Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Remove some stack frames from `.async` calls #66398

Merged
merged 1 commit into from Nov 15, 2019

Conversation

@sfackler
Copy link
Member

sfackler commented Nov 14, 2019

The Context argument is currently smuggled through TLS for
async-generated futures. The current infrastructure is closure-based,
and results in an extra 6 stack frames when .awaiting an async-generated
future!

  12: foo::async_b::{{closure}}
             at src/main.rs:10
  13: <std::future::GenFuture<T> as core::future::future::Future>::poll::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  14: std::future::set_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:79
  15: <std::future::GenFuture<T> as core::future::future::Future>::poll
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  16: std::future::poll_with_tls_context::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  17: std::future::get_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:111
  18: std::future::poll_with_tls_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  19: foo::async_a::{{closure}}
             at src/main.rs:6

While the long (medium?) term solution is to remove the use of TLS
entirely, we can improve things a bit in the meantime. In particular,
this commit does 2 things:

  1. get_task_context has been inlined into poll_with_tls_context,
    removing 2 frames (16 and 17 above).
  2. set_task_context now returns a guard type that resets the TLS
    rather than taking a closure, removing 2 frames (13 and 14 above).

We can also remove frame 18 by removing poll_with_tls_context in favor
of a get_task_context function which returns a guard, but that
requires adjusting the code generated for .await, so I've left that off
for now.

The `Context` argument is currently smuggled through TLS for
async-generated futures. The current infrastructure is closure-based,
and results in an extra 6 stack frames when .awaiting an async-generated
future!

```
  12: foo::async_b::{{closure}}
             at src/main.rs:10
  13: <std::future::GenFuture<T> as core::future::future::Future>::poll::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  14: std::future::set_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:79
  15: <std::future::GenFuture<T> as core::future::future::Future>::poll
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  16: std::future::poll_with_tls_context::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  17: std::future::get_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:111
  18: std::future::poll_with_tls_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  19: foo::async_a::{{closure}}
             at src/main.rs:6
```

While the long (medium?) term solution is to remove the use of TLS
entirely, we can improve things a bit in the meantime. In particular,
this commit does 2 things:

1. `get_task_context` has been inlined into `poll_with_tls_context`,
    removing 2 frames (16 and 17 above).
2. `set_task_context` now returns a guard type that resets the TLS
    rather than taking a closure, removing 2 frames (13 and 14 above).

We can also remove frame 18 by removing `poll_with_tls_context` in favor
of a `get_task_context` function which returns a guard, but that
requires adjusting the code generated for .await, so I've left that off
for now.
@rust-highfive

This comment has been minimized.

Copy link
Collaborator

rust-highfive commented Nov 14, 2019

r? @rkruppe

(rust_highfive has picked a reviewer for you, use r? to override)

@Centril

This comment has been minimized.

Copy link
Member

Centril commented Nov 14, 2019

Thanks! r? @Centril @bors r+

@bors

This comment has been minimized.

Copy link
Contributor

bors commented Nov 14, 2019

📌 Commit 3fe7cfc has been approved by Centril

@rust-highfive rust-highfive assigned Centril and unassigned rkruppe Nov 14, 2019
tmandry added a commit to tmandry/rust that referenced this pull request Nov 14, 2019
Remove some stack frames from `.async` calls

The `Context` argument is currently smuggled through TLS for
async-generated futures. The current infrastructure is closure-based,
and results in an extra 6 stack frames when .awaiting an async-generated
future!

```
  12: foo::async_b::{{closure}}
             at src/main.rs:10
  13: <std::future::GenFuture<T> as core::future::future::Future>::poll::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  14: std::future::set_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:79
  15: <std::future::GenFuture<T> as core::future::future::Future>::poll
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  16: std::future::poll_with_tls_context::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  17: std::future::get_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:111
  18: std::future::poll_with_tls_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  19: foo::async_a::{{closure}}
             at src/main.rs:6
```

While the long (medium?) term solution is to remove the use of TLS
entirely, we can improve things a bit in the meantime. In particular,
this commit does 2 things:

1. `get_task_context` has been inlined into `poll_with_tls_context`,
    removing 2 frames (16 and 17 above).
2. `set_task_context` now returns a guard type that resets the TLS
    rather than taking a closure, removing 2 frames (13 and 14 above).

