http://tinywrex.wordpress.com/2008/03/07/success-lc-working-on-32bit-rom/

It worked for the Forester XT EDM

Now it works for the 2004 STI USDM

Enjoy

:banana: (http://www.youtube.com/watch?v=cYU4ugI_VAc)

Thanks xswrex

click the dancing banana

thanks for testing Nuke ;)

:D

http://tinywrex.wordpress.com/2008/03/08/success-x2-lc-working-on-usdm-my-04-sti-32bit-rom/

Aces!

I will be waiting! ;)

awesome...too bad i have no reason to use it :(

Finally the wait is almost over!!!

Can't wait. Just in time for my six speed swap.

Welcome to last year. I've been running LC and FFS on my 2005 STI (Open Source code) for almost over a year. You'll end up experiencing the same things that I did (and that Cobb guys do) .... learned ignition retard that makes your car slower.

t

Welcome to last year. I've been running LC and FFS on my 2005 STI (Open Source code) for almost over a year. You'll end up experiencing the same things that I did (and that Cobb guys do) .... learned ignition retard that makes your car slower.

t

First of all, this thread is for LC, nothing more and nothing less. If you experience knock during FFS thats another story and does not belong here. Did you see anything about FFS in this thread? I dont understand what is the point of your post.

Second, i havent seen any open source LC or FFS available to the public. How can you call a questionably hacked ROM that you use open source code? Or you are one of the few chosen ones that have this feature running on their cars? Yeah, thats really open source...

Despite all of the above, if what you are running is open source code where is the source code? I do provide the code, as my patches are THE ACTUAL CODE.

You have not tried my version of LC, yet you have an opinion and compare it to something that is not available to anyone.

Really i can go on and on but it is not worth my time.
I really would like to see some consistent arguments from you, even if i doubt there are any.

Thank you for your appreciation on my work

PS: welcome to this year!

First of all, this thread is for LC, nothing more and nothing less. If you experience knock during FFS thats another story and does not belong here. Did you see anything about FFS in this thread? I dont understand what is the point of your post.

Second, i havent seen any open source LC or FFS available to the public. How can you call a questionably hacked ROM that you use open source code? Or you are one of the few chosen ones that have this feature running on their cars? Yeah, thats really open source...

Despite all of the above, if what you are running is open source code where is the source code? I do provide the code, as my patches are THE ACTUAL CODE.

You have not tried my version of LC, yet you have an opinion and compare it to something that is not available to anyone.

Really i can go on and on but it is not worth my time.
I really would like to see some consistent arguments from you, even if i doubt there are any.

Thank you for your appreciation on my work

PS: welcome to this year!
I appreciate your work. Keep it up.

I Know mine is a 16bit but that is one of your first patches. I have nothing but good things to say about this LC i log my car every day and my KC hasn't change a bit. Keep the good work.

Welcome to last year. I've been running LC and FFS on my 2005 STI (Open Source code) for almost over a year. You'll end up experiencing the same things that I did (and that Cobb guys do) .... learned ignition retard that makes your car slower.

t

You might as well give yourself a pat in the back and move on...

You want Open Source? Fine. Here ya go. Have fun.


Different between:
First file: "C:\Documents and Settings\Tim\Desktop\Tuning\WolfPlayer\042407.hex"
Second file: "C:\Documents and Settings\Tim\Desktop\Tuning\WolfPlayer\042407_2ste p.hex"
Shift: 0
------------------------------------------------------------------------

First file: "C:\Documents and Settings\Tim\Desktop\Tuning\WolfPlayer\042407.hex"
Second file: "C:\Documents and Settings\Tim\Desktop\Tuning\WolfPlayer\042407_2ste p.hex"
Shift: 0
Shift: 0


------------------------------------------------------------------------

00000FF8 | 90 96 | 00000FF8 | 91 95 |
------------------------------------------------------------------------
0002BA40 | D3 F4 38 D2 1D | 0002BA40 | D2 42 2B 00 09 |
0002BA48 | D1 1E | 0002BA48 | D2 1C |
------------------------------------------------------------------------
0002BA58 | D3 1A | 0002BA58 | 00 09 |
------------------------------------------------------------------------
0002BAB8 | FF FF 50 68 | 0002BAB8 | 00 0F A0 00 |
------------------------------------------------------------------------
000CD720 | FF FF FF FF FF FF FF FF | 000CD720 | 40 4D F3 B6 45 7A 00 00 |
000CD728 | FF FF FF FF FF FF FF FF | 000CD728 | 45 76 E0 00 45 AB E0 00 |
000CD730 | FF FF FF FF | 000CD730 | 45 A8 C0 00 |
------------------------------------------------------------------------
000FA000 | FF FF FF FF FF FF FF FF | 000FA000 | D3 16 D0 17 F4 38 60 00 |
000FA008 | FF FF FF FF FF FF FF | 000FA008 | C8 80 00 29 70 60 0B |
000FA010 | FF FF FF FF FF FF FF FF | 000FA010 | 88 01 8F 0F 00 09 D1 0A |
000FA018 | FF FF FF FF FF FF FF FF | 000FA018 | D2 0A F5 18 F3 28 F5 35 |
000FA020 | FF FF FF FF FF FF FF FF | 000FA020 | 8D 04 00 09 D1 08 D3 09 |
000FA028 | FF FF FF FF FF FF FF FF | 000FA028 | A0 06 00 09 D1 0D D3 0E |
000FA030 | FF FF FF FF FF FF FF FF | 000FA030 | A0 02 00 09 D1 06 D3 07 |
000FA038 | FF FF FF FF FF FF | 000FA038 | D2 07 42 2B 00 09 |
000FA040 | FF FF FF FF FF FF | 000FA040 | 50 54 00 0C D7 20 |
000FA048 | FF FF FF FF FF FF FF FF | 000FA048 | 00 0C D7 24 00 0C D7 28 |
000FA050 | FF FF FF FF FF FF FF FF | 000FA050 | 00 0C 60 38 00 0C 60 3C |
000FA058 | FF FF FF FF FF FF | 000FA058 | 00 02 BA 46 50 68 |
000FA060 | FF FF FF FF FF FF | 000FA060 | 50 39 00 0C D7 2C |
000FA068 | FF FF FF FF FF FF FF FF | 000FA068 | 00 0C D7 30 00 01 03 01 |
------------------------------------------------------------------------


Nuff said.

Wow, why would you attack someone who is helping the general opensource community?

xswrex is not trying to tune cars hes only writing code for those of us that dont have a background in computer programing and would other wise still be looking to cobb or ecutek for any tuning period.

im pretty sure there is a ways to change the fine learned tables to reduce the effects of the negative learned timing.

Thanx xswrex for the hard work and all the other Romraider Guys :)

For the record, my main point wasn't 'welcome to last year'. However, I do understand why you guys are focusing on that. For that - I apologize. Here is what I was thinking in my head:

Here we go again. People are going to get really excited about launch control. It relies on the factory fuel cut and will pull timing. People don't realize that this has beat to death on various forums about how this will actually end up causing you to run slower. This is for LC as well as FFS - although, admittedly, with LC you won't see it as much. The fuel cut LC works so great on UTEC because it controls the timing.

Just my 0.02.

t

For the record, my main point wasn't 'welcome to last year'. However, I do understand why you guys are focusing on that. For that - I apologize. Here is what I was thinking in my head:

Here we go again. People are going to get really excited about launch control. It relies on the factory fuel cut and will pull timing. People don't realize that this has beat to death on various forums about how this will actually end up causing you to run slower. This is for LC as well as FFS - although, admittedly, with LC you won't see it as much. The fuel cut LC works so great on UTEC because it controls the timing.

Just my 0.02.

t

If that was your point maybe you should have said it in a way that didn't seem like an attack on someone who is trying to help those of us who don't have the time or ability to do these things.

what you say it true it can ultimately cause timing to be pulled, but right now most are happy with the current progress being made

Im not a programmer but i would imagine its very difficult to hack the timing controls.

Thanks for the clarification WolfPlayer, I will say, you should've said that in the first place :P

Anyways, at least some progress are being made, I'm pretty sure those who are doing as much as they can, can try to figure it out.

Thanks for the clarification WolfPlayer, I will say, you should've said that in the first place :P




Yes, I should have - and for that, I am sorry.

t

Welcome to last year. I've been running LC and FFS on my 2005 STI (Open Source code) for almost over a year. You'll end up experiencing the same things that I did (and that Cobb guys do) .... learned ignition retard that makes your car slower.


Where was this code being distributed last year? I thought I knew where to look for this stuff...

Did you come up with the code changes? If not, who?

Thanks!

Also, the "source" in "open source" refers to "source code." Hexadecimal values don't count. :)

Not that I have any complaints about "free software," of course.

Dude, hex is source code. haven't you ever heard of assembly?

^^^ What he said. Hex is the actual source code.

No, I did not come up with it.

t

Yes, actually I am quite familiar with assembly, it's been a while since I've used it but once upon a time even got paid to write it for the 65816.


An assembly language is a low-level language for programming computers. It implements a symbolic representation of the numeric machine codes and other constants needed to program a particular CPU architecture. This representation is usually defined by the hardware manufacturer, and is based on abbreviations (called mnemonics) that help the programmer remember individual instructions, registers, etc. An assembly language is thus specific to a certain physical or virtual computer architecture (as opposed to most high-level languages, which are portable).

Assembly languages were first developed in the 1950s, when they were referred to as second generation programming languages. They eliminated much of the error-prone and time-consuming first-generation programming needed with the earliest computers, freeing the programmer from tedium such as remembering numeric codes and calculating addresses.


Emphasis mine.

Who did write it?

Was it shared before now? If so, where? No, I'm not setting you up for pedantic quibble about the meaning of "open," I'm just curious whether there's another forum somewhere that I should be following.

Thanks!

Who did write it?

Was it shared before now? If so, where? No, I'm not setting you up for pedantic quibble about the meaning of "open," I'm just curious whether there's another forum somewhere that I should be following.

Thanks!
I wrote that a year ago as a quick test for clutch-based launch control with FFS for the 32-bit ECU. I haven't messed with it since then. I wanted to test the basic functionality so that I could later integrate it with real-time tuning along with per gear wg comp and maybe a few other features. I never planned on releasing it separately as it is easier to develop these as a package than having to deal with older version/revisions as new features are released (esp. with RT tuning).

At the time, RomRaider development was going well and it appeared that java development for real-time tuning interface was just a matter of time. But, a lot of the java developers gradually dropped out and progress eventually became non-existent, especially as related to real-time tuning. So, that is where it is today. However, I am working on a separate, standalone app to allow for real-time tuning without worrying about integrating into RomRaider (there's a thread about it on the RomRaider site).

I think it is great that tinywrex is working on this stuff for the community. There really aren't too many hackers in the Subaru community. I don't think Wolfplayer meant any disrespect, he was just making a comment about the issues with running FFS using the rev limit fuel cut not realizing that tinywrex wasn't adding FFS. It might be possible to overcome the negative FLKC from using FFS (I have a few ideas) as long as it was false knock, but that is a discussion for another thread.

I don't think Wolfplayer meant any disrespect, he was just making a comment about the issues with running FFS using the rev limit fuel cut not realizing that tinywrex wasn't adding FFS.

^^^ Yes. I've spent countless time discussing the pulled timing. Again, I never meant any disrespect. It all came out wrong. I apologized twice. Sometimes you'll have that - sometimes you'll have those days when your thoughts just don't come out right. Yesterday was one of those days. Thanks for everybody setting me on fire :diaf: ;) My buttocks needed to be regulated.

I'm willing to test whatever. I tested the LC/FFS stuff as posted before and it works so incredibly well except for the learned timing. I truly love FFS. With a FMIC it makes everything so much better (no need to refill all that piping because the throttle plate is held Wide Open on shifts and boost is maintained!).

t

Ok I have a degree in Computer Sci, however I was not fortunate enough to get much training on assembly and hex. Could someone point me in a good direction to start. I learn pretty quick and have a few things I would like to play with. Thanks!

