13 files changed, 192 insertions, 0 deletions
diff --git a/testroms/branch_timing_tests/1.Branch_Basics.nes b/testroms/branch_timing_tests/1.Branch_Basics.nes
new file mode 100644
index 0000000..e7f81f0
--- /dev/null
+++ b/testroms/branch_timing_tests/1.Branch_Basics.nes
diff --git a/testroms/branch_timing_tests/2.Backward_Branch.nes b/testroms/branch_timing_tests/2.Backward_Branch.nes
new file mode 100644
index 0000000..6e02064
--- /dev/null
+++ b/testroms/branch_timing_tests/2.Backward_Branch.nes
diff --git a/testroms/branch_timing_tests/3.Forward_Branch.nes b/testroms/branch_timing_tests/3.Forward_Branch.nes
new file mode 100644
index 0000000..4e701b2
--- /dev/null
+++ b/testroms/branch_timing_tests/3.Forward_Branch.nes
diff --git a/testroms/branch_timing_tests/readme.txt b/testroms/branch_timing_tests/readme.txt
new file mode 100644
index 0000000..1f05e8e
--- /dev/null
+++ b/testroms/branch_timing_tests/readme.txt
@@ -0,0 +1,74 @@
+NES 6502 Branch Timing Test ROMs
+--------------------------------
+These ROMs test timing of the branch instruction, including edge cases
+which an emulator might get wrong. When run on a NES they all give a
+passing result. Each ROM runs several tests and reports the result on
+screen and by beeping a number of times. See below for the meaning of
+failure codes for each test. THE TESTS MUST BE RUN (*AND* *PASS*) IN
+ORDER, because some earlier ROMs test things that later ones assume will
+work properly.
+
+Source code for each test is included, and most tests are clearly
+divided into sections. Support code is also included, but it runs on a
+custom devcart and assembler so it will require some effort to assemble.
+Contact me if you'd like assistance porting them to your setup.
+
+
+Branch Timing Summary
+---------------------
+An untaken branch takes 2 clocks. A taken branch takes 3 clocks. A taken
+branch that crosses a page takes 4 clocks. Page crossing occurs when the
+high byte of the branch target address is different than the high byte
+of address of the next instruction:
+
+branch_target:
+	...
+	bne branch_target
+next_instruction:
+	nop
+	...
+branch_target:
+
+
+1.Branch_Basics
+---------------
+Tests branch timing basics and PPU NMI timing, which is needed for the
+tests
+
+2) NMI period is too short
+3) NMI period is too too long
+4) Branch not taken is too long
+5) Branch not taken is too short
+6) Branch taken is too long
+7) Branch taken is too short
+
+
+2.Backward_Branch
+-----------------
+Tests backward (negative) branch timing.
+
+2) Branch from $E4FD to $E4FC is too long
+3) Branch from $E4FD to $E4FC is too short
+4) Branch from $E5FE to $E5FD is too long
+5) Branch from $E5FE to $E5FD is too short
+6) Branch from $E700 to $E6FF is too long
+7) Branch from $E700 to $E6FF is too short
+8) Branch from $E801 to $E800 is too long
+9) Branch from $E801 to $E800 is too short
+
+
+3.Forward_Branch
+----------------
+Tests forward (positive) branch timing.
+
+2) Branch from $E5FC to $E5FF is too long
+3) Branch from $E5FC to $E5FF is too short
+4) Branch from $E6FD to $E700 is too long
+5) Branch from $E6FD to $E700 is too short
+6) Branch from $E7FE to $E801 is too long
+7) Branch from $E7FE to $E801 is too short
+8) Branch from $E8FF to $E902 is too long
+9) Branch from $E8FF to $E902 is too short
+
+-- 
+Shay Green <[email protected]>
diff --git a/testroms/nestest.nes b/testroms/nestest.nes
new file mode 100644
index 0000000..fc2a88c
--- /dev/null
+++ b/testroms/nestest.nes
diff --git a/testroms/vbl_nmi_timing/1.frame_basics.nes b/testroms/vbl_nmi_timing/1.frame_basics.nes
new file mode 100644
index 0000000..8eece87
--- /dev/null
+++ b/testroms/vbl_nmi_timing/1.frame_basics.nes
diff --git a/testroms/vbl_nmi_timing/2.vbl_timing.nes b/testroms/vbl_nmi_timing/2.vbl_timing.nes
new file mode 100644
index 0000000..3a45ec8
--- /dev/null
+++ b/testroms/vbl_nmi_timing/2.vbl_timing.nes
diff --git a/testroms/vbl_nmi_timing/3.even_odd_frames.nes b/testroms/vbl_nmi_timing/3.even_odd_frames.nes
new file mode 100644
index 0000000..535cb44
--- /dev/null
+++ b/testroms/vbl_nmi_timing/3.even_odd_frames.nes
diff --git a/testroms/vbl_nmi_timing/4.vbl_clear_timing.nes b/testroms/vbl_nmi_timing/4.vbl_clear_timing.nes
new file mode 100644
index 0000000..939f452
--- /dev/null
+++ b/testroms/vbl_nmi_timing/4.vbl_clear_timing.nes
diff --git a/testroms/vbl_nmi_timing/5.nmi_suppression.nes b/testroms/vbl_nmi_timing/5.nmi_suppression.nes
new file mode 100644
index 0000000..5677690
--- /dev/null
+++ b/testroms/vbl_nmi_timing/5.nmi_suppression.nes
diff --git a/testroms/vbl_nmi_timing/6.nmi_disable.nes b/testroms/vbl_nmi_timing/6.nmi_disable.nes
new file mode 100644
index 0000000..75a3a2c
--- /dev/null
+++ b/testroms/vbl_nmi_timing/6.nmi_disable.nes
diff --git a/testroms/vbl_nmi_timing/7.nmi_timing.nes b/testroms/vbl_nmi_timing/7.nmi_timing.nes
new file mode 100644
index 0000000..f777a9c
--- /dev/null
+++ b/testroms/vbl_nmi_timing/7.nmi_timing.nes
diff --git a/testroms/vbl_nmi_timing/readme.txt b/testroms/vbl_nmi_timing/readme.txt
new file mode 100644
index 0000000..03728ec
--- /dev/null
+++ b/testroms/vbl_nmi_timing/readme.txt
@@ -0,0 +1,118 @@
+NTSC NES PPU VBL/NMI Timing Tests
+---------------------------------
+These ROMs test the timing of the VBL flag and NMI to an accuracy of a
+single PPU clock, and also check special cases. They have been tested on
+an actual NES and all give a passing result. Sometimes the NES starts up
+with a different PPU timing that causes some of the tests to fail; these
+tests don't check that timing arrangement.
