|
Boost Testing : |
From: AlisdairM (alisdair.meredith_at_[hidden])
Date: 2006-11-16 15:29:45
AlisdairM wrote:
> Anthony Williams wrote:
>
> > One thing possibly worth noting: I'm running these tests with
> > bjam_options="release".
>
> That could well turn up a new set of errors on the Borland compiler.
And just to confirm the strangest Borland error I have seen for a while:
the int.cpp test in MPL is failing only when function inlining is
enabled. This is a compile time failure, trying to parse a
BOOST_STATIC_CONSTANT declaration. I am still trying to nail a minimal
test case for this, maybe some unforeseen mixing of the preprocessor
and inliner?! But in the meantime, I suggest we need to turn off
inlining until we get 1.34 out the door. I am hoping that with that
change, you will have the same set of test results as metacomm
(although it seems you also get unexpected passes as well as fails in
this mode - certainly something to investigate further)
-- AlisdairM