Re: stacktrace from withn an exec upon error condition

2002-04-02 Thread Lassi A. Tuura
> I haven't looked at it recently, but I really doubt that the code would > deal with the stabs symbol table that gcc/gas/ld use and allow symbolic > representation of symbols. My code certainly doesn't, it only deals with basic Win32. You get symbols for addresses only if imagehlp library can g

Re: stacktrace from withn an exec upon error condition

2002-04-02 Thread Christopher Faylor
On Tue, Apr 02, 2002 at 03:37:24PM +0200, Lassi A. Tuura wrote: >There's been code in MSJ (Microsoft Systems Journal) ages ago by Matt >Pietrek, in several articles. Do a google search for "Pietrek >stackwalk" for instance. > >http://iguana.cern.ch/lxr/source/IGUANA/src/Ig_Imports/classlib/src/de

Re: stacktrace from withn an exec upon error condition

2002-04-02 Thread Lassi A. Tuura
There's been code in MSJ (Microsoft Systems Journal) ages ago by Matt Pietrek, in several articles. Do a google search for "Pietrek stackwalk" for instance. http://iguana.cern.ch/lxr/source/IGUANA/src/Ig_Imports/classlib/src/debug.cxx#1099 has stack trace code you can use on a large number of sy

RE: stacktrace from withn an exec upon error condition

2002-03-29 Thread Robert Collins
> -Original Message- > From: Hans Horn [mailto:[EMAIL PROTECTED]] > Sent: Saturday, March 30, 2002 1:37 AM > p.s. Do you consider me fish? As I have nipples, I'd rather > be referred to a mammal. Hans, the references to fish in this list are not suggesting that anyone is a fish, but

Re: stacktrace from withn an exec upon error condition

2002-03-29 Thread Christopher Faylor
On Fri, Mar 29, 2002 at 06:37:23AM -0800, Hans Horn wrote: >p.s. Do you consider me fish? As I have nipples, I'd rather be >referred to a mammal. http://cygwin.com/cgi-bin/htsearch?method=and&sort=time&restrict=%2Fml%2Fcygwin%2F&exclude=&words=fish+teach cgf -- Unsubscribe info: http://c

Re: stacktrace from withn an exec upon error condition

2002-03-29 Thread Hans Horn
Dear Chris, this looks like it would take more like a hardcore system programmer to do, rather than naive me. In any case, I'm very grateful that you have taken so much of your precious time to respond to my posting. thanks a lot Hans p.s. Do you consider me fish? As I have nipples, I'd rather

RE: stacktrace from withn an exec upon error condition

2002-03-29 Thread Heribert Dahms
Hi Hans, I sometimes manually lookup like: $ gdb foo.exe l *0xdeadbeef Bye, Heribert ([EMAIL PROTECTED]) -Original Message- From: Christopher Faylor [mailto:[EMAIL PROTECTED]] Sent: Freitag, 29. März 2002 05:06 To: [EMAIL PROTECTED] Subject: Re: stacktrace from withn an exec upon

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Christopher Faylor
On Thu, Mar 28, 2002 at 11:05:32PM -0500, Christopher Faylor wrote: >For the record, I don't know anything about what is required to >interface with bfd. I'm offering this information so that Robert Randall >Schulz won't accuse me of bei

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Christopher Faylor
On Thu, Mar 28, 2002 at 07:40:03PM -0800, Hans Horn wrote: >I just tried Chris' little test and got: something like this: Wow, hens' teeth prevail. >This however, is of little help (at least not to me). > >What I had in mind, though, was to get a list of the function names >(demangled of course)

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Hans Horn
Dear all, I just tried Chris' little test and got: something like this: Stack trace: Frame Function Args 0022FE84 00401073 (00401244, , 0022FEA4, 00401063) 0022FE94 00401073 (610A0368, 0001, 0022FEB4, 00401054) 0022FEA4 00401063 (00D0, , 0022FF10, 61003F42) 0022

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Christopher Faylor
On Thu, Mar 28, 2002 at 06:46:52PM -0800, Richard Hitt wrote: >Is it unreasonable to find appalling the suggestion to use an >undocumented function? Are you volunteering to write documentation? That would be great! We really need someone to work on documentation! It seems that people frequently

