aboutsummaryrefslogtreecommitdiff
path: root/arch/x86_64/kernel
diff options
context:
space:
mode:
authorAndi Kleen <ak@suse.de>2005-04-16 15:25:15 -0700
committerLinus Torvalds <torvalds@ppc970.osdl.org>2005-04-16 15:25:15 -0700
commit3dd9d514846cdca1dcef2e4fce666d85e199e844 (patch)
tree28e60a8b733db213e88b0aee8ef3861a93a6fa48 /arch/x86_64/kernel
parentf2ea2750fbe56867bc8e0eb595115b14195a3e5e (diff)
[PATCH] x86_64: add support for Intel dual-core detection and displaying
Appended patch adds the support for Intel dual-core detection and displaying the core related information in /proc/cpuinfo. It adds two new fields "core id" and "cpu cores" to x86 /proc/cpuinfo and the "core id" field for x86_64("cpu cores" field is already present in x86_64). Number of processor cores in a die is detected using cpuid(4) and this is documented in IA-32 Intel Architecture Software Developer's Manual (vol 2a) (http://developer.intel.com/design/pentium4/manuals/index_new.htm#sdm_vol2a) This patch also adds cpu_core_map similar to cpu_sibling_map. Slightly hacked by AK. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Diffstat (limited to 'arch/x86_64/kernel')
-rw-r--r--arch/x86_64/kernel/setup.c64
-rw-r--r--arch/x86_64/kernel/smpboot.c24
2 files changed, 72 insertions, 16 deletions
diff --git a/arch/x86_64/kernel/setup.c b/arch/x86_64/kernel/setup.c
index 57f619052ba..ef74564eede 100644
--- a/arch/x86_64/kernel/setup.c
+++ b/arch/x86_64/kernel/setup.c
@@ -774,7 +774,7 @@ static void __init detect_ht(struct cpuinfo_x86 *c)
{
#ifdef CONFIG_SMP
u32 eax, ebx, ecx, edx;
- int index_lsb, index_msb, tmp;
+ int index_msb, tmp;
int cpu = smp_processor_id();
if (!cpu_has(c, X86_FEATURE_HT))
@@ -786,7 +786,6 @@ static void __init detect_ht(struct cpuinfo_x86 *c)
if (smp_num_siblings == 1) {
printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
} else if (smp_num_siblings > 1) {
- index_lsb = 0;
index_msb = 31;
/*
* At this point we only support two siblings per
@@ -798,21 +797,33 @@ static void __init detect_ht(struct cpuinfo_x86 *c)
return;
}
tmp = smp_num_siblings;
- while ((tmp & 1) == 0) {
- tmp >>=1 ;
- index_lsb++;
- }
- tmp = smp_num_siblings;
while ((tmp & 0x80000000 ) == 0) {
tmp <<=1 ;
index_msb--;
}
- if (index_lsb != index_msb )
+ if (smp_num_siblings & (smp_num_siblings - 1))
index_msb++;
phys_proc_id[cpu] = phys_pkg_id(index_msb);
printk(KERN_INFO "CPU: Physical Processor ID: %d\n",
phys_proc_id[cpu]);
+
+ smp_num_siblings = smp_num_siblings / c->x86_num_cores;
+
+ tmp = smp_num_siblings;
+ index_msb = 31;
+ while ((tmp & 0x80000000) == 0) {
+ tmp <<=1 ;
+ index_msb--;
+ }
+ if (smp_num_siblings & (smp_num_siblings - 1))
+ index_msb++;
+
+ cpu_core_id[cpu] = phys_pkg_id(index_msb);
+
+ if (c->x86_num_cores > 1)
+ printk(KERN_INFO "CPU: Processor Core ID: %d\n",
+ cpu_core_id[cpu]);
}
#endif
}
@@ -829,7 +840,28 @@ static void __init sched_cmp_hack(struct cpuinfo_x86 *c)
smp_num_siblings = 1;
#endif
}
-
+
+/*
+ * find out the number of processor cores on the die
+ */
+static int __init intel_num_cpu_cores(struct cpuinfo_x86 *c)
+{
+ unsigned int eax;
+
+ if (c->cpuid_level < 4)
+ return 1;
+
+ __asm__("cpuid"
+ : "=a" (eax)
+ : "0" (4), "c" (0)
+ : "bx", "dx");
+
+ if (eax & 0x1f)
+ return ((eax >> 26) + 1);
+ else
+ return 1;
+}
+
static void __init init_intel(struct cpuinfo_x86 *c)
{
/* Cache sizes */
@@ -847,6 +879,7 @@ static void __init init_intel(struct cpuinfo_x86 *c)
c->x86_cache_alignment = c->x86_clflush_size * 2;
if (c->x86 >= 15)
set_bit(X86_FEATURE_CONSTANT_TSC, &c->x86_capability);
