wasmtime / Issue #2662 Surprisingly slow module compilation · git-wasmtime

Stream: git-wasmtime

Topic: wasmtime / Issue #2662 Surprisingly slow module compilation

Wasmtime GitHub notifications bot (Feb 17 2021 at 21:45):

Thanks for opening a bug report! Please answer the questions below
if they're relevant and delete this text before submitting.

What are the steps to reproduce the issue?

I have pushed a repository at https://github.com/wycats/wand/tree/str-wasmtime. The branch str-wasmtime is self-contained.

The script ./script/all.sh runs all of the build steps. The final step, ./script/run.sh, runs an executable, which is a Rust executable that uses the wasmtime embedding API. Its source is located in ./crates/wand-cli. The Rust source for the wasm is located in ./src.

The build scripts use wasm-pack to compile and optimize the wasm. They copy the .wasm file into a data directory in the wand-cli crate, which uses include_bytes! to get the bytes.

The executable has a bunch of println!s that list out the execution steps and how much time has elapsed.

What do you expect to happen? What does actually happen? Does it panic, and
if so, with which assertion?

I expect the Module::new step to be relatively fast. Instead, it takes several seconds on my machine.

Which Wasmtime version / commit hash / branch are you using?

0.22.

If relevant, can you include some extra information about your environment?
(Rust version, operating system, architecture...)
$ rustc --version
rustc 1.52.0-nightly (d1206f950 2021-02-15)
$ cargo --version
cargo 1.51.0-nightly (ab64d1393 2021-02-10)
$ uname -a
Linux XPS-13-2019 4.19.84-microsoft-standard #1 SMP Wed Nov 13 11:44:37 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description:    Ubuntu 20.04.2 LTS
Release:        20.04
Codename:       focal
I'm on Ubuntu 20.04 inside of WSL2 on Windows 10.

Wasmtime GitHub notifications bot (Feb 17 2021 at 21:45):

wycats labeled Issue #2662:

Thanks for opening a bug report! Please answer the questions below
if they're relevant and delete this text before submitting.

What are the steps to reproduce the issue?

I have pushed a repository at https://github.com/wycats/wand/tree/str-wasmtime. The branch str-wasmtime is self-contained.

The script ./script/all.sh runs all of the build steps. The final step, ./script/run.sh, runs an executable, which is a Rust executable that uses the wasmtime embedding API. Its source is located in ./crates/wand-cli. The Rust source for the wasm is located in ./src.

The build scripts use wasm-pack to compile and optimize the wasm. They copy the .wasm file into a data directory in the wand-cli crate, which uses include_bytes! to get the bytes.

The executable has a bunch of println!s that list out the execution steps and how much time has elapsed.

What do you expect to happen? What does actually happen? Does it panic, and
if so, with which assertion?

I expect the Module::new step to be relatively fast. Instead, it takes several seconds on my machine.

Which Wasmtime version / commit hash / branch are you using?

0.22.

If relevant, can you include some extra information about your environment?
(Rust version, operating system, architecture...)
$ rustc --version
rustc 1.52.0-nightly (d1206f950 2021-02-15)
$ cargo --version
cargo 1.51.0-nightly (ab64d1393 2021-02-10)
$ uname -a
Linux XPS-13-2019 4.19.84-microsoft-standard #1 SMP Wed Nov 13 11:44:37 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description:    Ubuntu 20.04.2 LTS
Release:        20.04
Codename:       focal
I'm on Ubuntu 20.04 inside of WSL2 on Windows 10.

Wasmtime GitHub notifications bot (Feb 17 2021 at 22:53):

alexcrichton commented on Issue #2662:

Thanks for the report! To help minimize this a bit further, this is the wasm file in question and as you've mentioned the slower-than-expected part here is Module::new, aka compiling the module.

Executing your project I see compile times of ~2s, and to poke around a bit I made a smaller script which simply compiles this module. The timings on the main branch (soon to be 0.23.0) I get are:

old backend, opt speed - 634.ms

old backend, opt size - 641ms

old backend, opt none - 561ms

new backend, opt speed - 766ms

new backend, opt size - 752ms

new backend, opt none - 869ms

So it looks like the new backend is not actually a win here. I'm not entirely sure how to explain the discrepancy of your project taking 2.5s to compile and an isolated script taking ~2x faster. If I use the release commit of 0.22.0 I get slightly slower compile times by ~250ms, but still not in the ~2s range.

