在WIN从零开始在QMUE上添加一块自己的开发板（二）

文章目录

  • 一、前言
    • 往期回顾
  • 二、CPU虚拟化
    • （一）相关源码
    • （二）举个例子
    • （三）测试
  • 三、内存虚拟化
    • （一）相关源码
    • （二）举个例子
    • 测试
  • 参考资料

一、前言

笔者这篇博客作为平时学习时的笔记记录，如有不对还望指正，本博客大量借鉴资料，笔者只是拾人牙慧的小屁孩。
QEMU是一种通用的开源计算机仿真器和虚拟器。而QUME内置支持了一些开发板，我们可以基于这些内置的板子来做操作系统等软件的配置，但是实际市面上很多板子QUME中是没有提供支持的，这需要我们根据QUME的源码自定义一些开发板，然后再重新编译。

往期回顾

在WIN从零开始在QMUE上添加一块自己的开发板（一）

二、CPU虚拟化

（一）相关源码

QEMU中RISC-V CPU的支持
QOM的TYPE定义
target/riscv/cpu.h:

#define TYPE_RISCV_CPU "riscv-cpu"#define RISCV_CPU_TYPE_SUFFIX "-" TYPE_RISCV_CPU
#define RISCV_CPU_TYPE_NAME(name) (name RISCV_CPU_TYPE_SUFFIX)
#define CPU_RESOLVING_TYPE TYPE_RISCV_CPU#define TYPE_RISCV_CPU_ANY RISCV_CPU_TYPE_NAME("any")
#define TYPE_RISCV_CPU_BASE32 RISCV_CPU_TYPE_NAME("rv32")
#define TYPE_RISCV_CPU_BASE64 RISCV_CPU_TYPE_NAME("rv64")
#define TYPE_RISCV_CPU_IBEX RISCV_CPU_TYPE_NAME("lowrisc-ibex")
#define TYPE_RISCV_CPU_SIFIVE_E31 RISCV_CPU_TYPE_NAME("sifive-e31")
#define TYPE_RISCV_CPU_SIFIVE_E34 RISCV_CPU_TYPE_NAME("sifive-e34")
#define TYPE_RISCV_CPU_SIFIVE_E51 RISCV_CPU_TYPE_NAME("sifive-e51")
#define TYPE_RISCV_CPU_SIFIVE_U34 RISCV_CPU_TYPE_NAME("sifive-u34")
#define TYPE_RISCV_CPU_SIFIVE_U54 RISCV_CPU_TYPE_NAME("sifive-u54")

TYPE用于hash表的key（GHashTable.key），而hash表的值也就是ObjectClass、Object。

Class
RISCVCPUClass继承于CPUClass，然后继承于DeviceClass。
CPUClass中有很多接口:

struct CPUClass {DeviceClass parent_class;ObjectClass *(*class_by_name)();void (*parse_features)();int reset_dump_flags;bool (*has_work)();bool (*virtio_is_big_endian)();int (*memory_rw_debug)();void (*dump_state)();GuestPanicInformation* (*get_crash_info)();void (*dump_statistics)();int64_t (*get_arch_id)();bool (*get_paging_enabled)();void (*get_memory_mapping)();void (*set_pc)();hwaddr (*get_phys_page_debug)();hwaddr (*get_phys_page_attrs_debug)();int (*asidx_from_attrs)();int (*gdb_read_register)();int (*gdb_write_register)();int (*write_elf64_note)();int (*write_elf64_qemunote)();int (*write_elf32_note)();int (*write_elf32_qemunote)();const VMStateDescription *vmsd;const char *gdb_core_xml_file;gchar * (*gdb_arch_name)();const char * (*gdb_get_dynamic_xml)();void (*disas_set_info)();const char *deprecation_note;int gdb_num_core_regs;bool gdb_stop_before_watchpoint;struct AccelCPUClass *accel_cpu;struct TCGCPUOps *tcg_ops;
};

我们着重看一下struct TCGCPUOps：

struct TCGCPUOps {void (*initialize)(void);void (*synchronize_from_tb)(CPUState *cpu, const TranslationBlock *tb);void (*cpu_exec_enter)(CPUState *cpu);void (*cpu_exec_exit)(CPUState *cpu);bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);void (*do_interrupt)(CPUState *cpu);bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,MMUAccessType access_type, int mmu_idx,bool probe, uintptr_t retaddr);void (*debug_excp_handler)(CPUState *cpu);
#ifdef NEED_CPU_H
#ifdef CONFIG_SOFTMMUvoid (*do_transaction_failed)(CPUState *cpu, hwaddr physaddr, vaddr addr,unsigned size, MMUAccessType access_type,int mmu_idx, MemTxAttrs attrs,MemTxResult response, uintptr_t retaddr);void (*do_unaligned_access)(CPUState *cpu, vaddr addr,MMUAccessType access_type,int mmu_idx, uintptr_t retaddr);vaddr (*adjust_watchpoint_address)(CPUState *cpu, vaddr addr, int len);bool (*debug_check_watchpoint)(CPUState *cpu, CPUWatchpoint *wp);bool (*io_recompile_replay_branch)(CPUState *cpu,const TranslationBlock *tb);
#endif /* CONFIG_SOFTMMU */
#endif /* NEED_CPU_H */
};

可以看见里面有很多CPU运行时的接口。
在不同架构中，我们进行相应的实现，对于RISCVCPU，已经有相应的实现了：

static struct TCGCPUOps riscv_tcg_ops = {.initialize = riscv_translate_init,.synchronize_from_tb = riscv_cpu_synchronize_from_tb,.cpu_exec_interrupt = riscv_cpu_exec_interrupt,.tlb_fill = riscv_cpu_tlb_fill,
#ifndef CONFIG_USER_ONLY.do_interrupt = riscv_cpu_do_interrupt,.do_transaction_failed = riscv_cpu_do_transaction_failed,.do_unaligned_access = riscv_cpu_do_unaligned_access,
#endif /* !CONFIG_USER_ONLY */
};

