Re: [PATCH V3 0/4] mm/debug_vm_pgtable: Add some more tests

[Anshuman Khandual <anshuman.khandual@arm.com>]

On 06/24/2020 08:10 PM, Alexander Gordeev wrote:
> On Wed, Jun 24, 2020 at 01:48:08PM +0200, Gerald Schaefer wrote:
>> On Wed, 24 Jun 2020 13:05:39 +0200
>> Alexander Gordeev <agordeev@linux.ibm.com> wrote:
>>
>>> On Wed, Jun 24, 2020 at 08:43:10AM +0530, Anshuman Khandual wrote:
>>>
>>> [...]
>>>
>>>> Hello Gerald/Christophe/Vineet,
>>>>
>>>> It would be really great if you could give this series a quick test
>>>> on s390/ppc/arc platforms respectively. Thank you.
>>>
>>> That worked for me with the default and debug s390 configurations.
>>> Would you like to try with some particular options or combinations
>>> of the options?
>>
>> It will be enabled automatically on all archs that set
>> ARCH_HAS_DEBUG_VM_PGTABLE, which we do for s390 unconditionally.
>> Also, DEBUG_VM has to be set, which we have only in the debug config.
>> So only the s390 debug config will have it enabled, you can check
>> dmesg for "debug_vm_pgtable" to see when / where it was run, and if it
>> triggered any warnings.
> 
> Yes, that is what I did ;)
> 
> I should have been more clear. I wonder whether Anshuman has in
> mind other options which possibly makes sense to set or unset
> and check how it goes with non-standard configurations.

After enabling CONFIG_DEBUG_VM either explicitly or via DEBUG_VM, ideally
any memory config combination on s390 which can change platform page table
helpers (validated with CONFIG_DEBUG_VM) should also get tested. Recently,
there was a kernel crash on ppc64 [1] and a build failure on ppc32 [2] for
some particular configs. Hence it will be great if you could run this test
on multiple s390 configurations.

[1] 787d563b8642f35c5 ("mm/debug_vm_pgtable: fix kernel crash by checking for THP support")
[2] 9449c9cb420b249eb ("mm/debug_vm_pgtable: fix build failure with powerpc 8xx")

- Anshuman