Cortex M3 Hardfault exception

[shubha]

Posted on April 17, 2011 at 17:05

Cortex M3 Hardfault exception

[Tesla DeLorean]

Posted on May 17, 2011 at 14:31

The value in LR at the Hard Fault handler is a magic value which imparts information to the processor about how to return the the prior system/user state. It will not be the faulting instruction.

Implement a REAL Hard Fault Handler to decode the system state and faulting PC. Joseph Yiu has posted several real good examples of this, and his books on the Cortex-M3 is recommended.

[Gawie]

Posted on May 17, 2011 at 14:31

Hi

clive1's suggestion is the way to go. Have a look at the post below - it will give you more info on what he was talking about in his previous post.

[DEAD LINK /public/STe2ecommunities/mcu/Lists/ARM%20CortexM3%20STM32/Flat.aspx?RootFolder=/public/STe2ecommunities/mcu/Lists/ARM CortexM3 STM32/HardFault Exception Why oh why!&currentviews=3001]https://my.st.com/public/STe2ecommunities/mcu/Lists/ARM%20CortexM3%20STM32/Flat.aspx?RootFolder=%2Fpublic%2FSTe2ecommunities%2Fmcu%2FLists%2FARM%20CortexM3%20STM32%2FHardFault%20Exception%20Why%20oh%20why%21&currentviews=3001

Cheers,

Gawie

[shubha]

Posted on May 17, 2011 at 14:31

thanx.

I had attached a .bmp file indicating the stack contents at the time of hardfault. Can we interpret it offline(not thro program)?

[Tesla DeLorean]

Posted on May 17, 2011 at 14:31

Can we interpret it offline(not thro program)?

If you refer to the materials cited, you could interpret it yourself. To do it properly you need to interact with the debugger to review the faulting code, and the register context with respect to that code and what it was doing. Go look at the Thumb2 code you have at around 0x080066CF (the instruction immediately before) You can't review the core's register from a BMP

[aqueisser]

Posted on May 17, 2011 at 14:31

This is also a good overview of how to decode the hardfault stack info: 

http://www.keil.com/appnotes/files/apnt209.pdf

http://www.keil.com/appnotes/files/apnt209.pdf

And I think there was a typo, the bitmap suggests the fault should be just before address 0x08006CCF.

Andrew

[Tesla DeLorean]

Posted on May 17, 2011 at 14:31

Yes 0x08006CCF, my transcription error