All of lore.kernel.org
 help / color / mirror / Atom feed
* PATCH:2.4:CONFIG_BINFMT_IRIX
@ 2003-08-05 18:25 Michael Pruznick
  2003-08-07 18:02 ` PATCH:2.4:CONFIG_BINFMT_IRIX Michael Pruznick
  0 siblings, 1 reply; 3+ messages in thread
From: Michael Pruznick @ 2003-08-05 18:25 UTC (permalink / raw)
  To: linux-mips


All this does is put "CONFIG_BINFMT_IRIX is not set" into the
config files so that when switching dual-endian systems from LE
to BE this will default to "n" instead of "y".

The main problem with this is that, in general, there is no
need/desire to have CONFIG_BINFMT_IRIX included just because
the kernel is BE.  Rather than being forced to disable this,
I think the default should be off.

In some older kernels, it this causes compile errors, but that
problem doesn't seam to exit in the latest 2.4 tree.

I tested this with menuconfig and xconfig.

I'm not an expert in all the subtle dependencies issues with
the config.in files so there may be a better way do to this.


cvs diff -uN arch/mips/config-shared.in
Index: arch/mips/config-shared.in
===================================================================
RCS file: /home/cvs/linux/arch/mips/Attic/config-shared.in,v
retrieving revision 1.1.2.80
diff -u -r1.1.2.80 config-shared.in
--- arch/mips/config-shared.in  5 Aug 2003 11:13:39 -0000       1.1.2.80
+++ arch/mips/config-shared.in  5 Aug 2003 17:07:24 -0000
@@ -817,6 +817,8 @@
 
 if [ "$CONFIG_CPU_LITTLE_ENDIAN" = "n" ]; then
    bool 'Include IRIX binary compatibility' CONFIG_BINFMT_IRIX
+else
+   define_bool CONFIG_BINFMT_IRIX n
 fi
 
 if [ "$CONFIG_CPU_R10000" = "y" ]; then

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: PATCH:2.4:CONFIG_BINFMT_IRIX
  2003-08-05 18:25 PATCH:2.4:CONFIG_BINFMT_IRIX Michael Pruznick
@ 2003-08-07 18:02 ` Michael Pruznick
  2003-08-07 18:37   ` PATCH:2.4:CONFIG_BINFMT_IRIX Guido Guenther
  0 siblings, 1 reply; 3+ messages in thread
From: Michael Pruznick @ 2003-08-07 18:02 UTC (permalink / raw)
  To: linux-mips

This seams to be a better way to eliminate the irix
stuff from being automatically included when switching
a board from le to be.

I tested this with config, oldconfig, menuconfig
and xconfig and in all cases the default for
CONFIG_BINFMT_IRIX is now N.


2.4
Index: arch/mips/defconfig
===================================================================
RCS file: /home/cvs/linux/arch/mips/defconfig,v
retrieving revision 1.117.2.52
diff -u -r1.117.2.52 defconfig
--- arch/mips/defconfig 16 Jul 2003 19:27:30 -0000      1.117.2.52
+++ arch/mips/defconfig 7 Aug 2003 17:44:27 -0000
@@ -112,7 +112,7 @@
 # General setup
 #
 # CONFIG_CPU_LITTLE_ENDIAN is not set
-CONFIG_BINFMT_IRIX=y
+# CONFIG_BINFMT_IRIX is not set
 CONFIG_ARC_CONSOLE=y
 # CONFIG_IP22_EISA is not set
 CONFIG_NET=y


2.6
Index: arch/mips/defconfig
===================================================================
RCS file: /home/cvs/linux/arch/mips/defconfig,v
retrieving revision 1.211
diff -u -r1.211 defconfig
--- arch/mips/defconfig 31 Jul 2003 17:28:07 -0000      1.211
+++ arch/mips/defconfig 7 Aug 2003 17:55:32 -0000
@@ -130,7 +130,7 @@
 CONFIG_BINFMT_ELF=y
 # CONFIG_BINFMT_MISC is not set
 CONFIG_TRAD_SIGNALS=y
-CONFIG_BINFMT_IRIX=y
+# CONFIG_BINFMT_IRIX is not set
 
 #
 # Memory Technology Devices (MTD)



Michael Pruznick wrote:
> 
> All this does is put "CONFIG_BINFMT_IRIX is not set" into the
> config files so that when switching dual-endian systems from LE
> to BE this will default to "n" instead of "y".
> 
> The main problem with this is that, in general, there is no
> need/desire to have CONFIG_BINFMT_IRIX included just because
> the kernel is BE.  Rather than being forced to disable this,
> I think the default should be off.
> 
> In some older kernels, it this causes compile errors, but that
> problem doesn't seam to exit in the latest 2.4 tree.
> 
> I tested this with menuconfig and xconfig.
> 
> I'm not an expert in all the subtle dependencies issues with
> the config.in files so there may be a better way do to this.
> 
> cvs diff -uN arch/mips/config-shared.in
> Index: arch/mips/config-shared.in
> ===================================================================
> RCS file: /home/cvs/linux/arch/mips/Attic/config-shared.in,v
> retrieving revision 1.1.2.80
> diff -u -r1.1.2.80 config-shared.in
> --- arch/mips/config-shared.in  5 Aug 2003 11:13:39 -0000       1.1.2.80
> +++ arch/mips/config-shared.in  5 Aug 2003 17:07:24 -0000
> @@ -817,6 +817,8 @@
> 
>  if [ "$CONFIG_CPU_LITTLE_ENDIAN" = "n" ]; then
>     bool 'Include IRIX binary compatibility' CONFIG_BINFMT_IRIX
> +else
> +   define_bool CONFIG_BINFMT_IRIX n
>  fi
> 
>  if [ "$CONFIG_CPU_R10000" = "y" ]; then

-- 
Michael Pruznick, michael_pruznick@mvista.com, www.mvista.com
MontaVista Software, 1237 East Arques Ave, Sunnyvale, CA 94085
direct voice/fax:970-266-1108, main office:408-328-9200

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: PATCH:2.4:CONFIG_BINFMT_IRIX
  2003-08-07 18:02 ` PATCH:2.4:CONFIG_BINFMT_IRIX Michael Pruznick
@ 2003-08-07 18:37   ` Guido Guenther
  0 siblings, 0 replies; 3+ messages in thread
From: Guido Guenther @ 2003-08-07 18:37 UTC (permalink / raw)
  To: Michael Pruznick; +Cc: linux-mips

[-- Attachment #1: Type: text/plain, Size: 315 bytes --]

On Thu, Aug 07, 2003 at 12:02:09PM -0600, Michael Pruznick wrote:
> This seams to be a better way to eliminate the irix
> stuff from being automatically included when switching
> a board from le to be.
What about leaving Irix compatibility off by default anyway. I don't
think it matters much these days.
 -- Guido

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2003-08-08 12:50 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-08-05 18:25 PATCH:2.4:CONFIG_BINFMT_IRIX Michael Pruznick
2003-08-07 18:02 ` PATCH:2.4:CONFIG_BINFMT_IRIX Michael Pruznick
2003-08-07 18:37   ` PATCH:2.4:CONFIG_BINFMT_IRIX Guido Guenther

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.