diff options
| author |   Linus Torvalds <torvalds@home.transmeta.com> | 2003-04-02 12:22:54 -0700 |
|---|---|---|
| committer |   Linus Torvalds <torvalds@ppc970.osdl.org> | 2005-04-07 21:00:03 -0700 |
| commit | 67c88166eede42619bddd51c055b8e2c2c099334 (patch) | |
| tree | 15a58dcf37fe76bd99d93bea08a7039051780154 /FAQ | |
| parent | Don't drop SYM_NODE information when evaluating expression types. (diff) | |
| download | sparse-67c88166eede42619bddd51c055b8e2c2c099334.tar.gz sparse-67c88166eede42619bddd51c055b8e2c2c099334.tar.bz2 sparse-67c88166eede42619bddd51c055b8e2c2c099334.zip | |
Add a note in the FAQ about the lack of support for old K&R-style
untyped constructs.
Diffstat (limited to 'FAQ')
| -rw-r--r-- | FAQ | 18 |
1 files changed, 18 insertions, 0 deletions
@@ -66,3 +66,21 @@ A. I don't know yet. I originally thought it would be LGPL, but I'm people can write back-ends for it without having to make those back-ends available under the OSL. + +Q. Does it really parse C? + +A. Yeah, well... It parses a fairly complete subset of "extended C" as + defined by gcc. HOWEVER, since I don't believe in K&R syntax for + function declarations or in giving automatic integer types, it + doesn't do that. If you don't give types to your variables, they + won't have any types, and you can't use them. + + Similarly, it will be very unhappy about undeclared functions, + rather than just assuming they have type "int". + + Note that a large rationale for me doing this project is for type + following, which to some degree explains why the thing is type-anal + and refuses to touch the old-style pre-ANSI non-typed (or weakly + typed) constructs. Maybe somebody else who is working on projects + where pre-ANSI C makes sense might be more inclined to care about + ancient C. It's open source, after all. Go wild. |