[RFA] Darwin/x86 port (v4 - part 0/4)

[RFA] Darwin/x86 port (v4 - part 0/4)

[Tristan Gingold]

Hi,

here is a new version (v4) of the Darwin/x86 port.
I think I have addressed all comments:

* The typos (missing '>', "to detach from", "the one", "mach-o")
* Dead code has been removed or enabled on debug option
* No more direct calls to printf
* User messages are enclosed in _()
* A new section in the GDB user manual has been added to document  
darwin specific commands
* NEWS entry added.

(Thanks to Eli for his comments).

Waiting for Stan's review :-)

Tristan.


gdb/
2008-11-14  Tristan Gingold  <gingold@adacore.com>

	* NEWS: Add entry for new native configuration: Darwin.
	* configure.host: Add Darwin host.
	* configure.tgt: Add Darwin target.
	* defs.h (enum gdb_osabi): Add GDB_OSABI_DARWIN.
	* osabi.c (gdb_osabi_names): Add name for Darwin abi.
	* i386-darwin-nat.c: New file.
	* i386-darwin-tdep.c: New file.
	* i386-darwin-tdep.h: New file.
	* machoread.c: New file.
	* darwin-nat-info.c: New file.
	* darwin-nat.c: New file.
	* darwin-nat.h: New file.
	* darwin.defs: New file.
	* config/i386/darwin.mh: New file.

gdb/doc/
2008-11-14  Tristan Gingold  <gingold@adacore.com>

	* gdb.texinfo (Darwin): Document Darwin specific features.

diff -c -r1.292 NEWS
*** gdb/NEWS	22 Oct 2008 22:14:21 -0000	1.292
--- gdb/NEWS	14 Nov 2008 16:17:20 -0000
***************
*** 169,174 ****
--- 169,178 ----
     These allow macros to be defined, undefined, and listed
     interactively.

+ * New native configurations
+
+ x86 Darwin			i[34567]86-*-darwin*
+
   * New targets

   x86 DICOS			i[34567]86-*-dicos*
diff -c -r1.101 configure.host
*** gdb/configure.host	19 Jan 2008 15:03:50 -0000	1.101
--- gdb/configure.host	14 Nov 2008 16:17:20 -0000
***************
*** 62,67 ****
--- 62,69 ----

   case "${host}" in

+ *-*-darwin*)		gdb_host=darwin ;;
+
   alpha*-*-osf[3456789]*)	gdb_host=alpha-osf3 ;;
   alpha*-*-linux*)	gdb_host=alpha-linux ;;
   alpha*-*-freebsd* | alpha*-*-kfreebsd*-gnu)
diff -c -r1.204 configure.tgt
*** gdb/configure.tgt	2 Oct 2008 15:48:06 -0000	1.204
--- gdb/configure.tgt	14 Nov 2008 16:17:20 -0000
***************
*** 145,150 ****
--- 145,155 ----
   	gdb_target_obs="hppa-tdep.o"
   	;;

+ i[34567]86-*-darwin*)
+ 	# Target: Darwin/i386
+ 	gdb_target_obs="amd64-tdep.o i386-tdep.o i387-tdep.o \
+ 			i386-darwin-tdep.o"
+ 	;;
   i[34567]86-*-dicos*)
   	# Target: DICOS/i386
   	gdb_target_obs="i386-tdep.o i387-tdep.o \
diff -c -r1.241 defs.h
*** gdb/defs.h	5 Nov 2008 20:23:07 -0000	1.241
--- gdb/defs.h	14 Nov 2008 16:17:20 -0000
***************
*** 957,962 ****
--- 957,963 ----
     GDB_OSABI_CYGWIN,
     GDB_OSABI_AIX,
     GDB_OSABI_DICOS,
+   GDB_OSABI_DARWIN,

