Warning: Couldn't restore frame in current thread

Warning: Couldn't restore frame in current thread

[Holger Vogt]

Hi,

after upgrading Insight from CYGWIN distribution (20001125-1), I get
a warning message box with

"Couldn't restore frame in current thread, at frame 0"

every time I hit Step(S) or Next(N) in the Source Window.

What should I do to avoid this message?

Holger

Re: Warning: Couldn't restore frame in current thread

[Keith Seitz]

On Tue, 1 May 2001, Holger Vogt wrote:

> after upgrading Insight from CYGWIN distribution (20001125-1), I get
> a warning message box with
>
> "Couldn't restore frame in current thread, at frame 0"
>
> every time I hit Step(S) or Next(N) in the Source Window.

Two things:

If you run "gdb -nw" (i.e., without Insight), I'll presume that the
problem still exists. Please verify and send output of this happening to
gdb@sources.redhat.com.

Keith

Re: Warning: Couldn't restore frame in current thread

[Holger Vogt]

gdb -nw is working without problems.

In Insight either typing n, s, si to the console window or mouse-clicking to
the source window will generate the warning message. It appears for the
first time  after typing 'run' to the console window.

Holger



Keith Seitz schrieb:

> On Tue, 1 May 2001, Holger Vogt wrote:
>
> > after upgrading Insight from CYGWIN distribution (20001125-1), I get
> > a warning message box with
> >
> > "Couldn't restore frame in current thread, at frame 0"
> >
> > every time I hit Step(S) or Next(N) in the Source Window.
>
> Two things:
>
> If you run "gdb -nw" (i.e., without Insight), I'll presume that the
> problem still exists. Please verify and send output of this happening to
> gdb@sources.redhat.com.
>
> Keith

Re: Warning: Couldn't restore frame in current thread

[Keith Seitz]

On Wed, 2 May 2001, Holger Vogt wrote:

> gdb -nw is working without problems.
>
> In Insight either typing n, s, si to the console window or mouse-clicking to
> the source window will generate the warning message. It appears for the
> first time  after typing 'run' to the console window.

Wow, that is weird. The warning message you are displaying comes from gdb.
I would not expect Insight to show this message at all. Something fishy
must be going on...

What does gdb -v show? (If you're on cygwin, can you include the output of
"gcc -v" and "uname -a", too?)

Keith

Re: Warning: Couldn't restore frame in current thread

[Fernando Nasser]

Keith Seitz wrote:
> 
> On Wed, 2 May 2001, Holger Vogt wrote:
> 
> > gdb -nw is working without problems.
> >
> > In Insight either typing n, s, si to the console window or mouse-clicking to
> > the source window will generate the warning message. It appears for the
> > first time  after typing 'run' to the console window.
> 
> Wow, that is weird. The warning message you are displaying comes from gdb.
> I would not expect Insight to show this message at all. Something fishy
> must be going on...
> 
> What does gdb -v show? (If you're on cygwin, can you include the output of
> "gcc -v" and "uname -a", too?)
> 

Don't worry.  I have a patch already to cope with the change in the way
the type is created for (void *).  It is backward compatible, so we can
just make the change.

I decided to allow for (char *) to be dereferenced now, as this works
fine (like in the command line interface).  Only (void *) cannot be
dereferenced.

Please try the attached patch.


I still wonder why (void *) was tested used the name field and not the
type_code and whether the change (i.e., the removal of the name field of
VOID types) was intentional or a bug. 

For the ones that did not follow the thread, TYPE_CODE_VOID types now
have the name field empty (i.e. TYPE_NAME == NULL) -- previously it was
"void".  At least when they are the target_type of a TYPE_CODE_PTR type.


