Bug 228752 - [webkitpy] Narrow set of terminated simulator processes
Summary: [webkitpy] Narrow set of terminated simulator processes
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Tools / Tests (show other bugs)
Version: WebKit Nightly Build
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Jonathan Bedard
URL:
Keywords: InRadar
Depends on:
Blocks:
 
Reported: 2021-08-03 14:06 PDT by Jonathan Bedard
Modified: 2021-08-03 17:56 PDT (History)
7 users (show)

See Also:


Attachments
Patch (3.44 KB, patch)
2021-08-03 14:11 PDT, Jonathan Bedard
ews-feeder: commit-queue-
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Jonathan Bedard 2021-08-03 14:06:12 PDT
We are terminating too many simulator processes in our simulator management code, and we're not recovering when failing to delete a simulated device.
Comment 1 Radar WebKit Bug Importer 2021-08-03 14:08:41 PDT
<rdar://problem/81479508>
Comment 2 Jonathan Bedard 2021-08-03 14:11:39 PDT
Created attachment 434861 [details]
Patch
Comment 3 Aakash Jain 2021-08-03 14:28:23 PDT
Comment on attachment 434861 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=434861&action=review

> Tools/Scripts/webkitpy/xcode/simulated_device.py:-508
> -        if SimulatedDeviceManager._managing_simulator_app:

why don't we need this anymore?
Comment 4 Aakash Jain 2021-08-03 14:29:33 PDT
Overall, how do you come up with this? How do we know that we won't be killing too less (or different thing than needed) after this?
Comment 5 Jonathan Bedard 2021-08-03 14:40:15 PDT
Comment on attachment 434861 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=434861&action=review

>> Tools/Scripts/webkitpy/xcode/simulated_device.py:-508
>> -        if SimulatedDeviceManager._managing_simulator_app:
> 
> why don't we need this anymore?

Because CoreSimulator is mature enough now that we don't need to do this any more, we end up killing processes before they finish tearing themselves down.
Comment 6 Jonathan Bedard 2021-08-03 14:46:28 PDT
(In reply to Aakash Jain from comment #4)
> Overall, how do you come up with this? How do we know that we won't be
> killing too less (or different thing than needed) after this?

We will be killing less, in every case. That's why I did Simulator.app instead of Simulator. I started to play around killing less processes and seeing what that did, it made interacting with simctl much more stable. Which isn't particularly surprising, we shouldn't really need to kill CoreSimulator processes.
Comment 7 Alexey Proskuryakov 2021-08-03 14:58:42 PDT
Comment on attachment 434861 [details]
Patch

Seems worth trying!
Comment 8 EWS 2021-08-03 15:17:20 PDT
Committed r280618 (240231@main): <https://commits.webkit.org/240231@main>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 434861 [details].