Repository navigation
AML fixes found running laptops' battery methods - #377
Merged
Merged
Conversation
`If (Local0)` failed with ObjectNotOfExpectedType when the local held a reference, such as a field's value: the predicate was read as an integer without going through the reference. `While` already unwrapped it. Found on the Framework Laptop 13 (AMD)'s battery methods.
Namespace::search let the internal LevelDoesNotExist out for a path like \A.B.C with no B, where a missing last segment gives ObjectDoesNotExist. Callers that check for a missing object didn't expect it: CondRefOf (\A.B.C) was an error that stopped loading the whole table instead of false, and evaluate_if_present returned an error instead of None. ACPICA returns AE_NOT_FOUND for both. Found on the Framework Laptop 13 (AMD), whose SSDTs test for optional devices that way (\_SB.PCI0.GPPC.XHC0.RHUB.GHBX).
The resulting string kept the null byte that ends the buffer's string: `split_inclusive` includes the separator. Found on the Framework Laptop 13 (AMD)'s battery methods.
…eplaces the element `Index (PKG, n)` returned a RefOf reference, so storing to it converted the value to the element's current type: a string stored over an integer element became 0. When the Index reference is the target of the store, and the element isn't itself a reference, the element is now replaced, as do_store's documentation says and as uACPI does (store_to_reference). Through a Local that holds it, the store still goes to the element with an implicit cast. CopyObject to an Index reference was a todo!(). Found on the Framework Laptop 13 (AMD), whose battery methods store strings into a package's integer elements.
martin-hughes
requested changes
Nov 9, 2026
martin-hughes
left a comment
martin-hughes
left a comment
Contributor
There was a problem hiding this comment.
@ogoffart: Great work! Thanks for taking the time to fix all these.
"Request changes" seems a bit mean given that this is a great PR, but please would you add one extra test per my comment?
Happy enough to take this all in one PR since they're all straightforward fixes. Thanks for splitting each fix to one commit, it was very easy to check each one individually.
Some random comments:
- I also have a Dell XPS 13 7390 laptop 🥳 (it travels around the world when I'm away from home! I've never tested my own code on it though 😆)
- You've touched on a few areas that Isaac and I know need a bit more work (CopyObject, Index references, Notify) but from what I see your work is all a nice improvement.
…ct it names `DerefOf (Index (PKG, n))` with a string element looked the string up as a name in the namespace. Only a string given to DerefOf directly is a path; through a reference, it's the value. Found on the Framework Laptop 13 (AMD)'s battery methods. Fixes rust-osdev#372.
read_buffer_field compared the field's length, in bits, with the integer size, in bytes: only fields of up to 8 bits became integers, and a CreateDWordField read gave a 4-byte buffer. Compared with an integer, the integer is then converted to an 8-byte buffer, and they differ. Found on a Dell XPS 13 7390, whose battery's _STA compares such a field with an integer.
CreateDWordField (Arg0, Arg1, FLD) failed: the index is a reference to the argument, not an integer. CreateField's bit index and size had the same problem. Found on a Dell XPS 13 7390: every call to its firmware's SMI methods failed.
Notify was unimplemented, and its arguments didn't even match the types they're resolved to (an object and an integer, not a name string), so any method notifying failed. With no way to tell the operating system yet, it's logged and ignored, as ACPICA does without a notify handler. That's safe: Notify only queues an event for the operating system's handlers, which run later; it returns nothing and changes nothing in the namespace, so the method computes the same. Found on a Dell XPS 13 7390, whose embedded controller's _REG notifies the lid.
ogoffart
force-pushed
the
olivier/aml-fixes
branch
from
November 10, 2026 07:40
35e66df to
849c2a7
Compare
martin-hughes
approved these changes
Nov 10, 2026
martin-hughes
left a comment
martin-hughes
left a comment
Contributor
There was a problem hiding this comment.
Great, thank you again @ogoffart.
Really good to see how the crate is being used in the wild.
This file contains hidden or 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
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.
I discovered more bug when trying read the battery status on more laptops.
I made one PR because it is easier for me, but i can also split in several PR if you prefer.
See the individual commit messages for details