We can also remove frame 18 by removing `poll_with_tls_context` in favor
of a `get_task_context` function which returns a guard, but that
requires adjusting the code generated for .await, so I've left that off
for now.
bors added a commit that referenced this pull request Nov 14, 2019
Rollup of 8 pull requests

Successful merges:

 - #65557 (rename Error::iter_chain() and remove Error::iter_sources())
 - #66197 (Push `ast::{ItemKind, ImplItemKind}::OpaqueTy` hack down into lowering)
 - #66306 (Remove cannot mutate statics in initializer of another static error)
 - #66338 (Update mdbook.)
 - #66394 (Fix two OOM issues related to `ConstProp`)
 - #66396 (Make a test compatible across python versions.)
 - #66398 (Remove some stack frames from `.async` calls)
 - #66410 (miri: helper methods for max values of machine's usize/isize)

Failed merges:

r? @ghost
tmandry added a commit to tmandry/rust that referenced this pull request Nov 14, 2019
Remove some stack frames from `.async` calls

The `Context` argument is currently smuggled through TLS for
async-generated futures. The current infrastructure is closure-based,
and results in an extra 6 stack frames when .awaiting an async-generated
future!

```
  12: foo::async_b::{{closure}}
             at src/main.rs:10
  13: <std::future::GenFuture<T> as core::future::future::Future>::poll::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  14: std::future::set_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:79
  15: <std::future::GenFuture<T> as core::future::future::Future>::poll
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  16: std::future::poll_with_tls_context::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  17: std::future::get_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:111
  18: std::future::poll_with_tls_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  19: foo::async_a::{{closure}}
             at src/main.rs:6
```

While the long (medium?) term solution is to remove the use of TLS
entirely, we can improve things a bit in the meantime. In particular,
this commit does 2 things:

1. `get_task_context` has been inlined into `poll_with_tls_context`,
    removing 2 frames (16 and 17 above).
2. `set_task_context` now returns a guard type that resets the TLS
    rather than taking a closure, removing 2 frames (13 and 14 above).

We can also remove frame 18 by removing `poll_with_tls_context` in favor
of a `get_task_context` function which returns a guard, but that
requires adjusting the code generated for .await, so I've left that off
for now.
bors added a commit that referenced this pull request Nov 14, 2019
Rollup of 9 pull requests

Successful merges:

 - #65557 (rename Error::iter_chain() and remove Error::iter_sources())
 - #66197 (Push `ast::{ItemKind, ImplItemKind}::OpaqueTy` hack down into lowering)
 - #66306 (Remove cannot mutate statics in initializer of another static error)
 - #66338 (Update mdbook.)
 - #66391 (Do not ICE in `if` without `else` in `async fn`)
 - #66396 (Make a test compatible across python versions.)
 - #66398 (Remove some stack frames from `.async` calls)
 - #66410 (miri: helper methods for max values of machine's usize/isize)
 - #66418 (Link to tracking issue in HIR const-check error code description)

Failed merges:

r? @ghost
JohnTitor added a commit to JohnTitor/rust that referenced this pull request Nov 15, 2019
Remove some stack frames from `.async` calls

The `Context` argument is currently smuggled through TLS for
async-generated futures. The current infrastructure is closure-based,
and results in an extra 6 stack frames when .awaiting an async-generated
future!

```
  12: foo::async_b::{{closure}}
             at src/main.rs:10
  13: <std::future::GenFuture<T> as core::future::future::Future>::poll::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  14: std::future::set_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:79
  15: <std::future::GenFuture<T> as core::future::future::Future>::poll
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  16: std::future::poll_with_tls_context::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  17: std::future::get_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:111
  18: std::future::poll_with_tls_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  19: foo::async_a::{{closure}}
             at src/main.rs:6
```

