Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:51 +0000 (18:23 -0800)]
powerpc/vas: Export HVWC to debugfs
Export the VAS Window context information to debugfs.
We need to hold a mutex when closing the window to prevent a race
with the debugfs read(). Rather than introduce a per-instance mutex,
we use the global vas_mutex for now, since it is not heavily contended.
The window->cop field is only relevant to a receive window so we were
not setting it for a send window (which is is paired to a receive window
anyway). But to simplify reporting in debugfs, set the 'cop' field for the
send window also.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:50 +0000 (18:23 -0800)]
powerpc/vas, nx-842: Define and use chip_to_vas_id()
Define a helper, chip_to_vas_id() to map a given chip id to corresponding
vas id.
Normally, callers of vas_rx_win_open() and vas_tx_win_open() want the VAS
window to be on the same chip where the calling thread is executing. These
callers can pass in -1 for the VAS id.
This interface will be useful if a thread running on one chip wants to open
a window on another chip (like the NX-842 driver does during start up).
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:49 +0000 (18:23 -0800)]
powerpc/vas: Create cpu to vas id mapping
Create a cpu to vasid mapping so callers can specify -1 instead of
trying to find a VAS id.
Changelog[v2]
[Michael Ellerman] Use per-cpu variables to simplify code.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:48 +0000 (18:23 -0800)]
powerpc/vas: poll for return of window credits
Normally, the NX driver waits for the CRBs to be processed before closing
the window. But it is better to ensure that the credits are returned before
the window gets reassigned later.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:47 +0000 (18:23 -0800)]
powerpc/vas: Save configured window credits
Save the configured max window credits for a window in the vas_window
structure. We will need this when polling for return of window credits.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:46 +0000 (18:23 -0800)]
powerpc/vas: Reduce polling interval for busy state
A VAS window is normally in "busy" state for only a short duration.
Reduce the time we wait for the window to go to "not-busy" state to
speed-up vas_win_close() a bit.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:45 +0000 (18:23 -0800)]
powerpc/vas: Use helper to unpin/close window
Use a helper to have the hardware unpin and mark a window closed.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:44 +0000 (18:23 -0800)]
powerpc/vas: Drop poll_window_cast_out().
Polling for window cast out is listed in the spec, but turns out that
it is not strictly necessary and slows down window close. Making it a
stub for now.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:43 +0000 (18:23 -0800)]
powerpc/vas: Cleanup some debug code
Clean up vas.h and the debug code around ifdef vas_debug.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:42 +0000 (18:23 -0800)]
powerpc/vas: Validate window credits
NX-842, the only user of VAS, sets the window credits to default values
but VAS should check the credits against the possible max values.
The VAS_WCREDS_MIN is not needed and can be dropped.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sukadev Bhattiprolu [Wed, 8 Nov 2017 02:23:41 +0000 (18:23 -0800)]
powerpc/vas: init missing fields from [rt]xattr
Initialize a few missing window context fields from the window attributes
specified by the caller. These fields are currently set to their default
values by the caller (NX-842), but would be good to apply them anyway.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 24 Oct 2017 11:44:44 +0000 (21:44 +1000)]
powerpc/64: Set DSCR default initially from SPR
Take the DSCR value set by firmware as the dscr_default value,
rather than zero.
POWER9 recommends DSCR default to a non-zero value.
Signed-off-by: From: Nicholas Piggin <npiggin@gmail.com>
[mpe: Make record_spr_defaults() __init]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Mon, 23 Oct 2017 08:05:07 +0000 (18:05 +1000)]
powerpc/powernv: Avoid waiting for secondary hold spinloop with OPAL
OPAL boot does not insert secondaries at 0x60 to wait at the secondary
hold spinloop. Instead they are started later, and inserted at
generic_secondary_smp_init(), which is after the secondary hold
spinloop.
Avoid waiting on this spinloop when booting with OPAL firmware. This
wait always times out that case.
This saves 100ms boot time on powernv, and 10s of seconds of real time
when booting on the simulator in SMP.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 7 Nov 2017 07:53:09 +0000 (18:53 +1100)]
powerpc/64s/radix: Improve TLB flushing for page table freeing
Unmaps that free page tables always flush the entire PID, which is
sub-optimal. Provide TLB range flushing with an additional PWC flush
that can be use for va range invalidations with PWC flush.
Time to munmap N pages of memory including last level page table
teardown (after mmap, touch), local invalidate:
N 1 2 4 8 16 32 64
vanilla 3.2us 3.3us 3.4us 3.6us 4.1us 5.2us 7.2us
patched 1.4us 1.5us 1.7us 1.9us 2.6us 3.7us 6.2us
Global invalidate:
N 1 2 4 8 16 32 64
vanilla 2.2us 2.3us 2.4us 2.6us 3.2us 4.1us 6.2us
patched 2.1us 2.5us 3.4us 5.2us 8.7us 15.7us 6.2us
Local invalidates get much better across the board. Global ones have
the same issue where multiple tlbies for va flush do get slower than
the single tlbie to invalidate the PID. None of this test captures
the TLB benefits of avoiding killing everything.
Global gets worse, but it is brought in to line with global invalidate
for munmap()s that do not free page tables.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 7 Nov 2017 07:53:08 +0000 (18:53 +1100)]
powerpc/64s/radix: Introduce local single page ceiling for TLB range flush
The single page flush ceiling is the cut-off point at which we switch
from invalidating individual pages, to invalidating the entire process
address space in response to a range flush.
Introduce a local variant of this heuristic because local and global
tlbie have significantly different properties:
- Local tlbiel requires 128 instructions to invalidate a PID, global
tlbie only 1 instruction.
- Global tlbie instructions are expensive broadcast operations.
The local ceiling has been made much higher, 2x the number of
instructions required to invalidate the entire PID (i.e., 256 pages).
Time to mprotect N pages of memory (after mmap, touch), local invalidate:
N 32 34 64 128 256 512
vanilla 7.4us 9.0us 14.6us 26.4us 50.2us 98.3us
patched 7.4us 7.8us 13.8us 26.4us 51.9us 98.3us
The behaviour of both is identical at N=32 and N=512. Between there,
the vanilla kernel does a PID invalidate and the patched kernel does
a va range invalidate.
At N=128, these require the same number of tlbiel instructions, so
the patched version can be sen to be cheaper when < 128, and more
expensive when > 128. However this does not well capture the cost
of invalidated TLB.
The additional cost at 256 pages does not seem prohibitive. It may
be the case that increasing the limit further would continue to be
beneficial to avoid invalidating all of the process's TLB entries.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 7 Nov 2017 07:53:07 +0000 (18:53 +1100)]
powerpc/64s/radix: Optimize flush_tlb_range
Currently for radix, flush_tlb_range flushes the entire PID, because
the Linux mm code does not tell us about page size here for THP vs
regular pages. This is quite sub-optimal for small mremap / mprotect
/ change_protection.
So implement va range flushes with two flush passes, one for each
page size (regular and THP). The second flush has an order of matnitude
fewer tlbie instructions than the first, so it is a relatively small
additional cost.
There is still room for improvement here with some changes to generic
APIs, particularly if there are mostly THP pages to be invalidated,
the small page flushes could be reduced.
Time to mprotect 1 page of memory (after mmap, touch):
vanilla 2.9us 1.8us
patched 1.2us 1.6us
Time to mprotect 30 pages of memory (after mmap, touch):
vanilla 8.2us 7.2us
patched 6.9us 17.9us
Time to mprotect 34 pages of memory (after mmap, touch):
vanilla 9.1us 8.0us
patched 9.0us 8.0us
34 pages is the point at which the invalidation switches from va
to entire PID, which tlbie can do in a single instruction. This is
why in the case of 30 pages, the new code runs slower for this test.
This is a deliberate tradeoff already present in the unmap and THP
promotion code, the idea is that the benefit from avoiding flushing
entire TLB for this PID on all threads in the system.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 7 Nov 2017 07:53:06 +0000 (18:53 +1100)]
powerpc/64s/radix: Implement _tlbie(l)_va_range flush functions
Move the barriers and range iteration down into the _tlbie* level,
which improves readability.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 7 Nov 2017 07:53:05 +0000 (18:53 +1100)]
powerpc/64s/radix: Optimize TLB range flush barriers
Short range flushes issue a sequences of tlbie(l) instructions for
individual effective addresses. These do not all require individual
barrier sequences, only one covering all tlbie(l) instructions.
Commit
f7327e0ba3 ("powerpc/mm/radix: Remove unnecessary ptesync")
made a similar optimization for tlbiel for PID flushing.
For tlbie, the ISA says:
The tlbsync instruction provides an ordering function for the
effects of all tlbie instructions executed by the thread executing
the tlbsync instruction, with respect to the memory barrier
created by a subsequent ptesync instruction executed by the same
thread.
Time to munmap 30 pages of memory (after mmap, touch):
local global
vanilla 10.9us 22.3us
patched 3.4us 14.4us
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Fri, 10 Nov 2017 09:55:03 +0000 (20:55 +1100)]
Merge branch 'fixes' into next
We have some dependencies & conflicts between patches in fixes and
things to go in next, both in the radix TLB flush code and the IMC PMU
driver. So merge fixes into next.
Gustavo Romero [Wed, 1 Nov 2017 19:23:41 +0000 (15:23 -0400)]
selftests/powerpc: Check FP/VEC on exception in TM
Add a self test to check if FP/VEC/VSX registers are sane (restored
correctly) after a FP/VEC/VSX unavailable exception is caught during a
transaction.
This test checks all possibilities in a thread regarding the combination
of MSR.[FP|VEC] states in a thread and for each scenario raises a
FP/VEC/VSX unavailable exception in transactional state, verifying if
vs0 and vs32 registers, which are representatives of FP/VEC/VSX reg
sets, are not corrupted.
Signed-off-by: Gustavo Romero <gromero@linux.vnet.ibm.com>
Signed-off-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Balbir Singh [Mon, 30 Oct 2017 11:01:12 +0000 (22:01 +1100)]
powerpc/xmon: Support dumping software pagetables
It would be nice to be able to dump page tables in a particular
context.
eg: dumping vmalloc space:
0:mon> dv 0xd00037fffff00000
pgd @ 0xc0000000017c0000
pgdp @ 0xc0000000017c00d8 = 0x00000000f10b1000
pudp @ 0xc0000000f10b13f8 = 0x00000000f10d0000
pmdp @ 0xc0000000f10d1ff8 = 0x00000000f1102000
ptep @ 0xc0000000f1102780 = 0xc0000000f1ba018e
Maps physical address = 0x00000000f1ba0000
Flags = Accessed Dirty Read Write
This patch does not replicate the complex code of dump_pagetable and
has no support for bolted linear mapping, thats why I've it's called
dump virtual page table support. The format of the PTE can be expanded
even further to add more useful information about the flags in the PTE
if required.
Signed-off-by: Balbir Singh <bsingharora@gmail.com>
[mpe: Bike shed the output format, show the pgdir, fix build failures]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michal Suchanek [Tue, 11 Jul 2017 11:26:26 +0000 (13:26 +0200)]
powerpc/mm/hash: Remove stale comment.
In commit
e6f81a92015b ("powerpc/mm/hash: Support 68 bit VA") the
masking is folded into ASM_VSID_SCRAMBLE but the comment about masking
is removed only from the firt use of ASM_VSID_SCRAMBLE.
Signed-off-by: Michal Suchanek <msuchanek@suse.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Alexey Kardashevskiy [Tue, 7 Nov 2017 03:43:01 +0000 (14:43 +1100)]
powerpc/powernv/ioda: Remove explicit max window size check
DMA windows can only have a size of power of two on IODA2 hardware and
using memory_hotplug_max() to determine the upper limit won't work
correcly if it returns not power of two value.
This removes the check as the platform code does this check in
pnv_pci_ioda2_setup_default_config() anyway; the other client is VFIO
and that thing checks against locked_vm limit which prevents the userspace
from locking too much memory.
It is expected to impact DPDK on machines with non-power-of-two RAM size,
mostly. KVM guests are less likely to be affected as usually guests get
less than half of hosts RAM.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Shriya [Fri, 13 Oct 2017 04:36:41 +0000 (10:06 +0530)]
powerpc/powernv/cpufreq: Fix the frequency read by /proc/cpuinfo
The call to /proc/cpuinfo in turn calls cpufreq_quick_get() which
returns the last frequency requested by the kernel, but may not
reflect the actual frequency the processor is running at. This patch
makes a call to cpufreq_get() instead which returns the current
frequency reported by the hardware.
Fixes: fb5153d05a7d ("powerpc: powernv: Implement ppc_md.get_proc_freq()")
Signed-off-by: Shriya <shriyak@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Fri, 3 Nov 2017 04:13:21 +0000 (15:13 +1100)]
powerpc/64s/idle: avoid POWER9 DD1 and DD2.0 PMU workaround on DD2.1
DD2.1 does not have to save MMCR0 for all state-loss idle states,
only after deep idle states (like other PMU registers).
Reviewed-by: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Fri, 3 Nov 2017 04:13:20 +0000 (15:13 +1100)]
powerpc/64s/idle: avoid POWER9 DD1 and DD2.0 ERAT workaround on DD2.1
DD2.1 does not have to flush the ERAT after a state-loss idle.
Performance testing was done on a DD2.1 using only the stop0 idle state
(the shallowest state which supports state loss), using context_switch
selftest configured to ping-poing between two threads on the same core
and two different cores.
Performance improvement for same core is 7.0%, different cores is 14.8%.
Reviewed-by: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Fri, 3 Nov 2017 04:13:19 +0000 (15:13 +1100)]
powerpc: add POWER9_DD20 feature
Cc: Michael Neuling <mikey@neuling.org>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Thu, 2 Nov 2017 03:09:06 +0000 (14:09 +1100)]
powerpc: Remove facility loadups on transactional {fp, vec, vsx} unavailable
After handling a transactional FP, Altivec or VSX unavailable exception.
The return to userspace code will detect that the TIF_RESTORE_TM bit is
set and call restore_tm_state(). restore_tm_state() will call
restore_math() to ensure that the correct facilities are loaded.
This means that all the loadup code in {fp,altivec,vsx}_unavailable_tm()
is doing pointless work and can simply be removed.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Thu, 2 Nov 2017 03:09:05 +0000 (14:09 +1100)]
powerpc: Always save/restore checkpointed regs during treclaim/trecheckpoint
Lazy save and restore of FP/Altivec means that a userspace process can
be sent to userspace with FP or Altivec disabled and loaded only as
required (by way of an FP/Altivec unavailable exception). Transactional
Memory complicates this situation as a transaction could be started
without FP/Altivec being loaded up. This causes the hardware to
checkpoint incorrect registers. Handling FP/Altivec unavailable
exceptions while a thread is transactional requires a reclaim and
recheckpoint to ensure the CPU has correct state for both sets of
registers.
tm_reclaim() has optimisations to not always save the FP/Altivec
registers to the checkpointed save area. This was originally done
because the caller might have information that the checkpointed
registers aren't valid due to lazy save and restore. We've also been a
little vague as to how tm_reclaim() leaves the FP/Altivec state since it
doesn't necessarily always save it to the thread struct. This has lead
to an (incorrect) assumption that it leaves the checkpointed state on
the CPU.
tm_recheckpoint() has similar optimisations in reverse. It may not
always reload the checkpointed FP/Altivec registers from the thread
struct before the trecheckpoint. It is therefore quite unclear where it
expects to get the state from. This didn't help with the assumption
made about tm_reclaim().
These optimisations sit in what is by definition a slow path. If a
process has to go through a reclaim/recheckpoint then its transaction
will be doomed on returning to userspace. This mean that the process
will be unable to complete its transaction and be forced to its failure
handler. This is already an out if line case for userspace. Furthermore,
the cost of copying 64 times 128 bits from registers isn't very long[0]
(at all) on modern processors. As such it appears these optimisations
have only served to increase code complexity and are unlikely to have
had a measurable performance impact.
Our transactional memory handling has been riddled with bugs. A cause
of this has been difficulty in following the code flow, code complexity
has not been our friend here. It makes sense to remove these
optimisations in favour of a (hopefully) more stable implementation.
This patch does mean that some times the assembly will needlessly save
'junk' registers which will subsequently get overwritten with the
correct value by the C code which calls the assembly function. This
small inefficiency is far outweighed by the reduction in complexity for
general TM code, context switching paths, and transactional facility
unavailable exception handler.
0: I tried to measure it once for other work and found that it was
hiding in the noise of everything else I was working with. I find it
exceedingly likely this will be the case here.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Thu, 2 Nov 2017 03:09:04 +0000 (14:09 +1100)]
powerpc: Force reload for recheckpoint during tm {fp, vec, vsx} unavailable exception
Lazy save and restore of FP/Altivec means that a userspace process can
be sent to userspace with FP or Altivec disabled and loaded only as
required (by way of an FP/Altivec unavailable exception). Transactional
Memory complicates this situation as a transaction could be started
without FP/Altivec being loaded up. This causes the hardware to
checkpoint incorrect registers. Handling FP/Altivec unavailable
exceptions while a thread is transactional requires a reclaim and
recheckpoint to ensure the CPU has correct state for both sets of
registers.
tm_reclaim() has optimisations to not always save the FP/Altivec
registers to the checkpointed save area. This was originally done
because the caller might have information that the checkpointed
registers aren't valid due to lazy save and restore. We've also been a
little vague as to how tm_reclaim() leaves the FP/Altivec state since it
doesn't necessarily always save it to the thread struct. This has lead
to an (incorrect) assumption that it leaves the checkpointed state on
the CPU.
tm_recheckpoint() has similar optimisations in reverse. It may not
always reload the checkpointed FP/Altivec registers from the thread
struct before the trecheckpoint. It is therefore quite unclear where it
expects to get the state from. This didn't help with the assumption
made about tm_reclaim().
This patch is a minimal fix for ease of backporting. A more correct fix
which removes the msr parameter to tm_reclaim() and tm_recheckpoint()
altogether has been upstreamed to apply on top of this patch.
Fixes: dc3106690b20 ("powerpc: tm: Always use fp_state and vr_state to
store live registers")
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Thu, 2 Nov 2017 03:09:03 +0000 (14:09 +1100)]
powerpc: Don't enable FP/Altivec if not checkpointed
Lazy save and restore of FP/Altivec means that a userspace process can
be sent to userspace with FP or Altivec disabled and loaded only as
required (by way of an FP/Altivec unavailable exception). Transactional
Memory complicates this situation as a transaction could be started
without FP/Altivec being loaded up. This causes the hardware to
checkpoint incorrect registers. Handling FP/Altivec unavailable
exceptions while a thread is transactional requires a reclaim and
recheckpoint to ensure the CPU has correct state for both sets of
registers.
Lazy save and restore of FP/Altivec cannot be done if a process is
transactional. If a facility was enabled it must remain enabled whenever
a thread is transactional.
Commit
dc16b553c949 ("powerpc: Always restore FPU/VEC/VSX if hardware
transactional memory in use") ensures that the facilities are always
enabled if a thread is transactional. A bug in the introduced code may
cause it to inadvertently enable a facility that was (and should remain)
disabled. The problem with this extraneous enablement is that the
registers for the erroneously enabled facility have not been correctly
recheckpointed - the recheckpointing code assumed the facility would
remain disabled.
Further compounding the issue, the transactional {fp,altivec,vsx}
unavailable code has been incorrectly using the MSR to enable
facilities. The presence of the {FP,VEC,VSX} bit in the regs->msr simply
means if the registers are live on the CPU, not if the kernel should
load them before returning to userspace. This has worked due to the bug
mentioned above.
This causes transactional threads which return to their failure handler
to observe incorrect checkpointed registers. Perhaps an example will
help illustrate the problem:
A userspace process is running and uses both FP and Altivec registers.
This process then continues to run for some time without touching
either sets of registers. The kernel subsequently disables the
facilities as part of lazy save and restore. The userspace process then
performs a tbegin and the CPU checkpoints 'junk' FP and Altivec
registers. The process then performs a floating point instruction
triggering a fp unavailable exception in the kernel.
The kernel then loads the FP registers - and only the FP registers.
Since the thread is transactional it must perform a reclaim and
recheckpoint to ensure both the checkpointed registers and the
transactional registers are correct. It then (correctly) enables
MSR[FP] for the process. Later (on exception exist) the kernel also
(inadvertently) enables MSR[VEC]. The process is then returned to
userspace.
Since the act of loading the FP registers doomed the transaction we know
CPU will fail the transaction, restore its checkpointed registers, and
return the process to its failure handler. The problem is that we're
now running with Altivec enabled and the 'junk' checkpointed registers
are restored. The kernel had only recheckpointed FP.
This patch solves this by only activating FP/Altivec if userspace was
using them when it entered the kernel and not simply if the process is
transactional.
Fixes: dc16b553c949 ("powerpc: Always restore FPU/VEC/VSX if hardware
transactional memory in use")
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:46 +0000 (13:41 +1100)]
mtd: powernv_flash: Use opal_async_wait_response_interruptible()
The OPAL calls performed in this driver shouldn't be using
opal_async_wait_response() as this performs a wait_event() which, on
long running OPAL calls could result in hung task warnings. wait_event()
prevents timely signal delivery which is also undesirable.
This patch also attempts to quieten down the use of dev_err() when
errors haven't actually occurred and also to return better information up
the stack rather than always -EIO.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Acked-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:45 +0000 (13:41 +1100)]
powerpc/powernv: Add OPAL_BUSY to opal_error_code()
Also export opal_error_code() so that it can be used in modules
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:44 +0000 (13:41 +1100)]
powerpc/opal: Add opal_async_wait_response_interruptible() to opal-async
This patch adds an _interruptible version of opal_async_wait_response().
This is useful when a long running OPAL call is performed on behalf of
a userspace thread, for example, the opal_flash_{read,write,erase}
functions performed by the powernv-flash MTD driver.
It is foreseeable that these functions would take upwards of two
minutes causing the wait_event() to block long enough to cause hung
task warnings. Furthermore, wait_event_interruptible() is preferable
as otherwise there is no way for signals to stop the process which is
going to be confusing in userspace.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Stewart Smith [Fri, 3 Nov 2017 02:41:43 +0000 (13:41 +1100)]
powernv/opal-sensor: remove not needed lock
Parallel sensor reads could run out of async tokens due to
opal_get_sensor_data grabbing tokens but then doing the sensor
read behind a mutex, essentially serializing the (possibly
asynchronous and relatively slow) sensor read.
It turns out that the mutex isn't needed at all, not only
should the OPAL interface allow concurrent reads, the implementation
is certainly safe for that, and if any sensor we were reading
from somewhere isn't, doing the mutual exclusion in the kernel
is the wrong place to do it, OPAL should be doing it for the kernel.
So, remove the mutex.
Additionally, we shouldn't be printing out an error when we don't
get a token as the only way this should happen is if we've been
interrupted in down_interruptible() on the semaphore.
Reported-by: Robert Lippert <rlippert@google.com>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:42 +0000 (13:41 +1100)]
powerpc/opal: Rework the opal-async interface
Future work will add an opal_async_wait_response_interruptible()
which will call wait_event_interruptible(). This work requires extra
token state to be tracked as wait_event_interruptible() can return and
the caller could release the token before OPAL responds.
Currently token state is tracked with two bitfields which are 64 bits
big but may not need to be as OPAL informs Linux how many async tokens
there are. It also uses an array indexed by token to store response
messages for each token.
The bitfields make it difficult to add more state and also provide a
hard maximum as to how many tokens there can be - it is possible that
OPAL will inform Linux that there are more than 64 tokens.
Rather than add a bitfield to track the extra state, rework the
internals slightly.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
[mpe: Fix __opal_async_get_token() when no tokens are free]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:41 +0000 (13:41 +1100)]
powerpc/opal: Make __opal_async_{get, release}_token() static
There are no callers of both __opal_async_get_token() and
__opal_async_release_token().
This patch also removes the possibility of "emergency through
synchronous call to __opal_async_get_token()" as such it makes more
sense to initialise opal_sync_sem for the maximum number of async
tokens.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:40 +0000 (13:41 +1100)]
mtd: powernv_flash: Don't return -ERESTARTSYS on interrupted token acquisition
Because the MTD core might split up a read() or write() from userspace
into several calls to the driver, we may fail to get a token but already
have done some work, best to return -EINTR back to userspace and have
them decide what to do.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Acked-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:39 +0000 (13:41 +1100)]
mtd: powernv_flash: Remove pointless goto in driver init
powernv_flash_probe() has pointless goto statements which jump to the
end of the function to simply return a variable. Rather than checking
for error and going to the label, just return the error as soon as it is
detected.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Acked-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:38 +0000 (13:41 +1100)]
mtd: powernv_flash: Don't treat OPAL_SUCCESS as an error
While this driver expects to interact asynchronously, OPAL is well
within its rights to return OPAL_SUCCESS to indicate that the operation
completed without the need for a callback. We shouldn't treat
OPAL_SUCCESS as an error rather we should wrap up and return promptly to
the caller.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Acked-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Fri, 3 Nov 2017 02:41:37 +0000 (13:41 +1100)]
mtd: powernv_flash: Use WARN_ON_ONCE() rather than BUG_ON()
BUG_ON() should be reserved in situations where we can not longer
guarantee the integrity of the system. In the case where
powernv_flash_async_op() receives an impossible op, we can still
guarantee the integrity of the system.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Acked-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
William A. Kennington III [Fri, 22 Sep 2017 23:58:00 +0000 (16:58 -0700)]
powerpc/opal: Fix EBUSY bug in acquiring tokens
The current code checks the completion map to look for the first token
that is complete. In some cases, a completion can come in but the
token can still be on lease to the caller processing the completion.
If this completed but unreleased token is the first token found in the
bitmap by another tasks trying to acquire a token, then the
__test_and_set_bit call will fail since the token will still be on
lease. The acquisition will then fail with an EBUSY.
This patch reorganizes the acquisition code to look at the
opal_async_token_map for an unleased token. If the token has no lease
it must have no outstanding completions so we should never see an
EBUSY, unless we have leased out too many tokens. Since
opal_async_get_token_inrerruptible is protected by a semaphore, we
will practically never see EBUSY anymore.
Fixes: 8d7248232208 ("powerpc/powernv: Infrastructure to support OPAL async completion")
Signed-off-by: William A. Kennington III <wak@google.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Arnd Bergmann [Sat, 4 Nov 2017 21:26:52 +0000 (22:26 +0100)]
powerpc/eeh: Stop using do_gettimeofday()
This interface is inefficient and deprecated because of the y2038
overflow.
ktime_get_seconds() is an appropriate replacement here, since it
has sufficient granularity but is more efficient and uses monotonic
time.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Acked-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Vaibhav Jain [Wed, 11 Oct 2017 12:30:20 +0000 (18:00 +0530)]
cxl: Rework the implementation of cxl_stop_trace_psl9()
Presently the PSL9 specific cxl_stop_trace_psl9() only stops the RX0
traces on the CXL adapter when a PSL error irq is triggered. The patch
updates the function to stop all the traces arrays and move them to
the FIN state. The implementation issues the mmio to TRACECFG register
to stop the trace array iff it already not in FIN state. This prevents
the issue of trace data being reset in case of multiple stop mmio
issued for a single trace array.
Also the patch does some refactoring of existing cxl_stop_trace_psl9()
and cxl_stop_trace_psl8() functions by moving them to 'pci.c' from
'debugfs.c' file and marking them as static.
Signed-off-by: Vaibhav Jain <vaibhav@linux.vnet.ibm.com>
Acked-by: Frederic Barrat <fbarrat@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Sandipan Das [Fri, 1 Sep 2017 18:53:01 +0000 (00:23 +0530)]
bpf: take advantage of stack_depth tracking in powerpc JIT
Take advantage of stack_depth tracking, originally introduced for
x64, in powerpc JIT as well. Round up allocated stack by 16 bytes
to make sure it stays aligned for functions called from JITed bpf
program.
Signed-off-by: Sandipan Das <sandipan@linux.vnet.ibm.com>
Reviewed-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Thu, 12 Oct 2017 04:45:25 +0000 (15:45 +1100)]
powerpc/tm: Don't check for WARN in TM Bad Thing handling
Currently when we take a TM Bad Thing program check exception, we
search the bug table to see if the program check was generated by a
WARN/WARN_ON etc.
That makes no sense, the WARN macros use trap instructions, which
should never generate a TM Bad Thing exception. If they ever did that
would be a bug and we should oops.
We do have some hand-coded bugs in tm.S, using EMIT_BUG_ENTRY, but
those are all BUGs not WARNs, and they all use trap instructions
anyway. Almost certainly this check was incorrectly copied from the
REASON_TRAP handling in the same function.
Remove it.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Acked-By: Michael Neuling <mikey@neuling.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Tue, 24 Oct 2017 15:48:49 +0000 (17:48 +0200)]
powerpc/mm: Add a CONFIG option to choose if radix is used by default
Currently if the hardware supports the radix MMU we will use
it, *unless* "disable_radix" is passed on the kernel command line.
However some users would like the reverse semantics. ie. The kernel
uses the hash MMU by default, unless radix is explicitly requested on
the command line.
So add a CONFIG option to choose whether we use radix by default or
not, and expand the disable_radix command line option to allow
"disable_radix=no" which *enables* radix.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Thu, 19 Oct 2017 04:08:43 +0000 (15:08 +1100)]
powerpc/64s: Replace CONFIG_PPC_STD_MMU_64 with CONFIG_PPC_BOOK3S_64
CONFIG_PPC_STD_MMU_64 indicates support for the "standard" powerpc MMU
on 64-bit CPUs. The "standard" MMU refers to the hash page table MMU
found in "server" processors, from IBM mainly.
Currently CONFIG_PPC_STD_MMU_64 is == CONFIG_PPC_BOOK3S_64. While it's
annoying to have two symbols that always have the same value, it's not
quite annoying enough to bother removing one.
However with the arrival of Power9, we now have the situation where
CONFIG_PPC_STD_MMU_64 is enabled, but the kernel is running using the
Radix MMU - *not* the "standard" MMU. So it is now actively confusing
to use it, because it implies that code is disabled or inactive when
the Radix MMU is in use, however that is not necessarily true.
So s/CONFIG_PPC_STD_MMU_64/CONFIG_PPC_BOOK3S_64/, and do some minor
formatting updates of some of the affected lines.
This will be a pain for backports, but c'est la vie.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Thu, 19 Oct 2017 04:08:19 +0000 (15:08 +1100)]
powerpc/64: Free up CPU_FTR_ICSWX
The last user of CPU_FTR_ICSWX was removed in commit
6ff4d3e96652 ("powerpc: Remove old unused icswx based coprocessor
support"), so free the bit up for future use.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Aneesh Kumar K.V [Mon, 16 Oct 2017 07:01:40 +0000 (12:31 +0530)]
powerpc/mm/hash: Add pr_fmt() to hash_utils64.c
Make the printks look a bit nicer by adding a prefix.
Radix config now do
radix-mmu: Page sizes from device-tree:
radix-mmu: Page size shift = 12 AP=0x0
radix-mmu: Page size shift = 16 AP=0x5
radix-mmu: Page size shift = 21 AP=0x1
radix-mmu: Page size shift = 30 AP=0x2
This patch update hash config to do similar dmesg output. With the patch we have
hash-mmu: Page sizes from device-tree:
hash-mmu: base_shift=12: shift=12, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=0
hash-mmu: base_shift=12: shift=16, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=7
hash-mmu: base_shift=12: shift=24, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=56
hash-mmu: base_shift=16: shift=16, sllp=0x0110, avpnm=0x00000000, tlbiel=1, penc=1
hash-mmu: base_shift=16: shift=24, sllp=0x0110, avpnm=0x00000000, tlbiel=1, penc=8
hash-mmu: base_shift=20: shift=20, sllp=0x0111, avpnm=0x00000000, tlbiel=0, penc=2
hash-mmu: base_shift=24: shift=24, sllp=0x0100, avpnm=0x00000001, tlbiel=0, penc=0
hash-mmu: base_shift=34: shift=34, sllp=0x0120, avpnm=0x000007ff, tlbiel=0, penc=3
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Christophe Leroy [Wed, 18 Oct 2017 09:16:47 +0000 (11:16 +0200)]
powerpc/ipic: Fix status get and status clear
IPIC Status is provided by register IPIC_SERSR and not by IPIC_SERMR
which is the mask register.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Alexey Kardashevskiy [Wed, 27 Sep 2017 06:52:31 +0000 (16:52 +1000)]
powerpc/powernv: Reserve a hole which appears after enabling IOV
In order to make generic IOV code work, the physical function IOV BAR
should start from offset of the first VF. Since M64 segments share
PE number space across PHB, and some PEs may be in use at the time
when IOV is enabled, the existing code shifts the IOV BAR to the index
of the first PE/VF. This creates a hole in IOMEM space which can be
potentially taken by some other device.
This reserves a temporary hole on a parent and releases it when IOV is
disabled; the temporary resources are stored in pci_dn to avoid
kmalloc/free.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Tyrel Datwyler [Fri, 29 Sep 2017 00:19:20 +0000 (20:19 -0400)]
powerpc/pseries/vio: Dispose of virq mapping on vdevice unregister
When a vdevice is DLPAR removed from the system the vio subsystem
doesn't bother unmapping the virq from the irq_domain. As a result we
have a virq mapped to a hardware irq that is no longer valid for the
irq_domain. A side effect is that we are left with /proc/irq/<irq#>
affinity entries, and attempts to modify the smp_affinity of the irq
will fail.
In the following observed example the kernel log is spammed by
ics_rtas_set_affinity errors after the removal of a VSCSI adapter.
This is a result of irqbalance trying to adjust the affinity every 10
seconds.
rpadlpar_io: slot U8408.E8E.10A7ACV-V5-C25 removed
ics_rtas_set_affinity: ibm,set-xive irq=655385 returns -3
ics_rtas_set_affinity: ibm,set-xive irq=655385 returns -3
This patch fixes the issue by calling irq_dispose_mapping() on the
virq of the viodev on unregister.