Wasmtime GitHub notifications bot (Feb 17 2021 at 23:05):

alexcrichton commented on Issue #2662:

Aha I have found the source of the 2x slowdown over master. Your project enables logging with tracing-subscriber (no doubt because our wasi example does the same) but this has the adverse affect apparently of having log_enabled!(LogLevel::Debug) returning true. This means that very expensive debug logs like this are happening at runtime. That's accounting for a very large portion of the runtime in your example.

After commenting out the tracing-subscriber business it runs in the same amount of time I saw in my script (e.g. 2x faster)

Wasmtime GitHub notifications bot (Feb 17 2021 at 23:05):

alexcrichton edited a comment on Issue #2662:

Aha I have found the source of the 2x slowdown over main. Your project enables logging with tracing-subscriber (no doubt because our wasi example does the same) but this has the adverse affect apparently of having log_enabled!(LogLevel::Debug) returning true. This means that very expensive debug logs like this are happening at runtime. That's accounting for a very large portion of the runtime in your example.

After commenting out the tracing-subscriber business it runs in the same amount of time I saw in my script (e.g. 2x faster)

Wasmtime GitHub notifications bot (Feb 17 2021 at 23:53):

alexcrichton commented on Issue #2662:

For further speedups beyond the numbers above there's a zulip thread going now as well -- https://bytecodealliance.zulipchat.com/#narrow/stream/225524-cranelift-new-backend/topic/wasm.20file.20that.20is.20slower.20to.20compile.20in.20new.20BE.20than.20old.20BE

Wasmtime GitHub notifications bot (Feb 17 2021 at 23:55):

peterhuene commented on Issue #2662:

So I believe the way tracing-subscriber uses the tracing-log crate has the unfortunate consequence of formatting log strings for every level, regardless of whether or not the logger is "enabled" for that level (in this case, dependent upon the env filter respecting RUST_LOG).

The tracing-log feature is enabled by default in tracing-subscriber, which causes initialization of the global logger.

This uses a default Builder that ends up setting the maximum level to the most verbose level.

All the level-specific log macros end up as a invocation of the log! macro, which checks the level against log::max_level() to see if it needs to format the string and calls __private_api_log. __private_api_log calls Log::enabled on the logger to check to see if it wants the event; but at that point the string has already been formatted as only log::max_level is checked for that.

And since the max level is the most verbose, all strings are formatted, but the logger in tracing-log ends up telling the log crate it isn't interested in the event after all.

Wasmtime GitHub notifications bot (Feb 18 2021 at 01:10):

peterhuene edited a comment on Issue #2662:

So I believe the way tracing-subscriber uses the tracing-log crate has the unfortunate consequence of formatting log strings for every level, regardless of whether or not the logger is "enabled" for that level (in this case, dependent upon the env filter respecting RUST_LOG).

The tracing-log feature is enabled by default in tracing-subscriber, which causes initialization of the global logger.

This uses a default Builder that ends up setting the maximum level to the most verbose level.

All the level-specific log macros end up as a invocation of the log! macro, which checks the level against log::max_level() to see if it needs to format the string and calls __private_api_log. __private_api_log calls Log::enabled on the logger to check to see if it wants the event; but at that point the string has already been formatted as only log::max_level is checked for that.

And since the max level is the most verbose, all strings are formatted, but the logger in tracing-log ends up telling the log crate it isn't interested in the event after all.

As Alex also noted, this would also affect any explicit checks to see if the log level is enabled (log_enabled! has the same check against log::max_level), so extra work is also being performed outside of the macro string formatting too.

Wasmtime GitHub notifications bot (Feb 18 2021 at 18:42):

wycats commented on Issue #2662:

Your project enables logging with tracing-subscriber (no doubt because our wasi example does the same)

You're right! I stared at that line a bunch of times and kept telling myself "I copied it from an official example", and never got around to trying the whole thing with the line commented out.

Wasmtime GitHub notifications bot (Feb 18 2021 at 19:00):

peterhuene edited a comment on Issue #2662:

So I believe the way tracing-subscriber uses the tracing-log crate also has the unfortunate consequence of formatting log strings for every level, regardless of whether or not the logger is "enabled" for that level (in this case, dependent upon the env filter respecting RUST_LOG).

The tracing-log feature is enabled by default in tracing-subscriber, which causes initialization of the global logger.

