Maintainers file patch for M68HC1x port.

Maintainers file patch for M68HC1x port.

[Sean Keys]

[-- Attachment #1: Type: text/plain, Size: 130 bytes --]

Hi,

Attached is a patch, that adds my contact information, to the 
binutils/MAINTAINERS file for the M68HC1x port.

Thanks,
Sean

[-- Attachment #2: MAINTAINERS.diff --]
[-- Type: text/x-patch, Size: 570 bytes --]

diff --git binutils/MAINTAINERS binutils/MAINTAINERS
index 9850903..94f40e0 100644
--- binutils/MAINTAINERS
+++ binutils/MAINTAINERS
@@ -92,7 +92,6 @@ responsibility among the other maintainers.
   LM32             Jon Beniston <jon@beniston.com>
   M32R             Doug Evans <dje@sebabeach.org>
   M68HC11 M68HC12  Stephane Carrez <stcarrez@nerim.fr>
-  M68HC11 M68HC12  Sean Keys <skeys@ipdatasys.com>
   M88k		   Mark Kettenis <kettenis@gnu.org>
   MACH-O           Tristan Gingold <gingold@adacore.com>
   MAXQ		   Inderpreet Singh <inderpreetb@noida.hcltech.com>

Re: Maintainers file patch for M68HC1x port.

[James Murray]

On Tue, 2012-07-24 at 08:13 -0700, Sean Keys wrote:
> Hi,
> 
> Attached is a patch, that adds my contact information, to the 
> binutils/MAINTAINERS file for the M68HC1x port.
> 
> Thanks,
> Sean

That's rather contentious. We've both recently contributed to hc11.

Most of the addition by volume was from me. The code you have
contributed was to support linking in of the xgate additional target you
added.

(We discussed your other patch by email.) I'm still not sure I can think
of a case where a symbol would be created in XGATE space that needs to
be referenced by the S12, but the check you've added for that case is
likely harmless.

However, it would be worth adding some testcases to the testsuite to
confirm the new behaviour.  

regards

James

Re: Maintainers file patch for M68HC1x port.

[Fred Cooke]

James,

> I'm still not sure I can think
> of a case where a symbol would be created in XGATE space that needs to
> be referenced by the S12, but the check you've added for that case is
> likely harmless.

You may not be able to think of such a case, however dozens of people
are using such code as I write this, and have been for years. This is
a program architecture choice, and not something to be mandated by
you.

> However, it would be worth adding some testcases to the testsuite to
> confirm the new behaviour.

Agreed.

Regards,

Fred.

Re: Maintainers file patch for M68HC1x port.

[Sean Keys]

James,

 > That's rather contentious. We've both recently contributed to hc11.

We are both users/devs of this port, and are both therefore part of the 
same team. I'm after only one thing: contributing to this port and 
making it better.

 >Most of the addition by volume was from me. The code you have
 >contributed was to support linking in of the xgate additional target you
 >added.

The bulk of your patch was for XGATE support within the HC1x tools, not 
fixes or additions to the port itself. I have demonstrated continued 
commitment to development of this port. I've also spent enough time 
working with the code base to be completely comfortable maintaining it 
and that's the bottom line.

 >(We discussed your other patch by email.) I'm still not sure I can think
 >of a case where a symbol would be created in XGATE space that needs to
 >be referenced by the S12, but the check you've added for that case is
 >likely harmless.

It's a valid and reasonable use case, therefore in my opinion it should 
be left up to the developer. I ran the existing test-suite to make sure 
it was harmless to existing usage, of course.

 >However, it would be worth adding some testcases to the testsuite to
 >confirm the new behaviour.

Indeed, let me integrate my existing test into the test-suite. I'll post 
it to the mailing list, as soon as possible.

Thanks,
Sean

On 07/24/2012 12:14 PM, James Murray wrote:
> On Tue, 2012-07-24 at 08:13 -0700, Sean Keys wrote:
>> Hi,
>>
>> Attached is a patch, that adds my contact information, to the
>> binutils/MAINTAINERS file for the M68HC1x port.
>>
>> Thanks,
>> Sean
> That's rather contentious. We've both recently contributed to hc11.
>
> Most of the addition by volume was from me. The code you have
> contributed was to support linking in of the xgate additional target you
> added.
>
> (We discussed your other patch by email.) I'm still not sure I can think
> of a case where a symbol would be created in XGATE space that needs to
> be referenced by the S12, but the check you've added for that case is
> likely harmless.
>
> However, it would be worth adding some testcases to the testsuite to
> confirm the new behaviour.
>
> regards
>
> James

Re: Maintainers file patch for M68HC1x port.

[James Murray]

On Tue, 2012-07-24 at 13:49 -0700, Sean Keys wrote:
> The bulk of your patch was for XGATE support within the HC1x tools, not 
> fixes or additions to the port itself.

Erm? What about the addition of the S12X instruction support, objdump
output enhancements and HC11, HC12 testcase fixes ? Those all sound very
much like fixes and additions to the port itself.

>  >( I'm still not sure I can think
>  >of a case where a symbol would be created in XGATE space.. 
> 
> It's a valid and reasonable use case, therefore in my opinion it should 
> be left up to the developer. I ran the existing test-suite to make sure 
> it was harmless to existing usage, of course.

My question was based on the fact that I've not used it that way. (Isn't
asking questions and peer review a core part of open source software
such as binutils?) In my application all the variables are defined in
the S12 code and the XGATE code links against them.
Could you point me to the Freescale datasheet that you mentioned so I
can have a look at what they are doing.

regards

James

Re: Maintainers file patch for M68HC1x port.

[Fred Cooke]

James,

> > The bulk of your patch was for XGATE support within the HC1x tools, not
> > fixes or additions to the port itself.
>
> Erm? What about the addition of the S12X instruction support, objdump
> output enhancements and HC11, HC12 testcase fixes ? Those all sound very
> much like fixes and additions to the port itself.

Yes, they certainly are, however they're overshadowed by the size of
your XGATE-in-other-assembler hacks.

Regards,

Fred.

Re: Maintainers file patch for M68HC1x port.

[Sean Keys]

Again, It's not about how you or I currently use the tools, it's about desired/possible usage by all.

When I have some extra time I'll hunt for an example from Freescale. Even better, if one doesn't exist, and I never said that one definitely did, I will create one. Then we can say that binutils does something that the commercial tools don't.

-Sean


On 07/24/2012 02:03 PM, James Murray wrote:
> On Tue, 2012-07-24 at 13:49 -0700, Sean Keys wrote:
>> The bulk of your patch was for XGATE support within the HC1x tools, not
>> fixes or additions to the port itself.
> Erm? What about the addition of the S12X instruction support, objdump
> output enhancements and HC11, HC12 testcase fixes ? Those all sound very
> much like fixes and additions to the port itself.
>
>>   >( I'm still not sure I can think
>>   >of a case where a symbol would be created in XGATE space..
>>
>> It's a valid and reasonable use case, therefore in my opinion it should
>> be left up to the developer. I ran the existing test-suite to make sure
>> it was harmless to existing usage, of course.
> My question was based on the fact that I've not used it that way. (Isn't
> asking questions and peer review a core part of open source software
> such as binutils?) In my application all the variables are defined in
> the S12 code and the XGATE code links against them.
> Could you point me to the Freescale datasheet that you mentioned so I
> can have a look at what they are doing.
>
> regards
>
> James