     GDB_OSABI_INVALID		/* keep this last */
   };
diff -c -r1.43 osabi.c
*** gdb/osabi.c	1 May 2008 23:09:14 -0000	1.43
--- gdb/osabi.c	14 Nov 2008 16:17:21 -0000
***************
*** 73,78 ****
--- 73,79 ----
     "Cygwin",
     "AIX",
     "DICOS",
+   "Darwin",

     "<invalid>"
   };
diff -c -r1.532 gdb.texinfo
*** gdb/doc/gdb.texinfo	24 Oct 2008 21:04:22 -0000	1.532
--- gdb/doc/gdb.texinfo	14 Nov 2008 16:17:25 -0000
***************
*** 14675,14680 ****
--- 14675,14681 ----
   * Cygwin Native::		Features specific to the Cygwin port
   * Hurd Native::                 Features specific to @sc{gnu} Hurd
   * Neutrino::                    Features specific to QNX Neutrino
+ * Darwin::			Features specific to Darwin
   @end menu

   @node HP-UX
***************
*** 15456,15461 ****
--- 15457,15501 ----
   Show the current state of QNX Neutrino messages.
   @end table

+ @node Darwin
+ @subsection Darwin
+ @cindex Darwin
+
+ @value{GDBN} provides the following commands specific to the Darwin  
target:
+
+ @table @code
+ @item set debug inferior @var{num}
+ @kindex set debug inferior
+ When set to a non zero value, enables debugging messages specific to
+ the Darwin support.  The bigger the value is the more verbose the  
output is.
+
+ @item show debug inferior
+ @kindex show debug inferior
+ Show the current state of Darwin messages.
+
+ @item set debug mach-o @var{num}
+ @kindex set debug mach-o
+ When set to a non zero value, enables debugging messages specific to  
mach-o
+ object file processing.  The bigger the value is the more verbose the
+ output is.
+
+ @item show debug mach-o
+ @kindex show debug mach-o
+ Show the current state of mach-o file messages.
+
+ @item set mach-exceptions on
+ @itemx set mach-exceptions off
+ @kindex set mach-exceptions
+ Use this command to turn on trapping of exceptions in the inferior.   
The
+ default is off.  If on any inferior faults (but single stepping and
+ breakpoint) will be reported as a mach exception before being  
reported to
+ the kernel.
+
+ @item show mach-exceptions
+ @kindex show mach-exceptions
+ Show the current state of exceptions trapping.
+ @end table
+

   @node Embedded OS
   @section Embedded Operating Systems

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Pedro Alves]

On Friday 14 November 2008 16:26:23, Tristan Gingold wrote:
> + @table @code
> + @item set debug inferior @var{num}
> + @kindex set debug inferior
> + When set to a non zero value, enables debugging messages specific to
> + the Darwin support.  The bigger the value is the more verbose the  
> output is.

Please don't use "inferior" for darwin specific things.  It is a very
general term, and collides with what we may want to do with the
common gdb/inferior.c module.

-- 
Pedro Alves

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Tristan Gingold]

On Nov 14, 2008, at 5:41 PM, Pedro Alves wrote:

> On Friday 14 November 2008 16:26:23, Tristan Gingold wrote:
>> + @table @code
>> + @item set debug inferior @var{num}
>> + @kindex set debug inferior
>> + When set to a non zero value, enables debugging messages specific  
>> to
>> + the Darwin support.  The bigger the value is the more verbose the
>> output is.
>
> Please don't use "inferior" for darwin specific things.  It is a very
> general term, and collides with what we may want to do with the
> common gdb/inferior.c module.

Ok, will change to "maint darwin-debug"

Tristan.