Object
对于CPU中寄存器的定义都在Object中：

struct CPURISCVState {target_ulong gpr[32];uint64_t fpr[32]; /* assume both F and D extensions *//* vector coprocessor state. */uint64_t vreg[32 * RV_VLEN_MAX / 64] QEMU_ALIGNED(16);//vector regtarget_ulong pc;target_ulong misa;uint32_t features;/* Hypervisor CSRs *//* Virtual CSRs *//* HS Backup CSRs *//* temporary htif regs *//* physical memory protection *//* machine specific rdtime callback *//* True if in debugger mode. */bool debugger;float_status fp_status;/* Fields from here on are preserved across CPU reset. */QEMUTimer *timer; /* Internal timer */
};

实例化

	struct RISCVCPU {/*< private >*/CPUState parent_obj;/*< public >*/CPUNegativeOffsetState neg;CPURISCVState env;char *dyn_csr_xml;/* Configuration Settings */struct {……} cfg;
};

RISCV CPU TypeInfo注册：

.instance_init = riscv_cpu_init,
.class_init = riscv_cpu_class_init,

特殊的CPU使用特殊的函数进行实例

.instance_init = rvxx_sifive_e_cpu_init,
.class_init = riscv_cpu_class_init,

在特殊的函数中，将会针对不同CPU的特性进行个性化实例：

static void rvxx_sifive_e_cpu_init(Object *obj)
{CPURISCVState *env = &RISCV_CPU(obj)->env;set_misa(env, RVXLEN | RVI | RVM | RVA | RVC | RVU);set_priv_version(env, PRIV_VERSION_1_10_0);set_resetvec(env, 0x1004);qdev_prop_set_bit(DEVICE(obj), "mmu", false);
}

（二）举个例子

我们为之前创建的开发板增加CPU。
我们去target\riscv\cpu-qom.h，添加一个我们自己的CPU：

...
#define RISCV_CPU_TYPE_SUFFIX "-" TYPE_RISCV_CPU
#define RISCV_CPU_TYPE_NAME(name) (name RISCV_CPU_TYPE_SUFFIX)#define TYPE_RISCV_CPU_ANY              RISCV_CPU_TYPE_NAME("any")
#define TYPE_RISCV_CPU_MAX              RISCV_CPU_TYPE_NAME("max")
#define TYPE_RISCV_CPU_BASE32           RISCV_CPU_TYPE_NAME("rv32")
#define TYPE_RISCV_CPU_BASE64           RISCV_CPU_TYPE_NAME("rv64")
#define TYPE_RISCV_CPU_BASE128          RISCV_CPU_TYPE_NAME("x-rv128")
#define TYPE_RISCV_CPU_IBEX             RISCV_CPU_TYPE_NAME("lowrisc-ibex")
#define TYPE_RISCV_CPU_SHAKTI_C         RISCV_CPU_TYPE_NAME("shakti-c")
#define TYPE_RISCV_CPU_SIFIVE_E31       RISCV_CPU_TYPE_NAME("sifive-e31")
#define TYPE_RISCV_CPU_SIFIVE_E34       RISCV_CPU_TYPE_NAME("sifive-e34")
#define TYPE_RISCV_CPU_SIFIVE_E51       RISCV_CPU_TYPE_NAME("sifive-e51")
#define TYPE_RISCV_CPU_SIFIVE_U34       RISCV_CPU_TYPE_NAME("sifive-u34")
#define TYPE_RISCV_CPU_SIFIVE_U54       RISCV_CPU_TYPE_NAME("sifive-u54")
#define TYPE_RISCV_CPU_THEAD_C906       RISCV_CPU_TYPE_NAME("thead-c906")
#define TYPE_RISCV_CPU_VEYRON_V1        RISCV_CPU_TYPE_NAME("veyron-v1")
#define TYPE_RISCV_CPU_HOST             RISCV_CPU_TYPE_NAME("host")
/* 添加自己的CPU */
#define TYPE_RISCV_CPU_NUCLEI_N600       RISCV_CPU_TYPE_NAME("nuclei-n600")

并在./target/riscv/cpu.c中，添加其初始化函数：

#if defined(TARGET_RISCV32)
/* 自己的CPU */
static void rv32_nuclei_n_cpu_init(Object *obj)
{CPURISCVState *env = &RISCV_CPU(obj)->env;RISCVCPU *cpu = RISCV_CPU(obj);riscv_cpu_set_misa(env, MXL_RV32, RVI | RVM | RVA | RVC | RVF | RVD | RVU);env->priv_ver=  PRIV_VERSION_1_10_0;#ifndef CONFIG_USER_ONLYset_satp_mode_max_supported(cpu, VM_1_10_MBARE);
#endif/* inherited from parent obj via riscv_cpu_init() */cpu->cfg.ext_zifencei = true;cpu->cfg.ext_zicsr = true;cpu->cfg.pmp = true;
}
#endif

并在riscv_cpu_type_infos中添加DEFINE

static const TypeInfo riscv_cpu_type_infos[] = {{.name = TYPE_RISCV_CPU,.parent = TYPE_CPU,.instance_size = sizeof(RISCVCPU),.instance_align = __alignof(RISCVCPU),.instance_init = riscv_cpu_init,.instance_post_init = riscv_cpu_post_init,.abstract = true,.class_size = sizeof(RISCVCPUClass),.class_init = riscv_cpu_class_init,},{.name = TYPE_RISCV_DYNAMIC_CPU,.parent = TYPE_RISCV_CPU,.abstract = true,},DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_ANY,      riscv_any_cpu_init),DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_MAX,      riscv_max_cpu_init),
#if defined(TARGET_RISCV32)DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_BASE32,   rv32_base_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_IBEX,             rv32_ibex_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_SIFIVE_E31,       rv32_sifive_e_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_SIFIVE_E34,       rv32_imafcu_nommu_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_SIFIVE_U34,       rv32_sifive_u_cpu_init),/* 自己的CPU */DEFINE_CPU(TYPE_RISCV_CPU_NUCLEI_N600,      rv32_nuclei_n_cpu_init),#elif defined(TARGET_RISCV64)DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_BASE64,   rv64_base_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_SIFIVE_E51,       rv64_sifive_e_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_SIFIVE_U54,       rv64_sifive_u_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_SHAKTI_C,         rv64_sifive_u_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C906,       rv64_thead_c906_cpu_init),DEFINE_CPU(TYPE_RISCV_CPU_VEYRON_V1,        rv64_veyron_v1_cpu_init),DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_BASE128,  rv128_base_cpu_init),
#endif
};

