On Wed, 26 Jun 2019 at 18:19, Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx> wrote: > > Dear RT folks! > > I'm pleased to announce the v5.0.21-rt14 patch set. > > Changes since v5.0.21-rt13: > > - Fix a "scheduling while atomic" in zswap. Patch by Luis Claudio R. > Goncalves. > > - Replace a fix for for wait_for_completion() (introduced in > v5.0.14-rt9) by an alternative version as suggested by Peter > Zijlstra. > > - Major futex/rtmutex surgery. He Zhe reported that glibc's > tst-robustpi8.c triggers a BUG() statement. To address that bug the > hash bucket has been made a raw_spinlock_t. > > - Two x86 FPU patches from upstream to avoid fallout on 32bit and with > CRIU. > > Known issues > - rcutorture is currently broken on -RT. Reported by Juri Lelli. This does not boot for me, using the same kernel config as for 5.0.21-rt13, which works as normal. At boot it seems to hang for a while, before throwing an error, so this seems to be a timeout problem of some sort. I don't know if this helps: Call Trace: d_alloc_parallel+0x1c9/0x410 __lookup_slow+0x66/0x150 lookup_slow+0x31/0x50 walk_component+0x1bf/0x340 link_path_walk.part.0+0x2c1/0x540 path_openat+0xa4/0x1590 do_filp_open+0x8c/0x100 ? _cond_resched+0x10/0x20 ? _kmalloc_track_caller+0x167/0x200 do_open_execat+0x76/0x190 __do_execve_file.isra.0+0x6ad/0x890 ? kmemdup+0x20/0x30 ? umh_complete+0x30/0x30 do_execve+0x1c/0x20 call_usermodehelper_exec_async+0x17b/0x1a0 ret_from_fork+0x35/0x40 Regards, Arve
