- tm: Always reclaim in start_thread() for exec() class syscalls from Cyril Bur
- tm: Avoid SLB faults in treclaim/trecheckpoint when RI=0 from Michael Neuling
- eeh: Fix wrong argument passed to eeh_rmv_device() from Gavin Shan
- Initialise pci_io_base as early as possible from Darren Stevens
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=oEAY
-----END PGP SIGNATURE-----
Merge tag 'powerpc-4.7-5' into next
Pull in the fixes we sent during 4.7, we have code we want to merge into
next that depends on some of them.
When calling eeh_rmv_device() in eeh_reset_device() for partial hotplug
case, @rmv_data instead of its address is the proper argument.
Otherwise, the stack frame is corrupted when writing to
@rmv_data (actually its address) in eeh_rmv_device(). It results in
kernel crash as observed.
This fixes the issue by passing @rmv_data, not its address to
eeh_rmv_device() in eeh_reset_device().
Fixes: 67086e32b5 ("powerpc/eeh: powerpc/eeh: Support error recovery for VF PE")
Reported-by: Pridhiviraj Paidipeddi <ppaidipe@in.ibm.com>
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The PE primary bus cannot be got from its child devices when having
full hotplug in error recovery. The PE primary bus is cached, which
is done in commit <05ba75f84864> ("powerpc/eeh: Fix stale cached primary
bus"). In eeh_reset_device(), the flag (EEH_PE_PRI_BUS) is cleared
before the PCI hot remove. eeh_pe_bus_get() then returns NULL as the
PE primary bus in pnv_eeh_reset() and it crashes the kernel eventually.
This fixes the issue by clearing the flag (EEH_PE_PRI_BUS) before the
PCI hot add. With it, the PowerNV EEH reset backend (pnv_eeh_reset())
can get valid PE primary bus through eeh_pe_bus_get().
Fixes: 67086e32b5 ("powerpc/eeh: powerpc/eeh: Support error recovery for VF PE")
Reported-by: Pridhiviraj Paidipeddi <ppaiddipe@in.ibm.com>
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The function eeh_pe_reset_and_recover() is used to recover EEH
error when the passthrough device are transferred to guest and
backwards, meaning the device's driver is vfio-pci or none. In
both cases, the handlers triggered by eeh_report_reset() and
eeh_report_resume() shouldn't be called.
This ignores the error handlers from eeh_report_reset() and
eeh_report_resume().
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Reviewed-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The function eeh_pe_reset_and_recover() is used to recover EEH
error when the passthrou device are transferred to guest and
backwards. The content in the device's config space will be lost
on PE reset issued in the middle of the recovery. The function
saves/restores it before/after the reset. However, config access
to some adapters like Broadcom BCM5719 at this point will causes
fenced PHB. The config space is always blocked and we save 0xFF's
that are restored at late point. The memory BARs are totally
corrupted, causing another EEH error upon access to one of the
memory BARs.
This restores the config space on those adapters like BCM5719
from the content saved to the EEH device when it's populated,
to resolve above issue.
Fixes: 5cfb20b9 ("powerpc/eeh: Emulate EEH recovery for VFIO devices")
Cc: stable@vger.kernel.org #v3.18+
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Reviewed-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The function eeh_pe_reset_and_recover() is used to recover EEH
error when the passthrough device are transferred to guest and
backwards, meaning the device's driver is vfio-pci or none.
When the driver is vfio-pci that provides error_detected() error
handler only, the handler simply stops the guest and it's not
expected behaviour. On the other hand, no error handlers will
be called if we don't have a bound driver.
This ignores the error handler in eeh_pe_reset_and_recover()
that reports the error to device driver to avoid the exceptional
behaviour.
Fixes: 5cfb20b9 ("powerpc/eeh: Emulate EEH recovery for VFIO devices")
Cc: stable@vger.kernel.org #v3.18+
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Reviewed-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This renames pcibios_{add,remove}_pci_devices() to avoid conflicts
with names of the weak functions in PCI subsystem, which have the
prefix "pcibios". No logical changes introduced.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Reviewed-By: Alistair Popple <alistair@popple.id.au>
Reviewed-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When we have partial hotplug as part of the error recovery on PF,
the VFs that are bound with vfio-pci driver will experience hotplug.
That's not allowed.
This checks if the VF PE is passed or not. If it does, we leave
the VF without removing it.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Reviewed-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When EEH error happened to the parent PE of those PEs that have
been passed through to guest, the error is propagated to guest
domain and the VFIO driver's error handlers are called. It's not
correct as the error in the host domain shouldn't be propagated
to guests and affect them.
This adds one more limitation when calling EEH error handlers.
If the PE has been passed through to guest, the error handlers
won't be called.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Reviewed-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
PFs are enumerated on PCI bus, while VFs are created by PF's driver.
In EEH recovery, it has two cases:
1. Device and driver is EEH aware, error handlers are called.
2. Device and driver is not EEH aware, un-plug the device and plug it again
by enumerating it.
The special thing happens on the second case. For a PF, we could use the
original pci core to enumerate the bus, while for VF we need to record the
VFs which aer un-plugged then plug it again.
Also The patch caches the VF index in pci_dn, which can be used to
calculate VF's bus, device and function number. Those information helps to
locate the VF's PCI device instance when doing hotplug during EEH recovery
if necessary.
Signed-off-by: Wei Yang <weiyang@linux.vnet.ibm.com>
Acked-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
- eeh: Fix partial hotplug criterion from Gavin Shan
- mm: Clear the invalid slot information correctly from Aneesh Kumar K.V
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=bk3U
-----END PGP SIGNATURE-----
Merge tag 'powerpc-4.5-4' into next
Pull in our current fixes from 4.5, in particular the "Fix Multi hit
ERAT" bug is causing folks some grief when testing next.
During error recovery, the device could be removed as part of the
partial hotplug. The criterion used to come with partial hotplug
is: if the device driver provides error_detected(), slot_reset()
and resume() callbacks, it's immune from hotplug. Otherwise,
it's going to experience partial hotplug during EEH recovery. But
the criterion isn't correct enough: mlx4_core driver for Mellanox
adapters provides error_detected(), slot_reset() callbacks, but
resume() isn't there. Those Mellanox adapters won't be to involved
in the partial hotplug.
This fixes the criterion to a practical one: adpater with driver
that provides error_detected(), slot_reset() will be immune from
partial hotplug. resume() isn't mandatory.
Fixes: f2da4ccf ("powerpc/eeh: More relaxed hotplug criterion")
Cc: stable@vger.kernel.org #v4.4+
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When PE is created, its primary bus is cached to pe->bus. At later
point, the cached primary bus is returned from eeh_pe_bus_get().
However, we could get stale cached primary bus and run into kernel
crash in one case: full hotplug as part of fenced PHB error recovery
releases all PCI busses under the PHB at unplugging time and recreate
them at plugging time. pe->bus is still dereferencing the PCI bus
that was released.
This adds another PE flag (EEH_PE_PRI_BUS) to represent the validity
of pe->bus. pe->bus is updated when its first child EEH device is
online and the flag is set. Before unplugging in full hotplug for
error recovery, the flag is cleared.
Fixes: 8cdb2833 ("powerpc/eeh: Trace PCI bus from PE")
Cc: stable@vger.kernel.org #v3.11+
Reported-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Reported-by: Pradipta Ghosh <pradghos@in.ibm.com>
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Tested-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Bit 7 of the "Header Type" register indicates a multi-function device when
set. Bits 0-6 contain encoded values, where 0x1 indicates a PCI-PCI
bridge. It is incorrect to test this as though it were a mask.
For example, while the PCI 3.0 spec only defines values 0x0, 0x1, and 0x2,
it's conceivable that a future spec could define 0x3 to mean something
else; then tests for "(hdr_type & 0x7f) & PCI_HEADER_TYPE_BRIDGE" would
incorrectly succeed for this new 0x3 header type.
Test bits 0-6 of the Header Type for equality with PCI_HEADER_TYPE_BRIDGE.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
This reverts commit 527d10ef3a.
The reverted commit breaks cxlflash devices following an EEH reset (and
possibly other cxl devices, however this has not been tested).
The reverted commit changed the behaviour of eeh_reset_device() so that PHB
PEs are not unfrozen following the completion of the reset. This should not
be problematic, as no device resources should have been associated with the
PHB PE.
However, when attempting to load the cxlflash driver after a reset, the
driver attempts to read Vital Product Data through a call to
pci_read_vpd() (which is called on the physical cxl device, not on the
virtual AFU device). pci_read_vpd() in turn attempts to read from the cxl
device's config space. This fails, as the PE it's trying to read from is
still frozen. In turn, the driver gets an -ENODEV and fails to initialise.
It appears this issue only affects some parts of the VPD area, as "lspci
-vvv", which only reads a subset of the VPD bytes, is not broken by the
original patch.
At this stage, we don't fully understand why we're trying to read a frozen
PE, and we don't know how this affects other cxl devices. It is possible
that there is an underlying bug in the cxl driver or the powerpc CAPI
support code, or alternatively a bug in the PCI resource allocation/mapping
code that is incorrectly mapping resources to PE#0.
As such, this fix is incomplete, however it is necessary to prevent a
serious regression in CAPI support.
In the meantime, revert the commit, especially as it was intended to be a
non-functional change.
