'Restart' has encountered a problem

[lpcware]

Content originally posted in LPCWare by PKlammer on Fri Jul 29 10:53:16 MST 2011
[FONT=Helv][SIZE=2][FONT=Helv][SIZE=2]I keep losing debug capability; after a pause or breakpoint, Reset (Restart) becomes unavailable, I can only Red-Box close debugger and click the Bug which REBUILDS ALL (*sigh*); my try-breakpoint cycle is several MINUTES!
Target is LPC1114/301.
Tried:
Exit LPCXpresso IDE, confirm no process images running in TaskManager;
unplug target from power, unplug LPC-Link programmer from USB, move to on-case (no external hub) USB port.
Relaunch LPCXpresso IDE, set project to Release, rebuild, click one "step over" (silences beeper), then clicking Restart gets :
[IMG]http://knowledgebase.nxp.com/Image1.gif[/IMG]
[pop-up error box:
"""
Problem Occurred
[Red X ball]
'Restart' has encountered a problem.
Exception(s) occurred attempting to restart.
[OK] [[U]D[/U]etails >>]
"""
Clicking Details reveals:
"""
Exception(s) occurred attempting to restart.
  Target request failed: Cannot access memory at address 0x1feff8.
  Cannot access memory at address 0x1feff8
"""
This is after trying changeout of target boards (new board was OK for a while, then problem resumed); and changeout of LPC-Link with another (no change).
LPCXpresso 3.6.1 here.
Search on LPCXpresso KnowledgeBase community Forum for "'Restart' has encountered a problem" yields at least 500 hits (seems to be search limit parameter) so, no, I haven't scoured the KnowledgeBase yet.

[/SIZE][/FONT][/SIZE][/FONT]

[lpcware]

Content originally posted in LPCWare by Rob65 on Mon Aug 01 22:54:41 MST 2011
P,

be very carefull with chill spray.
If you spray chill spray directly on a hot IC there is a fair chance that you ruin the chip.
The shrinkage due to temparature changes may results in cracks inside the chip or damage the solder connections of a BGA chip.
The lpc3154 is soldered on the board with a lot of very tiny balls of soldering wire, sudden temperature changes can damage these easily.

Maybe (hopefully) you know - but just in case someone else tries he/she is warned ...

Regards,

Rob

[lpcware]

Content originally posted in LPCWare by PKlammer on Mon Aug 01 18:50:38 MST 2011

---

Quote: CodeRedSupport
It certainly sound like a problem with the board. Suggest you review especially the power circuit.

---

Well, I just tried chill spray, after 'Restart' failure, and it doesn't come back.  Checked chip beforehand and it was not hot.  Is there a chip self-test diagnostic that could be loaded to look around inside the die to check for non-working bits?

[lpcware]

Content originally posted in LPCWare by CodeRedSupport on Mon Aug 01 11:13:21 MST 2011
It certainly sound like a problem with the board. Suggest you review especially the power circuit.

[lpcware]

Content originally posted in LPCWare by PKlammer on Mon Aug 01 06:52:07 MST 2011

---

Quote: CodeRedSupport
What is your hardware setup? Is it a custom board? How is it powered? Are you using LPC-Link or Red Probe+ as the debug? Is it getting hot and that is when it is failing? When it fails, does the board continue to run, but you cannot debug it?

What is your software setup? Windows or Linux? Which version?

What is your firmware setup? What features of the target are you using? Clock setup?  Sleep modes? Are your muxing any of the debug pins?

---

It is our own layout, with regulator on board; using LPC-Link, from Windows; other setup details not in my hands.  The chip gets hot, but not too hot to leave finger on.  However, recovery after (unpowered) board rest is consistent with localized overheating.  Opening enclosure and applying wetted finger to chip restored 'Restart' without unpowered rest. It must be VERY localized, for reprogramming and normal execution have been unaffected when 'Restart' capability is lost. On the other hand, the 'Details' view in the error pop-up refers to memory access failure in a specific range, so it is perhaps understandable that overheating in one region of the die could affect only a limited range of memory.

[lpcware]

Content originally posted in LPCWare by Marco72 on Mon Aug 01 00:49:08 MST 2011
I have this problem with v4.0.5 [Build 123]  when I exceed the break-point quantity limit.
IMHO it is a bug of IDE.

Marco

[lpcware]

Content originally posted in LPCWare by CodeRedSupport on Sun Jul 31 02:17:04 MST 2011
What is your hardware setup? Is it a custom board? How is it powered? Are you using LPC-Link or Red Probe+ as the debug? Is it getting hot and that is when it is failing? When it fails, does the board continue to run, but you cannot debug it?

What is your software setup? Windows or Linux? Which version?

What is your firmware setup? What features of the target are you using? Clock setup?  Sleep modes? Are your muxing any of the debug pins?

[lpcware]

Content originally posted in LPCWare by PKlammer on Sat Jul 30 14:44:57 MST 2011
It can't be anything my code is doing, because even at a fresh reload, before even touching the 'Go' button, I get this error if the very first button I click is the 'Restart' icon.  And it goes away if I give the board a rest for a while (hours) and work on a different board (until it loses 'Restart' capability, too!). After laying aside for a few hours, that board will debug normally for a while, then lose it again.

[lpcware]

Content originally posted in LPCWare by Rob65 on Fri Jul 29 22:46:22 MST 2011

---

Quote: PKlammer

The first three don't even link to offending source code.
'Guess I need a v3 -> v4 update tutorial.

---

Strange ...
The update tutorial should be simple: export all your projects using the "export projects and references to archive" from the "Import and Export" Quick menu into one .zip file and import this using the "import archived projects" option from that same menu in the 4.0.5 Xpresso tools.

If you want to debug without reloading, check out the Code Red FAQ. Under "Debug"  you will find "Debug a running system".
Most of the times a sudden death of your debugger means you are doing things (in software) that stop debugging.
Think of:

[LIST=1]
[*]Switching the CPU clock to a very low frequency
[*]Entering Deep Sleep or Power Down modes
[*]Reprogramming the JTAG/SWD pins with another function
[/LIST]
Sometimes it helps to hit the reset button on your LPCXpresso base board (if you don't have a base board attach a seperate switch to do this).

If it is just loosing the restart option of the debugger, that happened to me once.
I made some stange settings in my project properties and that disabled the restart button.
Unfortunately the only option I got working was to import all .c/.h files into a new workspace and recompile everything again :(

Rob

[lpcware]

Content originally posted in LPCWare by Ex-Zero on Fri Jul 29 10:59:56 MST 2011
1. Update your LPCXpresso to V4.0.5 :)

2. Try vector catch http://support.code-red-tech.com/CodeRedWiki/DebugAccessChip