base/allocator/partition_alloc: make unittests lowmem friendly

base/allocator/partition_alloc: make unittests lowmem friendly

Partition Alloc, which has been recently moved to base/, has tests
that rely on the ability to successfully make large (512 MB)
allocations. Turns out this is not possible on small devices like
ChromeCast, that seem to either have little overcommit margins or
a heavily fragmented address space.

This CL skips those tests on devices that have < 2 GiB of physical
memory. For the same reason, it lowers down the chunk size of the
RepeatedReturnNullDirect test. The rationale of the test is to check
that repeated failures are handled consistently when performing
directly mapped allocations. The chunk size is irrelevant as long as
it is big enough to hit the directly-mmap codepaths (current
threshold = 1 MB).

BUG=675255
Committed: https://crrev.com/3ac3dbf4347193119fa3f31db087fd87db2e99dd
Cr-Commit-Position: refs/heads/master@{#439484}

[Primiano Tucci (use gerrit), 2016-12-19 14:04:42]

The CQ bit was checked by primiano@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-12-19 14:04:56]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Primiano Tucci (use gerrit), 2016-12-19 14:06:22]

primiano@chromium.org changed reviewers:
+ haraken@chromium.org, palmer@chromium.org

[Primiano Tucci (use gerrit), 2016-12-19 14:06:23]

haraken@ can you take a look at this? looks like palmer is ooo for a week and
this test is failing on chromecast

[commit-bot: I haz the power, 2016-12-19 14:14:54]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-12-19 14:14:55]

Dry run: Try jobs failed on following builders:
  linux_chromium_asan_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Primiano Tucci (use gerrit), 2016-12-19 14:48:11]

The CQ bit was checked by primiano@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-12-19 14:48:44]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[haraken, 2016-12-19 15:12:04]

LGTM

[commit-bot: I haz the power, 2016-12-19 15:24:04]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-12-19 15:24:05]

Dry run: This issue passed the CQ dry run.

[Primiano Tucci (use gerrit), 2016-12-19 15:58:42]

The CQ bit was checked by primiano@chromium.org

[commit-bot: I haz the power, 2016-12-19 15:59:04]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-12-19 16:02:59]

CQ is committing da patch.

Bot data: {"patchset_id": 20001, "attempt_start_ts": 1482163122269730,
"parent_rev": "712ab2951db7a2617ac63283084f58a337864cac", "commit_rev":
"1066996e7ca2f01bfa3b5bfaf0d888cb88938234"}

[commit-bot: I haz the power, 2016-12-19 16:03:24]

Description was changed from

==========
base/allocator/partition_alloc: make unittests lowmem friendly

Partition Alloc, which has been recently moved to base/, has tests
that rely on the ability to successfully make large (512 MB)
allocations. Turns out this is not possible on small devices like
ChromeCast, that seem to either have little overcommit margins or
a heavily fragmented address space.

This CL skips those tests on devices that have < 2 GiB of physical
memory. For the same reason, it lowers down the chunk size of the
RepeatedReturnNullDirect test. The rationale of the test is to check
that repeated failures are handled consistently when performing
directly mapped allocations. The chunk size is irrelevant as long as
it is big enough to hit the directly-mmap codepaths (current
threshold = 1 MB).

BUG=675255
==========

to

==========
base/allocator/partition_alloc: make unittests lowmem friendly

Partition Alloc, which has been recently moved to base/, has tests
that rely on the ability to successfully make large (512 MB)
allocations. Turns out this is not possible on small devices like
ChromeCast, that seem to either have little overcommit margins or
a heavily fragmented address space.

This CL skips those tests on devices that have < 2 GiB of physical
memory. For the same reason, it lowers down the chunk size of the
RepeatedReturnNullDirect test. The rationale of the test is to check
that repeated failures are handled consistently when performing
directly mapped allocations. The chunk size is irrelevant as long as
it is big enough to hit the directly-mmap codepaths (current
threshold = 1 MB).

BUG=675255

Review-Url: https://codereview.chromium.org/2589813002
==========

[commit-bot: I haz the power, 2016-12-19 16:03:25]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-12-19 16:06:25]

Description was changed from

==========
base/allocator/partition_alloc: make unittests lowmem friendly

Partition Alloc, which has been recently moved to base/, has tests
that rely on the ability to successfully make large (512 MB)
allocations. Turns out this is not possible on small devices like
ChromeCast, that seem to either have little overcommit margins or
a heavily fragmented address space.

This CL skips those tests on devices that have < 2 GiB of physical
memory. For the same reason, it lowers down the chunk size of the
RepeatedReturnNullDirect test. The rationale of the test is to check
that repeated failures are handled consistently when performing
directly mapped allocations. The chunk size is irrelevant as long as
it is big enough to hit the directly-mmap codepaths (current
threshold = 1 MB).

BUG=675255

Review-Url: https://codereview.chromium.org/2589813002
==========

to

==========
base/allocator/partition_alloc: make unittests lowmem friendly

Partition Alloc, which has been recently moved to base/, has tests
that rely on the ability to successfully make large (512 MB)
allocations. Turns out this is not possible on small devices like
ChromeCast, that seem to either have little overcommit margins or
a heavily fragmented address space.

This CL skips those tests on devices that have < 2 GiB of physical
memory. For the same reason, it lowers down the chunk size of the
RepeatedReturnNullDirect test. The rationale of the test is to check
that repeated failures are handled consistently when performing
directly mapped allocations. The chunk size is irrelevant as long as
it is big enough to hit the directly-mmap codepaths (current
threshold = 1 MB).

BUG=675255

Committed: https://crrev.com/3ac3dbf4347193119fa3f31db087fd87db2e99dd
Cr-Commit-Position: refs/heads/master@{#439484}
==========

[commit-bot: I haz the power, 2016-12-19 16:06:25]

Patchset 2 (id:??) landed as
https://crrev.com/3ac3dbf4347193119fa3f31db087fd87db2e99dd
Cr-Commit-Position: refs/heads/master@{#439484}