| Age | Commit message (Collapse) | Author | Lines |
|---|
|
Ideally, we'd probably print the closure substs themselves actually.
|
|
r=michaelwoerister
Add more profiler events
- Adds Start\Stop events for time spent loading incremental query results from disk.
- Adds Start\Stop events for time spent blocked waiting for queries to complete (when parallel queries are enabled).
r? @michaelwoerister
|
|
|
|
This captures time spent blocked when a query is waiting for another
query to finish executing in another thread.
|
|
|
|
|
|
Partially HirId-ify rustc
Another step towards https://github.com/rust-lang/rust/pull/57578.
|
|
Implement more detailed self profiling
Timing data and cache hits/misses are now recorded at the query level.
This allows us to show detailed per query information such as total time
for each query.
To see detailed query information in the summary pass the `-Z verbose`
flag. For example:
```
rustc -Z self-profile -Z verbose hello_world.rs
```
results in something like:
```md
Self profiling results:
| Phase | Time (ms) | Time (%) | Queries | Hits (%)
| ----------------------------------------- | -------------- | -------- | -------------- | --------
| Other | 177 | 54.97 | 8094 | 45.47
| - {time spent not running queries} | 113 | 35.09 | 0 | 0.00
| - const_eval | 16 | 4.97 | 26 | 11.54
| - type_of | 9 | 2.80 | 627 | 27.75
| - const_eval_raw | 8 | 2.48 | 22 | 0.00
| - adt_def | 7 | 2.17 | 381 | 11.55
| - visible_parent_map | 7 | 2.17 | 99 | 98.99
| - item_attrs | 6 | 1.86 | 698 | 50.14
| - item_children | 5 | 1.55 | 2815 | 0.00
| - adt_dtorck_constraint | 4 | 1.24 | 2 | 0.00
| - adt_destructor | 2 | 0.62 | 15 | 86.67
| TypeChecking | 53 | 16.46 | 2834 | 79.89
| - trait_impls_of | 9 | 2.80 | 65 | 86.15
| - evaluate_obligation | 7 | 2.17 | 80 | 2.50
| - const_is_rvalue_promotable_to_static | 6 | 1.86 | 1 | 0.00
| - is_copy_raw | 6 | 1.86 | 29 | 58.62
| - rvalue_promotable_map | 6 | 1.86 | 2 | 50.00
| - {time spent not running queries} | 6 | 1.86 | 0 | 0.00
| - typeck_item_bodies | 5 | 1.55 | 1 | 0.00
| - typeck_tables_of | 5 | 1.55 | 19 | 94.74
| - dropck_outlives | 2 | 0.62 | 1 | 0.00
| - layout_raw | 1 | 0.31 | 668 | 87.87
| Linking | 48 | 14.91 | 43 | 46.51
| - {time spent not running queries} | 48 | 14.91 | 0 | 0.00
| Codegen | 29 | 9.01 | 420 | 61.90
| - {time spent not running queries} | 16 | 4.97 | 0 | 0.00
| - collect_and_partition_mono_items | 11 | 3.42 | 13 | 92.31
| - mir_const | 1 | 0.31 | 1 | 0.00
| - mir_validated | 1 | 0.31 | 3 | 66.67
| Expansion | 14 | 4.35 | 0 | 0.00
| - {time spent not running queries} | 14 | 4.35 | 0 | 0.00
| BorrowChecking | 1 | 0.31 | 12 | 41.67
| - borrowck | 1 | 0.31 | 2 | 50.00
| Parsing | 0 | 0.00 | 0 | 0.00
Optimization level: No
Incremental: off
```
<details>
<summary>Rendered</summary>
Self profiling results:
| Phase | Time (ms) | Time (%) | Queries | Hits (%)
| ----------------------------------------- | -------------- | -------- | -------------- | --------
| **Other** | **177** | **54.97** | **8094** | **45.47**
| - {time spent not running queries} | 113 | 35.09 | 0 | 0.00
| - const_eval | 16 | 4.97 | 26 | 11.54
| - type_of | 9 | 2.80 | 627 | 27.75
| - const_eval_raw | 8 | 2.48 | 22 | 0.00
| - adt_def | 7 | 2.17 | 381 | 11.55
| - visible_parent_map | 7 | 2.17 | 99 | 98.99
| - item_attrs | 6 | 1.86 | 698 | 50.14
| - item_children | 5 | 1.55 | 2815 | 0.00
| - adt_dtorck_constraint | 4 | 1.24 | 2 | 0.00
| - adt_destructor | 2 | 0.62 | 15 | 86.67
| TypeChecking | 53 | 16.46 | 2834 | 79.89
| - trait_impls_of | 9 | 2.80 | 65 | 86.15
| - evaluate_obligation | 7 | 2.17 | 80 | 2.50
| - const_is_rvalue_promotable_to_static | 6 | 1.86 | 1 | 0.00
| - is_copy_raw | 6 | 1.86 | 29 | 58.62
| - rvalue_promotable_map | 6 | 1.86 | 2 | 50.00
| - {time spent not running queries} | 6 | 1.86 | 0 | 0.00
| - typeck_item_bodies | 5 | 1.55 | 1 | 0.00
| - typeck_tables_of | 5 | 1.55 | 19 | 94.74
| - dropck_outlives | 2 | 0.62 | 1 | 0.00
| - layout_raw | 1 | 0.31 | 668 | 87.87
