Remove !cold KASSERT - it does not compile on all kernels, and it is not the
right thing to test for anyway. XXX should we panic instead? Are "compatible" strings this long happening in real devices?
This commit is contained in:
parent
23444c94ea
commit
a111d20361
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: ofw_subr.c,v 1.38 2020/06/25 22:50:56 rin Exp $ */
|
||||
/* $NetBSD: ofw_subr.c,v 1.39 2020/06/26 10:14:32 martin Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright 1998
|
||||
@ -34,7 +34,7 @@
|
||||
*/
|
||||
|
||||
#include <sys/cdefs.h>
|
||||
__KERNEL_RCSID(0, "$NetBSD: ofw_subr.c,v 1.38 2020/06/25 22:50:56 rin Exp $");
|
||||
__KERNEL_RCSID(0, "$NetBSD: ofw_subr.c,v 1.39 2020/06/26 10:14:32 martin Exp $");
|
||||
|
||||
#include <sys/param.h>
|
||||
#include <sys/systm.h>
|
||||
@ -109,7 +109,6 @@ of_compatible(int phandle, const char * const *strings)
|
||||
return (-1);
|
||||
|
||||
if (len > sizeof(sbuf)) {
|
||||
KASSERT(!cold);
|
||||
buf = malloc(len, M_TEMP, M_WAITOK);
|
||||
allocated = 1;
|
||||
} else {
|
||||
|
Loading…
Reference in New Issue
Block a user