Hello!
I disagree with the assertion that programs which rely on PAGE_SIZE are
broken -- if they are architecture specific, they are perfectly correct to
assume that PAGE_SIZE is a compile-time constant and there is just no valid
reason for trying to hide PAGE_SIZE from them.
PAGE_SIZE has been a part of glibc API since ages (on architectures where
it's a constant) and I don't think that Debian should be the one who takes
decisions on what's in the API and what isn't, leaving us with packages which
build (on the given architecture) with every libc except for the Debian one.
Also, you have not hidden PAGE_SIZE, but replaced it by a non-constant
expression. This breaks programs which assume that if PAGE_SIZE exists,
it's a constant. So far, this was true on all architectures and although
it hasn't been codified anywhere, so technically you are not breaking
any standard, it was a common behavior and there are programs that
rely on that, so I don't think it should be changed without a really
strong reason.
Please revert this patch and leave the decision to upstream maintainers.
Martin
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
