Allocate and setup per node hub info structs. CPU 0/Node 0 hub info
is statically allocated to be accessible early in system startup. The
remaining hub info structs are allocated on the node's local memory,
and shared among the CPU's on that node. This leaves the small amount
of info unique to each CPU in the per CPU info struct.
Memory is saved by combining the common per node info fields to common
node local structs. In addtion, since the info is read only only after
setup, it should stay in the L3 cache of the local processor socket.
This should therefore improve the cache hit rate when a group of cpus
on a node are all interrupted for a common task.
Signed-off-by: Mike Travis <[email protected]>
Reviewed-by: Dimitri Sivanich <[email protected]>
Reviewed-by: Andrew Banman <[email protected]>
Tested-by: John Estabrook <[email protected]>
Tested-by: Gary Kroening <[email protected]>
---
arch/x86/include/asm/uv/uv_hub.h | 38 +++++++++++++++++++++--
arch/x86/kernel/apic/x2apic_uv_x.c | 59 +++++++++++++++++++++++++++----------
2 files changed, 78 insertions(+), 19 deletions(-)
--- linux.orig/arch/x86/include/asm/uv/uv_hub.h
+++ linux/arch/x86/include/asm/uv/uv_hub.h
@@ -163,10 +163,6 @@ struct uv_hub_info_s {
unsigned char n_val;
};
-DECLARE_PER_CPU(struct uv_hub_info_s, __uv_hub_info);
-#define uv_hub_info this_cpu_ptr(&__uv_hub_info)
-#define uv_cpu_hub_info(cpu) (&per_cpu(__uv_hub_info, cpu))
-
/* CPU specific info with a pointer to the hub common info struct */
struct uv_cpu_info_s {
void *p_uv_hub_info;
@@ -181,6 +177,40 @@ DECLARE_PER_CPU(struct uv_cpu_info_s, __
#define uv_scir_info (&uv_cpu_info->scir)
#define uv_cpu_scir_info(cpu) (&uv_cpu_info_per(cpu)->scir)
+/* Node specific hub common info struct */
+extern void **__uv_hub_info_list;
+static inline struct uv_hub_info_s *uv_hub_info_list(int node)
+{
+ void *p = __uv_hub_info_list[node];
+ return (struct uv_hub_info_s *)p;
+}
+
+static inline struct uv_hub_info_s *_uv_hub_info(void)
+{
+ void *p = uv_cpu_info->p_uv_hub_info;
+ return (struct uv_hub_info_s *)p;
+}
+#define uv_hub_info _uv_hub_info()
+
+static inline struct uv_hub_info_s *uv_cpu_hub_info(int cpu)
+{
+ void *p = uv_cpu_info_per(cpu)->p_uv_hub_info;
+ return (struct uv_hub_info_s *)p;
+}
+
+#define UV_HUB_INFO_VERSION 0x7150
+extern int uv_hub_info_version(void);
+static inline int uv_hub_info_check(int version)
+{
+ if (uv_hub_info_version() == version)
+ return 0;
+
+ pr_crit("UV: uv_hub_info version(%x) mismatch, expecting(%x)\n",
+ uv_hub_info_version(), version);
+ BUG();
+}
+#define _uv_hub_info_check() uv_hub_info_check(UV_HUB_INFO_VERSION)
+
/*
* HUB revision ranges for each UV HUB architecture.
* This is a software convention - NOT the hardware revision numbers in
--- linux.orig/arch/x86/kernel/apic/x2apic_uv_x.c
+++ linux/arch/x86/kernel/apic/x2apic_uv_x.c
@@ -50,6 +50,7 @@ int uv_min_hub_revision_id;
EXPORT_SYMBOL_GPL(uv_min_hub_revision_id);
static struct apic apic_x2apic_uv_x;
+static struct uv_hub_info_s uv_hub_info_node0;
/* Set this to use hardware error handler instead of kernel panic */
static int disable_uv_undefined_panic = 1;
@@ -144,6 +145,9 @@ static int __init uv_acpi_madt_oem_check
if (strncmp(oem_id, "SGI", 3) != 0)
return 0;
+ /* Setup early hub type field in uv_hub_info for Node 0 */
+ uv_cpu_info->p_uv_hub_info = &uv_hub_info_node0;
+
/*
* Determine UV arch type.
