RFE: configuration for compiling HotSpot with Clang

Chris Lattner clattner at apple.com
Tue Mar 1 15:22:04 PST 2011

On Mar 1, 2011, at 2:24 AM, Volker Simonis wrote:

> The new numbers are now online at:
> http://weblogs.java.net/blog/simonis/archive/2011/02/10/compiling-hotspot-vm-clang
> After my first blog I got a lot of feedback and it turned out that the
> comparison was unfair because the way I did it Clang didn't use
> Precompiled Headers correctly. Now with the new configuration, it does
> use PCH and the results are indeed pleasant: Clang has now 15% faster
> build times compared to GCC.

This is great progress!  If -fcheck-new is an important feature for you (and not a workaround for an old gcc problem), please feel free to file a bug at http://llvm.org/bugs/

In the LLVM 3.0 timeframe (~6 months from now) -flimit-debug-info should be on by default and additional debug info improvements will also land.  I don't know how interested you are in tracking down the cause of the performance differences you see, but if there is anything specific LLVM is doing and there is a way to reproduce it in a small testcase, I would also really appreciate a bug report :)


More information about the hotspot-dev mailing list