Databases
StackOverflow2013
Postsdbo
POOptimisation for a brainfuck interpreter
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POOptimisation for a brainfuck interpreter
    primarykey
    data
    text
    <p>As an exercise to help me learn about interpreters and optimisation, neither of which I know anything about, I have written a brainfuck interpreter in C. It appears to work flawlessly thus far, though it does not compete well in execution speed compared to other fast interpreters.</p> <p>What are some ways that I can change this interpreter to improve performance (or otherwise)?</p> <p>One interesting aspect of my interpreter (though most others probably do this as well) is that I run one loop that reads through the source input and converts each instruction into a</p> <pre><code>struct { long instruction; long loop; } </code></pre> <p>The <code>loop</code> value is the index of the matching <code>]</code> instruction, if the instruction is a <code>[</code>, and the index of the matching <code>[</code> instruction, if the instruction is a <code>]</code>, allowing quick jumping. I'd imagine that this 'parsing' process (which doesn't take long) improves execution times over doing redundant reparsing to find matching square brackets every time they are needed.</p> <p>An interesting test of brainfuck interpreter speed is this program:</p> <pre><code>++++++++[-&gt;-[-&gt;-[-&gt;-[-]&lt;]&lt;]&lt;]&gt;++++++++[&lt;++++++++++&gt;-]&lt;[&gt;+&gt;+&lt;&lt;-]&gt;-.&gt;-----.&gt; </code></pre> <ol> <li><a href="http://azabani.com/cgit.cgi/brief/plain/brief.c?id=0884ccb22287158fd3664eb83427444644aaa328" rel="nofollow noreferrer">the first version of the interpreter</a></li> <li><a href="http://azabani.com/cgit.cgi/brief/plain/brief.c?id=92a1dc40c549e25927831fee469429a9ab3be293" rel="nofollow noreferrer">the interpreter</a> after implementing <a href="https://stackoverflow.com/questions/6853548/optimisation-for-a-brainfuck-interpreter/6853750#6853750">Jerry Coffin's answer</a>, which removes the giant switch in the runtime loop, by making the <code>instruction</code> struct's <code>instruction</code> a direct pointer to an operation function - <em>this runs slower than the previous version (function call overhead?)</em></li> <li><a href="http://azabani.com/cgit.cgi/brief/plain/brief.c?id=c7fd6e04a88a60bd9ec480c00718ca7d91834898" rel="nofollow noreferrer">the interpreter</a> after <strong>reversing the previous change</strong> and adding an optimisation to 'collapse' multiple consecutive non-loop operations, reducing loop cycles - <em>this runs slightly faster than the original</em></li> </ol>
    singulars
    1. PO
      singulars
      1. PTAnswer
    1. This table or related slice is empty.
    1. PTQuestion
    1. USCommunity
    1. USDelan Azabani
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    2. PO
      singulars
      1. PTAnswer
    3. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POOptimisation for a brainfuck interpreter
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POOptimisation for a brainfuck interpreter
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. POOptimisation for a brainfuck interpreter
      1. This table or related slice is empty.
      1. VTUpMod
    1. COAs a side-note, your code (all versions, I think) has a bug. In the code for: "case BF_OP_LSTART:", you have: "ls = realloc(im, sizeof(long) * (la *= 2));", where I'm sure you intended: "ls = realloc(ls, sizeof(long) * (la *= 2));".
      singulars
      1. POOptimisation for a brainfuck interpreter
      1. USJerry Coffin
    2. COThanks for picking that up! The parser should be doubling and returning the allocation to the loop stack, not the instruction memory.
      singulars
      1. POOptimisation for a brainfuck interpreter
      1. USDelan Azabani
    3. COUsing `realloc` that way leads to memory leaks. You need to check the return value before assigning back over the original pointer.
      singulars
      1. POOptimisation for a brainfuck interpreter
      1. USR..
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload