On Mon, Sep 14, 2015 at 09:28:18PM +0200, Kurt Roeckx wrote: > On Mon, Sep 14, 2015 at 02:47:39PM -0400, Daniel Kahn Gillmor wrote: > > > > It looks to me like the hang is happening in the code that calibrates > > the s2k count in the agent. (from agent/protect.c) > > > > I've tried to extract the logic from this code and make a simple > > executable that just does it (without invoking gcrypt or anything else > > -- it's just calibrating cpu cycles and walltime clocks. > > So that makes me wonder if it's somehow related to #785557. > csail has just been updated on Saturday to have the workaround, > while the previous build was the day before. > > The qemu issue was seen om more hosts than csail, so I suspect > it's not the problem.
So I just ran an old t-protect binary I had around, and it works now. I'm just going to assume it was the qemu issue for now. Kurt