Thread overview
Compile time performance for metaprogramming is somewhat inconsistent
Mar 03, 2016
maik klein
Mar 03, 2016
cym13
Mar 03, 2016
maik klein
Mar 03, 2016
John Colvin
Mar 03, 2016
maik klein
Mar 03, 2016
Stefan Koch
March 03, 2016
Consider the following code

void main()
{
    import std.stdio;
    import std.range: iota, join;
    import std.algorithm.iteration: map;
    import std.conv: to;
    import std.meta: aliasSeqOf, staticMap, AliasSeq;
    enum types = "AliasSeq!(" ~ iota(0,10000).map!(i => to!string(i)).join(",") ~ ")";
    alias t = AliasSeq! (mixin(types));
    //alias t1 = aliasSeqOf!(iota(0, 10000));
}

't' compiles on my machine in ~3.5 seconds while 't1' needs ~1 minute to compile. It seems that mixins are just way more performant than template instantiations. Any ideas why? What causes the slowdown and what can I improve?

Also I compared some meta stuff in C++ and D.

For example filtering

enum isEven(alias i) = i % 2 is 0;
void main()
{
    import std.stdio;
    import std.range: iota, join;
    import std.algorithm.iteration: map;
    import std.conv: to;
    import std.meta: AliasSeq, Filter;
    enum types = "AliasSeq!(" ~ iota(0,10000).map!(i => to!string(i)).join(",") ~ ")";

    alias t = AliasSeq!(mixin(types));
    alias evenTypes = Filter!(isEven,t);
}

Someone was also so kind to create this in C++ though it looks a bit more crazy because he wanted to do roughly the same thing.

https://gist.github.com/ricejasonf/8c2b54c182e6038fd0ce

The D version compiles in ~14.5 seconds while the C++ version compiles in ~4.2 seconds. This was very surprising to me.

The more Hana like code is here https://github.com/boostorg/hana/blob/master/benchmark/filter/compile.hana.tuple.erb.cpp

*I wasn't yet able to run the Hana benchmarks yet because the build script doesn't detect my ruby executable.







March 03, 2016
On Thursday, 3 March 2016 at 02:03:01 UTC, maik klein wrote:
> Consider the following code
>
> void main()
> {
>     import std.stdio;
>     import std.range: iota, join;
>     import std.algorithm.iteration: map;
>     import std.conv: to;
>     import std.meta: aliasSeqOf, staticMap, AliasSeq;
>     enum types = "AliasSeq!(" ~ iota(0,10000).map!(i => to!string(i)).join(",") ~ ")";
>     alias t = AliasSeq! (mixin(types));
>     //alias t1 = aliasSeqOf!(iota(0, 10000));
> }
>
> [...]

As often, compiler-version-flags please?
March 03, 2016
On Thursday, 3 March 2016 at 02:26:09 UTC, cym13 wrote:
> On Thursday, 3 March 2016 at 02:03:01 UTC, maik klein wrote:
>> Consider the following code
>>
>> void main()
>> {
>>     import std.stdio;
>>     import std.range: iota, join;
>>     import std.algorithm.iteration: map;
>>     import std.conv: to;
>>     import std.meta: aliasSeqOf, staticMap, AliasSeq;
>>     enum types = "AliasSeq!(" ~ iota(0,10000).map!(i => to!string(i)).join(",") ~ ")";
>>     alias t = AliasSeq! (mixin(types));
>>     //alias t1 = aliasSeqOf!(iota(0, 10000));
>> }
>>
>> [...]
>
> As often, compiler-version-flags please?

default dub.sdl and `dub build` with dmd v2.070
March 03, 2016
On Thursday, 3 March 2016 at 02:03:01 UTC, maik klein wrote:
> Consider the following code
>
> void main()
> {
>     import std.stdio;
>     import std.range: iota, join;
>     import std.algorithm.iteration: map;
>     import std.conv: to;
>     import std.meta: aliasSeqOf, staticMap, AliasSeq;
>     enum types = "AliasSeq!(" ~ iota(0,10000).map!(i => to!string(i)).join(",") ~ ")";
>     alias t = AliasSeq! (mixin(types));
>     //alias t1 = aliasSeqOf!(iota(0, 10000));
> }
>
> 't' compiles on my machine in ~3.5 seconds while 't1' needs ~1 minute to compile. It seems that mixins are just way more performant than template instantiations. Any ideas why? What causes the slowdown and what can I improve?