Fixes: f2ab6219969f ("powerpc/pseries: Add PFO support to the VIO bus")
Signed-off-by: Tyrel Datwyler <tyreld@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 24 Oct 2017 13:06:54 +0000 (23:06 +1000)]
powerpc/64s/radix: Fix process table entry cache invalidation
According to the architecture, the process table entry cache must be
flushed with tlbie RIC=2.
Currently the process table entry is set to invalid right before the
PID is returned to the allocator, with no invalidation. This works on
existing implementations that are known to not cache the process table
entry for any except the current PIDR.
It is architecturally correct and cleaner to invalidate with RIC=2
after clearing the process table entry and before the PID is returned
to the allocator. This can be done in arch_exit_mmap that runs before
the final flush, and to ensure the final flush (fullmm) is always a
RIC=2 variant.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 24 Oct 2017 13:06:53 +0000 (23:06 +1000)]
powerpc/64s/radix: Improve preempt handling in TLB code
Preempt should be consistently disabled for mm_is_thread_local tests,
so bring the rest of these under preempt_disable().
Preempt does not need to be disabled for the mm->context.id tests,
which allows simplification and removal of gotos.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Mon, 23 Oct 2017 07:08:15 +0000 (17:08 +1000)]
powerpc/powernv: Use FIXUP_ENDIAN_HV in OPAL return
Close the recoverability gap for OPAL calls by using FIXUP_ENDIAN_HV
in the return path.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Mon, 23 Oct 2017 07:08:14 +0000 (17:08 +1000)]
powerpc/book3s: Add an HV variant of FIXUP_ENDIAN that is recoverable
Add an HV variant of FIXUP_ENDIAN which uses HSRR[01] and does not
clear MSR[RI], which improves recoverability.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Mon, 23 Oct 2017 07:08:13 +0000 (17:08 +1000)]
powerpc/book3s: Use label for FIXUP_ENDIAN macro branch
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Sat, 21 Oct 2017 07:56:06 +0000 (17:56 +1000)]
powerpc/64: Fix latency tracing for lazy irq replay
When returning from an exception to a soft-enabled context, pending
IRQs are replayed but IRQ tracing is not reset, so a number of them
can get chained together into the same IRQ-disabled trace.
Fix this by having __check_irq_replay re-set IRQ trace. This is
conceptually where we respond to the next interrupt, so it fits the
semantics of the IRQ tracer.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Sun, 5 Nov 2017 12:33:55 +0000 (23:33 +1100)]
KVM: PPC: Book3S HV: Handle host system reset in guest mode
If the host takes a system reset interrupt while a guest is running,
the CPU must exit the guest before processing the host exception
handler.
After this patch, taking a sysrq+x with a CPU running in a guest
gives a trace like this:
cpu 0x27: Vector: 100 (System Reset) at [
c000000fdf5776f0]
pc:
c008000010158b80: kvmppc_run_core+0x16b8/0x1ad0 [kvm_hv]
lr:
c008000010158b80: kvmppc_run_core+0x16b8/0x1ad0 [kvm_hv]
sp:
c000000fdf577850
msr:
9000000002803033
current = 0xc000000fdf4b1e00
paca = 0xc00000000fd4d680 softe: 3 irq_happened: 0x01
pid = 6608, comm = qemu-system-ppc
Linux version
4.14.0-rc7-01489-g47e1893a404a-dirty #26 SMP
[
c000000fdf577a00]
c008000010159dd4 kvmppc_vcpu_run_hv+0x3dc/0x12d0 [kvm_hv]
[
c000000fdf577b30]
c0080000100a537c kvmppc_vcpu_run+0x44/0x60 [kvm]
[
c000000fdf577b60]
c0080000100a1ae0 kvm_arch_vcpu_ioctl_run+0x118/0x310 [kvm]
[
c000000fdf577c00]
c008000010093e98 kvm_vcpu_ioctl+0x530/0x7c0 [kvm]
[
c000000fdf577d50]
c000000000357bf8 do_vfs_ioctl+0xd8/0x8c0
[
c000000fdf577df0]
c000000000358448 SyS_ioctl+0x68/0x100
[
c000000fdf577e30]
c00000000000b220 system_call+0x58/0x6c
--- Exception: c01 (System Call) at
00007fff76868df0
SP (
7fff7069baf0) is in userspace
Fixes: e36d0a2ed5 ("powerpc/powernv: Implement NMI IPI with OPAL_SIGNAL_SYSTEM_RESET")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Madhavan Srinivasan [Thu, 2 Nov 2017 12:12:26 +0000 (17:42 +0530)]
powerpc/perf: Fix core-imc hotplug callback failure during imc initialization
Call trace observed during boot:
nest_capp0_imc performance monitor hardware support registered
nest_capp1_imc performance monitor hardware support registered
core_imc memory allocation for cpu 56 failed
Unable to handle kernel paging request for data at address 0xffa400010
Faulting instruction address: 0xc000000000bf3294
0:mon> e
cpu 0x0: Vector: 300 (Data Access) at [
c000000ff38ff8d0]
pc:
c000000000bf3294: mutex_lock+0x34/0x90
lr:
c000000000bf3288: mutex_lock+0x28/0x90
sp:
c000000ff38ffb50
msr:
9000000002009033
dar:
ffa400010
dsisr: 80000
current = 0xc000000ff383de00
paca = 0xc000000007ae0000 softe: 0 irq_happened: 0x01
pid = 13, comm = cpuhp/0
Linux version 4.11.0-39.el7a.ppc64le (mockbuild@ppc-058.build.eng.bos.redhat.com) (gcc version 4.8.5
20150623 (Red Hat 4.8.5-16) (GCC) ) #1 SMP Tue Oct 3 07:42:44 EDT 2017
0:mon> t
[
c000000ff38ffb80]
c0000000002ddfac perf_pmu_migrate_context+0xac/0x470
[
c000000ff38ffc40]
c00000000011385c ppc_core_imc_cpu_offline+0x1ac/0x1e0
[
c000000ff38ffc90]
c000000000125758 cpuhp_invoke_callback+0x198/0x5d0
[
c000000ff38ffd00]
c00000000012782c cpuhp_thread_fun+0x8c/0x3d0
[
c000000ff38ffd60]
c0000000001678d0 smpboot_thread_fn+0x290/0x2a0
[
c000000ff38ffdc0]
c00000000015ee78 kthread+0x168/0x1b0
[
c000000ff38ffe30]
c00000000000b368 ret_from_kernel_thread+0x5c/0x74
While registering the cpuhoplug callbacks for core-imc, if we fails
in the cpuhotplug online path for any random core (either because opal call to
initialize the core-imc counters fails or because memory allocation fails for
that core), ppc_core_imc_cpu_offline() will get invoked for other cpus who
successfully returned from cpuhotplug online path.
But in the ppc_core_imc_cpu_offline() path we are trying to migrate the event
context, when core-imc counters are not even initialized. Thus creating the
above stack dump.
Add a check to see if core-imc counters are enabled or not in the cpuhotplug
offline path before migrating the context to handle this failing scenario.
Fixes: 885dcd709ba9 ("powerpc/perf: Add nest IMC PMU support")
Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Signed-off-by: Anju T Sudhakar <anju@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Naveen N. Rao [Mon, 30 Oct 2017 15:12:09 +0000 (20:42 +0530)]
powerpc/kprobes: Dereference function pointers only if the address does not belong to kernel text
This makes the changes introduced in commit
83e840c770f2c5
("powerpc64/elfv1: Only dereference function descriptor for non-text
symbols") to be specific to the kprobe subsystem.
We previously changed ppc_function_entry() to always check the provided
address to confirm if it needed to be dereferenced. This is actually
only an issue for kprobe blacklisted asm labels (through use of
_ASM_NOKPROBE_SYMBOL) and can cause other issues with ftrace. Also, the
additional checks are not really necessary for our other uses.
As such, move this check to the kprobes subsystem.
Fixes: 83e840c770f2 ("powerpc64/elfv1: Only dereference function descriptor for non-text symbols")
Cc: stable@vger.kernel.org # v4.13+
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Naveen N. Rao [Mon, 30 Oct 2017 15:12:08 +0000 (20:42 +0530)]
Revert "powerpc64/elfv1: Only dereference function descriptor for non-text symbols"
This reverts commit
83e840c770f2c5 ("powerpc64/elfv1: Only dereference
function descriptor for non-text symbols").