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Stan Shebs]

Tristan Gingold wrote:
>
> On Nov 14, 2008, at 5:41 PM, Pedro Alves wrote:
>
>> On Friday 14 November 2008 16:26:23, Tristan Gingold wrote:
>>> + @table @code
>>> + @item set debug inferior @var{num}
>>> + @kindex set debug inferior
>>> + When set to a non zero value, enables debugging messages specific to
>>> + the Darwin support.  The bigger the value is the more verbose the
>>> output is.
>>
>> Please don't use "inferior" for darwin specific things.  It is a very
>> general term, and collides with what we may want to do with the
>> common gdb/inferior.c module.
>
> Ok, will change to "maint darwin-debug"
>
We have "set debug" that is generally available; some of the options are 
clearly maintenance-only, such as "set debug infrun", so that seems a 
convenient place for a Darwin-specific flag. (We could add "maint set 
debug", but that seems a little too pedantic :-) .)

Stan

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Pedro Alves]

On Friday 14 November 2008 17:47:03, Stan Shebs wrote:
> Tristan Gingold wrote:
> >
> > On Nov 14, 2008, at 5:41 PM, Pedro Alves wrote:
> >
> >> On Friday 14 November 2008 16:26:23, Tristan Gingold wrote:
> >>> + @table @code
> >>> + @item set debug inferior @var{num}
> >>> + @kindex set debug inferior
> >>> + When set to a non zero value, enables debugging messages specific to
> >>> + the Darwin support.  The bigger the value is the more verbose the
> >>> output is.
> >>
> >> Please don't use "inferior" for darwin specific things.  It is a very
> >> general term, and collides with what we may want to do with the
> >> common gdb/inferior.c module.
> >
> > Ok, will change to "maint darwin-debug"
> >
> We have "set debug" that is generally available; some of the options are 
> clearly maintenance-only, such as "set debug infrun", so that seems a 
> convenient place for a Darwin-specific flag. (We could add "maint set 
> debug", but that seems a little too pedantic :-) .)

I'm OK with "set debug", just not "set debug inferior".

"set debug darwin" would also be fine with me.

-- 
Pedro Alves

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Eli Zaretskii]

> From: Tristan Gingold <gingold@adacore.com>
> Date: Fri, 14 Nov 2008 17:26:23 +0100
> 
> here is a new version (v4) of the Darwin/x86 port.
> I think I have addressed all comments:

Thanks.

> *** gdb/NEWS	22 Oct 2008 22:14:21 -0000	1.292
> --- gdb/NEWS	14 Nov 2008 16:17:20 -0000
> ***************
> *** 169,174 ****
> --- 169,178 ----
>      These allow macros to be defined, undefined, and listed
>      interactively.
> 
> + * New native configurations
> +
> + x86 Darwin			i[34567]86-*-darwin*
> +

This is okay, but I thought you also support x86_64?

>   The bigger the value is the more verbose the output is.

Suggest to rephrase:

  "Higher values produce more verbose output."

If you agree, please make this change in all the other places you use
a similar phrase.

> + @item set debug mach-o @var{num}
> + @kindex set debug mach-o
> + When set to a non zero value, enables debugging messages specific to  
> mach-o
> + object file processing.  The bigger the value is the more verbose the

Do all Darwin developers know what is "mach-o object file processing"?
If not, perhaps a sentence or two of explanations are due.

> +                 If on any inferior faults (but single stepping and
> + breakpoint) will be reported as a mach exception before being reported to
> + the kernel.

 "If on, any inferior faults (except single stepping and breakpoint
  traps) will be announced as mach exceptions before being reported to
  the kernel."

To tell the truth, I don't really understand the meaning of the
mach-exceptions option, nor when it will be useful to GDB users.  Can
you explain?

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Daniel Jacobowitz]

On Fri, Nov 14, 2008 at 06:14:23PM +0000, Pedro Alves wrote:
> "set debug darwin" would also be fine with me.

That's what I'd suggest, too.  There's already a 'set debug lin-lwp'
which is similar; the name is just historical at this point.

