Allow the board support code to register a raw notifier callback for
NMI, similar to what is done for CU2 exceptions.
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/2958/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
I am running SMP Linux 2.6.37-rc1 on BMIPS5000 (single core dual thread)
and observe some abnormalities when doing system suspend/resume which I
narrowed down to cpu hotplugging. The suspend brings the second thread
processor down and then restarts it, after which I see memory corruption
in userspace. I started digging and found out that problem occurs because
while doing execve() the child process is getting the same ASID as the
parent, which obviously corrupts parent's address space.
Further digging showed that activate_mm() calls get_new_mmu_context() to
get a new ASID, but at this time ASID field in entryHi is 1, and
asid_cache(cpu) is 0x100 (it was just reset to ASID_FIRST_VERSION when
the secondary TP was booting).
So, get_new_mmu_context() increments the asid_cache(cpu) value to
0x101, and thus puts 0x01 into entryHi. The result - ASID field does
not get changed as it was supposed to.
My solution is very simple - do not reset asid_cache(cpu) on TP warm
restart.
Patchwork: https://patchwork.linux-mips.org/patch/1797/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
None of these files are using modular infrastructure, and build
tests reveal that none of these files are really relying on any
implicit inclusions via. module.h either. So delete them.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
On preempt-rt this lock needs to be raw, so it does not get converted
to a sleeping spinlock. Trying to sleep in a panic is not really
desireable.
Signed-off-by: Wu Zhangjin <wuzhangjin@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/2636/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The nmi parameter indicated if we could do wakeups from the current
context, if not, we would set some state and self-IPI and let the
resulting interrupt do the wakeup.
For the various event classes:
- hardware: nmi=0; PMI is in fact an NMI or we run irq_work_run from
the PMI-tail (ARM etc.)
- tracepoint: nmi=0; since tracepoint could be from NMI context.
- software: nmi=[0,1]; some, like the schedule thing cannot
perform wakeups, and hence need 0.
As one can see, there is very little nmi=1 usage, and the down-side of
not using it is that on some platforms some software events can have a
jiffy delay in wakeup (when arch_irq_work_raise isn't implemented).
The up-side however is that we can remove the nmi parameter and save a
bunch of conditionals in fast paths.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Michael Cree <mcree@orcon.net.nz>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Deng-Cheng Zhu <dengcheng.zhu@gmail.com>
Cc: Anton Blanchard <anton@samba.org>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jason Wessel <jason.wessel@windriver.com>
Cc: Don Zickus <dzickus@redhat.com>
Link: http://lkml.kernel.org/n/tip-agjev8eu666tvknpb3iaj0fg@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
We were unconditionally sending SIGBUS with an empty siginfo on FP
emulator faults. This differs from what happens when real floating
point hardware would get a fault.
For most faults we need to send SIGSEGV with the faulting address
filled in in the struct siginfo.
Reported-by: Camm Maguire <camm@maguirefamily.org>
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
To: linux-mips@linux-mips.org
Cc: Camm Maguire <camm@maguirefamily.org>
Patchwork: https://patchwork.linux-mips.org/patch/1727/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Some MIPS32R1 processors implement UserLocal (RDHWR $29) to accelerate
programs that make extensive use of thread-local storage. Therefore,
setting up the HWRENA register should not depend on cpu_has_mips_r2.
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Software events are required as part of the measurable stuff by the
Linux performance counter subsystem. Here is the list of events added by
this patch:
PERF_COUNT_SW_PAGE_FAULTS
PERF_COUNT_SW_PAGE_FAULTS_MIN
PERF_COUNT_SW_PAGE_FAULTS_MAJ
PERF_COUNT_SW_ALIGNMENT_FAULTS
PERF_COUNT_SW_EMULATION_FAULTS
Signed-off-by: Deng-Cheng Zhu <dengcheng.zhu@gmail.com>
To: linux-mips@linux-mips.org
Cc: a.p.zijlstra@chello.nl
Cc: paulus@samba.org
Cc: mingo@elte.hu
Cc: acme@redhat.com
Cc: jamie.iles@picochip.com
Acked-by: David Daney <ddaney@caviumnetworks.com>
Reviewed-by: Matt Fleming <matt@console-pimps.org>
Patchwork: https://patchwork.linux-mips.org/patch/1686/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Add missing #inclusions of <linux/irq.h> to a whole bunch of files that should
really include it. Note that this can replace #inclusions of <asm/irq.h>.
This is required for the patch to sort out irqflags handling function naming to
compile on MIPS.
