CCSV5.1.0.07001 - Concurrent sessions of CCS interfering with each other?

Hi Mike,

MikeH said:

Is this (two concurrent instances on a single PC) a known issue?

Not as far as I know (as long as you have enough memory). I have not encountered any issues myself.

You are running two instances of the same CCSv5.1 install (using two separate workspaces), correct? Is the hang pretty random or does it occur after you do some specific task (loading programs, setting a breakpoint, etc)?

ki

Ki,

Ki-Soo Lee said:

You are running two instances of the same CCSv5.1 install (using two separate workspaces), correct?

Correct. However, I auto-load the same .out file on both instances so they run identical code.

It happens rather sporadically. I have yet to be able to pin down a cause.

Mike

I've been trying to reproduce this for a few days now. So far no luck. We need to be able to reproduce this on our end to be able to investigate it. If you can come up with more ideas on what may be the cause, that would be great.

Ki,

The "hang" does not appear to be related to two simultaneous debugging sessions. It happens with a single debugging session running on my XDS560v1STM. It appears to happen after stopping execution and then "Restart"-ing the execution. The hang is usually for about 15 seconds, after which everything returns to normal.

I am running into a number of other anomolies (however, v5.1 is *much* better than v4.x), including:

1. CCS shuts down (program closes) for no apparent reason. Will restart normally.

2. A few times I have gotten errors with "garbage collection", and one rather humorous "Internal error while reporting internal erorr"....:). After this I could not re-open the workspace because it was "in use or not available". After a while (1-2 minutes) it became available again. I suspected an Eclipse process was still running, but could not find it with Task Manger.

3. Once or twice the program will hang and not come out of it. This required a hard shutdown using Task Manager.

Hope this helps. Keep up the good work!

MikeH

Ki,

MikeH said:

The "hang" does not appear to be related to two simultaneous debugging sessions.

...sorry, not true. Starting two debugging sessions definitely makes this situation worse.

Again, I am using two different emulators. XDS560v2STM and XDS100 USB.

I'll try to narrow down exactly when this happens and report back later.

MikeH

MikeH said:

It happens with a single debugging session running on my XDS560v1STM. It appears to happen after stopping execution and then "Restart"-ing the execution. The hang is usually for about 15 seconds, after which everything returns to normal.

Does the hang occur AFTER you press the "resume" button to resume execution? If so, does the target execution actually resume first (it says "Running" in the Debug view) and then hang or does it hang first and then resume execution?

And when it hangs, does the GUI have repainting problems?

Thanks

ki

Ki-Soo Lee said:

Does the hang occur AFTER you press the "resume" button to resume execution?

Yes.

Ki-Soo Lee said:

If so, does the target execution actually resume first (it says "Running" in the Debug view) and then hang or does it hang first and then resume execution?

It hangs first, then resumes execution.

Ki-Soo Lee said:

And when it hangs, does the GUI have repainting problems?

Not quite sure what you mean. The GUI is unresponsive during the hang. This is what I mean by "hang", the program is unresponsive to any input.

MikeH

This is happening more and more frequently. I am in a debug session on both instances of CCS. When I attempt to suspend the session I get a very long pause before CCS actaully suspends running. Once the session is suspended, when I hit "Restart", CCS again becomes unresponsive. This same behavior happens on both instances of CCS at the same time, as though they are interfering with each other.

FYI....I suspect this has something to do with network congestion on my LAN-connected emulator (XDS560v2 STM). My internet speeds (speedtest.net)  tend to be slower and sporadic while I am running a debug session. When I terminate the debug session my internet speeds appear to be higher.

MikeH

....and this morning..

ICEPICK_C: Error connecting to the target: (Error -181 @ 0x0) The controller has detected a dead JTAG clock. The user must turn-on or connect the JTAG clock for the target. (Emulation package 5.0.470.0) 

ARM9_0: Failed to reset the emulator: Unified-SCIF is operating in multiple client mode. The operation being attempted is invalid in this mode. It is likely a problem with `SC_CMD' in SMG_call(). 

Mike,

Thanks for your sharing your findings on the cause of the issue! That is interesting that having multiple LAN connections is the issue. I disable the wireless when I am connected but I can try switching it back on to see if I can reproduce the issue.

Thanks

ki

Ki,

Ki-Soo Lee said:

Thanks for your sharing your findings on the cause of the issue!

No problem. Happy to help.

Ki-Soo Lee said:

That is interesting that having multiple LAN connections is the issue.

It may be premature to declare victory, but I have worked all day without the "pause" issue. I guess with the wireless NIC, CCS has multiple potential routes to the emulator. The wireless route would (probably) have additional delay. Perhaps the emulator driver is sensitive to delays.

Thx,

MikeH