FW: gdb seems to be broken on darwin 10.5.8 and later

FW: gdb seems to be broken on darwin 10.5.8 and later

[Anirban Sinha]

>Could you guys please rename this sub-thread to something about darwin?
>Thanks...

Done.


>
>Jack Howarth wrote:
>> On Wed, Sep 30, 2009 at 10:42:12AM -0700, Paul Pluzhnikov wrote:
>>> On Wed, Sep 30, 2009 at 10:31 AM, Jack Howarth
<howarth@bromo.med.uc.edu>
>wrote:
>>>
>>>> I assume 10.5.8 and Snow Leopard share the same problem. I used to
>>>> see the error...
>>> Quite possible. I don't have Snow Leopard, so I don't know what
happens
>>> there.
>>>
>>
>> I have a spare drive with Leopard (10.5.8) on it so I can double
>> check that gdb cvs shows the same behavior.
>>
>>> The kernel denies GDB access to debug port; I don't see what this
would
>>> have to do with Xcode.
>>
>> I was just wondering if the behavior was any different from Xcode
3.1.2
>> to Xcode 3.1.3.

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Paul Pluzhnikov]

FWIW, I have just built GDB from CVS Head on

System Version:	Mac OS X 10.5.8 (9L31a)
9.8.0 Darwin Kernel Version 9.8.0: Wed Jul 15 16:55:01 PDT 2009;
root:xnu-1228.15.4~1/RELEASE_I386 i386

and GDB appears to work just fine (when executed as root),
except for minor annoyances:

# ../gdb ../gdb
GNU gdb (GDB) 7.0.50.20091001-cvs
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=i386-apple-darwin9.7.0
--target=i386-apple-darwin9.8.0".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /Users/ppluzhnikov/gdb-cvs/build/gdb/gdb...
warning: section __DATA.__common not found in main.o

warning: section __DATA.__common not found in utils.o

warning: section __DATA.__common not found in exec.o

... many more of these ...
warning: section __DATA.__common not found in plural-exp.o
(no debugging symbols found)...done.
(gdb) run
`main.o' has disappeared; keeping its symbols.
`make-relative-prefix.o' has disappeared; keeping its symbols.
`utils.o' has disappeared; keeping its symbols.
... many more ....
`stab-syms.o' has disappeared; keeping its symbols.
`hex.o' has disappeared; keeping its symbols.
Starting program: /Users/ppluzhnikov/gdb-cvs/build/gdb/gdb
GNU gdb (GDB) 7.0.50.20091001-cvs
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=i386-apple-darwin9.7.0
--target=i386-apple-darwin9.8.0".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
(gdb)
Program received signal SIGINT, Interrupt.
0x91cdc6fa in select$DARWIN_EXTSN () from /usr/lib/libSystem.B.dylib
(gdb) bt
#0  0x91cdc6fa in select$DARWIN_EXTSN () from /usr/lib/libSystem.B.dylib
#1  0x0006765a in gdb_select (n=1, readfds=0x354ecc,
writefds=0x354f4c, exceptfds=0x354fcc, timeout=0x0) at
../../src/gdb/posix-hdep.c:52
#2  0x00101d98 in gdb_wait_for_event (block=1) at ../../src/gdb/event-loop.c:877
#3  0x001011ac in gdb_do_one_event (data=0x0) at ../../src/gdb/event-loop.c:455
#4  0x000f99b3 in catch_errors (func=0x1010e3 <gdb_do_one_event>,
func_args=0x0, errstring=0x29f7d0 "", mask=6) at
../../src/gdb/exceptions.c:510
#5  0x000542bf in tui_command_loop (data=0x0) at
../../src/gdb/tui/tui-interp.c:153
#6  0x000fa3b8 in current_interp_command_loop () at ../../src/gdb/interps.c:291
#7  0x000fad93 in captured_command_loop (data=0x0) at ../../src/gdb/main.c:226
#8  0x000f99b3 in catch_errors (func=0xfad81 <captured_command_loop>,
func_args=0x0, errstring=0x29f7d0 "", mask=6) at
../../src/gdb/exceptions.c:510
#9  0x000fc271 in captured_main (data=0xbffffcb0) at ../../src/gdb/main.c:902
#10 0x000f99b3 in catch_errors (func=0xfadd6 <captured_main>,
func_args=0xbffffcb0, errstring=0x29f7d0 "", mask=6) at
../../src/gdb/exceptions.c:510
#11 0x000fc318 in gdb_main (args=0xbffffcb0) at ../../src/gdb/main.c:911
#12 0x00001cc4 in main (argc=1, argv=0xbffffce8) at ../../src/gdb/gdb.c:33
(gdb) quit


--
Paul Pluzhnikov

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Jack Howarth]

