Asm error:error[118]

[Citius]

I'm in a rush and need to solve a frustrating problem, so I thought I'd pick your brains...Normally I like challenges, but time is against me!! I've no time to play Sherlock Holmes tonight...

I running 3.4.0.8 with 2.0.0.5 IDE on Windows Vista Home Premium programming a 18F45k20. I've been running 3.4.0.8 since it came out back in Octoberish.

I have been working on a project that I know inside out and had the code to the point that it is working flawlessly for over a month - But I opened it after a few days break to make some small tweaks. It won't compile at all, even previously working versions. It worked perfectly just last week. I get a series of errors as follows: ASM ERROR:ERROR[118] C:\USERS\ME\APPDATA\LOCAL\\TEMP\A.S 8643: Overwriting previous address contents (0000) etc...).

At first I though I had some stupid coding problem, such as deleting some vital line of code...but I've reduced it to the ridiculously simple, and compiler is still not working.

My code is somewhat long, but same code compiles just fine on my XP system and states it is only 35% full and RAM is at 25%.

So I reduced the code to the following, with the same results:ASM ERROR:ERROR[118]...

DEVICE 18F45K20 'PIC Microcontroller 18F45K20
Main:
DELAYMS 100
GOTO Main

I have un-installed both the ProtonIDE, and the Crownhill folder, and re-installed. I even tried un-installing MPLAB IDE, and then re-installed all starting with 3.4.0.5, then upgrading to 3.4.0.6, then adding in the patch 3.4.0.8. I also re-installed the IDE 2.0.0.5 - All of this TWICE.

Nothing has solved this yet. It recognizes my dongle just fine, as it complains when I unplug it and won't compile at all, saying it misses sparky...

Any suggestions (except using my ancient XP system, which allows me to take long coffee breaks between copiling small code sections )

I haven't made any major changes lately, and have used Proton for years with COMPLETE satisfaction and ease...

All seems to point to the MPASM, but why? and how can I fix this...

Time and help much appreciated...


Carl

P.S. - Every time I'm down in my lab working on a project and using Proton IDE with Proton Compiler, I wonder why anyone would use anything else...this is simply the best supported, best product out there. I read an old post the other day (the poll) that was questioning Proton...The post was annoying...Les truly has made me a lot of money...way way more than I've made him...and the forum members here are generous, wise, and most of all, entertaining...A great group that I enjoy interacting with...Just know you're all appreciated, even if I don't post all that often...

[Citius]

49 views and no responses...I take it no one else has seen this problem? Any suggestions? I'm getting ready to re-format my entire computer - but seems a drastic solution to a problem - it is my "nuclear option"

Any suggestions prior to this option would be greatly appreciated...

Carl

[Citius]

Here is more info:

This Does not compile!

Device 18F452
Main:
DelayMS 1
GoTo Main

This does compile just fine!!

Main:
DelayMS 1
GoTo Main

Error in case of first is this:

Message[308] C:\USERS\CARL\APPDATA\LOCAL\TEMP\A.S 11 : Warning level superseded by command Line value. (2)
Error[118]   C:\USERS\CARL\APPDATA\LOCAL\TEMP\A.S 1206 : Overwriting previous address contents (0000)
Error[118]   C:\USERS\CARL\APPDATA\LOCAL\TEMP\A.S 1206 : Overwriting previous address contents (0001)
Error[154]   C:\USERS\CARL\APPDATA\LOCAL\TEMP\A.S 1206 : Each object file section must be contiguous (section .config_000000_A.O)
Error[118]   C:\USERS\CARL\APPDATA\LOCAL\TEMP\A.S 1206 : Overwriting previous address contents (0000)
Error[118]   C:\USERS\CARL\APPDATA\LOCAL\TEMP\A.S 1206 : Overwriting previous address contents (0001)

Anyone...any suggestions???? Again, I'm running 3.4.0.8 on 2.0.0.5 - all was fine until 1-week ago

Carl

[wastrix]

I have compiled all the code you have said did not work and it compiled fine - no errors at all.

