Added test of Break/Trap, cleaned up macros.

Author: FluBBaOfWard

README.md

@@ -1,4 +1,4 @@
-# WonderSwan TimingTest V0.3.0 (20230118)
+# WonderSwan TimingTest V0.4.0 (20230909)
 
 This is a program to test the timing of the CPU in a Bandai WonderSwan.
 These results are not an absolute definition of how long an opcode takes, the cpu has an instruction pipeline which is flushed when branching, it takes 1 extra clock when branching to an odd address, it also seems to take more cycles depending on what the destination instruction is. I have tried to make the tests of jump/branch/call avoid these situations but I'm not sure I did it correctly everywhere.
@@ -42,6 +42,6 @@ Godzil for boot rom stubs. https://github.com/Godzil/NewOswan
 lidnariq for RTC info.
 ```
 
-Twitter @TheRealFluBBa
+Twitter/X @TheRealFluBBa
 
 <https://www.github.com/FluBBaOfWard>

TimingTest_todo.txt

@@ -1,3 +1,4 @@
 TimingTest Todo
 
+Check which port is used in the fill_prefetch macro for "in al, dx"
 Test different instructions as destination from JMP.

testcalls.asm

@@ -407,7 +407,7 @@ execute  7, 74, teststringEA8F, test_EA8F
 execute  8, 82, teststringEAC4, test_EAC4
 execute  9, 82, teststringEAC4, test_EAC42
 execute 10, 82, teststringEAC5, test_EAC5
-execute 11, 86, teststringEAC5, test_EAC52
+execute 11, 82, teststringEAC5, test_EAC52
 execute 12, 74, teststringEAFF, test_EAFF
 execute 13, 70, teststringA00, test_A00X
 execute 14, 66, teststringA04, test_A04X
@@ -422,6 +422,7 @@ execute  5,300, teststringEI37,test_EI37
 execute  6,289, teststringEI3E,test_EI3E
 execute  7,300, teststringEI3F,test_EI3F
 execute  8,293, teststringExIrq,test_ExIrq
+execute  9,261, teststringTrap,test_Trap
 test22end:
 
 testgroup 23