[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: The Mauve unicode testcase and VM performance
From: |
Mark Wielaard |
Subject: |
RE: The Mauve unicode testcase and VM performance |
Date: |
Tue, 13 Apr 2004 20:41:43 +0200 |
Hi,
On Tue, 2004-04-13 at 11:38, Jeroen Frijters wrote:
> This is what I did:
> #cd mauve
> #mono ../ikvm/bin/ikvmc.exe -out:mauve.exe
> -main:gnu.testlet.SimpleTestHarness `find | grep -v altered | grep
> .*class$` -reference:../ikvm/bin/classpath.dll
> #cp ../ikvm/bin/*.dll .
Aha! OK I actually did:
mono ../ikvm/bin/ikvmc.exe -out:mauve.exe \
-main:gnu.testlet.SimpleTestHarness \
gnu/testlet/*.class gnu/testlet/java/lang/Character/*.class \
-reference:../ikvm/bin/classpath.dll
Which seemed to give a slightly faster executable then when including
all mauve classes.
$ time echo gnu.testlet.java.lang.Character.unicode | mono mauve.exe -debug
Reading unicode database...
done
Benchmark : load:2944ms tests:1337ms
0 of 3578944 tests failed
real 0m4.905s
user 0m0.072s
sys 0m0.012s
(Best of three runs again)
load test overhead total
Old: 1703 1292 3174 6169
New: 2944 1337 624 4905
So the overhead/startup cost is way down, but both load and test take
much longer. Why is that do you think?
> > The other thing to try would be using Mono's --aot option
> > to pre-JIT (as it were) the generated exe and dlls.
>
> I couldn't get that to work.
OK seems that it simply:
$ mono -O=all --aot mauve.exe
Mono Ahead of Time compiler - compiling assembly mauve.exe
Executing the native assembler: as /tmp/mono_aot_hSe8Lr -o
/tmp/mono_aot_hSe8Lr.o
Executing the native linker: ld -shared -o mauve.exe.so
/tmp/mono_aot_hSe8Lr.o
Compiled 103 out of 104 methods (99%)
0 methods contain absolute addresses (0%)
1 methods contain wrapper references (0%)
0 methods contain lmf pointers (0%)
0 methods have other problems (0%)
AOT RESULT 0
You then still need the original "exe" for execution.
The man page says about --aot:
This option is used to precompile the CIL code in the specified
assembly to native code. The generated code is stored in a file
with the extension .so. This file will be automatically picked
up by the runtime when the assembly is executed.
This pre-compiles the methods, but the original assembly is
still required to execute as this one contains the metadata and
exception information which is not availble on the generated
file. When precompiling code, you might want to compile with
all optimizations (-O=all). Pre-compiled code is position inde-
pendent code.
$ time echo gnu.testlet.java.lang.Character.unicode | mono mauve.exe -debug
Reading unicode database...
done
Benchmark : load:2944ms tests:1312ms
0 of 3578944 tests failed
real 0m4.882s
user 0m0.068s
sys 0m0.017s
(Best of three again)
load test overhead total
Old: 1703 1292 3174 6169
New: 2944 1337 624 4905
Oat: 2944 1312 626 4882
So that doesn't seem to help in this case (or I did something wrong and
it isn't picking up the mauve.exe.so after all).
Cheers,
Mark
signature.asc
Description: This is a digitally signed message part
- Re: The Mauve unicode testcase and VM performance, (continued)
- RE: The Mauve unicode testcase and VM performance, Jeroen Frijters, 2004/04/13
- RE: The Mauve unicode testcase and VM performance,
Mark Wielaard <=
- Re: The Mauve unicode testcase and VM performance, Stuart Ballard, 2004/04/13
- Re: The Mauve unicode testcase and VM performance, Mark Wielaard, 2004/04/13
- Re: [Mono-list] Re: The Mauve unicode testcase and VM performance, Jonathan Pryor, 2004/04/13
- Re: [Mono-list] Re: The Mauve unicode testcase and VM performance, Anthony Green, 2004/04/14
- Re: [Mono-list] Re: The Mauve unicode testcase and VM performance, Jonathan Pryor, 2004/04/15
- Re: [Mono-list] Re: The Mauve unicode testcase and VM performance, Anthony Green, 2004/04/15
- Re: [Mono-list] Re: The Mauve unicode testcase and VM performance, Miguel de Icaza, 2004/04/15
- Re: [Mono-list] Re: The Mauve unicode testcase and VM performance, Miguel de Icaza, 2004/04/14