[RFC] ppc32: Report chip version in common /proc/cpuinfo handling

Tom,

You asked me to look at doing the following.  I just want to make sure 
this is what we want to do (and call it).  I'm a little concerned that 
'chip' is not necessary the right name in light of TSI10x and MV64x60 
being described by ppc_sys in the future.

- kumar

--

Moved reporting of chip revision from board specific to common handing of 
/proc/cpuinfo.  In light of numerous PPC system-on-chip devices, we report 
both the cpu version (reflects the core version) and the chip version 
(reflects the system-on-chip version).

Signed-off-by: Kumar Gala <redacted>

---
diff -Nru a/arch/ppc/kernel/setup.c b/arch/ppc/kernel/setup.c

--- a/arch/ppc/kernel/setup.c	2005-03-18 18:17:16 -06:00
+++ b/arch/ppc/kernel/setup.c	2005-03-18 18:17:16 -06:00

@@ -40,6 +40,7 @@
 #include <asm/nvram.h>
 #include <asm/xmon.h>
 #include <asm/ocp.h>
+#include <asm/ppc_sys.h>
 
 #if defined CONFIG_KGDB
 #include <asm/kgdb.h>

@@ -245,6 +246,11 @@
 
 	seq_printf(m, "bogomips\t: %lu.%02lu\n",
 		   lpj / (500000/HZ), (lpj / (5000/HZ)) % 100);
+
+#if defined (CONFIG_85xx) || defined (CONFIG_83xx)
+	if (cur_ppc_sys_spec->ppc_sys_name)
+		seq_printf(m, "chip\t\t: %s\n", cur_ppc_sys_spec->ppc_sys_name);
+#endif
 
 #ifdef CONFIG_SMP
 	seq_printf(m, "\n");

diff -Nru a/arch/ppc/platforms/83xx/mpc834x_sys.c b/arch/ppc/platforms/83xx/mpc834x_sys.c

--- a/arch/ppc/platforms/83xx/mpc834x_sys.c	2005-03-18 18:17:16 -06:00
+++ b/arch/ppc/platforms/83xx/mpc834x_sys.c	2005-03-18 18:17:16 -06:00

@@ -143,7 +143,6 @@
 	pvid = mfspr(SPRN_PVR);
 	svid = mfspr(SPRN_SVR);
 
-	seq_printf(m, "chip\t\t: MPC%s\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "Vendor\t\t: Freescale Inc.\n");
 	seq_printf(m, "Machine\t\t: mpc%s sys\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "core clock\t: %d MHz\n"

diff -Nru a/arch/ppc/platforms/85xx/mpc85xx_ads_common.c b/arch/ppc/platforms/85xx/mpc85xx_ads_common.c

--- a/arch/ppc/platforms/85xx/mpc85xx_ads_common.c	2005-03-18 18:17:16 -06:00
+++ b/arch/ppc/platforms/85xx/mpc85xx_ads_common.c	2005-03-18 18:17:16 -06:00

@@ -129,7 +129,6 @@
 	pvid = mfspr(SPRN_PVR);
 	svid = mfspr(SPRN_SVR);
 
-	seq_printf(m, "chip\t\t: MPC%s\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "Vendor\t\t: Freescale Semiconductor\n");
 	seq_printf(m, "Machine\t\t: mpc%sads\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "clock\t\t: %dMHz\n", freq / 1000000);

diff -Nru a/arch/ppc/platforms/85xx/mpc85xx_cds_common.c b/arch/ppc/platforms/85xx/mpc85xx_cds_common.c

--- a/arch/ppc/platforms/85xx/mpc85xx_cds_common.c	2005-03-18 18:17:16 -06:00
+++ b/arch/ppc/platforms/85xx/mpc85xx_cds_common.c	2005-03-18 18:17:16 -06:00

@@ -146,7 +146,6 @@
 	pvid = mfspr(SPRN_PVR);
 	svid = mfspr(SPRN_SVR);
 
-	seq_printf(m, "chip\t\t: MPC%s\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "Vendor\t\t: Freescale Semiconductor\n");
 	seq_printf(m, "Machine\t\t: CDS - MPC%s (%x)\n", cur_ppc_sys_spec->ppc_sys_name, cadmus[CM_VER]);
 	seq_printf(m, "clock\t\t: %dMHz\n", freq / 1000000);

diff -Nru a/arch/ppc/platforms/85xx/sbc85xx.c b/arch/ppc/platforms/85xx/sbc85xx.c

--- a/arch/ppc/platforms/85xx/sbc85xx.c	2005-03-18 18:17:16 -06:00
+++ b/arch/ppc/platforms/85xx/sbc85xx.c	2005-03-18 18:17:16 -06:00

