mandoline: Reland "Fix ICU initialization".

mandoline: Reland "Fix ICU initialization".

We need to complete the initialization of ICU before we raise the
sandbox. That means we can't pass a file descriptor to the ICU data file
across mojo pipes. Due to how Android handles resources files, we also
can't pass a file descriptor around inside the same process. So pass a
raw pointer to a memory mapped file during the sandbox warm-up phase,
and make a new option to initialize ICU from this raw pointer.

But that just uncovers a bigger issue: we don't always call what was the
sandbox warming code. If it's general initialization that needs to be
called before we run MojoMain(), we need to include initialization code
in most all main.cc implementations. This bakes a base initialize call
into the mojo application library.

This fixes a crash in the page cycler set, which happens when a page
tries to do date operations, which fail because of missing locale data.
This fixes several crashes in local runs.

[[This reland disables the sandbox on the page cycler bots because they
don't have a new enough kernel to be sandboxed(!). It also includes the
icu data in the mojo runner unittests.]]

BUG=546644
First Review URL: https://codereview.chromium.org/1425853003
TBR=jam@chromium.org,sky@chromium.org
R=yzshen@chromium.org

Committed: https://crrev.com/140735acbef15c56d9ddaadd2bc58b5e26692a48
Cr-Commit-Position: refs/heads/master@{#359349}

[Elliot Glaysher, 2015-11-12 00:27:43]

Description was changed from

==========
mandoline: Reland "Fix ICU initialization".

We need to complete the initialization of ICU before we raise the
sandbox. That means we can't pass a file descriptor to the ICU data file
across mojo pipes. Due to how Android handles resources files, we also
can't pass a file descriptor around inside the same process. So pass a
raw pointer to a memory mapped file during the sandbox warm-up phase,
and make a new option to initialize ICU from this raw pointer.

But that just uncovers a bigger issue: we don't always call what was the
sandbox warming code. If it's general initialization that needs to be
called before we run MojoMain(), we need to include initialization code
in most all main.cc implementations. This bakes a base initialize call
into the mojo application library.

This fixes a crash in the page cycler set, which happens when a page
tries to do date operations, which fail because of missing locale data.
This should let us raise the sandbox on the page cycler.

BUG=546644

Review URL: https://codereview.chromium.org/1425853003

Cr-Commit-Position: refs/heads/master@{#358904}
==========

to

==========
mandoline: Reland "Fix ICU initialization".

We need to complete the initialization of ICU before we raise the
sandbox. That means we can't pass a file descriptor to the ICU data file
across mojo pipes. Due to how Android handles resources files, we also
can't pass a file descriptor around inside the same process. So pass a
raw pointer to a memory mapped file during the sandbox warm-up phase,
and make a new option to initialize ICU from this raw pointer.

But that just uncovers a bigger issue: we don't always call what was the
sandbox warming code. If it's general initialization that needs to be
called before we run MojoMain(), we need to include initialization code
in most all main.cc implementations. This bakes a base initialize call
into the mojo application library.

This fixes a crash in the page cycler set, which happens when a page
tries to do date operations, which fail because of missing locale data.
This should let us raise the sandbox on the page cycler.

BUG=546644
First Review URL: https://codereview.chromium.org/1425853003
TBR=jam@chromium.org,sky@chromium.org
R=yzshen@chromium.org
==========

[Elliot Glaysher, 2015-11-12 00:27:43]

erg@chromium.org changed reviewers:
+ jam@chromium.org, sky@chromium.org, yzshen@chromium.org

[Elliot Glaysher, 2015-11-12 00:36:59]

Description was changed from

==========
mandoline: Reland "Fix ICU initialization".

We need to complete the initialization of ICU before we raise the
sandbox. That means we can't pass a file descriptor to the ICU data file
across mojo pipes. Due to how Android handles resources files, we also
can't pass a file descriptor around inside the same process. So pass a
raw pointer to a memory mapped file during the sandbox warm-up phase,
and make a new option to initialize ICU from this raw pointer.

But that just uncovers a bigger issue: we don't always call what was the
sandbox warming code. If it's general initialization that needs to be
called before we run MojoMain(), we need to include initialization code
in most all main.cc implementations. This bakes a base initialize call
into the mojo application library.

This fixes a crash in the page cycler set, which happens when a page
tries to do date operations, which fail because of missing locale data.
This should let us raise the sandbox on the page cycler.

BUG=546644
First Review URL: https://codereview.chromium.org/1425853003
TBR=jam@chromium.org,sky@chromium.org
R=yzshen@chromium.org
==========

to

==========
mandoline: Reland "Fix ICU initialization".

We need to complete the initialization of ICU before we raise the
sandbox. That means we can't pass a file descriptor to the ICU data file
across mojo pipes. Due to how Android handles resources files, we also
can't pass a file descriptor around inside the same process. So pass a
raw pointer to a memory mapped file during the sandbox warm-up phase,
and make a new option to initialize ICU from this raw pointer.

But that just uncovers a bigger issue: we don't always call what was the
sandbox warming code. If it's general initialization that needs to be
called before we run MojoMain(), we need to include initialization code
in most all main.cc implementations. This bakes a base initialize call
into the mojo application library.

This fixes a crash in the page cycler set, which happens when a page
tries to do date operations, which fail because of missing locale data.
This fixes several crashes in local runs.

[[This reland disables the sandbox on the page cycler bots because they
don't have a new enough kernel to be sandboxed(!). It also includes the
icu data in the mojo runner unittests.]]

BUG=546644
First Review URL: https://codereview.chromium.org/1425853003
TBR=jam@chromium.org,sky@chromium.org
R=yzshen@chromium.org
==========

[Elliot Glaysher, 2015-11-12 00:46:52]

yzshen: Sorry, I should have said what I wanted reviewed. Most of this patch was
already reviewed by sky and jam, but the only two places that I changed that
aren't straight merge fixes are the adding of no-sandbox to
tools/.../mandoline_browser_backend.py, and mojo/runner/BUILD.gn to include the
icu data in the android unit test bundles.

[jam, 2015-11-12 03:20:33]

lgtm

[yzshen1, 2015-11-12 18:37:36]

On 2015/11/12 03:20:33, jam wrote:
> lgtm

LGTM

[Elliot Glaysher, 2015-11-12 18:39:49]

The CQ bit was checked by erg@chromium.org

[commit-bot: I haz the power, 2015-11-12 18:41:36]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1431133003/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1431133003/100001

[commit-bot: I haz the power, 2015-11-12 18:52:05]

Committed patchset #6 (id:100001)

[commit-bot: I haz the power, 2015-11-12 20:08:40]

Patchset 6 (id:??) landed as
https://crrev.com/140735acbef15c56d9ddaadd2bc58b5e26692a48
Cr-Commit-Position: refs/heads/master@{#359349}