Re: Strange virtual/physical address of global data

• Previous message: David S. Miller: "Re: [patch *] VM deadlock fix"
• In reply to: Timur Tabi: "Strange virtual/physical address of global data"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

> unsigned long test_data;
>
> int init_module(void)
> {
> void *virt = &test_data;
> unsigned long phys = virt_to_phys(virt);
>
> When I run this and check the valur of virt and phys, it appears that phys is
> outside the range of physical memory! That is, if I have 512MB of RAM, then
> phys is equal to about 520M. However, if I make test_data a local variable:

Module data isnt guaranteed to be in dmaable memory (nor btw is the stack but
you were lucky). If you want to use virt_to_bus() and friends for stuff like
DMA you must kmalloc the buffers

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

• Next message: Alan Cox: "Re: Question: Using floating point in the kernel"
• Previous message: David S. Miller: "Re: [patch *] VM deadlock fix"
• In reply to: Timur Tabi: "Strange virtual/physical address of global data"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

This archive was generated by hypermail 2b29 : Thu Sep 21 2000 - 18:38:35 EDT