[petsc-dev] Using multiple mallocs with PETSc

Thu Mar 9 21:53:08 CST 2017

On Thu, Mar 9, 2017 at 7:04 PM, Zhang, Hong <hongzhang at anl.gov> wrote:

>
> On Mar 9, 2017, at 8:50 PM, Barry Smith <bsmith at mcs.anl.gov> wrote:
>
>
> On Mar 9, 2017, at 8:29 PM, Zhang, Hong <hongzhang at anl.gov> wrote:
>
>
> On Mar 9, 2017, at 7:38 PM, Barry Smith <bsmith at mcs.anl.gov> wrote:
>
>
> On Mar 9, 2017, at 7:18 PM, Richard Mills <richardtmills at gmail.com> wrote:
>
> Hi Barry,
>
> I like the sound of this, but I think we'd need to be careful about not
> messing up data alignment if we do this.  If we want a malloc that is going
> to let us put the start of an array on, say, a 64 byte alignment boundary,
> then we need to not mess that up by putting this integer value there.
>
>
> As I said the extra space is 64 bit. Now if you want 128 bit alignment we
> could put a 128 bit.
>
> We could pad with an extra 64 bytes internally, though that may be getting
> too wasteful.  I don't know how to get a malloc that gives us a starting
> address that is 64 bits *before* an alignment boundary (so that the memory
> the user sees from the malloc call indeed starts at the boundary), but
> maybe that's doable...
>
>
> What alignment boundaries are useful for Intel processes? 64 yup, 128,
> 256, 512 ? Does higher values provide better performance for SIMD etc?
>
>
> If the goal is to simply deal with allocations to high bandwidth memory on
> KNL, the memkind-provided free() will do the right thing with allocations
> in DRAM or MCDRAM.
>
>
> Hmm, Hong, how come we don't use this? I didn't realize it worked this
> way. This would shut Jed up immediately.
>
>
> This is what I have already mentioned in our discussions under the pull
> request. If we have smarter APIs in the future (and I believe we will),
> things will be much easier.
>
> However, hbw_free() in memkind does not work this way yet. If I do the
> allocation with normal malloc() and free it with hbw_free(),  I will get an
> SEGV error. If I pair hbw_malloc() with free(), I will also get an SEGV
> error.
>
> If you set the policy to HBW_POLICY_PREFERRED (the default), hbw_malloc()
will pefer high-bandwidth memory, but will resort to using normal DRAM if
that is all that is available.  But that doesn't mean that it starts using
regular malloc() -- it just starts fulfilling its allocation request by
using the lower bandwidth partition.  hbw_malloc() places different
bookkeeping information in the allocated space than malloc(), so calling
free() is going to fail.

>
>  Yes, but why are you using the hwb_ calls instead of using memkind calls?
> There could be good reasons I am not suggesting calling the memkind ...
>
>
> I started to play with memkind last summer. At that time, there were
> plenty of sayings online like this:
> "the *hbwmalloc* interface is stable but *memkind* interface is only
> partially stable."
>
> Perhaps I should try memkind calls since they may become much better.
>

The hbw_malloc() stuff is all just a wrapper for memkind.  I'd suggest
going ahead and using the slightly more complicated but more general
memkind interface.  We'll need to be using something like memkind, anyway,
when we get machines that have even more "kinds" of memory.  (Sorry, it
will happen: nonvolatile RAM technologies are coming.  Actually, we already
have more than one "kind", as "huge TLB" memory is a supported kind.)

>
> Hong (Mr.)
>
>
>
>
> What is interesting is that hbw_malloc() in memkind can automatically
> change its allocator to use DRAM if there is insufficient MCDRAM. So I
> would hope for hbw_free() could change to normal free() when handling data
> allocated by normal malloc().
>
> As explained above, memkind uses the *same* allocator for both DRAM and
MCRAM, it just does the allocation from a different heap partition.  On
machines that need memkind, the solution for getting a DRAM allocation
should not be to use the normal malloc()/free(), but to use memkind and ask
for MEMKIND_DEFAULT instead of MEMKIND_HBW.  memkind_free() will be able to
free allocations of either type if you specify a kind of "0".

--Richard

>
> Hong (Mr.)
>
>
> Sadly, I fear the answer is we don't use memkind because it sucks :-) Calm
> down Jeff, I didn't insult your mother.
>
>
>
> Barry
>
> But, as you say, there are issues in other cases, like with -malloc_debug.
>
> --Richard
>
> On Thu, Mar 9, 2017 at 4:19 PM, Barry Smith <bsmith at mcs.anl.gov> wrote:
>
> Using different mallocs for different objects/arrays in PETSc is very iffy
> because each free() has to match the malloc used for that memory. This is
> even true with just -malloc_debug in that certain initialization functions
> in PETSc need to use the raw malloc() because we cannot be sure if the
> (*PetscTrMalloc)() has been set yet and the raw free() that comes at
> PetscFinalize() time needs to be matched with it.
>
> Why not have PetscMalloc() ALWAYS allocate an extra 64 bit space at the
> beginning and put in an integer indicating the malloc family that has been
> used to get the space. PetscFree() would use this integer to determine the
> correct free() to use. A mechanism to register new malloc families could be
> easily done, for example
>
>  PetscMallocRegister(malloc,realloc,free,&basicmalloc);
>  PetscMallocRegister(PetscMallocDebug,PetscReallocDebug,
> PetscFreeDebug,&debugmalloc);
>  PetscMallocRegister(PetscMallocHBW,PetscReallocHBW,PetscFreeHBW,&
> hbwmalloc);
>
>  To change the malloc used you would do PetscMallocPush(debugmalloc);
>  PetscMalloc(....); .... PetscMallocPop();  Note that you can register
> additional malloc families at any time (it doesn't have to be as soon as
> the program starts up).
>
> What is wrong with the model and why shouldn't we use it?
>
> Barry
>
> Notes:
>
> It is easy to implement, so that is not a reason.
>
> The extra memory usage is trivial.
>
> The mapping from integer to malloc() or free() would be a bounds check and
> then accessing the function pointer from a little array so pretty cheap.
>
> if certain mallocs are missing (like PetscMallocHBW) the hbwmalloc
> variable could be set to the basicmalloc value (or some other) so one would
> not need to ifdef if if () code deciding which malloc to use in many places.
>
> It seems so simple something must be fundamentally flawed with it. Even
> with just PetscTrMallocDefault() and PetscMallocAlign() I feel like
> implementing it.
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20170309/f928f5cc/attachment.html>