Re: xf86UnloadSubModule is still buggy

2012-06-04 Thread Michal Suchanek
Hello, The patch just enables unloading modules. That is the right thing to do IMHO. If unloading siblings is not expected then it should not be done. That is not something this patch introduced, it has been there unnoticed for years. Thanks Michal

Re: xf86UnloadSubModule is still buggy

2012-06-04 Thread Michal Suchanek
Hello, I don't see why the unload hunk can't be backed out separately. It's just something that I found when looking at the loader. As for fixing the sibling unloading - the siblings are a FIFO so you should unload only the top one, the first loaded. This can be fixed in the driver. Try one IO

xf86UnloadSubModule is still buggy

2012-06-03 Thread Matthieu Herrb
Hi commit 0d4bb5442ceb8e8e4a8de6cfc4203cae469eee72 Author: Michal Suchanek hramr...@centrum.cz Date: Sat Oct 8 14:13:33 2011 +0200 Unload submodules. Signed-off-by: Michal Suchanek hramr...@centrum.cz Reviewed-by: Peter Hutterer peter.hutte...@who-t.net diff --git

Re: xf86UnloadSubModule is still buggy

2012-06-03 Thread Peter Hutterer
cc-ing Michal On Sun, Jun 03, 2012 at 10:39:37PM +0200, Matthieu Herrb wrote: Hi commit 0d4bb5442ceb8e8e4a8de6cfc4203cae469eee72 Author: Michal Suchanek hramr...@centrum.cz Date: Sat Oct 8 14:13:33 2011 +0200 Unload submodules. Signed-off-by: Michal Suchanek