~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- ccache now has a ``direct mode'' where it computes a hash of the source
- code (including all included files) and compiler arguments without
+ code (including all included files) and compiler options without
running the preprocessor. By not running the preprocessor, CPU usage is
reduced; the speed is somewhere between 1 and 5 times that of ccache
running in traditional mode, depending on the circumstances. The speedup
- Fixed NFS issues.
- Computation of the hash sum has been improved to decrease the risk of
- hash collisions. For instance, the compiler arguments `-X -Y` and `-X-Y`
+ hash collisions. For instance, the compiler options `-X -Y` and `-X-Y`
previously contributed equally to the hash sum.
- Bail out on too hard compiler options `--coverage`, `-fprofile-arcs`,