I also use v3.4.0.8.

Can you post the ASM or LST files? (If they exist)

[Citius]

Wastrix -

I knew I could count on you to offer up help...many thanks...your regular advice on the forum is always thoughtful and sincere...

I know the code snippet will compile on anyone's system...All of my much more complicated code compiles on my other system as well...but not them, nor even this simplest one, even under the same exact conditions on my best system (Vista)...It is like something in the install is missing, but I've re-installed many times. I'm not new...I've installed each new beta version as they've become available on 4 systems I regularly use [Work(Windows 2000), Home Lab#1(Vista), Home Lab#2 (XP) and Laptop(XP)] and never have any problems. The Beta's have worked great, particularly 3.4.0.8. It had been running fine on this system (Vista) since October (& almost every single day since) without a hitch or slightest sign of any flaws...until Sunday...Now nothing, no matter how simple, will compile.

List & ASM included

Your advice / suggestions are greatly appreciated!

Carl

[Tim]

Looking at the lst file its a config issue

Try adding your own configs

[Citius]

Tim -

Thanks for your input Tim - In practice, I always set my config words to avoid guessing what state they are in.

In the actual much more complicated code that compiles on 3 of my systems, but not on the fourth, I'm using the 18Fk4520. I use a hardware include file that I've not changed since early December. In that include file, I define the device, set my TRIS, ADCON, Define Aliases, etc and I've set every single config word to my desired state - carefully based off of the data sheet and plenty of prior experience with this micro.

Since I've not changed my config words in 2 months, but the code stopped working 3 days ago...I ruled out that possibility (I will certainly double check tonight when I get home, but I tend to suspect something else).

I guess I was hoping someone would say, "I've had this same problem before on my system and here is how I fixed it..."

As a side note, I've recently installed a 24-bit C compiler on my system...but I tend to doubt that could have screwed me here. I only mention it because I saw a post Bruce made a while back that hinted at such a problem.

Carl

[pverburg]

you might not have changed the file but another program or gremlin might have.

Why not copy the directory's from a working machine in to your problem one rename the original's first.

if it fixes it its in those if not then ?????

Paul V.

[Alan R.]

This happened to me.

An update some time ago changed some of the default congfig settings - I think it is actually in a Microchip file. This means that some programs that did not set the config fuses no longer work - this includes some of the sample programs. Bottom line is to set the config fuses every time.

-AR..

[Citius]

Paul -

Thanks...I like your idea and will try tonight and let you know the results. Thanks!

Alan -

Thanks for the input! Good thought, though I do set each config word...But it gives me a hint at where the problem could lay...

Here are my config words for my code:

Device 18F45K20      'PIC Microcontroller 18F45K20
'18F45K20 CONFIGURATION REGISTER (SET by CHS 09-14-2009)                 
Config_Start
    FOSC = INTIO67  'Internal oscillator block, port function ON RA6 AND RA7
    FCMEN = OFF     'Fail Safe Clock Monitor Disabled
    IESO = OFF      'Int/Ext Oscillator Switch Over Disabled
    
    PWRT = OFF      'Power Up Timer Disabled
    BOREN = OFF     'Brownout Reset Disabled
    BORV = 18       'Brownout RESET Voltage: Minimum setting (1.8v)
    
    WDTEN = OFF     'Watchdog Timer Disabled
    WDTPS = 256     'Watchdog Postscaler: 1:256
    
    MCLRE = OFF     'E3 INPUT pin Enabled/Disabled; MCLR Enabled/Disabled
    LPT1OSC = OFF	'Timer1 configured for higher power operation
    PBADEN = OFF    'PortB AD functions OFF (PORTB<4:0> pins are configured AS digital I/O ON RESET)
    CCP2MX = PORTC	'CCP2 input/output is multiplexed with RC1    
    LVP = OFF       'Low Voltage ICSP Disabled
    Debug = OFF     'Background Debugger Disabled
    XINST = OFF		'Instruction set extension and Indexed Addressing mode disabled
    STVREN = OFF    'Stack Overflow Reset Disabled
    
