mbox

[GIT,PULL] SoCFPGA fixes for v4.6-rc5

Message ID 1461684244-6112-1-git-send-email-dinguyen@kernel.org
State New
Headers show

Pull-request

git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux.git socfpga_fixes_for_v4.6

Message

Dinh Nguyen April 26, 2016, 3:24 p.m. UTC
Hi Arnd, Kevin, Olof:

Please pull in these fixes for v4.6-rc5. The first patch "ARM: SoCFPGA: Fix
secondary CPU startup in thumb2 kernel" is a true fix for enabling CPU1 in
thumb2 mode, but the second patch "ARM: SoCFPGA: simplify secondary trampoline"
may not be a true fix, but it is dependent on the first patch. I didn't want
to cause any conflicts.

Thanks,

Dinh

The following changes since commit 02da2d72174c61988eb4456b53f405e3ebdebce4:

  Linux 4.6-rc5 (2016-04-24 16:17:05 -0700)

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux.git socfpga_fixes_for_v4.6

for you to fetch changes up to 162332eb719bc822c950732c343044bb3158f5fe:

  ARM: SoCFPGA: simplify secondary trampoline (2016-04-26 10:13:49 -0500)

----------------------------------------------------------------
SoCFPGA fixes against v4.6-rc5
- Fix secondary CPU startup in Thumb2 mode
- Simplify secondary CPU startup

----------------------------------------------------------------
Sascha Hauer (2):
      ARM: SoCFPGA: Fix secondary CPU startup in thumb2 kernel
      ARM: SoCFPGA: simplify secondary trampoline

 arch/arm/mach-socfpga/core.h    |  4 ++--
 arch/arm/mach-socfpga/headsmp.S | 28 ++++++++--------------
 arch/arm/mach-socfpga/platsmp.c | 51 +++++++++++++++++++----------------------
 arch/arm/mach-socfpga/socfpga.c |  9 --------
 4 files changed, 34 insertions(+), 58 deletions(-)

Comments

Kevin Hilman April 27, 2016, 10:33 p.m. UTC | #1
Dinh Nguyen <dinguyen@kernel.org> writes:

> Please pull in these fixes for v4.6-rc5. The first patch "ARM: SoCFPGA: Fix
> secondary CPU startup in thumb2 kernel" is a true fix for enabling CPU1 in
> thumb2 mode, but the second patch "ARM: SoCFPGA: simplify secondary trampoline"
> may not be a true fix, but it is dependent on the first patch. I didn't want
> to cause any conflicts.

IIUC, the 2nd one (simplify) seems to be dependent on the first (fix),
but doesn't seem to be appropriate for this late stage of the -rc cycle,
so I've applied the first one only.  Please include the 2nd one with
your v4.7 updates.

I also verified on my socfpga board that the 2nd CPU does indeed come up
with this patch when building with THUMB2_KERNEL=y.

Thanks,

Kevin