接着我们回到我们的开发板文件，
这里讲个小技巧，也就是从include\hw\riscv\virt.h（官方虚拟开发板的例程）看相关的引用，从而确定相关头文件的所在位置和相关代码的实现。
在我们的nuclei_n.h中（请注意，为了方便理解，这里对往期变量及函数的名字进行了更改）：
引入头文件

#include "hw/riscv/riscv_hart.h"

并在SOC中添加CPU

/* CPU 定义 */
#define NUCLEI_N_CPU TYPE_RISCV_CPU_NUCLEI_N600typedef struct NucLeiNSoCState
{/*< private >*/SysBusDevice parent_obj;/*< public >*/RISCVHartArrayState cpus;
} NucLeiNSoCState;

以及一些TYPE：

#define NUCLEI_N_CPU 						TYPE_RISCV_CPU_NUCLEI_N600

之后我们在Machine的Class里面增加最小CPU个数和默认CPU的TYPE：

static void nuclei_machine_class_init(ObjectClass *oc, void *data)
{qemu_log(">>nuclei_machine_class_init \n");MachineClass *mc = MACHINE_CLASS(oc);mc->desc = "Nuclei MCU 200T FPGA Evaluation Kit";mc->init = nuclei_mcu_machine_init;mc->max_cpus = 1;mc->default_cpu_type = NUCLEI_N_CPU;
}

我们在SOC实例初始化函数中，对CPU初始化：

static void nuclei_n_soc_instance_init(Object *obj)
{qemu_log(">>nuclei_n_soc_instance_init \n");NucLeiNSoCState *s = NUCLEI_N_SOC(obj);object_initialize_child(obj, "cpus", &s->cpus, TYPE_RISCV_HART_ARRAY);  //初始化CPU
}

并在SOC实现中（nuclei_n_soc_realize）进行CPU的实现：

static void nuclei_n_soc_realize(DeviceState *dev, Error **errp)
{qemu_log(">>nuclei_n_soc_realize \n");MachineState *ms = MACHINE(qdev_get_machine());NucLeiNSoCState *s = NUCLEI_N_SOC(dev);object_property_set_str(OBJECT(&s->cpus), "cpu-type", ms->cpu_type, &error_abort);      object_property_set_int(OBJECT(&s->cpus), "num-harts", ms->smp.cpus, &error_abort);sysbus_realize(SYS_BUS_DEVICE(&s->cpus), &error_abort);  //CPU实例化
}

附上完整：

nuclei_n.h:

#include "hw/sysbus.h"
#include "hw/riscv/riscv_hart.h"#define TYPE_NUCLEI_N_SOC "riscv.nuclei.n.soc"
#define NUCLEI_N_SOC(obj) \OBJECT_CHECK(NucLeiNSoCState, (obj), TYPE_NUCLEI_N_SOC)/* CPU 定义 */
#define NUCLEI_N_CPU TYPE_RISCV_CPU_NUCLEI_N600typedef struct NucLeiNSoCState
{/*< private >*/SysBusDevice parent_obj;/*< public >*/RISCVHartArrayState cpus;
} NucLeiNSoCState;/* Machine state定义 */
#define TYPE_NUCLEI_MCU_FPGA_MACHINE MACHINE_TYPE_NAME("mcu_200t")
#define MCU_FPGA_MACHINE(obj) \OBJECT_CHECK(NucLeiNState, (obj), TYPE_NUCLEI_MCU_FPGA_MACHINE)typedef struct NucLeiNState
{/*< private >*/SysBusDevice parent_obj;/*< public >*/NucLeiNSoCState soc;
} NucLeiNState;

nuclei_n.c:

#include "qemu/osdep.h"
#include "qemu/log.h"
#include "qemu/error-report.h"
#include "qapi/error.h"
#include "hw/riscv/nuclei_n.h"
#include "hw/boards.h"static void nuclei_n_soc_instance_init(Object *obj)
{qemu_log(">>nuclei_n_soc_instance_init \n");NucLeiNSoCState *s = NUCLEI_N_SOC(obj);object_initialize_child(obj, "cpus", &s->cpus, TYPE_RISCV_HART_ARRAY);  //初始化CPU
}static void nuclei_n_soc_realize(DeviceState *dev, Error **errp)
{qemu_log(">>nuclei_n_soc_realize \n");MachineState *ms = MACHINE(qdev_get_machine());NucLeiNSoCState *s = NUCLEI_N_SOC(dev);object_property_set_str(OBJECT(&s->cpus), "cpu-type", ms->cpu_type, &error_abort);      object_property_set_int(OBJECT(&s->cpus), "num-harts", ms->smp.cpus, &error_abort);sysbus_realize(SYS_BUS_DEVICE(&s->cpus), &error_abort);            //CPU实例化
}
static void nuclei_n_soc_class_init(ObjectClass *oc, void *data)
{qemu_log(">>nuclei_n_soc_class_init \n");DeviceClass *dc = DEVICE_CLASS(oc);dc->realize = nuclei_n_soc_realize;dc->user_creatable = false;
}static const TypeInfo nuclei_n_soc_type_info = {.name = TYPE_NUCLEI_N_SOC,.parent = TYPE_DEVICE,.instance_size = sizeof(NucLeiNSoCState),.instance_init = nuclei_n_soc_instance_init,.class_init = nuclei_n_soc_class_init,
};
static void nuclei_n_soc_register_types(void)
{type_register_static(&nuclei_n_soc_type_info);
}
type_init(nuclei_n_soc_register_types)static void nuclei_mcu_machine_init(MachineState *machine)
{NucLeiNState *s = MCU_FPGA_MACHINE(machine);qemu_log(">>nuclei_mcu_machine_init \n");/* Initialize SOC */object_initialize_child(OBJECT(machine), "soc", &s->soc, TYPE_NUCLEI_N_SOC);qdev_realize(DEVICE(&s->soc), NULL, &error_abort);
}
static void nuclei_machine_instance_init(Object *obj)
{qemu_log(">>nuclei_machine_instance_init \n");
}
static void nuclei_machine_class_init(ObjectClass *oc, void *data)
{qemu_log(">>nuclei_machine_class_init \n");MachineClass *mc = MACHINE_CLASS(oc);mc->desc = "Nuclei MCU 200T FPGA Evaluation Kit";mc->init = nuclei_mcu_machine_init;mc->max_cpus = 1;mc->default_cpu_type = NUCLEI_N_CPU;
}static const TypeInfo nuclei_machine_typeinfo = {.name = TYPE_NUCLEI_MCU_FPGA_MACHINE,.parent = TYPE_MACHINE,.class_init = nuclei_machine_class_init,.instance_init = nuclei_machine_instance_init,.instance_size = sizeof(NucLeiNState),
};
static void nuclei_machine_init_register_types(void)
{type_register_static(&nuclei_machine_typeinfo);
}
type_init(nuclei_machine_init_register_types)

