Message for all curious about end of write interrupt

[radim]

Posted on April 23, 2007 at 13:24

Message for all curious about end of write interrupt

[radim]

Posted on March 30, 2007 at 07:42

Unfortunately the end of write interrupt doesn't work you need to test the relevant bit as managed in standard library.

Nice and efficient way would be also using timer so you won't waste your time by wait loop and you would get dedicated time frame for testing bit according to the ammount of data to be written.

I'm sorry about that information I beliave we'll solvve the issu in the next revision.

Regards

Radim

[kleshov]

Posted on April 02, 2007 at 06:37

Thank you for the information. It is good to see communication going between us, developers, and the MCU manufacturer.

Now would probably be a good time to issue an official errata document.

Best regards,

- mike

[kaouther]

Posted on April 02, 2007 at 11:46

Hello,

we are investigating on the issue and thanks for all your details.

We reproduced the problem of end of write operation but ONLY when using running in Debug Mode:

- If a Global run is applied, there is no problem: the

End Of Write Interrupt is serviced normally.

- If Breakpoints are inserted in the project and ‘Step

Over’ or ‘Step Into’ are used, the problem could occurs.

In stand-alone, there is no problem: the End Of Write Interrupt is serviced normally.

We are also using the toolchain IAR 4.40 and other tests are set to find precisely the root cause.

Thanks for your understanding. We will keep you informed with our last results and the root cause of this behavior.

[kleshov]

Posted on April 02, 2007 at 12:35

Hello,

I would suggest the following test:

1) Clear the flash channel interrupt enable flag: EIC_IER0.IER1 = 0

2) Perform a flash write to trigger the flash interrupt

3) Wait until the flash write operation is completed

4) Execute a delay, say 10 milliseconds to be sure

5) Set the flash channel interrupt enable flag: EIC_IER0.IER1 = 1

6) Observe

Expected result: after step 5 an interrupt should trigger

Actual result: no interrupt is triggered

Further tests:

Reduce the delay in step 4 above until the interrupt starts to trigger.

Needless to say, the tests should be run uninterrupted (by the debugger or anything else).

Best regards,

- mike

[najoua]

Posted on April 04, 2007 at 11:55

Hello Volius,

Could you please give me a compiled project reproducing the problem?

Regards,

Najoua.

[kleshov]

Posted on April 04, 2007 at 12:16

This might take me some time. I will try to put together a project reproducing the problem this week and post it to this thread.

Regards,

- mike

[najoua]

Posted on April 04, 2007 at 12:19

Ok,thank you Volius. I am waiting for your feedback.

Regards,

Najoua.

[kleshov]

Posted on April 05, 2007 at 15:10

Hello Najoua,

Please find attached a compiled project demonstrating the problem.

Regards,

- mike

________________

Attachments :

flash_test.zip : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HtHs&d=%2Fa%2F0X0000000aL6%2FbFgXhrftr26cROHQdUkPwW57lwEhiQvEXoW30A2q0B8&asPdf=false

[najoua]

Posted on April 06, 2007 at 05:48

Hello Volius,

Could you please provide me the iostr710.h header file because it doesn't exist in the IAR version I have.

Thank you in advance,

Regards,

Najoua.

[kleshov]

Posted on April 06, 2007 at 09:24

Sorry, I completely forgot about that file. I got the file from other sources, can't remember now where from exactly.

Here it is.

Regards,

- mike

________________

Attachments :

iostr710.h.zip : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HtKH&d=%2Fa%2F0X0000000aLc%2FZiySTSD800VWs7gIKhHgHJM5PRgcXmLZAo.jhkO.zsE&asPdf=false