通过代码随机进行单元测试调试步骤

Question

We've got a project which has recently been upgraded from being written under VS2010 (using .net Framework 3.5) to VS2012 (using .net Framework 4). We work with TFS, so older builds are available to work on.

When debugging the VS2010 versions of the build breakpointing the <TestMethod()> methods will work, stepping into and breakpointing on the called code works, and stepping through the code from start to finish works.

In the 2012 version, breakpoints only really work in the <TestMethod()> code. Once the code leaves the <TestMethod()> area, things get odd . Stepping into and through the code seems to simply step to random lines in the code. Values resolve strangely and the debugger seems to hop forward through the code largely at random. Once the debugger returns to the <TestMethod()> code area, things go back to normal, and the correct results are returned from the called functions even though the values shows during debugging were all strange.

I can't really think how to describe what's going on any better than that. I'm not in charge of building the software, and the people who are aren't aware of anything that they have changed that might have caused this issue.

Does any of this sound familiar enough to anyone that they can suggest what setting might have been changed to produce this behaviour, or what I need to set under .net 4 in order to get unit testing working again?

Answer 1

当调试器使用的源代码与用于编译要调试的程序集的源代码不同时，通常会发生这种情况。