MCS is Crashing

[fowardbias]

I have had the MCS shutdown several times (gray-out) with a error screen. I believe this may have something to do with the file saving path. If I manually save my work every so often I have not experienced any problems. If I don't manually save the file and an auto save works is where I believe things go awry. I usually save to the factory default area under "My PBP3 Programs". Most of the time I will save the new rewrite to a flash drive under "Save As" and use this to copy over the old file on the same computer and over the same file on another computer. I am using all the latest upgrades and have had this happen several times after having 3 pages of code written on Toshiba running W7 and a newer Dell running W10. I'm sorry I didn't copy the error code. ( It said an exception error has happened with an address and MCS must shut down). I hope this helps someone to fix this issue. It is very annoying. Now I manually save after every page rewrite. Cheers.

[HenrikOlsson]

Don't worry about the error message, it's been posted before but the developler doesn't seem to care all that much since he's never seen it happen and "we" can't provide a way to reproduce it at will. It's all a big mess and nothing seems to get done.

I've never thought it might have something to do with the save path or auto save or anyting like that. Lately I've been leaning towards the serial terminal beeing tha culprit but I don't know. Did you happen to have the serial terminal running when it crashed for you?

/Henrik.

[fowardbias]

No serial port was in use during programming writing. I would usually code a page and run test compile, and if good, add more lines. Sometimes I didn't use the thumb drive or any other connections to the laptop and it would still drop out. Hope this helps. So for the time being, no fix available and I'll manually backup.

[picster]

cross-thread posting, I have two Dell laptops (one I5, one I7) with W7x64 running. BOTH experience this problem. Sure would be nice to have a fix. Usually seems to happen while typing rem comments in - anyone else see this trend, or is it just me?

[HenrikOlsson]

Can't say I can relate to that. Have been coding on/off for short periods, 20 minutes here, 2 hours there during the last week, writing a lot of comments, and haven't had a single crash. Same file(s) as I've been working on Before, on the same machine etc etc but not a single crash now while it can crash 5 times in 10 minutes on a moody day.

/Henrik.

[picster]

Yeah I've seen that frustrating lack of consistency. Next time it crashes, see if you can make a mental note of WHAT you were typing. EVERY crash I have had occurs during typing, despite the fact that I can easily leave the program up and running for a day or two between sittings.

[mark_s]

I never had any problems with crashing, until recently. I followed Henrik's posts in the past, regarding this issue, but had nothing to add.

Last month it crashed on my work bench laptop. After rebooting the pc it seemed ok and no more crashes. It's a gateway 486 running the last version of XP. This pc is never connected to the net and been stable, running PBP2.6 and 3 for at least 5 years.

A few days ago, I plugged my usb flash drive into the laptop and opened a pbp file in MCS. I compiled the file without problem. Then I saved the file to my MSCX folder. I started to modify a line, and got the error message. After rebooting it would not allow me to modify that file without crashing.

Now days later, I can open the file from the MSCX folder without any crashing.

For me it seems to be a path problem as Forwardbias stated above. Both times that I had problems it was while using a flash drive.

One thing my case indicates is that it has nothing to do with which version of Windows you are running. Same problem on XP, 7 and 10

[picster]

NO crashes so far since I changed the default settings for comments to NON italics. All my crashes were during typing documentation remarks, and I figure it was processing them on the fly and changing the font with every keystroke... since I changed it to only change the COLOR (but not the font characteristic to italics), I have NOT experienced a crash.

I would appreciate if others could try this and continue to test concurrently with me.

[mark_s]

Ok, I switched to non-italics on both MCS and MCSX.

Thanks

[Demon]

Mark, sorry I missed this thread back in March.

I would have been curious to know what would have happened if you editted that bad line in NOTEPAD, and then compiled that from MCS.

Notepad removes all formatting.

Robert

[tumbleweed]

I would have been curious to know what would have happened if you editted that bad line in NOTEPAD, and then compiled that from MCS

Not much. Any formatting like fonts, colors, or italics is strictly in the IDE display. The source files are already plain-text.

[mark_s]

Hi Robert, it's actually Picster that beleives the crashing is from using italics in the IDE. I was trying to change a DEFINE OSC statement and every time I hit a key it would crash. This happened only to the file from my flash drive. I could edit files on my hard drive? I switched to non italics, just in case this actually is the cause.

I think a real test would be to leave italics checked. When a crash occurs switch to non-italics and see if it stops.

If you read through both threads on this subject. You will find there is no pattern or consistancy, so difficult to trouble shoot.

