[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20231118154707.3668-1-shijie@os.amperecomputing.com>
Date: Sat, 18 Nov 2023 23:47:07 +0800
From: Huang Shijie <shijie@...amperecomputing.com>
To: catalin.marinas@....com
Cc: will@...nel.org, gregkh@...uxfoundation.org, rafael@...nel.org,
arnd@...db.de, mark.rutland@....com, broonie@...nel.org,
keescook@...omium.org, linux-arm-kernel@...ts.infradead.org,
linux-kernel@...r.kernel.org, linux-arch@...r.kernel.org,
patches@...erecomputing.com,
Huang Shijie <shijie@...amperecomputing.com>
Subject: [PATCH v2] arm64: irq: set the correct node for VMAP stack
In current code, init_irq_stacks() will call cpu_to_node().
The cpu_to_node() depends on percpu "numa_node" which is initialized in:
arch_call_rest_init() --> rest_init() -- kernel_init()
--> kernel_init_freeable() --> smp_prepare_cpus()
But init_irq_stacks() is called in init_IRQ() which is before
arch_call_rest_init().
So in init_irq_stacks(), the cpu_to_node() does not work, it
always return 0. In NUMA, it makes the node 1 cpu accesses the IRQ stack which
is in the node 0.
This patch fixes it by exporting the early_cpu_to_node(), and use it
in the init_irq_stacks().
Signed-off-by: Huang Shijie <shijie@...amperecomputing.com>
---
v1 --> v2:
fix the !NUMA compiling error.
---
arch/arm64/kernel/irq.c | 3 ++-
drivers/base/arch_numa.c | 2 +-
include/asm-generic/numa.h | 2 ++
3 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/kernel/irq.c b/arch/arm64/kernel/irq.c
index 6ad5c6ef5329..5226030979ae 100644
--- a/arch/arm64/kernel/irq.c
+++ b/arch/arm64/kernel/irq.c
@@ -25,6 +25,7 @@
#include <asm/softirq_stack.h>
#include <asm/stacktrace.h>
#include <asm/vmap_stack.h>
+#include <asm/numa.h>
/* Only access this in an NMI enter/exit */
DEFINE_PER_CPU(struct nmi_ctx, nmi_contexts);
@@ -57,7 +58,7 @@ static void init_irq_stacks(void)
unsigned long *p;
for_each_possible_cpu(cpu) {
- p = arch_alloc_vmap_stack(IRQ_STACK_SIZE, cpu_to_node(cpu));
+ p = arch_alloc_vmap_stack(IRQ_STACK_SIZE, early_cpu_to_node(cpu));
per_cpu(irq_stack_ptr, cpu) = p;
}
}
diff --git a/drivers/base/arch_numa.c b/drivers/base/arch_numa.c
index eaa31e567d1e..90519d981471 100644
--- a/drivers/base/arch_numa.c
+++ b/drivers/base/arch_numa.c
@@ -144,7 +144,7 @@ void __init early_map_cpu_to_node(unsigned int cpu, int nid)
unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
EXPORT_SYMBOL(__per_cpu_offset);
-static int __init early_cpu_to_node(int cpu)
+int early_cpu_to_node(int cpu)
{
return cpu_to_node_map[cpu];
}
diff --git a/include/asm-generic/numa.h b/include/asm-generic/numa.h
index 1a3ad6d29833..16073111bffc 100644
--- a/include/asm-generic/numa.h
+++ b/include/asm-generic/numa.h
@@ -35,6 +35,7 @@ int __init numa_add_memblk(int nodeid, u64 start, u64 end);
void __init numa_set_distance(int from, int to, int distance);
void __init numa_free_distance(void);
void __init early_map_cpu_to_node(unsigned int cpu, int nid);
+int early_cpu_to_node(int cpu);
void numa_store_cpu_info(unsigned int cpu);
void numa_add_cpu(unsigned int cpu);
void numa_remove_cpu(unsigned int cpu);
@@ -46,6 +47,7 @@ static inline void numa_add_cpu(unsigned int cpu) { }
static inline void numa_remove_cpu(unsigned int cpu) { }
static inline void arch_numa_init(void) { }
static inline void early_map_cpu_to_node(unsigned int cpu, int nid) { }
+static inline int early_cpu_to_node(int cpu) { return 0; }
#endif /* CONFIG_NUMA */
--
2.40.1
Powered by blists - more mailing lists