• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
Re: Xcode preference gives me hope
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Xcode preference gives me hope


  • Subject: Re: Xcode preference gives me hope
  • From: Jim Ingham <email@hidden>
  • Date: Tue, 14 Mar 2006 14:38:11 -0800

I was going to change to subject to "Jim takes the hope away"... but I thought I'd let the body do that :-(

"Load CFM symbol information" does very little, and whatever it does do is already done in the command-line gdb unless you've specifically turned it off.

If "load CFM symbol information" is turned off, then gdb will arrange not to be notified of CFM fragment loads, and will not search the CFM fragment list to see what fragments are loaded. The only time gdb will read a CFM fragment is if it happens to show up in a backtrace. It is equivalent to setting "inferior-auto-load-cfm" to "off" in the gdb command line.

gdb does scan the trace-back tables in CFM fragments (when it reads them at all). So if there's traceback information in your CFM code, we will be able to reconstruct functions from there. But this isn't true symbol information. Reading this is slow however, since it amounts to scanning all your program text for consecutive null's...

gdb does in fact have a buggy, slow and completely unmaintained xSYM parser. So there is some chance that if you put your xSYM file in the same directory as the fragment, and pass the path to the fragment when it's loaded so we can find it, we might read some symbol information out of the xSYM file. It worked - though poorly - 4 years ago; but that's the last time I tried it.

We don't do anything to actively break the CFM support but OTOH, beyond what's needed for CodeWarrior (which due to the way it works is very little) we don't plan on putting any effort into supporting CFM. This was true before the Intel transition, and given that there's no support for CFM in the Intel Mac OS X, it is even more true now.

Sorry...

Jim

On Mar 14, 2006, at 1:48 PM, Mark Sanvitale wrote:

The Debugging pane of Xcode's preferences window includes the checkbox "Load CFM symbol information". Now, I am no CFM expert, but the "Carbon Porting Guide" discusses the limitations of using GDB to debug a CFM app:

" you cannot set breakpoints at CFM functions, including those in your application"
" GDB cannot use the symbol names in a CFM application"


which seems to contradict what I expect to happen from the mentioned Xcode preference.

So, what specifically (i.e. at the gdb level) does this preference do? Because if it does what I hope it does (i.e. remove above limitations) then I want to do it in my command-line GDB sessions for CFM apps.


Mark Sanvitale Electrical Geodesics, Inc. 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

_______________________________________________ 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
References: 
 >Xcode preference gives me hope (From: Mark Sanvitale <email@hidden>)

  • Prev by Date: Re: One byte bool
  • Next by Date: Re: One byte bool
  • Previous by thread: Xcode preference gives me hope
  • Next by thread: One byte bool
  • Index(es):
    • Date
    • Thread