This uses a default Builder that ends up setting the maximum level to the most verbose level.

All the level-specific log macros end up as a invocation of the log! macro, which checks the level against log::max_level() to see if it needs to format the string and calls __private_api_log. __private_api_log calls Log::enabled on the logger to check to see if it wants the event; but at that point the string has already been formatted as only log::max_level is checked for that.

And since the max level is the most verbose, all strings are formatted, but the logger in tracing-log ends up telling the log crate it isn't interested in the event after all.

As Alex also noted, this would also affect any explicit checks to see if the log level is enabled (log_enabled! has the same check against log::max_level), so extra work is also being performed outside of the macro string formatting too.

Wasmtime GitHub notifications bot (Feb 18 2021 at 21:41):

peterhuene edited a comment on Issue #2662:

So I believe the way tracing-subscriber uses the tracing-log crate also has the unfortunate consequence of formatting log strings for every level, regardless of whether or not the logger is "enabled" for that level (in this case, dependent upon the env filter respecting RUST_LOG).

~~The tracing-log feature is enabled by default in tracing-subscriber, which causes initialization of the global logger.~~

~~This uses a default Builder that ends up setting the maximum level to the most verbose level.~~

All the level-specific log macros end up as a invocation of the log! macro, which checks the level against log::max_level() to see if it needs to format the string and calls __private_api_log. __private_api_log calls Log::enabled on the logger to check to see if it wants the event; but at that point the string has already been formatted as only log::max_level is checked for that.

~~And since the max level is the most verbose, all strings are formatted, but the logger in tracing-log ends up telling the log crate it isn't interested in the event after all.~~

As Alex also noted, this would also affect any explicit checks to see if the log level is enabled (log_enabled! has the same check against log::max_level), so extra work is also being performed outside of the macro string formatting too.

Edit: my analysis was wrong as I was looking at the wrong function. String formatting is delayed until the logger is asked if the level is enabled.

This only affects the log_enabled! macro it seems.

Wasmtime GitHub notifications bot (Feb 18 2021 at 21:52):

peterhuene edited a comment on Issue #2662:

So I believe the way tracing-subscriber uses the tracing-log crate also has the unfortunate consequence of formatting log strings for every level, regardless of whether or not the logger is "enabled" for that level (in this case, dependent upon the env filter respecting RUST_LOG).

~~The tracing-log feature is enabled by default in tracing-subscriber, which causes initialization of the global logger.~~

~~This uses a default Builder that ends up setting the maximum level to the most verbose level.~~

All the level-specific log macros end up as a invocation of the log! macro, which checks the level against log::max_level() to see if it needs to format the string and calls __private_api_log. __private_api_log calls Log::enabled on the logger to check to see if it wants the event; but at that point the string has already been formatted as only log::max_level is checked for that.

~~And since the max level is the most verbose, all strings are formatted, but the logger in tracing-log ends up telling the log crate it isn't interested in the event after all.~~

As Alex also noted, this would also affect any explicit checks to see if the log level is enabled (log_enabled! has the same check against log::max_level), so extra work is also being performed outside of the macro string formatting too.

Edit: my analysis was wrong as I was looking at the wrong function (accidentally was looking at __private_api_enabled and saw the &str arg and my brain assumed a formatted message rather than the target name). String formatting is delayed until the logger is asked if the level is enabled.

This only affects the log_enabled! macro it seems.

Wasmtime GitHub notifications bot (Feb 21 2021 at 21:34):

softprops commented on Issue #2662:

I was just going to reach out about this when I found this open issue.

I was wondering if anyone has done a baseline comparison with wasmer. I recently came across this article which mentions a breakthrough in compile time performance. I was wondering how wasmtime compares.

The context for me is a small project that runs fastly wasm apps inside a local http server https://github.com/softprops/fasttime

The compilation step where I load the wasm module can take seconds. I was wondering if work was being done to improve this

Wasmtime GitHub notifications bot (Feb 21 2021 at 21:44):

bjorn3 commented on Issue #2662:

That article mentions parallel compilation as reason that it is faster. This is already the default with Wasmtime.

Wasmtime GitHub notifications bot (Feb 22 2021 at 15:38):

alexcrichton commented on Issue #2662:

@softprops do you have a wasm module that you can share so we can dig in?

Wasmtime GitHub notifications bot (Feb 22 2021 at 17:15):

softprops commented on Issue #2662:

Sure. Where's the best place to upload it?

