Re: query on Apple crashreporter logs
Re: query on Apple crashreporter logs
- Subject: Re: query on Apple crashreporter logs
- From: Arnab Ganguly <email@hidden>
- Date: Wed, 13 Jan 2010 00:07:48 +0530
Thanks for the update.Does it have any dependency with ASLR set ?So far I haven't come across crash report where I get the address space overlap of modules but when I try to resolve symbols on a different machine with respect to program counter passed from the crashed machine , in some cases I am getting overlap and it is not matching with Crash report.
I am able to find out module name , line number and function name correctly but the range of load address in some cases I am getting an overlap. Any pointers or inputs would be very much appreciated.
Thanks in advance.
-A
On Tue, Jan 12, 2010 at 8:24 PM, Shantonu Sen
<email@hidden> wrote:
For the read-only __TEXT segment, there should be no overlaps. Otherwise function calls intended for one of the images at that address would start executing another image's instruction stream, probably resulting in crashes.
For dylibs in the shared cache, the writable segments (like __DATA) are not necessarily laid out directly after __TEXT, and can slide independently. So using the __TEXT segments load addresses to infer the __DATA load addresses may result in incorrect assumptions and overlaps, because that's not actually where __DATA is.
Shantonu
Sent from my MacBook
On Jan 12, 2010, at 3:52 AM, Arnab Ganguly wrote:
>
> Hi All,
> I have a very basic query about the binary images loaded in the process as shown by Apple crash report.
>
> Is there any chance that any of the binaries can have overlapping address space regions?
>
>
> 0x90a71000 - 0x90e2ffea libLAPACK.dylib ??? (???) /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libLAPACK.dylib
>
> 0x90e30000 - 0x90e39fff com.apple.speech.recognition.framework 3.7.24 (3.7.24) <6a6518b392d3d41ace3dcea69d6809d9> /System/Library/Frameworks/Carbon.framework/Versions/A/Frameworks/SpeechRecognition.framework/Versions/A/SpeechRecognition
>
> 0x90e3a000 - 0x90e46ffe libGL.dylib ??? (???) /System/Library/Frameworks/OpenGL.framework/Versions/A/Libraries/libGL.dylib
>
>
> The above address spaces are unique for each binary so is there a case where we may get some overlapping region or it is absurd.
> Thanks in advance.
> -A
> Do not post admin requests to the list. They will be ignored.
> Darwin-dev 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.
Darwin-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden