lumiera_/tests/basics/diagnostic-context-test.cpp
Ichthyostega a90b9e5f16 Library: uniform definition scheme for error-IDs
In the Lumiera code base, we use C-String constants as unique error-IDs.
Basically this allows to create new unique error IDs anywhere in the code.

However, definition of such IDs in arbitrary namespaces tends to create
slight confusion and ambiguities, while maintaining the proper use statements
requires some manual work.

Thus I introduce a new **standard scheme**
 * Error-IDs for widespread use shall be defined _exclusively_ into `namespace lumiera::error`
 * The shorthand-Macro `LERR_()` can now be used to simplify inclusion and referral
 * (for local or single-usage errors, a local or even hidden definition is OK)
2024-03-21 19:57:34 +01:00

233 lines
6.8 KiB
C++

/*
DiagnosticContext(Test) - verify thread local stack for collecting diagnostics
Copyright (C) Lumiera.org
2011, Hermann Vosseler <Ichthyostega@web.de>
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License as
published by the Free Software Foundation; either version 2 of
the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
* *****************************************************/
/** @file diagnostic-context-test.cpp
** unit test \ref DiagnosticContext_test
*/
#include "lib/test/run.hpp"
#include "lib/test/test-helper.hpp"
#include "lib/diagnostic-context.hpp"
#include "lib/iter-explorer.hpp"
#include "lib/thread.hpp"
#include <vector>
#include <chrono>
#include <array>
using std::this_thread::sleep_for;
using std::chrono_literals::operator ""us;
namespace lib {
namespace test{
namespace error = lumiera::error;
namespace { // private test setup...
/* WARNING: memory hungry */
const uint NUM_THREADS = 75;
const uint MAX_RAND = 100*1000;
auto isOdd = [](auto val) { return bool (val % 2); };
} // (End) test setup....
using lib::ThreadJoinable;
using LERR_(LOGIC);
using std::rand;
/**
* Subject of this test:
* a thread-local stack of int values
*/
typedef DiagnosticContext<uint> Marker;
typedef std::vector<uint> VecI;
/******************************************************************************//**
* @test verify a diagnostic facility to collect and access contextual information.
* DiagnosticContext frames are placed into automatic storage (as local
* variable within some function scope). Based on thread-local storage,
* the next reachable frame can be accessed from anywhere within
* the callstack. This feature is useful for collecting
* information regarding features cross-cutting
* the usual dependency hierarchy.
* @see lib::DiagnosticContext
* @see lib::ThreadJoinable
*/
class DiagnosticContext_test : public Test
{
virtual void
run (Arg)
{
verify_simpleAccess();
verify_heavilyParallelUsage();
}
/** @test create nested scopes and place a
* DiagnosticContext frame into each.
* Access the next reachable frame.
*/
void
verify_simpleAccess()
{
VERIFY_ERROR (LOGIC, Marker::access());
VecI loggedValues;
Marker zero(0);
CHECK (0 == zero);
CHECK (0 == Marker::access());
{ // nested scope
CHECK (0 == Marker::access());
Marker one(1);
CHECK (1 == Marker::access());
CHECK (1 == one);
CHECK (0 == zero);
{ // nested scope
CHECK (1 == Marker::access());
Marker two(2);
CHECK (2 == Marker::access());
CHECK (2 == two);
CHECK (1 == one);
CHECK (0 == zero);
loggedValues = Marker::extractStack();
}
CHECK (1 == Marker::access());
}
CHECK (0 == Marker::access());
CHECK (3 == loggedValues.size());
CHECK (2 == loggedValues[0]);
CHECK (1 == loggedValues[1]);
CHECK (0 == loggedValues[2]);
}
/** @test verify the thread local property of ContextStack.
* Create several threads, each of which processes a sequence
* of numbers starting from a random initial value down to one.
* Whenever a simple division by two leads to an odd value, this
* value is placed onto the ContextStack. At recursion end, we
* take a snapshot of the full ContextStack and then unwind.
* Thus the captured numbers must from a decreasing sequence
* of odd values.
* @warning this test case seems to cause memory pressure.
* When running the test suite with VSize limit 500MB,
* we frequently got aborts even with 40 threads.
* This is surprising, since all of the lists
* generated in the individual threads are
* of size below 20 elements.
*/
void
verify_heavilyParallelUsage()
{
auto verifyResult = [](VecI sequence)
{
uint prev = 0;
for (uint val : sequence)
{
CHECK (isOdd(val) and val > prev);
prev = val;
}
};
std::array<TestThread, NUM_THREADS> testcases;
auto results = lib::explore(testcases)
.transform([](TestThread& t){ return t.join(); })
.effuse();
for (auto& res : results)
verifyResult (res);
}
/** build a call stack within separate thread and capture diagnostics */
struct TestThread
: ThreadJoinable<VecI>
{
TestThread()
: ThreadJoinable("test context stack"
,&verifyDiagnosticStack)
{ }
};
/** the actual test operation running in a separate thread
* produces a descending number sequence, and only odd values
* will be captured into the diagnostic stack
*/
static VecI
verifyDiagnosticStack()
{
uint seed (1 + rand() % MAX_RAND);
return descend (seed);
}
static VecI
descend (uint current)
{
if (current < 2)
return Marker::extractStack();
sleep_for (500us);
if (isOdd(current))
{
Marker remember(current);
return descend (current+1);
}
else
return descend (current/2);
}
};
/** Register this test class... */
LAUNCHER (DiagnosticContext_test, "function common");
}} // namespace vault::test