We have 16 architectural exception vectors, and depending on kernel
configuration we handle 8 or 12 of these with C code, with the remaining
8 or 4 of these handled as special cases in the entry assembly.
It would be nicer if the entry assembly were uniform for all exceptions,
and we deferred any specific handling of the exceptions to C code. This
way the entry assembly can be more easily templated without ifdeffery or
special cases, and it's easier to modify the handling of these cases in
future (e.g. to dump additional registers other context).
This patch reworks the entry code so that we always have a C handler for
every architectural exception vector, with the entry assembly being
completely uniform. We now have to handle exceptions from EL1t and EL1h,
and also have to handle exceptions from AArch32 even when the kernel is
built without CONFIG_COMPAT. To make this clear and to simplify
templating, we rename the top-level exception handlers with a consistent
naming scheme:
asm: <el+sp>_<regsize>_<type>
c: <el+sp>_<regsize>_<type>_handler
.. where:
<el+sp> is `el1t`, `el1h`, or `el0t`
<regsize> is `64` or `32`
<type> is `sync`, `irq`, `fiq`, or `error`
... e.g.
asm: el1h_64_sync
c: el1h_64_sync_handler
... with lower-level handlers simply using "el1" and "compat" as today.
For unexpected exceptions, this information is passed to
__panic_unhandled(), so it can report the specific vector an unexpected
exception was taken from, e.g.
| Unhandled 64-bit el1t sync exception
For vectors we never expect to enter legitimately, the C code is
generated using a macro to avoid code duplication. The exceptions are
handled via __panic_unhandled(), replacing bad_mode() (which is
removed).
The `kernel_ventry` and `entry_handler` assembly macros are updated to
handle the new naming scheme. In theory it should be possible to
generate the entry functions at the same time as the vectors using a
single table, but this will require reworking the linker script to split
the two into separate sections, so for now we have separate tables.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Joey Gouly <joey.gouly@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-15-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
Our use of bad_mode() has a few rough edges:
* AArch64 doesn't use the term "mode", and refers to "Execution
states", "Exception levels", and "Selected stack pointer".
* We log the exception type (SYNC/IRQ/FIQ/SError), but not the actual
"mode" (though this can be decoded from the SPSR value).
* We use bad_mode() as a second-level handler for unexpected synchronous
exceptions, where the "mode" is legitimate, but the specific exception
is not.
* We dump the ESR value, but call this "code", and so it's not clear to
all readers that this is the ESR.
... and all of this can be somewhat opaque to those who aren't extremely
familiar with the code.
Let's make this a bit clearer by having bad_mode() log "Unhandled
${TYPE} exception" rather than "Bad mode in ${TYPE} handler", using
"ESR" rather than "code", and having the final panic() log "Unhandled
exception" rather than "Bad mode".
In future we'd like to log the specific architectural vector rather than
just the type of exception, so we also split the core of bad_mode() out
into a helper called __panic_unhandled(), which takes the vector as a
string argument.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Joey Gouly <joey.gouly@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-13-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
In subsequent patches we'll rework the way bad_mode() is called by
exception entry code. In preparation for this, let's move bad_mode()
itself into entry-common.c.
Let's also mark it as noinstr (e.g. to prevent it being kprobed), and
let's also make the `handler` array a local variable, as this is only
use by bad_mode(), and will be removed entirely in a subsequent patch.
There should be no functional change as a result of this patch.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Joey Gouly <joey.gouly@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-12-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
In subsequent patches we'll rename the entry handlers based on their
original EL, register width, and exception class. To do so, we need to
make all 3 mandatory arguments to the `kernel_ventry` macro, and
distinguish EL1h from EL1t.
In preparation for this, let's make the current set of arguments
mandatory, and move the `regsize` column before the branch label suffix,
making the vectors easier to read column-wise.
There should be no functional change as a result of this patch.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Joey Gouly <joey.gouly@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-10-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
In entry.S we have two comments which distinguish EL0 and EL1 exception
handlers, but the code isn't actually laid out to match, and there are a
few other inconsistencies that would be good to clear up.
