device_utils.Install: Calculate host md5s at the same time as device md5s

device_utils.Install: Calculate host md5s at the same time as device md5s

For ChromeModern.apk, this made adb_install_apk.py ~350ms faster on my
machine

BUG=

Committed: https://crrev.com/1374ea5aacd0423b91def81904752cb0c28b7394
Cr-Commit-Position: refs/heads/master@{#342879}

[agrieve, 2015-08-11 17:31:24]

agrieve@chromium.org changed reviewers:
+ jbudorick@chromium.org

[agrieve, 2015-08-11 17:31:25]

Another minor optimization

[jbudorick, 2015-08-11 17:33:21]

This is dependent on your other CL, right? (or it conflicts with it?)

[agrieve, 2015-08-11 17:43:54]

On 2015/08/11 17:33:21, jbudorick wrote:
> This is dependent on your other CL, right? (or it conflicts with it?)

Sorry fixed now. It is.

[jbudorick, 2015-08-11 18:06:40]

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
File build/android/pylib/device/device_utils.py (right):

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
build/android/pylib/device/device_utils.py:1097: def DoComputeHostChecksums():
nit: I've generally been trying to keep local functions within pylib as
lower_case_with_underscores to distinguish them.

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
build/android/pylib/device/device_utils.py:1098: host_checksums_holder[0] =
md5sum.CalculateHostMd5Sums(host_apk_paths)
nit: line break after function definition

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
build/android/pylib/device/device_utils.py:1099: host_thread =
threading.Thread(target=DoComputeHostChecksums)
I'd prefer it if you used a timeout_retry thread:
https://code.google.com/p/chromium/codesearch#chromium/src/build/android/pyli...

(this doesn't need retries, but a timeout would be nice in case something
strange happens)

[agrieve, 2015-08-11 18:15:00]

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
File build/android/pylib/device/device_utils.py (right):

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
build/android/pylib/device/device_utils.py:1097: def DoComputeHostChecksums():
On 2015/08/11 18:06:40, jbudorick wrote:
> nit: I've generally been trying to keep local functions within pylib as
> lower_case_with_underscores to distinguish them.

Done.

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
build/android/pylib/device/device_utils.py:1098: host_checksums_holder[0] =
md5sum.CalculateHostMd5Sums(host_apk_paths)
On 2015/08/11 18:06:40, jbudorick wrote:
> nit: line break after function definition

Done.

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
build/android/pylib/device/device_utils.py:1099: host_thread =
threading.Thread(target=DoComputeHostChecksums)
On 2015/08/11 18:06:40, jbudorick wrote:
> I'd prefer it if you used a timeout_retry thread:
>
https://code.google.com/p/chromium/codesearch#chromium/src/build/android/pyli...
> 
> (this doesn't need retries, but a timeout would be nice in case something
> strange happens)

Since this is a private helper, and all public methods use their own timeout
logic, having a timeout within a timeout seems like overkill.

[jbudorick, 2015-08-11 18:16:54]

lgtm

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
File build/android/pylib/device/device_utils.py (right):

https://codereview.chromium.org/1286763002/diff/40001/build/android/pylib/dev...
build/android/pylib/device/device_utils.py:1099: host_thread =
threading.Thread(target=DoComputeHostChecksums)
On 2015/08/11 at 18:15:00, agrieve wrote:
> On 2015/08/11 18:06:40, jbudorick wrote:
> > I'd prefer it if you used a timeout_retry thread:
> >
https://code.google.com/p/chromium/codesearch#chromium/src/build/android/pyli...
> > 
> > (this doesn't need retries, but a timeout would be nice in case something
> > strange happens)
> 
> Since this is a private helper, and all public methods use their own timeout
logic, having a timeout within a timeout seems like overkill.

ah, right.

[agrieve, 2015-08-11 19:17:27]

The CQ bit was checked by agrieve@chromium.org

[commit-bot: I haz the power, 2015-08-11 19:17:46]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1286763002/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1286763002/60001

[commit-bot: I haz the power, 2015-08-11 20:23:39]

Committed patchset #4 (id:60001)

[commit-bot: I haz the power, 2015-08-11 20:24:21]

Patchset 4 (id:??) landed as
https://crrev.com/1374ea5aacd0423b91def81904752cb0c28b7394
Cr-Commit-Position: refs/heads/master@{#342879}