Hi Ben, On Sun, Feb 10, 2019 at 05:07:07PM +0000, Ben Hutchings wrote: > On Tue, 3 Jul 2018 18:51:49 -0700 Ryan Tandy <r...@nardis.ca> wrote: > > Control: found -1 4.17.3-1 > > > > Hi, > > > > On Mon, Jul 02, 2018 at 08:42:45PM -0700, Ryan Tandy wrote: > > >However it looks like it might be resolved with the kernel in > > >unstable. I will run some more iterations and let you know. > > > > The issue still reproduces with stretch userland and the current > > unstable kernel. > > > > It cannot be reproduced with buster's glibc because lock elision was > > disabled in #878071, citing this issue. > > This bug has been open for over 18 months with no sign of a solution, > which is really not acceptable for a release architecture. I propose > to disable CONFIG_PPC_TRANSACTIONAL_MEMORY in both stretch and > unstable, until a fix is available for TM.
We're getting pretty close to the release. Are you planning to disable CONFIG_PPC_TRANSACTIONAL_MEMORY for buster? Thanks, Ivo
