In this exercise you will implement and benchmark different types of memory management in a threaded parallel application.
Create an application membench T N S which launches a configurable (program launch parameter T) number of threads.
Each thread should perform a given number of memory allocations and deallocations (N) of a given size in bytes (S).
For example
./membench 8 10000 1024
would launch 8 threads, each of which performs 10000 allocations of 1 kiB each.
Implement a naive best fit memory allocator (see lecture slides) and use it in your application.
The allocator should provide its own versions of the malloc and free functions.
In this task, use a single global allocator (which of course needs to be locked when used) that is shared by all threads.
Compare the performance of your allocator to the system-provided malloc and free function using "membench", with a few different numbers of threads and allocation sizes, and a number of allocations/deallocations that gives a reasonable total runtime (more than a second and less than a minute).
For this task, instead of a single global allocator, create a separate local allocator for each thread.
For creating local instances of the allocator, you can use thread-local storage in GCC using __thread (see e.g. http://gcc.gnu.org/onlinedocs/gcc-7.3.0/gcc/Thread-Local.html for further information)
Compare the performance of Task 2 and 3 in "membench", with a few different numbers of threads and allocation sizes, and a number of allocations/deallocations that gives a reasonable total runtime (more than a second and less than a minute).