The problem is that these files require access to things like setup_irq() -
which isn't available by #including <linux/interrupt.h>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Ralf Baechle <ralf@linux-mips.org>
The sixth argument of notify_die() is a signal number, the fifth is a
trap number.
Instead of passing a signal number in a randomly selected argument,
pass it in the sixth. Extract the exception code from regs and pass
that as the trap number.
Get rid of redundant cast, and remove some gratuitous spaces.
Nobody actually does anything with the signal number or trap number,
but we might as well populate them with sensible values.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
To: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/1532/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Breaking here dropped us to the default code which always sends a SIGILL
to the current process, no matter what the CU2 notifier says.
[Ralf: Currently this only hurts on Cavium and possibly some out of tree
platforms.]
Signed-off-by: Jesper Nilsson <jesper@jni.nu>
To: linux-mips@linux-mips.org
To: linux-kernel@vger.kernel.org
Patchwork: http://patchwork.linux-mips.org/patch/1391/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The only way the debugger can handle a trap in inside rcu_lock,
notify_die, or atomic_notifier_call_chain without a recursive fault is
to have a low level "first opportunity handler" do_trap_or_bp() handler.
Generally this will be something the vast majority of folks will not
need, but for those who need it, it is added as a kernel .config
option called KGDB_LOW_LEVEL_TRAP.
Also added was a die notification for oops such that kdb can catch an
oops for analysis.
There appeared to be no obvious way to pass the struct pt_regs from
the original exception back to the stack back tracer, so a special
case was added to show_stack() for when kdb is active because you
generally desire to generally look at the back trace of the original
exception.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Acked-by: Ralf Baechle <ralf@linux-mips.org>
"MIPS: Calculate proper ebase value for 64-bit kernels"
9af43ea080dd5d6c7b34f38261780e5dd43537bc (lmo) rsp.
f6be75d03c (kernel.org) broke some 64-bit
MIPS systems.
Before this we were using XKPHYS/cached as ebase and computed the uncached
xphsys/unchached address for that area. After that commit ebase became a
32-bit compat address and convert does not work anymore. We now should use
CKSEG1 for this. CKSEG1ADDR does just that in 32-bit and 64-bit.
Signed-off-by: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
To: Ralf Baechle <ralf@linux-mips.org>
Patchwork: http://patchwork.linux-mips.org/patch/1149/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The MIPS implementation of die() forgets to call notify_die() and thus notifiers
registered via register_die_notifier() are not called. This results in kgdb not
being activated on exceptions.
The only subtlety is that notify_die declares its regs argument w/o const, so
the const had to be removed from mips die() as well.
[Ralf: Fixed build error for SGI IP22 and IP28 platforms.]
Signed-off-by: Yury Polyanskiy <ypolyans@princeton.edu>
Cc: linux-mips@linux-mips.org
Patchworks: http://patchwork.linux-mips.org/patch/1142/
Acked-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---
The ebase is relative to CKSEG0 not CAC_BASE. On a 32-bit kernel they
are the same thing, for a 64-bit kernel they are not.
It happens to kind of work on a 64-bit kernel as they both reference
the same physical memory. However since the CPU uses the CKSEG0 base,
determining if a J instruction will reach always gives the wrong result
unless we use the same number the CPU uses.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
To: linux-mips@linux-mips.org
Patchwork: http://patchwork.linux-mips.org/patch/1093/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
AR7 has a larger physical offset than other MIPS based systems and therefore
needs to setup its handlers beyond the usual KSEG0 range. When running the
kernel in mapped mode this modification is also required. Remove function
comment which is now incorrect.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Eugene Konev <ejka@imfi.kspu.ru>
Signed-off-by: Florian Fainelli <florian@openwrt.org>
To: linux-mips@linux-mips.org
To: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: linux-mips@linux-mips.org
Patchwork: http://patchwork.linux-mips.org/patch/889/
Patchwork: http://patchwork.linux-mips.org/patch/932/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
All call sites of set_except_vector are already annotated with __init, so
annotate that one too.
Signed-off-by: Regards, Florian Fainelli <florian@openwrt.org>
To: linux-mips@linux-mips.org
To: David Daney <ddaney@caviumnetworks.com>
Patchwork: http://patchwork.linux-mips.org/patch/888/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
As reported by Maxime Bizon, the commit "MIPS: PowerTV: Fix support for
timer interrupts with > 64 external IRQs" have broken the r4k timer
since it didn't initialize the cp0_compare_irq_shift variable used in
c0_compare_int_pending() on the architectures whose cpu_has_mips_r2 is
false.