[MichelJasmin]

Do you copy some text from the datasheet PDF and then paste to MCS? Not sure but I used to had crashes when doing it...

[picster]

Do you copy some text from the datasheet PDF and then paste to MCS? Not sure but I used to had crashes when doing it...

Maybe the windows clipboard is including the font formatting and it's also messing up the in-line font processing?

[Demon]

I just did a test on Windows 10 64bit home edition. I always used italics and colours for comments with Win XP 32bit, Win7 64bit and now Win10 64bit (default MCSP settings), never a problem.

Copied in text from PDFs, made them comments, compiled no problem.
Copied in fuse settings and made them code, compiled no problem.

I checked my MCSP shortcut, I do not run it as administrator or any other compatibility setting.

There must be more involved, we're still missing something.

Robert
MCSP v5.0.0.5
PBPX v3.0.8.4

[Demon]

i have only one idea, did you install MCSP and PBP as administrator?

I might have, don't remember. It might be possible for the install to be incomplete if the installation package did not have necessary admin rights. I do know several software have varying levels of problems with Win10.

I just cannot install properly PaintShop Pro XI using Win10 64bit. Looks like I'll be forced to bite that bullet and get and upgrade even though I am very content with that old version.

Robert

[picster]

I just did a test on Windows 10 64bit home edition. I always used italics and colours for comments with Win XP 32bit, Win7 64bit and now Win10 64bit (default MCSP settings), never a problem.

Copied in text from PDFs, made them comments, compiled no problem.
Copied in fuse settings and made them code, compiled no problem.

I checked my MCSP shortcut, I do not run it as administrator or any other compatibility setting.

There must be more involved, we're still missing something.

Robert
MCSP v5.0.0.5
PBPX v3.0.8.4

Compiling has never caused problems for me, NEVER had any when sitting idle, and ***always always always WHILE TYPING!!!*** This strongly suggests to me that there's some issue with processing the keystrokes, and the text on the screen is obviously continuously "reassessed" to display it in the configured font/color. That's my "working theory". As I recall, there was a suggestion about video drivers as well, a while back - Nvidia being questioned as potentially part of the issue, but I couldn't test it on a non-Nvidia machine (although I tried running with my alternate "slow" graphics card on the dual-graphics equipped machine and it still crashed).

[picster]

Hi Robert, it's actually Picster that beleives the crashing is from using italics in the IDE. I was trying to change a DEFINE OSC statement and every time I hit a key it would crash. This happened only to the file from my flash drive. I could edit files on my hard drive? I switched to non italics, just in case this actually is the cause.

I think a real test would be to leave italics checked. When a crash occurs switch to non-italics and see if it stops.

If you read through both threads on this subject. You will find there is no pattern or consistancy, so difficult to trouble shoot.

Note that a DEFINE statement is also likely in a specified "different" font/color, causing it to be processed "on the fly" as you're typing... same problrem, different manifestation. Were you using italics for Defines?

[Demon]

I've always used default MCSP settings.

My laptop used is a HP Pavillion i5 Intel/nVidia.

I'll try later typing a long-winded comment and see how that goes.

Robert

[Demon]

Nope, long-winded comments, a DEFINE OSC 10 thrown in, more comments, no problem.

Robert

[picster]

Nope, long-winded comments, a DEFINE OSC 10 thrown in, more comments, no problem.

Robert

Perhaps I'm not approaching this with the same attack...

If this was a "sure thing" that ALWAYS happened, it would be pretty easy to resolve. However, my experience is it's very hit and miss - hence I don't believe a crash can be "forced", there are other undetermined variables involved, but... again... because it ALWAYS when I'm typing, I believe that says something. It can sit idle for DAYS and nothing happens. I'm wondering if others who are experiencing this can avoid crashes by trying this change, thereby contributing to the validation of the theory - not try to FORCE another crash when it's unpredictable and not a "sure thing".

[Demon]

Well then, I guess I'll just keep an eye open and report back f anything happens. But MCSP has never bombed on me in over a year on this laptop.

I honestly cannot see how graphic drivers have anything to do with data entry, especially in software that emulates Notepad (no special characters).

But then again, my specialty was mainframe programming. I only program PCs when I sleep at the Holiday Inn.

Robert

[picster]

Just crashed, not using italics. So much for that theory. HOWEVER, it still crashed while I was typing in a comment field, which has always been the case for me.

