Re: [PATCH v1] hw/loongarch: Add numa support

[maobibo]

在 2023/5/18 17:06, Song Gao 写道:
> 1. Implement some functions for LoongArch numa support;
> 2. Implement fdt_add_memory_node() for fdt;
> 3. build_srat() fills node_id and adds build numa memory.
> 
> Base-on:
> 20230518014115.117869-1-gaosong@loongson.cn/">https://patchew.org/QEMU/20230518014115.117869-1-gaosong@loongson.cn/
> 
> Signed-off-by: Song Gao <gaosong@loongson.cn>
> ---
>  hw/loongarch/acpi-build.c | 42 ++++++++++++-----
>  hw/loongarch/virt.c       | 96 ++++++++++++++++++++++++++++++++++-----
>  2 files changed, 116 insertions(+), 22 deletions(-)
> 
> diff --git a/hw/loongarch/acpi-build.c b/hw/loongarch/acpi-build.c
> index 232344e1c7..bb5adb9c1e 100644
> --- a/hw/loongarch/acpi-build.c
> +++ b/hw/loongarch/acpi-build.c
> @@ -163,11 +163,12 @@ build_madt(GArray *table_data, BIOSLinker *linker, 
> LoongArchMachineState *lams)
>  static void
>  build_srat(GArray *table_data, BIOSLinker *linker, MachineState *machine)
>  {
> -    int i, arch_id;
> +    int i, arch_id, node_id;
> +    uint64_t mem_len, mem_base;
> +    int nb_numa_nodes = machine->numa_state->num_nodes;
>      LoongArchMachineState *lams = LOONGARCH_MACHINE(machine);
> -    MachineState *ms = MACHINE(lams);
> -    MachineClass *mc = MACHINE_GET_CLASS(ms);
> -    const CPUArchIdList *arch_ids = mc->possible_cpu_arch_ids(ms);
> +    MachineClass *mc = MACHINE_GET_CLASS(lams);
> +    const CPUArchIdList *arch_ids = mc->possible_cpu_arch_ids(machine);
>      AcpiTable table = { .sig = "SRAT", .rev = 1, .oem_id = lams->oem_id,
>                          .oem_table_id = lams->oem_table_id };
>  
> @@ -177,12 +178,13 @@ build_srat(GArray *table_data, BIOSLinker *linker, 
> MachineState *machine)
>  
>      for (i = 0; i < arch_ids->len; ++i) {
>          arch_id = arch_ids->cpus[i].arch_id;
> +        node_id = arch_ids->cpus[i].props.node_id;
>  
>          /* Processor Local APIC/SAPIC Affinity Structure */
>          build_append_int_noprefix(table_data, 0, 1);  /* Type  */
>          build_append_int_noprefix(table_data, 16, 1); /* Length */
>          /* Proximity Domain [7:0] */
> -        build_append_int_noprefix(table_data, 0, 1);
> +        build_append_int_noprefix(table_data, node_id, 1);
>          build_append_int_noprefix(table_data, arch_id, 1); /* APIC ID */
>          /* Flags, Table 5-36 */
>          build_append_int_noprefix(table_data, 1, 4);
> @@ -192,15 +194,33 @@ build_srat(GArray *table_data, BIOSLinker *linker, 
> MachineState *machine)
>          build_append_int_noprefix(table_data, 0, 4); /* Reserved */
>      }
>  
> +    /* Node0 */
>      build_srat_memory(table_data, VIRT_LOWMEM_BASE, VIRT_LOWMEM_SIZE,
>                        0, MEM_AFFINITY_ENABLED);
> +    mem_base = VIRT_HIGHMEM_BASE;
> +    if (!nb_numa_nodes) {
> +        mem_len = machine->ram_size - VIRT_LOWMEM_SIZE;
> +    } else {
> +        mem_len = machine->numa_state->nodes[0].node_mem - VIRT_LOWMEM_SIZE;
> +    }
> +    build_srat_memory(table_data, mem_base, mem_len, 0, 
> MEM_AFFINITY_ENABLED);
> +
> +    /* Node1 - Nodemax */
> +    if (nb_numa_nodes) {
> +        mem_base += mem_len;
> +        for (i = 1; i < nb_numa_nodes; ++i) {
> +            if (machine->numa_state->nodes[i].node_mem > 0) {
> +                build_srat_memory(table_data, mem_base,
> +                                  machine->numa_state->nodes[i].node_mem, i,
> +                                  MEM_AFFINITY_ENABLED);
> +                mem_base += machine->numa_state->nodes[i].node_mem;
> +            }
> +        }
> +    }
>  
> -    build_srat_memory(table_data, VIRT_HIGHMEM_BASE, machine->ram_size - 
> VIRT_LOWMEM_SIZE,
> -                      0, MEM_AFFINITY_ENABLED);
> -
> -    if (ms->device_memory) {
> -        build_srat_memory(table_data, ms->device_memory->base,
> -                          memory_region_size(&ms->device_memory->mr),
> +    if (machine->device_memory) {
> +        build_srat_memory(table_data, machine->device_memory->base,
> +                          memory_region_size(&machine->device_memory->mr),
>                            0, MEM_AFFINITY_HOTPLUGGABLE | 
> MEM_AFFINITY_ENABLED);
>      }
>  
> diff --git a/hw/loongarch/virt.c b/hw/loongarch/virt.c
> index 6e1c42fb2b..c9235f740e 100644
> --- a/hw/loongarch/virt.c
> +++ b/hw/loongarch/virt.c
> @@ -164,11 +164,18 @@ static void fdt_add_cpu_nodes(const 
> LoongArchMachineState *lams)
>      for (num = smp_cpus - 1; num >= 0; num--) {
>          char *nodename = g_strdup_printf("/cpus/cpu@%d", num);
>          LoongArchCPU *cpu = LOONGARCH_CPU(qemu_get_cpu(num));
> +        CPUState *cs = CPU(cpu);
>  
>          qemu_fdt_add_subnode(ms->fdt, nodename);
>          qemu_fdt_setprop_string(ms->fdt, nodename, "device_type", "cpu");
>          qemu_fdt_setprop_string(ms->fdt, nodename, "compatible",
>                                  cpu->dtb_compatible);
> +
> +        if (ms->possible_cpus->cpus[cs->cpu_index].props.has_node_id) {
> +            qemu_fdt_setprop_cell(ms->fdt, nodename, "numa-node-id",
> +                ms->possible_cpus->cpus[cs->cpu_index].props.node_id);
> +        }
> +