不用忘记在./target/riscv/cpu.c去定义CPU哦。

（三）测试

执行run.sh:

SHELL_FOLDER=$(cd "$(dirname "$0")";pwd)
$SHELL_FOLDER/output/qemu/qemu-system-riscv32.exe \
-M mcu_200t

得到以下效果：
CPU虚拟化

三、内存虚拟化

（一）相关源码

有了之前的开发经验，我们这次直接看相关API：

Types of regionsinitialize
RAMmemory_region_init_ram()
MMIOmemory_region_init_io()
ROMmemory_region_init_rom()
ROM_evicememory_region_init_rom_device()
IOMMU regionmemory_region_init_iommu()
containermemory_region_init()
aliasmemory_region_init_alias()
reservation regionmemory_region_init_io()

其次是添加硬件的地址和映射的地址长度的结构体：
（注意这里的长度不能为0，不然会报错）

typedef struct MemMapEntry {hwaddr base; //基址hwaddr size; //长度
} MemMapEntry;

然后是关于ROM的指令初始化：

/* reset vector */
uint32_t reset_vec[8] = {0x00000297, /* 1: auipc t0, %pcrel_hi(dtb) */0x02028593, /* addi a1, t0, %pcrel_lo(1b) */0xf1402573, /* csrr a0, mhartid */#if defined(TARGET_RISCV32)0x0182a283, /* lw t0, 24(t0) */#elif defined(TARGET_RISCV64)0x0182b283, /* ld t0, 24(t0) */#endif0x00028067, /* jr t0 */0x00000000,start_addr, /* start: .dword DRAM_BASE */0x00000000,
};
/* copy in the reset vector in little_endian byte order */
for (i = 0; i < sizeof(reset_vec) >> 2; i++)
{reset_vec[i] = cpu_to_le32(reset_vec[i]);
}
rom_add_blob_fixed_as("mrom.reset", reset_vec, sizeof(reset_vec),memmap[NUCLEI_N_ROM].base, &address_space_memory);
/* boot rom */
if (machine->kernel_filename)
{riscv_load_kernel(machine>kernel_filename, start_addr, NULL);
}

这种是手动执行指令初始化，当然，因为我们使用的RISCV架构，直接使用riscv_setup_rom_reset_vec进行指令初始化也是可以的。

这里我们再讲一下内存模拟的一个步骤：

  1. 执行初始化函数，例如ROM的就是memory_region_init_rom；
  2. 分配/挂载，memory_region_add_subregion，当然还会用上系统根节点获取get_system_memory；
  3. ROM设置（指令初始化）或者加载kernel

（二）举个例子

根据SOC指定地址编写MemMapEntry
MemMapEntry

enum
{NUCLEI_N_DEBUG,NUCLEI_N_ROM,NUCLEI_N_TIMER,NUCLEI_N_ECLIC,NUCLEI_N_GPIO,NUCLEI_N_UART0,NUCLEI_N_QSPI0,NUCLEI_N_PWM0,NUCLEI_N_UART1,NUCLEI_N_QSPI1,NUCLEI_N_PWM1,NUCLEI_N_QSPI2,NUCLEI_N_PWM2,NUCLEI_N_XIP,NUCLEI_N_DRAM,NUCLEI_N_ILM,NUCLEI_N_DLM
};static MemMapEntry nuclei_n_memmap[] = {[NUCLEI_N_DEBUG] 	= 	{0x0, 0x1000},[NUCLEI_N_ROM] 		= 	{0x1000, 0x1000},[NUCLEI_N_TIMER] 	= 	{0x2000000, 0x1000},[NUCLEI_N_ECLIC] 	= 	{0xc000000, 0x10000},[NUCLEI_N_GPIO] 	= 	{0x10012000, 0x1000},[NUCLEI_N_UART0] 	= 	{0x10013000, 0x1000},[NUCLEI_N_QSPI0] 	= 	{0x10014000, 0x1000},[NUCLEI_N_PWM0] 	= 	{0x10015000, 0x1000},[NUCLEI_N_UART1] 	= 	{0x10023000, 0x1000},[NUCLEI_N_QSPI1] 	= 	{0x10024000, 0x1000},[NUCLEI_N_PWM1] 	= 	{0x10025000, 0x1000},[NUCLEI_N_QSPI2] 	= 	{0x10034000, 0x1000},[NUCLEI_N_PWM2] 	= 	{0x10035000, 0x1000},[NUCLEI_N_XIP] 		= 	{0x20000000, 0x10000000},[NUCLEI_N_DRAM] 	= 	{0xa0000000, 0x0},[NUCLEI_N_ILM] 		= 	{0x80000000, 0x20000},[NUCLEI_N_DLM] 		= 	rub{0x90000000, 0x20000},
};

之后我们初始化ROM地址：

/* Internal ROM */memory_region_init_rom(&s->internal_rom, OBJECT(obj), "riscv.nuclei.n.irom", memmap[NUCLEI_N_ROM].size, &error_fatal);memory_region_add_subregion(sys_mem, memmap[NUCLEI_N_ROM].base, &s->internal_rom);

