Hi, GL,
I apologize for the frustration you're experiencing around this issue. As the folks above allude to, the above information isn't enough for us to pinpoint precisely what's going on. We really do want to help!
Based on above comments, I'm guessing that there are related forum threads on this that I haven't read, so I'm not sure of the history of your problem. So, just to get everything on the table, I want to ask a couple of questions that you may have already answered (and my apologies if this is the case):
- When the compiler crashes, is it prompting you to send a report to us (and is it being sent) Assuming that's the case, the bug will exist in our database and it will be investigated.
- Have you filed a report at http://lab.msdn.microsoft.com/productfeedback/ This makes it easier for us to connect the crash report to the people who are experiencing it, and also allows us you to enter repro steps (which is super important -- the compiler gets a lot of input from various places, all funneled through a couple of channels, and this allows us to better pinpoint the culprit code. This also allows you to track the progress of the investigation.
The fixes we released earlier addressed specific problems. They were not panaceas for all compiler crashes, and so it is certainly possible that other unrelated crashes are lurking there (though we do our best to make this not be the case). If your specific case is unique (and so far I haven't heard other similar issues since the previous fixes we released), then it's going to take some digging to get to the bottom of it. All information you can provide (beyind the error message, which you've already posted) will be very helpful -- the scenario, the code that was running when the crash occured, etc. The error message unfortunately can only tell us so much -- it tells us what happened "now" (i.e. crash in a certain area of code) but nothing about the events that led up to that state. With only the "now" information, devs can sometimes make an empirical fix (e.g., don't try to dereference a NULL pointer) but not a root-cause fix (e.g., "So what is it that's causing the code to send a NULL-reference pointer anyway That's not right! I wonder what else is affected ").
--Matt--*
|