-
Notifications
You must be signed in to change notification settings - Fork 1.5k
feat(unused_var): analyze global variables inside function body #2944
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
Conversation
|
Thanks! I will look this evening.. |
danmar
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
small nits.. I will test it out a bit..
danmar
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
hmm I have tried it out and it works better than I thought!
| // check if pointer to global variable assigned to another variable (another_var = &global_var) | ||
| if (Token::simpleMatch(bodyToken->tokAt(-1), "&") && Token::simpleMatch(bodyToken->tokAt(-2), "=")) { | ||
| const Token* assigned_var_token = bodyToken->tokAt(-3); | ||
| if (assigned_var_token && assigned_var_token->variable()) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
if this condition is true then you don't care that the address is used somehow.. would probably be safer to return false then.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
As I see, isVariableChanged() works fine enough with "pointer-to-global" variables. I prefer to leave it as is, if you not insist.
Do you see any test cases which will force this code work not as expected?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think it works well.. right now I am only worried about the todo test case. Do you know it it's complex to fix that?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Probably, it should be not so complex. But it can affect existing tests and I will need some time to figure out with them.
Also I can not estimate a free time amount I will be able to put on this work in the nearest time. =)
How about this one - I will return false for now and update astutils.cpp and this piece of code in next PR?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It's fine to make a partial fix.. and finish later. but then I feel more comfortable if there are no false positives in the meantime. So well.. could you return false if there is a global array/pointer variable.
lib/checkunusedvar.cpp
Outdated
| } | ||
| // check if global variable is changed | ||
| if (bodyVariable->isGlobal() || | ||
| (std::find(pointersToGlobals.begin(), pointersToGlobals.end(), bodyVariable) != pointersToGlobals.end())) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Now we can use pointersToGlobals.find(bodyVariable) != pointersToGlobals.end() instead here.
|
Thanks! |
Welcome. |
Before this PR, any global variable found in function body marked this function as side-effect one. Analyzing global variables more precisely from now.
Any help with missed test cases will be highly appreciated.