>          qemu_fdt_setprop_cell(ms->fdt, nodename, "reg", num);
>          qemu_fdt_setprop_cell(ms->fdt, nodename, "phandle",
>                                qemu_fdt_alloc_phandle(ms->fdt));
> @@ -280,6 +287,22 @@ static void fdt_add_irqchip_node(LoongArchMachineState 
> *lams)
>      g_free(nodename);
>  }
>  
> +static void fdt_add_memory_node(MachineState *ms,
> +                                uint64_t base, uint64_t size, int node_id)
> +{
> +    char *nodename = g_strdup_printf("/memory@%lx", base);
> +
> +    qemu_fdt_add_subnode(ms->fdt, nodename);
> +    qemu_fdt_setprop_cells(ms->fdt, nodename, "reg", 2, base, 2, size);
> +    qemu_fdt_setprop_string(ms->fdt, nodename, "device_type", "memory");
> +
> +    if (ms->numa_state && ms->numa_state->num_nodes) {
> +        qemu_fdt_setprop_cell(ms->fdt, nodename, "numa-node-id", node_id);
> +    }
> +
> +    g_free(nodename);
> +}
> +
>  #define PM_BASE 0x10080000
>  #define PM_SIZE 0x100
>  #define PM_CTRL 0x10
> @@ -766,14 +789,17 @@ static void loongarch_init(MachineState *machine)
>      const char *cpu_model = machine->cpu_type;
>      ram_addr_t offset = 0;
>      ram_addr_t ram_size = machine->ram_size;
> -    uint64_t highram_size = 0;
> +    uint64_t highram_size = 0, phyAddr = 0;
>      MemoryRegion *address_space_mem = get_system_memory();
>      LoongArchMachineState *lams = LOONGARCH_MACHINE(machine);
> +    int nb_numa_nodes = machine->numa_state->num_nodes;
> +    NodeInfo *numa_info = machine->numa_state->nodes;
>      int i;
>      hwaddr fdt_base;
>      const CPUArchIdList *possible_cpus;
>      MachineClass *mc = MACHINE_GET_CLASS(machine);
>      CPUState *cpu;
> +    char *ramName = NULL;
>  
>      if (!cpu_model) {
>          cpu_model = LOONGARCH_CPU_TYPE_NAME("la464");
> @@ -798,17 +824,45 @@ static void loongarch_init(MachineState *machine)
>          machine->possible_cpus->cpus[i].cpu = OBJECT(cpu);
>      }
>      fdt_add_cpu_nodes(lams);
> -    /* Add memory region */
> -    memory_region_init_alias(&lams->lowmem, NULL, "loongarch.lowram",
> -                             machine->ram, 0, 256 * MiB);
> -    memory_region_add_subregion(address_space_mem, offset, &lams->lowmem);
> -    offset += 256 * MiB;
> -    memmap_add_entry(0, 256 * MiB, 1);
> -    highram_size = ram_size - 256 * MiB;
> -    memory_region_init_alias(&lams->highmem, NULL, "loongarch.highmem",
> +
> +    memory_region_add_subregion(address_space_mem, 0, machine->ram);
> +
> +    /* Node0 memory */
> +    memmap_add_entry(VIRT_LOWMEM_BASE, VIRT_LOWMEM_SIZE, 1);
> +    fdt_add_memory_node(machine, VIRT_LOWMEM_BASE, VIRT_LOWMEM_SIZE, 0);
> +    memory_region_init_alias(&lams->lowmem, NULL, "loongarch.node0.lowram",
> +                             machine->ram, offset, VIRT_LOWMEM_SIZE);
> +    memory_region_add_subregion(address_space_mem, phyAddr, &lams->lowmem);
> +
> +    offset += VIRT_LOWMEM_SIZE;
> +    if (nb_numa_nodes > 0) {
> +        assert(numa_info[0].node_mem > VIRT_LOWMEM_SIZE);
> +        highram_size = numa_info[0].node_mem - VIRT_LOWMEM_SIZE;
> +    } else {
> +        highram_size = ram_size - VIRT_LOWMEM_SIZE;
> +    }
> +    phyAddr = VIRT_HIGHMEM_BASE;
> +    memmap_add_entry(phyAddr, highram_size, 1);
> +    fdt_add_memory_node(machine, phyAddr, highram_size, 0);
> +    memory_region_init_alias(&lams->highmem, NULL, "loongarch.node0.highram",
>                               machine->ram, offset, highram_size);
> -    memory_region_add_subregion(address_space_mem, 0x90000000, 
> &lams->highmem);
> -    memmap_add_entry(0x90000000, highram_size, 1);
> +    memory_region_add_subregion(address_space_mem, phyAddr, &lams->highmem);
> +
> +    /* Node1 - Nodemax memory */
> +    offset += highram_size;
> +    phyAddr += highram_size;
> +
> +    for (i = 1; i < nb_numa_nodes; i++) {
> +        MemoryRegion *nodemem = g_new(MemoryRegion, 1);
> +        ramName = g_strdup_printf("loongarch.node%d.ram", i);
> +        memory_region_init_alias(nodemem, NULL, ramName, machine->ram,
> +                                 offset,  numa_info[i].node_mem);
> +        memory_region_add_subregion(address_space_mem, phyAddr, nodemem);
> +        memmap_add_entry(phyAddr, numa_info[i].node_mem, 1);
> +        fdt_add_memory_node(machine, phyAddr, numa_info[i].node_mem, i);
> +        offset += numa_info[i].node_mem;
> +        phyAddr += numa_info[i].node_mem;
> +    }
>  
>      /* initialize device memory address space */
>      if (machine->ram_size < machine->maxram_size) {
> @@ -1051,6 +1105,21 @@ static const CPUArchIdList 
> *virt_possible_cpu_arch_ids(MachineState *ms)
>      return ms->possible_cpus;
>  }
>  
> +static CpuInstanceProperties
> +virt_cpu_index_to_props(MachineState *ms, unsigned cpu_index)
> +{
> +    MachineClass *mc = MACHINE_GET_CLASS(ms);
> +    const CPUArchIdList *possible_cpus = mc->possible_cpu_arch_ids(ms);
> +
> +    assert(cpu_index < possible_cpus->len);
> +    return possible_cpus->cpus[cpu_index].props;
> +}
> +
> +static int64_t virt_get_default_cpu_node_id(const MachineState *ms, int idx)
> +{
> +    return idx % ms->numa_state->num_nodes;
> +}
The logic has some problem, or we should add support for packageid, or
do something like riscv.

