Profiling
When profiling Servo or troubleshooting performance issues, make sure your build is optimised while still allowing for accurate profiling data.
./mach build --profile profiling --with-frame-pointer
- –profile profiling builds Servo with our profiling configuration
- –with-frame-pointer builds Servo with stack frame pointers on all platforms
The matching profile needs to be selected when running Servo:
./mach run --profile profiling http://example.org
Several ways to get profiling information about Servo’s runs:
Tracing with Perfetto
Tracing works by instrumenting specific functions (or portions of code) with explicit annotations such as time_profile! and servo_tracing::*.
It deterministically records every call through instrumented code, but does not provide any visibility into code that is not instrumented.
In contrast, sampling profilers can see everything but only probabilistically and thus get more accurate with long-running loops.
To use tracing, enable related compile-time features:
./mach build --profile profiling --features tracing tracing-perfetto
Then run Servo with the SERVO_TRACING environment variable set to EnvFilter directives to select which traces to enable:
SERVO_TRACING=… ./mach run --profile profiling http://example.org
For example:
SERVO_TRACING=offdisables all tracing (this is the default)SERVO_TRACING=traceenables all tracing (yields massive traces due tocompositing)SERVO_TRACING=[{servo_profiling}]does the same, since we implicitly filter onservo_profilingSERVO_TRACING=infowould enable only theinfolevel and above, but we don’t yet use levelsSERVO_TRACING=layoutenables tracing in thelayoutcrate onlySERVO_TRACING=trace,compositing=offenables all tracing except in thecompositingcrateSERVO_TRACING=[handle_reflow]enables tracing in spans named “handle_reflow” or their descendants
This creates a servo.pftrace file in the current directory, which can be visualized in ui.perfetto.dev.
Interval Profiling
Using the -p option followed by a number (time period in seconds), you can spit out profiling information to the terminal periodically. To do so, run Servo on the desired site (URLs and local file paths are both supported) with profiling enabled:
./mach run --profile profiling http://example.org -p 5
In the example above, while Servo is still running (AND is processing new passes), the profiling information is printed to the terminal every 5 seconds.
Once the page has loaded, hit ESC (or close the app) to exit. Profiling output will be provided, broken down by area of the browser and URL. For example, you might get output of the form below:
_category_ _incremental?_ _iframe?_ _url_ _mean (ms)_ _median (ms)_ _min (ms)_ _max (ms)_ _events_
Painting N/A N/A N/A 6.8177 0.9512 0.0035 30.7573 6
Layout yes yes http://example.org/ 0.0016 0.0016 0.0016 0.0016 1
Layout no yes http://example.org/ 14.4966 14.4966 14.4966 14.4966 1
ScriptParseHTML no yes http://example.org/ 0.8507 1.7009 0.0004 1.7009 2
TimeToFirstPaint no yes http://example.org/ 0.0000 0.0000 0.0000 0.0000 1
TimeToFirstContentfulPaint no yes http://example.org/ 0.0000 0.0000 0.0000 0.0000 1
_url_ _blocked layout queries_
In this example, when loading the page we performed one full layout and one incremental layout passes.
TSV Profiling
Using the -p option followed by a file name, you can spit out profiling information of Servo’s execution to a TSV (tab-separated because certain url contained commas) file.
The information is written to the file only upon Servo’s termination.
This works well with the -x, -z, and -o options so that performance information can be collected during automated runs.
Example usage:
./mach run --profile profiling http://example.org -zxo out.png -p out.tsv
The formats of the profiling information in the Interval and TSV Profiling options are the essentially the same; the url names are not truncated in the TSV Profiling option.
Generating Timelines
Add the --profiler-trace-path /timeline/output/path.html flag to output the profiling data as a self contained HTML timeline.
Because it is a self contained file (all CSS and JS is inline), it is easy to share, upload, or link to from bug reports.
./mach run --profile profiling http://example.org -p 5 --profiler-trace-path trace.html
Usage:
-
Use the mouse wheel or trackpad scrolling, with the mouse focused along the top of the timeline, to zoom the viewport in or out.
-
Grab the selected area along the top and drag left or right to side scroll.
-
Hover over a trace to show more information.
Hacking
The JS, CSS, and HTML for the timeline comes from fitzgen/servo-trace-dump and there is a script in that repo for updating servo’s copy.
All other code is in the components/profile/ directory.
