Dynamically detect support for Wasmtime's pooling allocator #2508
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This commit is intended to address fermyon#2119 and mirror bytecodealliance/wasmtime#8610. The base problem is that some systems are configured with smaller amounts of virtual memory than other systems, for example some aarch64 and riscv64 systems are shown to have only 39 bits of virtual address space rather than the 48 by default on x86_64. This means that the pooling allocator can't be initialized on these platforms since it needs more virtual memory than that. This changes Spin to dynamically choosing whether to use the pooling allocator. It's still used by default in Wasmtime but a dynamic probe is performed to determine whether it's going to work first. While here I also added an env var to control this behavior for an escape hatch if that's needed in the future too. Closes fermyon#2119 Signed-off-by: Alex Crichton <alex@alexcrichton.com>
alexcrichton
force-pushed
the
probe-for-the-pooling-allocator
branch
from
a2cdd07 to
a06b6b3
Compare
lann
approved these changes
May 14, 2024
|
Also fixes #2343 |
rylev
reviewed
May 14, 2024
crates/core/src/lib.rs
Outdated
| match std::env::var("SPIN_WASMTIME_POOLING") { | ||
| Ok(s) if s == "1" => return true, | ||
| Ok(s) if s == "0" => return false, | ||
| _ => {} |
There was a problem hiding this comment.
I think we should respect SPIN_WASMTIME_POOLING if it's set no matter what. If it's set to "0" or "false", we can turn it off, and otherwise enable it.
There was a problem hiding this comment.
Sure yeah, I opted to panic for non 0/1 values to avoid making this too too sprawling (as it's probably not going to be used that much anyway)
Signed-off-by: Alex Crichton <alex@alexcrichton.com>
rylev
approved these changes
May 14, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This commit is intended to address #2119 and mirror bytecodealliance/wasmtime#8610. The base problem is that some systems are configured with smaller amounts of virtual memory than other systems, for example some aarch64 and riscv64 systems are shown to have only 39 bits of virtual address space rather than the 48 by default on x86_64. This means that the pooling allocator can't be initialized on these platforms since it needs more virtual memory than that.
This changes Spin to dynamically choosing whether to use the pooling allocator. It's still used by default in Wasmtime but a dynamic probe is performed to determine whether it's going to work first. While here I also added an env var to control this behavior for an escape hatch if that's needed in the future too.
Closes #2119
Closes #2343