-- 
Daniel Jacobowitz
CodeSourcery

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Tristan Gingold]

On Nov 14, 2008, at 7:42 PM, Eli Zaretskii wrote:
>> *** gdb/NEWS	22 Oct 2008 22:14:21 -0000	1.292
>> + * New native configurations
>> +
>> + x86 Darwin			i[34567]86-*-darwin*
>> +
>
> This is okay, but I thought you also support x86_64?

Right.  What about:

* New native configurations

x86/x86_64 Darwin		i[34567]86-*-darwin*


I realized that we don't support configuring as x86_64-*-darwin*.   
Maybe I should
add it ?

>>  The bigger the value is the more verbose the output is.
>
> Suggest to rephrase:
>
>  "Higher values produce more verbose output."
>
> If you agree, please make this change in all the other places you use
> a similar phrase.

Sure.  English is not my native language and I sometimes don't pick  
the right word.

>> + @item set debug mach-o @var{num}
>> + @kindex set debug mach-o
>> + When set to a non zero value, enables debugging messages specific  
>> to
>> mach-o
>> + object file processing.  The bigger the value is the more verbose  
>> the
>
> Do all Darwin developers know what is "mach-o object file processing"?
> If not, perhaps a sentence or two of explanations are due.

Ok.

>> +                 If on any inferior faults (but single stepping and
>> + breakpoint) will be reported as a mach exception before being  
>> reported to
>> + the kernel.
>
> "If on, any inferior faults (except single stepping and breakpoint
>  traps) will be announced as mach exceptions before being reported to
>  the kernel."
>
> To tell the truth, I don't really understand the meaning of the
> mach-exceptions option, nor when it will be useful to GDB users.  Can
> you explain?

I tried to explain better in the documentation.  Basically it is  
useful to understand the
reason of why a signal was sent.

Here is the updated doc patch.

Thank you for your comments,
Tristan.


+ @node Darwin
+ @subsection Darwin
+ @cindex Darwin
+
+ @value{GDBN} provides the following commands specific to the Darwin  
target:
+
+ @table @code
+ @item set debug darwin @var{num}
+ @kindex set debug darwin
+ When set to a non zero value, enables debugging messages specific to
+ the Darwin support.  Higher values produce more verbose output.
+
+ @item show debug darwin
+ @kindex show debug darwin
+ Show the current state of Darwin messages.
+
+ @item set debug mach-o @var{num}
+ @kindex set debug mach-o
+ When set to a non zero value, enables debugging messages while gdb is
+ reading Darwin object files.  The object and executable file format
+ for Darwin is named mach-o.  Higher values produce more verbose
+ output.
+
+ @item show debug mach-o
+ @kindex show debug mach-o
+ Show the current state of mach-o file messages.
+
+ @item set mach-exceptions on
+ @itemx set mach-exceptions off
+ @kindex set mach-exceptions
+ On Darwin, faults are first reported as a mach exception and then as a
+ signal.  Use this command to turn on trapping of mach exceptions in
+ the inferior.  This might be seldom useful to better understand the
+ cause of a fault. The default is off.
+
+ @item show mach-exceptions
+ @kindex show mach-exceptions
+ Show the current state of exceptions trapping.
+ @end table
+

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Tristan Gingold]

On Nov 17, 2008, at 12:36 AM, Daniel Jacobowitz wrote:

> On Fri, Nov 14, 2008 at 06:14:23PM +0000, Pedro Alves wrote:
>> "set debug darwin" would also be fine with me.
>
> That's what I'd suggest, too.  There's already a 'set debug lin-lwp'
> which is similar; the name is just historical at this point.

Fine with me.

Tristan.

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Eli Zaretskii]

> Cc: Eli Zaretskii <eliz@gnu.org>
> From: Tristan Gingold <gingold@adacore.com>
> Date: Mon, 17 Nov 2008 10:38:32 +0100
> 
> > This is okay, but I thought you also support x86_64?
> 
> Right.  What about:
> 
> * New native configurations
> 
> x86/x86_64 Darwin		i[34567]86-*-darwin*

Fine with me.

