[analyzer] Fix a regression (from r 165079): compare canonical types.

Suppresses a leak false positive (radar://12663777).

In addition, we'll need to rewrite the adjustReturnValue() method not to
return UnknownVal by default, but rather assert in cases we cannot
handle. To make it possible, we need to correctly handle some of the
edge cases we already know about.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@167762 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Anna Zaks 2012-11-12 22:06:24 +00:00
Родитель 5ece32e335
Коммит e7ad14e182
2 изменённых файлов: 18 добавлений и 0 удалений

Просмотреть файл

@ -137,6 +137,8 @@ static SVal adjustReturnValue(SVal V, QualType ExpectedTy, QualType ActualTy,
return V;
// If the types already match, don't do any unnecessary work.
ExpectedTy = ExpectedTy.getCanonicalType();
ActualTy = ActualTy.getCanonicalType();
if (ExpectedTy == ActualTy)
return V;

Просмотреть файл

@ -343,5 +343,21 @@ void test_test_return_inline_2(char *bytes) {
CFRelease(str);
}
extern CFStringRef getString(void);
CFStringRef testCovariantReturnType(void) __attribute__((cf_returns_retained));
void usetestCovariantReturnType() {
CFStringRef S = ((void*)0);
S = testCovariantReturnType();
if (S)
CFRelease(S);
}
CFStringRef testCovariantReturnType() {
CFStringRef Str = ((void*)0);
Str = getString();
if (Str) {
CFRetain(Str);
}
return Str;
}