Trigger old-space GC when scavange made the external heap of old-space larger than max old-space si…

Trigger old-space GC when scavange made the external heap of old-space larger than max old-space size.

BUG=
R=koda@google.com

Committed: https://code.google.com/p/dart/source/detail?r=34647

[Anders Johnsen, 2014-04-01 08:05:00]

I'm not 100% this is the right approach, so feel free to create your own CL for
it.

The issue I'm trying to solve, is that scavanges can make the external-size of
old-space grow larger than max old-space size. When reading a lot of large
files, this can lead to C++ allocation failure (OOM).

[koda, 2014-04-01 16:58:07]

https://codereview.chromium.org/220113004/diff/1/runtime/vm/heap.cc
File runtime/vm/heap.cc (right):

https://codereview.chromium.org/220113004/diff/1/runtime/vm/heap.cc#newcode219
runtime/vm/heap.cc:219: (FLAG_new_gen_heap_size * MBInWords))) {
Thanks for this fix. It does address the immediate problem with repeated
promotion overflowing old space. However, I now realize there is still an issue
with direct allocation in old space not being tested against the growth policy
and max capacity until something else triggers GC.

I will work on a CL that tries to unify growth control for external with rest of
old space. (It's slightly tricky since the current policy is based on
page-granularity.) But until that's done this patch gives some protection.

As for the code: You meant *old* gen heap size here, right? Also, please
consider encapsulating this check into a new method like
PageSpace::NeedExternalGC() {
return ExternalInWords() > max_capacity_in_words_;
}

[Anders Johnsen, 2014-04-02 11:16:28]

PTAL.

Looking forward to your CL :)

https://codereview.chromium.org/220113004/diff/1/runtime/vm/heap.cc
File runtime/vm/heap.cc (right):

https://codereview.chromium.org/220113004/diff/1/runtime/vm/heap.cc#newcode219
runtime/vm/heap.cc:219: (FLAG_new_gen_heap_size * MBInWords))) {
On 2014/04/01 16:58:07, koda wrote:
> Thanks for this fix. It does address the immediate problem with repeated
> promotion overflowing old space. However, I now realize there is still an
issue
> with direct allocation in old space not being tested against the growth policy
> and max capacity until something else triggers GC.
> 
> I will work on a CL that tries to unify growth control for external with rest
of
> old space. (It's slightly tricky since the current policy is based on
> page-granularity.) But until that's done this patch gives some protection.
> 
> As for the code: You meant *old* gen heap size here, right? Also, please
> consider encapsulating this check into a new method like
> PageSpace::NeedExternalGC() {
> return ExternalInWords() > max_capacity_in_words_;
> }

Yes, refactor error - should be old. Changed to your suggestion, though I use 

    used_in_words_ + ExternalInWords() > max_capacity_in_words_;

as internal + external should be < max.

[koda, 2014-04-02 14:12:16]

lgtm

[Anders Johnsen, 2014-04-02 14:25:23]

Committed patchset #2 manually as r34647 (presubmit successful).