This patch organizes the entry handers consistently:
* The handlers are laid out in order of the vectors, to make them easier
to navigate.
* The inconsistently-applied alignment is removed
* The handlers are consistently marked with SYM_CODE_START_LOCAL()
rather than SYM_CODE_START_LOCAL_NOALIGN(), giving them the same
default alignment as other assembly code snippets.
There should be no functional change as a result of this patch.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Joey Gouly <joey.gouly@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-9-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
For various reasons we'd like to convert the bulk of arm64's exception
triage logic to C. As a step towards that, this patch converts the EL1
and EL0 IRQ+FIQ triage logic to C.
Separate C functions are added for the native and compat cases so that
in subsequent patches we can handle native/compat differences in C.
Since the triage functions can now call arm64_apply_bp_hardening()
directly, the do_el0_irq_bp_hardening() wrapper function is removed.
Since the user_exit_irqoff macro is now unused, it is removed. The
user_enter_irqoff macro is still used by the ret_to_user code, and
cannot be removed at this time.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Joey Gouly <joey.gouly@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-8-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
When handling IRQ/FIQ exceptions the entry assembly may transition from
a task's stack to a CPU's IRQ stack (and IRQ shadow call stack).
In subsequent patches we want to migrate the IRQ/FIQ triage logic to C,
and as we want to perform some actions on the task stack (e.g. EL1
preemption), we need to switch stacks within the C handler. So that we
can do so, this patch adds a helper to call a function on a CPU's IRQ
stack (and shadow stack as appropriate).
Subsequent patches will make use of the new helper function.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-7-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
Currently portions of our preempt logic are written in C while other
parts are written in assembly. Let's clean this up a little bit by
moving the NMI preempt checks to C. For now, the preempt count (and
need_resched) checking is left in assembly, and will be converted
with the body of the IRQ handler in subsequent patches.
Other than the increased lockdep coverage there should be no functional
change as a result of this patch.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Joey Gouly <joey.gouly@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-6-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
For non-fatal exceptions taken from EL0, we expect that at some point
during exception handling it is possible to return to a regular process
context with all exceptions unmasked (e.g. as we do in
do_notify_resume()), and we generally aim to unmask exceptions wherever
possible.
While handling SError and debug exceptions from EL0, we need to leave
some exceptions masked during handling. Handling SError requires us to
mask SError (which also requires masking IRQ+FIQ), and handing debug
exceptions requires us to mask debug (which also requires masking
SError+IRQ+FIQ).
Once do_serror() or do_debug_exception() has returned, we no longer need
to mask exceptions, and can unmask them all, which is what we did prior
to commit:
9034f62515 ("arm64: Do not enable IRQs for ct_user_exit")
... where we had to mask IRQs as for context_tracking_user_exit()
expected IRQs to be masked.
Since then, we realised that our context tracking wasn't entirely
correct, and reworked the entry code to fix this. As of commit:
23529049c6 ("arm64: entry: fix non-NMI user<->kernel transitions")
... we replaced the call to context_tracking_user_exit() with a call to
user_exit_irqoff() as part of enter_from_user_mode(), which occurs
earlier, before we run the body of the handler and unmask exceptions in
DAIF.
When we return to userspace, we go via ret_to_user(), which masks
exceptions in DAIF prior to calling user_enter_irqoff() as part of
exit_to_user_mode().
Thus, there's no longer a reason to leave IRQs or FIQs masked at the end
of the EL0 debug or error handlers, as neither the user exit context
tracking nor the user entry context tracking requires this. Let's bring
these into line with other EL0 exception handlers and ensure that IRQ
and FIQ are unmasked in DAIF at some point during the handler.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Joey Gouly <joey.gouly@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20210607094624.34689-3-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
Kprobes has a counter 'nmissed', that is used to count the number of
times a probe handler was not called. This generally happens when we hit
a kprobe while handling another kprobe.
However, if one of the probe handlers causes a fault, we are currently
incrementing 'nmissed'. The comment in fault handler indicates that this
can be used to account faults taken by the probe handlers. But, this has
never been the intention as is evident from the comment above 'nmissed'
in 'struct kprobe':
/*count the number of times this probe was temporarily disarmed */
unsigned long nmissed;
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Masami Hiramatsu <mhiramat@kernel.org>
Link: https://lkml.kernel.org/r/20210601120150.672652-1-naveen.n.rao@linux.vnet.ibm.com
The only user we have of Normal Write-Through memory is in the ACPI code
when mapping memory regions advertised as EFI_MEMORY_WT. Since most (all?)
CPUs treat write-through as non-cacheable under the hood, don't bother
with the extra memory type here and just treat EFI_MEMORY_WT the same way
as EFI_MEMORY_WC by mapping it to the Normal-NC memory type instead and
emitting a warning if we have failed to find an alternative EFI memory
type.
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20210527110319.22157-3-will@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
The reason for kprobe::fault_handler(), as given by their comment:
* We come here because instructions in the pre/post
* handler caused the page_fault, this could happen
* if handler tries to access user space by
* copy_from_user(), get_user() etc. Let the
* user-specified handler try to fix it first.
Is just plain bad. Those other handlers are ran from non-preemptible
context and had better use _nofault() functions. Also, there is no
upstream usage of this.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Acked-by: Masami Hiramatsu <mhiramat@kernel.org>
Link: https://lore.kernel.org/r/20210525073213.561116662@infradead.org
Aarch64 instruction set encoding and decoding logic can prove useful
for some features/tools both part of the kernel and outside the kernel.
Isolate the function dealing only with encoding/decoding instructions,
with minimal dependency on kernel utilities in order to be able to reuse
that code.
Code was only moved, no code should have been added, removed nor
modifier.
Signed-off-by: Julien Thierry <jthierry@redhat.com>
Link: https://lore.kernel.org/r/20210303170536.1838032-5-jthierry@redhat.com
Signed-off-by: Will Deacon <will@kernel.org>
Now that we have a consistent place to initialize CPU context registers
early in the boot path, let's also initialize the per-cpu offset here.
This makes the primary and secondary boot paths more consistent, and
allows for the use of per-cpu operations earlier, which will be
necessary for instrumentation with KCSAN.
Note that smp_prepare_boot_cpu() still needs to re-initialize CPU0's
offset as immediately prior to this the per-cpu areas may be
reallocated, and hence the boot-time offset may be stale. A comment is
added to make this clear.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Suzuki Poulose <suzuki.poulose@arm.com>
Cc: Will Deacon <will@kernel.org>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20210520115031.18509-7-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
Once we enable the MMU, we have to initialize:
* SP_EL0 to point at the active task
* SP to point at the active task's stack
* SCS_SP to point at the active task's shadow stack
For all tasks (including init_task), this information can be derived
from the task's task_struct.
Let's unify __primary_switched and __secondary_switched to consistently
acquire this information from the relevant task_struct. At the same
time, let's fold this together with initializing a task's final frame.
There should be no functional change as a result of this patch.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Suzuki Poulose <suzuki.poulose@arm.com>
Cc: Will Deacon <will@kernel.org>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20210520115031.18509-6-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
SMCCC v1.2 allows x8-x17 to be used as parameter registers and x4—x17
to be used as result registers in SMC64/HVC64. Arm Firmware Framework
for Armv8-A specification makes use of x0-x7 as parameter and result
registers. There are other users like Hyper-V who intend to use beyond
x0-x7 as well.
Current SMCCC interface in the kernel just use x0-x7 as parameter and
x0-x3 as result registers as required by SMCCCv1.0. Let us add new
interface to support this extended set of input/output registers namely
x0-x17 as both parameter and result registers.
Acked-by: Mark Rutland <mark.rutland@arm.com>
Tested-by: Michael Kelley <mikelley@microsoft.com>
Reviewed-by: Michael Kelley <mikelley@microsoft.com>
Cc: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20210518163618.43950-1-sudeep.holla@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
Although naming across the codebase isn't that consistent, it
tends to follow certain patterns. Moreover, the term "flush"
isn't defined in the Arm Architecture reference manual, and might
be interpreted to mean clean, invalidate, or both for a cache.
Rename arm64-internal functions to make the naming internally
consistent, as well as making it consistent with the Arm ARM, by
specifying whether it applies to the instruction, data, or both
caches, whether the operation is a clean, invalidate, or both.
Also specify which point the operation applies to, i.e., to the
point of unification (PoU), coherency (PoC), or persistence
(PoP).
This commit applies the following sed transformation to all files
under arch/arm64:
"s/\b__flush_cache_range\b/caches_clean_inval_pou_macro/g;"\
"s/\b__flush_icache_range\b/caches_clean_inval_pou/g;"\
"s/\binvalidate_icache_range\b/icache_inval_pou/g;"\
"s/\b__flush_dcache_area\b/dcache_clean_inval_poc/g;"\
"s/\b__inval_dcache_area\b/dcache_inval_poc/g;"\
"s/__clean_dcache_area_poc\b/dcache_clean_poc/g;"\
"s/\b__clean_dcache_area_pop\b/dcache_clean_pop/g;"\
"s/\b__clean_dcache_area_pou\b/dcache_clean_pou/g;"\
"s/\b__flush_cache_user_range\b/caches_clean_inval_user_pou/g;"\
"s/\b__flush_icache_all\b/icache_inval_all_pou/g;"
Note that __clean_dcache_area_poc is deliberately missing a word
boundary check at the beginning in order to match the efistub
symbols in image-vars.h.
Also note that, despite its name, __flush_icache_range operates
on both instruction and data caches. The name change here
reflects that.
No functional change intended.
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Fuad Tabba <tabba@google.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20210524083001.2586635-19-tabba@google.com
Signed-off-by: Will Deacon <will@kernel.org>
To be consistent with other functions with similar names and
functionality in cacheflush.h, cache.S, and cachetlb.rst, change
to specify the range in terms of start and end, as opposed to
start and size.
Because the code is shared with __dma_clean_area, it changes the
parameters for that as well. However, __dma_clean_area is local to
cache.S, so no other users are affected.
No functional change intended.
Reported-by: Will Deacon <will@kernel.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Fuad Tabba <tabba@google.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20210524083001.2586635-14-tabba@google.com
Signed-off-by: Will Deacon <will@kernel.org>
To be consistent with other functions with similar names and
functionality in cacheflush.h, cache.S, and cachetlb.rst, change
to specify the range in terms of start and end, as opposed to
start and size.
Because the code is shared with __dma_inv_area, it changes the
parameters for that as well. However, __dma_inv_area is local to
cache.S, so no other users are affected.
No functional change intended.
Reported-by: Will Deacon <will@kernel.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Fuad Tabba <tabba@google.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20210524083001.2586635-11-tabba@google.com
Signed-off-by: Will Deacon <will@kernel.org>
Many comments refer to the function flush_icache_range, where the
intent is in fact __flush_icache_range. Fix these comments to
refer to the intended function.
That's probably due to commit 3b8c9f1cdf ("arm64: IPI each
CPU after invalidating the I-cache for kernel mappings"), which
renamed flush_icache_range() to __flush_icache_range() and added
a wrapper.
No functional change intended.
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Fuad Tabba <tabba@google.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20210524083001.2586635-10-tabba@google.com
Signed-off-by: Will Deacon <will@kernel.org>
Reliable stacktracing requires that we identify when a stacktrace is
terminated early. We can do this by ensuring all tasks have a final
frame record at a known location on their task stack, and checking
that this is the final frame record in the chain.
We'd like to use task_pt_regs(task)->stackframe as the final frame
record, as this is already setup upon exception entry from EL0. For
kernel tasks we need to consistently reserve the pt_regs and point x29
at this, which we can do with small changes to __primary_switched,
__secondary_switched, and copy_process().
Since the final frame record must be at a specific location, we must
create the final frame record in __primary_switched and
__secondary_switched rather than leaving this to start_kernel and
secondary_start_kernel. Thus, __primary_switched and
__secondary_switched will now show up in stacktraces for the idle tasks.
Since the final frame record is now identified by its location rather
than by its contents, we identify it at the start of unwind_frame(),
before we read any values from it.
External debuggers may terminate the stack trace when FP == 0. In the
pt_regs->stackframe, the PC is 0 as well. So, stack traces taken in the
debugger may print an extra record 0x0 at the end. While this is not
pretty, this does not do any harm. This is a small price to pay for
having reliable stack trace termination in the kernel. That said, gdb
does not show the extra record probably because it uses DWARF and not
frame pointers for stack traces.
Signed-off-by: Madhavan T. Venkataraman <madvenka@linux.microsoft.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
[Mark: rebase, use ASM_BUG(), update comments, update commit message]
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Link: https://lore.kernel.org/r/20210510110026.18061-1-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
As pointed out by commit
de9b8f5dcb ("sched: Fix crash trying to dequeue/enqueue the idle thread")
init_idle() can and will be invoked more than once on the same idle
task. At boot time, it is invoked for the boot CPU thread by
sched_init(). Then smp_init() creates the threads for all the secondary
CPUs and invokes init_idle() on them.
As the hotplug machinery brings the secondaries to life, it will issue
calls to idle_thread_get(), which itself invokes init_idle() yet again.
In this case it's invoked twice more per secondary: at _cpu_up(), and at
bringup_cpu().
Given smp_init() already initializes the idle tasks for all *possible*
CPUs, no further initialization should be required. Now, removing
init_idle() from idle_thread_get() exposes some interesting expectations
with regards to the idle task's preempt_count: the secondary startup always
issues a preempt_disable(), requiring some reset of the preempt count to 0
between hot-unplug and hotplug, which is currently served by
idle_thread_get() -> idle_init().
Given the idle task is supposed to have preemption disabled once and never
see it re-enabled, it seems that what we actually want is to initialize its
preempt_count to PREEMPT_DISABLED and leave it there. Do that, and remove
init_idle() from idle_thread_get().
Secondary startups were patched via coccinelle:
@begone@
@@
-preempt_disable();
...
cpu_startup_entry(CPUHP_AP_ONLINE_IDLE);
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20210512094636.2958515-1-valentin.schneider@arm.com
Pull more arm64 updates from Catalin Marinas:
"A mix of fixes and clean-ups that turned up too late for the first
pull request:
- Restore terminal stack frame records. Their previous removal caused
traces which cross secondary_start_kernel to terminate one entry
too late, with a spurious "0" entry.
- Fix boot warning with pseudo-NMI due to the way we manipulate the
PMR register.
- ACPI fixes: avoid corruption of interrupt mappings on watchdog
probe failure (GTDT), prevent unregistering of GIC SGIs.
- Force SPARSEMEM_VMEMMAP as the only memory model, it saves with
having to test all the other combinations.
- Documentation fixes and updates: tagged address ABI exceptions on
brk/mmap/mremap(), event stream frequency, update booting
requirements on the configuration of traps"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: kernel: Update the stale comment
arm64: Fix the documented event stream frequency
arm64: entry: always set GIC_PRIO_PSR_I_SET during entry
arm64: Explicitly document boot requirements for SVE
arm64: Explicitly require that FPSIMD instructions do not trap
arm64: Relax booting requirements for configuration of traps
arm64: cpufeatures: use min and max
arm64: stacktrace: restore terminal records
arm64/vdso: Discard .note.gnu.property sections in vDSO
arm64: doc: Add brk/mmap/mremap() to the Tagged Address ABI Exceptions
psci: Remove unneeded semicolon
ACPI: irq: Prevent unregistering of GIC SGIs
ACPI: GTDT: Don't corrupt interrupt mappings on watchdow probe failure
arm64: Show three registers per line
arm64: remove HAVE_DEBUG_BUGVERBOSE
arm64: alternative: simplify passing alt_region
arm64: Force SPARSEMEM_VMEMMAP as the only memory management model
arm64: vdso32: drop -no-integrated-as flag
Zenghui reports that booting a kernel with "irqchip.gicv3_pseudo_nmi=1"
on the command line hits a warning during kernel entry, due to the way
we manipulate the PMR.
