CCM and GDB oddity

[jpeacock2399]

Posted on August 26, 2013 at 22:06

While dumping some memory from the CCM on an STM32F405 at address 0x1000 FF00 I noticed the GDB memory browser window cut off at 0x1000 FFEF as the end of valid memory.  The last 16 bytes of CCM didn't show.  Has anyone else seen this?

I use a Segger J-Link, GDB and Eclipse.  The J-Link is a few weeks behind on updates, GDB 7.2.50.20100908.   Jack Peacock #ccm-gdb

[Tesla DeLorean]

Posted on August 26, 2013 at 22:48

I've certainly observed boundary issues, in other debuggers, where the block it's trying to read spans the end so the whole read fails, they are typically not reading a byte at a time. In these case one parks the view window so it doesn't exceed the allowed scope.