Patch series "improvements to the nmi_backtrace code" v9.
This patch series modifies the trigger_xxx_backtrace() NMI-based remote
backtracing code to make it more flexible, and makes a few small
improvements along the way.
The motivation comes from the task isolation code, where there are
scenarios where we want to be able to diagnose a case where some cpu is
about to interrupt a task-isolated cpu. It can be helpful to see both
where the interrupting cpu is, and also an approximation of where the
cpu that is being interrupted is. The nmi_backtrace framework allows us
to discover the stack of the interrupted cpu.
I've tested that the change works as desired on tile, and build-tested
x86, arm, mips, and sparc64. For x86 I confirmed that the generic
cpuidle stuff as well as the architecture-specific routines are in the
new cpuidle section. For arm, mips, and sparc I just build-tested it
and made sure the generic cpuidle routines were in the new cpuidle
section, but I didn't attempt to figure out which the platform-specific
idle routines might be. That might be more usefully done by someone
with platform experience in follow-up patches.
This patch (of 4):
Currently you can only request a backtrace of either all cpus, or all
cpus but yourself. It can also be helpful to request a remote backtrace
of a single cpu, and since we want that, the logical extension is to
support a cpumask as the underlying primitive.
This change modifies the existing lib/nmi_backtrace.c code to take a
cpumask as its basic primitive, and modifies the linux/nmi.h code to use
the new "cpumask" method instead.
The existing clients of nmi_backtrace (arm and x86) are converted to
using the new cpumask approach in this change.
The other users of the backtracing API (sparc64 and mips) are converted
to use the cpumask approach rather than the all/allbutself approach.
The mips code ignored the "include_self" boolean but with this change it
will now also dump a local backtrace if requested.
Link: http://lkml.kernel.org/r/1472487169-14923-2-git-send-email-cmetcalf@mellanox.com
Signed-off-by: Chris Metcalf <cmetcalf@mellanox.com>
Tested-by: Daniel Thompson <daniel.thompson@linaro.org> [arm]
Reviewed-by: Aaron Tomlin <atomlin@redhat.com>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: David Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We get 1 warning when building kernel with W=1:
drivers/char/mem.c:220:12: warning: no previous prototype for 'phys_mem_access_prot_allowed' [-Wmissing-prototypes]
int __weak phys_mem_access_prot_allowed(struct file *file,
In fact, its declaration is spreading to several header files in
different architecture, but need to be declare in common header file.
So this patch moves phys_mem_access_prot_allowed() to pgtable.h.
Link: http://lkml.kernel.org/r/1473751597-12139-1-git-send-email-baoyou.xie@linaro.org
Signed-off-by: Baoyou Xie <baoyou.xie@linaro.org>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull KVM updates from Radim Krčmář:
"All architectures:
- move `make kvmconfig` stubs from x86
- use 64 bits for debugfs stats
ARM:
- Important fixes for not using an in-kernel irqchip
- handle SError exceptions and present them to guests if appropriate
- proxying of GICV access at EL2 if guest mappings are unsafe
- GICv3 on AArch32 on ARMv8
- preparations for GICv3 save/restore, including ABI docs
- cleanups and a bit of optimizations
MIPS:
- A couple of fixes in preparation for supporting MIPS EVA host
kernels
- MIPS SMP host & TLB invalidation fixes
PPC:
- Fix the bug which caused guests to falsely report lockups
- other minor fixes
- a small optimization
s390:
- Lazy enablement of runtime instrumentation
- up to 255 CPUs for nested guests
- rework of machine check deliver
- cleanups and fixes
x86:
- IOMMU part of AMD's AVIC for vmexit-less interrupt delivery
- Hyper-V TSC page
- per-vcpu tsc_offset in debugfs
- accelerated INS/OUTS in nVMX
- cleanups and fixes"
* tag 'kvm-4.9-1' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (140 commits)
KVM: MIPS: Drop dubious EntryHi optimisation
KVM: MIPS: Invalidate TLB by regenerating ASIDs
KVM: MIPS: Split kernel/user ASID regeneration
KVM: MIPS: Drop other CPU ASIDs on guest MMU changes
KVM: arm/arm64: vgic: Don't flush/sync without a working vgic
KVM: arm64: Require in-kernel irqchip for PMU support
KVM: PPC: Book3s PR: Allow access to unprivileged MMCR2 register
KVM: PPC: Book3S PR: Support 64kB page size on POWER8E and POWER8NVL
KVM: PPC: Book3S: Remove duplicate setting of the B field in tlbie
KVM: PPC: BookE: Fix a sanity check
KVM: PPC: Book3S HV: Take out virtual core piggybacking code
KVM: PPC: Book3S: Treat VTB as a per-subcore register, not per-thread
ARM: gic-v3: Work around definition of gic_write_bpr1
KVM: nVMX: Fix the NMI IDT-vectoring handling
KVM: VMX: Enable MSR-BASED TPR shadow even if APICv is inactive
KVM: nVMX: Fix reload apic access page warning
kvmconfig: add virtio-gpu to config fragment
config: move x86 kvm_guest.config to a common location
arm64: KVM: Remove duplicating init code for setting VMID
ARM: KVM: Support vgic-v3
...
Convert the MIPS SEAD-3 board support to be a generic board, supported
by generic kernels.
Because the SEAD-3 boot protocol was defined long ago and we don't want
to force a switch to the UHI protocol, SEAD-3 is added as a legacy board
which is detected by reading the REVISION register. This may technically
not be a valid memory read & future work will include attempting to
handle that gracefully. In practice since SEAD-3 is the only legacy
board supported by the generic kernel so far the read will only happen
on SEAD-3 boards, and even once Malta is converted the same REVISION
register exists there too. Other boards such as Boston, Ci20 & Ci40 will
use the UHI boot protocol & thus not run any of the legacy board detect
functions.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14354/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Introduce a "generic" platform, which aims to be board-agnostic by
making use of device trees passed by the boot protocol defined in the
MIPS UHI (Universal Hosting Interface) specification. Provision is made
for supporting boards which use a legacy boot protocol that can't be
changed, but adding support for such boards or any others is left to
followon patches.
Right now the built kernels expect to be loaded to 0x80100000, ie. in
kseg0. This is fine for the vast majority of MIPS platforms, but
nevertheless it would be good to remove this limitation in the future by
mapping the kernel via the TLB such that it can be loaded anywhere & map
itself appropriately.
Configuration is handled by dynamically generating configs using
scripts/kconfig/merge_config.sh, somewhat similar to the way powerpc
makes use of it. This allows for variations upon the configuration, eg.
differing architecture revisions or subsets of driver support for
differing boards, to be handled without having a large number of
defconfig files.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14353/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Add support for generating kernel images in the Flattened Image Tree
(.itb) format as supported by U-Boot. This format is essentially a
Flattened Device Tree binary containing images (kernels, DTBs, ramdisks)
and configurations which link those images together. The big advantages
of FIT images over the uImage format are:
- We can include FDTs in the kernel image in a way that the bootloader
can extract it & manipulate it before providing it to the kernel.
Thus we can ship FDTs as part of the kernel giving us the advantages
of being able to develop & maintain the DT within the kernel tree,
but also have the benefits of the bootloader being able to
manipulate the FDT. Example uses for this would be to inject the
kernel command line into the chosen node, or to fill in the correct
memory size.
- We can include multiple configurations in a single kernel image.
This means that a single FIT image can, given appropriate
bootloaders, be booted on different boards with the bootloader
selecting an appropriate configuration & providing the correct FDT
to the kernel.
- We can support a multitude of hashes over the data.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14352/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
On MIPS64 we define the default CAC_BASE as one of the xkphys regions of
the virtual address space. Since the CCA is encoded in bits 61:59 of
xkphys addresses, fixing CAC_BASE to any particular one prevents us from
dynamically changing the CCA as we do for MIPS32 where CAC_BASE is
placed within kseg0. In order to make the kernel more generic, drop the
current kludge that gives CAC_BASE CCA=3 if CONFIG_DMA_NONCOHERENT is
selected (disregarding CONFIG_DMA_MAYBE_COHERENT) & CCA=5 (which is not
standardised by the architecture) otherwise. Instead read Config.K0 and
generate the appropriate offset into xkphys, presuming that either the
bootloader or early kernel code will have configured Config.K0
appropriately. This seems like the best option for a generic
implementation.
The ip27 spaces.h is adjusted to set its former value of CAC_BASE, since
it's the only user of CAC_BASE from assembly (in its smp_slave_setup
macro). This allows the generic case to focus solely on C code without
breaking ip27.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14351/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
If a bus error occurs on a system with a MIPS Coherence Manager (CM)
then the CM may hold useful diagnostic information. Printing this out
has so far been left up to boards, with the requirement that they
register a board_be_handler function & call mips_cm_error_decode() from
there.
In order to avoid boards other than Malta needing to duplicate this
code, call mips_cm_error_decode() automatically if the board registers
no board_be_handler, and remove the Malta implementation of that.
This patch results in no functional change, but removes a further piece
of platform-specific code.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14350/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
On some MIPS systems, a subset of devices may have DMA coherent with CPU
caches. For example in systems including a MIPS I/O Coherence Unit
(IOCU), some devices may be connected to that IOCU whilst others are
not.
Prior to this patch, we have a plat_device_is_coherent() function but no
implementation which does anything besides return a global true or
false, optionally chosen at runtime. For devices such as those described
above this is insufficient.
Fix this by tracking DMA coherence on a per-device basis with a
dma_coherent field in struct dev_archdata. Setting this from
arch_setup_dma_ops() takes care of devices which set the dma-coherent
property via device tree, and any PCI devices beneath a bridge described
in DT, automatically.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14349/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
There are no cases where plat_device_is_coherent() will return zero
whilst hw_coherentio is non-zero, and acting any differently in such a
case doesn't make much sense - if a device is non-coherent with the CPU
caches then access to memory "coherent" with DMA must be uncached. Clean
up the nonsensical case.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14348/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The coherentio variable has previously been used as a boolean value,
indicating whether the user specified that coherent I/O should be
enabled or disabled. It failed to take into account the case where the
user does not specify any preference, in which case it makes sense that
we should default to coherent I/O if the hardware supports it
(hw_coherentio is non-zero).
Introduce an enum to clarify the 3 different values of coherentio & use
it throughout the code, modifying plat_device_is_coherent() &
r4k_cache_init() to take into account the default case.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: Paul Burton <paul.burton@imgtec.com>
Patchwork: https://patchwork.linux-mips.org/patch/14347/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Introduce support for PCI drivers using only functionality provided
generically by the PCI subsystem, by adding the minimum arch-provided
functions required.
The driver this has been developed for & tested with the xilinx-pcie on
a MIPS Boston development board.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14346/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Introduce 2 Kconfig symbols, CONFIG_PCI_DRIVERS_GENERIC &
CONFIG_PCI_DRIVERS_LEGACY, which indicate whether the system should be
built to for PCI drivers using the MIPS-specific struct pci_controller
API (hereafter "legacy" drivers) or more generic drivers using only
functionality provided by the PCI core (hereafter "generic" drivers).
The Kconfig entries are created such that platforms have to select
CONFIG_PCI_DRIVERS_GENERIC if they wish to use it - that is, the default
is CONFIG_PCI_DRIVERS_LEGACY so that existing platforms need no
modification.
The functions declared in pci.h are rearranged with those provided only
by pci-legacy.c being guarded by an #ifdef CONFIG_PCI_DRIVERS_LEGACY to
ensure they are only used in configurations where they are implemented.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14345/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Introduce support for CONFIG_PCI_DOMAINS_GENERIC, allowing for platforms
to make use of generic PCI domains instead of the MIPS-specific
implementation. The set_pci_need_domain_info function is introduced to
abstract away the removed need_domain_info field in struct
pci_controller, and pcibios_scanbus is adjusted to use the pci_domain_nr
accessor instead of directly accessing the index field.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14341/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Current instruction decoder for uprobe/kprobe handler only handles
branches with delay slots. For compact branches the behaviour is rather
unpredictable - and depending on the encoding of a compact branch
instruction may result in one (or more) of:
- executing an instruction that follows a branch which wasn't in a delay
slot and shouldn't have been executed
- incorrectly emulating a branch leading to a jump to a wrong location
- unexpected branching out of the single-stepped code and never reaching
the breakpoint that should terminate the probe handler
Results of these actions are generally unpredictable, but can end up
with a probed application or kernel crash, so disable placing probes on
compact branches until they are handled properly.
Signed-off-by: Marcin Nowakowski <marcin.nowakowski@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14336/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Currently both kprobes and uprobes code have definitions of the
insn_has_delay_slot method. Move it to a separate header as an inline
method that each probe-specific method can later use.
No functional change intended, although the methods slightly varied in
the constraints they set for the methods - the uprobes one was chosen as
it is slightly more specific when filtering opcode fields.
Signed-off-by: Marcin Nowakowski <marcin.nowakowski@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14335/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Malta boards can have more than 256MB DDR available, but we have
previously only made use of up to 256MB (ie. the DDR accessible via
kseg0) by default, without the user manually specifying mem= kernel
parameters. This patch causes all available DDR, as reported by the
bootloader via the ememsize or memsize environment variables or
optionally on the command line, to be used when possible without the
user needing to manually provide the memory ranges.
Malta now has 2 subtly different memory maps which have to be taken into
account when setting this up. The original memory map (referred to by
the code as v1) has up to 2GB of DDR aliased in both the upper & lower
halves of the 32 bit physical address space, with a 256MB I/O region
obscuring 0x10000000-0x1fffffff only in the lower alias. The revised v2
memory map is flat with up to 4GB DDR starting from 0x0, and the I/O
region obscures 256MB of DDR which becomes inacessible. The memory map
in use is indicated by a register provided by the rocit2 system
controller, which is checked in order to set up the kernels memory
ranges accordingly.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: Masahiro Yamada <yamada.masahiro@socionext.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14273/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The default i8259 polling function (i8259_irq) is nicely generic but is
fairly costly. Platforms often provide an alternative means of polling
for an i8259 interrupt, and when using the i8259 without device tree
have typically just chained its parent interrupt to their own handler
function. In order to allow for platform-specific polling functions to
be used in cases where the driver is probed via device tree, provide an
i8259_set_poll function that accepts a pointer to an alternative poll
function that will override the default.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14270/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The 2 line * 16 character LCD display on the SEAD3 board has no real use
as a framebuffer device. It's far too small to produce any meaningful
output if used as the kernel console, SEAD3 is a development board that
will essentially always have a far more useful UART connection & the
code in sead3-display.c will overwrite whatever's on the display every
second anyway. Remove this unused code.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14059/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
more victims of indirect include chains - au1200fb
lasat/picvue_proc and watchdog/ath79_wdt
... as well as tb0219, spotted by Sudip Mukherjee
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
