On Thu, Mar 6, 2025 at 2:26 PM bibo mao maobibo@loongson.cn wrote:
On 2025/3/6 下午12:06, Huacai Chen wrote:
Hi, Bibo,
On Thu, Mar 6, 2025 at 11:53 AM Bibo Mao maobibo@loongson.cn wrote:
The current max_pfn equals to zero. In this case, it caused users cannot get some page information through /proc such as kpagecount. The following message is displayed by stress-ng test suite with the command "stress-ng --verbose --physpage 1 -t 1".
# stress-ng --verbose --physpage 1 -t 1 stress-ng: error: [1691] physpage: cannot read page count for address 0x134ac000 in /proc/kpagecount, errno=22 (Invalid argument) stress-ng: error: [1691] physpage: cannot read page count for address 0x7ffff207c3a8 in /proc/kpagecount, errno=22 (Invalid argument) stress-ng: error: [1691] physpage: cannot read page count for address 0x134b0000 in /proc/kpagecount, errno=22 (Invalid argument) ...
After applying this patch, the kernel can pass the test. # stress-ng --verbose --physpage 1 -t 1 stress-ng: debug: [1701] physpage: [1701] started (instance 0 on CPU 3) stress-ng: debug: [1701] physpage: [1701] exited (instance 0 on CPU 3) stress-ng: debug: [1700] physpage: [1701] terminated (success)
Cc: stable@vger.kernel.org Fixes: ff6c3d81f2e8 ("NUMA: optimize detection of memory with no node id assigned by firmware")
Signed-off-by: Bibo Mao maobibo@loongson.cn
arch/loongarch/kernel/setup.c | 1 + 1 file changed, 1 insertion(+)
diff --git a/arch/loongarch/kernel/setup.c b/arch/loongarch/kernel/setup.c index edcfdfcad7d2..a9c1184ab893 100644 --- a/arch/loongarch/kernel/setup.c +++ b/arch/loongarch/kernel/setup.c @@ -390,6 +390,7 @@ static void __init arch_mem_init(char **cmdline_p) if (usermem) pr_info("User-defined physical RAM map overwrite\n");
max_low_pfn = max_pfn = PHYS_PFN(memblock_end_of_DRAM());max_low_pfn is already calculated for all three cases, so here just need "max_pfn = max_low_pfn".
In theory it should be.
However there are potential problems, it should be recalculated in function early_parse_mem() also if commandline "mem=" is added.
Yes, you are right, thanks.
Huacai
The other thing is that calculation init_numa_memory() is unnecessary since it is already calculated in memblock_init(). Memory block information comes from UEFI table or FDT table, and ACPI srat information only adds node information.
Regards Bibo Mao
Huacai
check_kernel_sections_mem(); /*base-commit: 848e076317446f9c663771ddec142d7c2eb4cb43
2.39.3