spapr_numa.c: parametrize FORM1 macros
The next preliminary step to introduce NUMA FORM2 affinity is to make the existing code independent of FORM1 macros and values, i.e. MAX_DISTANCE_REF_POINTS, NUMA_ASSOC_SIZE and VCPU_ASSOC_SIZE. This patch accomplishes that by doing the following: - move the NUMA related macros from spapr.h to spapr_numa.c where they are used. spapr.h gets instead a 'NUMA_NODES_MAX_NUM' macro that is used to refer to the maximum number of NUMA nodes, including GPU nodes, that the machine can support; - MAX_DISTANCE_REF_POINTS and NUMA_ASSOC_SIZE are renamed to FORM1_DIST_REF_POINTS and FORM1_NUMA_ASSOC_SIZE. These FORM1 specific macros are used in FORM1 init functions; - code that uses MAX_DISTANCE_REF_POINTS now retrieves the max_dist_ref_points value using get_max_dist_ref_points(). NUMA_ASSOC_SIZE is replaced by get_numa_assoc_size() and VCPU_ASSOC_SIZE is replaced by get_vcpu_assoc_size(). These functions are used by the generic device tree functions and h_home_node_associativity() and will allow them to switch between FORM1 and FORM2 without changing their core logic. Reviewed-by: Greg Kurz <groug@kaod.org> Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com> Message-Id: <20210920174947.556324-4-danielhb413@gmail.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This commit is contained in:
parent
afa3b3c9ee
commit
3a6e4ce684
|
@ -19,6 +19,33 @@
|
||||||
/* Moved from hw/ppc/spapr_pci_nvlink2.c */
|
/* Moved from hw/ppc/spapr_pci_nvlink2.c */
|
||||||
#define SPAPR_GPU_NUMA_ID (cpu_to_be32(1))
|
#define SPAPR_GPU_NUMA_ID (cpu_to_be32(1))
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Retrieves max_dist_ref_points of the current NUMA affinity.
|
||||||
|
*/
|
||||||
|
static int get_max_dist_ref_points(SpaprMachineState *spapr)
|
||||||
|
{
|
||||||
|
return FORM1_DIST_REF_POINTS;
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Retrieves numa_assoc_size of the current NUMA affinity.
|
||||||
|
*/
|
||||||
|
static int get_numa_assoc_size(SpaprMachineState *spapr)
|
||||||
|
{
|
||||||
|
return FORM1_NUMA_ASSOC_SIZE;
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Retrieves vcpu_assoc_size of the current NUMA affinity.
|
||||||
|
*
|
||||||
|
* vcpu_assoc_size is the size of ibm,associativity array
|
||||||
|
* for CPUs, which has an extra element (vcpu_id) in the end.
|
||||||
|
*/
|
||||||
|
static int get_vcpu_assoc_size(SpaprMachineState *spapr)
|
||||||
|
{
|
||||||
|
return get_numa_assoc_size(spapr) + 1;
|
||||||
|
}
|
||||||
|
|
||||||
static bool spapr_numa_is_symmetrical(MachineState *ms)
|
static bool spapr_numa_is_symmetrical(MachineState *ms)
|
||||||
{
|
{
|
||||||
int src, dst;
|
int src, dst;
|
||||||
|
@ -96,7 +123,7 @@ static void spapr_numa_define_FORM1_domains(SpaprMachineState *spapr)
|
||||||
* considered a match with associativity domains of node 0.
|
* considered a match with associativity domains of node 0.
|
||||||
*/
|
*/
|
||||||
for (i = 1; i < nb_numa_nodes; i++) {
|
for (i = 1; i < nb_numa_nodes; i++) {
|
||||||
for (j = 1; j < MAX_DISTANCE_REF_POINTS; j++) {
|
for (j = 1; j < FORM1_DIST_REF_POINTS; j++) {
|
||||||
spapr->numa_assoc_array[i][j] = cpu_to_be32(i);
|
spapr->numa_assoc_array[i][j] = cpu_to_be32(i);
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
@ -134,7 +161,7 @@ static void spapr_numa_define_FORM1_domains(SpaprMachineState *spapr)
|
||||||
*
|
*
|
||||||
* The Linux kernel will assume that the distance between src and
|
* The Linux kernel will assume that the distance between src and
|
||||||
* dst, in this case of no match, is 10 (local distance) doubled
|
* dst, in this case of no match, is 10 (local distance) doubled
|
||||||
* for each NUMA it didn't match. We have MAX_DISTANCE_REF_POINTS
|
* for each NUMA it didn't match. We have FORM1_DIST_REF_POINTS
|
||||||
* levels (4), so this gives us 10*2*2*2*2 = 160.
