https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88942
Bug ID: 88942
Summary: [openacc, testsuite] lib-82.c does not wait for all
streams before calling acc_shutdown
Product: gcc
Version: 9.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: testsuite
Assignee: unassigned at gcc dot gnu.org
Reporter: vries at gcc dot gnu.org
Target Milestone: ---
[ Spin-off bug from PR88941 - "[nvptx, openacc, libgomp] Assertion
`!s->map->active' failed for empty asynchronous parallel" ]
In lib-82.c, we have:
...
acc_wait_all_async (0);
...
acc_shutdown (acc_device_nvidia);
...
At the acc_shutdown documentation, we read:
..
- This routine may not be called during execution of an accelerator compute
region.
- If the program attempts to execute a compute region or access any device
data on such a device, the behavior is undefined.
...
The lib-82.c testcase launches kernels asynchronously (not by using parallel
async, but by using cuLaunchKernel).
The documentation of acc_shutdown seems to imply that we need to wait for all
those streams to finish before calling acc_shutdown.
There is a wait call before acc_shutdown:
...
acc_wait_all_async (0);
...
but the semantics for that one is:
...
The acc_wait_all_async routine enqueues wait operations on one async queue
for the operations previously enqueued on all other async queues.
...
ISTM that this can't guarantee that all queues have finished.
OTOH, using acc_wait_all, with semantics:
...
The acc_wait_all routine waits for completion of all asynchronous operations.
...
seems to guarantee that.
We should fix the test-case.
[ It would be nice have a warning or error for this, that is, when calling
shutdown, to report on unfinished stream. ]