On Wed, Sep 30, 2009 at 08:34:21PM -0700, Paul Pluzhnikov wrote:
> FWIW, I have just built GDB from CVS Head on
> 
> System Version:	Mac OS X 10.5.8 (9L31a)
> 9.8.0 Darwin Kernel Version 9.8.0: Wed Jul 15 16:55:01 PDT 2009;
> root:xnu-1228.15.4~1/RELEASE_I386 i386
> 
> and GDB appears to work just fine (when executed as root),
> except for minor annoyances:
> 
> # ../gdb ../gdb
> GNU gdb (GDB) 7.0.50.20091001-cvs
> Copyright (C) 2009 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
> and "show warranty" for details.
> This GDB was configured as "--host=i386-apple-darwin9.7.0
> --target=i386-apple-darwin9.8.0".
> For bug reporting instructions, please see:
> <http://www.gnu.org/software/gdb/bugs/>...
> Reading symbols from /Users/ppluzhnikov/gdb-cvs/build/gdb/gdb...
> warning: section __DATA.__common not found in main.o
> 
> warning: section __DATA.__common not found in utils.o
> 
> warning: section __DATA.__common not found in exec.o
> 
> ... many more of these ...
> warning: section __DATA.__common not found in plural-exp.o
> (no debugging symbols found)...done.
> (gdb) run
> `main.o' has disappeared; keeping its symbols.
> `make-relative-prefix.o' has disappeared; keeping its symbols.
> `utils.o' has disappeared; keeping its symbols.
> ... many more ....
> `stab-syms.o' has disappeared; keeping its symbols.
> `hex.o' has disappeared; keeping its symbols.
> Starting program: /Users/ppluzhnikov/gdb-cvs/build/gdb/gdb
> GNU gdb (GDB) 7.0.50.20091001-cvs
> Copyright (C) 2009 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
> and "show warranty" for details.
> This GDB was configured as "--host=i386-apple-darwin9.7.0
> --target=i386-apple-darwin9.8.0".
> For bug reporting instructions, please see:
> <http://www.gnu.org/software/gdb/bugs/>.
> (gdb)
> Program received signal SIGINT, Interrupt.
> 0x91cdc6fa in select$DARWIN_EXTSN () from /usr/lib/libSystem.B.dylib
> (gdb) bt
> #0  0x91cdc6fa in select$DARWIN_EXTSN () from /usr/lib/libSystem.B.dylib
> #1  0x0006765a in gdb_select (n=1, readfds=0x354ecc,
> writefds=0x354f4c, exceptfds=0x354fcc, timeout=0x0) at
> ../../src/gdb/posix-hdep.c:52
> #2  0x00101d98 in gdb_wait_for_event (block=1) at ../../src/gdb/event-loop.c:877
> #3  0x001011ac in gdb_do_one_event (data=0x0) at ../../src/gdb/event-loop.c:455
> #4  0x000f99b3 in catch_errors (func=0x1010e3 <gdb_do_one_event>,
> func_args=0x0, errstring=0x29f7d0 "", mask=6) at
> ../../src/gdb/exceptions.c:510
> #5  0x000542bf in tui_command_loop (data=0x0) at
> ../../src/gdb/tui/tui-interp.c:153
> #6  0x000fa3b8 in current_interp_command_loop () at ../../src/gdb/interps.c:291
> #7  0x000fad93 in captured_command_loop (data=0x0) at ../../src/gdb/main.c:226
> #8  0x000f99b3 in catch_errors (func=0xfad81 <captured_command_loop>,
> func_args=0x0, errstring=0x29f7d0 "", mask=6) at
> ../../src/gdb/exceptions.c:510
> #9  0x000fc271 in captured_main (data=0xbffffcb0) at ../../src/gdb/main.c:902
> #10 0x000f99b3 in catch_errors (func=0xfadd6 <captured_main>,
> func_args=0xbffffcb0, errstring=0x29f7d0 "", mask=6) at
> ../../src/gdb/exceptions.c:510
> #11 0x000fc318 in gdb_main (args=0xbffffcb0) at ../../src/gdb/main.c:911
> #12 0x00001cc4 in main (argc=1, argv=0xbffffce8) at ../../src/gdb/gdb.c:33
> (gdb) quit
> 
> 
> --
> Paul Pluzhnikov

Paul,
   Actually current gdb cvs with the macho changes is now behaving fine as long
as I run as root. Is there any chance that we can get the macho changes into
gdb cvs since the remaining issues with permissions seem to be completely
orthogonal to those changes.
                  Jack

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Paul Pluzhnikov]

On Thu, Oct 1, 2009 at 6:55 PM, Jack Howarth <howarth@bromo.med.uc.edu> wrote:

> Is there any chance that we can get the macho changes into
> gdb cvs since the remaining issues with permissions seem to be completely
> orthogonal to those changes.

Here are the last messages on this patch:
http://sourceware.org/ml/gdb-patches/2009-09/msg00430.html

I am not sure how Joel wants to handle this WRT the 7.0 branch; none
of the other maintainers chimed in :-(


-- 
Paul Pluzhnikov

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Joel Brobecker]

> I am not sure how Joel wants to handle this WRT the 7.0 branch; none
> of the other maintainers chimed in :-(

Did Jack confirm that it does not work without applying your patch?
So far, I said that AdaCore was going to use that patch for sure,
just to be on the safe side, but that I can live without it in the
FSF tree if no one else is interested in it. If GDB is completely
broken without it, then I think we have to include it in the FSF
sources too.