[Amoque]

I find it highly unlikely that it has anything at all to do with PBP (other than the obvious interruption to it). If so, I would think every user (having the exact same files loaded) would have the exact same issue - and, one would suppose, it would not be difficult to duplicate for the same reasons. I think that it must be some interaction with another process or service - perhaps a background application using the com port, USB access or internet connection- there are dozens of running programs active from the moment the power is turned on.

Has anyone disabled the files that start at power up? It is one of my pet peeves is manufacturers who seem to believe they own all of my computer's resources because I use one of their applications. I have found HP, Nvidia, and Adobe to be nearly criminal in their dis-care, but they are far from the only culprits. Starting with a minimal set-up and adding one line or application at a time may help to isolate what resource is causing the issue?

[picster]

crashed again this morning, while typing a comment (same as always).

Faulting application name: CodeStudioPlus.exe, version: 5.0.0.5, time stamp: 0x2a425e19
Faulting module name: unknown, version: 0.0.0.0, time stamp: 0x00000000
Exception code: 0x00000000
Fault offset: 0x00000000
Faulting process id: 0x1fb0
Faulting application start time: 0x01d2d7bae18ebaac
Faulting application path: C:\Program Files (x86)\Mecanique\MCSPX\CodeStudioPlus.exe
Faulting module path: unknown
Report Id: 1f0308d8-43b0-11e7-a3d0-ac72892a9ffd

and 23 seconds later,

Faulting application name: CodeStudioPlus.exe, version: 5.0.0.5, time stamp: 0x2a425e19
Faulting module name: ntdll.dll, version: 6.1.7601.23796, time stamp: 0x59028db3
Exception code: 0xc0000025
Fault offset: 0x000904b0
Faulting process id: 0x1fb0
Faulting application start time: 0x01d2d7bae18ebaac
Faulting application path: C:\Program Files (x86)\Mecanique\MCSPX\CodeStudioPlus.exe
Faulting module path: C:\Windows\SysWOW64\ntdll.dll
Report Id: 2d149eb5-43b0-11e7-a3d0-ac72892a9ffd

Any further suggestions?

This is the ONLY program that halts and displays an error, ALWAYS when typing in comments.

[mark_s]

Just a quick scan through the error list and this sticks out "Faulting module name: ntdll.dll." A quick Google of "ntdll.dll" and found this, along with a lot of other hits.

https://www.lifewire.com/how-to-fix-...errors-2624474

Maybe a solution?

[Scampy]

NTDLL.DLL is a windows system file containing library data for the core windows kernel. This would suggest that the issue is more likely a conflict between it and other system files rather than MSC itself crashing. Typical suspects are driver files, with prim suspects being graphics drivers. ATI and NVidia both have been know to have drivers that can cause this issue (as I stated in you other thread).

Often ensuring that your system is as up to day as possible is recommend, however I have found that on the odd occasion using an older driver, such as the one supplied with the graphics card has resolved a similar issue. Other possible causes can be a corrupt registry, but only a totally clean re-install on a freshly low level formatted drive will resolve that, but is often the last resort.

If you have access to another machine, try installing MCS etc and see if the issue is easily replicated. Unless the two machines use identical components I doubt that the problem will occur

[HenrikOlsson]

NTDLL.DLL is a windows system file containing library data for the core windows kernel. This would suggest that the issue is more likely a conflict between it and other system files rather than MSC itself crashing

If it's conflict between ntdll.dll and other system files then why is MCS the only program crashing with that particular error message? Do you believe MCS is the only program using a particular function exported by ntdll.dll?

MCS is the only program I've seen do this and it's done it to me hundreds of times on 3 different computers, it's done it to many other users as well yet the notion that it's got nothing to do with MCS keeps comming back.

The enduser (me) doesn't really care if the actual error is in the MCS source or in any of the supporting files MCS is using, it's MCS that crashes. If it does that due to a bug in MCS code itself or the way it calls some library function or an actual bug IN that library it ought to be in the developers interest to fix it but but he's already got my money and doesn't seem to care that much despite multiple threads and reports for YEARS. And MeLabs claims they can't do anything about it since they're not the developers, handy isn't it?

[richard]

And MeLabs claims they can't do anything about it since they're not the developers

they however could have spent some resources getting mplabx integration sorted , with proper language syntax plugin and source level
debugging .
I would have paid twice the 3.1 upgrade price just for that, there is little point in supporting newer chips when common "command line"
programmers can't program them . also the new chips with pin select modules can't take advantage of the "microchip code configurator "
without mplabx

still dreaming

[HenrikOlsson]

Indeed, so would I.

