It seems to me that there is a small error in cpu_getmcontext(),
but I suspect it is quite harmless and does not break anything...

Anyway, here's a patch:

Index: machdep.c
===================================================================
RCS file: /cvsroot/netbsd/src/sys/arch/sparc64/sparc64/machdep.c,v
retrieving revision 1.171
diff -r1.171 machdep.c
1892c1892
< 	(void)memset(mcp, '0', sizeof (*mcp));
---
> 	(void)memset(mcp, '\0', sizeof (*mcp));