    CP0 = OFF ' Code Protection Block 0 Disabled
    CP1 = OFF ' Code Protection Block 1 Disabled
    CP2 = OFF ' Code Protection Block 2 Disabled
    
    CPB = OFF ' Boot Block Code Protection Disabled
    CPD = OFF ' Data EEPROM Code Protection Disabled
    
    WRT0 = OFF ' Write Protection Block 0 Disabled
    WRT1 = OFF ' Write Protection Block 1 Disabled
    WRT2 = OFF ' Write Protection Block 2 Disabled
    
    WRTB = OFF ' Boot Block Write Protection Disabled
    WRTC = OFF ' Configuration Register Write Protection Disabled
    WRTD = OFF ' Data EEPROM Write Protection Disabled
    
    EBTR0 = OFF ' Table Read Protection Block 0 Disabled
    EBTR1 = OFF ' Table Read Protection Block 1 Disabled
    EBTR2 = OFF ' Table Read Protection Block 2 Disabled
    
    EBTRB = OFF ' Boot Block Table Read Protection Disabled
Config_End

As you can see, I address each config word, and settings have worked just fine for some time.

I will review data sheet again tonight to see if I can see a problem, but I suspect it is elsewhere...

Thanks both for you time and thoughts...

Carl

[Citius]

I've tried all suggestions...all without luck. Compiler will not compile even the Sample Programs in the PDS folder. (I should clarify that it compiles, but does not assemble)

I have un-installed & re-installed compiler & IDE 5 times now. As a last resort, I even tried to copy over the compiler directory from one of my good systems to replace the entire compiler directory on my Vista system - this did not work either. Same beta and IDE are installed on all 3 of my other systems and work perfectly.

My Vista system that has run the Beta 3.4.0.8 with IDE 2.0.0.5 flawlessly almost every single day since October simply will simply not compile code now.

I am at a loss and can see no hope except to re-format my entire computer and hope that when that is done, that the install process will work.

Hope this never happens to any of you, as it is extremely frustrating.

Any last thoughts before I nuke my entire system? Any help or suggestions would be greatly appreciated!

Carl

[billyminor]

Why not try a system restore first before reformatting ,just pick a restore point about 2 weeks ago

[Citius]

Brian -

Thanks for the suggestion, but I've already tried this as well, restoring my system to Jan 26, the day before my compiler stopped working - to no avail.

I may give it one more shot to an earlier date (out of desperation - not because I think it will work). But I appreciate the suggestion none-the-less.

Carl

[billyminor]

Sorry that didn't work,if you do reformat don't forget to unplug that dongle before re-installing anything

[pic-ignorant]

Is it possible that the C compiler has messed up MPasmwin, or changed some defaults?

John

[Citius]

John -

Thanks for your time John...I consider that a possibility, though somewhat slight, since I installed the 24-bit C compiler (mikroE) 3-weeks before my good compiler stopped working. I have disabled it (& finally removed it), but that did not help. I've also un-installed and re-installed all previous version of Microchip's IDE, and re-installed the latest version (version 8-billion-430-thousand-something).

I ran a registry cleaner program shortly before I had my malfunctions, but I assume my 6 attempts at un-installing & re-installing (starting from my Proton CD (3.1.0.9), Sparky 64-bit driver, IDE 2.0.0.4 then 3.4.0.5 then IDE 2.0.0.5 then 3.0.4.6 patch, then 3.4.0.8 patch) would have overcome any registry errors.

I suppose I could have somehow buggered up my install , but 6 times?...doesn't seem likely, and I've done this on all my other systems without a hitch. (I will be thrilled when Les releases an install package!!)

Also my system restore was to prior to my Registry Clean - and I un-cleaned my registry as well.

