s390x assembly pack: implement OPENSSL_rdtsc as STCKF

.. if avalable. STCK has an artificial delay to ensure uniqueness
which can result in a performance penalty if used heavily
concurrently.
Signed-off-by: NPatrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: NAndy Polyakov <appro@openssl.org>
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5284)

crypto/s390xcpuid.pl

@@ -129,6 +129,14 @@ OPENSSL_s390x_facilities:
 .type	OPENSSL_rdtsc,\@function
 .align	16
 OPENSSL_rdtsc:
+	larl	%r4,OPENSSL_s390xcap_P
+	tm	S390X_STFLE+3(%r4),0x40	# check for store-clock-fast facility
+	jz	.Lstck
+
+	.long	0xb27cf010	# stckf 16($sp)
+	lg	%r2,16($sp)
+	br	$ra
+.Lstck:
 	stck	16($sp)
 	lg	%r2,16($sp)
 	br	$ra