Sampling profiler
Servo includes a sampling profiler which generates profiles that can be opened in the Gecko profiling tools. To use them:
- Run Servo, loading the page you wish to profile
- Press Ctrl+P (or Cmd+P on macOS) to start the profiler (the console should show “Enabling profiler”)
- Press Ctrl+P (or Cmd+P on macOS) to stop the profiler (the console should show “Stopping profiler”)
- Keep Servo running until the symbol resolution is complete (the console should show a final “Resolving N/N”)
- Run
python etc/profilicate.py samples.json >gecko_samples.jsonto transform the profile into a format that the Gecko profiler understands - Load
gecko_samples.jsoninto https://profiler.firefox.com/
To control the output filename, set the PROFILE_OUTPUT environment variable.
To control the sampling rate (default 10ms), set the SAMPLING_RATE environment variable.
Memory Profiling
- Run Servoshell normally
- Open a new tab
- Navigate to
about:memory - Click
Measure - See a report similar to this:
115.15 MiB -- explicit
101.15 MiB -- jemalloc-heap-unclassified
14.00 MiB -- url(http://example.org/)
10.01 MiB -- layout-thread
10.00 MiB -- font-context
0.00 MiB -- stylist
0.00 MiB -- display-list
4.00 MiB -- js
2.75 MiB -- malloc-heap
1.00 MiB -- gc-heap
0.56 MiB -- decommitted
0.35 MiB -- used
0.06 MiB -- unused
0.02 MiB -- admin
0.25 MiB -- non-heap
0.00 MiB -- memory-cache
0.00 MiB -- private
0.00 MiB -- public
121.89 MiB -- jemalloc-heap-active
111.16 MiB -- jemalloc-heap-allocated
203.02 MiB -- jemalloc-heap-mapped
272.61 MiB -- resident
Using macOS Instruments
Xcode has a instruments tool to profile easily.
First, you need to install Xcode instruments:
xcode-select --install
Second, install cargo-instruments via Homebrew:
brew install cargo-instruments
Then, you can simply run it via CLI:
cargo instruments -t Allocations
Here are some links and resources for help with Instruments (Some will stream only on Safari):
- cargo-instruments on crates.io
- Using Time Profiler in Instruments
- Profiling in Depth
- System Trace in Depth
- Threads, virtual memory, and locking
- Core Data Performance Optimization and Debugging
- Learning Instruments
Profiling WebRender
When running Servoshell, press CTRL+F12 to show (or hide) the WebRender overlay.
Webpage snapshots
It is possible to use the mitmproxy tool to intercept servo traffic and create local snapshot (dump) of an arbitrary web-page, to then serve locally for profiling purposes.
mitmproxy support several ways to intersept the traffic including a proxy mode at port :8080, so you can set the browser to just:
./target/release/servo \
--pref=network_http_proxy_uri=http://127.0.0.1:8080 \
--ignore-certificate-errors
[!info] The default
mitmproxycerts are in the~/.mitmproxyor you can generate some usingmitmproxy, but I have just set my browser to ignore cert errors
[!warning] ignoring certs is easy, but be cautious of risks
Default mitmproxy
On a default network, the mitmproxy creates a local proxy server at :8080 and by setting it in browser or passing http_proxy=locahost:8080 and/or https_proxy=localhost:8080 (and by optionally unsetting the no_proxy) you can dump and serve the traffic.
Creating a dump
mitmproxy -w <dumpfile>
Serving a dump
mitmproxy --serve-replay <dumpfile>
The resulted dump file is about ~5MB per page, so it can get large pretty fast, as the tool is very verbose and can store pictures.
Chain-proxy
In case of a another primary proxy connection, we need to pass the upstream to the main proxy from mitmproxy and if the main proxy also has custom certificates, it is crucial to pass them, or to ignore them
[!warning] ignoring certs is easy, but be cautious of risks
Creating a dump
mitmproxy --mode upstream:${http_proxy} -w <dump-path>\
--set ssl_insecure=true
#### Serving the dump
```bash
mitmproxy -v --server-replay ~/dev/recodings/servo_org_3.dump \
--set server_replay_extra=404 \
--set server_replay_ignore_host=true \
--set connection_strategy=lazy \
--set server_replay_reuse=true
[!info] the
replay_extraandreplay_reuseare optional, and may cause unexpected behaviour
OpenHarmony
It is possible to use the tool to intersept remote phone traffic including OpenHarmony targets. Open a reverse proxy port using hdc and then run the servo with proxy and certs set up.
reverse port
hdc rport tcp:8080 tcp:8080
run with args
hdc shell aa start -a EntryAbility \
-b org.servo.servo -U https://servo.org \
--psn=--pref=network_http_proxy_uri=http://127.0.0.1:8080 \
--psn=--ignore-certificate-errors