[GIT PULL] Please pull powerpc/linux.git powerpc-5.10-1 tag

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Hi Linus,

Please pull powerpc updates for 5.10.

Just two minor conflicts I'm aware of. The only slight subtlety is the conflict
in kasan_init() where "int ret" needs to move out of the for_each_mem_range()
and up to the function scope.

Notable out of area changes:
  arch/sparc/kernel/smp_64.c	# bafb056ce279 sparc64: remove mm_cpumask clearing to fix kthread_use_mm race
  fs/exec.c			# d53c3dfb23c4 mm: fix exec activate_mm vs TLB shootdown and lazy tlb switching race
  include/uapi/linux/mman.h	# e47168f3d1b1 powerpc/8xx: Support 16k hugepages with 4k pages
  include/uapi/asm-generic/hugetlb_encode.h	# as above

cheers


The following changes since commit d012a7190fc1fd72ed48911e77ca97ba4521bccd:

  Linux 5.9-rc2 (2020-08-23 14:08:43 -0700)

are available in the git repository at:

  https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux.git tags/powerpc-5.10-1

for you to fetch changes up to ffd0b25ca049a477cb757e5bcf2d5e1664d12e5d:

  Revert "powerpc/pci: unmap legacy INTx interrupts when a PHB is removed" (2020-10-15 13:42:49 +1100)


- ------------------------------------------------------------------
powerpc updates for 5.10

 - A series from Nick adding ARCH_WANT_IRQS_OFF_ACTIVATE_MM & selecting it for
   powerpc, as well as a related fix for sparc.

 - Remove support for PowerPC 601.

585 more lines

On Thu, Oct 15, 2020 at 8:24 PM Michael Ellerman [off-list ref] wrote:

Just two minor conflicts I'm aware of. The only slight subtlety is the conflict
in kasan_init() where "int ret" needs to move out of the for_each_mem_range()
and up to the function scope.

Well, there was also a conflict for the dependencies of OCXL.

I resolved that by ruthlessly simplifying the dependency:

-       depends on PPC_POWERNV && PCI && EEH && HOTPLUG_PCI_POWERNV
 -      depends on PPC_POWERNV && PCI && EEH && PPC_XIVE_NATIVE
++      depends on HOTPLUG_PCI_POWERNV

because all the other dependencies seem to be pointless.

HOTPLUG_PCI_POWERNV already has a

        depends on PPC_POWERNV && EEH

so there's no point in repeating those.

And PPC_XIVE_NATIVE is selected by PPC_POWERNV, so if PPC_POWERNV, we
know PPC_XIVE_NATIVE is set.

Maybe I missed something strange, so I'm just letting you know so you
can blame me if I broke something.

              Linus

The pull request you sent on Fri, 16 Oct 2020 14:24:39 +1100:

https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux.git tags/powerpc-5.10-1

has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/96685f8666714233d34abb71b242448c80077536

Thank you!

-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/prtracker.html

Linus Torvalds [off-list ref] writes:

On Thu, Oct 15, 2020 at 8:24 PM Michael Ellerman [off-list ref] wrote:

quoted

Just two minor conflicts I'm aware of. The only slight subtlety is the conflict
in kasan_init() where "int ret" needs to move out of the for_each_mem_range()
and up to the function scope.

Well, there was also a conflict for the dependencies of OCXL.

Yep, it was so trivial I didn't think it was worth calling out.

I resolved that by ruthlessly simplifying the dependency:

-       depends on PPC_POWERNV && PCI && EEH && HOTPLUG_PCI_POWERNV
 -      depends on PPC_POWERNV && PCI && EEH && PPC_XIVE_NATIVE
++      depends on HOTPLUG_PCI_POWERNV

because all the other dependencies seem to be pointless.

HOTPLUG_PCI_POWERNV already has a

        depends on PPC_POWERNV && EEH

so there's no point in repeating those.

And PPC_POWERNV selects FORCE_PCI, which takes care of the dependency on
PCI.

And PPC_XIVE_NATIVE is selected by PPC_POWERNV, so if PPC_POWERNV, we
know PPC_XIVE_NATIVE is set.

Maybe I missed something strange, so I'm just letting you know so you
can blame me if I broke something.

No, that looks good to me. Thanks.

cheers