They tout PBP as a professional-level development tool and I can't really say anything bad about the compiler itself but the supporting / surounding stuff does leave a lot to desire. We're way past writing programs in Notepad, compiling from the commandline, loading the hex file into MPLAB and programming with the PICStart+

Proper and seamless integration into MPLABX is what's needed for the advanced users. After seeing Charles video a month or so ago I was able to compile a program within MPLAB but wasn't even able to set a breakpoint.

[richard]

you need to go back to mplabx 2.35 or earlier for breakpoints to be placed easily, but then I can't use my 16f1619

[picster]

NTDLL.DLL is a windows system file containing library data for the core windows kernel. This would suggest that the issue is more likely a conflict between it and other system files rather than MSC itself crashing. Typical suspects are driver files, with prim suspects being graphics drivers. ATI and NVidia both have been know to have drivers that can cause this issue (as I stated in you other thread).

Often ensuring that your system is as up to day as possible is recommend, however I have found that on the odd occasion using an older driver, such as the one supplied with the graphics card has resolved a similar issue. Other possible causes can be a corrupt registry, but only a totally clean re-install on a freshly low level formatted drive will resolve that, but is often the last resort.

If you have access to another machine, try installing MCS etc and see if the issue is easily replicated. Unless the two machines use identical components I doubt that the problem will occur

This isn't directed at you Scampy, but it's an acknowledgement of a lot of wasted time and effort, caused by ONE program that the developer isn't fixing.

The problem has occurred on 3 of 3 machines of mine, different graphics card, ALWAYS while running MCS and ALWAYS while typing comments. No other programs are crashing with this error. I repeat for clarity, NO other programs are crashing with this error, my system is rock solid stable otherwise. Seeing as MCS is always the offending (crashing) party, I would thing the manufacturer would be held somewhat accountable. I know there is a number of users who have reported this issue, and are met with frustration and denial "it's peculiar to your system" - "and yours" "and yours". More than one (I have THREE) means it's not a single system. More than one operating system means the same. It all points to MCS being the uncooperative player, since nothing else is crashing, all the other kids "play nicely together".

There is no guarantee that any of the "fixes" noted will work, when this isn't a windows problem, it's a MCS problem. There is some aspect of it that's doing this or a number of other users and I would all see this same error on OTHER programs on our multiple (non-identical) systems as well. Yet there is NO cooperation from the software developer because they "cannot replicate the problem". Too many of us have it, and there should be some accountability for paid software that not only crashes, but loses our work. That really sucks.

This isn't directed toward any individual - except perhaps the developer - but the whole ostrich thing of "one off" isn't helpful when you are one of a NUMBER of "one-offs" that this is recurring with, and you've tried it on no less than 3 different laptop system configurations with the same results.

[Scampy]

Picster, I can understand your frustration, and again, this is not aimed at anyone personally.

Although I don't have official figures, would have thought that there would be quite a few thousands of PCs with MCS installed and there would be a lot more reported cases if the issue was a bug in their code, and the developers would (hopefully) take notice. But a google search for "microcode studio crashing" or words to that effect doesn't throw up stacks of results (apart from a couple of threads on this forum) suggesting its not as wide spread is thought.

It would be interesting to see if there was any log file that cold be used to see if there are any similarities between those machines or software versions etc, but the variables are so great and it would probably take someone with expertise in programming and computing to analyse. The fact you have had three PCs with different hardware crash and all at the same point of typing comments (which is strange as it's just text, unlike when it registers a word as a command and turns bold) is strange. I've not gone back through the rest of the thread, but does this happen with a certain version of MSC, or PBP. I've been using MCS 4.0.0 with PBP 2.60c on different builds of PC for the passed 7 or so years and can't say I've ever experienced a crash. Maybe the issue is with later versions that were patched in a hurry when PBP3 came out ??

One thing I have noticed is how quiet this forum is these days, which to me suggests that programming in PBP is less popular than it was 5 - 10 years ago, before the likes of the arduino and other similar development systems appearing on the market. Maybe the market has changed and the sales of PBP have dropped off to the extent that it's no longer a viable option for the developers of MSC to fix issues, which in their eyes is not a global bug... Either way I agree that for someone with this issue it sucks.

Is there any alternative to MSC from a different software house as I doubt that you will get anywhere with this.

[HenrikOlsson]

The alternative would be MPLABX but you know, I just payed $50 for the PBP3.1 upgrade but in order to properly use MPLABX one, apparently, has to use a several year old version - in which the chips I payed $50 to get compiler support for aren't available at the assembler level (ie they can't be used). And even when using an older version it apparently can't debug anything but BYTE variables properly without running thru hoops.

