-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Test updated release/1.10-ci with prefast fixes #5308
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Test updated release/1.10-ci with prefast fixes #5308
Conversation
|
Replaces #5281 |
|
@dotnet-bot test ci please |
|
@dotnet-bot test this please |
|
@dotnet-bot |
|
@dotnet-bot |
We had signed/unsigned mixups between the format string and the args.
Note that this includes functional changes to one implementation, as it could in some cases only set the getter or setter and leave the other uninitialized. There's also a couple other function annotation alignment fixes in this commit.
Unfortunately there's a good number of holes, which seem to be due to how the PREfast analysis of the locking works. Someone familiar with the mechanics may be able to do this with fewer suppressions.
A pre-defined function is added to the chakra etw header that does not initialize a buffer before passing it around. It looks ok, but the warning is in a file that we have limited control over. Here I disabled the warning for the whole included file, which seems like the closest we can scope it.
Due to template expansion this check can be redundant, so prefast warns on it.
Due to tightly scoped includes this file wasn't getting the warning disables that are used in the rest of the code base to handle a few situations that we accept (e.g. constant boolean logic, likely done due to flag handling).
PREfast checks for this now and decides that the pointer may be null, so the additional if statement protects against that.
PREfast complains if you add bools together; inserting the ternary expressions around them resolves this.
Unfortunately, this needs the suppression despire the assert. Here we copy things into a buffer, and PREfast thinks that we may write beyond the buffer bounds, despite the analysis asserts.
f474361 to
da901d2
Compare
|
@dotnet-bot |
|
@Penguinwizzard it looks like there still some issues with prefast, as well as Win7/Dev14 and Win10/Dev15 build failures. |
|
@dotnet-bot test ci please |
|
@dotnet-bot |
See #5307 for CI update and #5272 by @Penguinwizzard for fixes.