We need confirmation from Jack pronto, because the 7.0 release is
approaching really fast.

-- 
Joel

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Tom Tromey]

>>>>> "Joel" == Joel Brobecker <brobecker@adacore.com> writes:

>> I am not sure how Joel wants to handle this WRT the 7.0 branch; none
>> of the other maintainers chimed in :-(

Joel> Did Jack confirm that it does not work without applying your patch?

I read the referenced email.  If the only issue with the patch is that
it is not the best possible way to fix the problem, then I think it
might as well go in.  I say that because I think it generally preferable
to commit a functioning patch rather than wait for a more ideal patch.

Tom

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Jack Howarth]

On Fri, Oct 02, 2009 at 10:00:28AM -0600, Tom Tromey wrote:
> >>>>> "Joel" == Joel Brobecker <brobecker@adacore.com> writes:
> 
> >> I am not sure how Joel wants to handle this WRT the 7.0 branch; none
> >> of the other maintainers chimed in :-(
> 
> Joel> Did Jack confirm that it does not work without applying your patch?
> 
> I read the referenced email.  If the only issue with the patch is that
> it is not the best possible way to fix the problem, then I think it
> might as well go in.  I say that because I think it generally preferable
> to commit a functioning patch rather than wait for a more ideal patch.
> 
> Tom

Tom,
   I agree since a functional gdb (even if only as root) will encourage more
folks to work the remaining issues. It also allows darwin users to test gdb
against the new debug support in gcc trunk.
               Jack

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Joel Brobecker]

> >> I am not sure how Joel wants to handle this WRT the 7.0 branch; none
> >> of the other maintainers chimed in :-(
> 
> Joel> Did Jack confirm that it does not work without applying your patch?
> 
> I read the referenced email.  If the only issue with the patch is that
> it is not the best possible way to fix the problem, then I think it
> might as well go in.  I say that because I think it generally preferable
> to commit a functioning patch rather than wait for a more ideal patch.

Jack came back to me a couple of days ago, and confirmed that the debugger
apperas to work even without the patch. We can try to do without, at least
until we confirm that it really does help.

-- 
Joel

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Jack Howarth]

On Sun, Oct 04, 2009 at 07:57:00PM -0700, Joel Brobecker wrote:
> > >> I am not sure how Joel wants to handle this WRT the 7.0 branch; none
> > >> of the other maintainers chimed in :-(
> > 
> > Joel> Did Jack confirm that it does not work without applying your patch?
> > 
> > I read the referenced email.  If the only issue with the patch is that
> > it is not the best possible way to fix the problem, then I think it
> > might as well go in.  I say that because I think it generally preferable
> > to commit a functioning patch rather than wait for a more ideal patch.
> 
> Jack came back to me a couple of days ago, and confirmed that the debugger
> apperas to work even without the patch. We can try to do without, at least
> until we confirm that it really does help.
> 
> -- 
> Joel

Joel,
  I ran the make check on gdb cvs last night (without an ada compiler
installed) and the test results seemed reasonable. Does gdb have anything
like the contrib/test_summary script from gcc so that a shorter summary
can be generated for the gdb-testers mailing list? I thought I would
post some full test results in case any particular failures jumped out
as being low hanging fruit for fixing.
               Jack

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Joel Brobecker]

> I ran the make check on gdb cvs last night (without an ada compiler
> installed) and the test results seemed reasonable. Does gdb have anything
> like the contrib/test_summary script from gcc so that a shorter summary
> can be generated for the gdb-testers mailing list? I thought I would
> post some full test results in case any particular failures jumped out
> as being low hanging fruit for fixing.

Unfortunately not. People usually post a gzip'ed version of the gdb.sum
file. The gdb.log file might be too big to post, I don't know. You look
very interested in gdb on MacOS, perhaps you should consider helping us
fix the issues you're finding, because there aren't many people who are.
We can help you through the learning curve, if you have questions
(join us at #gdb on irc.freenode.net).

-- 
Joel

Re: FW: gdb seems to be broken on darwin 10.5.8 and later

[Ulrich Weigand]

Jack Howarth wrote:

>   I ran the make check on gdb cvs last night (without an ada compiler
> installed) and the test results seemed reasonable. Does gdb have anything
> like the contrib/test_summary script from gcc so that a shorter summary
> can be generated for the gdb-testers mailing list? I thought I would
> post some full test results in case any particular failures jumped out
> as being low hanging fruit for fixing.

You can simply use GCC's test_summary script; this works just fine
for GDB as well ...

Bye,
Ulrich


-- 
  Dr. Ulrich Weigand
  GNU Toolchain for Linux on System z and Cell BE
  Ulrich.Weigand@de.ibm.com