Re: [PATCH v3 15/17] crash: Simplify CRASH_MAX_MEMORY_RANGES handling

[Jinjie Ruan]

在 2026/9/2 4:14, Mike Rapoport 写道:
> Hi,
> 
>> Set CRASH_MAX_MEMORY_RANGES to 0 when MEMORY_HOTPLUG is disabled, then
>> remove the now unnecessary IS_ENABLED(CONFIG_MEMORY_HOTPLUG) checks in
>> powerpc and x86 arch_crash_get_elfcorehdr_size().
>>
>> No functional change intended.
>>
>> Signed-off-by: Jinjie Ruan <ruanjinjie@xxxxxxxxxx>
>>
>> diff --git a/arch/powerpc/kexec/crash.c b/arch/powerpc/kexec/crash.c
>> index fc0105c7af4c3..c47e3f867ea1f 100644
>> --- a/arch/powerpc/kexec/crash.c
>> +++ b/arch/powerpc/kexec/crash.c
>> @@ -474,9 +474,7 @@ unsigned int arch_crash_get_elfcorehdr_size(void)
>>  	unsigned long phdr_cnt;
>>  
>>  	/* A program header for possible CPUs + vmcoreinfo */
>> -	phdr_cnt = num_possible_cpus() + 1;
>> -	if (IS_ENABLED(CONFIG_MEMORY_HOTPLUG))
>> -		phdr_cnt += CONFIG_CRASH_MAX_MEMORY_RANGES;
>> +	phdr_cnt = num_possible_cpus() + 1 + CONFIG_CRASH_MAX_MEMORY_RANGES;
>>  
>>  	return elf64_phdr_size(phdr_cnt);
>>  }
>> diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c
>> index 99eb8583d9321..116410ac90fb4 100644
>> --- a/arch/x86/kernel/crash.c
>> +++ b/arch/x86/kernel/crash.c
>> @@ -407,9 +407,7 @@ unsigned int arch_crash_get_elfcorehdr_size(void)
>>  	unsigned int sz;
>>  
>>  	/* kernel_map, VMCOREINFO and maximum CPUs */
>> -	sz = 2 + num_possible_cpus();
>> -	if (IS_ENABLED(CONFIG_MEMORY_HOTPLUG))
>> -		sz += CONFIG_CRASH_MAX_MEMORY_RANGES;
>> +	sz = 2 + num_possible_cpus() + CONFIG_CRASH_MAX_MEMORY_RANGES;
>>  	sz *= sizeof(Elf64_Phdr);
>>  	return sz;
>>  }
>> diff --git a/kernel/Kconfig.kexec b/kernel/Kconfig.kexec
>> index a97ed9605602f..e6251dfe451a8 100644
>> --- a/kernel/Kconfig.kexec
>> +++ b/kernel/Kconfig.kexec
>> @@ -160,7 +160,8 @@ config CRASH_HOTPLUG
>>  
>>  config CRASH_MAX_MEMORY_RANGES
>>  	int "Specify the maximum number of memory regions for the elfcorehdr"
>> -	default 8192
>> +	default 8192 if MEMORY_HOTPLUG
>> +	default 0 if !MEMORY_HOTPLUG
> 
> This only sets the default, but a user can change the value explicitly.

You're right. Let me drop this patch for now. Thanks for the review!

>