In ci, I build a test app and then run it though some integration tests. It's uploaded as an artifact so I think you can download it directly.

Here's a recent GitHub actions job https://github.com/softprops/fasttime/actions/runs/586883035

Alternatively if there's anything I can do on my end, trace logging etc, I'd be happy to help. Wasmtime is such an amazing project!

Wasmtime GitHub notifications bot (Feb 22 2021 at 17:21):

alexcrichton commented on Issue #2662:

That works! You can also upload directly here in your comment by compressing it with gzip (GitHub doesn't allow raw wasm file uploads).

I did some timing though and your wasm module compiles in 86ms for me. That means that something deeper may be at play here, perhaps you could open a separate issue for this?

Wasmtime GitHub notifications bot (Feb 23 2021 at 05:48):

softprops commented on Issue #2662:

holy smokes that's fast! could it be the wasmtime version I'm using (0.22) by chance?

I'm mainly getting a sense from this code which loads the wasm file from a path.

Wasmtime GitHub notifications bot (Feb 23 2021 at 15:09):

alexcrichton commented on Issue #2662:

What sort of timings are you seeing for compiling that module? I don't think we had major perf improvements in 0.22, so if you're an order of magnitude different or so from what I'm seeing then there's probably a different issue.

Wasmtime GitHub notifications bot (Feb 23 2021 at 20:30):

pchickey commented on Issue #2662:

The latest tracing-subscriber 0.2.16 resolves the root cause of the issue Alex determined early in this thread. Downstream users should be able to cargo update to get this fix.

Wasmtime GitHub notifications bot (Feb 27 2021 at 05:29):

softprops commented on Issue #2662:

What sort of timings are you seeing for compiling that module? I don't think we had major perf improvements in 0.22, so if you're an order of magnitude different or so from what I'm seeing then there's probably a different issue.

I'll look into upgrading to wasmtime@0.23 this weekend but with wasmtime@0.22 is not uncommon for a 2.7mb wasm file to take multiple seconds to compile with Module::from_file(&engine, file)?; is about 6s on average for me which is why my head started spinning when you mentioned 82ms
◌  Loading module...
 ✔ Loaded module in 6.273588s
I'm using the default Engine instance and calling Module::from_file once here. I haven't yet tried loading the file separately and then calling Module::new with just the bytes.

If its helpful for extra context I'm on a mac using rust 1.49
rustc --version
rustc 1.49.0 (e1884a8e3 2020-12-29)

cargo --version
cargo 1.49.0 (d00d64df9 2020-12-05)
 uname -a
Darwin doug-mbp 20.3.0 Darwin Kernel Version 20.3.0: Thu Jan 21 00:07:06 PST 2021; root:xnu-7195.81.3~1/RELEASE_X86_64 x86_64
If I could get my 6s module compile time down to your 82ms it would be game changing for turn around time on testing for me.

Wasmtime GitHub notifications bot (Feb 27 2021 at 06:30):

softprops commented on Issue #2662:

Upgrading to 0.23 for me didn't seem to make a different in module compile times for me

Wasmtime GitHub notifications bot (Feb 27 2021 at 11:24):

lqd commented on Issue #2662:

@softprops I'm not exactly sure about these steps to reproduce but I cloned fasttime just now (but locally changed user-agent-parser to depend on onig this way to make my life easier on windows), built the tests/app for wasm32-wasi, and have these results:
$ ll -h tests/app/target/wasm32-wasi/release/app.wasm
-rw-r--r-- 2 lqd 197121 3.2M Feb 27 11:02 tests/app/target/wasm32-wasi/release/app.wasm

$ cargo run --release -- -w ./tests/app/target/wasm32-wasi/release/app.wasm
    Finished release [optimized] target(s) in 0.30s
     Running `target\release\fasttime.exe -w ./tests/app/target/wasm32-wasi/release/app.wasm`
 ◌  Loading module...
 ✔ Loaded module in 150.6526ms ✨
DEBUG: {}
 ● Listening on http://127.0.0.1:3000

Wasmtime GitHub notifications bot (Feb 27 2021 at 11:47):

lqd edited a comment on Issue #2662:

@softprops I'm not exactly sure about these steps to reproduce but I cloned fasttime just now (but locally changed user-agent-parser to depend on onig this way to make my life easier on windows), built the tests/app for wasm32-wasi, and have these results:
$ ll -h tests/app/target/wasm32-wasi/release/app.wasm
-rw-r--r-- 2 lqd 197121 3.2M Feb 27 11:02 tests/app/target/wasm32-wasi/release/app.wasm