As I wrote earlier, the compiler is solid and I love it but the surrounding/supporting stuff quite frankly sucks compared to what's offered elsewhere. It's 2017, we're past the 16F84 and a glorified Notepad. Give me a solid IDE with syntax highlighting, code folding and hardware level debug that works please.

There are several threads with several users reporting the exact same issue and it's been going on for years. I've seen on it several computers as well. I haven't been coding much on my newest machine with W10, it'll be interestning to see if does it there as well.

[picster]

Hi Scampy,

To answer your question regarding versions, it has happened with PBP 2.60x and 3.0 (it's not while compiling so I don't think this is relevant), and with all versions of MCS that I've used since 4 (I don't remember if it was prior to 4).

You're right, it's not as if the parser is processing reserved words and bolding them or coloring them in differently on-the-fly, the entire line is being colored (or at least the portion after the ')

The only thing worse than a bug is an intermittent bug.

[mark_s]

Old story

MCSX started crashing yesterday again. Did not matter which code file I tried to edited, any change in text caused the error message to pop up. I found that you can ignore the message and still compile the code and make changes. Out of frustration, I reinstalled pbp3 and mcsx. That did not help. On this laptop there was an old copy pbp 2.47. So I opened the file in this version without crashing. The only problem was 2.47 did not support my pic. I dug out my 2.60 cd which supports my pic and then patched it with the last update before pbp3. The 2.60 cd had MCS 3.3, now I'm back in business and on this pc done with pbp3/MSCX.

If anybody wants me to run any experiments let me know. Because it can be repeated. It would be nice if MEL could un-marry pbp3 from MCSX and allow the use of older versions of MCS. I am sure there is reason it can't be done.

System is a Gateway laptop, XP, not connected to the net for over 2 years. not sure what changed, other than the date.

[pedja089]

I have problem in one of my programs that didn't handled regional settings well... Try to set all to EN-US.

[Scampy]

Old story

MCSX started crashing yesterday again. Did not matter which code file I tried to edited, any change in text caused the error message to pop up. I found that you can ignore the message and still compile the code and make changes. Out of frustration, I reinstalled pbp3 and mcsx. That did not help. On this laptop there was an old copy pbp 2.47. So I opened the file in this version without crashing. The only problem was 2.47 did not support my pic. I dug out my 2.60 cd which supports my pic and then patched it with the last update before pbp3. The 2.60 cd had MCS 3.3, now I'm back in business and on this pc done with pbp3/MSCX.

If anybody wants me to run any experiments let me know. Because it can be repeated. It would be nice if MEL could un-marry pbp3 from MCSX and allow the use of older versions of MCS. I am sure there is reason it can't be done.

System is a Gateway laptop, XP, not connected to the net for over 2 years. not sure what changed, other than the date.

Mark, that's interesting,

I'm running version 4.0.0.0 of MCS, version 2.60c of PBP and an old version of mpsam suit (5.07 I think) and I've been very lucky as I've never had a crash whilst coding or compiling etc. Yes I know the restrictions this combination has, but for me and the projects I undertake, the supported chipset is fine.

I don't have the needs to upgrade to PBP3 so my input may be irrelevant, but I hope that someone can sort out this issue for those who have had to upgrade in order to remain competitive

Oh and the PC is an AMD 6 core processor running at 3ghz, 12gb DDR3 ram, and windows 7 professional 64bit .... Are those experiencing this issue running windows 7 or later versions... maybe (clutching at straws) it only affects those running windows 8 or 10 ??

[HenrikOlsson]

As I've said, I've had it crash at home and at work. One machine running W7 Home 64bit Swedish with the bought MCSPX, the other running W7 Pro 64bit English with the free MCSX. Both PBP3.0.x.x

At home I've upgraded my personal license to PBP3.1 on a brand new machine running W10 Pro but haven't been doing much with it lately so it's still low mileage (very much so).

I'm not a PC software developer but I'm pretty sure it's got nothing to do with the compiler, it's not when compiling/assembling that it crashes, it's when typing in your program. I've got screencapture (video) of it doing it if that would help....

[mark_s]

As Henrik points out it's not pbp3, the problem is MCSX. MCS is stable but pbp3 will only run on MCSX. A lot of software developers like to blame their bugs on WIN7,8,10. I am using XP and this pc is only used for programming pics. My home system is still stable so if I need something not supported by 2.60 it can get done.