Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Development mailing list

Subject: Re: [OMPI devel] MALLOC_MMAP_MAX (and MALLOC_MMAP_THRESHOLD)
From: Eugene Loh (Eugene.Loh_at_[hidden])
Date: 2010-01-09 11:55:33


Jeff Squyres wrote:

>I'm not sure I follow -- are you saying that Open MPI is disabling the large mmap allocations, and we shouldn't?
>
>
Basically the reverse. The default (I think this means Linux, whether
with gcc, gfortran, Sun f90, etc.) is to use mmap to malloc large
allocations. We don't change this, but arguably we should.

Try this:

#include <stdlib.h>
#include <stdio.h>

int main(int argc, char **argv) {
  size_t size, nextsize;
  void *ptr, *nextptr;

  size = 1;
  ptr = malloc(size);
  while ( size < 1000000 ) {
    nextsize = 1.1 * size + 1;
    nextptr = malloc(nextsize);
    printf("%9ld %18lx %18lx %18lx\n", size, size, nextptr - ptr, ptr);
    size = nextsize;
    ptr = nextptr ;
  }

  return 0;
}

Here is sample output:

   # bytes #bytes (hex) #bytes ptr (hex)
                                       to next ptr
                                          (hex)

    58279 e3a7 e3b0 58f870
    64107 fa6b fa80 59dc20
    70518 11376 11380 5ad6a0
    77570 12f02 12f10 5bea20
    85328 14d50 14d60 5d1930
    93861 16ea5 16eb0 5e6690
   103248 19350 19360 5fd540
   113573 1bba5 1bbb0 6168a0
   124931 1e803 2b3044655bc0 632450
   137425 218d1 22000 2b3044c88010
   151168 24e80 25000 2b3044caa010
   166285 2898d 29000 2b3044ccf010
   182914 2ca82 2d000 2b3044cf8010
   201206 311f6 294000 2b3044d25010
   221327 3608f 37000 2b3044fb9010
   243460 3b704 3c000 2b3044ff0010

So, below 128K allocations, pointers are allocated at successively
higher addresses, each one just barely far enough to make room for the
allocation. E.g., an allocation of 0xE3A7 will push the "high-water
mark" up 0xE3B0 further.

Beyond 128K allocations, allocations are page aligned. The pointers all
end in 0x010. That is, whole numbers of pages are allocated and the
returned address is 16 bytes (0x10) into the first page. The size of
the allocations are the requested amount, plus a few bytes of padding,
rounded up to the nearest whole page size multiple.

The motivation to change, in my case, is performance. I don't know how
widespread this problem is, but...

>On Jan 8, 2010, at 9:25 AM, Sylvain Jeaugey wrote:
>
>
>>On Thu, 7 Jan 2010, Eugene Loh wrote:
>>
>>>setenv MALLOC_MMAP_MAX_ 0
>>>setenv MALLOC_TRIM_THRESHOLD_ -1
>>>
>>>
>>But yes, this set of settings is the number one tweak on HPC code that I'm
>>aware of.
>>
>>
Wow! I might vote for "compiling with -O", but let's not pick nits here.