Memory leaks?

[isuruf]

Running an FMM with,

for _ in range(2):
    timing_data = {}
    drive_fmm(sumpy_wrangler, (charge_device,), timing_data=timing_data)
    queue.finish()
    gc.collect()

leads to a memory allocation failure with cuda in the second loop.

[inducer]

When you say "memory allocation", what precisely do you mean? How do you measure that?

[inducer]

Possibly related: inducer/pymbolic#125 (comment)

[inducer]

Running

 pycl test_fmm.py 'test_sumpy_fmm(_acf, LaplaceKernel(3), LinearPDEConformingVolumeTaylorLocalExpansion, LinearPDEConformingVolumeTaylorMultipoleExpansion, False, False, False, None)'

with this patch applied leads to no appreciable memory growth. 🤔

Could you provide a reproducer?

[inducer]

Similarly, no appreciable memory growth when run using mprof: