Hi Tom,

> On 09/10/09 13:53, Ralph Corderoy wrote:
> > Because Allegro is used, valgrind and cachegrind can't be used;
> > Allegro doesn't play nicely with them.  :-(
> 
> What's the problem with valgrind and allegro?

    $ valgrind ./rpcemu
    ==9782== Memcheck, a memory error detector.
    ==9782== Copyright (C) 2002-2007, and GNU GPL'd, by Julian Seward et al.
    ==9782== Using LibVEX rev 1804, a library for dynamic binary translation.
    ==9782== Copyright (C) 2004-2007, and GNU GPL'd, by OpenWorks LLP.
    ==9782== Using valgrind-3.3.0-Debian, a dynamic binary instrumentation 
framework.
    ==9782== Copyright (C) 2000-2007, and GNU GPL'd, by Julian Seward et al.
    ==9782== For more details, rerun with: -v
    ==9782== 
    rpcemu: version=0.8.5 _DEBUG=no DYNAREC=yes
    Shutting down Allegro due to signal #11
    ==9782== 
    ==9782== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 76 from 1)
    ==9782== malloc/free: in use at exit: 39,084 bytes in 730 blocks.
    ==9782== malloc/free: 5,886 allocs, 5,156 frees, 365,567 bytes allocated.
    ==9782== For counts of detected errors, rerun with: -v
    ==9782== searching for pointers to 730 not-freed blocks.
    ==9782== checked 12,604,404 bytes.
    ==9782== 
    ==9782== LEAK SUMMARY:
    ==9782==    definitely lost: 0 bytes in 0 blocks.
    ==9782==      possibly lost: 0 bytes in 0 blocks.
    ==9782==    still reachable: 39,084 bytes in 730 blocks.
    ==9782==         suppressed: 0 bytes in 0 blocks.
    ==9782== Rerun with --leak-check=full to see details of leaked memory.
    Segmentation fault
    $ 

This is package liballegro4.2.  I found

    http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=395909

And http://www.allegro.cc/forums/thread/598026 suggests if Allegro is
re-built without assembler then things improve.

Is valgrind and friends how you've profiled in the past?

Cheers,


Ralph.


_______________________________________________
Rpcemu mailing list
[email protected]
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

Reply via email to