|
From: Julian S. <js...@ac...> - 2007-08-29 09:14:23
|
ok, it's done. Your test program works correctly it seems. Check out and build the trunk as per http://valgrind.org/downloads/repository.html and let me know if it does/doesn't work w/ your application. J On Wednesday 29 August 2007 09:09, Daniel Frampton wrote: > I believe so yes. > > The hack I did to get JikesRVM to run under valgrind did exactly this, > although in an inelegant way (by simply performing a broken operation - > load from address 0). > > Cheers, > Daniel. > > > -----Original Message----- > > From: Julian Seward [mailto:js...@ac...] > > Sent: Wednesday, 29 August 2007 5:33 PM > > To: Daniel Frampton > > Cc: Steve Blackburn; val...@li...; Chris > > January; Robin Garner > > Subject: Re: [Valgrind-developers] Supporting the x86 INT > > instructionin > > > Valgrind > > > > On Friday 29 June 2007 11:06, Daniel Frampton wrote: > > > To add to what Steve said, I saw the implementation of INT3 on the > > > trunk, but unfortunately we use INT (0xCD) 0x40 - 0x43, and not INT3 > > > (0xCC). > > > > Just to check what needs to happen: you're expecting Valgrind to > > deliver > > SIGSEGV to your application when encountering INT (0xCD) 0x40 - 0x43, > > yes? > > > > J |