Ensure forward progress on trace exit to BC_ITERN.

Also use a safer way to force a static dispatch for BC_RET*.
Reported by Bartel Eerdekens. Analyzed by Peter Cawley. #1000 #1045

src/vm_arm.dasc

@@ -2196,8 +2196,8 @@ static void build_subroutines(BuildCtx *ctx)
   |.if JIT
   |  ldr L, SAVE_L
   |1:
-  |  cmp CARG1, #0
-  |  blt >9				// Check for error from exit.
+  |  cmn CARG1, #LUA_ERRERR
+  |  bhs >9				// Check for error from exit.
   |   lsl RC, CARG1, #3
   |  ldr LFUNC:CARG2, [BASE, FRAME_FUNC]
   |   str RC, SAVE_MULTRES
@@ -2213,6 +2213,8 @@ static void build_subroutines(BuildCtx *ctx)
   |   ldr INS, [PC], #4
   |     lsl MASKR8, MASKR8, #3		// MASKR8 = 255*8.
   |    st_vmstate CARG4
+  |  cmn CARG1, #17			// Static dispatch?
+  |  beq >5
   |  cmp OP, #BC_FUNCC+2		// Fast function?
   |  bhs >4
   |2:
@@ -2238,6 +2240,17 @@ static void build_subroutines(BuildCtx *ctx)
   |  ldr KBASE, [CARG3, #PC2PROTO(k)]
   |  b <2
   |
+  |5:  // Dispatch to static entry of original ins replaced by BC_JLOOP.
+  |  ldr CARG1, [DISPATCH, #DISPATCH_J(trace)]
+  |  decode_RD RC, INS
+  |  ldr TRACE:CARG1, [CARG1, RC, lsl #2]
+  |  ldr INS, TRACE:CARG1->startins
+  |  decode_OP OP, INS
+  |   decode_RA8 RA, INS
+  |  add OP, DISPATCH, OP, lsl #2
+  |   decode_RD RC, INS
+  |  ldr pc, [OP, #GG_DISP2STATIC]
+  |
   |9:  // Rethrow error from the right C frame.
   |  rsb CARG2, CARG1, #0
   |  mov CARG1, L