Re: [PATCH 0/9] Avoid populating unbounded num of ptes with mmap_sem held

[Michel Lespinasse]

On Fri, Dec 21, 2012 at 5:09 PM, Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote:
> On Fri, Dec 21, 2012 at 4:59 PM, Michel Lespinasse <walken@xxxxxxxxxx> wrote:
>> On Fri, Dec 21, 2012 at 4:36 PM, Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote:
>>> Something's buggy here.  My evil test case is stuck with lots of
>>> threads spinning at 100% system time.
>>>
>>> The tasks in question use MCL_FUTURE but not MAP_POPULATE.  These
>>> tasks are immune to SIGKILL.
>>
>> Looking into it.
>>
>> There seems to be a problem with mlockall - the following program
>> fails in an unkillable way even before my changes:
>>
>> #include <sys/mman.h>
>> #include <stdio.h>
>> #include <stdint.h>
>>
>> int main(void) {
>>   void *p = mmap(NULL, 0x100000000000,
>>                  PROT_READ | PROT_WRITE,
>>                  MAP_PRIVATE | MAP_ANON | MAP_NORESERVE,
>>                  -1, 0);
>>   printf("p: %p\n", p);
>>   mlockall(MCL_CURRENT);
>>   return 0;
>> }
>>
>> I think my changes propagate this existing problem so it now shows up
>> in more places :/

So in my test case, the issue was caused by the mapping being 2^32
pages, which overflowed the integer 'nr_pages' argument to
__get_user_pages, which caused an infinite loop as __get_user_pages()
would return 0 so __mm_populate() would make no progress.

When dropping one zero from that humongous size in the test case, the
test case becomes at least killable.

> Hmm.  I'm using MCL_FUTURE with MAP_NORESERVE, but those mappings are
> not insanely large.  Should MAP_NORESERVE would negate MCL_FUTURE?
> I'm doing MAP_NORESERVE, PROT_NONE to prevent pages from being
> allocated in the future -- I have no intention of ever using them.

MAP_NORESERVE doesn't prevent page allocation, but PROT_NONE does
(precisely because people use it the same way as you do :)

> The other odd thing I do is use MAP_FIXED to replace MAP_NORESERVE pages.
Yes, I've seen people do that here too.

Could you share your test case so I can try reproducing the issue
you're seeing ?

Thanks,

-- 
Michel "Walken" Lespinasse
A program is never fully debugged until the last user dies.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/