Ok I have a degree in Computer Sci, however I was not fortunate enough to get much training on assembly and hex. Could someone point me in a good direction to start. I learn pretty quick and have a few things I would like to play with. Thanks!

Romraider.com

I have a USDM 05 STi with rom A2ZJB11J. Just wondering when LC and mabe even FFS might be released for this rom?

I have a USDM 05 STi with rom A2ZJB11J. Just wondering when LC and mabe even FFS might be released for this rom?

You'll just have to hop onto the waiting boat along with everyone else who wants LC for their cars. :)

I have a USDM 05 STi with rom A2ZJB11J. Just wondering when LC and mabe even FFS might be released for this rom?

Don't worry Ryan. I've been monitoring the progress and we'll test it out when its ready ;)

No freebies for the 32 bit guys, huh?

Any timeframe for 07' USDM STI?

Will you be releasing the source code, or is this not open-source?

No freebies for the 32 bit guys, huh?

Any timeframe for 07' USDM STI?

Will you be releasing the source code, or is this not open-source?

Currently 04 STi, 06 WRX are available, 05 STi due this week.
Patches for most USDM models will be available until end of March.
Source code is always included in the patch package

Several people have implemented this before, though no one has made a bunch of ROMs available for download. I didn't see any download links. Maybe I missed them. You seem to say here that you do have download links. I'm just curious to take a peak and see what you did and compared to what I had tried.

I was never satisfied with what I came up with for an obvious hack using the factory rev limiter, either. I did run it, but it didn't work that well and I felt better off just launching normally.

Don't take my post as negative, please.

Here's the last ASM file I could find. I'm not 100% sure this is what I last tried. It's been over a year. This can be compiled with KPIT GNU SH assembler (free download), then needs to be pasted in with a hex editor, plus the call to the rev limiter routine needs to be changed to the location you paste to. This is also for the A2ZJ710J ECU, since I run the same car.
http://freon.shackspace.com/car/ecu%20and%20tuning/openecu/enhancements/revlimit/revlimit.asm

You seem to say here that you do have download links. I'm just curious to take a peak and see what you did and compared to what I had tried.

as i have stated some time ago, the 32bit patches are not free of charge, thus it doesnt feel right for me to post links in here that point to my online store. You can always visit my blog where i have posted many many free 16bit patches.


Don't take my post as negative, please.
I dont see a reason to ;)

Code i've written has the same logic used in the 16bit patches. I too use GNU SH assembler

Make sure before you use the LC feature that you change your (fuel resume boost) table.
if not your going to launch good then when you hit 10psi that car is going to shut off.

just a little reminder.
other than that seems to work ok.

^^^ very good point

Make sure before you use the LC feature that you change your (fuel resume boost) table.
if not your going to launch good then when you hit 10psi that car is going to shut off.

just a little reminder.
other than that seems to work ok.

Can you elaborate on this?

Can you elaborate on this?

Correct me if I'm wrong
It should be the 'Rev Limit Fuel Resume' table under 'Misc'
I'm looking at the default on the sti's map and it's set at 9.67psi

So I'm guessing he's recommending to lower the boost?

That won't be necessary unless you're making more than ~10 psi using the LC, which doesn't seem probable.

I just grabbed the 32bit LC for my car, but won't be able to test it out until my ankle heals. :(

Ironically, my XP box had some issues patching it (wouldn't recognize TinyPatcher.exe as an executable :lol:) while my vista laptop had no problems. :lol:

Also, are there any plans to add a clutch switch to the 32bit for FFS?

the car will 2 step fine but if you want to launch control with a mph set higher than 0 you will need to rise the (fuel resume) under the misc tab on the boost or the car will easily reach more than 9psi. if you just sit and 2 step if it will reach can reach more than 9psi, and if you are launching the car hard and go through the gear it should hit more than 9psi and the logic will wait for the boost to go under 9psi before the injectors fire again. its like hitting boost cut on your launch.

if you dont change it you will notice it.

O but i did launch the car after fixing it and almost smoked the tires though all of first :devil:

the car will 2 step fine but if you want to launch control with a mph set higher than 0 you will need to rise the (fuel resume) under the misc tab on the boost or the car will easily reach more than 9psi. if you just sit and 2 step if it will reach can reach more than 9psi, and if you are launching the car hard and go through the gear it should hit more than 9psi and the logic will wait for the boost to go under 9psi before the injectors fire again. its like hitting boost cut on your launch.

if you dont change it you will notice it.

O but i did launch the car after fixing it and almost smoked the tires though all of first :devil:

Besides the confusing grammar, I got it. Makes sense.

What kind of boost are you building, and at whats your LC RPM?

5000rpm probably 7psi within a second on a 18g.
yeah i was just kinda thinking and typing on the grammer

also you may want to adjust the (learned timing column ranges), so that if it does learn to pull timing, it will only pull timing for a couple 100rpm.

32bit ECU Open Source LC FTW!!!

Alright looks like I got an error right off the bat. I unzipped the tiny patcher A2ZJB11J (USDM 05 STI) files then put a copy of my rom labled .hex at the end in the folder then opened the bat. file and get an error saying "Invalid command line arguments". What have a I done wrong?

the rom label should be
A2ZJB11J.hex

not any other label

Alright looks like I got an error right off the bat. I unzipped the tiny patcher A2ZJB11J (USDM 05 STI) files then put a copy of my rom labled .hex at the end in the folder then opened the bat. file and get an error saying "Invalid command line arguments". What have a I done wrong?

the filename of your ROM can be anything but it must not contain spaces!

ooh, did not know u changed it

the rom label should be
A2ZJB11J.hex

not any other label

the filename of your ROM can be anything but it must not contain spaces!

Alright first it's in my own name with .hex and it has a space. I'll correct this and give it another try thanks much. Cant wait to try it out!

