Fix ____longjmp_chk for ppc(64)

[Andreas Schwab]

Richard Henderson <rth@twiddle.net> writes:

> Ah.  Then perhaps a trap insn might be more
> appropriate (and slightly less confusing)?

How about this?

 	bl	HIDDEN_JUMPTARGET (__fortify_fail);	\
+	/* __fortify_fail doesn't return, but we need	\
+	   an insn to attach the cfi to it.  */		\
+	nop;						\
+	cfi_adjust_cfa_offset (-112);			\
+	cfi_same_value (lr);				\
 .Lok:

Andreas.

-- 
Andreas Schwab, aschwab@redhat.com
GPG Key fingerprint = D4E8 DBE3 3813 BB5D FA84  5EC7 45C6 250E 6F00 984E
"And now for something completely different."