|
* levels (4), so this gives us 10*2*2*2*2 = 160.
|
||||||
*
|
*
|
||||||
* This logic can be seen in the Linux kernel source code, as of
|
* This logic can be seen in the Linux kernel source code, as of
|
||||||
|
@ -169,7 +196,7 @@ static void spapr_numa_FORM1_affinity_init(SpaprMachineState *spapr,
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* For all associativity arrays: first position is the size,
|
* For all associativity arrays: first position is the size,
|
||||||
* position MAX_DISTANCE_REF_POINTS is always the numa_id,
|
* position FORM1_DIST_REF_POINTS is always the numa_id,
|
||||||
* represented by the index 'i'.
|
* represented by the index 'i'.
|
||||||
*
|
*
|
||||||
* This will break on sparse NUMA setups, when/if QEMU starts
|
* This will break on sparse NUMA setups, when/if QEMU starts
|
||||||
|
@ -177,8 +204,8 @@ static void spapr_numa_FORM1_affinity_init(SpaprMachineState *spapr,
|
||||||
* 'i' will be a valid node_id set by the user.
|
* 'i' will be a valid node_id set by the user.
|
||||||
*/
|
*/
|
||||||
for (i = 0; i < nb_numa_nodes; i++) {
|
for (i = 0; i < nb_numa_nodes; i++) {
|
||||||
spapr->numa_assoc_array[i][0] = cpu_to_be32(MAX_DISTANCE_REF_POINTS);
|
spapr->numa_assoc_array[i][0] = cpu_to_be32(FORM1_DIST_REF_POINTS);
|
||||||
spapr->numa_assoc_array[i][MAX_DISTANCE_REF_POINTS] = cpu_to_be32(i);
|
spapr->numa_assoc_array[i][FORM1_DIST_REF_POINTS] = cpu_to_be32(i);
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
|
@ -192,15 +219,15 @@ static void spapr_numa_FORM1_affinity_init(SpaprMachineState *spapr,
|
||||||
max_nodes_with_gpus = nb_numa_nodes + NVGPU_MAX_NUM;
|
max_nodes_with_gpus = nb_numa_nodes + NVGPU_MAX_NUM;
|
||||||
|
|
||||||
for (i = nb_numa_nodes; i < max_nodes_with_gpus; i++) {
|
for (i = nb_numa_nodes; i < max_nodes_with_gpus; i++) {
|
||||||
spapr->numa_assoc_array[i][0] = cpu_to_be32(MAX_DISTANCE_REF_POINTS);
|
spapr->numa_assoc_array[i][0] = cpu_to_be32(FORM1_DIST_REF_POINTS);
|
||||||
|
|
||||||
for (j = 1; j < MAX_DISTANCE_REF_POINTS; j++) {
|
for (j = 1; j < FORM1_DIST_REF_POINTS; j++) {
|
||||||
uint32_t gpu_assoc = smc->pre_5_1_assoc_refpoints ?
|
uint32_t gpu_assoc = smc->pre_5_1_assoc_refpoints ?
