https://llvm.org/bugs/show_bug.cgi?id=28285
Bug ID: 28285
Summary: std::async with default policy moves their arguments
twice if std::thread cannot be created
Product: libc++
Version: unspecified
Hardware: PC
OS: Linux
Status: NEW
Severity: normal
Priority: P
Component: All Bugs
Assignee: unassignedclangb...@nondot.org
Reporter: kariya_mits...@hotmail.com
CC: llvm-bugs@lists.llvm.org, mclow.li...@gmail.com
Classification: Unclassified
In the libc++'s current implementation of std::async with default policy (i.e.
async | deferred),
first, the argument function is launched in the context of a new thread
asynchronously (async).
If it was failed, the function is launched in the context of the current thread
later (deffered).
But if the first launch was failed (ex. std::thread creation trows an
exception), std::async's
arguments was already moved.
So, the arguments have no valid values when second launch is executing.
The sample code below simulates std::thread creation failure.
============================== sample code ==============================
#include <iostream>
#include <future>
#include <string>
#include <system_error>
struct S {
S() : b(false) {}
// to simulate thread creation failure, it throws
resource_unavailable_try_again when 'o' is first move target
S(S&& o) : b(true) {
if (!o.b) {
o.b = true;
throw
std::system_error(std::make_error_code(std::errc::resource_unavailable_try_again));
}
}
S(const S&) = delete;
S& operator=(const S&) = delete;
S&& operator=(S&&) = delete;
bool b;
};
int main()
{
auto f = std::async([](std::string&& s, S&&){ std::cout << "s = '" << s <<
'\'' << std::endl; }, std::string{"str"}, S{});
f.get();
}
============================== sample code ==============================
============================== output ==============================
s = ''
============================== output ==============================
cf. http://melpon.org/wandbox/permlink/IZDRW8D7i0wifN91
I think that it should output "s = 'str'" or should output nothing,
whether exception is thrown or not.
--
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
