| Message ID | 1376636809-10159-2-git-send-email-tangchen@cn.fujitsu.com (mailing list archive) |
|---|---|
| State | RFC, archived |
| Headers | show |
On Fri, 2013-08-16 at 15:06 +0800, Tang Chen wrote: > save_add_info() is defined as: > > #ifdef CONFIG_MEMORY_HOTPLUG > static inline int save_add_info(void) {return 1;} > #else > static inline int save_add_info(void) {return 0;} > #endif > > which means it is true when memory hotplug is configured. > > In acpi_numa_memory_affinity_init(), it checks the memory hotplug > flag in SRAT memory affinity and save_add_info() like this: > > if ((ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) && !save_add_info()) > goto out_err; > ...... > node = setup_node(pxm); > numa_add_memblk(node, start, end); > ...... > > which means if the memory range is hotpluggable, but memory hotplug is not > configured, it won't add these memory to numa_meminfo. > > After this, numa_meminfo_cover_memory() will fail, which will finally cause > numa_init() to fail. > > numa_init() > |->numa_register_memblks() > |->numa_meminfo_cover_memory() > > When numa_init() fails, it will fallback to numa_init(dummy_numa_init), and > all numa architecture will not be setup. > > This is nonsense. Even if memory hotplug is not configured, we can also use > numa architecture. > > Actually, save_add_info() is added by commit 71efa8fdc55e70ec6687c897a30759f0a2c2ad7e > in 2006. And now it is useless. > > So this patch kill save_add_info() and the nonsense checking. > > Signed-off-by: Tang Chen <tangchen@cn.fujitsu.com> Good catch! Acked-by: Toshi Kani <toshi.kani@hp.com> Thanks, -Toshi -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/arch/x86/mm/srat.c b/arch/x86/mm/srat.c index cdd0da9..1613c02 100644 --- a/arch/x86/mm/srat.c +++ b/arch/x86/mm/srat.c @@ -135,12 +135,6 @@ acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa) pxm, apic_id, node); } -#ifdef CONFIG_MEMORY_HOTPLUG -static inline int save_add_info(void) {return 1;} -#else -static inline int save_add_info(void) {return 0;} -#endif - /* Callback for parsing of the Proximity Domain <-> Memory Area mappings */ int __init acpi_numa_memory_affinity_init(struct acpi_srat_mem_affinity *ma) @@ -154,8 +148,6 @@ acpi_numa_memory_affinity_init(struct acpi_srat_mem_affinity *ma) goto out_err_bad_srat; if ((ma->flags & ACPI_SRAT_MEM_ENABLED) == 0) goto out_err; - if ((ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) && !save_add_info()) - goto out_err; start = ma->base_address; end = start + ma->length;
save_add_info() is defined as: #ifdef CONFIG_MEMORY_HOTPLUG static inline int save_add_info(void) {return 1;} #else static inline int save_add_info(void) {return 0;} #endif which means it is true when memory hotplug is configured. In acpi_numa_memory_affinity_init(), it checks the memory hotplug flag in SRAT memory affinity and save_add_info() like this: if ((ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) && !save_add_info()) goto out_err; ...... node = setup_node(pxm); numa_add_memblk(node, start, end); ...... which means if the memory range is hotpluggable, but memory hotplug is not configured, it won't add these memory to numa_meminfo. After this, numa_meminfo_cover_memory() will fail, which will finally cause numa_init() to fail. numa_init() |->numa_register_memblks() |->numa_meminfo_cover_memory() When numa_init() fails, it will fallback to numa_init(dummy_numa_init), and all numa architecture will not be setup. This is nonsense. Even if memory hotplug is not configured, we can also use numa architecture. Actually, save_add_info() is added by commit 71efa8fdc55e70ec6687c897a30759f0a2c2ad7e in 2006. And now it is useless. So this patch kill save_add_info() and the nonsense checking. Signed-off-by: Tang Chen <tangchen@cn.fujitsu.com> --- arch/x86/mm/srat.c | 8 -------- 1 files changed, 0 insertions(+), 8 deletions(-)