The other change to my system was Microsoft's update to fix the IE6 G-mail error, but others would have seen an error too if that was the cause.

Anyone have any more ideas? All your help is GREATLY appreciated ...I can't wait to move beyond this issue!


Carl

[C_Moore]

Have you tried to run it in compatibility mode?

[pverburg]

what I sometimes do is get a new disk install windows and the app if that works your software is corrupt if not hardware problem.

other option can you run the asembler manualy ?

Paul V.

[wastrix]

How did you uninstall?

Did you actually delete the Crownhill folder or did you use the uninstaller in the "Add Remove Programs"/"Programs and Features" thing?

What happens when you run \Program Files\Crownhill\PDS\Includes\MpasmWin.exe ?

BTW, did you set the fuses manually or use the Fuse Configurator?

Oh, and maybe have a look at the MPLAB.INI file in the same directory as the above.

[Citius]

WASTRIX!!!

You did it my friend! It was the stupid Mpasmwin.exe file in the PDS\INCLUDES file. I saw two files in there

1.) Mpasmwin.exe (8.108MB)
2.) Mpasmwin-orig.exe (7.544MB)

I re-named the Mpasmwin.exe to Mpasmwin_junk.exe, then re-named the Mpasmwin_orig.exe to Mpasmwin.exe....Compiled my code and it compiled and assembled correctly!!!

I don't understand why, as this is part of the file I downloaded from the Beta install and , but I'm so happy to have my system back that I don't give a damn - maybe running his in compatibility mode as Charlie suggested would work as well, though why the assembler worked from October to Jan27th, is a mystery!

Thanks to all your your valuable time, and thoughtful/helpful suggestions.

By the way, I always use JGB's Fuse Configurator Plugin, a brilliant app, and another reason I knew it wasn't my config words.

So to anyone who ever has the following error:

Error[118]   C:\USERS\CARL\APPDATA\LOCAL\TEMP\A.S 1206 : Overwriting previous address contents (0000)

Consider the above fix as a possible solution. (Also note that this same error could be caused by incorrect ORG statements - though this was not the case for me.)

There are 3 reasons this is the greatest compiler on the market:

1.) A VERY STRONG and well written compiler that is constantly updated and strengthened thanks to Les, to whom I feel indebted to every time I get paid for a project.

2.) A FANTASTIC forum that is second to none thanks to all of you who are generous in your time and energy and thought-out suggestions. The time and effort you put into this forum to me is the greatest asset. What good is a great compiler that you don't know how to use and have no one to turn to for advice.

3.) Community supported Library manager, pluging aps, etc that are world class thanks to swright, JGB, Tim, normnet, Toley and others, that make our life so much easier.

(not to mention floating point math, graphic LCD support, and all the other things that we all love so much about this compiler)

It is ironic...I started PICs in 2001 with Assembly, then moved to PBP, but left that after 2 long years after frustration with the poorly supported compiler and more importantly to me, the forum, where people were out to prove how stupid you were. Then I found Proton, where the two weaknesses of the other were the strengths of this one.

Whenever I'm stuck on a problem, I come to the forum, not for a solution to my particular problem, but more to be in a community surrounded by people who face difficult problems and with force of mind and effort, solve their problems. It is always inspiring, and I can return to my problem with confidence that any problem can be solved with will and effort (and enough coffee)

Cheers to all!!

Carl

[Citius]

As a follow-on note, here is what I've seen:

MPASM version 5.30 (current version that installs with 3.4.0.8) does not assemble correctly

MPASM version 5.35 (M'chip's latest version) does not assemble correctly

MPASM version 5.21 (labled MPASM_orig in the 3.4.0.8) works perfectly

Carl

[wastrix]

Good to hear. Will have to remember this in case it happens to anyone else.

The older MPASM will not work on some certain new devices (not sure which ones these are) but the old one is a lot faster, and as you say, it is the one that actually works [].