[PATCH][RFC] -EB / -EL don't properly affect gcc predefined symbols

[Hiroki Kaminaga]

Hi,

Regarding the PR:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29413

Is below patch adequate to fix the problem?
Works fine with me, but haven't done any regression test.


Thanks in Advance

(Hiroki Kaminaga)
t
--


Index: gcc/gcc/config/mips/mips.h
===
--- gcc.orig/gcc/config/mips/mips.h
+++ gcc/gcc/config/mips/mips.h
@@ -1152,7 +1152,8 @@ extern const struct mips_cpu_info *mips_
 #define SUBTARGET_CPP_SPEC ""
 #endif
 
-#define CPP_SPEC "%(subtarget_cpp_spec)"
+#define CPP_SPEC "%(subtarget_cpp_spec) %{EB:-meb} %{EL:-mel} \
+%{EB:%{EL:%emay not use both -EB and -EL}}"
 
 /* This macro defines names of additional specifications to put in the specs
that can be used in various specifications like CC1_SPEC.  Its definition

Re: [PATCH][RFC] -EB / -EL don't properly affect gcc predefined symbols

[Hiroki Kaminaga]

From: Andrew Pinski <[EMAIL PROTECTED]>
> As I mentioned in the bug report, the problem is that CC1_SPEC
> is the wrong one.
(snip)
> The attached patch is the more correct way to solve this.

Thanks alot!


(Hiroki Kaminaga)
t
--

Re: [PATCH][RFC] -EB / -EL don't properly affect gcc predefined symbols

[Hiroki Kaminaga]

From: Andrew Pinski <[EMAIL PROTECTED]>
> On Thu, 2006-10-12 at 17:06 +0900, Hiroki Kaminaga wrote:
> > Hi,
> > 
> > Regarding the PR:
> > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29413
> > 
> > Is below patch adequate to fix the problem?
> > Works fine with me, but haven't done any regression test.
> 
> As I mentioned in the bug report, the problem is that CC1_SPEC
> is the wrong one.
> 
> Adding EB/EL to CPP_SPEC will only fix the C/C++/Objective-C front-ends
> and no others.
> 
> The attached patch is the more correct way to solve this.

Your patch would cause `-profile' option not to set `-p' to cc1. Is this OK?


Best Regards,

(Hiroki Kaminaga)
t
--

Re: [PATCH][RFC] -EB / -EL don't properly affect gcc predefined symbols

[Hiroki Kaminaga]

From: Andrew Pinski <[EMAIL PROTECTED]>
> > Your patch would cause `-profile' option not to set `-p' to cc1. Is this OK?
> Not really, the following patch on top of the previous fixes the above
> problem:
> Index: config/mips/linux.h
> ===
> --- config/mips/linux.h (revision 117683)
> +++ config/mips/linux.h (working copy)
> @@ -49,6 +49,8 @@ Boston, MA 02110-1301, USA.  */
>  #undef MD_EXEC_PREFIX
>  #undef MD_STARTFILE_PREFIX
> 
> +#define SUBTARGET_CC1_SPEC "%{profile:-p}"
> +
>  /* If we don't set MASK_ABICALLS, we can't default to PIC.  */
>  #undef TARGET_DEFAULT
>  #define TARGET_DEFAULT MASK_ABICALLS

How about appending to CC1_SPEC?

-#ifndef CC1_SPEC
+#undef CC1_SPEC
 #define CC1_SPEC "\
 %{gline:%{!g:%{!g0:%{!g1:%{!g2: -g1} \
 %{G*} %{EB:-meb} %{EL:-mel} %{EB:%{EL:%emay not use both -EB and -EL}} \
 %{save-temps: } \
-%(subtarget_cc1_spec)"
+%(subtarget_cc1_spec) %{profile:-p}"
-#endif

Other architecture seems to set %{profile:-p} to CC1_SPEC.


Best Regards,

(Hiroki Kaminaga)
t
--