Alright first it's in my own name with .hex and it has a space. I'll correct this and give it another try thanks much. Cant wait to try it out!

it will work fine if you remove the space(s) ;)

thanks for catching this bug in the batch file!

Hey Freon, I had a few questions about the LC assembly you posted:


/* rev limiter
modified code from Subaru Denso/Hitachi/Renesas 7055 ECU
rev limiter code
*/

rev_limit:
mov.w (rpm), r3 ! h'A78C rpm
mov.w (kph), r4 ! h'A778 kph
mova (kph_lc_limit), r0 ! kph limit for lc/normal
fmov.s @r4, fr2
fmov.s @r0, fr5
mov.w (map), r2 ! h'A4E0 ; FFFFA4E0 map
fcmp/gt fr5, fr2 ! is kph_lc_limit < kph
mov.w (rvlmbit), r4 ! h'AD58 ; FFFFAD58 rev limiter bit
bt/s over_lcspeed
fmov.s @r3, fr4 ! fr4 = FFFFA78C rpm

mova (upper_lim_lc), r0
fmov.s @r0, fr3 ! fr3 = upper limit
mova (lower_lim_lc), r0
bra check
nop

over_lcspeed:
mova (upper_lim), r0 ! rev limit upper
fmov.s @r0, fr3 ! fr3 = upper limit
mova (lower_lim), r0 ! rev limit reenable rpm

check:
fcmp/gt fr4, fr3 ! kph < upper_limit ?
bt/s enable_inj ! rev limit reenable
nop

mov.b @r4, r0
or #0x80, r0 ! turn injectors off, period
mov.b r0, @r4
bra End_Rvlm ! we're done here
nop

! ---------------------------------------------------------------------------

enable_inj:
fmov.s @r0, fr2 ! fr5 = lower limit
fcmp/gt fr4, fr2 ! is actual rpm lower than the injector re-enable value?
bf/s End_Rvlm
nop

fmov.s @r2, fr5 ! load MAP manifold pressure into fr5 mmHG
mova (map_lim), r0 ! load MAP
fmov.s @r0, fr1
fcmp/gt fr5, fr1 ! if MAP under 1.6 bar?
bf/s End_Rvlm ! makes sure MAP has fallen as well as rpm
nop

mov.b @r4, r0
and #0x7F, r0 ! turn injectors back on
mov.b r0, @r4

End_Rvlm:
mov.l (exit_revlimit), r2
jmp @r2 ! loc_22918
nop ! QED

rpm:
.word 0xA78C ! rpm memory location

map:
.word 0xA4E0 ! manifold absolute pressure memory location

rvlmbit:
.word 0xAD58 ! bit location for rev limiter bit (LSB high = disable injectors)

kph:
.word 0xA778 ! vehicle speed kph memory location

.align 2
upper_lim:
.long 0x45E5B000 ! 0x45E5B000 = 7350

lower_lim:
.long 0x45E29000 ! 0x45E29000 = 7250

upper_lim_lc:
.long 0x459F6000 ! 0x459F6000 = 5100

lower_lim_lc:
.long 0x459C4000 ! 0x459C4000 = 5000

kph_lc_limit:
.long 0x41000000 ! 0x41000000 = 8

map_lim:
.long 0x449D8000 ! 0x449D8000 = 1260mmHG (~1.6 bar)

exit_revlimit:
.long 0x22918 ! subroutine that is supposed to be called after this subroutine


I figured that the OR and AND operations on the rvlmbit set it to an on/off value...

From a hex logic table I found online, I get rvlmbit(injectors off) = AD58 OR 0080 = ADD8 and rvlmbit(injectors on) = AD58 AND 007F =AD58 Correct?

Could elaborate a little on the rvlmbit and the details of how it functions (I assume another fueling subroutine checks it before fueling) and the on/off values.

Also, I can't figure out the reason for the mov.b r0, @r4 after the AND/OR operations, and why they are sign extended.

Finally, why are RAM addresses in word and lacking the preceding 'FFFF'? How does the ecu determine ROM or RAM in an address below 0xFFFF?

And another question about definitions... <switch id="S63" name="Clutch Switch" desc="" byte="0x000121" bit="7"/> is 0x000121 the actual address in RAM? How is it referenced in assembly? Would you simply AND, OR, then compare to check it?

LSB = least significant bit. Actually that's an error. 0x80 is the most significant bit in the byte and that's what I had actually intended... whoops. I had used some incorrect nomenclature when I started disassembling and there are some remnants. This code is pretty old, well over a year.

The mov.b @r4, r0 puts what is in the address location r4 (in this case, the memory location of the rev limit on/off bit) into r0, so an OR operation can be performed on the first bit. Then it is stored back into @r4. The OR is to set the bit to the desired value. Since the injectors are either just on or off, they are controlled by a single bit.

"OR" and "AND" operations with a bitmask (usually an immediate value, like 0x7F or 0x80) are used to set bits on or off. Clutch on/off switch, defroster switches, etc. all do this.

This pattern for bit setting is used everywhere. The bits are often packed, like for CELs and such. Bit7 (MSB) might be the on/off bit for P0301, Bit6 might be for P0302, etc. The OR and AND operations to set them would vary. 0x80, 0x7F would be for the MSB, bit7. 0x40 would be for bit6, 0x20 for bit5, 0x10 for bit4, 0x08 for bit3.... 0x01 for bit0. Inversely, 0x7F, 0xCF.

For this to make sense you have to unpack it to a bit string.
h'80 = b'1000 0000
h'7F = b'0111 1111
etc.
Then the OR and AND instructions make sense. OR x, h'80 would set the first bit high, even if it already was set high, and not touch any of the remaining bits. AND x, h'7F would set the first bit low without touching any of the remaining bits.

You need to look at what is stored in r4 earlier in the code for it to make sense. r4 is first filled with a memory address. Then @r4 means it is actually going to the memory location. Literally, "the value at memory address r4". Not r4 itself. The entire ECU is based on this principle to address memory.