|
||||||
SPAPR_GPU_NUMA_ID : cpu_to_be32(i);
|
SPAPR_GPU_NUMA_ID : cpu_to_be32(i);
|
||||||
spapr->numa_assoc_array[i][j] = gpu_assoc;
|
spapr->numa_assoc_array[i][j] = gpu_assoc;
|
||||||
}
|
}
|
||||||
|
|
||||||
spapr->numa_assoc_array[i][MAX_DISTANCE_REF_POINTS] = cpu_to_be32(i);
|
spapr->numa_assoc_array[i][FORM1_DIST_REF_POINTS] = cpu_to_be32(i);
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
|
@ -234,13 +261,15 @@ void spapr_numa_write_associativity_dt(SpaprMachineState *spapr, void *fdt,
|
||||||
{
|
{
|
||||||
_FDT((fdt_setprop(fdt, offset, "ibm,associativity",
|
_FDT((fdt_setprop(fdt, offset, "ibm,associativity",
|
||||||
spapr->numa_assoc_array[nodeid],
|
spapr->numa_assoc_array[nodeid],
|
||||||
sizeof(spapr->numa_assoc_array[nodeid]))));
|
get_numa_assoc_size(spapr) * sizeof(uint32_t))));
|
||||||
}
|
}
|
||||||
|
|
||||||
static uint32_t *spapr_numa_get_vcpu_assoc(SpaprMachineState *spapr,
|
static uint32_t *spapr_numa_get_vcpu_assoc(SpaprMachineState *spapr,
|
||||||
PowerPCCPU *cpu)
|
PowerPCCPU *cpu)
|
||||||
{
|
{
|
||||||
uint32_t *vcpu_assoc = g_new(uint32_t, VCPU_ASSOC_SIZE);
|
int max_distance_ref_points = get_max_dist_ref_points(spapr);
|
||||||
|
int vcpu_assoc_size = get_vcpu_assoc_size(spapr);
|
||||||
|
uint32_t *vcpu_assoc = g_new(uint32_t, vcpu_assoc_size);
|
||||||
int index = spapr_get_vcpu_id(cpu);
|
int index = spapr_get_vcpu_id(cpu);
|
||||||
|
|
||||||
/*
|
/*
|
||||||
|
@ -249,10 +278,10 @@ static uint32_t *spapr_numa_get_vcpu_assoc(SpaprMachineState *spapr,
|
||||||
* 0, put cpu_id last, then copy the remaining associativity
|
* 0, put cpu_id last, then copy the remaining associativity
|
||||||
* domains.
|
* domains.
|
||||||
*/
|
*/
|
||||||
vcpu_assoc[0] = cpu_to_be32(MAX_DISTANCE_REF_POINTS + 1);
|
vcpu_assoc[0] = cpu_to_be32(max_distance_ref_points + 1);
|
||||||
vcpu_assoc[VCPU_ASSOC_SIZE - 1] = cpu_to_be32(index);
|
vcpu_assoc[vcpu_assoc_size - 1] = cpu_to_be32(index);
|
||||||
memcpy(vcpu_assoc + 1, spapr->numa_assoc_array[cpu->node_id] + 1,
|
memcpy(vcpu_assoc + 1, spapr->numa_assoc_array[cpu->node_id] + 1,
|
||||||
(VCPU_ASSOC_SIZE - 2) * sizeof(uint32_t));
|
(vcpu_assoc_size - 2) * sizeof(uint32_t));
|
||||||
|
|
||||||
return vcpu_assoc;
|
return vcpu_assoc;
|
||||||
}
|
}
|
||||||
|
@ -261,12 +290,13 @@ int spapr_numa_fixup_cpu_dt(SpaprMachineState *spapr, void *fdt,
|
||||||
int offset, PowerPCCPU *cpu)
|
int offset, PowerPCCPU *cpu)
|
||||||
{
|
{
|
||||||
g_autofree uint32_t *vcpu_assoc = NULL;
|
g_autofree uint32_t *vcpu_assoc = NULL;
|
||||||
|
int vcpu_assoc_size = get_vcpu_assoc_size(spapr);
|
||||||
|
|
||||||
vcpu_assoc = spapr_numa_get_vcpu_assoc(spapr, cpu);
|
vcpu_assoc = spapr_numa_get_vcpu_assoc(spapr, cpu);
|
||||||
|
|
||||||
/* Advertise NUMA via ibm,associativity */
|
/* Advertise NUMA via ibm,associativity */
|
||||||
return fdt_setprop(fdt, offset, "ibm,associativity", vcpu_assoc,
|
return fdt_setprop(fdt, offset, "ibm,associativity", vcpu_assoc,
|
||||||
VCPU_ASSOC_SIZE * sizeof(uint32_t));
|
vcpu_assoc_size * sizeof(uint32_t));
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|
||||||
|
@ -274,17 +304,18 @@ int spapr_numa_write_assoc_lookup_arrays(SpaprMachineState *spapr, void *fdt,
|
||||||
int offset)
|
int offset)
|
||||||
{
|
{
|
||||||
MachineState *machine = MACHINE(spapr);
|
MachineState *machine = MACHINE(spapr);
|
||||||
|
