Le 08/08/2017 à 04:16, Michael Ellerman a écrit :

Christophe LEROY [off-list ref] writes:

quoted

Le 07/08/2017 à 12:41, Michael Ellerman a écrit :

quoted

On Wed, 2017-07-19 at 04:49:23 UTC, Benjamin Herrenschmidt wrote:

quoted

This will allow simplifying the returns from do_page_fault

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Series applied to powerpc next, thanks.

https://git.kernel.org/powerpc/c/7afad422ac61067473d5f3d20bbd54

Boot failure on the 8xx:

[    6.029556] Failed to execute /init (error -14)
[    6.034623] Starting init: /bin/sh exists but couldn't execute it
(error -14)
[    6.041489] Kernel panic - not syncing: No working init found.  Try
passing init= option to kernel. See Linux
Documentation/admin-guide/init.rst for guidance.
[    6.055518] CPU: 0 PID: 1 Comm: init Not tainted
4.13.0-rc3-s3k-dev-00143-g7aa62e972a56 #56
[    6.063745] Call Trace:
[    6.066224] [c60f1ed0] [c001a624] panic+0x108/0x250 (unreliable)
[    6.072140] [c60f1f30] [c0002640] rootfs_mount+0x0/0x58
[    6.077311] [c60f1f40] [c000cb80] ret_from_kernel_thread+0x5c/0x64
[    6.083405] Rebooting in 180 seconds..


Bisected to c433ec0455f921eaf8dd0262a718ce6f8ad62ea2 ("powerpc/mm:
Pre-filter SRR1 bits before do_page_fault()")

Sorry about that. I don't have a way to test 8xx.

Looks like I concluded too quickly yesterday night, indeed the above 
commit is the last good one. The faulty one is 
d300627c6a53693fb01479b59b0cdd293761b1fa("powerpc/6xx: Handle DABR match 
before calling do_page_fault")

quoted

Will see tomorrow what's wrong with that.

Thanks.

This is because the 'bl do_page_fault' has been enclosed inside the 
#ifdef CONFIG_6xx

See patch at https://patchwork.ozlabs.org/patch/799039/

Christophe