On Fri, Jan 05, 2018 at 12:55:54PM +0000, Jamie Iles wrote:

Hi Greg,

Another panic in the 4.4 PTI backports with CONFIG_PARAVIRT_CLOCK=y:

[ 83.008457] init[1]: segfault at ffffffffff5ff0c0 ip 00007ffdcbf609c5 sp 00007ffdcbeb2fa0 error 5 [ 83.012895] Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b [ 83.012895] [ 83.013764] CPU: 3 PID: 1 Comm: init Not tainted 4.4.110-rc1+ #84 [ 83.014345] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.1-1ubuntu1 04/01/2014 [ 83.015219] ffff880119207b68 ffff8801191579b8 ffffffff8159b31d ffffffff820441c0 [ 83.015981] ffff880119157a90 ffff880119157a80 ffffffff8120cc1a 0000000041b58ab3 [ 83.016724] ffffffff825a770d ffffffff8120cae0 ffff8801191487b8 0000000000000010 [ 83.017464] Call Trace: [ 83.017708] [<ffffffff8159b31d>] dump_stack+0x86/0xc9 [ 83.018197] [<ffffffff8120cc1a>] panic+0x13a/0x283 [ 83.018695] [<ffffffff8120cae0>] ? set_ti_thread_flag+0xf/0xf [ 83.019259] [<ffffffff81109a52>] ? mark_held_locks+0x22/0xb0 [ 83.019802] [<ffffffff81e6bfec>] ? _raw_write_unlock_irq+0x2c/0x40 [ 83.020391] [<ffffffff810964ef>] do_exit+0x96f/0x1480 [ 83.020885] [<ffffffff81095b80>] ? release_task+0x8e0/0x8e0 [ 83.021437] [<ffffffff81109d48>] ? trace_hardirqs_on_caller+0x268/0x2a0 [ 83.022089] [<ffffffff8109932a>] do_group_exit+0xda/0x160 [ 83.022607] [<ffffffff810ad7b3>] get_signal+0xa93/0xba0 [ 83.023132] [<ffffffff81109a52>] ? mark_held_locks+0x22/0xb0 [ 83.023684] [<ffffffff81007e31>] do_signal+0x91/0xab0 [ 83.024171] [<ffffffff8107abaf>] ? force_sig_info_fault.constprop.19+0xef/0x120 [ 83.024887] [<ffffffff81007da0>] ? setup_sigcontext+0x270/0x270 [ 83.025450] [<ffffffff8107aac0>] ? is_prefetch.isra.16+0x260/0x260 [ 83.026041] [<ffffffff8120d3c2>] ? printk+0x99/0xb5 [ 83.026520] [<ffffffff8120d329>] ? power_down+0xa9/0xa9 [ 83.027057] [<ffffffff811027df>] ? up_read+0x1f/0x40 [ 83.027559] [<ffffffff8120d3c2>] ? printk+0x99/0xb5 [ 83.028030] [<ffffffff8107bb55>] ? __bad_area_nosemaphore+0x265/0x2a0 [ 83.028647] [<ffffffff81109a52>] ? mark_held_locks+0x22/0xb0 [ 83.029199] [<ffffffff810023a2>] ? exit_to_usermode_loop+0x52/0xd0 [ 83.029789] [<ffffffff810023c3>] exit_to_usermode_loop+0x73/0xd0 [ 83.030365] [<ffffffff81003840>] prepare_exit_to_usermode+0x60/0x80 [ 83.031067] [<ffffffff81e6cfc8>] retint_user+0x8/0x3c [ 83.031699] Kernel Offset: disabled [ 83.032067] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b [ 83.032067]

This one is from the PVCLOCK_FIXMAP_BEGIN fixmap page, and I can fix it by additionally picking:

6b078f5de7fc (x86, vdso, pvclock: Simplify and speed up the vdso pvclock reader) dac16fba6fc5 (x86/vdso: Get pvclock data from the vvar VMA instead of the fixmap)

onto the stable-rc linux-4.4.y branch.

Nice! That might explain a bunch of this as to why SLES12 works, but this tree doesn't, as those patches are in SLES12. Along with a ton of other ones, digging through 20k of patches is "fun"...

I'll go queue them up now.

greg k-h