+
+Each ROM runs several tests and reports the result on screen and by
+beeping a number of times. See below for the meaning of failure codes
+for each test. It's best to run the tests in order, because later ROMs
+depend on things tested by earlier ROMs and will give erroneous results
+if any earlier ones failed.
+
+Source code for each test is included, and most tests are clearly
+divided into sections. Support code is also included, but it runs on a
+custom devcart and assembler so it will require some effort to assemble.
+Contact me if you'd like assistance porting them to your setup.
+
+
+1.frame_basics
+--------------
+Tests basic VBL flag operation and general timing of PPU frames.
+
+2) VBL flag isn't being set
+3) VBL flag should be cleared after being read
+4) PPU frame with BG enabled is too short
+5) PPU frame with BG enabled is too long
+6) PPU frame with BG disabled is too short
+7) PPU frame with BG disabled is too long
+
+
+2.vbl_timing
+------------
+Tests timing of VBL being set, and special case where reading VBL flag
+as it would be set causes it to not be set for that frame.
+
+2) Flag should read as clear 3 PPU clocks before VBL
+3) Flag should read as set 0 PPU clocks after VBL
+4) Flag should read as clear 2 PPU clocks before VBL
+5) Flag should read as set 1 PPU clock after VBL
+6) Flag should read as clear 1 PPU clock before VBL
+7) Flag should read as set 2 PPU clocks after VBL
+8) Reading 1 PPU clock before VBL should suppress setting
+
+
+3.even_odd_frames
+-----------------
+Test clock skipped when BG is enabled on odd PPU frames. Tests
+enable/disable BG during 5 consecutive frames, then see how many clocks
+were skipped. Patterns are shown as XXXXX, where each X can either be B
+(BG enabled) or - (BG disabled).
+
+2) Pattern ----- should not skip any clocks
+3) Pattern BB--- should skip 1 clock
+4) Pattern B--B- (one even, one odd) should skip 1 clock
+5) Pattern -B--B (one odd, one even) should skip 1 clock
+6) Pattern BB-BB (two pairs) should skip 2 clocks
+
+
+4.vbl_clear_timing
+------------------
+Tests timing of VBL flag clearing.
+
+2) Cleared 3 or more PPU clocks too early
+3) Cleared 2 PPU clocks too early
+4) Cleared 1 PPU clock too early 
+5) Cleared 3 or more PPU clocks too late
+6) Cleared 2 PPU clocks too late
+7) Cleared 1 PPU clock too late
+
+
+5.nmi_suppression
+-----------------
+Tests timing of NMI suppression when reading VBL flag just as it's set,
+and that this doesn't occur when reading one clock before or after.
+
+2) Reading flag 3 PPU clocks before set shouldn't suppress NMI
+3) Reading flag when it's set should suppress NMI
+4) Reading flag 3 PPU clocks after set shouldn't suppress NMI
+5) Reading flag 2 PPU clocks before set shouldn't suppress NMI
+6) Reading flag 1 PPU clock after set should suppress NMI
+7) Reading flag 4 PPU clocks after set shouldn't suppress NMI
+8) Reading flag 4 PPU clocks before set shouldn't suppress NMI
+9) Reading flag 1 PPU clock before set should suppress NMI
+10)Reading flag 2 PPU clocks after set shouldn't suppress NMI
+
+
+6.nmi_disable
+-------------
+Tests NMI occurrence when disabling NMI just as VBL flag is set, and
+just after.
+
+2) NMI shouldn't occur when disabled 0 PPU clocks after VBL
+3) NMI should occur when disabled 3 PPU clocks after VBL
+4) NMI shouldn't occur when disabled 1 PPU clock after VBL
+5) NMI should occur when disabled 4 PPU clocks after VBL
+6) NMI shouldn't occur when disabled 1 PPU clock before VBL
+7) NMI should occur when disabled 2 PPU clocks after VBL
+
+
+7.nmi_timing
+------------
+Tests timing of NMI and immediate occurrence when enabled with VBL flag
+already set.
+
+2) NMI occurred 3 or more PPU clocks too early
+3) NMI occurred 2 PPU clocks too early
+4) NMI occurred 1 PPU clock too early
+5) NMI occurred 3 or more PPU clocks too late
+6) NMI occurred 2 PPU clocks too late
+7) NMI occurred 1 PPU clock too late
+8) NMI should occur if enabled when VBL already set
+9) NMI enabled when VBL already set should delay 1 instruction
+10)NMI should be possible multiple times in VBL
+
+-- 
+Shay Green <hotpop.com@blargg> (swap to e-mail)