int max_distance_ref_points = get_max_dist_ref_points(spapr);
|
||||||
int nb_numa_nodes = machine->numa_state->num_nodes;
|
int nb_numa_nodes = machine->numa_state->num_nodes;
|
||||||
int nr_nodes = nb_numa_nodes ? nb_numa_nodes : 1;
|
int nr_nodes = nb_numa_nodes ? nb_numa_nodes : 1;
|
||||||
uint32_t *int_buf, *cur_index, buf_len;
|
uint32_t *int_buf, *cur_index, buf_len;
|
||||||
int ret, i;
|
int ret, i;
|
||||||
|
|
||||||
/* ibm,associativity-lookup-arrays */
|
/* ibm,associativity-lookup-arrays */
|
||||||
buf_len = (nr_nodes * MAX_DISTANCE_REF_POINTS + 2) * sizeof(uint32_t);
|
buf_len = (nr_nodes * max_distance_ref_points + 2) * sizeof(uint32_t);
|
||||||
cur_index = int_buf = g_malloc0(buf_len);
|
cur_index = int_buf = g_malloc0(buf_len);
|
||||||
int_buf[0] = cpu_to_be32(nr_nodes);
|
int_buf[0] = cpu_to_be32(nr_nodes);
|
||||||
/* Number of entries per associativity list */
|
/* Number of entries per associativity list */
|
||||||
int_buf[1] = cpu_to_be32(MAX_DISTANCE_REF_POINTS);
|
int_buf[1] = cpu_to_be32(max_distance_ref_points);
|
||||||
cur_index += 2;
|
cur_index += 2;
|
||||||
for (i = 0; i < nr_nodes; i++) {
|
for (i = 0; i < nr_nodes; i++) {
|
||||||
/*
|
/*
|
||||||
|
@ -293,8 +324,8 @@ int spapr_numa_write_assoc_lookup_arrays(SpaprMachineState *spapr, void *fdt,
|
||||||
*/
|
*/
|
||||||
uint32_t *associativity = spapr->numa_assoc_array[i];
|
uint32_t *associativity = spapr->numa_assoc_array[i];
|
||||||
memcpy(cur_index, ++associativity,
|
memcpy(cur_index, ++associativity,
|
||||||
sizeof(uint32_t) * MAX_DISTANCE_REF_POINTS);
|
sizeof(uint32_t) * max_distance_ref_points);
|
||||||
cur_index += MAX_DISTANCE_REF_POINTS;
|
cur_index += max_distance_ref_points;
|
||||||
}
|
}
|
||||||
ret = fdt_setprop(fdt, offset, "ibm,associativity-lookup-arrays", int_buf,
|
ret = fdt_setprop(fdt, offset, "ibm,associativity-lookup-arrays", int_buf,
|
||||||
(cur_index - int_buf) * sizeof(uint32_t));
|
(cur_index - int_buf) * sizeof(uint32_t));
|
||||||
|
@ -383,6 +414,7 @@ static target_ulong h_home_node_associativity(PowerPCCPU *cpu,
|
||||||
target_ulong procno = args[1];
|
target_ulong procno = args[1];
|
||||||
PowerPCCPU *tcpu;
|
PowerPCCPU *tcpu;
|
||||||
int idx, assoc_idx;
|
int idx, assoc_idx;
|
||||||
|
int vcpu_assoc_size = get_vcpu_assoc_size(spapr);
|
||||||
|
|
||||||
/* only support procno from H_REGISTER_VPA */
|
/* only support procno from H_REGISTER_VPA */
|
||||||
if (flags != 0x1) {
|
if (flags != 0x1) {
|
||||||
|
@ -401,7 +433,7 @@ static target_ulong h_home_node_associativity(PowerPCCPU *cpu,
|
||||||
* 12 associativity domains for vcpus. Assert and bail if that's
|
* 12 associativity domains for vcpus. Assert and bail if that's
|
||||||
* not the case.
|
* not the case.
|
||||||
*/
|
*/
|
||||||
G_STATIC_ASSERT((VCPU_ASSOC_SIZE - 1) <= 12);
|
g_assert((vcpu_assoc_size - 1) <= 12);
|
||||||
|
|
||||||
vcpu_assoc = spapr_numa_get_vcpu_assoc(spapr, tcpu);
|
vcpu_assoc = spapr_numa_get_vcpu_assoc(spapr, tcpu);
|
||||||
/* assoc_idx starts at 1 to skip associativity size */
|
/* assoc_idx starts at 1 to skip associativity size */
|
||||||
|
@ -422,9 +454,9 @@ static target_ulong h_home_node_associativity(PowerPCCPU *cpu,
|
||||||
* macro. The ternary will fill the remaining registers with -1
|
* macro. The ternary will fill the remaining registers with -1
|
||||||
* after we went through vcpu_assoc[].
