The further a society drifts from the truth, the more it will hate those who speak it. ... In a time of deceit, telling the truth is a revolutionary act. George Orwell

Saturday, August 6, 2011

Warning MS C++ 2010 runtime upgrade breaks ATI graphics drivers! (off-topic)

I appologise for this quick off-topic warning post, but this info might save many of you from throwing out a perfectly good ATI Radeon 4xxx series that suddenly stopped working.  I use ATI HD4670 in Win7 32bit, but this may apply also to other series of ATI cards.

A very recent upgrade of Microsoft C++ run-time library version 2011.0.30xx to 2011.0.40xx (Win7 32bit) breaks ATI Radeon graphics drivers. The symptoms are similar to when a graphic card breaks due to a hardware failure, that is screen goes suddenly black during a boot up, in high resolution large screen mode, while it may still be working when booted in the "safe mode".

I haven't tracked down yet where that .40 upgrade came from but I have determined it by multiple trial and errors beyond reasonable doubt that it is caused by the Windows 7 system run-time library incompatibility.

To fix the problem:

1) If your screen is already black, shut down, remove ATI Radeon graphics card, plug in some other card (Nvidia or something much older, even a standard VGA will do). Reboot.

2) Open Control Panel, Programs and Features - Check if you have MS C++ 2010.0.40xx Run-time installed. If yes then uninstall it.

3) Go to go to ATI Radeon downloads section, download and install "ATI Catalyst Center" (this will install all necessary graphics drivers for the ATI Radeon series) . I installed an older version (10.12) but the most recent one (11.7) should also work They all install MS C++ 2010.0.30 run-time library, as part of their overall installation process - verify the log that it was installed OK! If you forgot to uninstall version 40 then this step of "ATI Catalyst" installation will report a failure!

4) Shut down the PC, remove the old graphics card, reinstall ATI Radeon, restart.


dav0 said...

Very strange. You would imagine that the runtime libs could potentially cause a problem after Win is loaded. Does it go dark after BIOS initialisation before running Windows? I'm not nailing my flag to a mast but it is also possible that the ATI driver is not strictly working by the book and perhaps using some user level resources it shouldn't. It will be interesting to see how it pans out, you should log an MS support request and see if it generates a KB article. As you know, kernel mode code is like a pact with the devil. If you do everything by the book it should work but if you break the contract, you're doomed. My guess is that they're relying on some user mode callback that doesn't exist. 0.02$, Dave.

john said...

Stan, do you have--and not mind sharing--an e-mail address?

Stan (Heretic) said...

at hotmail
dot com

PigeonOrStatue said...

Also off topic.
Have you seen
This is why I thought of you: "The real question is do electrons from different macronutrients have specific quantum biologic effects?"

Stan (Heretic) said...

This article is a bit confused. Einstein did not formulate the quantum non-locality theory, he fought it tooth and nail! Quantum Non-Locality ideas really became fully accepted as the physical reality (as opposed to being just a computational shortcuts) by the mainstream academics, in the 1970-ties following Alan Aspect's experimentally proving the so-called Bell's Inequality. This basically, experimentally invalidated any theory that excluded or avoided the Non-Locality (e.g. "Hidden Variable" theories etc). It also proved Einstein wrong, at least in his opposition to quantum mechanics.

Electrons do not have individual identities, according to the known physics, as far as we know. Electrons are undistinguishable from each other!

In a sense the Universe behaves as if there were realy One Electron, weaving its way around the Universe(s), wrapping its trajectory around it a very large number of times (like 10^40 or 10^80 times); forward in time as electron, and backwards in time as positron.