The problem is this is a generic error message. A "batchbuffer" is basically just a chunk of graphics data, which gets sent to the kernel graphics processing unit (GPU). X will submit these chunks to the data for display. X sets a timer to check back for if this data was taken and processed by the kernel. If the kernel doesn't handle it within that amount of time, X concludes that the kernel has gone to lunch. In other words, it decides the GPU is hung since the hang check timer has expired. That's why you see these kinds of error messages:
Hangcheck timer elapsed... GPU hung It just means something broke in the kernel, and X doesn't know what to do. Messages about 'hardware wedged', 'GPU lockup', and 'failed to submit batchbuffer' all mean the same thing. In all these cases, the GPU has hung due to some flaw in the kernel's intel drm driver. And just as there can be many ways to crash a car, there can be myriad ways to lockup a GPU. The error messages are generic, and don't really tell *why* the crash happened. Indeed, those 60 different people could be seeing 60 completely different root bugs. (Well, probably not, but it's likely to be a variety.) The bad news is that these GPU lockups were relatively common in Lucid, and kernel tools for analyzing them were in their infancy. In other words, there are not ways to get useful debugging information on these, without a TON of work. The good news is that over the last couple years Intel has added a number of useful tools and processes for isolating and solving these kinds of bugs. And doubly good is that they've utilized those tools to solve a vast number of known GPU lockup issues, such that in Oneiric and Precise, GPU lockups on -intel are a lot less common. So, if you are just looking for a workaround, I would suggest installing a newer kernel. We also have newer X drivers and newer mesa drivers, which sometimes can help - see the x-updates PPA, xorg-edgers, or other PPAs with newer drivers. On the other hand, if you're looking to help us improve Lucid itself, the steps that need to be taken are to do a git bisect on the kernel (or mesa or the -intel DDX) and isolate what git commit id marks the change between broken and fixed. With that info, we can consider an SRU of the patch to Lucid. Meanwhile, I will move this bug to the kernel team. Since GPU lockups are generally due to bugs in linux intel drm, that is the right place for these kinds of bug reports. ** Package changed: xserver-xorg-video-intel (Ubuntu) => linux (Ubuntu) ** Changed in: linux (Ubuntu) Status: Confirmed => New ** Summary changed: - [g41] GPU hung: Failed to submit batchbuffer: Input/output error + [g41] GPU hung: Failed to submit batchbuffer: Input/output error [lucid] -- You received this bug notification because you are a member of Ubuntu-X, which is subscribed to xserver-xorg-video-intel in Ubuntu. https://bugs.launchpad.net/bugs/555573 Title: [g41] GPU hung: Failed to submit batchbuffer: Input/output error [lucid] To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/555573/+subscriptions _______________________________________________ Mailing list: https://launchpad.net/~ubuntu-x-swat Post to : ubuntu-x-swat@lists.launchpad.net Unsubscribe : https://launchpad.net/~ubuntu-x-swat More help : https://help.launchpad.net/ListHelp