| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
There is a certain amount of smp_prepare_cpus() which doesn't belong
in the platform support code - that is, code which is invariant to the
SMP implementation. Move this code into arch/arm/kernel/smp.c, and
add a platform_ prefix to the original function.
Signed-off-by: Russell King <[email protected]>
|
|
Wait for CPUs to indicate that they've stopped, after sending the
stop IPI, rather than blindly continuing on and hoping that they've
stopped in time. Print a warning if we fail to stop the other CPUs.
Signed-off-by: Russell King <[email protected]>
|
|
Use r0,r3-r6 rather than r0,r3,r4,r6,r7, which makes it easier to
understand which registers can be modified. Also document which
registers hold values which must be preserved.
Signed-off-by: Russell King <[email protected]>
|
|
The IPI and local timer interrupts weren't being properly accounted
for in /proc/stat. Collect them from the irq_stat structure, and
return their sum.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
This separates out the individual IPI interrupt counts from the
total IPI count, which allows better visibility of what IPIs are
being used for.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
iwmmxt is used in XScale, XScale3, Mohawk and PJ4 core. But the instructions
of accessing CP0 and CP1 is changed in PJ4. Append more files to support
iwmmxt in PJ4 core.
Signed-off-by: Zhou Zhu <[email protected]>
Signed-off-by: Haojian Zhuang <[email protected]>
Acked-by: Nicolas Pitre <[email protected]>
Signed-off-by: Eric Miao <[email protected]>
|
|
As per x86, align the initial column according to how many IRQs we
have. Also, provide an english explaination for the 'LOC:' and
'IPI:' lines.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
Move the ipi_count into irq_stat, which allows the ipi_data structure
to be entirely removed.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
Provide __inc_irq_stat() and __get_irq_stat() to increment and
read the irq stat counters.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
send_ipi_message() does nothing except call smp_cross_call(). As
this is a static function, nothing external to this file calls it,
so we can easily clean up this now unnecessary indirection.
Signed-off-by: Russell King <[email protected]>
|
|
devel-stable
|
|
We should not be incrementing mm_users when we startup a secondary
CPU - doing so results in mm_users incrementing by one each time we
hotplug a CPU, which will eventually wrap, and will cause problems.
Other architectures such as x86 do not increment mm_users, but only
mm_count, so we follow that pattern.
Signed-off-by: Russell King <[email protected]>
|
|
Extend the perf_pmu_register() interface to allow for named and
dynamic pmu types.
Because we need to support the existing static types we cannot use
dynamic types for everything, hence provide a type argument.
If we want to enumerate the PMUs they need a name, provide one.
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
|
|
The debug registers can only be manipulated from software if monitor
debug mode is enabled. On some cores, this can never be enabled (i.e.
the corresponding bit in the DSCR is RAZ/WI).
This patch ensures we can handle this hardware configuration and fail
gracefully, rather than blow up the kernel during boot.
Reported-by: Cyril Chemparathy <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
|
|
Avoid adding nasty genirq-specific code to local timers to enable PPI
interrupts. Instead, provide a gic function to do this.
Signed-off-by: Russell King <[email protected]>
|
|
Merge reason: Pick up the latest -rc.
Signed-off-by: Ingo Molnar <[email protected]>
|
|
sparse doesn't like per-cpu accesses such as:
static DEFINE_PER_CPU(struct perf_event *, foo[MAXLEN]);
struct perf_event **bar = __get_cpu_var(foo);
and shouts quite loudly about it:
| warning: incorrect type in assignment (different modifiers)
| expected struct perf_event **slots
| got struct perf_event *[noderef] *<noident>
This patch adds casts to these sorts of assignments in hw_breakpoint.c
in order to silence the warnings.
Reported-by: Russell King <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
|
|
This patch fixes a trivial style issue in ptrace.c.
Signed-off-by: Will Deacon <[email protected]>
|
|
Single-stepping a breakpoint requires us to disable it temporarily so that
we don't get stuck in a recursive debug trap. With per-cpu breakpoints this
presents a problem where an interrupt can be taken before the single-step has
completed and a new task is eventually scheduled. This new task will not
hit the breakpoint because it will have been disabled during the previous
handling code.
This patch disallows per-cpu breakpoints on ARM when an overflow handler
is not present. A similar effect can be created by placing breakpoints on
a shell and then running applications there.
