LLVM Bugzilla is read-only and represents the historical archive of all LLVM issues filled before November 26, 2021. Use github to submit LLVM bugs

Bug 21857 - [fuzz] Assertion `!R.empty() && "lookup for a constructor or assignment operator was empty"' failed.
Summary: [fuzz] Assertion `!R.empty() && "lookup for a constructor or assignment opera...
Status: REOPENED
Alias: None
Product: clang
Classification: Unclassified
Component: -New Bugs (show other bugs)
Version: trunk
Hardware: PC Linux
: P normal
Assignee: Unassigned Clang Bugs
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-12-10 16:01 PST by Sami Liedes
Modified: 2015-02-10 19:53 PST (History)
3 users (show)

See Also:
Fixed By Commit(s):

Attachments
test case (70 bytes, text/x-c++src)
2014-12-10 16:01 PST, Sami Liedes 		Details
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Sami Liedes 2014-12-10 16:01:34 PST 
Created attachment 13498 [details]
test case

Fuzzing discovered that the attached test case, when given as input to

  clang -fno-crash-diagnostics -std=c++11 -xc++ -c -emit-llvm

causes this assertion failure:

clang-3.6: tools/clang/lib/Sema/SemaLookup.cpp:2504: Sema::SpecialMemberOverloadResult *clang::Sema::LookupSpecialMember(clang::CXXRecordDecl *, clang::Sema::CXXSpecialMember, bool, bool, bool, bool, bool): Assertion `!R.empty() && "lookup for a constructor or assignment operator was empty"' failed.
Comment 1 Eric Niebler 2015-02-01 17:08:43 PST 
This assertion is triggered by the following well-formed (AFAIK) code (tested on trunk):

template<typename Fn>
struct fun : Fn
{
    fun() = default;
    fun(Fn fn) : Fn(fn) {}
    using Fn::operator();
};

template<typename Fn>
fun<Fn> wrap(Fn fn)
{
    return fun<Fn>(fn);
}

int main()
{
    wrap([](){});
}

Bit of a PITA. I haven't found a work-around.
Comment 2 Richard Smith 2015-02-10 19:48:56 PST 
Fixed in r228785.
Comment 3 Richard Smith 2015-02-10 19:53:03 PST 
Eric's rejects-valid with lambdas is fixed, the testcase produced by the fuzzer still fails, and looks like an unrelated error-recovery issue.