Bugfix: mark outputs as early clobber in scalar x86_64 asm

In the existing code, the compiler is allowed to allocate the RSI register
for outputs m0, m1, or m2, which are written to before the input in RSI is
read from. Fix this by marking them as early clobber.

Reported by ehoffman2 in bitcoin-core/secp256k1#766

Author: sipa

src/scalar_4x64_impl.h

@@ -383,7 +383,7 @@ static void secp256k1_scalar_reduce_512(secp256k1_scalar *r, const uint64_t *l)
     "movq %%r10, %q5\n"
     /* extract m6 */
     "movq %%r8, %q6\n"
-    : "=g"(m0), "=g"(m1), "=g"(m2), "=g"(m3), "=g"(m4), "=g"(m5), "=g"(m6)
+    : "=&g"(m0), "=&g"(m1), "=&g"(m2), "=g"(m3), "=g"(m4), "=g"(m5), "=g"(m6)
     : "S"(l), "i"(SECP256K1_N_C_0), "i"(SECP256K1_N_C_1)
     : "rax", "rdx", "r8", "r9", "r10", "r11", "r12", "r13", "r14", "cc");