Hi, guys!
I've made DMD build using MSVC(ICC has some performance troubles) and tcmalloc, so it can compile more than 2x faster.

http://acomirei.ru/u/dmd.7z
It uses lastest DMD, druntime and Phobos.

Do you have a source for this? I don't trust random binaries, especially ones that are hosted on an .ru site..

On 8/31/13, Temtaime <temtaime@gmail.com> wrote:
> Hi, guys!
> I've made DMD build using MSVC(ICC has some performance troubles)
> and tcmalloc, so it can compile more than 2x faster.
>
> http://acomirei.ru/u/dmd.7z
> It uses lastest DMD, druntime and Phobos.
>

You can get DMD source on DMD's github.

I've read DMD backend license, so it forbids to distribute DMD itself.

I've removed the archive. I'll investigate in LDC.

On 8/31/13, Temtaime <temtaime@gmail.com> wrote:
> You can get DMD source on DMD's github.

I'm asking about the modified version which uses tcmalloc. You've said it compiles 2x faster, but faster to what? The regular MSVC build is already known to be 2x faster than the one built with DMC.

Am 31.08.2013 20:58, schrieb Temtaime:
> You can get DMD source on DMD's github.
>
> I've read DMD backend license, so it forbids to distribute DMD
> itself.
>
> I've removed the archive. I'll investigate in LDC.
>

it seems that the old malloc implementation was the source of
the 2x speed difference between dmc and msvc build

http://forum.dlang.org/thread/ktju8h$efb$1@digitalmars.com

so you're tcmalloc version is 2x faster or 4x?

Am 31.08.2013 20:58, schrieb Temtaime:
> You can get DMD source on DMD's github.
>
> I've read DMD backend license, so it forbids to distribute DMD
> itself.
>
> I've removed the archive. I'll investigate in LDC.
>

btw: nedmalloc stated on its own homepage...
http://www.nedprod.com/programs/portable/nedmalloc/

"If you're running on an older operating system (e.g. Windows XP, Linux 2.4 series, FreeBSD 6 series, Mac OS X 10.4 or earlier) you will probably find it significantly improves your application's performance (Windows 7, Linux 3.x, FreeBSD 8, Mac OS X 10.6 all contain state-of-the-art allocators and no third party allocator is likely to significantly improve on them in real world results)."

so you're running an older OS?

On 9/1/13, dennis luehring <dl.soluz@gmx.net> wrote:
> it seems that the old malloc implementation was the source of the 2x speed difference between dmc and msvc build

Not really, the MSVC build is still faster. :)

On 01/09/2013 13:46, Andrej Mitrovic wrote:
> On 9/1/13, dennis luehring <dl.soluz@gmx.net> wrote:
>> it seems that the old malloc implementation was the source of
>> the 2x speed difference between dmc and msvc build
>
> Not really, the MSVC build is still faster. :)
>

It made up the bulk of the difference for the algorithm unit tests at least - The MSVC build is still faster, but only a bit rather than several times.

I'm using Windows 8.1.

2x faster than DMD from downloads section(seems to be built with DMC).

I'm only removed overloading for operators new and delete from DMD frontend and link statically tcmalloc 2.0.

Am 02.09.2013 18:37, schrieb Temtaime:
> I'm using Windows 8.1.
>
> 2x faster than DMD from downloads section(seems to be built with
> DMC).
>
> I'm only removed overloading for operators new and delete from
> DMD frontend and link statically tcmalloc 2.0.
>

and what is the speed gain comparing your msc_dmd_tcmalloc with plain msc_dmd?

test both with

dmd std\algorithm -unittest –main