This patch fixes it via initializing the cp0_compare_irq_shift as the
cp0_compare_irq used in the old c0_compare_int_pending().
Reported-by: Maxime Bizon <mbizon@freebox.fr>
Signed-off-by: Wu Zhangjin <wuzhangjin@gmail.com>
Cc: David VomLehn <dvomlehn@cisco.com>
Cc: mbizon@freebox.fr
Cc: linux-mips@linux-mips.org
Patchwork: http://patchwork.linux-mips.org/patch/922/
Tested-by: Shane McDonald <mcdonald.shane@gmail.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The MIPS processor is limited to 64 external interrupt sources. Using a
greater number without IRQ sharing requires reading platform-specific
registers. On such platforms, reading the IntCtl register to determine
which interrupt corresponds to a timer interrupt will not work.
On MIPSR2 systems there is a solution - the TI bit in the Cause register,
specifically indicates that a timer interrupt has occured. This patch uses
that bit to detect interrupts for MIPSR2 processors, which may be expected
to work regardless of how the timer interrupt may be routed in the hardware.
Signed-off-by: David VomLehn (dvomlehn@cisco.com)
To: linux-mips@linux-mips.org
Patchwork: http://patchwork.linux-mips.org/patch/804/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Away with the daemons of ifdef; get ready for future COP2 users.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Patchwork: http://patchwork.linux-mips.org/patch/708/
We had an ugly #ifdef for Cavium Octeon hwrena bits in traps.c, remove
it to mach-cavium-octeon/cpu-feature-overrides.h
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Some CPUs have implementation dependent rdhwr registers. Allow them
to be enabled on a per CPU basis.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Commit 566f74f6b2 had a change that
incorrectly modified ebase. This backs out the lines that modified
ebase.
In addition, the ebase exception vector is now allocated with correct
alignment and the ebase register updated according to the architecture
specification.
Based on original patch by David VomLehn <dvomlehn@cisco.com>.
Signed-off-by: David VomLehn <dvomlehn@cisco.com>
Signed-off-by: Chris Dearman <chris@mips.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
If a context switch occurred between the watch exception and reading the
watch registers, it would be possible for the new process to corrupt their
state. Enabling interrupts only after the watch registers are read avoids
this race.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
If on Cavium, be aware of cop2 and hwrena during do_cpu().
Signed-off-by: Tomaso Paoletti <tpaoletti@caviumnetworks.com>
Signed-off-by: Paul Gortmaker <Paul.Gortmaker@windriver.com>
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Arguably using the address error handler has always been ugly. But with
processors that handle unaligned loads and stores in hardware the
current mechanism ceases to work so switch it to a BREAK instruction and
allocate break code 514 to the FPU emulator.
Yoichi Yuasa provided a build fix for CONFIG_BUG=n.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
It just so happens to be zero on all currently supported systems so this
hasn't bitten yet ...
[Ralf: Original patch from Cavium; handling of set_uncached_handler() and
de-ifdef'ed trap_init() implementation by me.]
Signed-off-by: Tomaso Paoletti <tpaoletti@caviumnetworks.com>
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Here we hook up the watch exception handler so that it sends SIGTRAP when
the hardware watch registers are triggered.
Signed-off-by: David Daney <ddaney@avtrex.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
If an interrupt happened between checking of NEED_RESCHED and WAIT
instruction, adjust EPC to restart from checking of NEED_RESCHED.
Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
trap_init issues flush_icache_range(), which uses ipi functions to
get icache flushing done on all cpus. But this is done before interrupts
are enabled and caused WARN_ON messages. This changeset introduces
a new local_flush_icache_range() and uses it before interrupts (and
additional CPUs) are enabled to avoid this problem.
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
With -ffunction-section the entries in __dbe_table aren't no longer
sorted, so the lookup of exception addresses in do_be() failed for
some addresses. To avoid this we now sort __dbe_table.
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The new kgdb architecture specific handler registers and unregisters
dynamically for exceptions depending on when you configure a kgdb I/O
driver.
Aside from initializing the exceptions earlier in the boot process,
kgdb should have no impact on a device when it is compiled in so long
as an I/O module is not configured for use.
There have been quite a number of contributors during the existence of
this patch (see arch/mips/kernel/kgdb.c). Most recently Jason
re-wrote the mips kgdb logic to use the die notification handlers.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
It is not used anywhere in tree.
Signed-off-by: David Daney <ddaney@avtrex.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The newly added check for valid stack pointer address breaks at least for
64bit kernels. Use __get_user() for accessing stack content to avoid crashes,
when doing the backtrace.
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