$ cargo run --release -- -w ./tests/app/target/wasm32-wasi/release/app.wasm
    Finished release [optimized] target(s) in 0.30s
     Running `target\release\fasttime.exe -w ./tests/app/target/wasm32-wasi/release/app.wasm`
 ◌  Loading module...
 ✔ Loaded module in 150.6526ms ✨
DEBUG: {}
 ● Listening on http://127.0.0.1:3000
EDIT: those measurements are on nightly from a couple days ago.

On 1.46.0, the wasm is smaller (the 2.7mb size you mentioned) and fasttime loads it here in 113ms as well)

Wasmtime GitHub notifications bot (Feb 27 2021 at 11:48):

lqd edited a comment on Issue #2662:

@softprops I'm not exactly sure about these steps to reproduce but I cloned fasttime just now (but locally changed user-agent-parser to depend on onig this way to make my life easier on windows), built the tests/app for wasm32-wasi, and have these results:
$ ll -h tests/app/target/wasm32-wasi/release/app.wasm
-rw-r--r-- 2 lqd 197121 3.2M Feb 27 11:02 tests/app/target/wasm32-wasi/release/app.wasm

$ cargo run --release -- -w ./tests/app/target/wasm32-wasi/release/app.wasm
    Finished release [optimized] target(s) in 0.30s
     Running `target\release\fasttime.exe -w ./tests/app/target/wasm32-wasi/release/app.wasm`
 ◌  Loading module...
 ✔ Loaded module in 150.6526ms ✨
DEBUG: {}
 ● Listening on http://127.0.0.1:3000
EDIT: those measurements are on a nightly from a couple days ago.

On 1.46.0, the wasm is smaller (the 2.7mb size you mentioned) and fasttime loads it here in 113ms as well)

Wasmtime GitHub notifications bot (Feb 27 2021 at 12:36):

softprops commented on Issue #2662:

Wow. I wonder if this is just a macOS specific issue

Wasmtime GitHub notifications bot (Feb 27 2021 at 12:37):

softprops edited a comment on Issue #2662:

Upgrading to 0.23 for me didn't seem to make a difference in module compile times for me

Wasmtime GitHub notifications bot (Feb 27 2021 at 13:55):

lqd edited a comment on Issue #2662:

@softprops I'm not exactly sure about these steps to reproduce but I cloned fasttime just now (but locally changed user-agent-parser to depend on onig this way to make my life easier on windows), built the tests/app for wasm32-wasi, and have these results:
$ ll -h tests/app/target/wasm32-wasi/release/app.wasm
-rw-r--r-- 2 lqd 197121 3.2M Feb 27 11:02 tests/app/target/wasm32-wasi/release/app.wasm

$ cargo run --release -- -w ./tests/app/target/wasm32-wasi/release/app.wasm
    Finished release [optimized] target(s) in 0.30s
     Running `target\release\fasttime.exe -w ./tests/app/target/wasm32-wasi/release/app.wasm`
 ◌  Loading module...
 ✔ Loaded module in 150.6526ms ✨
DEBUG: {}
 ● Listening on http://127.0.0.1:3000
EDIT: those measurements are on a nightly from a couple days ago.

On 1.46.0, the wasm is smaller (the 2.7mb size you mentioned) and fasttime loads it here in 113ms as well)

EDIT 2: I've tested this on linux as well and the numbers are similar

Wasmtime GitHub notifications bot (Feb 27 2021 at 14:17):

lqd edited a comment on Issue #2662:

@softprops I'm not exactly sure about these steps to reproduce but I cloned fasttime just now (but locally changed user-agent-parser to depend on onig this way to make my life easier on windows), built the tests/app for wasm32-wasi, and have these results:
$ ll -h tests/app/target/wasm32-wasi/release/app.wasm
-rw-r--r-- 2 lqd 197121 3.2M Feb 27 11:02 tests/app/target/wasm32-wasi/release/app.wasm

$ cargo run --release -- -w ./tests/app/target/wasm32-wasi/release/app.wasm
    Finished release [optimized] target(s) in 0.30s
     Running `target\release\fasttime.exe -w ./tests/app/target/wasm32-wasi/release/app.wasm`
 ◌  Loading module...
 ✔ Loaded module in 150.6526ms ✨
DEBUG: {}
 ● Listening on http://127.0.0.1:3000
EDIT: those measurements are on a nightly from a couple days ago.