What happens if you add a few extra branches to std.meta.aliasSeqOf, e.g. https://github.com/D-Programming-Language/phobos/commit/5d2cdf103bd697b8ff1a939c204dd2ed0eec0b59

Only a linear improvement but maybe worth a try?
March 03, 2016
On Thursday, 3 March 2016 at 02:03:01 UTC, maik klein wrote:
> Consider the following code
>
> void main()
> {
>     import std.stdio;
>     import std.range: iota, join;
>     import std.algorithm.iteration: map;
>     import std.conv: to;
>     import std.meta: aliasSeqOf, staticMap, AliasSeq;
>     enum types = "AliasSeq!(" ~ iota(0,10000).map!(i => to!string(i)).join(",") ~ ")";
>     alias t = AliasSeq! (mixin(types));
>     //alias t1 = aliasSeqOf!(iota(0, 10000));
> }
>
> [...]

Templates are by default more work then mixins.
If you want me to I can elaborate on that.
March 03, 2016
On Thursday, 3 March 2016 at 11:40:29 UTC, John Colvin wrote:
> On Thursday, 3 March 2016 at 02:03:01 UTC, maik klein wrote:
>> Consider the following code
>>
>> void main()
>> {
>>     import std.stdio;
>>     import std.range: iota, join;
>>     import std.algorithm.iteration: map;
>>     import std.conv: to;
>>     import std.meta: aliasSeqOf, staticMap, AliasSeq;
>>     enum types = "AliasSeq!(" ~ iota(0,10000).map!(i => to!string(i)).join(",") ~ ")";
>>     alias t = AliasSeq! (mixin(types));
>>     //alias t1 = aliasSeqOf!(iota(0, 10000));
>> }
>>
>> 't' compiles on my machine in ~3.5 seconds while 't1' needs ~1 minute to compile. It seems that mixins are just way more performant than template instantiations. Any ideas why? What causes the slowdown and what can I improve?
>
> What happens if you add a few extra branches to std.meta.aliasSeqOf, e.g. https://github.com/D-Programming-Language/phobos/commit/5d2cdf103bd697b8ff1a939c204dd2ed0eec0b59
>
> Only a linear improvement but maybe worth a try?

I have tried the same thing in general and stuff like this is always a huge improvement.

In this case it goes down from ~60 seconds to ~3.8 seconds. I have done the same thing with my compile time map function, which gave me a drastic improvement.

I think recursion is just really bad in general for compile time stuff, for example your version

alias t1 = aliasSeqOf!(iota(0, 10000));

compiles in 3.8 seconds and uses roughly 600mb of ram while

alias t1 = aliasSeqOf!(iota(0, 20000));

compiles in 10.2 seconds and uses 1.9gb ram.

The mixin version is always the fastest but it also consumes way more memory and explodes before 20k elements


March 03, 2016
On 3/2/16 9:33 PM, maik klein wrote:
> On Thursday, 3 March 2016 at 02:26:09 UTC, cym13 wrote:
>> On Thursday, 3 March 2016 at 02:03:01 UTC, maik klein wrote:
>>> Consider the following code
>>>
>>> void main()
>>> {
>>>     import std.stdio;
>>>     import std.range: iota, join;
>>>     import std.algorithm.iteration: map;
>>>     import std.conv: to;
>>>     import std.meta: aliasSeqOf, staticMap, AliasSeq;
>>>     enum types = "AliasSeq!(" ~ iota(0,10000).map!(i =>
>>> to!string(i)).join(",") ~ ")";
>>>     alias t = AliasSeq! (mixin(types));
>>>     //alias t1 = aliasSeqOf!(iota(0, 10000));
>>> }
>>>
>>> [...]
>>
>> As often, compiler-version-flags please?
>
> default dub.sdl and `dub build` with dmd v2.070

Try dub build -b release

-Steve