Signed-off-by: Will Deacon <[email protected]>
|
|
The single-stepping code is currently different depending on whether
we are stepping over a breakpoint or a watchpoint. There is no good
reason for this, so let's sort it out.
This patch adds functions for enabling/disabling single-step for
a particular hw_breakpoint and integrates this with the exception
handling code.
Signed-off-by: Will Deacon <[email protected]>
|
|
The watchpoint single-stepping code calls register_user_hw_breakpoint to
register a mismatch breakpoint for stepping over the watchpoint. This is
performed with preemption disabled, which is unsafe as we may end up scheduling
whilst in_atomic(). Furthermore, using the perf API is rather overkill since
we are already in the hw-breakpoint backend and only require access to reserved
breakpoints anyway.
This patch reworks the watchpoint stepping code so that we don't require
another perf_event for the mismatch breakpoint. Instead, we hold a separate
arch_hw_breakpoint_ctrl struct inside the watchpoint which is used exclusively
for stepping. We can check whether or not stepping is enabled when installing
or uninstalling the watchpoint and operate on the breakpoint accordingly.
Signed-off-by: Will Deacon <[email protected]>
|
|
To permit handling of watchpoint exceptions without signalling a
debugger, it is necessary to reserve breakpoint registers for in-kernel
use only.
This patch ensures that we record and subtract the number of reserved
breakpoints from the number of usable breakpoint registers that we
advertise to userspace via the ptrace API.
Signed-off-by: Will Deacon <[email protected]>
|
|
On ARM, debug exceptions occur in the form of data or prefetch aborts.
One difference is that debug exceptions require access to per-cpu banked
registers and data structures which are not saved in the low-level exception
code. For kernels built with CONFIG_PREEMPT, there is an unlikely scenario
that the debug handler ends up running on a different CPU from the one
that originally signalled the event, resulting in random data being read
from the wrong registers.
This patch adds a debug_entry macro to the low-level exception handling
code which checks whether the taken exception is a debug exception. If
it is, the preempt count for the faulting process is incremented. After
the debug handler has finished, the count is decremented.
Acked-by: Catalin Marinas <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
|
|
The current hw_breakpoint code tries to fix up the alignment of
breakpoints so that we can make use of sparse byte-address-select
bits in the control register and give the illusion that we can
set breakpoints on unaligned addresses.
Although this works on v6 cores, v7 forbids this behaviour, instead
requiring breakpoints to be set on aligned addresses and have contiguous
byte-address-select ranges depending on the instruction set in use.
For ARM the only supported size is 4 bytes, whilst Thumb-2 also permits
2 byte breakpoints (watchpoints can be of 1, 2, 4 or 8 bytes long).
This patch simplifies the alignment fixup code so that we require
addresses to be aligned to the size of the corresponding breakpoint.
This allows us to handle the common case of breaking on a half-word
aligned Thumb-2 instruction and also allows us to set byte watchpoints
on arbitrary addresses.
Signed-off-by: Will Deacon <[email protected]>
|
|
The ARMv7 debug architecture doesn't make any guarantees about the
contents of debug control registers following a debug logic reset.
This patch ensures that we reset the control registers when a cpu
comes ONLINE (for example, with hotplug) so that when we enable
monitor mode while inserting a breakpoint we won't exhibit random
behaviour.
Signed-off-by: Will Deacon <[email protected]>
|
|
ARMv7 architects a system for saving and restoring the debug registers
across low-power modes. At the heart of this system is a lock register
which, when set, forbids writes to the debug registers. While locked,
writes to debug registers via the co-processor interface will result
in undefined instruction traps. Linux currently doesn't make use of
this feature because we update the debug registers on context switch
anyway, however the status of the lock is IMPLEMENTATION DEFINED on
reset.
This patch ensures that the lock is cleared during boot so that we
can write to the debug registers safely.
Signed-off-by: Will Deacon <[email protected]>
|
|
Signed-off-by: Russell King <[email protected]>
|
|
As our SMP implementation uses MESI protocols. Grouping together data
which is mostly only read together means that we avoid unnecessary
cache line bouncing when this code shares a cache line with other data.
In other words, cache lines associated with read-mostly data are
expected to spend most of their time in shared state.
