Re: [PATCH][3/4] Add hotplug support to drivers/acpi/numa.c

From: Keiichiro Tokunaga <tokunaga.keiich_at_jp.fujitsu.com>
Date: 2004-09-29 16:26:29
On Mon, 27 Sep 2004 13:06:16 -0700 Keshavamurthy Anil S wrote:
> On Fri, Sep 24, 2004 at 01:32:55AM +0900, Keiichiro Tokunaga wrote:
> > +void acpi_numa_node_init(acpi_handle handle)
> Why is this function returning void? I expect
> this to return int, what do you think?
> > +
> > +	if (acpi_bus_get_device(handle, &node_dev)) {
> > +		printk(KERN_ERR"Unknown handle.\n");
> > +		return_VOID;
> > +	}
> Why do you need to call acpi_bus_get_device?

I wrote it for printk()s and ACPI_DEBUG_PRINT() to log.

        if (! _cnt) {                                                           
                ACPI_DEBUG_PRINT((ACPI_DB_INFO,                                 
                                  "nid of <%s> is not detected.\n",             
                                  acpi_device_bid(node_dev))); 
                goto cancel;                                                    
        }
        ...
        status = acpi_attach_data(handle, acpi_numa_data_handler, data);        
        if (ACPI_FAILURE(status)) {                                             
                printk(KERN_ERR"Failed to attach NUMA data for <%s>.\n",        
                       acpi_device_bid(node_dev));                              
                goto cancel;                                                    
        }

        printk(KERN_INFO"Container <%s> is NUMA node.\n",                       
               acpi_device_bid(node_dev));

> > +	acpi_walk_namespace(ACPI_TYPE_PROCESSOR,
> > +			    handle,
> > +			    (u32) 1,
> > +			    find_processor,
> > +			    data,
> > +			    (void **)&cnt);
> Why are you looking for processor device here?
> Please remove this acpi_walk_namespace function.

The reason why the acpi_walk_namespace() is used here was to
find a container object which is identical to a NUMA node.  My code
was assuming that a container having CPU and/or memory was
NUMA node sinece the current Linux seemed to assume so.

> > +	/*
> > diff -puN /dev/null include/acpi/numa.h
> > +#ifndef MAX_PXM_DOMAINS
> > +#define MAX_PXM_DOMAINS (256)
> > +#endif
> Why defining it again, It is already defined in asm-ia64/acpi.h file

Sorry, that's a stuff that I forgot to remove.  I will remove it.

Thanks,
Keiichiro Tokunaga
-
To unsubscribe from this list: send the line "unsubscribe linux-ia64" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Received on Wed Sep 29 02:30:44 2004

This archive was generated by hypermail 2.1.8 : 2005-08-02 09:20:31 EST