Comment 22 for bug 1058303

I've reporter this upstream. It looks like commit 62444b7462a2b98bc78d68736c03a7c4e66ba7e2 ("drm/radeon: properly handle
mc_stop/mc_resume on evergreen+ (v2)") should fix it without the need to revert the offending commit. Thus, I've built a test kernel that I would like you to try. Unfortunately, the commit had a few dependencies and I had to cherry-pick a few other commits. Here's the list of commits included in the test kernel:

3ae19b750bdc09ce233e1504348320141593ffda drm/radeon/kms: add wait_for_vblank asic callback
89e5181f3f79fbe46dbf811f9bc470f524704450 drm/radeon/kms: add a radeon asic callback for mc idle
8848f759c0de139bada271aa446f48637c98ca1a drm/radeon/kms: add trinity (TN) chip family
4a15903db02026728d0cf2755c6fabae16b8db6a drm/radeon/dce4+: don't use radeon_crtc for vblank callback
62444b7462a2b98bc78d68736c03a7c4e66ba7e2 drm/radeon: properly handle mc_stop/mc_resume on evergreen+ (v2)

Here's the link to the kernel:

http://people.canonical.com/~henrix/lp1058303/v5/