For instance, on the A2ZJ710J, FFFFA78C is the location of RPM (4byte float). Notice only the lower word is put into r3, but this results in r3 = 0xFFFFA78C . Then "@r3" means RPM.

Yes, the 32bit ECU does not always use the preceding FFFF for memory addressing. When you move a word into a register, the two highest bytes are set to FFFF anyway. I'm not sure why they do this. Maybe to save space. The later ROMs do not do this at all it seems.

"rvlmbit" is just referring to the "rvlmbit:" at the bottom of the code. The SH-2 is based on CODE/DATA, CODE/DATA blocks. So there is the CODE, and that code uses the DATA block immediately after to operate. This is because most instructions use relative addressing. Actual disassembly doesn't look like that. The "rvlmbit" gets replaced during assembly with a value. The value is the difference in relative location of that instruction and the location of the "rvlmbit:" in the data block. The code doesn't have to be written this way, I could count the difference myself and put in the value, but if I add anything in between that command and the data I would have to remember to update the relative address. So it is best to use a tag and let the assembler do this for me.

This will explain most everything:
http://freon.shackspace.com/car/ecu%20and%20tuning/openecu/renesas%20manuals/hitachi%20sh2%20software%20manual.pdf

This is also on Renesas's website. It is a free download.

Thanks for clearing the @ and logic operators.

I read a bunch of the software manual and I've got a much better grasp on SH, and now completely understand the assembly you posted.

One thing I couldn't find a definite explanation on was the bt instruction delay. The manual doesn't list it as delayed in the operation description, but since you followed one with a nop and only moved the rpm into the FPU once, I assume it is.

So, then this gets compiled and inserted where the hook at the original revlimiter code points, correct?

How is the hook implemented? Just "jsr (revlimiter disp)" followed up with a nop? Is it necessary to clear the rest of the old subroutine?

bt and bt/s are different. bt/s is the delay slot instruction. bt uses up the CPU cycles anyway because the CPU is pipelined. bt/s followed by nop is roughly the same as just bt when executed. It's not terribly important. There is more than one way to fry an egg, for sure.

To get the code into the actual ROM I use UltraEdit to paste my assembled object code into a previous unused segment (usually near the 7/8ths point or so). Then I go modify the call to go to my code instead of the old revlimiter routine. I do that part manually. In this case, I modify the value at 0x1122C from 0x29B54 to whereever I paste the code. 0x1122C is where the call to the revlimiter is. Sometimes this takes more doing, but that's roughly it.

Part of it is also making sure the modification doesn't modify any registers that might be passed through.

Alright sorry havent had a chance to get back to patching this untill today. Well just took out the spaces and got past that error just to run into another... This one reads "Source file is corrupt at byte 0x2bac5 (0xc expected, 0xff found)" . Tried some of my other roms with the same error. Let me know what I can do with this thanks.

Part of it is also making sure the modification doesn't modify any registers that might be passed through.

I suppose the only way to do this is to check a few subroutines after for using registers they don't assign, and then trial and error.

btw, not sure if you're aware but your shackspace is down.

Alright sorry havent had a chance to get back to patching this untill today. Well just took out the spaces and got past that error just to run into another... This one reads "Source file is corrupt at byte 0x2bac5 (0xc expected, 0xff found)" . Tried some of my other roms with the same error. Let me know what I can do with this thanks.

any chance the roms you are using are already modified (even though you might not be aware of it)? my patches work only on COMPLETELY STOCK FACTORY CODE. The patch will fail to merge into the rom even if 1 byte is not matching. If the patching tool says there is a difference, you can rest assure there is a difference.

any chance the roms you are using are already modified (even though you might not be aware of it)? my patches work only on COMPLETELY STOCK FACTORY CODE. The patch will fail to merge into the rom even if 1 byte is not matching. If the patching tool says there is a difference, you can rest assure there is a difference.

Hmm yeah my maps are stage 2+ street tunes but havent had any thing like LC added... I just tried my bone stock map and it seems to have worked except I cant use it since my car is modded and I need the street tune maps to keep it in line. I guess I'll just have to copy all my tables to the LC enabled stock map huh?

Hmm yeah my maps are stage 2+ street tunes but havent had any thing like LC added... I just tried my bone stock map and it seems to have worked except I cant use it since my car is modded and I need the street tune maps to keep it in line. I guess I'll just have to copy all my tables to the LC enabled stock map huh?

Yeah... that'll do it, ST is a big change in code. If you want LC with realtime, call cobb's garbage service and tell them to actually support their paying customers. I doubt you'll get anywhere though, as they don't even have the common courtesy to keep tabs on their StreetTuner support forums.

I patched my 06 wrx rom without any problems and it's anything but stock.

I patched my 06 wrx rom without any problems and it's anything but stock.

The data tables don't matter, its the ecu code.

I patched my 06 wrx rom without any problems and it's anything but stock.

stock refers to program logic.
Tuning boost, fuel, timing etc etc are data and these do not change the program logic.

stock refers to program logic.
Tuning boost, fuel, timing etc etc are data and these do not change the program logic.

Hmm yeah thats weird thats all I've messed around with is Boost, fuel, timing etc, and It kept comeing up with that error. Even tried a different 05 STi A2ZJB11J Rom with minor tunning with the same result it has to be a totally unaltered stock rom for the patch to work. But no biggy I can easily enough switch it all over and try it out. Thanks much.

I suppose the only way to do this is to check a few subroutines after for using registers they don't assign, and then trial and error.

It doesn't have to be trial and error. You can look back and forward for all references to each register. It's not that bad. Most of the time it is pretty clear because some register will be used for some menial task in one of the routines before the last jump. I minimize register use. If I need to use an extra register I'll have to check it. Most of the time I can use the existing registers in the subroutine I'm modifying. So as long as when I'm done the last value I set a register to is the same as before it is good to go. It is actually pretty rare a register is passed through too much code. r0-r8 and fr0-fr8 are so commonly used it doesn't take much searching. Sometimes it takes more to

btw, not sure if you're aware but your shackspace is down.
Shackspace is still working for me? I checked it yesterday briefly after you posted as well and it was up. It almost never goes down. I'd be surprised. The only trick is files over 25MB require (free) registration. Are you still having problems? Either way, you can go to Renesas's website and just search for the SH-2 sofware manual and 7055f/7058 hardware manuals. Free download there.