Chandan reported that on newer kernels, trying to enable function_graph
tracer on ppc64 (BE) locks up the system with the following trace:
Unable to handle kernel paging request for data at address 0x600000002fa30010
Faulting instruction address: 0xc0000000001f1300
Thread overran stack, or stack corrupted
Oops: Kernel access of bad area, sig: 11 [#1]
BE SMP NR_CPUS=2048 DEBUG_PAGEALLOC NUMA pSeries
Modules linked in:
CPU: 1 PID: 6586 Comm: bash Not tainted
4.14.0-rc3-00162-g6e51f1f-dirty #20
task:
c000000625c07200 task.stack:
c000000625c07310
NIP:
c0000000001f1300 LR:
c000000000121cac CTR:
c000000000061af8
REGS:
c000000625c088c0 TRAP: 0380 Not tainted (
4.14.0-rc3-00162-g6e51f1f-dirty)
MSR:
8000000000001032 <SF,ME,IR,DR,RI> CR:
28002848 XER:
00000000
CFAR:
c0000000001f1320 SOFTE: 0
...
NIP [
c0000000001f1300] .__is_insn_slot_addr+0x30/0x90
LR [
c000000000121cac] .kernel_text_address+0x18c/0x1c0
Call Trace:
[
c000000625c08b40] [
c0000000001bd040] .is_module_text_address+0x20/0x40 (unreliable)
[
c000000625c08bc0] [
c000000000121cac] .kernel_text_address+0x18c/0x1c0
[
c000000625c08c50] [
c000000000061960] .prepare_ftrace_return+0x50/0x130
[
c000000625c08cf0] [
c000000000061b10] .ftrace_graph_caller+0x14/0x34
[
c000000625c08d60] [
c000000000121b40] .kernel_text_address+0x20/0x1c0
[
c000000625c08df0] [
c000000000061960] .prepare_ftrace_return+0x50/0x130
...
[
c000000625c0ab30] [
c000000000061960] .prepare_ftrace_return+0x50/0x130
[
c000000625c0abd0] [
c000000000061b10] .ftrace_graph_caller+0x14/0x34
[
c000000625c0ac40] [
c000000000121b40] .kernel_text_address+0x20/0x1c0
[
c000000625c0acd0] [
c000000000061960] .prepare_ftrace_return+0x50/0x130
[
c000000625c0ad70] [
c000000000061b10] .ftrace_graph_caller+0x14/0x34
[
c000000625c0ade0] [
c000000000121b40] .kernel_text_address+0x20/0x1c0
This is because ftrace is using ppc_function_entry() for obtaining the
address of return_to_handler() in prepare_ftrace_return(). The call to
kernel_text_address() itself gets traced and we end up in a recursive
loop.
Fixes: 83e840c770f2 ("powerpc64/elfv1: Only dereference function descriptor for non-text symbols")
Cc: stable@vger.kernel.org # v4.13+
Reported-by: Chandan Rajendra <chandan@linux.vnet.ibm.com>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 24 Oct 2017 13:06:52 +0000 (23:06 +1000)]
powerpc/64s/radix: Fix preempt imbalance in TLB flush
Fixes: 424de9c6e3f8 ("powerpc/mm/radix: Avoid flushing the PWC on every flush_tlb_range")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Paul Mackerras [Wed, 25 Oct 2017 07:16:53 +0000 (18:16 +1100)]
powerpc: Fix check for copy/paste instructions in alignment handler
Commit
07d2a628bc00 ("powerpc/64s: Avoid cpabort in context switch
when possible", 2017-06-09) changed the definition of PPC_INST_COPY
and in so doing inadvertently broke the check for copy/paste
instructions in the alignment fault handler. The check currently
matches no instructions.
This fixes it by ANDing both sides of the comparison with the mask.
Fixes: 07d2a628bc00 ("powerpc/64s: Avoid cpabort in context switch when possible")
Cc: stable@vger.kernel.org # v4.13+
Reported-by: Markus Trippelsdorf <markus@trippelsdorf.de>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Guilherme G. Piccoli [Thu, 19 Oct 2017 19:41:33 +0000 (17:41 -0200)]
powerpc/perf: Fix IMC allocation routine
When setting nr_cpus=1, we observed a crash in IMC code during boot
due to a missing allocation: basically, IMC code is taking the number
of threads into account in imc_mem_init() and if we manually set
nr_cpus for a value that is not multiple of the number of threads per
core, an integer division in that function will discard the decimal
portion, leading IMC to not allocate one mem_info struct. This causes
a NULL pointer dereference later, on is_core_imc_mem_inited().
This patch just rounds that division up, fixing the bug.
Signed-off-by: Guilherme G. Piccoli <gpiccoli@linux.vnet.ibm.com>
Acked-by: Anju T Sudhakar <anju@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Markus Elfring [Wed, 18 Oct 2017 18:48:52 +0000 (20:48 +0200)]
powerpc/pseries: Cleanup error handling in iommu_pseries_alloc_group()
Although kfree(NULL) is legal, it's a bit lazy to rely on that to
implement the error handling. So do it the normal Linux way using
labels for each failure path.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
[mpe: Squash a few patches and rewrite change log]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Markus Elfring [Tue, 17 Oct 2017 11:31:42 +0000 (13:31 +0200)]
powerpc-opal: Fix a typo in a comment line of two file headers
Fix a word in these descriptions.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Markus Elfring [Tue, 5 Sep 2017 16:47:02 +0000 (18:47 +0200)]
powerpc/axonram: Drop unnecessary variable initialisation
The local variable "rc" will eventually be set only to an error code.
Thus omit the explicit initialisation at the beginning.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Geert Uytterhoeven [Fri, 2 Jun 2017 12:38:47 +0000 (14:38 +0200)]
powerpc: dts: acadia: DT fix s/#interrupts-parent/#interrupt-parent/
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Mon, 9 Oct 2017 10:52:44 +0000 (21:52 +1100)]
powerpc/perf/hv-24x7: Fix incorrect comparison in memord
In the hv-24x7 code there is a function memord() which tries to
implement a sort function return -1, 0, 1. However one of the
conditions is incorrect, such that it can never be true, because we
will have already returned.
I don't believe there is a bug in practice though, because the
comparisons are an optimisation prior to calling memcmp().
Fix it by swapping the second comparision, so it can be true.
Reported-by: David Binderman <dcb314@hotmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Mon, 9 Oct 2017 10:54:05 +0000 (21:54 +1100)]
powerpc: Disable the fast-endian switch syscall by default
Back in 2008 we added support for "fast little-endian switch" in the
syscall path. This added a special case syscall number 0x1ebe, which
is caught very early in the system call exception and switches endian
with as little overhead as possible. See commit
745a14cc264b
("[POWERPC] Add fast little-endian switch system call") for full
details.
Although it is fast, it's also completely non standard. The "syscall
number" is out of the range of normal syscalls, it can't be traced or
audited, and it's a bit of a wart. To the best of our knowledge it was
only used by one program, now long since discontinued.
So in an effort to shake out any current users, put it behind a config
option, and make it default n. If anyone *is* using it they can
quickly reinstate it with a rebuild, and we can flip it to default y.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Mon, 9 Oct 2017 10:54:04 +0000 (21:54 +1100)]
powerpc/64s: Move the two FAST_ENDIAN macros next to each other
So we can #ifdef them in the next patch.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Mon, 9 Oct 2017 10:59:32 +0000 (21:59 +1100)]
powerpc/xmon: Add kstack base to paca dump
When dumping the paca in xmon we currently show kstack. Although it's
not hard it's a bit fiddly to work out what the bounds of the kernel
stack should be based on the kstack value.
To make life easier and "kstack_base" which is the base (lowest
address) of the kernel stack, eg:
kstack = 0xc0000000f1a7be30 (0x258)
kstack_base = 0xc0000000f1a78000
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Andrew Donnellan [Thu, 12 Oct 2017 02:16:59 +0000 (13:16 +1100)]
powerpc/configs: Enable I2C_CHARDEV for pseries and powernv
i2c-dev provides an interface for userspace programs to interact with I2C
devices, and is very helpful for I2C-related debugging.
Enable it in pseries_defconfig and powernv_defconfig. It's already enabled
in many other powerpc defconfigs.
Signed-off-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Mon, 16 Oct 2017 07:11:00 +0000 (12:41 +0530)]
powerpc/mm/radix: Drop unneeded NULL check
We call these functions with non-NULL mm or vma. Hence we can skip the
NULL check in these functions. We also remove now unused function
__local_flush_hugetlb_page().
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
[mpe: Drop the checks with is_vm_hugetlb_page() as noticed by Nick]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Breno Leitao [Tue, 17 Oct 2017 18:20:18 +0000 (16:20 -0200)]
powerpc/xmon: Check before calling xive functions
Currently xmon could call XIVE functions from OPAL even if the XIVE is
disabled or does not exist in the system, as in POWER8 machines. This
causes the following exception:
1:mon> dx
cpu 0x1: Vector: 700 (Program Check) at [
c000000423c93450]
pc:
c00000000009cfa4: opal_xive_dump+0x50/0x68
lr:
c0000000000997b8: opal_return+0x0/0x50
This patch simply checks if XIVE is enabled before calling XIVE
functions.
