This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Repost: RFA [4.1]: improvement to if-conversion and cross-jumping (PR20070)


On Dec 14, 2005, at 7:45 AM, Bernd Schmidt wrote:
Bernd asked for *_p names for these functions in:
http://gcc.gnu.org/ml/gcc-patches/2005-11/msg02031.html .
Can you please discuss this with him and then tell me what consensus you reached (if any).

I'm open for better suggestions. I see your point that a predicate shouldn't modify its inputs, but for the most part, this function _does_ verify that the two rtxes are equivalent.

bla_p names are reserved for a small class of functions that never hurt anything. If you can tell it did something, then, it can't be called bla_p. An example of a reasonable use of _p would be if the object contains some bits that are modified that serve as a cache or a speed up to make things go faster. And example of something that isn't ok would be if you merge bits from two delcs into one of them and return whether or not they were the same.


A cursory look at the code appears to be a modification of at least one of the operands. unify/merge/smash come to mind for names, not _p.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]