We have seen this approach as it is supported by arms

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: h the debugger can use to allow the user to debug the program from a source-level view, addressing variables by their source names rather than by their memory address. This 'source-level debugging' is more powerful and requires less detailed knowledge of the machine environment than object-level debugging. A common weakness in software debuggers is the lack of a 'watchpoint' facility. A watchpoint is a memory address which halts execution if it is accessed as a data transfer address. Since many processors have no support for trapping on a particular The ARM debug architecture 233 address (apart, perhaps, from a memory management page fault, which is rather coarse for this purpose) this functionality is often omitted. This is a pity, since a very common source of error in a C program is corrupted data caused by an errant pointer in some unrelated part of the program, and this can be very hard to track down without a watchpoint facility. Embedd ed debugging Debugging becomes significantly more difficult when the target system is embedded. Now there is probably no user interface in the system, so the debugger must run on a remote host through some sort of communication link to the target. If the code is in ROM, instructions cannot simply be replaced by calls to the debugger since the locations are not writeable. The standard solution here is the In-Circuit Emulator (ICE). The processor in the target system is removed and replaced by a connection to an emulator. The emulator may be based around the same processor chip, or a variant with more pins (and more visibility of its internal state), but it will also incorporate buffers to copy the bus activity off to a 'trace buffer' (which stores the signals on all the pins in each clock cycle for some number of cycles) and various hardware resources which can watch for particular events, such as execution passing through a breakpoint. The trace buffer and hardware resources are managed by software running on a host desktop system. When a 'trigger' event occurs, the trace buffer is frozen so the user can observe activity around the point of interest. The host software will present the trace buffer data, giv...
View Full Document

This document was uploaded on 10/30/2011 for the course CSE 378 380 at SUNY Buffalo.

Ask a homework question - tutors are online