diff options
| author | Michael Ellerman <mpe@ellerman.id.au> | 2017-11-24 14:51:02 +1100 |
|---|---|---|
| committer | Michael Ellerman <mpe@ellerman.id.au> | 2017-11-24 16:49:37 +1100 |
| commit | 2621e945fbf1d6df5f3f0ba7be5bae3d2cf9b6a5 (patch) | |
| tree | a6129c61ec962f7aeeccb1f166051c02d5c2c1d1 /arch/powerpc | |
| parent | a3961f824cdbe7eb431254dc7d8f6f6767f474aa (diff) | |
| download | op-kernel-dev-2621e945fbf1d6df5f3f0ba7be5bae3d2cf9b6a5.zip op-kernel-dev-2621e945fbf1d6df5f3f0ba7be5bae3d2cf9b6a5.tar.gz | |
powerpc/kexec: Fix kexec/kdump in P9 guest kernels
The code that cleans up the IAMR/AMOR before kexec'ing failed to
remember that when we're running as a guest AMOR is not writable, it's
hypervisor privileged.
They symptom is that the kexec stops before entering purgatory and
nothing else is seen on the console. If you examine the state of the
system all threads will be in the 0x700 program check handler.
Fix it by making the write to AMOR dependent on HV mode.
Fixes: 1e2a516e89fc ("powerpc/kexec: Fix radix to hash kexec due to IAMR/AMOR")
Cc: stable@vger.kernel.org # v4.10+
Reported-by: Yilin Zhang <yilzhang@redhat.com>
Debugged-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Acked-by: Balbir Singh <bsingharora@gmail.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Tested-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Diffstat (limited to 'arch/powerpc')
| -rw-r--r-- | arch/powerpc/kernel/misc_64.S | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S index 8ac0bd2..3280953 100644 --- a/arch/powerpc/kernel/misc_64.S +++ b/arch/powerpc/kernel/misc_64.S @@ -623,7 +623,9 @@ BEGIN_FTR_SECTION * NOTE, we rely on r0 being 0 from above. */ mtspr SPRN_IAMR,r0 +BEGIN_FTR_SECTION_NESTED(42) mtspr SPRN_AMOR,r0 +END_FTR_SECTION_NESTED_IFSET(CPU_FTR_HVMODE, 42) END_FTR_SECTION_IFSET(CPU_FTR_ARCH_300) /* save regs for local vars on new stack. |
