nomaddo

// Module with 1 kernels, global data with 0 words (64-bit each), starting at offset 1 words and 0 words of stack-frame
// Kernel 'backprop_floats' with 2305 instructions, offset 2, with following parameters: float learningRateMultiplier (4 B, 1 items), int batchSize (4 B, 1 items), __global const float* gradOutput (4 B, 1 items), __global const float* images (4 B, 1 items), __global out float* gradWeights (4 B, 1 items) (lSize, lids, gidX, offX)
// label: %start_of_function
or r1, unif, unif
or ra1, unif, unif
or r0, unif, unif
or r2, unif, unif
or ra14, unif, unif
or ra13, unif, unif
or ra12, unif, unif

nomaddo

// Module with 1 kernels, global data with 0 words (64-bit each), starting at offset 1 words and 0 words of stack-frame
// Kernel 'backprop_floats_withscratch_dobias_striped' with 2502 instructions, offset 2, with following parameters: float learningRateMultiplier (4 B, 1 items), int batchSize (4 B, 1 items), __global const float* gradOutput (4 B, 1 items), __global const float* images (4 B, 1 items), __global out float* gradWeights (4 B, 1 items), __local out float* _errorStripe (4 B, 1 items), __local out float* _imageStripe (4 B, 1 items) (lSize, lids, gidX)
// label: %start_of_function
or rb13, unif, unif
or rb21, unif, unif
or r2, unif, unif
or rb11, unif, unif
or rb12, unif, unif
or rb7, unif, unif
or rb8, unif, unif

nomaddo

/*
hoge.cpp: In function ‘int main()’:
hoge.cpp:35:37: warning: dynamic_cast of ‘Base b’ to ‘class A*’ can never succeed
     if (auto x = dynamic_cast<A*>(&b))
*/

#include <iostream>

class Base {
public:

nomaddo

// Module with 1 kernels, global data with 0 words (64-bit each), starting at offset 1 words and 0 words of stack-frame
// Kernel 'sum_f16', offset 2, with following parameters: __global float* a (4 B, 1 items), __global float* b (4 B, 1 items), __global float* c (4 B, 1 items)
or -, unif, unif
or r1, unif, unif
or ra4, unif, unif
or -, unif, unif
or -, unif, unif
or -, unif, unif
or r0, unif, unif
or -, unif, unif

nomaddo

# see https://github.com/nineties/py-videocore
#
# pi@nomaddo:~/py-videocore$ sudo python add.py 40960
# threads: 12
# numpy: 0.0107 sec, 46.0525 Mflops
# GPU: 0.0050 sec, 99.2960 Mflops
#
# pi@nomaddo:~/py-videocore$ sudo python add.py 20480
# threads: 12
# numpy: 0.0054 sec, 45.1429 Mflops

nomaddo

// Module with 1 kernels, global data with 0 words (64-bit each), starting at offset 1 words and 0 words of stack-frame
// Kernel 'add', offset 2, with following parameters: __global float* a (4 B, 1 items), __global float* b (4 B, 1 items)
or -, unif, unif
or r1, unif, unif
or ra3, unif, unif
or -, unif, unif
or -, unif, unif
or -, unif, unif
or r0, unif, unif
or -, unif, unif

nomaddo

        .global main
        .type   main, %function
main:
        push {r3, lr}
        bl   test
        mov  r0, #0
        pop  {r3, pc}

        .global test
        .type   test, %function

nomaddo

#include <stdlib.h>
#include <stdbool.h>
#include <stdio.h>

#define dfunc(f) ((void (void *)) f)            \

typedef struct set_t {
  int cap;
  int num;
  void ** t;

nomaddo

(* ocamlfind ocamlopt -package mparser -linkall -linkpkg mparser-test.ml *)
(* どんどんひどくなっていっている *)

(* ./a.out "if 0 then 1 else 2" の結果は…… StackOverFlow*)
open MParser

let decimal =
  many_chars digit |>> int_of_string |>> (fun i -> `Int i)

let space =

nomaddo

(* ocamljava -o caml.jar -runtime-parameter runtime-lock=off -I +concurrent -I +javalib javalib.cmja -java-extensions concurrent.cmja unix.cmja ./testjava.ml *)

open JavaPervasives

let size = 1000000

let a = Array.make size 100

let rec tarai x y z =
  if x <= y then y