-- 
Fernando Nasser
Red Hat Canada Ltd.                     E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9
Index: varobj.c
===================================================================
RCS file: /cvs/src/src/gdb/varobj.c,v
retrieving revision 1.17
diff -c -p -r1.17 varobj.c
*** varobj.c	2001/03/27 20:36:24	1.17
--- varobj.c	2001/05/02 02:35:16
*************** c_number_of_children (struct varobj *var
*** 1761,1767 ****
      case TYPE_CODE_PTR:
        /* This is where things get compilcated. All pointers have one child.
           Except, of course, for struct and union ptr, which we automagically
!          dereference for the user and function ptrs, which have no children. */
        switch (TYPE_CODE (target))
  	{
  	case TYPE_CODE_STRUCT:
--- 1761,1773 ----
      case TYPE_CODE_PTR:
        /* This is where things get compilcated. All pointers have one child.
           Except, of course, for struct and union ptr, which we automagically
!          dereference for the user and function ptrs, which have no children.
!          We also don't dereference void* as we don't know what to show.
!          We can show char* so we allow it to be dereferenced.  If you decide
!          to test for it, please mind that a little magic is necessary to
!          properly identify it: char* has TYPE_CODE == TYPE_CODE_INT and 
!          TYPE_NAME == "char" */
! 
        switch (TYPE_CODE (target))
  	{
  	case TYPE_CODE_STRUCT:
*************** c_number_of_children (struct varobj *var
*** 1770,1786 ****
  	  break;
  
  	case TYPE_CODE_FUNC:
  	  children = 0;
  	  break;
  
  	default:
! 	  /* Don't dereference char* or void*. */
! 	  if (TYPE_NAME (target) != NULL
! 	      && (STREQ (TYPE_NAME (target), "char")
! 		  || STREQ (TYPE_NAME (target), "void")))
! 	    children = 0;
! 	  else
! 	    children = 1;
  	}
        break;
  
--- 1776,1787 ----
  	  break;
  
  	case TYPE_CODE_FUNC:
+ 	case TYPE_CODE_VOID:
  	  children = 0;
  	  break;
  
  	default:
! 	  children = 1;
  	}
        break;
  

Re: Warning: Couldn't restore frame in current thread

[Holger Vogt]

The messages requested are:

$ gcc -v
Reading specs from /bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-4/specs
gcc version 2.95.3-4 (cygwin special)

$ uname -a
CYGWIN_98-4.10 A7W7R0 1.3.1(0.38/3/2) 2001-04-24 20:01 i586 unknown

My insight history is:

downloaded sources insight-5_0_tar.bz2 at 07/23/00
compiled with CYGWIN version actual at that time out of the box,
running configure, make, make install: absolutely no problems!

upgraded CYGWIN form time to time, also including gdb, by setup.exe, only
binaries. Actual gdb: gdb-20001125-1.tar. Warning message appeared around that
time ?!

Removed the old insight installation from /usr/local: still warning message.

Recompiled the July sources with CC='gcc -mwin32' configure, moved only gdb.exe to
/bin: gdb -nw is o.k., gdb exits without displaying any window.

Is the email from Fernando Nasser directly related to my problem? Where could I
find the actual sources to apply this patch?

Holger




Keith Seitz schrieb:

> On Wed, 2 May 2001, Holger Vogt wrote:
>
> > gdb -nw is working without problems.
> >
> > In Insight either typing n, s, si to the console window or mouse-clicking to
> > the source window will generate the warning message. It appears for the
> > first time  after typing 'run' to the console window.
>
> Wow, that is weird. The warning message you are displaying comes from gdb.
> I would not expect Insight to show this message at all. Something fishy
> must be going on...
>
> What does gdb -v show? (If you're on cygwin, can you include the output of
> "gcc -v" and "uname -a", too?)
>
> Keith

Re: Warning: Couldn't restore frame in current thread

[Keith Seitz]

On Thu, 3 May 2001, Holger Vogt wrote:

> The messages requested are:
>
> $ gcc -v
> Reading specs from /bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-4/specs
> gcc version 2.95.3-4 (cygwin special)
>
> $ uname -a
> CYGWIN_98-4.10 A7W7R0 1.3.1(0.38/3/2) 2001-04-24 20:01 i586 unknown

You are using newer versions of cygwin and gcc than I am, but I don't
think that is the problem...

> Recompiled the July sources with CC='gcc -mwin32' configure, moved only gdb.exe to
> /bin: gdb -nw is o.k., gdb exits without displaying any window.

I would graqb the snapshot and rebuild and reinstall the whole thing.
Development on Insight is beginning to pick up, and I think that you may
be running into an imcompatibility between the binaries and the tcl code.
(Have you tried reinstalling ALL of insight instead of just copying the
exe?)

> Is the email from Fernando Nasser directly related to my problem? Where could I
> find the actual sources to apply this patch?