While the long (medium?) term solution is to remove the use of TLS
entirely, we can improve things a bit in the meantime. In particular,
this commit does 2 things:

1. `get_task_context` has been inlined into `poll_with_tls_context`,
    removing 2 frames (16 and 17 above).
2. `set_task_context` now returns a guard type that resets the TLS
    rather than taking a closure, removing 2 frames (13 and 14 above).

We can also remove frame 18 by removing `poll_with_tls_context` in favor
of a `get_task_context` function which returns a guard, but that
requires adjusting the code generated for .await, so I've left that off
for now.
bors added a commit that referenced this pull request Nov 15, 2019
Rollup of 12 pull requests

Successful merges:

 - #65557 (rename Error::iter_chain() and remove Error::iter_sources())
 - #66013 (Avoid hashing the key twice in `get_query()`.)
 - #66306 (Remove cannot mutate statics in initializer of another static error)
 - #66338 (Update mdbook.)
 - #66388 (Do not ICE on recovery from unmet associated type bound obligation)
 - #66390 (Fix ICE when trying to suggest `Type<>` instead of `Type()`)
 - #66391 (Do not ICE in `if` without `else` in `async fn`)
 - #66398 (Remove some stack frames from `.async` calls)
 - #66407 (Add more tests for fixed ICEs)
 - #66410 (miri: helper methods for max values of machine's usize/isize)
 - #66418 (Link to tracking issue in HIR const-check error code description)
 - #66419 (Don't warn labels beginning with `_` on unused_labels lint)

Failed merges:

r? @ghost
JohnTitor added a commit to JohnTitor/rust that referenced this pull request Nov 15, 2019
Remove some stack frames from `.async` calls

The `Context` argument is currently smuggled through TLS for
async-generated futures. The current infrastructure is closure-based,
and results in an extra 6 stack frames when .awaiting an async-generated
future!

```
  12: foo::async_b::{{closure}}
             at src/main.rs:10
  13: <std::future::GenFuture<T> as core::future::future::Future>::poll::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  14: std::future::set_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:79
  15: <std::future::GenFuture<T> as core::future::future::Future>::poll
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  16: std::future::poll_with_tls_context::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  17: std::future::get_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:111
  18: std::future::poll_with_tls_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  19: foo::async_a::{{closure}}
             at src/main.rs:6
```

While the long (medium?) term solution is to remove the use of TLS
entirely, we can improve things a bit in the meantime. In particular,
this commit does 2 things:

1. `get_task_context` has been inlined into `poll_with_tls_context`,
    removing 2 frames (16 and 17 above).
2. `set_task_context` now returns a guard type that resets the TLS
    rather than taking a closure, removing 2 frames (13 and 14 above).

We can also remove frame 18 by removing `poll_with_tls_context` in favor
of a `get_task_context` function which returns a guard, but that
requires adjusting the code generated for .await, so I've left that off
for now.
bors added a commit that referenced this pull request Nov 15, 2019
Rollup of 12 pull requests

Successful merges:

 - #65557 (rename Error::iter_chain() and remove Error::iter_sources())
 - #66013 (Avoid hashing the key twice in `get_query()`.)
 - #66306 (Remove cannot mutate statics in initializer of another static error)
 - #66338 (Update mdbook.)
 - #66388 (Do not ICE on recovery from unmet associated type bound obligation)
 - #66390 (Fix ICE when trying to suggest `Type<>` instead of `Type()`)
 - #66391 (Do not ICE in `if` without `else` in `async fn`)
 - #66394 (Fix two OOM issues related to `ConstProp`)
 - #66398 (Remove some stack frames from `.async` calls)
 - #66410 (miri: helper methods for max values of machine's usize/isize)
 - #66418 (Link to tracking issue in HIR const-check error code description)
 - #66419 (Don't warn labels beginning with `_` on unused_labels lint)