On 1.46.0, the wasm is smaller (the 2.7mb size you mentioned) and fasttime loads it here in 113ms as well)

EDIT 2: I've tested this on linux on the same machine as well and the numbers are similar

Wasmtime GitHub notifications bot (Feb 27 2021 at 15:45):

lqd commented on Issue #2662:

Maybe.

At this point this looks different enough from the OP that it's probably better to open a dedicated issue.

FWIW, I myself can't reproduce a Module::from_file slowdown with your wasm test app on windows, linux, or a 10-year old macbook AIR

Wasmtime GitHub notifications bot (Feb 27 2021 at 16:00):

lqd edited a comment on Issue #2662:

Maybe.

At this point this looks different enough from the OP that it's probably better to open a dedicated issue.

FWIW, I myself can't reproduce a Module::from_file slowdown with your wasm test app on windows, linux, or a 10-year old MacBook Air

Wasmtime GitHub notifications bot (Mar 01 2021 at 15:46):

alexcrichton commented on Issue #2662:

@softprops Checking out the latest softprops/fasttime@94faa0e7bd351b67e8d1444e0bc09c32764111b2 on macOS I get the same timings as @lqd got above (using the same app as well). Loading the module you referenced above it says it compiles in ~90ms.

To confirm, are you sure that you're building fasttime with --release? 60ms to 6s sounds like a Rust release mode thing that would cause the slowdown. Otherwise do you perhaps have any environment variables configured liked RUST_LOG or similar?

Wasmtime GitHub notifications bot (Mar 06 2021 at 06:21):

softprops commented on Issue #2662:

@alexcrichton you win! there's a world of a difference with wasm compile times between debug and release mode. I tried with both and I go from ~6s to ~300ms which is a night and day improvement for me. That's magic!
./target/debug/fasttime -w tests/app/target/wasm32-wasi/release/app.wasm
 ◌  Loading module...
 ✔ Loaded module in 6.793324s ✨
DEBUG: {}
 ● Listening on http://127.0.0.1:3000
target/release/fasttime -w tests/app/target/wasm32-wasi/release/app.wasm
 ◌  Loading module...
 ✔ Loaded module in 309.38ms ✨
DEBUG: {}
 ● Listening on http://127.0.0.1:3000

Wasmtime GitHub notifications bot (Mar 08 2021 at 15:02):

alexcrichton commented on Issue #2662:

Ok great! I'm gonna go ahead and close this issue since I believe this was all largely handled.

Wasmtime GitHub notifications bot (Mar 08 2021 at 15:02):

alexcrichton closed Issue #2662:

Thanks for opening a bug report! Please answer the questions below
if they're relevant and delete this text before submitting.

What are the steps to reproduce the issue?

I have pushed a repository at https://github.com/wycats/wand/tree/str-wasmtime. The branch str-wasmtime is self-contained.

The script ./script/all.sh runs all of the build steps. The final step, ./script/run.sh, runs an executable, which is a Rust executable that uses the wasmtime embedding API. Its source is located in ./crates/wand-cli. The Rust source for the wasm is located in ./src.

The build scripts use wasm-pack to compile and optimize the wasm. They copy the .wasm file into a data directory in the wand-cli crate, which uses include_bytes! to get the bytes.

The executable has a bunch of println!s that list out the execution steps and how much time has elapsed.

What do you expect to happen? What does actually happen? Does it panic, and
if so, with which assertion?

I expect the Module::new step to be relatively fast. Instead, it takes several seconds on my machine.

Which Wasmtime version / commit hash / branch are you using?

0.22.

If relevant, can you include some extra information about your environment?
(Rust version, operating system, architecture...)
$ rustc --version
rustc 1.52.0-nightly (d1206f950 2021-02-15)
$ cargo --version
cargo 1.51.0-nightly (ab64d1393 2021-02-10)
$ uname -a
Linux XPS-13-2019 4.19.84-microsoft-standard #1 SMP Wed Nov 13 11:44:37 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description:    Ubuntu 20.04.2 LTS
Release:        20.04
Codename:       focal
I'm on Ubuntu 20.04 inside of WSL2 on Windows 10.

Last updated: Oct 23 2024 at 20:03 UTC