Actually, I have no idea what Fernando is talking about. I would wait for
a response from him...

Keith

Re: Warning: Couldn't restore frame in current thread

[Holger Vogt]

Hi,

the problem is identified:

when I issue the command "info thread" in the console window, the warning message
appears (only once, when I use this command again without leaving the console window, it
will not show up a second time). If I leave the console window and enter it again, it
will again appear.

Some time ago I had opened the window "Processes" (via "View" and "Thread list" in the
Source window), which has the effect that every time the process is stepped further, the
warning message appears.

Close the Processes Window, and the warning message is gone.


I have downloaded the recent snapshot. Compilation stopped with an error and the
following message:

gcc -c  -I. -I. -I/cygdrive/g/insight+dejagnu-20010503/tcl/generic -I/cygdrive/g
/insight+dejagnu-20010503/tk/generic -I/cygdrive/g/insight+dejagnu-20010503/tk/x
lib -DEXP_VERSION=\"5.26\" -DSCRIPTDIR=\"/usr/local/share\" -DEXECSCRIPTDIR=\"/u
sr/local/lib\" -DTCL_DEBUGGER -g -O2  -DCYGWIN_ALTTCL -DDFLT_STTY="\"sane\""  ex
p_command.c
In file included from /usr/include/w32api/windef.h:145,
                 from /usr/include/w32api/windows.h:98,
                 from exp_command.c:626:
/usr/include/w32api/winnt.h:48: parse error before `CCHAR'
/usr/include/w32api/winnt.h:48: warning: data definition has no type or storage
class
/usr/include/w32api/winnt.h:68: parse error before `*'
/usr/include/w32api/winnt.h:68: warning: data definition has no type or storage
class
/usr/include/w32api/winnt.h:69: parse error before `*'
/usr/include/w32api/winnt.h:69: warning: data definition has no type or storage
class
/usr/include/w32api/winnt.h:79: parse error before `TCHAR'
/usr/include/w32api/winnt.h:79: warning: data definition has no type or storage
class
/usr/include/w32api/winnt.h:82: parse error before `TBYTE'
/usr/include/w32api/winnt.h:82: warning: data definition has no type or storage
class
/usr/include/w32api/winnt.h:83: parse error before `*'
/usr/include/w32api/winnt.h:83: warning: data definition has no type or storage
class
/usr/include/w32api/winnt.h:84: parse error before `*'
/usr/include/w32api/winnt.h:84: warning: data definition has no type or storage
class
...
...
many more error messages


Adding a line
#undef VOID
as line 41 in winnt.h solved the problem (probably not a useful general patch!?). This
also happened with the July 2000 Insight sources, so it may be a compatibility problem
with the new CYGWIN.

Holger




Keith Seitz schrieb:

> On Thu, 3 May 2001, Holger Vogt wrote:
>
> > The messages requested are:
> >
> > $ gcc -v
> > Reading specs from /bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-4/specs
> > gcc version 2.95.3-4 (cygwin special)
> >
> > $ uname -a
> > CYGWIN_98-4.10 A7W7R0 1.3.1(0.38/3/2) 2001-04-24 20:01 i586 unknown
>
> You are using newer versions of cygwin and gcc than I am, but I don't
> think that is the problem...
>
> > Recompiled the July sources with CC='gcc -mwin32' configure, moved only gdb.exe to
> > /bin: gdb -nw is o.k., gdb exits without displaying any window.
>
> I would graqb the snapshot and rebuild and reinstall the whole thing.
> Development on Insight is beginning to pick up, and I think that you may
> be running into an imcompatibility between the binaries and the tcl code.
> (Have you tried reinstalling ALL of insight instead of just copying the
> exe?)
>
> > Is the email from Fernando Nasser directly related to my problem? Where could I
> > find the actual sources to apply this patch?
>
> Actually, I have no idea what Fernando is talking about. I would wait for
> a response from him...
>
> Keith

CYGWIN 1.3.1 compile error

[Holger Vogt]

