4 ms·

I have my own toy benchmark too :-) https://hoult.org/primes.txt https://hoult.org/primes.txt

by brucehoult 1y ago

kragen 1y ago
That seems like a significantly better microbenchmark than dumbfib!
- brucehoult 1y ago
  The main weakness is -- opposite of dumbfib -- it has no function calls at all! But I find it correlates well with other CPU core & L1 cache/SRAM benchmarks such as Dhrystone and Coremark. The Embench guys added a cut down version -- smaller arrays, shorter execution time [1] to their benchmark suite a couple of year back. [1] baseline ~4s on typical microcontrollers ... I'd started at ~5s on Ivy Bridge, 10-30s on the Arm boards I had back then
  - kragen 1y ago
    It does have loops, integer multiplication, and array indexing (mostly reading), and all of those are pretty important. Subroutine calls are particularly difficult to benchmark because they are so critical to real-world performance and so easy for compilers to remove from a benchmark if they can guess that a particular callsite will be hot. (LuaJIT takes this to the extreme.) You'd think non-tail recursion would stymie this, but I've seen GCC unroll a recursive loop several layers deep and constant-fold away several layers over the base case. Impressive, useful, and completely invalidated my dumb benchmark. I would guess that most programs have a subroutine call on the order of once every 100 instructions, spending something like 15% of their time on subroutine call and return. This is a very difficult number to quantify because of the pervasive indirect costs of subroutine calls: less efficient register allocation, forgone opportunities for constant folding and specialization, objects stored in memory rather than in registers, allocation on the heap rather than a stack, etc.
    - kragen 1y ago
      I wasn't able to find good information about how many instructions per subroutine call computers typically run these days. Probably I'm not using the right search terms, but here's what I found. I just ran Emacs a few times under valgrind --tool=callgrind. Perhaps surprisingly, it was usably fast once it started up. I analyzed callgrind's output file with the following highly sophisticated methodology to determine the total number of subroutine calls (and presumably returns): print(sum(int(mo.group(1)) for line in sys.stdin for mo in [re.match(r'^calls=(\d+)', line)] if mo)) This, plus some division by numbers of executed instructions reported, revealed that it had a subroutine call for about every 55–59 instructions, and about a hundred and fifty thousand extra subroutine calls (and 7.7 million extra instructions) per character typed. This suggests that, at least in Emacs, the cost of subroutine call and return is probably higher than 15%, maybe 25%. Other programs I tested: - GNU ls: 79 instructions per call - PARI/GP: 360 instructions per call - CPython with Numpy: 88 instructions per call - Minetest (rather unplayable): 88 instructions per call
      - brucehoult 1y ago
        Interesting results. Are you able to figure out how many of those are calls of leaf functions? On modern ISAs and ABIs calling a small leaf function is very cheap -- just the `jal` and `ret`, plus getting the arguments into the right registers first. In many cases memory is not touched at all (unless the code in the function necessarily touches memory), no stack frame is created etc.
        kragen 1y ago
        Hmm, I think that ought to be pretty doable with the callgrind output actually. I'm using amd64 rather than a modern ISA, so the calls do necessarily touch "memory", but leaf functions are still especially cheap—they can use temporary registers for local variables. In the case of Emacs most of the functions are compiled Lisp, and I doubt the Lisp compiler is smart enough to exploit leafness in its register allocator.
        kragen 1y ago

I have my own toy benchmark too :-) https://hoult.org/primes.txt https://hoult.org/primes.txt

kragen 1y ago

That seems like a significantly better microbenchmark than dumbfib!

brucehoult 1y ago
The main weakness is -- opposite of dumbfib -- it has no function calls at all! But I find it correlates well with other CPU core & L1 cache/SRAM benchmarks such as Dhrystone and Coremark. The Embench guys added a cut down version -- smaller arrays, shorter execution time [1] to their benchmark suite a couple of year back. [1] baseline ~4s on typical microcontrollers ... I'd started at ~5s on Ivy Bridge, 10-30s on the Arm boards I had back then
- kragen 1y ago
  It does have loops, integer multiplication, and array indexing (mostly reading), and all of those are pretty important. Subroutine calls are particularly difficult to benchmark because they are so critical to real-world performance and so easy for compilers to remove from a benchmark if they can guess that a particular callsite will be hot. (LuaJIT takes this to the extreme.) You'd think non-tail recursion would stymie this, but I've seen GCC unroll a recursive loop several layers deep and constant-fold away several layers over the base case. Impressive, useful, and completely invalidated my dumb benchmark. I would guess that most programs have a subroutine call on the order of once every 100 instructions, spending something like 15% of their time on subroutine call and return. This is a very difficult number to quantify because of the pervasive indirect costs of subroutine calls: less efficient register allocation, forgone opportunities for constant folding and specialization, objects stored in memory rather than in registers, allocation on the heap rather than a stack, etc.
  - kragen 1y ago
    I wasn't able to find good information about how many instructions per subroutine call computers typically run these days. Probably I'm not using the right search terms, but here's what I found. I just ran Emacs a few times under valgrind --tool=callgrind. Perhaps surprisingly, it was usably fast once it started up. I analyzed callgrind's output file with the following highly sophisticated methodology to determine the total number of subroutine calls (and presumably returns): print(sum(int(mo.group(1)) for line in sys.stdin for mo in [re.match(r'^calls=(\d+)', line)] if mo)) This, plus some division by numbers of executed instructions reported, revealed that it had a subroutine call for about every 55–59 instructions, and about a hundred and fifty thousand extra subroutine calls (and 7.7 million extra instructions) per character typed. This suggests that, at least in Emacs, the cost of subroutine call and return is probably higher than 15%, maybe 25%. Other programs I tested: - GNU ls: 79 instructions per call - PARI/GP: 360 instructions per call - CPython with Numpy: 88 instructions per call - Minetest (rather unplayable): 88 instructions per call
    - brucehoult 1y ago
      Interesting results. Are you able to figure out how many of those are calls of leaf functions? On modern ISAs and ABIs calling a small leaf function is very cheap -- just the `jal` and `ret`, plus getting the arguments into the right registers first. In many cases memory is not touched at all (unless the code in the function necessarily touches memory), no stack frame is created etc.
      - kragen 1y ago
        Hmm, I think that ought to be pretty doable with the callgrind output actually. I'm using amd64 rather than a modern ISA, so the calls do necessarily touch "memory", but leaf functions are still especially cheap—they can use temporary registers for local variables. In the case of Emacs most of the functions are compiled Lisp, and I doubt the Lisp compiler is smart enough to exploit leafness in its register allocator.
      - kragen 1y ago