Skip to content

fix: clear stray SAL annotation _Must_inspect_result_ from req.irql on 14 IddCx functions#1662

Open
cristeigabriela wants to merge 1 commit into
MicrosoftDocs:stagingfrom
cristeigabriela:fix/iddcx-irql-empty
Open

fix: clear stray SAL annotation _Must_inspect_result_ from req.irql on 14 IddCx functions#1662
cristeigabriela wants to merge 1 commit into
MicrosoftDocs:stagingfrom
cristeigabriela:fix/iddcx-irql-empty

Conversation

@cristeigabriela
Copy link
Copy Markdown
Contributor

@cristeigabriela cristeigabriela commented May 11, 2026

Context: This is part of a series of pull requests.

I am working on a sister project to sparse, the sdk-api parser, and in the process, I automated finding issues in windows-driver-docs-ddi (I've also done this for sdk-api.)

I'm now submitting patches for everything I've found for review. All the changes were tested in a branch where all of them are applied, and every single YAML frontmatter parses correctly.

All 12 PRs in this series — merging guide
# PR Branch Files
1 #1660 fix/typos-frontmatter 5
2 #1661 fix/acxmisc-acxobjectbagretrieveblob-irql 1
3 #1662 fix/iddcx-irql-empty 14
4 #1663 fix/silo-and-reparse-irql 8
5 #1664 fix/portcls-na-dll 2
6 #1665 fix/irql-apc-level-spacing 41
7 #1666 fix/irql-passive-abbreviation 2
8 #1667 fix/irql-dispatch-level-spacing 301
9 #1668 fix/ntoskrnl-lib-misfiled-as-dll 3
10 #1669 fix/strip-irql-prose-prefix 80
11 #1670 fix/irql-trailing-period-and-passive-case 133
12 #1671 fix/irql-prose-to-operator 239

All of the 12 PRs have zero pairwise file overlap. They can be merged individually, in any order, and there will not be merge conflicts between them.

After every PR is merged, all 10,915 nf-*.md frontmatter blocks will still parse cleanly as YAML.

A reference branch with all of the PRs merged at once is at integration/all-open-prs.

These functions have no IRQL restraints in the respective IddCx.h headers, and so I've removed the IRQL specification from here as well.

I assume the current values are a parser error (I can imagine how, in Clang I would imagine the way this would be implemented is looking at the AST children of the function and expecting a specific order. Or maybe it's something simpler.)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant