Commit Graph

692141 Commits

Author SHA1 Message Date
Nicholas Piggin
bded070643 powerpc/pseries/le: Work around a firmware quirk
Some PowerVM firmware when delivering a system reset interrupt to a
little endian OS will mess up SRR registers. They are byteswapped, and
SRR1 is incorrect. An example from a crash:

  NIP: 14dd0900000000c0
  MSR: 1000000200000080

It's possible to detect this pattern in SRR1 (that would never happen
in normal operation), and at least fix the NIP. After this patch, the
same interrupt reports NIP properly:

  NIP [c00000000009dd14] plpar_hcall_norets+0x1c/0x28

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-31 14:26:02 +10:00
Nicholas Piggin
a3b2cb30f2 powerpc: Do not call ppc_md.panic in fadump panic notifier
If fadump is not registered, and no other crash or debug handlers are
registered, the powerpc panic handler stops the guest before the
generic panic code can push out debug information to the console.

Currently, system reset injection causes the guest to silently stop.

Stop calling ppc_md.panic in the panic notifier. crash_fadump already
does rtas_os_term() to terminate the guest if fadump is registered.

Remove ppc_md.panic. Move fadump panic notifier into fadump code.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-31 14:26:01 +10:00
Nicholas Piggin
70412c55d4 powerpc/64: Fix watchdog configuration regressions
This fixes a couple more bits of fallout from the new hard lockup watchdog
patch.

It restores the required hw_nmi_get_sample_period() function for the
perf watchdog, and removes some function declarations on 64e that are only
defined for 64s. This fixes the 64e build when the hardlockup detector is
enabled.

It restores the default behaviour of disabling the perf watchdog, and also
fixes disabling the 64s watchdog when running as a guest.

Fixes: 2104180a53 ("powerpc/64s: implement arch-specific hardlockup watchdog")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-31 14:26:00 +10:00
Nicholas Piggin
b68b1d7487 powerpc/64s/radix: Do not allocate SLB shadow structures
These are unused in radix mode.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-31 14:25:59 +10:00
Nicholas Piggin
d55071905e powerpc/64s/radix: Remove bolted-SLB address limit for per-cpu stacks
Radix MMU does not take SLB or TLB interrupts when accessing kernel
linear address. Remove this restriction for radix mode.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-31 14:25:59 +10:00
Nicholas Piggin
76b42e28be powerpc/powernv: powernv platform is not constrained by RMA
Remove incorrect comment about real mode address restrictions on
powernv (bare metal), and unnecessary clamping to ppc64_rma_size.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-31 14:25:58 +10:00
Nicholas Piggin
72b0d51d97 powerpc/64s: idle POWER9 can execute stop in virtual mode
The hardware can execute stop in any context, and KVM does not
require real mode because siblings do not share MMU state. This
saves a switch to real-mode when going idle.

Acked-by: Gautham R. Shenoy <ego@linux.vnet.ibm.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-29 21:42:14 +10:00
Nicholas Piggin
65dbbe812f powerpc/64s: Drop no longer used IDLE_STATE_ENTER_SEQ
There are no longer any callers of IDLE_STATE_ENTER_SEQ, all callers
use IDLE_STATE_ENTER_SEQ_NORET. So drop the former.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch, write change log]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-29 21:41:44 +10:00
Nicholas Piggin
56ee52408e powerpc/64s: POWER9 can execute stop without a sync sequence
We don't need to use IDLE_STATE_ENTER_SEQ_NORET on Power9.

Reviewed-by: Gautham R. Shenoy <ego@linux.vnet.ibm.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-29 21:39:07 +10:00
Nicholas Piggin
aafc8a8300 powerpc/64s: Move IDLE_STATE_ENTER_SEQ[_NORET] into idle_book3s.S
This macro is only used in idle_book3s.S, move it in there and add a
more descriptive comment.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch and write change log]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-29 21:38:47 +10:00
Michael Ellerman
82b7fcc005 Merge branch 'topic/ppc-kvm' into next
Merge Nicks commit to rework the KVM thread management, shared with the
KVM tree via the ppc-kvm topic branch.
2017-08-29 21:26:30 +10:00
Nicholas Piggin
94a04bc25a KVM: PPC: Book3S HV: POWER9 does not require secondary thread management
POWER9 CPUs have independent MMU contexts per thread, so KVM does not
need to quiesce secondary threads, so the hwthread_req/hwthread_state
protocol does not have to be used. So patch it away on POWER9, and patch
away the branch from the Linux idle wakeup to kvm_start_guest that is
never used.