Signed-off-by: Russell King <[email protected]>
|
|
For kernels built with PREEMPT_RT, critical sections protected
by standard spinlocks are preemptible. This is not acceptable
on perf as (a) we may be scheduled onto a different CPU whilst
reading/writing banked PMU registers and (b) the latency when
reading the PMU registers becomes unpredictable.
This patch upgrades the pmu_lock spinlock to a raw_spinlock
instead.
Reported-by: Jamie Iles <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
Russell reported a number of warnings coming from sparse when
checking the ARM perf_event.c files:
| perf_event.c seems to also have problems too:
|
| CHECK arch/arm/kernel/perf_event.c
| arch/arm/kernel/perf_event.c:37:1: warning: symbol 'pmu_lock' was not declared. Should it be static?
| arch/arm/kernel/perf_event.c:70:1: warning: symbol 'cpu_hw_events' was not declared. Should it be static?
| arch/arm/kernel/perf_event.c:1006:1: warning: symbol 'armv6pmu_enable_event' was not declared. Should it be static?
| arch/arm/kernel/perf_event.c:1113:1: warning: symbol 'armv6pmu_stop' was not declared. Should it be static?
| arch/arm/kernel/perf_event.c:1956:6: warning: symbol 'armv7pmu_enable_event' was not declared. Should it be static?
| arch/arm/kernel/perf_event.c:3072:14: warning: incorrect type in argument 1 (different address spaces)
| arch/arm/kernel/perf_event.c:3072:14: expected void const volatile [noderef] <asn:1>*<noident>
| arch/arm/kernel/perf_event.c:3072:14: got struct frame_tail *tail
| arch/arm/kernel/perf_event.c:3074:49: warning: incorrect type in argument 2 (different address spaces)
| arch/arm/kernel/perf_event.c:3074:49: expected void const [noderef] <asn:1>*from
| arch/arm/kernel/perf_event.c:3074:49: got struct frame_tail *tail
This patch resolves these issues so we can live in silence
again.
Reported-by: Russell King <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
When kexec is used to start a crash kernel the other cores
are notified. These non-crashing cores will save their state
in the crash notes and then do nothing.
Signed-off-by: Per Fransson <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
The existing code invokes the syscall with rubbish in r7,
due to what looks like an incorrect literal load idiom.
Reviewed-by: Will Deacon <[email protected]>
Signed-off-by: Dave Martin <[email protected]>
Acked-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
As we've now removed the spinlock and bitmask, we have nothing left
which requires interrupts to be disabled when sending an IPI. All
current IPI-sending implementations use the GIC, which also does not
require interrupts disabled when calling gic_raise_softirq().
Remove the now unnecessary IRQ disable.
Signed-off-by: Russell King <[email protected]>
|
|
Avoid using bitmasks and locks in the percpu area for IPIs, and instead
use individual software generated interrupts to identify the reason for
the IPI. This avoids the problems of having spinlocks in the percpu
area.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
This allows us to use smp_cross_call() to trigger a number of different
software generated interrupts, rather than combining them all on one
SGI. Recover the SGI number via do_IPI.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
If a section is not marked with SHF_ALLOC, it will be discarded
by the module code. Therefore, it is not correct to register
the unwind tables.
Signed-off-by: Russell King <[email protected]>
|
|
There's no need to keep pointers to the ELF sections available while
the module is loaded - we only need the section pointers while we're
finding and registering the unwind tables, which can all be done during
the finalize stage of loading.
Signed-off-by: Russell King <[email protected]>
|
|
Thumb-2.
The 32-bit conditional branches in Thumb-2 have a shorter range
(+/-512K) than their ARM counterparts (+/-32MB). The linker does
not currently generate trampolines to extend the range of these
Thumb-2 conditional branches, resulting in link errors when vmlinux
is sufficiently large, e.g.:
head.o:(.text+0x464): relocation truncated to fit: R_ARM_THM_JUMP19
This patch forces the longer-range, unconditional branch encoding
by use of an explicit IT instruction. The resulting branches are
triggered on the same conditions as before.
Signed-off-by: Dave Martin <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
kernel/head.S
Directives such as .long and .word do not magically cause the
assembler location counter to become aligned in gas. As a result,
using these directives in code sections can result in misaligned
data words when building a Thumb-2 kernel (CONFIG_THUMB2_KERNEL).
