public inbox for insight@sourceware.org
 help / color / mirror / Atom feed
* Insight Bug?
@ 2000-05-29  6:32 Paul_Heijman
  2000-05-29  8:12 ` Fernando Nasser
  2000-05-30  5:48 ` Finish [was: Insight Bug?] Fernando Nasser
  0 siblings, 2 replies; 7+ messages in thread
From: Paul_Heijman @ 2000-05-29  6:32 UTC (permalink / raw)
  To: insight

Hi folks,

I encountered the following, of what I think is a bug of Insight:

When I use the 'finish' button to complete a function, Insight apparently uses
breakpoints for this (they are hidden someway).
The problem is that when I press the 'continue' button afterwards, the 'hidden
breakpoint' GDB used for the 'finish' function, is still there.
Result: the program stops again.

I hope I explained the problem well enough for you, if I didn't, let me know.

Kind regards,
Paul Heijmann
OMRON ETC-NL
-




^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: Insight Bug?
  2000-05-29  6:32 Insight Bug? Paul_Heijman
@ 2000-05-29  8:12 ` Fernando Nasser
  2000-05-30  5:48 ` Finish [was: Insight Bug?] Fernando Nasser
  1 sibling, 0 replies; 7+ messages in thread
From: Fernando Nasser @ 2000-05-29  8:12 UTC (permalink / raw)
  To: Paul_Heijman; +Cc: insight, fnasser

Paul_Heijman@eu.omron.com wrote:
> 
> Hi folks,
> 
> I encountered the following, of what I think is a bug of Insight:
> 
> When I use the 'finish' button to complete a function, Insight apparently uses
> breakpoints for this (they are hidden someway).
> The problem is that when I press the 'continue' button afterwards, the 'hidden
> breakpoint' GDB used for the 'finish' function, is still there.
> Result: the program stops again.
> 
> I hope I explained the problem well enough for you, if I didn't, let me know.
> 

First of all: which host and which target are you talking about?
Which version of insight (gdb) are you using?  Where did you get it from?
How was it configured/compiled?  What is your O.S. version?

Te underlying debugger mechanism is very dependent on the architecture, OS
support and other details.  Problems may arise on a specific platform and not
on others.  Sometimes just in a specific OS version.  For remote targets,
sometimes is a problem with communication or withing a certain version of a
ROM monitor.

If you tell what your platform is, people with similar setups may be able to
help you with problems that the others may have no clue.


Another very important test one should do when encountering a problem while
using the GUI is to try running gdb with the command line interface.

Invoke gdb with "gdb -nw yourprog" (assuming you are debugging natively on a host)
and try the "finish" command at the same situation as above (try repeating the
corresponding actions you performed with the GUI).

This way we will know if it is a gdb problem or a GUI problem.  In you case, it must
be a gdb gdb probem as the GUI works fine on many systems.  If this is indeed the
case, please post your question to the gdb list as there may be people who have
seem your problem before or have some idea on how to investigate it further.  This is
the GUI list and people who sign to it are more into the user interface bits.

Regards.
F.


-- 
Fernando Nasser
Cygnus Solutions (a Red Hat company)    E-Mail:  fnasser@cygnus.com

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Finish [was: Insight Bug?]
  2000-05-29  6:32 Insight Bug? Paul_Heijman
  2000-05-29  8:12 ` Fernando Nasser
@ 2000-05-30  5:48 ` Fernando Nasser
  1 sibling, 0 replies; 7+ messages in thread
From: Fernando Nasser @ 2000-05-30  5:48 UTC (permalink / raw)
  To: Paul_Heijman; +Cc: insight

Paul_Heijman@eu.omron.com wrote:
> 
> Hi folks,
> 
> I encountered the following, of what I think is a bug of Insight:
> 
> When I use the 'finish' button to complete a function, Insight apparently uses
> breakpoints for this (they are hidden someway).
> The problem is that when I press the 'continue' button afterwards, the 'hidden
> breakpoint' GDB used for the 'finish' function, is still there.
> Result: the program stops again.
> 
> I hope I explained the problem well enough for you, if I didn't, let me know.
> 
> Here are the details:
> - GDB: 20000105
> - OS: i586-pc-cygwin (WINNT)
> - TARGET: m68k-coff
>
> I tried it in GDB with -nw option:
> (...)
> This works - that is, the application stays running without breaking on the end
> of the functions you see in the stack above.
> It must be an Insight bug then.

Is anyone else seeing a similar problem with the "finish" button?
I would like very much to know if this happens in some other platform besides
Paul's cross m68k?

So far we know it does not happen with the command line interface.
Please let me know if it fails in yours and what is your host and target.

Thanks.

-- 
Fernando Nasser
Red Hat - Toronto                       E-Mail:  fnasser@cygnus.com
2323 Yonge Street, Suite #300           Tel:  416-482-2661 ext. 311
Toronto, Ontario   M4P 2C9              Fax:  416-482-6299

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: Insight Bug?
       [not found]   ` <opsfn5rmzge8k94p@mail.thevertigo.com>
@ 2004-10-10 19:40     ` Keith Seitz
  0 siblings, 0 replies; 7+ messages in thread
From: Keith Seitz @ 2004-10-10 19:40 UTC (permalink / raw)
  To: Kip Warner; +Cc: insight

