[Testing] Be more aggressive when killing android processes via ADB.

[Testing] Be more aggressive when killing android processes via ADB.

Without -9, this was not killing the forwarder.

BUG=237820
NOTRY=true

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=198212

[Ted C, 2013-05-03 16:47:53]

michaelbai|pilard - PTAL

This was failing for me locally while trying to repro a build failure on our
release bot.

[craigdh, 2013-05-03 16:53:40]

lgtm.

[commit-bot: I haz the power, 2013-05-03 17:38:46]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tedchoc@chromium.org/14721005/1

[commit-bot: I haz the power, 2013-05-03 22:10:52]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tedchoc@chromium.org/14721005/1

[commit-bot: I haz the power, 2013-05-03 22:13:04]

Change committed as 198212

[Philippe, 2013-05-06 16:56:55]

On 2013/05/03 22:13:04, I haz the power (commit-bot) wrote:
> Change committed as 198212

I think this might cause us some problems soon if it's not already the case.
Some processes including forwarder need to do some work when they get killed. If
there is a problem with forwarder I can look into it but I don't think we should
move to a SIGKILL behavior by default. To me this should only be done in last
resort (after a failed SIGTERM attempt).

What do you guys think?

[craigdh, 2013-05-06 17:07:51]

On 2013/05/06 16:56:55, Philippe wrote:
> On 2013/05/03 22:13:04, I haz the power (commit-bot) wrote:
> > Change committed as 198212
> 
> I think this might cause us some problems soon if it's not already the case.
> Some processes including forwarder need to do some work when they get killed.
If
> there is a problem with forwarder I can look into it but I don't think we
should
> move to a SIGKILL behavior by default. To me this should only be done in last
> resort (after a failed SIGTERM attempt).
> 
> What do you guys think?

A SIGTERM first sounds reasonable, but won't we hit issues anyway if the
forwarder is hanging around?

[Ted C, 2013-05-06 17:08:24]

On 2013/05/06 16:56:55, Philippe wrote:
> On 2013/05/03 22:13:04, I haz the power (commit-bot) wrote:
> > Change committed as 198212
> 
> I think this might cause us some problems soon if it's not already the case.
> Some processes including forwarder need to do some work when they get killed.
If
> there is a problem with forwarder I can look into it but I don't think we
should
> move to a SIGKILL behavior by default. To me this should only be done in last
> resort (after a failed SIGTERM attempt).

I think it would be fine to move to a attempt to shut it down more gracefully,
but then force it if that fails.

Seems like we could just parameterize these two methods with a force kill flag
and leave it up to the caller to do as they choose.  In the case of the
forwarder that was giving me problems, I'd still rather it force kill the app
internally acknowledging a unclean shutdown than me having to dig into the phone
myself and kill it manually.

> 
> What do you guys think?