Alright so after sitting on my butt for a couple hours scratching my head I've figured out the problem. My car has the A2ZJB11J Rom and the stock 05 STI Rom I was able to patch the file to has the A2ZJB10J.... I downloaded the A2ZJB11J Patch... So I think you just labled the patches wrong. So if I PM you my e-mail could you please send me the (As it is right now) A2ZJB10J patch? Thanks.

why not just move everything over to the a2zb11j rom instead?
Should not take you anymore than 10 minutes, if.
And patch that rom.

why not just move everything over to the a2zb11j rom instead?
Should not take you anymore than 10 minutes, if.
And patch that rom.

Yeah I could but I'm slow with Romraider and just dont have the time. In no real rush anyway.

Alright so after sitting on my butt for a couple hours scratching my head I've figured out the problem. My car has the A2ZJB11J Rom and the stock 05 STI Rom I was able to patch the file to has the A2ZJB10J.... I downloaded the A2ZJB11J Patch... So I think you just labled the patches wrong. So if I PM you my e-mail could you please send me the (As it is right now) A2ZJB10J patch? Thanks.

patches for A2ZJB10J and A2ZJB11J are exactly the same, i provide them individually so it will be always 1<->1 relationship in order to avoid problems.

I'll say it one more time, the problem you have is that the ROM you are trying to patch HAS MODIFIED CODE, be it cobb, ecutek or whatever!

I'll say it one more time, the problem you have is that the ROM you are trying to patch HAS MODIFIED CODE, be it cobb, ecutek or whatever!

:eek::eek::eek:

patches for A2ZJB10J and A2ZJB11J are exactly the same, i provide them individually so it will be always 1<->1 relationship in order to avoid problems.

I'll say it one more time, the problem you have is that the ROM you are trying to patch HAS MODIFIED CODE, be it cobb, ecutek or whatever!

Alright sorry man I really wasnt trying to attack you just had an idea that I thought I'ed run by you. But if they're the same yeah I guess theres nothing I can do but manually transfer all the tables. Very weird though I know the code has never been messed with... I've always used Enginuity for the tunning and if that cant change the code I'm clueless as to what did. :confused:

Just copy over, if you're lazy, have someone else do it for you, lol

Alright sorry man I really wasnt trying to attack you just had an idea that I thought I'ed run by you. But if they're the same yeah I guess theres nothing I can do but manually transfer all the tables. Very weird though I know the code has never been messed with... I've always used Enginuity for the tunning and if that cant change the code I'm clueless as to what did. :confused:

Street Tuner changes the code........................................

fwiw, if you were tuning a ST map with enginuity, you were only changing the basemap.

Alright sorry man I really wasnt trying to attack you just had an idea that I thought I'ed run by you. But if they're the same yeah I guess theres nothing I can do but manually transfer all the tables. Very weird though I know the code has never been messed with... I've always used Enginuity for the tunning and if that cant change the code I'm clueless as to what did. :confused:

:huh: i didnt consider any of your saying as attack! I am only trying to make you understand what the real problem is and save you the time of looking into other things...

waiting for 07 wrx :)
weeee.

Street Tuner changes the code........................................

fwiw, if you were tuning a ST map with enginuity, you were only changing the basemap.

Yeah never used street tuner either. I did have an AP V.2 way back and used a stg. 2 map but never street tuned it. Just unmarried it and sold it and started tuning with Enginuity. So mabe the AP did it... Oh well.

:huh: i didnt consider any of your saying as attack! I am only trying to make you understand what the real problem is and save you the time of looking into other things...

Ok got it. ya I'm just haveing my tuner transfer all the tables over for me. I'm just not familiar enough with Enginuity to transfer everything over in a timely manner. So I might get to try it out tonight thanks for all your help.

I'm not too sure about this but, I believe when the AP puts the map back on, it's not the same exact map?

I'm not too sure about this but, I believe when the AP puts the map back on, it's not the same exact map?

I cant think of anything else that would have change the codeing around.... :confused:
But on a good note the LC works!!! Something needs adjusting though. Initially the car 2 steps quikly then boost comes on (Right up to 10psi) the limiter slows way down as it drops more RPM before kicking the the fuel back on. Then you release the clutch and she falls on her face from major fuel cut. To start off I set the disable MPH to 5 and RPM threshold to 2200 for a nice 5k limiter. So what needs some tweaking?

Yeah... that'll do it, ST is a big change in code. If you want LC with realtime, call cobb's garbage service and tell them to actually support their paying customers. I doubt you'll get anywhere though, as they don't even have the common courtesy to keep tabs on their StreetTuner support forums.

Yeah never used street tuner either. I did have an AP V.2 way back and used a stg. 2 map but never street tuned it. Just unmarried it and sold it and started tuning with Enginuity. So mabe the AP did it... Oh well.



Ok got it. ya I'm just haveing my tuner transfer all the tables over for me. I'm just not familiar enough with Enginuity to transfer everything over in a timely manner. So I might get to try it out tonight thanks for all your help.

My bad, thought you meant Street Tuner when you said 'street tunes'.

I find it easier to copy tables using ecuflash. Just load both roms, open a table you have modified on both roms, and copy & paste. I've done it a few times and I have tuned alot of tables, doesn't take too long once you get the hang of it.


I just had a friend test my 32bit (AD8H200X 06 WRX) Launch control since I'm in a cast now...