On Sun, 2004-10-10 at 12:29, Kip Warner wrote:
> > On Thu, 2004-10-07 at 21:08, Kip Warner wrote:
> >> >> (gdb) info func main
> >> >> All functions matching regular expression "main":
> >> >>Non-debugging symbols:
> >> >> 0x00401000  mainCRTStartup
> >> >>(gdb) list main
> >> >> Error: No line number known for main.
> >>
> >> Hey Keith,
> >>
> >> objdump gives me a ton of output when I pass it the -g switch and the  
> >> path
> >> to my win32 executable. Any ideas? I think it may be a bug in Insights  
> >> as
> >> I have heard of many others having this problem.
> >
> > Okay, I believe your problem is in gdb, but to be sure, I need to ask
> > another question. MinGW doesn't use a "main" function, so that was a
> > silly test. Can you list the contents of a function for which you know
> > you have debug info? Open a console and enter "list MYFUNC" or "info
> > func MYFUNC".
> >
> > Keith
> >
> >
> 
> Hey Keith,
> 
> I added...
> 
> void foo(void)
> {
>      puts("foo");
>      return;
> }
> 
> ...to my source and called it in WinMain to ensure that was compiled in  
> there and invoked. I compile with -g3 and no optimizations or strips. I  
> open Insight (--version gives: GNU gdb 2003-09-20-cvs (cygwin-special))  
> and select function "foo." It gives only assembly listing still, even when  
> I select "source" feature in listbox.
> 
> I then opened the gdb console from the menu and performed the following:
> 
> (gdb) list foo
> 9	in WinMain.cpp
> 
> (gdb) info func foo
> All functions matching regular expression "foo":
> 
> File WinMain.cpp:
> void foo();
> 
> (gdb)
> 
> I hope this is helping. There are many people having the same problem. You  
> are a major help. The entire MinGW community that uses a GUI debugger  
> depends on you =)

Okay, before I ask for a pre-compiled testcase, let me bark down a
hunch. First, let me ask if we've tried the whole debug window thing?
("tk ManagedWin::open DebugWin" in console window; then switch to
function by typing it into function combobox; any errors show up in
debug window?)

After that, try this: open a console window and type "maint print
psymbols OUTFILE". Look for the function "foo"  in OUTFILE. What does it
say? Win32 paths, I'll bet?m

Keith

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: Insight Bug?
       [not found] <opsfi9s0h3e8k94p@mail.thevertigo.com>
@ 2004-10-08 14:29 ` Keith Seitz
       [not found]   ` <opsfn5rmzge8k94p@mail.thevertigo.com>
  0 siblings, 1 reply; 7+ messages in thread
From: Keith Seitz @ 2004-10-08 14:29 UTC (permalink / raw)
  To: Kip Warner; +Cc: insight

On Thu, 2004-10-07 at 21:08, Kip Warner wrote:
> >> (gdb) info func main
> >> All functions matching regular expression "main":
> >>Non-debugging symbols:
> >> 0x00401000  mainCRTStartup
> >>(gdb) list main
> >> Error: No line number known for main.
> 
> Hey Keith,
> 
> objdump gives me a ton of output when I pass it the -g switch and the path  
> to my win32 executable. Any ideas? I think it may be a bug in Insights as  
> I have heard of many others having this problem.

Okay, I believe your problem is in gdb, but to be sure, I need to ask
another question. MinGW doesn't use a "main" function, so that was a
silly test. Can you list the contents of a function for which you know
you have debug info? Open a console and enter "list MYFUNC" or "info
func MYFUNC".

Keith

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: Insight Bug?
  2004-10-03  3:12 Insight Bug? Kip Warner
@ 2004-10-03  3:43 ` Keith Seitz
  0 siblings, 0 replies; 7+ messages in thread
From: Keith Seitz @ 2004-10-03  3:43 UTC (permalink / raw)
  To: Kip Warner; +Cc: insight

On Sat, 2004-10-02 at 20:12, Kip Warner wrote:
> I have tried using two different MinGW builds of Insight only to find that  
> neither can view source mode of my GCC produced executables. All I can see  
> is assembly listing. I am compiling with -g3, full debugging information  
> and I am not stripping them with the linker. Any ideas?

As far as I know, it is not possible to compile insight with Mingw. You
must have a cygwin build (or something with a bunch of patches outside
the normal distribution).

All that aside, if you type "info func main" in the console window, what
does it show? Can you type "list main"? I am guessing that either gdb or
insight cannot deal with the non-POSIX paths that mingw-gcc outputs.

Keith

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Insight Bug?
@ 2004-10-03  3:12 Kip Warner
  2004-10-03  3:43 ` Keith Seitz
  0 siblings, 1 reply; 7+ messages in thread
From: Kip Warner @ 2004-10-03  3:12 UTC (permalink / raw)
  To: insight

Greetings,

I have tried using two different MinGW builds of Insight only to find that  
neither can view source mode of my GCC produced executables. All I can see  
is assembly listing. I am compiling with -g3, full debugging information  
and I am not stripping them with the linker. Any ideas?

-- 
Kip Warner
Vertigo
http://TheVertigo.com

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2004-10-10 19:40 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2000-05-29  6:32 Insight Bug? Paul_Heijman
2000-05-29  8:12 ` Fernando Nasser
2000-05-30  5:48 ` Finish [was: Insight Bug?] Fernando Nasser
2004-10-03  3:12 Insight Bug? Kip Warner
2004-10-03  3:43 ` Keith Seitz
     [not found] <opsfi9s0h3e8k94p@mail.thevertigo.com>
2004-10-08 14:29 ` Keith Seitz
     [not found]   ` <opsfn5rmzge8k94p@mail.thevertigo.com>
2004-10-10 19:40     ` Keith Seitz

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).