On 20 August 2014 13:08, Renato Golin <renato.golin at linaro.org> wrote: > 850 is bad, I'm testing 846 now. 846 and 845 are both bad, and ever since it first regressed, 20 new failures have been creeping in... :( --renato