Set the LC limit to 4700, it made 10 Psi, and I didn't see any timing pulled. :D

Can't wait to get healed up and get it to the track. :)

But on a good note the LC works!!! Something needs adjusting though. Initially the car 2 steps quikly then boost comes on (Right up to 10psi) the limiter slows way down as it drops more RPM before kicking the the fuel back on. Then you release the clutch and she falls on her face from major fuel cut. To start off I set the disable MPH to 5 and RPM threshold to 2200 for a nice 5k limiter. So what needs some tweaking?

this is what you need to tweak ;) http://forums.nasioc.com/forums/showpost.php?p=21558460&postcount=42

this is what you need to tweak ;) http://forums.nasioc.com/forums/showpost.php?p=21558460&postcount=42

Awsome thats exactly whats happening. Boost builds up to 9-10psi and it seems to start limiting slower waiting for boost to come down. Thanks again. Let me know if you you come out with FFS later on down the road. ;)

LC FTW! btw the whole car shakes like mad!

LC FTW! btw the whole car shakes like mad!
My wagon is shaking like crazy also lol

LC FTW! btw the whole car shakes like mad!

Chicken! :banana:

Following patches are now available

LAUNCH CONTROL FOR USDM WRX MODELS

* SUBARU IMPREZA WRX MY06 A8DH200X
* SUBARU IMPREZA WRX MY06 A8DH201X
* SUBARU IMPREZA WRX MY07 A8DK100X

LAUNCH CONTROL FOR USDM STI MODELS

* SUBARU IMPREZA STi MY04 A2ZJ710J
* SUBARU IMPREZA STi MY05 A2ZJB10J
* SUBARU IMPREZA STi MY05 A2ZJB11J
* SUBARU IMPREZA STi MY06 A2ZJE11J
* SUBARU IMPREZA STi MY07 A2UG000J
* SUBARU IMPREZA STi MY07 A2UJ000J

http://tinywrex.wordpress.com/2008/03/28/new-32bit-ecu-patches-available/

Awaiting the 06' LGT :)

Awaiting the '05 LGT (MT) :)

(I'll post the firmware ID shortly. Or the image itself if that would help.)

USDM Legacy GT roms that i am working
05
A2WC500N
A2WC510N
A2WC521N
A2WC522N
06
A2WF200N

these are the ones i currently have, if there are any other revisions available let me know

On the LGT roms, A2WC521N, A2WC522N, and A2WF200N, merchgod modified the rom to eliminate the time delay in the WGDC. I don't think he changed the logic. I think he just changed values in multiple locations.

For the 32 bit ecu's are you selling pre-modified roms? Or a patching program to modify our existing rom?

For the 32 bit ecu's are you selling pre-modified roms? Or a patching program to modify our existing rom?

patching program and patch file, so you can modify your existing rom

Patches for Legacy GT models now available

LAUNCH CONTROL FOR USDM LEGACY GT MODELS

* SUBARU LEGACY GT MY05 A2WC500N
* SUBARU LEGACY GT MY05 A2WC510N
* SUBARU LEGACY GT MY05 A2WC521N
* SUBARU LEGACY GT MY05 A2WC522N
* SUBARU LEGACY GT MY06 A2WF200N

Patches for Legacy GT models now available

LAUNCH CONTROL FOR USDM LEGACY GT MODELS

* SUBARU LEGACY GT MY05 A2WC500N
* SUBARU LEGACY GT MY05 A2WC510N
* SUBARU LEGACY GT MY05 A2WC521N
* SUBARU LEGACY GT MY05 A2WC522N
* SUBARU LEGACY GT MY06 A2WF200N

:D.....

Finally got it installed and working on my 07, set to 4500RPMS building 5-7psi :)

Thanks so much tinywrex.

Haven't actually gotten the balls to try it yet.

Alright sorry man I really wasnt trying to attack you just had an idea that I thought I'ed run by you. But if they're the same yeah I guess theres nothing I can do but manually transfer all the tables. Very weird though I know the code has never been messed with... I've always used Enginuity for the tunning and if that cant change the code I'm clueless as to what did. :confused:

There may have been some left over residual code from when I altered the actual HEX file to protect my tune in the past. I though I had removed it from Forsfed25's ROM but I might have missed a bit somewhere.

I copied every table into a stock ROM and he is off and launching!

I don't know what the problem is, but for the life of me I can not edit the definition for my ecu in ecuflash :( Anyone know what I am doing wrong. Says file path cannot be found....

I don't know what the problem is, but for the life of me I can not edit the definition for my ecu in ecuflash :( Anyone know what I am doing wrong. Says file path cannot be found....

Ok help us help you, where do you see the error message File path cannot be found?

Ok help us help you, where do you see the error message File path cannot be found?
As I go to add the commands for my rom in the ecuflash defs. I open my XML file with notepad but it will not let me save the updated file. Using Vista by the way.

As I go to add the commands for my rom in the ecuflash defs. I open my XML file with notepad but it will not let me save the updated file. Using Vista by the way.

Vista is the :devil:

Windows Vista sets access permissions on the Program Files directories and will not allow a user to write to those directories. Ecuflash happens to be under Program Files so no write access for you :)
Turn UAC back on, just go to Start > Control Panel > Security > Security Center > Other Security Settings.

Vista is the :devil:

Windows Vista sets access permissions on the Program Files directories and will not allow a user to write to those directories. Ecuflash happens to be under Program Files so no write access for you :)
Turn UAC back on, just go to Start > Control Panel > Security > Security Center > Other Security Settings.
haha
Thanks, I figured it was a Vista related thing, I just didn't have the time to load the needed programs to my PC and try it last night. I will try changing the security settings or maybe just the whole program to a different location all together. Thanks again, looking forward to neck snapping launches :banana:

I downloaded the 32 bit patch for my USDM 07 STI today, order number 080401-152145-3844.

However, when I go to use it, it gives me the following error:

Source file is corrupt at byte 0x32158 (0xd1 expected, 0x4f found)!

I think I may have ordered the wrong one.I got the A2UG000J patch. Looks like I may have an A2UJ000J ROM, instead. I am currently using the XPT stage 2 tweaked by Gabedude. I'm just not sure if I got the wrong patch, or if XPT maps are tweaked in a weird way...

Is there any way I could get the other patcher? If so, please let me know. Thanks for all the work you put into this, I'm looking forward to trying it out.