| Linking | 48 | 14.91 | 43 | 46.51
| - {time spent not running queries} | 48 | 14.91 | 0 | 0.00
| Codegen | 29 | 9.01 | 420 | 61.90
| - {time spent not running queries} | 16 | 4.97 | 0 | 0.00
| - collect_and_partition_mono_items | 11 | 3.42 | 13 | 92.31
| - mir_const | 1 | 0.31 | 1 | 0.00
| - mir_validated | 1 | 0.31 | 3 | 66.67
| Expansion | 14 | 4.35 | 0 | 0.00
| - {time spent not running queries} | 14 | 4.35 | 0 | 0.00
| BorrowChecking | 1 | 0.31 | 12 | 41.67
| - borrowck | 1 | 0.31 | 2 | 50.00
| Parsing | 0 | 0.00 | 0 | 0.00
Optimization level: No
Incremental: off
</details>
cc @nikomatsakis @michaelwoerister @Zoxc
Fixes #54141
|
|
Co-Authored-By: Gabriel Smith <yodaldevoid@users.noreply.github.com>
|
|
|
|
|
|
Timing data and cache hits/misses are now recorded at the query level.
This allows us to show detailed per query information such as total time
for each query.
To see detailed query information in the summary pass the `-Z verbose`
flag. For example:
```
rustc -Z self-profile -Z verbose hello_world.rs
```
|
|
|
|
|
|
Co-authored-by: Alexander Regueiro <alexreg@me.com>
|
|
std: Force `Instant::now()` to be monotonic
This commit is an attempt to force `Instant::now` to be monotonic
through any means possible. We tried relying on OS/hardware/clock
implementations, but those seem buggy enough that we can't rely on them
in practice. This commit implements the same hammer Firefox recently
implemented (noted in #56612) which is to just keep whatever the lastest
`Instant::now()` return value was in memory, returning that instead of
the OS looks like it's moving backwards.
Closes #48514
Closes #49281
cc #51648
cc #56560
Closes #56612
Closes #56940
|
|
|
|
|
|
This commit is an attempt to force `Instant::now` to be monotonic
through any means possible. We tried relying on OS/hardware/clock
implementations, but those seem buggy enough that we can't rely on them
in practice. This commit implements the same hammer Firefox recently
implemented (noted in #56612) which is to just keep whatever the lastest
`Instant::now()` return value was in memory, returning that instead of
the OS looks like it's moving backwards.
Closes #48514
Closes #49281
cc #51648
cc #56560
Closes #56612
Closes #56940
|
|
Add support for trait-objects without a principal
The hard-error version of #56481 - should be merged after we do something about the `traitobject` crate.
Fixes #33140.
Fixes #57057.
r? @nikomatsakis
|
|
Don't emit `Unevaluated` from `const_eval`
cc @eddyb @RalfJung
|
|
This makes sure they are printed in a compiler-version-independent
order, avoiding ui test instability.
|
|
|
|
should be a pure refactoring.
|
|
It used to display as just `impl`
|
|
|
|
|
|
|
|
|
|
In NLL, ReVid is all there is, but I might want to repurpose.
|
|
|
|
Say "trait" instead of "type" in diagnostics for `dyn Trait`
|
|
|
|
|
|
|
|
Use a function to access the Hir map to be able to turn it into a query later
r? @eddyb
|
|
Send textual profile data to stderr, not stdout
This makes it possible to run with RUSTFLAGS="-Zself-profile" without
having to redirect compiler output; otherwise Cargo will error out due
to non-empty compiler stdout.
r? @wesleywiser
|
|
|
|
This makes it possible to run with RUSTFLAGS="-Zself-profile" without
having to redirect compiler output; otherwise Cargo will error out due
to non-empty compiler stdout.
|
|
|
|
(compiler front-ends).
|
|
|
|
Fix self profiler ICE on Windows
Fixes #51648
|
|
On Windows, the high-resolution timestamp api doesn't seem to always be
monotonic. This can cause panics when the self-profiler uses the
`Instant` api to find elapsed time.
Work around this by detecting the case where now is less than the start
time and just use 0 elapsed ticks as the measurement.
Fixes #51648
|
|
|
|
|
|
|
|
Fix json output in the self-profiler
Fix missing ',' array element separators and convert NaN's to 0.
cc @Mark-Simulacrum
|
|
|
|
Fix missing ',' array element separators and convert NaN's to 0.
|
