* Changes:
* Venkatesh Pallipadi : Adding cache identification through cpuid(4)
* Ashok Raj <ashok.raj@intel.com>: Work with CPU hotplug infrastructure.
+ * Andi Kleen : CPUID4 emulation on AMD.
*/
#include <linux/init.h>
cpumask_t shared_cpu_map;
};
-static unsigned short num_cache_leaves;
+unsigned short num_cache_leaves;
+
+/* AMD doesn't have CPUID4. Emulate it here to report the same
+ information to the user. This makes some assumptions about the machine:
+ No L3, L2 not shared, no SMT etc. that is currently true on AMD CPUs.
+
+ In theory the TLBs could be reported as fake type (they are in "dummy").
+ Maybe later */
+union l1_cache {
+ struct {
+ unsigned line_size : 8;
+ unsigned lines_per_tag : 8;
+ unsigned assoc : 8;
+ unsigned size_in_kb : 8;
+ };
+ unsigned val;
+};
+
+union l2_cache {
+ struct {
+ unsigned line_size : 8;
+ unsigned lines_per_tag : 4;
+ unsigned assoc : 4;
+ unsigned size_in_kb : 16;
+ };
+ unsigned val;
+};
+
+static const unsigned short assocs[] = {
+ [1] = 1, [2] = 2, [4] = 4, [6] = 8,
+ [8] = 16,
+ [0xf] = 0xffff // ??
+ };
+static const unsigned char levels[] = { 1, 1, 2 };
+static const unsigned char types[] = { 1, 2, 3 };
+
+static void __cpuinit amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax,
+ union _cpuid4_leaf_ebx *ebx,
+ union _cpuid4_leaf_ecx *ecx)
+{
+ unsigned dummy;
+ unsigned line_size, lines_per_tag, assoc, size_in_kb;
+ union l1_cache l1i, l1d;
+ union l2_cache l2;
+
+ eax->full = 0;
+ ebx->full = 0;
+ ecx->full = 0;
+
+ cpuid(0x80000005, &dummy, &dummy, &l1d.val, &l1i.val);
+ cpuid(0x80000006, &dummy, &dummy, &l2.val, &dummy);
+
+ if (leaf > 2 || !l1d.val || !l1i.val || !l2.val)
+ return;
+
+ eax->split.is_self_initializing = 1;
+ eax->split.type = types[leaf];
+ eax->split.level = levels[leaf];
+ eax->split.num_threads_sharing = 0;
+ eax->split.num_cores_on_die = current_cpu_data.x86_max_cores - 1;
+
+ if (leaf <= 1) {
+ union l1_cache *l1 = leaf == 0 ? &l1d : &l1i;
+ assoc = l1->assoc;
+ line_size = l1->line_size;
+ lines_per_tag = l1->lines_per_tag;
+ size_in_kb = l1->size_in_kb;
+ } else {
+ assoc = l2.assoc;
+ line_size = l2.line_size;
+ lines_per_tag = l2.lines_per_tag;
+ /* cpu_data has errata corrections for K7 applied */
+ size_in_kb = current_cpu_data.x86_cache_size;
+ }
+
+ if (assoc == 0xf)
+ eax->split.is_fully_associative = 1;
+ ebx->split.coherency_line_size = line_size - 1;
+ ebx->split.ways_of_associativity = assocs[assoc] - 1;
+ ebx->split.physical_line_partition = lines_per_tag - 1;
+ ecx->split.number_of_sets = (size_in_kb * 1024) / line_size /
+ (ebx->split.ways_of_associativity + 1) - 1;
+}
static int __cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
{
- unsigned int eax, ebx, ecx, edx;
- union _cpuid4_leaf_eax cache_eax;
+ union _cpuid4_leaf_eax eax;
+ union _cpuid4_leaf_ebx ebx;
+ union _cpuid4_leaf_ecx ecx;
+ unsigned edx;
- cpuid_count(4, index, &eax, &ebx, &ecx, &edx);
- cache_eax.full = eax;
- if (cache_eax.split.type == CACHE_TYPE_NULL)
