Add a comment indicating that the current method of enabling high vectors
requires that the CPU control vector be properly readable. I believe that all CPUs that have high vector support have a readable CPU control register, but if we ever encounter one that does not, then we'll have to adjust this code.
This commit is contained in:
parent
11cc5eb2c6
commit
78b1b81e74
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: arm32_machdep.c,v 1.31 2003/04/18 11:08:24 scw Exp $ */
|
||||
/* $NetBSD: arm32_machdep.c,v 1.32 2003/04/18 22:30:05 thorpej Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1994-1998 Mark Brinicombe.
|
||||
@ -151,6 +151,13 @@ arm32_vector_init(vaddr_t va, int which)
|
||||
* cpu_setup()) because the vector page needs to be
|
||||
* accessible *before* cpu_startup() is called.
|
||||
* Think ddb(9) ...
|
||||
*
|
||||
* NOTE: If the CPU control register is not readable,
|
||||
* this will totally fail! We'll just assume that
|
||||
* any system that has high vector support has a
|
||||
* readable CPU control register, for now. If we
|
||||
* ever encounter one that does not, we'll have to
|
||||
* rethink this.
|
||||
*/
|
||||
cpu_control(CPU_CONTROL_VECRELOC, CPU_CONTROL_VECRELOC);
|
||||
}
|
||||
|
Loading…
Reference in New Issue
Block a user