Steven Malis
Steven Malis
Would it be possible to move the serde dependency to be behind a feature flag, instead of being always required?
join_absolute_paths is not always right on Windows when dealing with just drive letter paths ("C:")
### Summary When dealing with a Path that consists solely of a drive letter and a colon (i.e. "C:"), joining it with a separator actually results in a meaningfully different...
It would be handy to have a FromBytes::read_from_(prefix/suffix)_split that returns a tuple of the T and the unused portion of the byte slice. Ref::new_from_prefix isn't quite the same in case...
I stumbled on https://github.com/davidlattimore/duplicate-function-checker today and ran it over a relatively large work project that contains heavy use of async-task. Despite building our binary with fat LTO and 1 codegen-unit...
As the underlying code is performing fetch_(or/and)s this is perfectly valid to do. These could be moved into a new replace_aliased instead if desired, I have no real preference, this...
Supersedes https://github.com/microsoft/openvmm/pull/111. Since WHP is the only case in which we can hit this path, make it a bit clearer with a new per-backing method. Open for bikeshedding of course
Given that most of our backings halt in kernel mode instead of user mode, the previous doc comment was rather misleading.
Pretty small this time around.
Example run https://github.com/microsoft/openvmm/actions/runs/11337175381/job/31528850443?pr=81 ``` 2024-10-15T01:11:59.758365Z INFO linux_log: [ 32.716659] usbcore: registered new interface driver usblp 2024-10-15T01:11:59.882830Z INFO linux_log: [ 32.776050] scsi 0:0:0:0: Direct-Access Msft Virtual Disk 1.0 PQ: 0 ANSI:...
The assumption that it is safe to perform wider-than-size non-atomic reads and writes on a &[AtomicFoo] is rather deeply baked into our codebase. However, under the current Rust safety rules...