+ if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)
+ amd_cpuid4(index, &eax, &ebx, &ecx);
+ else
+ cpuid_count(4, index, &eax.full, &ebx.full, &ecx.full, &edx);
+ if (eax.split.type == CACHE_TYPE_NULL)
return -EIO; /* better error ? */
- this_leaf->eax.full = eax;
- this_leaf->ebx.full = ebx;
- this_leaf->ecx.full = ecx;
- this_leaf->size = (this_leaf->ecx.split.number_of_sets + 1) *
- (this_leaf->ebx.split.coherency_line_size + 1) *
- (this_leaf->ebx.split.physical_line_partition + 1) *
- (this_leaf->ebx.split.ways_of_associativity + 1);
+ this_leaf->eax = eax;
+ this_leaf->ebx = ebx;
+ this_leaf->ecx = ecx;
+ this_leaf->size = (ecx.split.number_of_sets + 1) *
+ (ebx.split.coherency_line_size + 1) *
+ (ebx.split.physical_line_partition + 1) *
+ (ebx.split.ways_of_associativity + 1);
return 0;
}
unsigned int new_l1d = 0, new_l1i = 0; /* Cache sizes from cpuid(4) */
unsigned int new_l2 = 0, new_l3 = 0, i; /* Cache sizes from cpuid(4) */
unsigned int l2_id = 0, l3_id = 0, num_threads_sharing, index_msb;
-#ifdef CONFIG_SMP
+#ifdef CONFIG_X86_HT
unsigned int cpu = (c == &boot_cpu_data) ? 0 : (c - cpu_data);
#endif
if (new_l2) {
l2 = new_l2;
-#ifdef CONFIG_SMP
+#ifdef CONFIG_X86_HT
cpu_llc_id[cpu] = l2_id;
#endif
}
if (new_l3) {
l3 = new_l3;
-#ifdef CONFIG_SMP
+#ifdef CONFIG_X86_HT
cpu_llc_id[cpu] = l3_id;
#endif
}
if (num_cache_leaves == 0)
return -ENOENT;
- cpuid4_info[cpu] = kmalloc(
+ cpuid4_info[cpu] = kzalloc(
sizeof(struct _cpuid4_info) * num_cache_leaves, GFP_KERNEL);
if (unlikely(cpuid4_info[cpu] == NULL))
return -ENOMEM;
- memset(cpuid4_info[cpu], 0,
- sizeof(struct _cpuid4_info) * num_cache_leaves);
oldmask = current->cpus_allowed;
retval = set_cpus_allowed(current, cpumask_of_cpu(cpu));
return -ENOENT;
/* Allocate all required memory */
- cache_kobject[cpu] = kmalloc(sizeof(struct kobject), GFP_KERNEL);
+ cache_kobject[cpu] = kzalloc(sizeof(struct kobject), GFP_KERNEL);
if (unlikely(cache_kobject[cpu] == NULL))
goto err_out;
- memset(cache_kobject[cpu], 0, sizeof(struct kobject));
- index_kobject[cpu] = kmalloc(
+ index_kobject[cpu] = kzalloc(
sizeof(struct _index_kobject ) * num_cache_leaves, GFP_KERNEL);
if (unlikely(index_kobject[cpu] == NULL))
goto err_out;
- memset(index_kobject[cpu], 0,
- sizeof(struct _index_kobject) * num_cache_leaves);
return 0;
return;
}
-static int cacheinfo_cpu_callback(struct notifier_block *nfb,
+static int __cpuinit cacheinfo_cpu_callback(struct notifier_block *nfb,
unsigned long action, void *hcpu)
{
unsigned int cpu = (unsigned long)hcpu;
return NOTIFY_OK;
}
-static struct notifier_block cacheinfo_cpu_notifier =
+static struct notifier_block __cpuinitdata cacheinfo_cpu_notifier =
{
.notifier_call = cacheinfo_cpu_callback,
};
if (num_cache_leaves == 0)
return 0;
- register_cpu_notifier(&cacheinfo_cpu_notifier);
+ register_hotcpu_notifier(&cacheinfo_cpu_notifier);
for_each_online_cpu(i) {
cacheinfo_cpu_callback(&cacheinfo_cpu_notifier, CPU_ONLINE,
projects / linux-drm-fsl-dcu.git / blobdiff