+ c->x86_num_cores = intel_num_cpu_cores(c);
}
void __init get_cpu_vendor(struct cpuinfo_x86 *c)
@@ -1153,13 +1186,16 @@ static int show_cpuinfo(struct seq_file *m, void *v)
seq_printf(m, " [%d]", i);
}
}
- seq_printf(m, "\n");
- if (c->x86_num_cores > 1)
- seq_printf(m, "cpu cores\t: %d\n", c->x86_num_cores);
-
- seq_printf(m, "\n\n");
+ seq_printf(m, "\n");
+#ifdef CONFIG_SMP
+ /* Put new fields at the end to lower the probability of
+ breaking user space parsers. */
+ seq_printf(m, "core id\t\t: %d\n", cpu_core_id[c - cpu_data]);
+ seq_printf(m, "cpu cores\t: %d\n", c->x86_num_cores);
+#endif
+ seq_printf(m, "\n");
return 0;
}
diff --git a/arch/x86_64/kernel/smpboot.c b/arch/x86_64/kernel/smpboot.c
index e5714090b00..7e2de8925b0 100644
--- a/arch/x86_64/kernel/smpboot.c
+++ b/arch/x86_64/kernel/smpboot.c
@@ -58,7 +58,10 @@
int smp_num_siblings = 1;
/* Package ID of each logical CPU */
u8 phys_proc_id[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
+/* Core ID of each logical CPU */
+u8 cpu_core_id[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
EXPORT_SYMBOL(phys_proc_id);
+EXPORT_SYMBOL(cpu_core_id);
/* Bitmask of currently online CPUs */
cpumask_t cpu_online_map;
@@ -71,6 +74,7 @@ static cpumask_t smp_commenced_mask;
struct cpuinfo_x86 cpu_data[NR_CPUS] __cacheline_aligned;
cpumask_t cpu_sibling_map[NR_CPUS] __cacheline_aligned;
+cpumask_t cpu_core_map[NR_CPUS] __cacheline_aligned;
/*
* Trampoline 80x86 program as an array.
@@ -713,6 +717,7 @@ static void __init smp_boot_cpus(unsigned int max_cpus)
io_apic_irqs = 0;
cpu_online_map = cpumask_of_cpu(0);
cpu_set(0, cpu_sibling_map[0]);
+ cpu_set(0, cpu_core_map[0]);
phys_cpu_present_map = physid_mask_of_physid(0);
if (APIC_init_uniprocessor())
printk(KERN_NOTICE "Local APIC not detected."
@@ -740,6 +745,7 @@ static void __init smp_boot_cpus(unsigned int max_cpus)
io_apic_irqs = 0;
cpu_online_map = cpumask_of_cpu(0);
cpu_set(0, cpu_sibling_map[0]);
+ cpu_set(0, cpu_core_map[0]);
phys_cpu_present_map = physid_mask_of_physid(0);
disable_apic = 1;
goto smp_done;
@@ -756,6 +762,7 @@ static void __init smp_boot_cpus(unsigned int max_cpus)
io_apic_irqs = 0;
cpu_online_map = cpumask_of_cpu(0);
cpu_set(0, cpu_sibling_map[0]);
+ cpu_set(0, cpu_core_map[0]);
phys_cpu_present_map = physid_mask_of_physid(0);
disable_apic = 1;
goto smp_done;
@@ -833,10 +840,13 @@ static void __init smp_boot_cpus(unsigned int max_cpus)
* Construct cpu_sibling_map[], so that we can tell the
* sibling CPU efficiently.
*/
- for (cpu = 0; cpu < NR_CPUS; cpu++)
+ for (cpu = 0; cpu < NR_CPUS; cpu++) {
cpus_clear(cpu_sibling_map[cpu]);
+ cpus_clear(cpu_core_map[cpu]);
+ }
for (cpu = 0; cpu < NR_CPUS; cpu++) {
+ struct cpuinfo_x86 *c = cpu_data + cpu;
int siblings = 0;
int i;
if (!cpu_isset(cpu, cpu_callout_map))
@@ -846,7 +856,7 @@ static void __init smp_boot_cpus(unsigned int max_cpus)
for (i = 0; i < NR_CPUS; i++) {
if (!cpu_isset(i, cpu_callout_map))
continue;
- if (phys_proc_id[cpu] == phys_proc_id[i]) {
+ if (phys_proc_id[cpu] == cpu_core_id[i]) {
siblings++;
cpu_set(i, cpu_sibling_map[cpu]);
}
@@ -862,6 +872,16 @@ static void __init smp_boot_cpus(unsigned int max_cpus)
siblings, cpu, smp_num_siblings);
smp_num_siblings = siblings;
}
+ if (c->x86_num_cores > 1) {
+ for (i = 0; i < NR_CPUS; i++) {
+ if (!cpu_isset(i, cpu_callout_map))
+ continue;
+ if (phys_proc_id[cpu] == phys_proc_id[i]) {
+ cpu_set(i, cpu_core_map[cpu]);
+ }
+ }
+ } else
+ cpu_core_map[cpu] = cpu_sibling_map[cpu];
}
Dprintk("Boot done.\n");