Early in the entry sequence, we call lockdep_hardirqs_off() to inform
lockdep that interrupts have been masked (as the HW sets DAIF wqhen
entering an exception). Architecturally PMR_EL1 is not affected by
exception entry, and we don't set GIC_PRIO_PSR_I_SET in the PMR early in
the exception entry sequence, so early in exception entry the PMR can
indicate that interrupts are unmasked even though they are masked by
DAIF.
If DEBUG_LOCKDEP is selected, lockdep_hardirqs_off() will check that
interrupts are masked, before we set GIC_PRIO_PSR_I_SET in any of the
exception entry paths, and hence lockdep_hardirqs_off() will WARN() that
something is amiss.
We can avoid this by consistently setting GIC_PRIO_PSR_I_SET during
exception entry so that kernel code sees a consistent environment. We
must also update local_daif_inherit() to undo this, as currently only
touches DAIF. For other paths, local_daif_restore() will update both
DAIF and the PMR. With this done, we can remove the existing special
cases which set this later in the entry code.
We always use (GIC_PRIO_IRQON | GIC_PRIO_PSR_I_SET) for consistency with
local_daif_save(), as this will warn if it ever encounters
(GIC_PRIO_IRQOFF | GIC_PRIO_PSR_I_SET), and never sets this itself. This
matches the gic_prio_kentry_setup that we have to retain for
ret_to_user.
The original splat from Zenghui's report was:
| DEBUG_LOCKS_WARN_ON(!irqs_disabled())
| WARNING: CPU: 3 PID: 125 at kernel/locking/lockdep.c:4258 lockdep_hardirqs_off+0xd4/0xe8
| Modules linked in:
| CPU: 3 PID: 125 Comm: modprobe Tainted: G W 5.12.0-rc8+ #463
| Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015
| pstate: 604003c5 (nZCv DAIF +PAN -UAO -TCO BTYPE=--)
| pc : lockdep_hardirqs_off+0xd4/0xe8
| lr : lockdep_hardirqs_off+0xd4/0xe8
| sp : ffff80002a39bad0
| pmr_save: 000000e0
| x29: ffff80002a39bad0 x28: ffff0000de214bc0
| x27: ffff0000de1c0400 x26: 000000000049b328
| x25: 0000000000406f30 x24: ffff0000de1c00a0
| x23: 0000000020400005 x22: ffff8000105f747c
| x21: 0000000096000044 x20: 0000000000498ef9
| x19: ffff80002a39bc88 x18: ffffffffffffffff
| x17: 0000000000000000 x16: ffff800011c61eb0
| x15: ffff800011700a88 x14: 0720072007200720
| x13: 0720072007200720 x12: 0720072007200720
| x11: 0720072007200720 x10: 0720072007200720
| x9 : ffff80002a39bad0 x8 : ffff80002a39bad0
| x7 : ffff8000119f0800 x6 : c0000000ffff7fff
| x5 : ffff8000119f07a8 x4 : 0000000000000001
| x3 : 9bcdab23f2432800 x2 : ffff800011730538
| x1 : 9bcdab23f2432800 x0 : 0000000000000000
| Call trace:
| lockdep_hardirqs_off+0xd4/0xe8
| enter_from_kernel_mode.isra.5+0x7c/0xa8
| el1_abort+0x24/0x100
| el1_sync_handler+0x80/0xd0
| el1_sync+0x6c/0x100
| __arch_clear_user+0xc/0x90
| load_elf_binary+0x9fc/0x1450
| bprm_execve+0x404/0x880
| kernel_execve+0x180/0x188
| call_usermodehelper_exec_async+0xdc/0x158
| ret_from_fork+0x10/0x18
Fixes: 23529049c6 ("arm64: entry: fix non-NMI user<->kernel transitions")
Fixes: 7cd1ea1010 ("arm64: entry: fix non-NMI kernel<->kernel transitions")
Fixes: f0cd5ac1e4 ("arm64: entry: fix NMI {user, kernel}->kernel transitions")
