Reputation: 14928
I found that --suppress=unmatchedSuppression
only suppresses unmatched suppression types in cppcheck options, but NOT unmatched inline suppressions.
Is this the expected behavior?
test.c
Line 4 is wrong. It should be warned arrayIndexOutOfBounds
Line 7 is okay. It should NOT be warned arrayIndexOutOfBounds
I have inline cppcheck-suppress
for both lines.
1 void f() {
2 char arr[5];
3 // cppcheck-suppress arrayIndexOutOfBounds
4 arr[10] = 0;
5
6 // cppcheck-suppress arrayIndexOutOfBounds
7 const char ok[] = "this line is ok";
8 }
Situation 1
Suppress cstyleCast
, which does NOT exist in code.
cppcheck --inline-suppr --force --enable=all
--xml-version=2 --suppress=cstyleCast test.c
2>cppcheckresults.xml
I get warned about (among other irrelevant warnings)
unmatchedSuppression: arrayIndexOutOfBounds
in test.c
line 7
(as expected)
unmatchedSuppression: cstyleCast
in *
line 0
(as expected)
Situation 2
Same as situation 1, but with additional --suppress=unmatchedSuppression
option
cppcheck --inline-suppr --force --enable=all
--xml-version=2 --suppress=cstyleCast --suppress=unmatchedSuppressiontest.c
2>cppcheckresults.xml
I expect both previous unmatchedSuppression
warnings to go away. But I still get
unmatchedSuppression
in test.c
line 7
(NOT expected)Upvotes: 11
Views: 12812
Reputation: 16242
The problem with the order solution is that you have to do this for every file that has the unmatched suppression.
A. In the command line use the option --suppress=unmatchedSuppression:{}
. (note the curly braces)
B. What I discovered is that you can also suppress the unmatched suppression itself in place, for example:
// cppcheck-suppress[assertWithSideEffect,unmatchedSuppression]
assert(...);
(tested with cppcheck 2.3 and 2.5)
This is necessary for example when running cppcheck 2.3 which didn't have the assertWithSideEffect
feature, later availabe in cppcheck 2.5.
Ideally, as you bump the cppcheck version you can start removing the unnecessary unmatchedSuppression
.
Needless to say, this all needs the command line --inline-suppr
.
Upvotes: 2
Reputation: 7459
What I recently found is that an unmatchedSuppression
warning from inline suppressions cannot be suppressed by a generic --suppress=unmatchedSuppression
or putting just that warning ID in a --suppressions-list
file. You have to qualify it with the filename. For example --suppress=unmatchedSuppression:test.c
.
Upvotes: 5