I downloaded the 32 bit patch for my USDM 07 STI today, order number 080401-152145-3844.

However, when I go to use it, it gives me the following error:

Source file is corrupt at byte 0x32158 (0xd1 expected, 0x4f found)!

I think I may have ordered the wrong one.I got the A2UG000J patch. Looks like I may have an A2UJ000J ROM, instead. I am currently using the XPT stage 2 tweaked by Gabedude. I'm just not sure if I got the wrong patch, or if XPT maps are tweaked in a weird way...

Is there any way I could get the other patcher? If so, please let me know. Thanks for all the work you put into this, I'm looking forward to trying it out.

YHPM

for any issue regarding the patches you can contact me by email

I downloaded the 32 bit patch for my USDM 07 STI today, order number 080401-152145-3844.

However, when I go to use it, it gives me the following error:

Source file is corrupt at byte 0x32158 (0xd1 expected, 0x4f found)!

I think I may have ordered the wrong one.I got the A2UG000J patch. Looks like I may have an A2UJ000J ROM, instead. I am currently using the XPT stage 2 tweaked by Gabedude. I'm just not sure if I got the wrong patch, or if XPT maps are tweaked in a weird way...

Is there any way I could get the other patcher? If so, please let me know. Thanks for all the work you put into this, I'm looking forward to trying it out.

I had the same prob when patching my 05 STI. Turned out that the Access Port I had on it a few months back changed the codeing in my rom. Luckily I had a Stock rom off a 05 STI that had never been married to an AP and that rom patched fine. So if you've ever used an AP on your car thats prob why your getting that error, or mabe your XPT maps are doing the same thing. So try and find a 07 STi rom thats never been altered. ;)

try scoobypedia.co.uk for stock roms

FFS+LC is in testing phase. Next step once tests are complete is to make all parameters changeable in real-time (via PC)

FFS+LC is in testing phase. Next step once tests are complete is to make all parameters changeable in real-time (via PC)

Awesome!

How is the logic set up?

FFS+LC is in testing phase. Next step once tests are complete is to make all parameters changeable in real-time (via PC)

Get at me if needed! :D

xswrex,

anyword on the A2ZJ700J rom for 2004 STI?

I'm assuming this means you're actually changing LC over to a proper clutch switch rather than a speed threshold?

I'm interested in this development, just because it's cool (I have a UTEC, so it doesn't really matter :p)

Awesome!

How is the logic set up?


! if vehspeed < speed threshold then
! lc limiter
! else
! if clutch then
! ffs limiter
! else
! stock limiter

Get at me if needed! :D

will get back to you if needed :lol:

xswrex,

anyword on the A2ZJ700J rom for 2004 STI?

yes, the 2 older 04 STi revisions will be available end of next week.

I'm assuming this means you're actually changing LC over to a proper clutch switch rather than a speed threshold?

I'm interested in this development, just because it's cool (I have a UTEC, so it doesn't really matter :p)

maybe i am wrong but clutch for LC is not needed, but the speed threshold is

Oh it's certainly not needed, it's just the 'proper' way to implement it, because it aviods a situation where, through clutch slip, the LC rev limiter would be hit but the speed threshold is not yet hit. For most people, the speed threshold way is most likely more than good enough.

And of course you'd still have a speed threshold (else how does it know to LC limit and not FFS limit?), it would just be hard-coded to like 3MPH, rather than being user tunable.

Oh it's certainly not needed, it's just the 'proper' way to implement it, because it aviods a situation where, through clutch slip, the LC rev limiter would be hit but the speed threshold is not yet hit. For most people, the speed threshold way is most likely more than good enough.

And of course you'd still have a speed threshold (else how does it know to LC limit and not FFS limit?), it would just be hard-coded to like 3MPH, rather than being user tunable.

The difference is that you consider the speed threshold as the toggle between LC/FFS, whereas i ALSO consider it a parameter for launching. The car will hold the LC limiter until it breaks the speed threshold,
whereas
what you say is that the LC limiter will only work when you have the clutch pushed down. The moment you release the clutch (doesn't matter if you side step or slip, the switch will toggle) the normal limiter will engage.

Yep. The question is which system is better? As I said, I'm sure for 90% of the people out there, the speed limit is more than enough, but for some, they need to keep the RPM's down for first contact with the clutch to avoid wheelspin, but then will need them to increase as they slip the clutch to get a good launch (say if they have a carbon clutch, which gets stickier with slip rather than less sticky).

There's also the question of where the car considers the cltuch to be depressed. I don't know where this point is in terms of how disengaged the clutch is.

xswrex,

anyword on the A2ZJ700J rom for 2004 STI?

A2ZJ500J and A2ZJ700J are now available

Alright finally got around to takeing some videos of Tinywrex's LC enabled on my USDM 05 STi. WORKS GREAT!

Cant wait to get a hold of the FFS when it's released! Thanks agian.

http://youtube.com/watch?v=KufRWTYISRs

http://youtube.com/watch?v=Tjncrh6xd0s

Any plans for LC on the USDM 05+ RS (non-turbo) models?

thanks as always!

Any plans for LC on the USDM 05+ RS (non-turbo) models?

thanks as always!

why don't you email me this type of ROM? then i can try to make an LC patch for it ;)

yes, the 2 older 04 STi revisions will be available end of next week.

Great work! Where can we get them?

Great work! Where can we get them?


http://tinywrex.x10hosting.com/online/index.php?act=viewProd&productId=47

http://tinywrex.x10hosting.com/online/index.php?act=viewProd&productId=48

Hows the FFS development going?

This should be moved to the "Commercial ECU Reflashes" forum.

I'm not debating that this was the correct place to post when the thread was originally started.

i disagree. there's nothing commercial about this. simply one man wanting compensation for his knowledge & efforts.

Any plans for LC on the USDM 05+ RS (non-turbo) models?

thanks as always!

LC now complete for the USDM 2005 RS! I can't wait to try it on my customer's car.

bump for being released!!! in need for my 07wrx!!!