAMP The user specifies which component he wishes to configure and provides a pointer to a structure (located in its process space) that contains the relevant configuration. The common code copies the relevant data from the user-space to kernel space and then calls the ASIC-specific function to do the H/W configuration. After the configuration is done, which is usually composed of several IOCTL calls depending on what the user wanted to trace, the user can start executing the topology. The trace data will be written to the user's area in the device's DRAM. After the tracing operation is complete, and user will call the IOCTL again to disable the tracing operation. The user also need to read values from registers for some of the components (e.g. the size of the trace data in the device's DRAM). In that case, the user will provide a pointer to an "output" structure in user-space, which the IOCTL code will fill according the to selected component. Signed-off-by: Omer Shpigelman <oshpigelman@habana.ai> Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
2019-04-01 13:31:22 -06:00
int goya_debug_coresight(struct hl_device *hdev, void *data)
{
struct hl_debug_params *params = data;
u32 val;
int rc = 0;
switch (params->op) {
case HL_DEBUG_OP_STM:
rc = goya_config_stm(hdev, params);
break;
case HL_DEBUG_OP_ETF:
rc = goya_config_etf(hdev, params);
break;
case HL_DEBUG_OP_ETR:
rc = goya_config_etr(hdev, params);
break;
case HL_DEBUG_OP_FUNNEL:
rc = goya_config_funnel(hdev, params);
break;
case HL_DEBUG_OP_BMON:
rc = goya_config_bmon(hdev, params);
break;
case HL_DEBUG_OP_SPMU:
rc = goya_config_spmu(hdev, params);
break;
case HL_DEBUG_OP_TIMESTAMP:
/* Do nothing as this opcode is deprecated */
break;
default:
dev_err(hdev->dev, "Unknown coresight id %d\n", params->op);
return -EINVAL;
}
/* Perform read from the device to flush all configuration */
val = RREG32(mmPCIE_DBI_DEVICE_ID_VENDOR_ID_REG);
return rc;
habanalabs: add new IOCTL for debug, tracing and profiling Habanalabs ASICs use the ARM coresight infrastructure to support debug, tracing and profiling of neural networks topologies. Because the coresight is configured using register writes and reads, and some of the registers hold sensitive information (e.g. the address in the device's DRAM where the trace data is written to), the user must go through the kernel driver to configure this mechanism. This patch implements the common code of the IOCTL and calls the ASIC-specific function for the actual H/W configuration. The IOCTL supports configuration of seven coresight components: ETR, ETF, STM, FUNNEL, BMON, SPMU and TIMESTAMP The user specifies which component he wishes to configure and provides a pointer to a structure (located in its process space) that contains the relevant configuration. The common code copies the relevant data from the user-space to kernel space and then calls the ASIC-specific function to do the H/W configuration. After the configuration is done, which is usually composed of several IOCTL calls depending on what the user wanted to trace, the user can start executing the topology. The trace data will be written to the user's area in the device's DRAM. After the tracing operation is complete, and user will call the IOCTL again to disable the tracing operation. The user also need to read values from registers for some of the components (e.g. the size of the trace data in the device's DRAM). In that case, the user will provide a pointer to an "output" structure in user-space, which the IOCTL code will fill according the to selected component. Signed-off-by: Omer Shpigelman <oshpigelman@habana.ai> Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
2019-04-01 13:31:22 -06:00
}
void goya_halt_coresight(struct hl_device *hdev)
{
struct hl_debug_params params = {};
int i, rc;
for (i = GOYA_ETF_FIRST ; i <= GOYA_ETF_LAST ; i++) {
params.reg_idx = i;
rc = goya_config_etf(hdev, &params);
if (rc)
dev_err(hdev->dev, "halt ETF failed, %d/%d\n", rc, i);
}
rc = goya_config_etr(hdev, &params);
if (rc)
dev_err(hdev->dev, "halt ETR failed, %d\n", rc);
}