<div dir="auto">Hi Stephan,<div dir="auto"><br></div><div dir="auto">Sorry i didn't wrote all explanation yet, it's execution time in seconds to convert the ODT standart from odt to pdf, I know that's not a complete performance test but i will add more test.</div><div dir="auto"><br></div><div dir="auto">I didn' measure build time, i guess clang is better but not sure and that's not what i was checking</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">Le mar. 23 nov. 2021 à 17:58, Stephan Bergmann <<a href="mailto:sbergman@redhat.com">sbergman@redhat.com</a>> a écrit :<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On 22/11/2021 17:28, Arnaud Versini wrote:<br>
> I just tested libreoffice performance between GCC + gold + libstdc++, <br>
> clang + lld + libstdc++ and clang + lld + libc++ and noticed that gcc <br>
> builds are slower (at least in this case). I will write a blog article <br>
> with all the process to reproduce but small version, it just converts <br>
> openDocument standart files (ODF 1.0 to 1.3) to PDF. There is no <br>
> significant difference between both clang builds (changing lib C++ from <br>
> libstdc++ to libc++) but gcc is slower than clang. Of course all builds <br>
> don't use system libraries in this case to compare between lib C++, and <br>
> builds are optimised and use LTO.<br>
> <br>
> <br>
> OpenSUSE 7.2 Clang + libc++ Clang + std Gcc + std<br>
> Run 1 50,923045275 45,456441587 45,878275287 49,487550577<br>
> Run 2 50,360027388 43,83441037 44,303120572 47,898307652<br>
> Run 3 50,258628137 43,552443057 44,332110013 47,596628518<br>
> Run 4 50,062155622 43,587905602 44,616728628 47,430735675<br>
> Run 5 53,567235331 43,548408389 44,561502427 47,727024813<br>
> Run 6 50,354870307 43,877387065 44,107891362 47,444552655<br>
> Run 7 49,663738973 43,92048805 44,583599634 47,293633635<br>
> Run 8 49,780141995 43,843555496 44,055990454 47,561677627<br>
> Run 9 49,763425026 43,398042427 44,011950682 47,65275752<br>
> Run 10 50,197944032 44,153460819 44,303763557 47,974880802<br>
> Average 50,4931212086 43,9172542862 44,4754932616 47,8067749474<br>
<br>
What are you even measuring here, wall-clock time to build LO with the <br>
various toolchains, or wall-clock time to execute something from the <br>
resulting buils, or...? (And what is the unit for those numbers?)<br>
<br>
</blockquote></div>