alexcrichton opened PR #8823 from alexcrichton:match-linux-for-fd-pwrite
to bytecodealliance:main
:
This commit updates the implementation of
fd_pwrite
in WASI to match Linux semantics for an under-specified corner of WASI. Specifically iffd_pwrite
is used the offset specified is ignored if the file is opened in append mode and the bytes are instead appended.This commit additionally refactors
fd_write
andfd_pwrite
to have basically the same code with only a minor branch internally when the final write is being performed to help deduplicate more logic.Closes #8817
<!--
Please make sure you include the following information:
If this work has been discussed elsewhere, please include a link to that
conversation. If it was discussed in an issue, just mention "issue #...".Explain why this change is needed. If the details are in an issue already,
this can be brief.Our development process is documented in the Wasmtime book:
https://docs.wasmtime.dev/contributing-development-process.htmlPlease ensure all communication follows the code of conduct:
https://github.com/bytecodealliance/wasmtime/blob/main/CODE_OF_CONDUCT.md
-->
alexcrichton requested wasmtime-core-reviewers for a review on PR #8823.
alexcrichton requested fitzgen for a review on PR #8823.
fitzgen requested pchickey for a review on PR #8823.
fitzgen commented on PR #8823:
Redirecting to @pchickey
alexcrichton updated PR #8823.
alexcrichton has enabled auto merge for PR #8823.
alexcrichton has disabled auto merge for PR #8823.
pchickey submitted PR review.
alexcrichton has enabled auto merge for PR #8823.
alexcrichton updated PR #8823.
alexcrichton updated PR #8823.
alexcrichton commented on PR #8823:
CI turned up an interesting phenomena which makes sense in retrospect. With wasi-common the behavior is subject to whatever the platform OS provides which makes Linux the odd-one-out. Papering over OS differences is easy enough in
wasmtime-wasi
but is not trivial inwasi-common
.@pchickey how do you feel about sticking with these semantics and just ignoring the whole test in wasi-common? Alternatives could include not landing this in Wasmtime and taking it up in an issue with WASI and/or adding more stuff to wasi-common.
pchickey commented on PR #8823:
I think it is fine to just leave this bug in wasi-common and ignore the test - I don't think there is much value in bringing bug fixes for corner cases like this, if users need the fix the real question is why they havent upgraded to wasmtime-wasi.
alexcrichton merged PR #8823.
Last updated: Nov 22 2024 at 17:03 UTC