2024-01-23 13:13:18

by Thomas Gleixner

[permalink] [raw]
Subject: [patch V2 01/22] x86/cpu/topology: Make the APIC mismatch warnings complete

From: Thomas Gleixner <[email protected]>

Detect all possible combinations of mismatch right in the CPUID evaluation
code.

Signed-off-by: Thomas Gleixner <[email protected]>

---
arch/x86/include/asm/apic.h | 5 ++---
arch/x86/kernel/cpu/common.c | 15 ++-------------
arch/x86/kernel/cpu/topology_common.c | 12 ++++++++++++
3 files changed, 16 insertions(+), 16 deletions(-)
---
--- a/arch/x86/include/asm/apic.h
+++ b/arch/x86/include/asm/apic.h
@@ -46,16 +46,15 @@ extern void x86_32_probe_apic(void);
static inline void x86_32_probe_apic(void) { }
#endif

-#ifdef CONFIG_X86_LOCAL_APIC
+extern u32 cpuid_to_apicid[];

+#ifdef CONFIG_X86_LOCAL_APIC
extern int apic_verbosity;
extern int local_apic_timer_c2_ok;

extern bool apic_is_disabled;
extern unsigned int lapic_timer_period;

-extern u32 cpuid_to_apicid[];
-
extern enum apic_intr_mode_id apic_intr_mode;
enum apic_intr_mode_id {
APIC_PIC,
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -1715,22 +1715,11 @@ static void generic_identify(struct cpui
#endif
}

-/*
- * Validate that ACPI/mptables have the same information about the
- * effective APIC id and update the package map.
- */
-static void validate_apic_and_package_id(struct cpuinfo_x86 *c)
+static void update_package_map(struct cpuinfo_x86 *c)
{
#ifdef CONFIG_SMP
unsigned int cpu = smp_processor_id();
- u32 apicid;

- apicid = apic->cpu_present_to_apicid(cpu);
-
- if (apicid != c->topo.apicid) {
- pr_err(FW_BUG "CPU%u: APIC id mismatch. Firmware: %x APIC: %x\n",
- cpu, apicid, c->topo.initial_apicid);
- }
BUG_ON(topology_update_package_map(c->topo.pkg_id, cpu));
BUG_ON(topology_update_die_map(c->topo.die_id, cpu));
#else
@@ -1921,7 +1910,7 @@ void identify_secondary_cpu(struct cpuin
#ifdef CONFIG_X86_32
enable_sep_cpu();
#endif
- validate_apic_and_package_id(c);
+ update_package_map(c);
x86_spec_ctrl_setup_ap();
update_srbds_msr();
if (boot_cpu_has_bug(X86_BUG_GDS))
--- a/arch/x86/kernel/cpu/topology_common.c
+++ b/arch/x86/kernel/cpu/topology_common.c
@@ -177,6 +177,18 @@ void cpu_parse_topology(struct cpuinfo_x

parse_topology(&tscan, false);

+ if (IS_ENABLED(CONFIG_X86_LOCAL_APIC)) {
+ if (c->topo.initial_apicid != c->topo.apicid) {
+ pr_err(FW_BUG "CPU%4u: APIC ID mismatch. CPUID: 0x%04x APIC: 0x%04x\n",
+ cpu, c->topo.initial_apicid, c->topo.apicid);
+ }
+
+ if (c->topo.apicid != cpuid_to_apicid[cpu]) {
+ pr_err(FW_BUG "CPU%4u: APIC ID mismatch. Firmware: 0x%04x APIC: 0x%04x\n",
+ cpu, cpuid_to_apicid[cpu], c->topo.apicid);
+ }
+ }
+
for (dom = TOPO_SMT_DOMAIN; dom < TOPO_MAX_DOMAIN; dom++) {
if (tscan.dom_shifts[dom] == x86_topo_system.dom_shifts[dom])
continue;



2024-01-25 12:43:12

by Ashok Raj

[permalink] [raw]
Subject: Re: [patch V2 01/22] x86/cpu/topology: Make the APIC mismatch warnings complete

On Tue, Jan 23, 2024 at 02:10:04PM +0100, Thomas Gleixner wrote:
> From: Thomas Gleixner <[email protected]>
>
> Detect all possible combinations of mismatch right in the CPUID evaluation
> code.
>
> Signed-off-by: Thomas Gleixner <[email protected]>
>
> ---
> arch/x86/include/asm/apic.h | 5 ++---
> arch/x86/kernel/cpu/common.c | 15 ++-------------
> arch/x86/kernel/cpu/topology_common.c | 12 ++++++++++++
> 3 files changed, 16 insertions(+), 16 deletions(-)

[snip]

> --- a/arch/x86/kernel/cpu/topology_common.c
> +++ b/arch/x86/kernel/cpu/topology_common.c
> @@ -177,6 +177,18 @@ void cpu_parse_topology(struct cpuinfo_x
>
> parse_topology(&tscan, false);
>
> + if (IS_ENABLED(CONFIG_X86_LOCAL_APIC)) {
> + if (c->topo.initial_apicid != c->topo.apicid) {
> + pr_err(FW_BUG "CPU%4u: APIC ID mismatch. CPUID: 0x%04x APIC: 0x%04x\n",
> + cpu, c->topo.initial_apicid, c->topo.apicid);
> + }
> +
> + if (c->topo.apicid != cpuid_to_apicid[cpu]) {
> + pr_err(FW_BUG "CPU%4u: APIC ID mismatch. Firmware: 0x%04x APIC: 0x%04x\n",
> + cpu, cpuid_to_apicid[cpu], c->topo.apicid);
> + }

Should we consider tainting the kernel when there is any mismatch?

2024-02-12 15:39:21

by Thomas Gleixner

[permalink] [raw]
Subject: Re: [patch V2 01/22] x86/cpu/topology: Make the APIC mismatch warnings complete

On Thu, Jan 25 2024 at 01:53, Ashok Raj wrote:
> On Tue, Jan 23, 2024 at 02:10:04PM +0100, Thomas Gleixner wrote:
>> From: Thomas Gleixner <[email protected]>
>>
>> Detect all possible combinations of mismatch right in the CPUID evaluation
>> code.
>>
>> Signed-off-by: Thomas Gleixner <[email protected]>
>>
>> ---
>> arch/x86/include/asm/apic.h | 5 ++---
>> arch/x86/kernel/cpu/common.c | 15 ++-------------
>> arch/x86/kernel/cpu/topology_common.c | 12 ++++++++++++
>> 3 files changed, 16 insertions(+), 16 deletions(-)
>
> [snip]
>
>> --- a/arch/x86/kernel/cpu/topology_common.c
>> +++ b/arch/x86/kernel/cpu/topology_common.c
>> @@ -177,6 +177,18 @@ void cpu_parse_topology(struct cpuinfo_x
>>
>> parse_topology(&tscan, false);
>>
>> + if (IS_ENABLED(CONFIG_X86_LOCAL_APIC)) {
>> + if (c->topo.initial_apicid != c->topo.apicid) {
>> + pr_err(FW_BUG "CPU%4u: APIC ID mismatch. CPUID: 0x%04x APIC: 0x%04x\n",
>> + cpu, c->topo.initial_apicid, c->topo.apicid);
>> + }
>> +
>> + if (c->topo.apicid != cpuid_to_apicid[cpu]) {
>> + pr_err(FW_BUG "CPU%4u: APIC ID mismatch. Firmware: 0x%04x APIC: 0x%04x\n",
>> + cpu, cpuid_to_apicid[cpu], c->topo.apicid);
>> + }
>
> Should we consider tainting the kernel when there is any mismatch?

No strong opinion about that.