<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Thu, Mar 9, 2017 at 5:38 PM, Barry Smith <span dir="ltr"><<a href="mailto:bsmith@mcs.anl.gov" target="_blank">bsmith@mcs.anl.gov</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span class="gmail-"><br>
> On Mar 9, 2017, at 7:18 PM, Richard Mills <<a href="mailto:richardtmills@gmail.com">richardtmills@gmail.com</a>> wrote:<br>
><br>
> Hi Barry,<br>
><br>
> 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.<br>
<br>
</span> As I said the extra space is 64 bit. Now if you want 128 bit alignment we could put a 128 bit.<br></blockquote><div><br></div><div>KNL wants 64 *byte* alignment, not 64 bit.<br> <br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<span class="gmail-"><br>
> 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...<br>
<br>
</span> What alignment boundaries are useful for Intel processes? 64 yup, 128, 256, 512 ? Does higher values provide better performance for SIMD etc?<br></blockquote><div><br></div><div>KNL has 64 byte cache lines, hence the preference for 64 byte alignment. This isn't as big an issue as it was on the previous generation "Knights Corner" (KNC) Xeon Phi, which really suffered from alignment issues and required two instructions for an unaligned load. On KNL only one instruction is needed. You are using the cache better, of course, if you can honor those cache line boundaries, though.<br> <br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<span class="gmail-"><br>
><br>
> 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.<br>
<br>
</span> Hmm, Hong, how come we don't use this? I didn't realize it worked this way. This would shut Jed up immediately.<br>
<br>
Sadly, I fear the answer is we don't use memkind because it sucks :-) Calm down Jeff, I didn't insult your mother.<br></blockquote><div><br></div><div>You just need to use the slightly more complicated memkind_malloc (or memkind_posix_memalign) instead of hbw_malloc(). memkind_free() does ask for type of memory, but according to the man page:<br><br>"In cases where the kind is unknown in the context of the call to memkind_free() 0 can be given as the kind specified to memkind_free() but this will require a look up that can be bypassed by specifying a non-zero value."<br><br></div><div>This seems like a non-issue -- how often is a PETSc code going to be doing performance critical free() calls?<br><br></div><div>--Richard<br><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<span class="gmail-HOEnZb"><font color="#888888"><br>
<br>
<br>
Barry<br>
</font></span><div class="gmail-HOEnZb"><div class="gmail-h5"><br>
> But, as you say, there are issues in other cases, like with -malloc_debug.<br>
><br>
> --Richard<br>
><br>
> On Thu, Mar 9, 2017 at 4:19 PM, Barry Smith <<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>> wrote:<br>
><br>
> 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.<br>
><br>
> 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<br>
><br>
> PetscMallocRegister(malloc,<wbr>realloc,free,&basicmalloc);<br>
> PetscMallocRegister(<wbr>PetscMallocDebug,<wbr>PetscReallocDebug,<wbr>PetscFreeDebug,&debugmalloc);<br>
> PetscMallocRegister(<wbr>PetscMallocHBW,<wbr>PetscReallocHBW,PetscFreeHBW,&<wbr>hbwmalloc);<br>
><br>
> 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).<br>
><br>
> What is wrong with the model and why shouldn't we use it?<br>
><br>
> Barry<br>
><br>
> Notes:<br>
><br>
> It is easy to implement, so that is not a reason.<br>
><br>
> The extra memory usage is trivial.<br>
><br>
> 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.<br>
><br>
> 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.<br>
><br>
> It seems so simple something must be fundamentally flawed with it. Even with just PetscTrMallocDefault() and PetscMallocAlign() I feel like implementing it.<br>
><br>
><br>
><br>
><br>
><br>
><br>
<br>
</div></div></blockquote></div><br></div></div>