On Tue, Feb 17, 2015 at 11:44:42AM -0800, John Johansen wrote: > On 02/17/2015 11:36 AM, Steve Beattie wrote: > > On Thu, Feb 12, 2015 at 12:40:41PM -0800, Seth Arnold wrote: > >> (I don't have an easy way to test the build with gcc-5, so this may not be > >> exhaustive.) > > > > Even with the fix applied, the parser still fails to build with gcc-5, > > with the errors: > > > > http://paste.ubuntu.com/10278117/ > > > > The following patch lets the parser build both under gcc-5 and gcc-4.9. > > I'm not sure why gcc-5 thinks the types returned by dump_XXXX are wrong. > > > > Signed-off-by: Steve Beattie <st...@nxnw.org> for both trunk and 2.9. > > The patch is fine but this sure feels like thrashing about in the dark, > it doesn't actually change the return value just stops using it.
I agree about thrashing around in the dark and I don't really like dropping the return value on the floor, either. > What of all the other places we have a similar pattern? Why are > they passing Well, we don't actually have other places that invoke a method that returns an ostream reference (unless I overlooked something). In a simple grep, it looked like every time we called a method in a << pipeline, what was returned was a data item to be output directly. I may have missed a variable of type ostream& (a la the 'os' variable that the commit Seth made addressed), but nothing jumped out at me. -- Steve Beattie <sbeat...@ubuntu.com> http://NxNW.org/~steve/
signature.asc
Description: Digital signature
-- AppArmor mailing list AppArmor@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/apparmor