31 Dec
2012
31 Dec
'12
3:43 a.m.
On Fri, 28 Dec 2012 18:31:22 +0800, Stefan Tauner stefan.tauner@student.tuwien.ac.at wrote:
On Fri, 28 Dec 2012 16:53:36 +0800 Roy roytam@gmail.com wrote:
Sure. But isn't pcilib yells the prefix path is NULL?
Yes, but *why* was not all that clear without the stack trace. :) Thanks Idwer for asking for it.
I am pretty sure that we try to undo the pci write after we tell pcilib to shutdown - obviously a bad idea and we should...
- guard against this behavior,
- check all existing drivers that they don't do the same nonsense (I
fell for this too in the past).
Roy, can you please try the attached patch (against 0.9.6.1 stable aka r1564; but nothing has changed since then in that file, so you can also use current HEAD.)?
segfault is fixed confirmed. Thanks.