I've got some free time again so I'm returning to this problem. 2 important facts: - A test C program, using one of the "undefined" functions (isnan) compiles with GCC and runs - I went in by hand and temporarily replaced each version of math.h on my computer with a version that has the necessary functions defined. Cabal-install still failed with the same error. This makes it seem like cabal doesn't have access to _any_ math.h file, but if this were the case, wouldn't I get many more undefined-function errors than I do? Thanks for the help, Tom On 1/20/12, Brandon Allbery <allbery.b@gmail.com> wrote:
On Fri, Jan 20, 2012 at 15:26, Tom Murphy <amindfv@gmail.com> wrote:
On 1/19/12, Brandon Allbery <allbery.b@gmail.com> wrote:
On Thu, Jan 19, 2012 at 16:37, Antoine Latter <aslatter@gmail.com> wrote:
On Thu, Jan 19, 2012 at 3:24 PM, Tom Murphy <amindfv@gmail.com> wrote:
Is it possible that I have a version of math.h which doesn't have these definitions ("gamma," "finite," "isinf," etc.) in it?
That's a good guess based on the error messages - or math.h isn't in whatever paths we're looking in.
If math.h wasn't in the paths at all, wouldn't I get an error about _every_ function being undefined?
The problem isn't math.h; these are link errors, so it's /usr/lib/libm.dylib that is somehow wrong. I can't help with this too much as I'm on Lion so it's a symlink to libSystem.dylib (and if that's messed up, you don't even *boot* much less compile stuff).
-- brandon s allbery allbery.b@gmail.com wandering unix systems administrator (available) (412) 475-9364 vm/sms