这里我们假设idlm和ROM都为Soc外设：
于是我们编写相关函数：

static void nuclei_n_soc_memory_create(Object *obj)
{NucLeiNSoCState *s = NUCLEI_N_SOC(obj);const MemMapEntry *memmap = nuclei_n_memmap;MemoryRegion *sys_mem = get_system_memory();/* Internal ROM */memory_region_init_rom(&s->internal_rom, OBJECT(obj), "riscv.nuclei.n.irom", memmap[NUCLEI_N_ROM].size, &error_fatal);memory_region_add_subregion(sys_mem, memmap[NUCLEI_N_ROM].base, &s->internal_rom);/* Initialize ilm dlm */memory_region_init_ram(&s->ilm, NULL, "riscv.nuclei.n.ilm", memmap[NUCLEI_N_ILM].size, &error_fatal);memory_region_add_subregion(sys_mem, memmap[NUCLEI_N_ILM].base, &s->ilm);memory_region_init_ram(&s->dlm, NULL, "riscv.nuclei.n.dlm", memmap[NUCLEI_N_DLM].size, &error_fatal);memory_region_add_subregion(sys_mem, memmap[NUCLEI_N_DLM].base, &s->dlm);/* SysTimer */create_unimplemented_device("riscv.nuclei.n.timer", memmap[NUCLEI_N_TIMER].base, memmap[NUCLEI_N_TIMER].size);/* Eclic */create_unimplemented_device("riscv.nuclei.n.eclic", memmap[NUCLEI_N_ECLIC].base, memmap[NUCLEI_N_ECLIC].size);/* GPIO */create_unimplemented_device("riscv.nuclei.n.gpio", memmap[NUCLEI_N_GPIO].base, memmap[NUCLEI_N_GPIO].size);
}

因为还没有实现一些设备，所以我们创建unimplemented设备来占用内存：

	 /* SysTimer */create_unimplemented_device("riscv.nuclei.n.timer", memmap[NUCLEI_N_TIMER].base, memmap[NUCLEI_N_TIMER].size);/* Eclic */create_unimplemented_device("riscv.nuclei.n.eclic", memmap[NUCLEI_N_ECLIC].base, memmap[NUCLEI_N_ECLIC].size);/* GPIO */create_unimplemented_device("riscv.nuclei.n.gpio", memmap[NUCLEI_N_GPIO].base, memmap[NUCLEI_N_GPIO].size);

这次我们把CPU的初始化和实例化也类似封装成一个函数：

static void nuclei_n_soc_cpu_create(Object *obj)
{MachineState *ms = MACHINE(qdev_get_machine());NucLeiNSoCState *s = NUCLEI_N_SOC(obj);object_initialize_child(obj, "cpus", &s->cpus, TYPE_RISCV_HART_ARRAY);  //初始化CPUobject_property_set_str(OBJECT(&s->cpus), "cpu-type", ms->cpu_type, &error_abort);      object_property_set_int(OBJECT(&s->cpus), "num-harts", ms->smp.cpus, &error_abort);sysbus_realize(SYS_BUS_DEVICE(&s->cpus), &error_abort);            		//CPU实例化
}

然后我们在nuclei_n_soc_instance_init中调用：

static void nuclei_n_soc_instance_init(Object *obj)
{/* SOC CPU */nuclei_n_soc_cpu_create(obj);/* SOC Memory */nuclei_n_soc_memory_create(obj);
}

其次是设置ROM和加载kernel，我们在整个Machine实例中进行初始化：

static void nuclei_mcu_machine_init(MachineState *machine)
{NucLeiNState *s = MCU_FPGA_MACHINE(machine);const MemMapEntry *memmap = nuclei_n_memmap;target_ulong start_addr;int i;/* Initialize SOC */object_initialize_child(OBJECT(machine), "soc", &s->soc, TYPE_NUCLEI_N_SOC);qdev_realize(DEVICE(&s->soc), NULL, &error_abort);//选择启动方式switch (s->msel){case MSEL_ILM:start_addr = memmap[NUCLEI_N_ILM].base;break;case MSEL_FLASH:start_addr = memmap[NUCLEI_N_XIP].base;break;case MSEL_FLASHXIP:start_addr = memmap[NUCLEI_N_XIP].base;break;case MSEL_DDR:start_addr = memmap[NUCLEI_N_DRAM].base;break;default:start_addr = memmap[NUCLEI_N_ILM].base;break;}/* reset vector */uint32_t reset_vec[8] = {0x00000297, /* 1:  auipc  t0, %pcrel_hi(dtb) */0x02028593, /*     addi   a1, t0, %pcrel_lo(1b) */0xf1402573, /*     csrr   a0, mhartid  */
#if defined(TARGET_RISCV32)0x0182a283, /*     lw     t0, 24(t0) */
#elif defined(TARGET_RISCV64)0x0182b283, /*     ld     t0, 24(t0) */
#endif0x00028067, /*     jr     t0 */0x00000000,start_addr, /* start: .dword DRAM_BASE */0x00000000,};/* copy in the reset vector in little_endian byte order */for (i = 0; i < sizeof(reset_vec) >> 2; i++){reset_vec[i] = cpu_to_le32(reset_vec[i]);}rom_add_blob_fixed_as("mrom.reset", reset_vec, sizeof(reset_vec),memmap[NUCLEI_N_ROM].base, &address_space_memory); //CPU初始化地址/* boot rom */if (machine->kernel_filename){riscv_load_kernel(machine, &s->soc.cpus, start_addr, true, NULL);	//将裸机代码加载到地址start_addr// riscv_load_kernel(machine->kernel_filename, start_addr, NULL);  }
}

测试

我们进行测试：
编译完成后运行程序并执行：

info mtree

测试结果

结尾附上完整代码：
nuclei_n.h：

#include "hw/boards.h"
#include "hw/riscv/riscv_hart.h"
#include "hw/sysbus.h"#define TYPE_NUCLEI_N_SOC "riscv.nuclei.n.soc"
#define NUCLEI_N_SOC(obj) \OBJECT_CHECK(NucLeiNSoCState, (obj), TYPE_NUCLEI_N_SOC)/* CPU 定义 */
#define NUCLEI_N_CPU TYPE_RISCV_CPU_NUCLEI_N600typedef struct NucLeiNSoCState
{/*< private >*/DeviceState parent_obj;/*< public >*/RISCVHartArrayState cpus;MemoryRegion internal_rom;MemoryRegion ilm;MemoryRegion dlm;MemoryRegion xip_mem;
} NucLeiNSoCState;/* Machine state定义 */
#define TYPE_NUCLEI_MCU_FPGA_MACHINE MACHINE_TYPE_NAME("mcu_200t")
#define MCU_FPGA_MACHINE(obj) \OBJECT_CHECK(NucLeiNState, (obj), TYPE_NUCLEI_MCU_FPGA_MACHINE)typedef struct NucLeiNState
{/*< private >*/MachineState parent;/*< public >*/NucLeiNSoCState soc;uint32_t msel;
} NucLeiNState;enum
{MSEL_ILM = 1,MSEL_FLASH = 2,MSEL_FLASHXIP = 3,MSEL_DDR = 4
};enum
{NUCLEI_N_DEBUG,NUCLEI_N_ROM,NUCLEI_N_TIMER,NUCLEI_N_ECLIC,NUCLEI_N_GPIO,NUCLEI_N_UART0,NUCLEI_N_QSPI0,NUCLEI_N_PWM0,NUCLEI_N_UART1,NUCLEI_N_QSPI1,NUCLEI_N_PWM1,NUCLEI_N_QSPI2,NUCLEI_N_PWM2,NUCLEI_N_XIP,NUCLEI_N_DRAM,NUCLEI_N_ILM,NUCLEI_N_DLM
};