@@ -129,7 +129,6 @@
 	pvid = mfspr(SPRN_PVR);
 	svid = mfspr(SPRN_SVR);
 
-	seq_printf(m, "chip\t\t: MPC%s\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "Vendor\t\t: Wind River\n");
 	seq_printf(m, "Machine\t\t: SBC%s\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "clock\t\t: %dMHz\n", freq / 1000000);

diff -Nru a/arch/ppc/platforms/85xx/stx_gp3.c b/arch/ppc/platforms/85xx/stx_gp3.c

--- a/arch/ppc/platforms/85xx/stx_gp3.c	2005-03-18 18:17:16 -06:00
+++ b/arch/ppc/platforms/85xx/stx_gp3.c	2005-03-18 18:17:16 -06:00

@@ -268,7 +268,6 @@
 
 	memsize = total_memory;
 
-	seq_printf(m, "chip\t\t: MPC%s\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "Vendor\t\t: RPC Electronics STx \n");
 	seq_printf(m, "Machine\t\t: GP3 - MPC%s\n", cur_ppc_sys_spec->ppc_sys_name);
 	seq_printf(m, "bus freq\t: %u.%.6u MHz\n", freq / 1000000,

Ok, in a possible solution we could call it 'chip(set)' this dealing=20
with either chips (85??, 82??, 52??, ....) or chipsets (MV64x60,=20
TSI10x, MPC10x, etc..)

- kumar

On Mar 18, 2005, at 6:26 PM, Kumar Gala wrote:

Tom,

You asked me to look at doing the following.=A0 I just want to make =

sure

this is what we want to do (and call it).=A0 I'm a little concerned =

that

'chip' is not necessary the right name in light of TSI10x and MV64x60
being described by ppc_sys in the future.

- kumar

--=20

Moved reporting of chip revision from board specific to common handing=20=

of
/proc/cpuinfo.=A0 In light of numerous PPC system-on-chip devices, we=20=

(lpj / (5000/HZ)) % 100);

%s\n",=20

18:17:16=20

-06:00
 +++ b/arch/ppc/platforms/83xx/mpc834x_sys.c=A0=A0=A0=A0 2005-03-18 =

18:17:16=20

2005-03-18=20

18:17:16 -06:00
 +++ b/arch/ppc/platforms/85xx/mpc85xx_ads_common.c=A0=A0=A0=A0=A0 =

2005-03-18=20

18:17:16 -06:00
 @@ -129,7 +129,6 @@
 =A0=A0=A0=A0=A0=A0=A0 pvid =3D mfspr(SPRN_PVR);
=A0=A0=A0=A0=A0=A0=A0 svid =3D mfspr(SPRN_SVR);
=A0
 -=A0=A0=A0=A0=A0=A0 seq_printf(m, "chip\t\t: MPC%s\n",=20
cur_ppc_sys_spec->ppc_sys_name);
=A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "Vendor\t\t: Freescale =

Semiconductor\n");

=A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "Machine\t\t: mpc%sads\n",=20
cur_ppc_sys_spec->ppc_sys_name);
=A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "clock\t\t: %dMHz\n", freq / =

1000000);

2005-03-18=20

18:17:16 -06:00
 +++ b/arch/ppc/platforms/85xx/mpc85xx_cds_common.c=A0=A0=A0=A0=A0 =

2005-03-18=20

18:17:16 -06:00
 @@ -146,7 +146,6 @@
 =A0=A0=A0=A0=A0=A0=A0 pvid =3D mfspr(SPRN_PVR);
=A0=A0=A0=A0=A0=A0=A0 svid =3D mfspr(SPRN_SVR);
=A0
 -=A0=A0=A0=A0=A0=A0 seq_printf(m, "chip\t\t: MPC%s\n",=20
cur_ppc_sys_spec->ppc_sys_name);
=A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "Vendor\t\t: Freescale =

Semiconductor\n");

=A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "Machine\t\t: CDS - MPC%s (%x)\n",=20=

cur_ppc_sys_spec->ppc_sys_name, cadmus[CM_VER]);
=A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "clock\t\t: %dMHz\n", freq / =

1000000);

1000000);

\n");

 =A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "Machine\t\t: GP3 - MPC%s\n",=20
cur_ppc_sys_spec->ppc_sys_name);
=A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "bus freq\t: %u.%.6u MHz\n", freq =

/ 1000000,

_______________________________________________
Linuxppc-embedded mailing list
 Linuxppc-embedded@ozlabs.org
https://ozlabs.org/mailman/listinfo/linuxppc-embedded

On Fri, Mar 18, 2005 at 06:26:52PM -0600, Kumar Gala wrote:

You asked me to look at doing the following.  I just want to make sure 
this is what we want to do (and call it).  I'm a little concerned that 
'chip' is not necessary the right name in light of TSI10x and MV64x60 
being described by ppc_sys in the future.

