There are likely several places that will accept you No Matter What for a masters, since you're paying them tuition and all. You can even do some masters programs online (I actually only know that's the case for U Washingtons Applied Math, but I bet there are physics programs too). They'll be basically worthless, but I'm sure someone will take you.
If you're intent is to get into a PhD program on the back of that masters, I have to tell you, "tough luck."* There's no way a year of graduate coursework (even with a 4.0) makes up for an awful undergraduate record. Your best bet would be (a) get another bachelors (in physics this time) in 2 years with a 3.8+ or (b) work with a (notable) professor from a PhD granting institution for free for a couple years, and apply on the strength of an OUTSTANDING recommendation from that professor. For this to work though you'd need to (a) publish, and (b) be extremely useful.
Honestly, though, as Riley asked, why switch? Physics is much more competitive than computer engineering (just based on jobs/number of applicants). If you're lazy, you'll fail at some point along your path.
*Getting a 990 or close on the PGRE might be enough to get into some school. But not the top 10, or even top 50. There are GPA minimums, even at the masters level. If you really think you can get in based exclusively on your brilliance, your best bet is to do some major work on some open question in physics, and get it published. I don't really think it's that easy to be a closet Feynman or Einstein in the days of computational theory though, so I highly doubt this is possible (my expectation value keeps coming out negative, actually, but I'm pretty sure it's machine error).