Fixes: 243e25112d06 ("powerpc/xive: Native exploitation of the XIVE interrupt controller")
Suggested-by: Guilherme G. Piccoli <gpiccoli@linux.vnet.ibm.com>
Signed-off-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Vaibhav Jain [Wed, 20 Sep 2017 06:23:22 +0000 (11:53 +0530)]
cxl: Provide debugfs access to PSL_DEBUG/XSL_DEBUG registers
Access to PSL/XSL_DEBUG registers on the adapter provides easy access
to the debug facilities provided by PSL/XSL. So this patch adds two
new files (debug, xsl-debug) to the cxl-adapter specific debugfs
folder located at /sys/kernel/debugfs/cxl/card<n>, which will provide
direct r/w access to corrosponding debug registers in the adapter
config-space.
Signed-off-by: Vaibhav Jain <vaibhav@linux.vnet.ibm.com>
Acked-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Acked-by: Frederic Barrat <fbarrat@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Neuling [Thu, 12 Oct 2017 10:17:19 +0000 (21:17 +1100)]
powerpc/tm: P9 disable transactionally suspended sigcontexts
Unfortunately userspace can construct a sigcontext which enables
suspend. Thus userspace can force Linux into a path where trechkpt is
executed.
This patch blocks this from happening on POWER9 by sanity checking
sigcontexts passed in.
ptrace doesn't have this problem as only MSR SE and BE can be changed
via ptrace.
This patch also adds a number of WARN_ON()s in case we ever enter
suspend when we shouldn't. This should not happen, but if it does the
symptoms are soft lockup warnings which are not obviously TM related,
so the WARN_ON()s should make it obvious what's happening.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Thu, 12 Oct 2017 10:17:18 +0000 (21:17 +1100)]
powerpc/powernv: Enable TM without suspend if possible
Some Power9 revisions can run in a mode where TM operates without
suspended state. If we find ourself on a CPU that might be in this
mode, we query OPAL to check, and if so we reenable TM in CPU
features, and enable a new user feature to signal to userspace that we
are in this mode.
We do not enable the "normal" user feature, PPC_FEATURE2_HTM, but we
do enable PPC_FEATURE2_HTM_NOSC because that indicates to userspace
that the kernel will abort transactions on syscall entry, which is
true regardless of the suspend mode.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Thu, 12 Oct 2017 10:17:17 +0000 (21:17 +1100)]
powerpc: Add PPC_FEATURE2_HTM_NO_SUSPEND
Some CPUs can operate in a mode where TM (Transactional Memory) is
enabled but the suspended state of TM is disabled. In this mode
tsuspend does not enter suspended state, instead the transaction is
aborted. Similarly any other event that would lead to suspended state
instead aborts the transaction.
There is also an ABI change, in that in this mode processes are not
allowed to sigreturn with an MSR that would lead to suspended state,
Linux will instead return an error to the sigreturn syscall.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Thu, 12 Oct 2017 10:17:16 +0000 (21:17 +1100)]
powerpc/tm: Add commandline option to disable hardware transactional memory
Currently the kernel relies on firmware to inform it whether or not the
CPU supports HTM and as long as the kernel was built with
CONFIG_PPC_TRANSACTIONAL_MEM=y then it will allow userspace to make
use of the facility.
There may be situations where it would be advantageous for the kernel
to not allow userspace to use HTM, currently the only way to achieve
this is to recompile the kernel with CONFIG_PPC_TRANSACTIONAL_MEM=n.
This patch adds a simple commandline option so that HTM can be
disabled at boot time.
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
[mpe: Simplify to a bool, move to prom.c, put doco in the right place.
Always disable, regardless of initial state, to avoid user confusion.]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Fri, 20 Oct 2017 00:10:30 +0000 (11:10 +1100)]
Merge branch 'topic/ppc-kvm' into next
Bring in some KVM commits we need (the TM one in particular).
Michael Ellerman [Thu, 12 Oct 2017 11:58:54 +0000 (22:58 +1100)]
KVM: PPC: Tie KVM_CAP_PPC_HTM to the user-visible TM feature
Currently we use CPU_FTR_TM to decide if the CPU/kernel can support
TM (Transactional Memory), and if it's true we advertise that to
Qemu (or similar) via KVM_CAP_PPC_HTM.
PPC_FEATURE2_HTM is the user-visible feature bit, which indicates that
the CPU and kernel can support TM. Currently CPU_FTR_TM and
PPC_FEATURE2_HTM always have the same value, either true or false, so
using the former for KVM_CAP_PPC_HTM is correct.
However some Power9 CPUs can operate in a mode where TM is enabled but
TM suspended state is disabled. In this mode CPU_FTR_TM is true, but
PPC_FEATURE2_HTM is false. Instead a different PPC_FEATURE2 bit is
set, to indicate that this different mode of TM is available.
It is not safe to let guests use TM as-is, when the CPU is in this
mode. So to prevent that from happening, use PPC_FEATURE2_HTM to
determine the value of KVM_CAP_PPC_HTM.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Paul Mackerras [Thu, 19 Oct 2017 04:14:20 +0000 (15:14 +1100)]
Revert "KVM: PPC: Book3S HV: POWER9 does not require secondary thread management"
This reverts commit
94a04bc25a2c6296bd0c5e82c10e8231c2b11f77.
In order to run HPT guests on a radix POWER9 host, we will have to run
the host in single-threaded mode, because POWER9 processors do not
currently support running some threads of a core in HPT mode while
others are in radix mode ("mixed mode").
That means that we will need the same mechanisms that are used on
POWER8 to make the secondary threads available to KVM, which were
disabled on POWER9 by commit
94a04bc25a2c.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Bringmann [Fri, 8 Sep 2017 20:47:56 +0000 (15:47 -0500)]
powerpc/vphn: Fix numa update end-loop bug
powerpc/vphn: On Power systems with shared configurations of CPUs
and memory, there are some issues with the association of additional
CPUs and memory to nodes when hot-adding resources. This patch
fixes an end-of-updates processing problem observed occasionally
in numa_update_cpu_topology().
Signed-off-by: Michael Bringmann <mwb@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Bringmann [Fri, 8 Sep 2017 20:47:47 +0000 (15:47 -0500)]
powerpc/hotplug: Improve responsiveness of hotplug change
powerpc/hotplug: On Power systems with shared configurations of CPUs
and memory, there are some issues with the association of additional
CPUs and memory to nodes when hot-adding resources. During hotplug
CPU operations, this patch resets the timer on topology update work
function to a small value to better ensure that the CPU topology is
detected and configured sooner.
Signed-off-by: Michael Bringmann <mwb@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Bringmann [Fri, 8 Sep 2017 20:47:36 +0000 (15:47 -0500)]
powerpc/vphn: Improve recognition of PRRN/VPHN
powerpc/vphn: On Power systems with shared configurations of CPUs
and memory, there are some issues with the association of additional
CPUs and memory to nodes when hot-adding resources. This patch
updates the initialization checks to independently recognize PRRN
or VPHN support.
Signed-off-by: Michael Bringmann <mwb@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Bringmann [Fri, 8 Sep 2017 20:47:27 +0000 (15:47 -0500)]
powerpc/vphn: Update CPU topology when VPHN enabled
powerpc/vphn: On Power systems with shared configurations of CPUs
and memory, there are some issues with the association of additional
CPUs and memory to nodes when hot-adding resources. This patch
corrects the currently broken capability to set the topology for
shared CPUs in LPARs. At boot time for shared CPU lpars, the
topology for each CPU was being set to node zero. Now when
numa_update_cpu_topology() is called appropriately, the Virtual
Processor Home Node (VPHN) capabilities information provided by the
pHyp allows the appropriate node in the shared configuration to be
selected for the CPU.
Signed-off-by: Michael Bringmann <mwb@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Balbir Singh [Fri, 29 Sep 2017 04:26:55 +0000 (14:26 +1000)]
powerpc/mce: hookup memory_failure for UE errors
If we are in user space and hit a UE error, we now have the
basic infrastructure to walk the page tables and find out
the effective address that was accessed, since the DAR
is not valid.
We use a work_queue content to hookup the bad pfn, any
other context causes problems, since memory_failure itself
can call into schedule() via lru_drain_ bits.
We could probably poison the struct page to avoid a race
between detection and taking corrective action.
Signed-off-by: Balbir Singh <bsingharora@gmail.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Balbir Singh [Fri, 29 Sep 2017 04:26:54 +0000 (14:26 +1000)]
powerpc/mce: Hookup ierror (instruction) UE errors
Hookup instruction errors (UE) for memory offling via memory_failure()
in a manner similar to load/store errors (derror). Since we have access
to the NIP, the conversion is a one step process in this case.
