[parisc-linux] Re: [parisc-linux-cvs] linux-2.6 jejb

Joel Soete soete.joel at tiscali.be
Mon May 3 02:57:34 MDT 2004

> Modified files:
> 	.              : Makefile 
> 	arch/parisc/kernel: asm-offsets.c entry.S head64.S 
> 	i
> clude/asm-parisc: page.h pgalloc.h pgtable.h 
> Log message:
> Increase the maximum physical and virtual space available to
> the 64 bit kernel from 512GB to 8TB

Sincerely nice job but how will it permit me to access at least to the only
all 4GB of the N4k I test when setup_bootmem() limits actually the usage
to the first 2Gb?

For remind:
static void __init setup_bootmem(void)
         * Throw out ranges that are too far apart (controlled by
         * MAX_GAP). If CONFIG_DISCONTIGMEM wasn't implemented so
         * poorly, we would recommend enabling that option, but,
         * until it is fixed, this is the best way to go.

        printk("npmem_ranges = %d.(line 171)\n", npmem_ranges);
        for (i = 1; i < npmem_ranges; i++) {
              printk("pmem_ranges[%d].start_pfn = %ld.\n", i-1, pmem_ranges[i-1].pages);
                printk("pmem_ranges[%d].pages = %ld.\n", i-1, pmem_ranges[i-1].pages);
                printk("pmem_ranges[%d].start_pfn = %ld.\n", i, pmem_ranges[i].pages);
And actual results would be:
<==== return by setup_bootmem() ====>
pmem_ranges[0].start_pfn = 0.
pmem_ranges[0].pages = 524288.
pmem_ranges[1].start_pfn = 1572864.

so 1572864-524288-0=1048576 greater then MAX_GAP = 262144.

I also investigate the choice of using CONFIG_DISCONTIGMEM but as far as
I understand this option is foreseen when each processor has its own ram
and can/want access the ram of the others (as on sun 6800): not the N4k case
where all ram is shared between all processor?

Thanks again for your attention,

Tiscali ADSL: 35 €/mois, la meilleure offre du marché!

More information about the parisc-linux mailing list