Add a warning and error out of kvmppc_grab_hwthread in case it is ever
called on POWER9.

This avoids a hwsync in the idle wakeup path on POWER9.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Acked-by: Paul Mackerras <paulus@ozlabs.org>
[mpe: Use WARN(...) instead of WARN_ON()/pr_err(...)]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-29 14:48:59 +10:00
Michael Ellerman
12c15a7e70 powerpc/configs/6xx: Drop removed CONFIG_USB_LED
In commit a335aaf312 ("usb: misc: remove outdated USB LED driver")
CONFIG_USB_LED was removed, so drop it from our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:25 +10:00
Michael Ellerman
b1f9a827e4 powerpc/configs/6xx: Drop no longer selectable CONFIG_BT_HCIUART_LL
Since commit 76c4969fec ("Bluetooth: hci_uart: fix kconfig
dependency") we can no longer select CONFIG_BT_HCIUART_LL.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:24 +10:00
Michael Ellerman
8f67600f21 powerpc/configs/c2k: Switch CONFIG_GEN_RTC from =m to =y
In commit 835ea93e9d ("char/genrtc: remove powerpc support"),
CONFIG_GEN_RTC switch from tristate to bool, update the defconfig to
match.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:24 +10:00
Michael Ellerman
2d8b1ca3d9 powerpc/configs/6xx: Switch CONFIG_USB_EHCI_FSL to =m
In commit ca07e1c1e4 ("drivers:usb:fsl:Make fsl ehci drv an
independent driver module"), CONFIG_USB_EHCI_FSL was switched from
built-in to modular. Update the defconfig.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:23 +10:00
Michael Ellerman
dcb5956154 powerpc/configs/6xx: Drop no longer needed CONFIG_BT_HCIUART_H4
Since commit 943cc59219 ("Bluetooth: bpa10x: Use h4_recv_buf helper
for frame reassembly") we no longer need to set CONFIG_BT_HCIUART_H4
in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:23 +10:00
Michael Ellerman
c29f9b31bb powerpc/configs/6xx: Drop no longer needed CONFIG_NETFILTER_XT_MATCH_SOCKET
Since commit 8db4c5be88 ("netfilter: move socket lookup
infrastructure to nf_socket_ipv{4,6}.c") we no longer need to set
CONFIG_NETFILTER_XT_MATCH_SOCKET in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:22 +10:00
Michael Ellerman
6946d5e1fa powerpc/configs/6xx: Reinstate CONFIG_CPU_FREQ_STAT
In commit 1aefc75b24 ("cpufreq: stats: Make the stats code
non-modular"), the CPU_FREQ_STAT code was made non-modular. Our
defconfig still said =m though, which meant we no longer got the
code at all. Switch the defconfig to =y.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:21 +10:00
Michael Ellerman
886a3bacb7 powerpc/configs/6xx: Drop no longer needed CONFIG_NF_CONNTRACK_PROC_COMPAT
Since commit adf0516845 ("netfilter: remove ip_conntrack* sysctl
compat code") we no longer need to set CONFIG_NF_CONNTRACK_PROC_COMPAT
in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:20 +10:00
Michael Ellerman
1d66e404b9 powerpc/configs/6xx: Drop removed CONFIG_BLK_DEV_HD
In commit 8e14be53f4 ("remove the obsolete hd driver") the
CONFIG_BLK_DEV_HD symbol was removed, so drop it from the defconfig.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:20 +10:00
Michael Ellerman
533141ae0e powerpc/configs/6xx: Clean up duplicate CONFIG_EXT4 values
We had two values for CONFIG_EXT4, =m and =y, just use =y.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:19 +10:00
Michael Ellerman
f623a54e83 powerpc/configs/6xx: Drop no longer needed CONFIG_TIMER_STATS
Since commit dfb4357da6 ("time: Remove CONFIG_TIMER_STATS") we no
longer need to set CONFIG_TIMER_STATS in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:19 +10:00
Michael Ellerman
f963937a16 powerpc/configs/6xx: Turn CONFIG_DRM_RADEON back on
In commit d92d9c3a14 ("drm: hide legacy drivers with CONFIG_DRM_LEGACY")
CONFIG_DRM_RADEON was moved behind CONFIG_DRM_LEGACY meaning it
stopped being enabled by ppc6xx_defconfig. Although no one has
noticed, given this is basically a legacy platform, it seems anyone
who is using it probably still wants this driver. So turn it back on
for now.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:18 +10:00
Michael Ellerman
22220b16df powerpc/configs/mpc5200: Drop no longer needed CONFIG_FB
Since commit a03fdcb186 ("drm: Add top level Kconfig option for DRM
fbdev emulation") we no longer need to set CONFIG_FB in our
defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:18 +10:00
Michael Ellerman
edff694087 powerpc/configs: Update for CONFIG_INPUT_MOUSEDEV=n
In commit 73d8ef7600 ("Input: mousedev - stop offering PS/2 to
userspace by default") the symbol INPUT_MOUSEDEV went from being
'default y' to 'default n' (implied).