* SGI: UV100/1000
@@ -199,8 +203,8 @@ int is_uv_system(void)
}
EXPORT_SYMBOL_GPL(is_uv_system);
-DEFINE_PER_CPU(struct uv_hub_info_s, __uv_hub_info);
-EXPORT_PER_CPU_SYMBOL_GPL(__uv_hub_info);
+void **__uv_hub_info_list;
+EXPORT_SYMBOL_GPL(__uv_hub_info_list);
DEFINE_PER_CPU(struct uv_cpu_info_s, __uv_cpu_info);
EXPORT_PER_CPU_SYMBOL_GPL(__uv_cpu_info);
@@ -220,6 +224,12 @@ EXPORT_SYMBOL_GPL(uv_possible_blades);
unsigned long sn_rtc_cycles_per_second;
EXPORT_SYMBOL(sn_rtc_cycles_per_second);
+extern int uv_hub_info_version(void)
+{
+ return UV_HUB_INFO_VERSION;
+}
+EXPORT_SYMBOL(uv_hub_info_version);
+
static int uv_wakeup_secondary(int phys_apicid, unsigned long start_rip)
{
unsigned long val;
@@ -948,9 +958,15 @@ void __init uv_system_init(void)
uv_init_hub_info(&hub_info);
- for(i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++)
- uv_possible_blades +=
- hweight64(uv_read_local_mmr( UVH_NODE_PRESENT_TABLE + i * 8));
+ pr_info("UV: NODE_PRESENT_DEPTH = %d\n", UVH_NODE_PRESENT_TABLE_DEPTH);
+ for (i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++) {
+ unsigned long np;
+
+ np = uv_read_local_mmr(UVH_NODE_PRESENT_TABLE + i * 8);
+ if (np)
+ pr_info("UV: NODE_PRESENT(%d) = 0x%016lx\n", i, np);
+ uv_possible_blades += hweight64(np);
+ }
/* uv_num_possible_blades() is really the hub count */
pr_info("UV: Found %d hubs, %d nodes, %d cpus\n",
@@ -976,6 +992,10 @@ void __init uv_system_init(void)
BUG_ON(!uv_cpu_to_blade);
memset(uv_cpu_to_blade, 255, bytes);
+ bytes = sizeof(void *) * uv_num_possible_blades();
+ __uv_hub_info_list = kzalloc(bytes, GFP_KERNEL);
+ BUG_ON(!__uv_hub_info_list);
+
blade = 0;
for (i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++) {
unsigned long present =
@@ -1000,28 +1020,37 @@ void __init uv_system_init(void)
uv_rtc_init();
for_each_present_cpu(cpu) {
+ struct uv_hub_info_s *new_hub = NULL;
int apicid = per_cpu(x86_cpu_to_apicid, cpu);
int nodeid = cpu_to_node(cpu);
- int lcpu;
- *uv_cpu_hub_info(cpu) = hub_info; /* common hub values */
- pnode = uv_apicid_to_pnode(apicid);
- blade = boot_pnode_to_blade(pnode);
- lcpu = uv_blade_info[blade].nr_possible_cpus;
- uv_blade_info[blade].nr_possible_cpus++;
+ /* Allocate new per hub info list */
+ if (uv_hub_info_list(nodeid) == NULL) {
+ if (cpu == 0)
+ __uv_hub_info_list[0] = &uv_hub_info_node0;
+ else
+ __uv_hub_info_list[nodeid] =
+ kzalloc_node(bytes, GFP_KERNEL, nodeid);
+
+ new_hub = uv_hub_info_list(nodeid);
+ BUG_ON(!new_hub);
+ *new_hub = hub_info;
+ blade = boot_pnode_to_blade(new_hub->pnode);
+ new_hub->pnode = uv_apicid_to_pnode(apicid);
+ new_hub->numa_blade_id = blade;
+ }
/* Any node on the blade, else will contain -1. */
uv_blade_info[blade].memory_nid = nodeid;
- uv_cpu_hub_info(cpu)->numa_blade_id = blade;
- uv_cpu_hub_info(cpu)->pnode = pnode;
uv_node_to_blade[nodeid] = blade;
uv_cpu_to_blade[cpu] = blade;
/* Initialize per cpu info list */
- uv_cpu_info_per(cpu)->p_uv_hub_info = uv_cpu_hub_info(cpu);
- uv_cpu_info_per(cpu)->blade_cpu_id = lcpu;
+ uv_cpu_info_per(cpu)->p_uv_hub_info = uv_hub_info_list(nodeid);
uv_cpu_info_per(cpu)->scir.offset = uv_scir_offset(apicid);
+ uv_cpu_info_per(cpu)->blade_cpu_id =
+ uv_blade_info[blade].nr_possible_cpus++;
}
/* Add blade/pnode info for nodes without cpus */
--