|
* after we went through vcpu_assoc[].
|
||||||
*/
|
*/
|
||||||
a = assoc_idx < VCPU_ASSOC_SIZE ?
|
a = assoc_idx < vcpu_assoc_size ?
|
||||||
be32_to_cpu(vcpu_assoc[assoc_idx++]) : -1;
|
be32_to_cpu(vcpu_assoc[assoc_idx++]) : -1;
|
||||||
b = assoc_idx < VCPU_ASSOC_SIZE ?
|
b = assoc_idx < vcpu_assoc_size ?
|
||||||
be32_to_cpu(vcpu_assoc[assoc_idx++]) : -1;
|
be32_to_cpu(vcpu_assoc[assoc_idx++]) : -1;
|
||||||
|
|
||||||
args[idx] = ASSOCIATIVITY(a, b);
|
args[idx] = ASSOCIATIVITY(a, b);
|
||||||
|
|
|
@ -100,23 +100,23 @@ typedef enum {
|
||||||
|
|
||||||
#define FDT_MAX_SIZE 0x200000
|
#define FDT_MAX_SIZE 0x200000
|
||||||
|
|
||||||
/*
|
/* Max number of GPUs per system */
|
||||||
* NUMA related macros. MAX_DISTANCE_REF_POINTS was taken
|
#define NVGPU_MAX_NUM 6
|
||||||
* from Linux kernel arch/powerpc/mm/numa.h. It represents the
|
|
||||||
* amount of associativity domains for non-CPU resources.
|
|
||||||
*
|
|
||||||
* NUMA_ASSOC_SIZE is the base array size of an ibm,associativity
|
|
||||||
* array for any non-CPU resource.
|
|
||||||
*
|
|
||||||
* VCPU_ASSOC_SIZE represents the size of ibm,associativity array
|
|
||||||
* for CPUs, which has an extra element (vcpu_id) in the end.
|
|
||||||
*/
|
|
||||||
#define MAX_DISTANCE_REF_POINTS 4
|
|
||||||
#define NUMA_ASSOC_SIZE (MAX_DISTANCE_REF_POINTS + 1)
|
|
||||||
#define VCPU_ASSOC_SIZE (NUMA_ASSOC_SIZE + 1)
|
|
||||||
|
|
||||||
/* Max number of these GPUsper a physical box */
|
/* Max number of NUMA nodes */
|
||||||
#define NVGPU_MAX_NUM 6
|
#define NUMA_NODES_MAX_NUM (MAX_NODES + NVGPU_MAX_NUM)
|
||||||
|
|
||||||
|
/*
|
||||||
|
* NUMA FORM1 macros. FORM1_DIST_REF_POINTS was taken from
|
||||||
|
* MAX_DISTANCE_REF_POINTS in arch/powerpc/mm/numa.h from Linux
|
||||||
|
* kernel source. It represents the amount of associativity domains
|
||||||
|
* for non-CPU resources.
|
||||||
|
*
|
||||||
|
* FORM1_NUMA_ASSOC_SIZE is the base array size of an ibm,associativity
|
||||||
|
* array for any non-CPU resource.
|
||||||
|
*/
|
||||||
|
#define FORM1_DIST_REF_POINTS 4
|
||||||
|
#define FORM1_NUMA_ASSOC_SIZE (FORM1_DIST_REF_POINTS + 1)
|
||||||
|
|
||||||
typedef struct SpaprCapabilities SpaprCapabilities;
|
typedef struct SpaprCapabilities SpaprCapabilities;
|
||||||
struct SpaprCapabilities {
|
struct SpaprCapabilities {
|
||||||
|
@ -249,7 +249,7 @@ struct SpaprMachineState {
|
||||||
unsigned gpu_numa_id;
|
unsigned gpu_numa_id;
|
||||||
SpaprTpmProxy *tpm_proxy;
|
SpaprTpmProxy *tpm_proxy;
|
||||||
|
|
||||||
uint32_t numa_assoc_array[MAX_NODES + NVGPU_MAX_NUM][NUMA_ASSOC_SIZE];
|
uint32_t numa_assoc_array[NUMA_NODES_MAX_NUM][FORM1_NUMA_ASSOC_SIZE];
|
||||||
|
|
||||||
Error *fwnmi_migration_blocker;
|
Error *fwnmi_migration_blocker;
|
||||||
};
|
};
|
||||||
|
|
Loading…
Reference in New Issue