That means we no longer need to explicitly disable it in our
defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:17 +10:00
Michael Ellerman
f6fe7a1583 powerpc/configs: Drop removed CONFIG_LOGFS
In commit 1d0fd57a50 ("logfs: remove from tree"), logfs was removed
from the tree, so we can drop it from our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:17 +10:00
Michael Ellerman
dc3f5d2168 powerpc/configs: Turn CONFIG_R128 back in pmac32_defconfig
In commit d92d9c3a14 ("drm: hide legacy drivers with
CONFIG_DRM_LEGACY") CONFIG_R128 was moved behind CONFIG_DRM_LEGACY
meaning it stopped being enabled by pmac32_defconfig. Although no one
has noticed, given this is basically a legacy platform, it seems
anyone who is using it probably still wants this driver. So turn it
back on for now.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:16 +10:00
Michael Ellerman
641f656db0 powerpc/configs: Drop no longer needed CONFIG_LIBCRC32C
Since commit 300ae14946 ("netfilter: select LIBCRC32C together with
SCTP conntrack") we no longer need to set CONFIG_LIBCRC32C in our
defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:16 +10:00
Michael Ellerman
ab306d1811 powerpc/configs: Drop unnecessary CONFIG_EDAC from ppc64e
There are no EDAC drivers for ppc64e.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:15 +10:00
Michael Ellerman
5593edb176 powerpc/configs: Drop no longer needed CONFIG_SCSI
Since commit 67f6d66559 ("powerpc: convert amigaone_defconfig to use
libata PATA drivers") we no longer need to set CONFIG_SCSI in our
defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:15 +10:00
Michael Ellerman
72ac99688d powerpc/configs: Drop no longer needed CONFIG_IPV6
Since commit de551f2eb2 ("net: Build IPv6 into kernel by default")
we no longer need to set CONFIG_IPV6 in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:14 +10:00
Michael Ellerman
aad004a945 powerpc/configs: Add CONFIG_RAS now required for CONFIG_EDAC
In commit e3c4ff6d8c ("EDAC: Remove EDAC_MM_EDAC") CONFIG_EDAC grew
a dependency on CONFIG_RAS. Some of our defconfigs don't have the
latter, which means we lose CONFIG_EDAC, so add CONFIG_RAS to fix
that.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:14 +10:00
Michael Ellerman
f3a45e560f powerpc/configs: Drop no longer needed CONFIG_AUDITSYSCALL
Since commit cb74ed278f ("audit: always enable syscall auditing when
supported and audit is enabled") we no longer need to set
CONFIG_AUDITSYSCALL in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:13 +10:00
Michael Ellerman
599f959fcf powerpc/configs: Drop CONFIG_SERIAL_TXX9_* from cell/ppc64
In commit bf4981a006 ("powerpc: Remove the celleb support") we
dropped the celleb support, which made these symbols unselectable
because we no longer select HAS_TX99_SERIAL. So drop them from the
defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:13 +10:00
Michael Ellerman
e9cb40a7d6 powerpc/configs: Drop MEMORY_HOTREMOVE from ppc64/cell
xxxx

In commit 577ec789a7 ("powerpc/cell: Drop select of MEMORY_HOTPLUG")
we removed the last traces of any dependency between

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:12 +10:00
Michael Ellerman
5a73b1a18a powerpc/configs: Drop unnecessary CONFIG_POWERNV_OP_PANEL
In commit 43a1dd9b5f ("powerpc/powernv: Add driver for operator
panel on FSP machines") we added CONFIG_POWERNV_OP_PANEL=m to the
powernv defconfig, but it's default m so that's no necessary.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:11 +10:00
Michael Ellerman
8b2ee33d88 powerpc/configs: Drop no longer needed PCI_MSI on powernv
In commit a311e738b6 ("powerpc/powernv: Make PCI non-optional") we
made PCI (and therefore PCI_MSI) non-optional on powernv, so it
doesn't need to be in the defconfig anymore.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:11 +10:00
Michael Ellerman
4ca2ddfc47 powerpc/configs: Drop no longer needed CONFIG_SMP for pseries/ppc64/powernv
In commit 40e275653e ("powerpc/powernv: Always enable SMP when
building powernv") and 270e2dc9b8 ("powerpc/pseries: Always enable
SMP when building pseries") we forced CONFIG_SMP on for some configs.
Therefore we don't need to set it in those configs anymore.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:10 +10:00
Michael Ellerman
b9bc4bbc7c powerpc/configs: Drop unnecessary CONFIG_UPROBE_EVENT
In commit 6b0b755142 ("perf/core: Rename CONFIG_[UK]PROBE_EVENT to
CONFIG_[UK]PROBE_EVENTS") it was renamed to CONFIG_UPROBE_EVENTS.