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Richard Hitt
Is it unreasonable to find appalling the suggestion to use an undocumented function? Richard On Thu, 28 Mar 2002 17:17:51 -0500, you wrote: >On Thu, Mar 28, 2002 at 08:44:27AM -0800, Hans Horn wrote: >>where does cygwin_stackdump() live? >>I searched through the entire cygwin inst tree, but cou

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Randall R Schulz
Chris, I'm sorry. I missed that suggestion. But I do enjoy jumping through hoops at times. Randall Schulz Mountain View, CA USA At 17:06 2002-03-28, you wrote: >On Thu, Mar 28, 2002 at 04:32:36PM -0800, Randall R Schulz wrote: > >If you're going to take the time to respond, the least you cou

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Christopher Faylor
On Thu, Mar 28, 2002 at 04:32:36PM -0800, Randall R Schulz wrote: >If you're going to take the time to respond, the least you could do is >give a useful answer. > >Since unlike Chris, I do not know the answer, I'll illustrate how I >discovered it. > >[snip] On Thu, Mar 28, 2002 at 05:17:51PM -050

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Randall R Schulz
Chris, If you're going to take the time to respond, the least you could do is give a useful answer. Since unlike Chris, I do not know the answer, I'll illustrate how I discovered it. Hans is correct that there is no prototype for this function in any file under /usr/include, though there i

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Larry Hall (RFK Partners, Inc)
At 06:05 PM 3/28/2002, Christopher Faylor wrote: >On Thu, Mar 28, 2002 at 03:13:15PM -0800, Hans Horn wrote: > >I'm afraid not, as I do not know what files to #include so the runtime will > >find cygwin_stackdump(). > >I guess you're out of luck then. A shadowy apparition appears, shrouded in a

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Christopher Faylor
On Thu, Mar 28, 2002 at 03:13:15PM -0800, Hans Horn wrote: >I'm afraid not, as I do not know what files to #include so the runtime will >find cygwin_stackdump(). I guess you're out of luck then. cgf -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http:/

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Hans Horn
Chris, I'm afraid not, as I do not know what files to #include so the runtime will find cygwin_stackdump(). H. -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http://cygwin.com/bugs.html Documentation: http://cygwin.com/docs.html FAQ:

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Christopher Faylor
On Thu, Mar 28, 2002 at 08:44:27AM -0800, Hans Horn wrote: >where does cygwin_stackdump() live? >I searched through the entire cygwin inst tree, but couldn't find any header >where it is defined. > >If it is a function I just invoke, where does it write the trace dump to? >stdout? stderr? I sugge

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Christopher Faylor
On Thu, Mar 28, 2002 at 06:30:39AM -0800, Hans Horn wrote: >is there a way to create (and print) a procedure traceback from within an >executable that is invoked upon error condition? >This is possible on many other platforms. >I've read about strace, but that seems to be not what I want, as to my

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Larry Hall (RFK Partners, Inc)
At 10:17 AM 3/28/2002, Hans Horn wrote: >Thanks a lot. >But I'm sorry - I've never dealt with the Win32 API. Don't you have to join >Micro$haft in order to access the MSDN site? >H. No, I didn't. Try http://msdn.microsoft.com/library. Poke around and see what you can find. Hm, try looking st

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Hans Horn
Thanks a lot. But I'm sorry - I've never dealt with the Win32 API. Don't you have to join Micro$haft in order to access the MSDN site? H. -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http://cygwin.com/bugs.html Documentation: http://cygwin

Re: stacktrace from withn an exec upon error condition

2002-03-28 Thread Larry Hall (RFK Partners, Inc)
At 09:30 AM 3/28/2002, Hans Horn wrote: >Dear all, > >is there a way to create (and print) a procedure traceback from within an >executable that is invoked upon error condition? >This is possible on many other platforms. >I've read about strace, but that seems to be not what I want, as to my >unde

stacktrace from withn an exec upon error condition

2002-03-28 Thread Hans Horn
Dear all, is there a way to create (and print) a procedure traceback from within an executable that is invoked upon error condition? This is possible on many other platforms. I've read about strace, but that seems to be not what I want, as to my understanding the executable needs to be run under