This is a Bad Thing, since the ABI permits the compiler to assume
that fundamental types of word size or above are word- aligned when
accessing them from C. If the data is not really word-aligned,
this can cause impaired performance and stray alignment faults in
some circumstances.
In general, the following rules should be applied when using data
word declaration directives inside code sections:
* .quad and .double:
.align 3
* .long, .word, .single, .float:
.align (or .align 2)
* .short:
No explicit alignment required, since Thumb-2
instructions are always 2 or 4 bytes in size.
immediately after an instruction.
Reviewed-by: Will Deacon <[email protected]>
Signed-off-by: Dave Martin <[email protected]>
Acked-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
Directives such as .long and .word do not magically cause the
assembler location counter to become aligned in gas. As a result,
using these directives in code sections can result in misaligned
data words when building a Thumb-2 kernel (CONFIG_THUMB2_KERNEL).
This is a Bad Thing, since the ABI permits the compiler to assume
that fundamental types of word size or above are word- aligned when
accessing them from C. If the data is not really word-aligned,
this can cause impaired performance and stray alignment faults in
some circumstances.
In general, the following rules should be applied when using data
word declaration directives inside code sections:
* .quad and .double:
.align 3
* .long, .word, .single, .float:
.align (or .align 2)
* .short:
No explicit alignment required, since Thumb-2
instructions are always 2 or 4 bytes in size.
immediately after an instruction.
Reviewed-by: Will Deacon <[email protected]>
Signed-off-by: Dave Martin <[email protected]>
Acked-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
|
|
Rather than passing the pte value to __pte_error, pass the raw pte_t
cookie instead. Do the same for pmd and pgd functions.
Signed-off-by: Russell King <[email protected]>
|
|
The perf hardware pmu got initialized at various points in the boot,
some before early_initcall() some after (notably arch_initcall).
The problem is that the NMI lockup detector is ran from early_initcall()
and expects the hardware pmu to be present.
Sanitize this by moving all architecture hardware pmu implementations to
initialize at early_initcall() and move the lockup detector to an explicit
initcall right after that.
Cc: paulus <[email protected]>
Cc: davem <[email protected]>
Cc: Michael Cree <[email protected]>
Cc: Deng-Cheng Zhu <[email protected]>
Acked-by: Paul Mundt <[email protected]>
Acked-by: Will Deacon <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <1290707759.2145.119.camel@laptop>
Signed-off-by: Ingo Molnar <[email protected]>
|
|
swp_emulate is only used on ARMv7+, and includes ARMv7+ assembly
instructions. Allow the assembler to accept ARMv7 instructions,
but leave the compiler's code generation options alone.
Signed-off-by: Russell King <[email protected]>
|
|
|
|
The ARM perf_event.c file contains all PMU backends and, as new PMUs
are introduced, will continue to grow.
This patch follows the example of x86 and splits the PMU implementations
into separate files which are then #included back into the main
file. Compile-time guards are added to each PMU file to avoid compiling
in code that is not relevant for the version of the architecture which
we are targetting.
Acked-by: Jean Pihet <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
|
|
Currently, perf uses the PMU ID as an index into a string table
to look up the name of a given PMU.
This patch encodes the name of a PMU directly into the arm_pmu
structure so that PMU-specific code can be factored out into
separate files.
Acked-by: Jamie Iles <[email protected]>
Acked-by: Jean Pihet <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
|
|
In preparation for separating the PMU-specific code, this patch adds
self-contained init functions to each PMU, therefore removing any
PMU-specific knowledge from the PMU-agnostic init_hw_perf_events
function.
Acked-by: Jamie Iles <[email protected]>
Acked-by: Jean Pihet <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
|
|
Unlike other pmu functions, armv6pmu_pmu_stop is not declared static.
This patch adds the missing keyword.
Acked-by: Jamie Iles <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
|
|
The functions for mapping PMU events (perf, cache and raw) are
common between all PMU types and differ only in the data on which
they operate.
This patch implements common definitions of these mapping functions
and changes the arm_pmu struct to hold pointers to the data which
they require. This is in anticipation of separating out the PMU-specific
code into separate files.
Acked-by: Jamie Iles <[email protected]>
Acked-by: Jean Pihet <[email protected]>
Signed-off-by: Will Deacon <[email protected]>
|