Additionally it's default y, and we have the prerequisites enabled, so
we don't need it in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:10 +10:00
Michael Ellerman
93a4a4ffeb powerpc/configs: Drop unnecessary CONFIG_NUMA_BALANCING_DEFAULT_ENABLED
In commit 9654f95a08 ("powerpc: Enable NUMA balancing in
pseries[_le]_defconfig") we added CONFIG_NUMA_BALANCING_DEFAULT_ENABLED
to our defconfigs. But it's already enabled by default, so drop it.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:09 +10:00
Michael Ellerman
76869cb569 powerpc/configs: Drop no longer needed CONFIG_DEVPTS_MULTIPLE_INSTANCES
Since commit eedf265aa0 ("devpts: Make each mount of devpts an
independent filesystem.") we no longer need to set
CONFIG_DEVPTS_MULTIPLE_INSTANCES in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:09 +10:00
Michael Ellerman
cd67a20f77 powerpc/configs: Drop no longer needed CONFIG_CRYPTO_GCM
Since commit 00b9cfa3ff ("mac80111: Add GCMP and GCMP-256 ciphers")
we no longer need to set CONFIG_CRYPTO_GCM in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:08 +10:00
Michael Ellerman
f0297310e5 powerpc/configs: Drop no longer needed CONFIG_CRYPTO_NULL in g5 / c2k
Since commit 3491244c62 ("crypto: echainiv - Set Kconfig default to
m") we no longer need to set CONFIG_CRYPTO_NULL in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:08 +10:00
Michael Ellerman
60ac25ba05 powerpc/configs: Drop no longer needed CONFIG_CRYPTO_NULL
Since commit 00b9cfa3ff ("mac80111: Add GCMP and GCMP-256 ciphers")
we no longer need to set CONFIG_CRYPTO_NULL in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:07 +10:00
Michael Ellerman
7cf5775cd0 powerpc/configs: Drop no longer needed CONFIG_CRYPTO_SHA256
Since commit 826775bbf3 ("crypto: drbg - Add select on sha256") we
no longer need to set CONFIG_CRYPTO_SHA256 in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:07 +10:00
Michael Ellerman
813f41352a powerpc/configs: Drop no longer needed CONFIG_CRYPTO_ECB
Since commit 12cb3a1c41 ("crypto: xts - Add ECB dependency") we no
longer need to set CONFIG_CRYPTO_ECB in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:06 +10:00
Michael Ellerman
8f56939009 powerpc/configs: Drop no longer needed CONFIG_CRYPTO_HMAC
Since commit 401e4238f3 ("crypto: rng - Make DRBG the default RNG")
we no longer need to set CONFIG_CRYPTO_HMAC in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:05 +10:00
Michael Ellerman
8f3e6bdf8c powerpc/configs: Drop no longer needed CONFIG_CRYPTO_DEV_VMX_ENCRYPT
Since commit ccf5c442a1 ("crypto: vmx - Convert to CPU feature based
module autoloading") we no longer need to set
CONFIG_CRYPTO_DEV_VMX_ENCRYPT in our defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:05 +10:00
Michael Ellerman
b8465a6ae5 powerpc/configs: Update for CONFIG_NF_CT_PROTO_(SCTP|UDPLITE)=y
In commit a85406afeb ("netfilter: conntrack: built-in support for
SCTP"), NF_CT_PROTO_SCTP switched from tristate to bool and became
default y. Similarly in commit 9b91c96c5d ("netfilter: conntrack:
built-in support for UDPlite"), NF_CT_PROTO_UDPLITE switched from
tristate to bool and became default y.

We had a few configs which set them to =m, which is no longer valid.
We don't need to change them to =y because both symbols are default y
and are enabled automatically based on the other symbols in the
affected defconfigs.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2017-08-28 22:10:04 +10:00