mmm....not really sure..... but perhaps the Planck scale is the ultimate (so far)... you may have problems if your photon wavelength is shorter than the Planck length or at least the rules of the game may change (?)
You may consider that a frequency higher than an inverse planck time is not meaningful in current theories, so that we cannot give a clear meaning to a photon energy higher than 1/sqrt (hbar * G / c^5) * h.
In current QFT, which is at the origin of the concept of what is a "photon", there is no official upper limit on what is the maximal energy of a photon.
However, there are two caveats:
- the first is that gravity (which is NOT included in QFT as we know it), suggests that there will be trouble around the Planck energy. In other words, we think that whatever QFT might say about arbitrarily high energies, it will not be correct because gravity will not only be not neglegible (as it is assumed in QFT), but it will rather be so dominant that it is meaningless to try to work with a theory that doesn't contain gravity
- QFT itself, as we know it, has a serious problem at high energies. We think that current QFT are "low energy effective" approximations to something else, and this approach is what justifies renormalisation schemes. This has nothing to do with Planck scales, but rather with the intrinsic structure of QFT.