Hi,

I have downloaded the recent snapshot. Compilation stopped with an error
and the following message:

gcc -c  -I. -I. -I/cygdrive/g/insight+dejagnu-20010503/tcl/generic
-I/cygdrive/g
/insight+dejagnu-20010503/tk/generic
-I/cygdrive/g/insight+dejagnu-20010503/tk/x
lib -DEXP_VERSION=\"5.26\" -DSCRIPTDIR=\"/usr/local/share\"
-DEXECSCRIPTDIR=\"/u
sr/local/lib\" -DTCL_DEBUGGER -g -O2  -DCYGWIN_ALTTCL
-DDFLT_STTY="\"sane\""  exp_command.c
In file included from /usr/include/w32api/windef.h:145,
                 from /usr/include/w32api/windows.h:98,
                 from exp_command.c:626:
/usr/include/w32api/winnt.h:48: parse error before `CCHAR'
/usr/include/w32api/winnt.h:48: warning: data definition has no type or
storage
class
/usr/include/w32api/winnt.h:68: parse error before `*'
/usr/include/w32api/winnt.h:68: warning: data definition has no type or
storage
class
...
...
many more error messages

I use the recent gcc:
$ gcc -v
Reading specs from /bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-4/specs
gcc version 2.95.3-4 (cygwin special)
$ uname -a
CYGWIN_98-4.10 A7W7R0 1.3.1(0.38/3/2) 2001-04-24 20:01 i586 unknown


Adding a line #undef VOID   to exp_command.c removed the error.

$ diff exp_command.c exp_command.c.bak
626d625
< #undef VOID

This also happened with the July 2000 Insight sources, so it may be a
compatibility problem with the new CYGWIN.

A similar error had been reported in
http://sources.redhat.com/ml/insight-prs/2001-q2/msg00000.html
I could not identify the patch there, which might be different

Holger

Re: CYGWIN 1.3.1 compile error

[Christopher Faylor]

On Sun, May 06, 2001 at 02:33:33PM +0200, Holger Vogt wrote:
>I have downloaded the recent snapshot. Compilation stopped with an error
>and the following message:
>
>gcc -c  -I. -I. -I/cygdrive/g/insight+dejagnu-20010503/tcl/generic
>-I/cygdrive/g
>/insight+dejagnu-20010503/tk/generic
>-I/cygdrive/g/insight+dejagnu-20010503/tk/x
>lib -DEXP_VERSION=\"5.26\" -DSCRIPTDIR=\"/usr/local/share\"
>-DEXECSCRIPTDIR=\"/u
>sr/local/lib\" -DTCL_DEBUGGER -g -O2  -DCYGWIN_ALTTCL
>-DDFLT_STTY="\"sane\""  exp_command.c
>In file included from /usr/include/w32api/windef.h:145,
>                 from /usr/include/w32api/windows.h:98,
>                 from exp_command.c:626:
>/usr/include/w32api/winnt.h:48: parse error before `CCHAR'
>/usr/include/w32api/winnt.h:48: warning: data definition has no type or
>storage
>class
>/usr/include/w32api/winnt.h:68: parse error before `*'
>/usr/include/w32api/winnt.h:68: warning: data definition has no type or
>storage
>class
>...
>...
>many more error messages
>
>I use the recent gcc:
>$ gcc -v
>Reading specs from /bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-4/specs
>gcc version 2.95.3-4 (cygwin special)
>$ uname -a
>CYGWIN_98-4.10 A7W7R0 1.3.1(0.38/3/2) 2001-04-24 20:01 i586 unknown
>
>
>Adding a line #undef VOID   to exp_command.c removed the error.
>
>$ diff exp_command.c exp_command.c.bak
>626d625
>< #undef VOID
>
>This also happened with the July 2000 Insight sources, so it may be a
>compatibility problem with the new CYGWIN.

Nope.  I'm the cygwin maintainer.  I created the gdb distribution
and the gcc distribution.  Guess what I use to create gdb?

I assume that you are using older insight/gdb sources.  You might
want to search the mailing list archives for the last 427 times this
has come up.  The short answer is upgrade your sources, or if this
is impossible for some inexplicable reason, configure using:

CC='gcc -mwin32' configure

cgf

Re: Warning: Couldn't restore frame in current thread

[Fernando Nasser]

Holger Vogt wrote:
> 
> > > Is the email from Fernando Nasser directly related to my problem? 

Nope!  Sorry, I've replied to the wrong message.

(That patch fixes the gdb/51 PR.  It applies to the current CVS
sources.)

-- 
Fernando Nasser
Red Hat Canada Ltd.                     E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9