[PATCH v5 09/15] arch/powerpc: don't initialize kasan if it's disabled

[Christophe Leroy (CS GROUP)]

Le 12/03/2026 à 12:12, Sourabh Jain a écrit :
> 
> 
> On 25/02/26 13:44, Baoquan He wrote:
>> Here, kasan is disabled if specified 'kasan=off' in kernel cmdline.
>>
>> This includes 32bit, book3s/64 and book3e/64.
>>
>> Signed-off-by: Baoquan He <bhe at redhat.com>
>> Cc: linuxppc-dev at lists.ozlabs.org
>> ---
>>   arch/powerpc/mm/kasan/init_32.c        | 6 +++++-
>>   arch/powerpc/mm/kasan/init_book3e_64.c | 4 ++++
>>   arch/powerpc/mm/kasan/init_book3s_64.c | 4 ++++
>>   3 files changed, 13 insertions(+), 1 deletion(-)
>>
>> diff --git a/arch/powerpc/mm/kasan/init_32.c b/arch/powerpc/mm/kasan/ 
>> init_32.c
>> index 1d083597464f..0ea2a636c992 100644
>> --- a/arch/powerpc/mm/kasan/init_32.c
>> +++ b/arch/powerpc/mm/kasan/init_32.c
>> @@ -141,6 +141,10 @@ void __init kasan_init(void)
>>       u64 i;
>>       int ret;
>> +    /* If KASAN is disabled via command line, don't initialize it. */
>> +    if (kasan_arg_disabled)
>> +        return;
>> +
>>       for_each_mem_range(i, &base, &end) {
>>           phys_addr_t top = min(end, total_lowmem);
>> @@ -170,7 +174,7 @@ void __init kasan_init(void)
>>   void __init kasan_late_init(void)
>>   {
>> -    if (IS_ENABLED(CONFIG_KASAN_VMALLOC))
>> +    if (IS_ENABLED(CONFIG_KASAN_VMALLOC) && kasan_enabled())
>>           kasan_unmap_early_shadow_vmalloc();
>>   }
>> diff --git a/arch/powerpc/mm/kasan/init_book3e_64.c b/arch/powerpc/mm/ 
>> kasan/init_book3e_64.c
>> index 0d3a73d6d4b0..fbe4c9a7e460 100644
>> --- a/arch/powerpc/mm/kasan/init_book3e_64.c
>> +++ b/arch/powerpc/mm/kasan/init_book3e_64.c
>> @@ -111,6 +111,10 @@ void __init kasan_init(void)
>>       u64 i;
>>       pte_t zero_pte = pfn_pte(virt_to_pfn(kasan_early_shadow_page), 
>> PAGE_KERNEL_RO);
>> +    /* If KASAN is disabled via command line, don't initialize it. */
>> +    if (kasan_arg_disabled)
>> +        return;
>> +
>>       for_each_mem_range(i, &start, &end)
>>           kasan_init_phys_region(phys_to_virt(start), phys_to_virt(end));
>> diff --git a/arch/powerpc/mm/kasan/init_book3s_64.c b/arch/powerpc/mm/ 
>> kasan/init_book3s_64.c
>> index dcafa641804c..f7906f9ef9be 100644
>> --- a/arch/powerpc/mm/kasan/init_book3s_64.c
>> +++ b/arch/powerpc/mm/kasan/init_book3s_64.c
>> @@ -54,6 +54,10 @@ void __init kasan_init(void)
>>       u64 i;
>>       pte_t zero_pte = pfn_pte(virt_to_pfn(kasan_early_shadow_page), 
>> PAGE_KERNEL);
>> +    /* If KASAN is disabled via command line, don't initialize it. */
>> +    if (kasan_arg_disabled)
>> +        return;
>> +
>>       if (!early_radix_enabled()) {
>>           pr_warn("KASAN not enabled as it requires radix!");
>>           return;
> 
> Should we log in the kernel log buffer that KASAN is disabled?
> 
> Right now, I don't see the kernel advertising the same.

When KASAN is enabled it is advertised with:

	pr_info("KernelAddressSanitizer initialized (generic)\n");

Isn't the absence of that message enough to understand KASAN is not there ?

> 
> $ dmesg | grep -i kasan
> [    0.000000] Kernel command line: BOOT_IMAGE=(ieee1275//vdevice/v- 
> scsi at 30000070/disk at 8100000000000000,msdos2)/vmlinuz-7.0.0-rc3+ 
> crashkernel=2G kasan=off
> 
> Tested this series on powerpc Pseries platform.
> So feel free to add:
> Tested-by: Sourabh Jain <sourabhjain at linux.ibm.com>
> 
> - Sourabh Jain
>