Description Ken Dyck 2012-08-27 16:20:21 EDT

If a memory monitor is created while the target is still running, it displays '??' for the block of addresses that are visible in the rendering. This is as expected, however, for the rest of the debug session, the memory in that initially visible region remains '??' even after the target is suspended. Memory that was not initially in the visible region updates as you'd expect.

To Reproduce:
1. Run the debug target.
2. Create a new memory monitor
3. Pause core.
4. Notice how the memory immediately before and after the current view is updated correctly, but the region initial region remains '??'.

Note: This bug was observed in a "GDB Hardware Debugging" environment (i.e. remote debugging of an embedded processor via GDB) and I have not attempted to reproduce it while debugging locally. I have, however, reproduced it on two variants of GDB (one for an ARM architecture, the other for a DSP architecture). So I guess that problem is in somewhere in CDT/Eclipse.