Re: Rosetta/XCode Problem
Re: Rosetta/XCode Problem
- Subject: Re: Rosetta/XCode Problem
- From: "shin kurokawa" <email@hidden>
- Date: Mon, 4 Dec 2006 00:55:11 -0500
I had something similar to this that crashed on
Macbooks running my UB program under Rosetta,
but working perfectly fine native (i386-on-intel
or ppc-on-g5/g4).
It took me a while to figure this out, but in
my case it was in xcode build setting:
the 64bit Integer flag...
-Shin
On 12/3/06, Bob Currier <email@hidden> wrote:
I've got a very strange problem with XCode and Rosetta and I'm looking
for some suggestions on how to track it down.
Some background:
I have a PPC app that has been successfully running for a long time.
Originally it used CodeWarrior, but I converted it to XCode some time
ago and it's been running fine throughout testing.
I changed it to generate both Intel and PowerPC code and it's been
testing out fine on both types of processors.
Now, because of some plug-in host issues (the host is not universal
yet), sometimes the PowerPC code will be run under Rosetta on Intel
machines. My old CodeWarrior-compiled version works fine in this
configuration. But the XCode-compiled version crashes.
I've tracked the crash down to local storage definition with an
initializer over 4 bytes in length. So these statements work:
UInt32 temp = 1234;
char array[10] = "ABC";
But a statement like this crashes every time:
char array[10] = "ABCDEFGH";
Making "array" a static makes the crash go away.
Compiling the same code in CodeWarrior generates a working version for
PPC and Rosetta. Running the XCode version on a PowerPC system works
fine. It has something to do with the specific code being emitted by
XCode that is crashing only in Rosetta.
The stack appears intact and of reasonable size when the crash occurs. A
stack trace looks completely normal: it shows execution up to the point
of the crash. I can find no evidence of memory corruption.
I'm using XCode 2.3 under 10.4.7.
If it works on PowerPC, it's supposed to work under Rosetta, right? I'm
sure there are still bugs lurking in Rosetta, but for something this
fundamental I think it far more likely to be something I'm doing.
Anyone have any ideas of where to look next?
-- Bob
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden