Re: [2342] 2009-06-18 Vladimir Serbinenko <phcoder@gmail.com>

Re: [2342] 2009-06-18 Vladimir Serbinenko <phcoder@gmail.com>

[Christian Franke]

Vladimir Serbinenko wrote:
> 	* configure.ac: remove a leftover AC_MSG_RESULT
> ...
> --- trunk/grub2/configure.ac	2009-06-18 13:51:06 UTC (rev 2341)
> +++ trunk/grub2/configure.ac	2009-06-18 14:02:23 UTC (rev 2342)
> @@ -308,7 +308,6 @@
>  AC_SUBST(TARGET_IMG_LDSCRIPT)
>  AC_SUBST(TARGET_IMG_LDFLAGS)
>  AC_SUBST(TARGET_IMG_CFLAGS)
> -AC_MSG_RESULT([$TARGET_IMG_LDFLAGS_AC])
>  
>   

This was the AC_MSG_RESULT(...) for the
AC_MSG_CHECKING([for option to link raw image])
which is still present for the "$grub_cv_apple_target_cc" != "yes"  case.

I would suggest to move the AC_MSG_CHECKING outside the 'if 
"$grub_cv_apple_target_cc" = "yes" ... fi' and re-add the AC_MSG_RESULT.

-- 
Regards,
Christian Franke

Re: [2342] 2009-06-18 Vladimir Serbinenko <phcoder@gmail.com>

[Pavel Roskin]

On Fri, 2009-06-19 at 22:05 +0200, Christian Franke wrote:

> This was the AC_MSG_RESULT(...) for the
> AC_MSG_CHECKING([for option to link raw image])
> which is still present for the "$grub_cv_apple_target_cc" != "yes"  case.
> 
> I would suggest to move the AC_MSG_CHECKING outside the 'if 
> "$grub_cv_apple_target_cc" = "yes" ... fi' and re-add the AC_MSG_RESULT.

Thanks for pointing it out.  Actually, the message is not needed at all.
There is no "checking" that needs to be communicated to the user.  It
would be appropriate if configure was running a compiler or looked for
some tool in the PATH.  In this case, it's just shell variable
assignments with on test for a file.

-- 
Regards,
Pavel Roskin

Re: [2342] 2009-06-18 Vladimir Serbinenko <phcoder@gmail.com>

[Christian Franke]

Pavel Roskin wrote:
> On Fri, 2009-06-19 at 22:05 +0200, Christian Franke wrote:
>   
>> This was the AC_MSG_RESULT(...) for the
>> AC_MSG_CHECKING([for option to link raw image])
>> which is still present for the "$grub_cv_apple_target_cc" != "yes"  case.
>>
>> I would suggest to move the AC_MSG_CHECKING outside the 'if 
>> "$grub_cv_apple_target_cc" = "yes" ... fi' and re-add the AC_MSG_RESULT.
>>     
>
> Thanks for pointing it out.  Actually, the message is not needed at all.
> There is no "checking" that needs to be communicated to the user.  It
> would be appropriate if configure was running a compiler or looked for
> some tool in the PATH.  In this case, it's just shell variable
> assignments with on test for a file.
>
>   

I would prefer an output here even if the test done isn't a 'real' 
configure check. Would also make sense when we later apply your
'Use common linker script for all i386-pc systems' suggestion.

But if you want to remove this messages it is also fine with me.

Same would then possibly also apply to the
AC_MSG_CHECKING([for command to convert module to ELF format])
a few lines below.

-- 
Regards,
Christian Franke

Re: [2342] 2009-06-18 Vladimir Serbinenko <phcoder@gmail.com>

[Pavel Roskin]

On Sat, 2009-06-20 at 09:04 +0200, Christian Franke wrote:

> I would prefer an output here even if the test done isn't a 'real' 
> configure check. Would also make sense when we later apply your
> 'Use common linker script for all i386-pc systems' suggestion.
> 
> But if you want to remove this messages it is also fine with me.
> 
> Same would then possibly also apply to the
> AC_MSG_CHECKING([for command to convert module to ELF format])
> a few lines below.

At least it doesn't produce a long line starting with a dash that looks
out of place.

-- 
Regards,
Pavel Roskin