int64_t riscv_numa_get_default_cpu_node_id(const MachineState *ms, int idx)
{   
    int64_t nidx = 0;
    
    if (ms->numa_state->num_nodes) {
        nidx = idx / (ms->smp.cpus / ms->numa_state->num_nodes);
        if (ms->numa_state->num_nodes <= nidx) {
            nidx = ms->numa_state->num_nodes - 1;
        }
    }
    
    return nidx;
}


Regards
Bibo, Mao 

> +
>  static void loongarch_class_init(ObjectClass *oc, void *data)
>  {
>      MachineClass *mc = MACHINE_CLASS(oc);
> @@ -1068,6 +1137,11 @@ static void loongarch_class_init(ObjectClass *oc, void 
> *data)
>      mc->default_boot_order = "c";
>      mc->no_cdrom = 1;
>      mc->possible_cpu_arch_ids = virt_possible_cpu_arch_ids;
> +    mc->cpu_index_to_instance_props = virt_cpu_index_to_props;
> +    mc->get_default_cpu_node_id = virt_get_default_cpu_node_id;
> +    mc->numa_mem_supported = true;
> +    mc->auto_enable_numa_with_memhp = true;
> +    mc->auto_enable_numa_with_memdev = true;
>      mc->get_hotplug_handler = virt_machine_get_hotplug_handler;
>      hc->plug = loongarch_machine_device_plug_cb;
>      hc->pre_plug = virt_machine_device_pre_plug;