Re: vmalloc comments patch

Mel (mel@csn.ul.ie)
Sun, 21 Apr 2002 21:54:59 +0100 (IST)


On Sun, 21 Apr 2002, Anton Blanchard wrote:

> > + * To give a safety margin, the linear address starts about 8MB after the end
> > + * of physical memory at VMALLOC_START. This is to try and catch memory
> > + * overruns.
>
> Thats architecture dependent. On ppc64 for example the kernel text
> starts at 0xC000000000000000 and the vmalloc region starts at
> 0xD000000000000000, 2^60 bits apart.
>

Changed to

* These are the functions for assigning a block of linear addresses for pages.
* To give a safety margin, the linear address starts at VMALLOC_START.
* This is at PAGE_OFFSET + VMALLOC_OFFSET which are all arch dependant
* values

Is that ok?

> > spin_unlock(&init_mm.page_table_lock);
> > page = alloc_page(gfp_mask);
> > spin_lock(&init_mm.page_table_lock);
>
> Since alloc_page can sleep we must drop the spinlock.
>

Changed to

/* The page table lock has to be released because alloc_page
* could sleep if memory is low
*/

All good?

Mel

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/