[mojo-edk] Initialize message buffers after alloc

[mojo-edk] Initialize message buffers after alloc

We weren't zeroing the allocated memory before filling
the buffer. Due to padding this means there could be
uninitialized bytes. Bad news. Fixed.

Should fix failures on memory bots:

https://build.chromium.org/p/chromium.memory.fyi/builders/Windows%20Content%20Browser%20%28DrMemory%20full%29%20%284%29/builds/7154

BUG=None
TBR=thestig@chromium.org,darin@chromium.org

Committed: https://crrev.com/9df1bc579dbac1e8cd51987bca64a8a326eb8a07
Cr-Commit-Position: refs/heads/master@{#371737}

[Ken Rockot(use gerrit already), 2016-01-27 04:20:37]

Description was changed from

==========
[mojo-edk] Initialize message buffers after alloc

We weren't zeroing the allocated memory before filling
the buffer. Due to padding this means there could be
uninitialized bytes. Bad news. Fixed.

BUG=None
TBR=thestig@chromium.org,darin@chromium.org
==========

to

==========
[mojo-edk] Initialize message buffers after alloc

We weren't zeroing the allocated memory before filling
the buffer. Due to padding this means there could be
uninitialized bytes. Bad news. Fixed.

Should fix failures on memory bots:

https://build.chromium.org/p/chromium.memory.fyi/builders/Windows%20Content%2...

BUG=None
TBR=thestig@chromium.org,darin@chromium.org
==========

[Ken Rockot(use gerrit already), 2016-01-27 04:20:47]

The CQ bit was checked by rockot@chromium.org

[commit-bot: I haz the power, 2016-01-27 04:21:20]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1641493002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1641493002/1

[darin (slow to review), 2016-01-27 05:08:55]

We had attempted to avoid clearing the memory buffer with the intent that the
memory would be completely overwritten subsequently as we write out the
messages. Maybe this won't show up in any profiles, and if that's true, then we
could simplify some other code that was attempting to zero out padding bytes.

[commit-bot: I haz the power, 2016-01-27 05:25:02]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-01-27 05:25:56]

Description was changed from

==========
[mojo-edk] Initialize message buffers after alloc

We weren't zeroing the allocated memory before filling
the buffer. Due to padding this means there could be
uninitialized bytes. Bad news. Fixed.

Should fix failures on memory bots:

https://build.chromium.org/p/chromium.memory.fyi/builders/Windows%20Content%2...

BUG=None
TBR=thestig@chromium.org,darin@chromium.org
==========

to

==========
[mojo-edk] Initialize message buffers after alloc

We weren't zeroing the allocated memory before filling
the buffer. Due to padding this means there could be
uninitialized bytes. Bad news. Fixed.

Should fix failures on memory bots:

https://build.chromium.org/p/chromium.memory.fyi/builders/Windows%20Content%2...

BUG=None
TBR=thestig@chromium.org,darin@chromium.org

Committed: https://crrev.com/9df1bc579dbac1e8cd51987bca64a8a326eb8a07
Cr-Commit-Position: refs/heads/master@{#371737}
==========

[commit-bot: I haz the power, 2016-01-27 05:25:56]

Patchset 1 (id:??) landed as
https://crrev.com/9df1bc579dbac1e8cd51987bca64a8a326eb8a07
Cr-Commit-Position: refs/heads/master@{#371737}

[Ken Rockot(use gerrit already), 2016-01-27 05:28:05]

Aye. Given that it's always a 7-byte span I'm pretty sure this is from
newly introduced padding in PortDescriptor since adding |peer_closed|. If
we have any issues here with performance (and you know, maybe even if we
don't), we can move away from alignment attributes and instead do
explicitly initialized padding as with other types, and static asserts to
enforce alignment.

On Tue, Jan 26, 2016 at 9:25 PM, commit-bot@chromium.org via
codereview.chromium.org <reply@chromiumcodereview-hr.appspotmail.com> wrote:

> Patchset 1 (id:??) landed as
> https://crrev.com/9df1bc579dbac1e8cd51987bca64a8a326eb8a07
> Cr-Commit-Position: refs/heads/master@{#371737}
>
> https://codereview.chromium.org/1641493002/
>

-- 
You received this message because you are subscribed to the Google Groups
"Chromium-reviews" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.