Fixes: 2a9b3e6ac6 ("arm64: entry: fix EL1 debug transitions")
Link: https://lore.kernel.org/r/f4012761-026f-4e51-3a0c-7524e434e8b3@huawei.com
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reported-by: Zenghui Yu <yuzenghui@huawei.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Will Deacon <will@kernel.org>
Acked-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210428111555.50880-1-mark.rutland@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Pull kvm updates from Paolo Bonzini:
"This is a large update by KVM standards, including AMD PSP (Platform
Security Processor, aka "AMD Secure Technology") and ARM CoreSight
(debug and trace) changes.
ARM:
- CoreSight: Add support for ETE and TRBE
- Stage-2 isolation for the host kernel when running in protected
mode
- Guest SVE support when running in nVHE mode
- Force W^X hypervisor mappings in nVHE mode
- ITS save/restore for guests using direct injection with GICv4.1
- nVHE panics now produce readable backtraces
- Guest support for PTP using the ptp_kvm driver
- Performance improvements in the S2 fault handler
x86:
- AMD PSP driver changes
- Optimizations and cleanup of nested SVM code
- AMD: Support for virtual SPEC_CTRL
- Optimizations of the new MMU code: fast invalidation, zap under
read lock, enable/disably dirty page logging under read lock
- /dev/kvm API for AMD SEV live migration (guest API coming soon)
- support SEV virtual machines sharing the same encryption context
- support SGX in virtual machines
- add a few more statistics
- improved directed yield heuristics
- Lots and lots of cleanups
Generic:
- Rework of MMU notifier interface, simplifying and optimizing the
architecture-specific code
- a handful of "Get rid of oprofile leftovers" patches
- Some selftests improvements"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (379 commits)
KVM: selftests: Speed up set_memory_region_test
selftests: kvm: Fix the check of return value
KVM: x86: Take advantage of kvm_arch_dy_has_pending_interrupt()
KVM: SVM: Skip SEV cache flush if no ASIDs have been used
KVM: SVM: Remove an unnecessary prototype declaration of sev_flush_asids()
KVM: SVM: Drop redundant svm_sev_enabled() helper
KVM: SVM: Move SEV VMCB tracking allocation to sev.c
KVM: SVM: Explicitly check max SEV ASID during sev_hardware_setup()
KVM: SVM: Unconditionally invoke sev_hardware_teardown()
KVM: SVM: Enable SEV/SEV-ES functionality by default (when supported)
KVM: SVM: Condition sev_enabled and sev_es_enabled on CONFIG_KVM_AMD_SEV=y
KVM: SVM: Append "_enabled" to module-scoped SEV/SEV-ES control variables
KVM: SEV: Mask CPUID[0x8000001F].eax according to supported features
KVM: SVM: Move SEV module params/variables to sev.c
KVM: SVM: Disable SEV/SEV-ES if NPT is disabled
KVM: SVM: Free sev_asid_bitmap during init if SEV setup fails
KVM: SVM: Zero out the VMCB array used to track SEV ASID association
x86/sev: Drop redundant and potentially misleading 'sev_enabled'
KVM: x86: Move reverse CPUID helpers to separate header file
KVM: x86: Rename GPR accessors to make mode-aware variants the defaults
...