"chip" or "chip(set)" or anything else is fine with me (I'm horrible at
names).  But I don't like:

ifdef'ing this.  If the field is set, we should print it.

-- 
Tom Rini
http://gate.crashing.org/~trini/

On Mar 21, 2005, at 10:39 AM, Tom Rini wrote:

On Fri, Mar 18, 2005 at 06:26:52PM -0600, Kumar Gala wrote:

quoted

You asked me to look at doing the following.=A0 I just want to make=20=

sure

quoted

this is what we want to do (and call it).=A0 I'm a little concerned=20=

that

quoted

'chip' is not necessary the right name in light of TSI10x and =

MV64x60

quoted

being described by ppc_sys in the future.

"chip" or "chip(set)" or anything else is fine with me (I'm horrible =

at

 names).=A0 But I don't like:

quoted

+#if defined (CONFIG_85xx) || defined (CONFIG_83xx)
+=A0=A0=A0=A0 if (cur_ppc_sys_spec->ppc_sys_name)
+=A0=A0=A0=A0 =A0=A0=A0=A0=A0=A0=A0 seq_printf(m, "chip\t\t: %s\n",=20=

cur_ppc_sys_spec->ppc_sys_name);

quoted

+#endif

ifdef'ing this.=A0 If the field is set, we should print it.

The only reason this is ifdef'd is that cur_ppc_sys_spec does not=20
always exist on all platforms built.

- kumar=

On Mon, Mar 21, 2005 at 11:21:28AM -0600, Kumar Gala wrote:

On Mar 21, 2005, at 10:39 AM, Tom Rini wrote:

quoted

On Fri, Mar 18, 2005 at 06:26:52PM -0600, Kumar Gala wrote:

quoted

You asked me to look at doing the following.? I just want to make 

sure

quoted

this is what we want to do (and call it).? I'm a little concerned 

that

quoted

'chip' is not necessary the right name in light of TSI10x and MV64x60
being described by ppc_sys in the future.

"chip" or "chip(set)" or anything else is fine with me (I'm horrible at
names).? But I don't like:

quoted

+#if defined (CONFIG_85xx) || defined (CONFIG_83xx)
+???? if (cur_ppc_sys_spec->ppc_sys_name)
+???? ??????? seq_printf(m, "chip\t\t: %s\n", 

cur_ppc_sys_spec->ppc_sys_name);

quoted

+#endif

ifdef'ing this.? If the field is set, we should print it.

The only reason this is ifdef'd is that cur_ppc_sys_spec does not 
always exist on all platforms built.

Oh boy, I sense an <asm-ppc/serial.h> in the making.  Please tell me
it's at least not possible that someone could stick an 85xx and an
MV64x60 together.  So we want the enum there to do sanity checking,
right?

-- 
Tom Rini
http://gate.crashing.org/~trini/

On Mar 21, 2005, at 12:05 PM, Tom Rini wrote:

On Mon, Mar 21, 2005 at 11:21:28AM -0600, Kumar Gala wrote:
 >

quoted

On Mar 21, 2005, at 10:39 AM, Tom Rini wrote:

 >

quoted

quoted

On Fri, Mar 18, 2005 at 06:26:52PM -0600, Kumar Gala wrote:

 > >
 > >>
 > >> You asked me to look at doing the following.? I just want to =

make

quoted

quoted

sure

 > >> this is what we want to do (and call it).? I'm a little =

concerned

quoted

quoted

that

 > >> 'chip' is not necessary the right name in light of TSI10x and=20
MV64x60
 > >> being described by ppc_sys in the future.
 > >
 > >"chip" or "chip(set)" or anything else is fine with me (I'm=20
horrible at
 > > names).? But I don't like:

quoted

quoted

 > >> +#if defined (CONFIG_85xx) || defined (CONFIG_83xx)

quoted

quoted

quoted

+???? if (cur_ppc_sys_spec->ppc_sys_name)
+???? ??????? seq_printf(m, "chip\t\t: %s\n",

cur_ppc_sys_spec->ppc_sys_name);

quoted

+#endif

 > >
 > >ifdef'ing this.? If the field is set, we should print it.
 >

quoted

The only reason this is ifdef'd is that cur_ppc_sys_spec does not
always exist on all platforms built.

Oh boy, I sense an <asm-ppc/serial.h> in the making.=A0 Please tell me
 it's at least not possible that someone could stick an 85xx and an
 MV64x60 together.=A0 So we want the enum there to do sanity checking,
 right?

It is not possible that this could happen, in the sense that you are=20
using the MV64x60 as a chipset, not a PCI device or something bizarre.

- kumar=