> I realized that we don't support configuring as x86_64-*-darwin*.
> Maybe I should add it ?

Probably, but I admit I don't know enough about the details.

> English is not my native language and I sometimes don't pick the
> right word.

Mine neither, so there's hope ;-)

> + When set to a non zero value, enables debugging messages while gdb is
> + reading Darwin object files.  The object and executable file format
> + for Darwin is named mach-o.

The last sentence above should probably be in parentheses, as that's a
minor note.  Also, I would rearrange slightly:

  (@dfn{Mach-o} is the file format used on Darwin for object and
  executable files.)

> + @kindex set mach-exceptions
> + On Darwin, faults are first reported as a mach exception and then as a
> + signal.  Use this command to turn on trapping of mach exceptions in
> + the inferior.  This might be seldom useful to better understand the
> + cause of a fault. The default is off.

So you are basically saying that mach exceptions are native Darwin
exceptions, and they are then mapped to Posix signals?  At least that
is what I understand from this explanation.

Also, please make sure you have 2 spaces after each period that ends a
sentence.

Thanks.

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Stan Shebs]

Tristan Gingold wrote:
> + @item set debug inferior @var{num}
We're changing this from "inferior" to "darwin", right?
> + When set to a non zero value, enables debugging messages specific to 
> mach-o
> + object file processing.  The bigger the value is the more verbose the
> + output is.
Capitalize "Mach-O" here. Also add a comment to the effect that this is 
mainly to diagnose problems internal to GDB, and should not be needed in 
normal usage.

With these changes, the bits here are OK for the trunk (and I'll be the 
first person to try them out once checked in!).

Stan

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Tristan Gingold]

Hi Stan,

thank you for the reviewing.  I have made the changes, the testsuite  
is now running.
I will post the new diffs soon.

I am replying to all the mail at least to double check I didn't forget  
anything.

Tristan.

On Nov 18, 2008, at 2:42 AM, Stan Shebs wrote:

> Tristan Gingold wrote:
>> + @item set debug inferior @var{num}
> We're changing this from "inferior" to "darwin", right?

Right.

>> + When set to a non zero value, enables debugging messages specific  
>> to mach-o
>> + object file processing.  The bigger the value is the more verbose  
>> the
>> + output is.
> Capitalize "Mach-O" here. Also add a comment to the effect that this  
> is mainly to diagnose problems internal to GDB, and should not be  
> needed in normal usage.

Done.

> With these changes, the bits here are OK for the trunk (and I'll be  
> the first person to try them out once checked in!).

Thank you for your interest!

Re: [RFA] Darwin/x86 port (v4 - part 0/4)

[Tristan Gingold]

Thank you again for your comments.

>> I realized that we don't support configuring as x86_64-*-darwin*.
>> Maybe I should add it ?
>
> Probably, but I admit I don't know enough about the details.

As this cannot be reported by config.guess, I won't add it now.  Could  
be of course
revisited later.

>> + When set to a non zero value, enables debugging messages while  
>> gdb is
>> + reading Darwin object files.  The object and executable file format
>> + for Darwin is named mach-o.
>
> The last sentence above should probably be in parentheses, as that's a
> minor note.  Also, I would rearrange slightly:
>
>  (@dfn{Mach-o} is the file format used on Darwin for object and
>  executable files.)

Ok.

>> + @kindex set mach-exceptions
>> + On Darwin, faults are first reported as a mach exception and then  
>> as a
>> + signal.  Use this command to turn on trapping of mach exceptions in
>> + the inferior.  This might be seldom useful to better understand the
>> + cause of a fault. The default is off.
>
> So you are basically saying that mach exceptions are native Darwin
> exceptions, and they are then mapped to Posix signals?  At least that
> is what I understand from this explanation.

Yes.  I have used your reworded sentence.

> Also, please make sure you have 2 spaces after each period that ends a
> sentence.

Sure.

Tristan.