Try and make sure the camera is exactly parallel to the wall and don't shoot at the widest aperture, stop it down to about f10 to give it a good chance to get the wall completely in focus.
With all due respect, stopping down to f/10 is not much of a test of any lens. It needs to work well wide open, and even better stopped down.
As an example, at 15mm and f/10, for a subject 3' away, your DOF runs from 1.7' to 12.8'. Well that gives a pretty massive safety margin for focus inaccuracy and sharpness.
I would suggest trying some of the settings (aperture and focal length) listed here -
http://www.photozone.de/canon-eos/3...-56-hsm-ex-dc-lab-test-report--review?start=1 - to put the lens through its paces. There's no need to go mad but I'd say nine shots should cover it - 3 focal lengths at three apertures each. Decent light, low ISO, a solid tripod and mirror lockup and remote release go without saying, of course
You can infer from the web page that things to look out for are distortion, vignetting, chromatic abberation and sharpness (at the centre, edges and corners). It's hard to check AF accuracy with a wide angle because the DOF is so huge, but there's no harm checking that things do look sharp when using AF, and for that to be effective you'll need to be wide open to get the narrowest DOF you can.
Taking things the extra mile, it is commonly said that a lens should be tested at distances that represent real world use, rather than a contrived environment. So, for example, if you're going to use the lens for landscapes then test its performance with the scene/subject some distance away. If you're going to use it close up for unusual distorted perspectives then test it close up (but make sure you are not closer than the MFD). In the absence of any specific intentions for use I've often read that you should test a lens at around 40X the focal length. So at 10mm FL you should be testing with a subject 40cm away and at 20mm your test subject should be 80cm away.
Just some food for thought....
