docs/task_scheduler_migration.md - Issue 2959563002: Sequence[d]TaskRunner

Side by Side Diff: docs/task_scheduler_migration.md

Issue 2959563002: Sequence[d]TaskRunner (Closed)

Patch Set: Created 3 years, 5 months ago

Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.

Jump to:

« no previous file with comments | « base/threading/thread_task_runner_handle.h ('k') | docs/threading_and_tasks.md » ('j') | no next file with comments »

OLD	NEW
1 # TaskScheduler Migration	1 # TaskScheduler Migration

2	2

3 [TOC]	3 [TOC]

4	4

5 ## Overview	5 ## Overview

6	6

7 [`base/task_scheduler/post_task.h`](https://cs.chromium.org/chromium/src/base/ta sk_scheduler/post_task.h)	7 [`base/task_scheduler/post_task.h`](https://cs.chromium.org/chromium/src/base/ta sk_scheduler/post_task.h)

8 was introduced to Chrome in Q1. The API is fully documented under [Threading and	8 was introduced to Chrome in Q1. The API is fully documented under [Threading and

9 Tasks in Chrome](threading_and_tasks.md). This page will go into more details	9 Tasks in Chrome](threading_and_tasks.md). This page will go into more details

10 about how to migrate callers of existing APIs to TaskScheduler.	10 about how to migrate callers of existing APIs to TaskScheduler.

(...skipping 63 matching lines...) Expand 10 before \| Expand all \| Expand 10 after Loading...
74 decided otherwise by a developer.	74 decided otherwise by a developer.

75	75

76 As a developer your goal is to get rid of all uses of BrowserThread::FOO in your	76 As a developer your goal is to get rid of all uses of BrowserThread::FOO in your

77 assigned files by:	77 assigned files by:

78 1. Splitting things into their own execution sequence (i.e. post to a TaskRunne r	78 1. Splitting things into their own execution sequence (i.e. post to a TaskRunne r

79 obtained from post_task.h -- see [Threading and Tasks in	79 obtained from post_task.h -- see [Threading and Tasks in

80 Chrome](threading_and_tasks.md) for details).	80 Chrome](threading_and_tasks.md) for details).

81 2. Removing the plumbing: if GetTaskRunnerForThread(BrowserThread::FOO) is	81 2. Removing the plumbing: if GetTaskRunnerForThread(BrowserThread::FOO) is

82 passed down into a component the prefered paradigm is to remove all of that	82 passed down into a component the prefered paradigm is to remove all of that

83 plumbing and simply have the leaf layers requiring a TaskRunner get it from	83 plumbing and simply have the leaf layers requiring a TaskRunner get it from

84 base::CreateSequenceTaskRunnerWithTraits() directly.	84 base::CreateSequencedTaskRunnerWithTraits() directly.

85 3. Ideally migrating from a single-threaded context to a	85 3. Ideally migrating from a single-threaded context to a

86 [much preferred](threading_and_tasks.md#Prefer-Sequences-to-Threads) sequenc ed context.	86 [much preferred](threading_and_tasks.md#Prefer-Sequences-to-Threads) sequenc ed context.

87 * Note: if your tasks use COM APIs (Component Object Model on Windows),	87 * Note: if your tasks use COM APIs (Component Object Model on Windows),

88 you'll need to use CreateCOMSTATaskRunnerWithTraits() and sequencing will	88 you'll need to use CreateCOMSTATaskRunnerWithTraits() and sequencing will

89 not be an option (there are DCHECKs in place that will fire if your task	89 not be an option (there are DCHECKs in place that will fire if your task

90 uses COM without being on a COM initialized TaskRunner).	90 uses COM without being on a COM initialized TaskRunner).

91	91

92 ## Relevant single-thread -> sequence mappings	92 ## Relevant single-thread -> sequence mappings

93	93

94 * base::SingleThreadTaskRunner -> base::SequencedTaskRunner	94 * base::SingleThreadTaskRunner -> base::SequencedTaskRunner

(...skipping 28 matching lines...) Expand all Loading...
123 * Everything in a file/component needs to run on the same sequence but there	123 * Everything in a file/component needs to run on the same sequence but there

124 isn't a clear place to own/access the common SequencedTaskRunner =>	124 isn't a clear place to own/access the common SequencedTaskRunner =>

125 base::Lazy(Sequenced\|SingleThread\|COMSTA)TaskRunner.	125 base::Lazy(Sequenced\|SingleThread\|COMSTA)TaskRunner.

126 * Mojo isn't sequence-friendly yet ([coming soon](https://crbug.com/678155)).	126 * Mojo isn't sequence-friendly yet ([coming soon](https://crbug.com/678155)).

127 Use base::CreateSingleThreadTaskRunnerWithTraits() instead of	127 Use base::CreateSingleThreadTaskRunnerWithTraits() instead of

128 base::CreateSequencedTaskRunnerWithTraits() for sequences that need to use	128 base::CreateSequencedTaskRunnerWithTraits() for sequences that need to use

129 mojo constructs for now (tag with TODO against https://crbug.com/678155).	129 mojo constructs for now (tag with TODO against https://crbug.com/678155).

130 * For anything else, ping [base/task_scheduler/OWNERS](https://cs.chromium.org/c hromium/src/base/task_scheduler/OWNERS)	130 * For anything else, ping [base/task_scheduler/OWNERS](https://cs.chromium.org/c hromium/src/base/task_scheduler/OWNERS)

131 or [scheduler-dev@chromium.org](https://groups.google.com/a/chromium.org/forum /#!forum/scheduler-dev),	131 or [scheduler-dev@chromium.org](https://groups.google.com/a/chromium.org/forum /#!forum/scheduler-dev),

132 thanks!	132 thanks!

OLD	NEW