[BUG] device memory leak in cuBLAS (matmul)

Each launch of a new threat leaves around 10KB of device memory allocated.

Description
===========
Launching a function with a matmul operation in a loop of consecutive threads on CUDA will result in memory overflow on the device, and exception: CUBLAS Error (3): CUBLAS_STATUS_ALLOC_FAILED. Running the same function in a loop on the main thread operates normally, even after 10,000 loops.
On OpenCL, the thread version as the main thread version operate normally.
Full trace logging, indicates the existing device buffers are reused as expected (also in the separate threads).

Arrayfire build: master 3.9.0 b05da694
Back-end: CUDA
Workaround available: No
Reproducibility: Yes
Logging: [Logging.txt](https://github.com/arrayfire/arrayfire/files/9490460/Logging.txt)
Device memory in main thread: 

Device memory in consecutive threads: 

Reproducible Code and/or Steps
------------------------------
```
int main() {
class trainer {
int device;

public:
trainer(const int device) : device(device){};
void train() {
af::setDevice(device);
const af::array a{af::iota(af::dim4(10, 10))};
// size has no impact
af::array c{af::matmul(a, a)};
};
};

try {
af::info();
trainer trainers{af::getDevice()};

for (int i{0}; i < 1000; ++i) {
std::cout << i << ", ";
// OK
trainers.train();

// trows exception
// std::thread t(&trainer::train, std::ref(trainers));
// t.join();
}
} catch (af::exception &ae) { std::cerr << ae.what() << std::endl; }
return 0;
}
```

System Information
------------------
1. ArrayFire version : master 3.9.0 b05da694
2. Devices installed on the system: GTX 750 Ti
3. (optional) Output from the af::info() function if applicable: see logging
4. Output from the following scripts:
[Output cmds.txt](https://github.com/arrayfire/arrayfire/files/9490479/Output.cmds.txt)

Checklist
---------

- [x] Using the latest available ArrayFire release
- [x] GPU drivers are up to date

Start by running the provided C++ reproducer with the CUDA backend, comparing af::matmul in the main thread with consecutive std::thread launches, and review Logging.txt. Trace device allocation and cleanup around af::setDevice and matmul; done means repeated threaded launches no longer consume device memory or end with CUBLAS_STATUS_ALLOC_FAILED.

由索引模型根据 Issue 内容生成。