nuclei_n.c：

#include "qemu/osdep.h"
#include "qemu/error-report.h"
#include "qapi/error.h"
#include "hw/riscv/nuclei_n.h"
#include "qapi/visitor.h"
#include "hw/boards.h"
#include "hw/loader.h"
#include "hw/sysbus.h"
#include "target/riscv/cpu.h"
#include "hw/misc/unimp.h"
#include "hw/riscv/riscv_hart.h"
#include "hw/riscv/boot.h"static MemMapEntry nuclei_n_memmap[] = {[NUCLEI_N_DEBUG] 	= 	{0x0, 0x1000},[NUCLEI_N_ROM] 		= 	{0x1000, 0x1000},[NUCLEI_N_TIMER] 	= 	{0x2000000, 0x1000},[NUCLEI_N_ECLIC] 	= 	{0xc000000, 0x10000},[NUCLEI_N_GPIO] 	= 	{0x10012000, 0x1000},[NUCLEI_N_UART0] 	= 	{0x10013000, 0x1000},[NUCLEI_N_QSPI0] 	= 	{0x10014000, 0x1000},[NUCLEI_N_PWM0] 	= 	{0x10015000, 0x1000},[NUCLEI_N_UART1] 	= 	{0x10023000, 0x1000},[NUCLEI_N_QSPI1] 	= 	{0x10024000, 0x1000},[NUCLEI_N_PWM1] 	= 	{0x10025000, 0x1000},[NUCLEI_N_QSPI2] 	= 	{0x10034000, 0x1000},[NUCLEI_N_PWM2] 	= 	{0x10035000, 0x1000},[NUCLEI_N_XIP] 		= 	{0x20000000, 0x10000000},[NUCLEI_N_DRAM] 	= 	{0xa0000000, 0x0},[NUCLEI_N_ILM] 		= 	{0x80000000, 0x20000},[NUCLEI_N_DLM] 		= 	{0x90000000, 0x20000},
};static void nuclei_n_soc_cpu_create(Object *obj)
{MachineState *ms = MACHINE(qdev_get_machine());NucLeiNSoCState *s = NUCLEI_N_SOC(obj);object_initialize_child(obj, "cpus", &s->cpus, TYPE_RISCV_HART_ARRAY);  //初始化CPUobject_property_set_str(OBJECT(&s->cpus), "cpu-type", ms->cpu_type, &error_abort);      object_property_set_int(OBJECT(&s->cpus), "num-harts", ms->smp.cpus, &error_abort);sysbus_realize(SYS_BUS_DEVICE(&s->cpus), &error_abort);            		//CPU实例化
}static void nuclei_n_soc_memory_create(Object *obj)
{NucLeiNSoCState *s = NUCLEI_N_SOC(obj);const MemMapEntry *memmap = nuclei_n_memmap;MemoryRegion *sys_mem = get_system_memory();/* Internal ROM */memory_region_init_rom(&s->internal_rom, OBJECT(obj), "riscv.nuclei.n.irom", memmap[NUCLEI_N_ROM].size, &error_fatal);memory_region_add_subregion(sys_mem, memmap[NUCLEI_N_ROM].base, &s->internal_rom);/* Initialize ilm dlm */memory_region_init_ram(&s->ilm, NULL, "riscv.nuclei.n.ilm", memmap[NUCLEI_N_ILM].size, &error_fatal);memory_region_add_subregion(sys_mem, memmap[NUCLEI_N_ILM].base, &s->ilm);memory_region_init_ram(&s->dlm, NULL, "riscv.nuclei.n.dlm", memmap[NUCLEI_N_DLM].size, &error_fatal);memory_region_add_subregion(sys_mem, memmap[NUCLEI_N_DLM].base, &s->dlm);/* SysTimer */create_unimplemented_device("riscv.nuclei.n.timer",memmap[NUCLEI_N_TIMER].base, memmap[NUCLEI_N_TIMER].size);/* Eclic */create_unimplemented_device("riscv.nuclei.n.eclic",memmap[NUCLEI_N_ECLIC].base, memmap[NUCLEI_N_ECLIC].size);/* GPIO */create_unimplemented_device("riscv.nuclei.n.gpio",memmap[NUCLEI_N_GPIO].base, memmap[NUCLEI_N_GPIO].size);
}static void nuclei_n_soc_instance_init(Object *obj)
{/* SOC CPU */nuclei_n_soc_cpu_create(obj);/* SOC Memory */nuclei_n_soc_memory_create(obj);
}static void nuclei_n_soc_class_init(ObjectClass *oc, void *data)
{DeviceClass *dc = DEVICE_CLASS(oc);dc->user_creatable = false;
}static const TypeInfo nuclei_n_soc_type_info = {.name = TYPE_NUCLEI_N_SOC,.parent = TYPE_DEVICE,.instance_size = sizeof(NucLeiNSoCState),.instance_init = nuclei_n_soc_instance_init,.class_init = nuclei_n_soc_class_init,
};
static void nuclei_n_soc_register_types(void)
{type_register_static(&nuclei_n_soc_type_info);
}
type_init(nuclei_n_soc_register_types)static void nuclei_mcu_machine_init(MachineState *machine)
{NucLeiNState *s = MCU_FPGA_MACHINE(machine);const MemMapEntry *memmap = nuclei_n_memmap;target_ulong start_addr;int i;/* Initialize SOC */object_initialize_child(OBJECT(machine), "soc", &s->soc, TYPE_NUCLEI_N_SOC);qdev_realize(DEVICE(&s->soc), NULL, &error_abort);//选择启动方式switch (s->msel){case MSEL_ILM:start_addr = memmap[NUCLEI_N_ILM].base;break;case MSEL_FLASH:start_addr = memmap[NUCLEI_N_XIP].base;break;case MSEL_FLASHXIP:start_addr = memmap[NUCLEI_N_XIP].base;break;case MSEL_DDR:start_addr = memmap[NUCLEI_N_DRAM].base;break;default:start_addr = memmap[NUCLEI_N_ILM].base;break;}/* reset vector */uint32_t reset_vec[8] = {0x00000297, /* 1:  auipc  t0, %pcrel_hi(dtb) */0x02028593, /*     addi   a1, t0, %pcrel_lo(1b) */0xf1402573, /*     csrr   a0, mhartid  */
#if defined(TARGET_RISCV32)0x0182a283, /*     lw     t0, 24(t0) */
#elif defined(TARGET_RISCV64)0x0182b283, /*     ld     t0, 24(t0) */
#endif0x00028067, /*     jr     t0 */0x00000000,start_addr, /* start: .dword DRAM_BASE */0x00000000,};/* copy in the reset vector in little_endian byte order */for (i = 0; i < sizeof(reset_vec) >> 2; i++){reset_vec[i] = cpu_to_le32(reset_vec[i]);}rom_add_blob_fixed_as("mrom.reset", reset_vec, sizeof(reset_vec),memmap[NUCLEI_N_ROM].base, &address_space_memory); //CPU初始化地址/* boot rom */if (machine->kernel_filename){riscv_load_kernel(machine, &s->soc.cpus, start_addr, true, NULL);	//将裸机代码加载到地址start_addr// riscv_load_kernel(machine->kernel_filename, start_addr, NULL);  }
}static void nuclei_machine_class_init(ObjectClass *oc, void *data)
{MachineClass *mc = MACHINE_CLASS(oc);mc->desc = "Nuclei MCU 200T FPGA Evaluation Kit";mc->init = nuclei_mcu_machine_init;mc->max_cpus = 1;mc->default_cpu_type = NUCLEI_N_CPU;
}static const TypeInfo nuclei_machine_typeinfo = {.name = TYPE_NUCLEI_MCU_FPGA_MACHINE,.parent = TYPE_MACHINE,.class_init = nuclei_machine_class_init,.instance_size = sizeof(NucLeiNState),
};
static void nuclei_machine_init_register_types(void)
{type_register_static(&nuclei_machine_typeinfo);
}
type_init(nuclei_machine_init_register_types)

