[PATCH v4 2/2] arch: arm: Show the serial number from devicetree in cpuinfo

[Paul Kocialkowski]

Le dimanche 07 juin 2015 à 18:03 +0100, Grant Likely a écrit :
> On Tue, 28 Apr 2015 09:29:56 +0200
> , Paul Kocialkowski <contact at paulk.fr>
>  wrote:
> > This grabs the serial number shown in cpuinfo from the serial-number devicetree
> > property in priority. When booting with ATAGs (and without device-tree), the
> > provided number is still shown instead.
> > 
> > Signed-off-by: Paul Kocialkowski <contact at paulk.fr>
> 
> One comment below, but otherwise:
> 
> Acked-by: Grant Likely <grant.likely at linaro.org>

I'm afraid this was merged in Russell's tree already:
http://ftp.arm.linux.org.uk/cgit/linux-arm.git/commit/?h=for-next&id=3f599875e5202986b350618a617527ab441bf206

Still, it might be useful to make another patch on top with your
comments!

> > ---
> >  arch/arm/include/asm/system_info.h |  1 +
> >  arch/arm/kernel/setup.c            | 23 +++++++++++++++++++++--
> >  2 files changed, 22 insertions(+), 2 deletions(-)
> > 
> > diff --git a/arch/arm/include/asm/system_info.h b/arch/arm/include/asm/system_info.h
> > index 720ea03..3860cbd40 100644
> > --- a/arch/arm/include/asm/system_info.h
> > +++ b/arch/arm/include/asm/system_info.h
> > @@ -17,6 +17,7 @@
> >  
> >  /* information about the system we're running on */
> >  extern unsigned int system_rev;
> > +extern const char *system_serial;
> >  extern unsigned int system_serial_low;
> >  extern unsigned int system_serial_high;
> >  extern unsigned int mem_fclk_21285;
> > diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c
> > index 1d60beb..b501754 100644
> > --- a/arch/arm/kernel/setup.c
> > +++ b/arch/arm/kernel/setup.c
> > @@ -93,6 +93,9 @@ unsigned int __atags_pointer __initdata;
> >  unsigned int system_rev;
> >  EXPORT_SYMBOL(system_rev);
> >  
> > +const char *system_serial;
> > +EXPORT_SYMBOL(system_serial);
> > +
> 
> Is there any need to export this symbol? It's only used by built-in
> code. Not by modules.

I agree, but I thought it would be consistent with the way
system_serial_high/low are exported. I'm not sure this was ever needed
either, but generally, I guess it makes sense to export parameters that
are passed to the kernel (it was through ATAGs before, now it's through
device-tree, but the logic remains the same).

> >  unsigned int system_serial_low;
> >  EXPORT_SYMBOL(system_serial_low);
> >  
> > @@ -821,8 +824,25 @@ arch_initcall(customize_machine);
> >  
> >  static int __init init_machine_late(void)
> >  {
> > +	struct device_node *root;
> > +	int ret;
> > +
> >  	if (machine_desc->init_late)
> >  		machine_desc->init_late();
> > +
> > +	root = of_find_node_by_path("/");
> > +	if (root) {
> > +		ret = of_property_read_string(root, "serial-number",
> > +					      &system_serial);
> 
> of_property_read_string() will not modify the argument on failure. If
> system_serial is initialized to NULL, then the failure path is
> unnecessary.

Makes sense.

> > +		if (ret)
> > +			system_serial = NULL;
> > +	}
> 
> Calls to of_find_node* functions increment the refcount for the node.
> Need an of_node_put() here, or just use the of_root pointer directly. It
> is safe to call of_property_read_string() with a NULL node pointer too,
> it will fail gracefully.
> 
> So, the whole thing can safely boil down to:
> 
> 	of_property_read_string(of_root, "serial-number", &system_serial);
> 	if (!system_serial)
> 		system_serial = kasprintf(GFP_KERNEL, "%08x%08x",
> 					  system_serial_high, system_serial_low);

Looks good to me, feel free to commit this change, on top of:
http://ftp.arm.linux.org.uk/cgit/linux-arm.git/commit/?h=for-next&id=3f599875e5202986b350618a617527ab441bf206
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: This is a digitally signed message part
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20150611/b66a37b3/attachment.sig>