Failed merges:

r? @ghost
JohnTitor added a commit to JohnTitor/rust that referenced this pull request Nov 15, 2019
Remove some stack frames from `.async` calls

The `Context` argument is currently smuggled through TLS for
async-generated futures. The current infrastructure is closure-based,
and results in an extra 6 stack frames when .awaiting an async-generated
future!

```
  12: foo::async_b::{{closure}}
             at src/main.rs:10
  13: <std::future::GenFuture<T> as core::future::future::Future>::poll::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  14: std::future::set_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:79
  15: <std::future::GenFuture<T> as core::future::future::Future>::poll
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:43
  16: std::future::poll_with_tls_context::{{closure}}
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  17: std::future::get_task_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:111
  18: std::future::poll_with_tls_context
             at /rustc/4560ea788cb760f0a34127156c78e2552949f734/src/libstd/future.rs:121
  19: foo::async_a::{{closure}}
             at src/main.rs:6
```

While the long (medium?) term solution is to remove the use of TLS
entirely, we can improve things a bit in the meantime. In particular,
this commit does 2 things:

1. `get_task_context` has been inlined into `poll_with_tls_context`,
    removing 2 frames (16 and 17 above).
2. `set_task_context` now returns a guard type that resets the TLS
    rather than taking a closure, removing 2 frames (13 and 14 above).

We can also remove frame 18 by removing `poll_with_tls_context` in favor
of a `get_task_context` function which returns a guard, but that
requires adjusting the code generated for .await, so I've left that off
for now.
bors added a commit that referenced this pull request Nov 15, 2019
Rollup of 12 pull requests

Successful merges:

 - #65557 (rename Error::iter_chain() and remove Error::iter_sources())
 - #66013 (Avoid hashing the key twice in `get_query()`.)
 - #66306 (Remove cannot mutate statics in initializer of another static error)
 - #66338 (Update mdbook.)
 - #66388 (Do not ICE on recovery from unmet associated type bound obligation)
 - #66390 (Fix ICE when trying to suggest `Type<>` instead of `Type()`)
 - #66391 (Do not ICE in `if` without `else` in `async fn`)
 - #66398 (Remove some stack frames from `.async` calls)
 - #66410 (miri: helper methods for max values of machine's usize/isize)
 - #66418 (Link to tracking issue in HIR const-check error code description)
 - #66419 (Don't warn labels beginning with `_` on unused_labels lint)
 - #66428 (Cleanup unused function)

Failed merges:

r? @ghost
@bors bors merged commit 3fe7cfc into rust-lang:master Nov 15, 2019
4 checks passed
4 checks passed
pr Build #20191114.7 succeeded
Details
pr (Linux mingw-check) Linux mingw-check succeeded
Details
pr (Linux x86_64-gnu-llvm-6.0) Linux x86_64-gnu-llvm-6.0 succeeded
Details
pr (Linux x86_64-gnu-tools) Linux x86_64-gnu-tools succeeded
Details
@sfackler sfackler deleted the sfackler:no-async-nesting branch Nov 16, 2019
@jakubadamw

This comment has been minimized.

Copy link
Contributor

jakubadamw commented Nov 29, 2019

@sfackler, was the visibility change to set_task_context from public to private intentional? There are a few crates out there relying on it (futures-async-stream and gen-stream, for example).

@sfackler

This comment has been minimized.

Copy link
Member Author

sfackler commented Nov 29, 2019

Yes, it was intentional. The TLS interface is not a public API and will be removed entirely in the future.

AerialX added a commit to AerialX/core-futures-tls that referenced this pull request Dec 6, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
6 participants
You can’t perform that action at this time.