参考资料

  1. [完结]从零开始的RISC-V模拟器开发·第一季·2021春季
  2. 新建quard-star开发板

本文来自互联网用户投稿，该文观点仅代表作者本人，不代表本站立场。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如若转载，请注明出处：http://www.rhkb.cn/news/242403.html

如若内容造成侵权/违法违规/事实不符，请联系长河编程网进行投诉反馈email:809451989@qq.com，一经查实，立即删除！

相关文章

雷盛红酒LEESON分享葡萄酒也有“社会责任感”？

雷盛红酒LEESON分享葡萄酒也有“社会责任感”？

葡萄酒文化从来都不仅仅是感官体验&#xff0c;一瓶佳酿的背后不但蕴含着风土人情、历史传承和文化交流&#xff0c;更反映了时代社会的变迁以及体现的社会责任意识。 目前葡萄酒生产商追求酒瓶越来越轻就是葡萄酒市场上的一个趋势&#xff0c;因为任何一个行业都在追求与世界共…
阅读更多...
c语言算法——大数相加

c语言算法——大数相加

C数据类型 类型与描述1基本数据类型 它们是算术类型&#xff0c;包括整型&#xff08;int&#xff09;、字符型&#xff08;char&#xff09;、浮点型&#xff08;float&#xff09;和双精度浮点型&#xff08;double&#xff09;。2枚举类型&#xff1a; 它们也是算术类型&am…
阅读更多...
Vue2的双向数据绑定

Vue2的双向数据绑定

Vue2的双向数据绑定 Observer&#xff1a;观察者&#xff0c;这里的主要工作是递归地监听对象上的所有属性&#xff0c;在属性值改变的时候&#xff0c;触发相应的watcher。 Watcher&#xff1a;订阅者&#xff0c;当监听的数据值修改时&#xff0c;执行响应的回调函数&#x…
阅读更多...
Spring：StopWatch

Spring：StopWatch

提示&#xff1a;文章写完后&#xff0c;目录可以自动生成&#xff0c;如何生成可参考右边的帮助文档 目录 一、输出总耗时 二、输出所有任务的耗时和占比 总结 提示&#xff1a;以下是本篇文章正文内容&#xff0c;下面案例可供参考 一、输出总耗时 public void stopWatc…
阅读更多...
ERP进出库+办公用品管理系统

ERP进出库+办公用品管理系统

系统架构 简介系统架构部分页面结构图UML逻辑图办公用品入出库 简介 本系统适用于ERP企业公司职员关于系统化的申请相关办公用品&#xff0c;提高整体系统整合行&#xff0c;加大上下级之间的联系&#xff0c;规避因人员过多&#xff0c;而浪费人力在简单重复的工作中&#xf…
阅读更多...
conda国内加速

conda国内加速

1、配置国内源 conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/ conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/ 2、显示源地址 conda config --set show_channel_urls yes
阅读更多...
【MongoDB】下载安装、指令操作

【MongoDB】下载安装、指令操作

目录 1.下载安装 2.指令 2.1.基础操作指令 2.2.增加 2.3.查询 2.4.修改 2.5.删除 前言&#xff1a; 关于MongoDB的核心概念请移步&#xff1a; 【文档数据库】ES和MongoDB的对比-CSDN博客 1.下载安装 本文以安装Windows版本的mongodb为例&#xff0c;Linux版本的其实…
阅读更多...
30岁的路口，这些90后选择离开大城市