Signed-off-by: Balbir Singh <bsingharora@gmail.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Balbir Singh [Fri, 29 Sep 2017 04:26:53 +0000 (14:26 +1000)]
powerpc/mce: Hookup derror (load/store) UE errors
Extract physical_address for UE errors by walking the page
tables for the mm and address at the NIP, to extract the
instruction. Then use the instruction to find the effective
address via analyse_instr().
We might have page table walking races, but we expect them to
be rare, the physical address extraction is best effort. The idea
is to then hook up this infrastructure to memory failure eventually.
Signed-off-by: Balbir Singh <bsingharora@gmail.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Balbir Singh [Fri, 29 Sep 2017 04:26:52 +0000 (14:26 +1000)]
powerpc/mce: Align the print of physical address better
Use the same alignment as Effective address.
Signed-off-by: Balbir Singh <bsingharora@gmail.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Balbir Singh [Fri, 29 Sep 2017 04:26:51 +0000 (14:26 +1000)]
powerpc/mce: Remove unused function get_mce_fault_addr()
There are no users of get_mce_fault_addr() since commit
1363875bdb63 ("powerpc/64s: fix handling of non-synchronous machine
checks") removed the last usage.
Signed-off-by: Balbir Singh <bsingharora@gmail.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Anju T Sudhakar [Fri, 13 Oct 2017 05:59:41 +0000 (11:29 +0530)]
powerpc/perf: Fix IMC initialization crash
Panic observed with latest firmware, and upstream kernel:
NIP init_imc_pmu+0x8c/0xcf0
LR init_imc_pmu+0x2f8/0xcf0
Call Trace:
init_imc_pmu+0x2c8/0xcf0 (unreliable)
opal_imc_counters_probe+0x300/0x400
platform_drv_probe+0x64/0x110
driver_probe_device+0x3d8/0x580
__driver_attach+0x14c/0x1a0
bus_for_each_dev+0x8c/0xf0
driver_attach+0x34/0x50
bus_add_driver+0x298/0x350
driver_register+0x9c/0x180
__platform_driver_register+0x5c/0x70
opal_imc_driver_init+0x2c/0x40
do_one_initcall+0x64/0x1d0
kernel_init_freeable+0x280/0x374
kernel_init+0x24/0x160
ret_from_kernel_thread+0x5c/0x74
While registering nest imc at init, cpu-hotplug callback
nest_pmu_cpumask_init() makes an OPAL call to stop the engine. And if
the OPAL call fails, imc_common_cpuhp_mem_free() is invoked to cleanup
memory and cpuhotplug setup.
But when cleaning up the attribute group, we are dereferencing the
attribute element array without checking whether the backing element
is not NULL. This causes the kernel panic.
Add a check for the backing element prior to dereferencing the
attribute element, to handle the failing case gracefully.
Signed-off-by: Anju T Sudhakar <anju@linux.vnet.ibm.com>
Reported-by: Pridhiviraj Paidipeddi <ppaidipe@linux.vnet.ibm.com>
[mpe: Trim change log]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Kamalesh Babulal [Tue, 10 Oct 2017 14:47:32 +0000 (20:17 +0530)]
powerpc/modules: Use WARN_ON() in stub_for_addr()
Use WARN_ON(), while running out of stubs in stub_for_addr()
and abort loading of the module instead of BUG_ON().
Signed-off-by: Kamalesh Babulal <kamalesh@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Vaibhav Jain [Wed, 11 Oct 2017 06:14:41 +0000 (11:44 +0530)]
cxl: Dump PSL_FIR register on PSL9 error irq
For PSL9 currently we aren't dumping the PSL FIR register when a
PSL error interrupt is triggered. Contents of this register are useful
in debugging AFU issues.
This patch fixes issue by adding a new service_layer_ops callback
cxl_native_err_irq_dump_regs_psl9() to dump the PSL_FIR registers on a
PSL error interrupt thereby bringing the behavior in line with PSL on
POWER-8. Also the existing service_layer_ops callback
for PSL8 has been renamed to cxl_native_err_irq_dump_regs_psl8().
Signed-off-by: Vaibhav Jain <vaibhav@linux.vnet.ibm.com>
Acked-by: Frederic Barrat <fbarrat@linux.vnet.ibm.com>
Acked-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Cyril Bur [Tue, 4 Jul 2017 01:21:15 +0000 (11:21 +1000)]
selftests/powerpc: context_switch: Fix pthread errors
Turns out pthreads returns an errno and doesn't set errno. This doesn't
play well with perror().
Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Anju T Sudhakar [Wed, 11 Oct 2017 12:57:39 +0000 (18:27 +0530)]
powerpc/perf: Add ___GFP_NOWARN flag to alloc_pages_node()
Stack trace output during a stress test:
[ 4.310049] Freeing initrd memory: 22592K
[ 4.310646] rtas_flash: no firmware flash support
[ 4.313341] cpuhp/64: page allocation failure: order:0, mode:0x14480c0(GFP_KERNEL|__GFP_ZERO|__GFP_THISNODE), nodemask=(null)
[ 4.313465] cpuhp/64 cpuset=/ mems_allowed=0
[ 4.313521] CPU: 64 PID: 392 Comm: cpuhp/64 Not tainted 4.11.0-39.el7a.ppc64le #1
[ 4.313588] Call Trace:
[ 4.313622] [
c000000f1fb1b8e0] [
c000000000c09388] dump_stack+0xb0/0xf0 (unreliable)
[ 4.313694] [
c000000f1fb1b920] [
c00000000030ef6c] warn_alloc+0x12c/0x1c0
[ 4.313753] [
c000000f1fb1b9c0] [
c00000000030ff68] __alloc_pages_nodemask+0xea8/0x1000
[ 4.313823] [
c000000f1fb1bbb0] [
c000000000113a8c] core_imc_mem_init+0xbc/0x1c0
[ 4.313892] [
c000000f1fb1bc00] [
c000000000113cdc] ppc_core_imc_cpu_online+0x14c/0x170
[ 4.313962] [
c000000f1fb1bc90] [
c000000000125758] cpuhp_invoke_callback+0x198/0x5d0
[ 4.314031] [
c000000f1fb1bd00] [
c00000000012782c] cpuhp_thread_fun+0x8c/0x3d0
[ 4.314101] [
c000000f1fb1bd60] [
c0000000001678d0] smpboot_thread_fn+0x290/0x2a0
[ 4.314169] [
c000000f1fb1bdc0] [
c00000000015ee78] kthread+0x168/0x1b0
[ 4.314229] [
c000000f1fb1be30] [
c00000000000b368] ret_from_kernel_thread+0x5c/0x74
[ 4.314313] Mem-Info:
[ 4.314356] active_anon:0 inactive_anon:0 isolated_anon:0
core_imc_mem_init() at system boot use alloc_pages_node() to get memory
and alloc_pages_node() throws this stack dump when tried to allocate
memory from a node which has no memory behind it. Add a ___GFP_NOWARN
flag in allocation request as a fix.
Signed-off-by: Anju T Sudhakar <anju@linux.vnet.ibm.com>
Reported-by: Michael Ellerman <mpe@ellerman.id.au>
Reported-by: Venkat R.B <venkatb3@in.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Anju T Sudhakar [Wed, 4 Oct 2017 06:50:52 +0000 (12:20 +0530)]
powerpc/perf: Fix for core/nest imc call trace on cpuhotplug
Nest/core pmu units are enabled only when it is used. A reference count is
maintained for the events which uses the nest/core pmu units. Currently in
*_imc_counters_release function a WARN() is used for notification of any
underflow of ref count.
The case where event ref count hit a negative value is, when perf session is
started, followed by offlining of all cpus in a given core.
i.e. in cpuhotplug offline path ppc_core_imc_cpu_offline() function set the
ref->count to zero, if the current cpu which is about to offline is the last
cpu in a given core and make an OPAL call to disable the engine in that core.
And on perf session termination, perf->destroy (core_imc_counters_release) will
first decrement the ref->count for this core and based on the ref->count value
an opal call is made to disable the core-imc engine.
Now, since cpuhotplug path already clears the ref->count for core and disabled
the engine, perf->destroy() decrementing again at event termination make it
negative which in turn fires the WARN_ON. The same happens for nest units.
Add a check to see if the reference count is alreday zero, before decrementing
the count, so that the ref count will not hit a negative value.
Signed-off-by: Anju T Sudhakar <anju@linux.vnet.ibm.com>
Reviewed-by: Santosh Sivaraj <santosh@fossix.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>