30岁的路口，这些90后选择离开大城市

#第一批90后今年34岁了#【30岁的路口&#xff0c;这些90后选择离开大城市】#第一批90后现状如何# 据惊蛰研究所&#xff1a;第一批90后今年34岁了。假如从2012年踏入职场&#xff0c;第一批90后如今已在职场摸爬滚打十年。十年之前&#xff0c;他们意气风发来到大城市&#xff…
阅读更多...
go语言（十一）----面向对象继承

go语言（十一）----面向对象继承

一、面向对象继承 写一个父类 package mainimport "fmt"type Human struct {name stringsex string }func (this *Human) Eat() {fmt.Println("Human.Eat()...") }func (this *Human) Walk() {fmt.Println("Human.Walk()...") }func main() {h…
阅读更多...
开源项目_大模型应用_Chat2DB

开源项目_大模型应用_Chat2DB

1 基本信息 项目地址&#xff1a;https://github.com/chat2db/Chat2DBStar&#xff1a;10.7K 2 功能 Chat2DB 是一个智能且多功能的 SQL 客户端和报表工具&#xff0c;适用于各种数据库。 对于那些平时会用到数据库&#xff0c;但又不是数据库专家的程序员来说&#xff0c;…
阅读更多...
CISSP 2024年考试大纲中文版

CISSP 2024年考试大纲中文版

2024 CISSP详细内容大纲及权重最终版(仅供公众使用) 最后编辑于2023年8月18日-生效日期2024年4月15日 分类 域/任务/子任务 权重 域1 安全和风险管理 16% 1.1 理解、坚持和促进职业道德(2-4项) 1.1.1 ISC2职业道德守则 1.1.2 组织道德守则 1.2 理解并应用安全概…
阅读更多...
【MATLAB源码-第119期】基于matlab的GMSK系统1bit差分解调误码率曲线仿真，输出各个节点的波形以及功率谱。

【MATLAB源码-第119期】基于matlab的GMSK系统1bit差分解调误码率曲线仿真，输出各个节点的波形以及功率谱。

操作环境&#xff1a; MATLAB 2022a 1、算法描述 GMSK&#xff08;高斯最小频移键控&#xff09;是一种数字调制技术&#xff0c;广泛应用于移动通信&#xff0c;例如GSM网络。它是一种连续相位调频制式&#xff0c;通过改变载波的相位来传输数据。GMSK的关键特点是其频谱的…
阅读更多...
华为FusionStorage Block、OceanStor 100D、OceanStor pacific的区别

华为FusionStorage Block、OceanStor 100D、OceanStor pacific的区别

华为FusionStorage Block、OceanStor 100D、OceanStor pacific的区别&#xff1f; 华为块存储到底是叫什么呢&#xff1f; 有接触过华为块存储产品的小伙伴肯定都有疑惑&#xff0c;在FusionStorage 、FusionStorage Block、OceanStor 100D、OceanStor pacific等等的名词中&a…
阅读更多...
【JSON2WEB】01 WEB管理信息系统架构设计

【JSON2WEB】01 WEB管理信息系统架构设计

WEB管理信息系统分三层设计&#xff0c;分别为DataBase数据库、REST2SQL后端、JSON2WEB前端&#xff0c;三层都可以单独部署。 1 DataBase数据库 数据库根据需要选型即可&#xff0c;不需要自己设计开发&#xff0c;一般管理信息系统都选关系数据库&#xff0c;比如Oracle、…
阅读更多...
Docker（四）操作容器

Docker（四）操作容器

作者主页&#xff1a; 正函数的个人主页 文章收录专栏&#xff1a; Docker 欢迎大家点赞 &#x1f44d; 收藏 ⭐ 加关注哦&#xff01; 操作 Docker 容器 容器是 Docker 又一核心概念。 简单的说&#xff0c;容器是独立运行的一个或一组应用&#xff0c;以及它们的运行态环境…
阅读更多...
matlab appdesigner系列-常用14-树（复选框）

matlab appdesigner系列-常用14-树（复选框）

之前系列常用9&#xff0c;为单个复选框。树&#xff0c;就是多个复选框形成的选项组 示例&#xff1a;列举湖北省的几个城市 湖北省 武汉 宜昌 襄阳 荆州 1&#xff09;将树&#xff08;复选框&#xff09;拖拽到画布上&#xff0c;方式1就是&#xff1a;文字可以在右侧…
阅读更多...
循序渐进学 JavaScript ＜二＞

循序渐进学 JavaScript ＜二＞

续 <一> 九、JavaScript常见内置类 9.1 原始类型的包装类 基本数据类型也可以调用属性 在理论上来说它们是没有办法获取属性或者调用方法的 原始类型是简单的值&#xff0c;默认并不能调用属性和方法js 为了可以使其获取属性和调用方法&#xff0c;对其封装了对应的包装…
阅读更多...
RT-Thread 瑞萨 智能家居网络开发：RA6M3 HMI Board 以太网+GUI技术实践

RT-Thread 瑞萨 智能家居网络开发：RA6M3 HMI Board 以太网+GUI技术实践

不用放大了&#xff0c; 我在包里找到张不小的…… 以太网HMI线下培训-环境准备 这是社群的文档&#xff1a;【腾讯文档】以太网线下培训&#xff08;HMI-Board&#xff09; https://docs.qq.com/doc/DY0FIWFVuTEpORlNn 先介绍周六的培训是啥&#xff0c;然后再介绍一下要准…
阅读更多...
GPT应用开发：编写插件获取实时天气信息

GPT应用开发：编写插件获取实时天气信息

欢迎阅读本系列文章&#xff01;我将带你一起探索如何利用OpenAI API开发GPT应用。无论你是编程新手还是资深开发者&#xff0c;都能在这里获得灵感和收获。 本文&#xff0c;我们将继续展示聊天API中插件的使用方法&#xff0c;让你能够轻松驾驭这个强大的工具。 插件运行效…
阅读更多...
Oracle21C + PLSQL Developer 15 + Oracle客户端21安装配置完整图文版

Oracle21C + PLSQL Developer 15 + Oracle客户端21安装配置完整图文版

一、Oracle21C PLSQL Developer 15 Oracle客户端文件下载 1、Oracl21C下载地址&#xff1a;Database Software Downloads | Oracle 中国 2、 PLSQL Developer 15下载地址&#xff1a;Registered download PL/SQL Developer - Allround Automations 3、 Oracle 客户端下载地址…
阅读更多...
最新文章
推荐文章

Copyright @ 2022~2023