Cc: Gavin Shan <gwshan@linux.vnet.ibm.com>
Cc: Ian Munsie <imunsie@au1.ibm.com>
Cc: Daniel Axtens <dja@axtens.net>
Signed-off-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
On fenced PHB, the error handlers in the drivers of its subordinate
devices could return PCI_ERS_RESULT_CAN_RECOVER, indicating no reset
will be issued during the recovery. It's conflicting with the fact
that fenced PHB won't be recovered without reset.
This limits the return value from the error handlers in the drivers
of the fenced PHB's subordinate devices to PCI_ERS_RESULT_NEED_NONE
or PCI_ERS_RESULT_NEED_RESET, to ensure reset will be issued during
recovery.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Reviewed-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Currently, we rely on the existence of struct pci_driver::err_handler
to decide if the corresponding PCI device should be unplugged during
EEH recovery (partially hotplug case). However that check is not
sufficient. Some device drivers implement only some of the EEH error
handlers to collect diag-data. That means the driver still expects a
hotplug to recover from the EEH error.
This makes the hotplug criterion more relaxed: if the device driver
doesn't provide all necessary EEH error handlers, it will experience
hotplug during EEH recovery.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
[mpe: Minor change log rewording]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
On PowerNV platform, the PE is kept in frozen state until the PE
reset is completed to avoid recursive EEH error caused by MMIO
access during the period of EEH reset. The PE's frozen state is
cleared after BARs of PCI device included in the PE are restored
and enabled. However, we needn't clear the frozen state for PHB PE
explicitly at this point as there is no real PE for PHB PE. As the
PHB PE is always binding with PE#0, we actually clear PE#0, which
is wrong. It doesn't incur any problem though.
This checks if the PE is PHB PE and doesn't clear the frozen state
if it is.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
To retrieve the PCI slot state, EEH driver would set a timeout for that.
While current comment is not aligned to what the code does.
This patch fixes those comments according to the code.
Signed-off-by: Wei Yang <weiyang@linux.vnet.ibm.com>
Acked-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The patch removes struct eeh_dev::dn and the corresponding helper
functions: eeh_dev_to_of_node() and of_node_to_eeh_dev(). Instead,
eeh_dev_to_pdn() and pdn_to_eeh_dev() should be used to get the
pdn, which might contain device_node on PowerNV platform.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
When PE's frozen count hits maximal allowed frozen times, which is
5 currently, it will be forced to be offline permanently. Once the
PE is removed permanently, rebooting machine is required to bring
the PE back. It's not convienent when testing EEH functionality.
The patch exports the maximal allowed frozen times through debugfs
entry (/sys/kernel/debug/powerpc/eeh_max_freezes).
Requested-by: Ryan Grimm <grimm@linux.vnet.ibm.com>
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The conditions that one specific PE's frozen count exceeds the maximal
allowed times (EEH_MAX_ALLOWED_FREEZES) and it's in isolated or recovery
state indicate the PE was removed permanently implicitly. The patch
introduces flag EEH_PE_REMOVED to indicate that explicitly so that we
don't depend on the fixed maximal allowed times, which can be varied as
we do in subsequent patch.
Flag EEH_PE_REMOVED is expected to be marked for the PE whose frozen
count exceeds the maximal allowed times, or just failed from recovery.
Requested-by: Ryan Grimm <grimm@linux.vnet.ibm.com>
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The patch introduces additional flag EEH_PE_RESET to indicate the
corresponding PE is under reset. In turn, the PE retrieval bakcend
on PowerNV platform can return unfrozen state for the EEH core to
moving forward. Flag EEH_PE_CFG_BLOCKED isn't the correct one for
the purpose.
In PCI passthrou case, the problem is more worse: Guest doesn't
recover 6th EEH error. The PE is left in isolated (frozen) and
config blocked state on Broadcom adapters. We can't retrieve the
PE's state correctly any more, even from the host side via sysfs
/sys/bus/pci/devices/xxx/eeh_pe_state.
Reported-by: Rajeshkumar Subramanian <rajeshkumars@in.ibm.com>
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
The flag EEH_PE_RESET indicates blocking config space of the PE
during reset time. We potentially need block PE's config space
other than reset time. So it's reasonable to replace it with
EEH_PE_CFG_BLOCKED to indicate its usage.
There are no substantial code or logic changes in this patch.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When enabling EEH functionality on passed through devices (PE)
with VFIO, the devices in the PE would be removed permanently
from guest side. In that case, the PE remains frozen state.
When returning PE to host, or restarting the guest again, we
had mechanism unfreezing the PE by clearing PESTA/B frozen
bits. However, that's not enough for some adapters, which are
indicated as following "lspci" shows. Those adapters require
hot reset on the parent bus to bring their firmware back to
workable state. Otherwise, those adaptrs won't be operative
and the host (for returning case) or the guest will fail to
load the drivers for those adapters without exception.
0000:01:00.0 Ethernet controller: Emulex Corporation OneConnect \
10Gb NIC (be3) (rev 02)
0000:01:00.0 0200: 19a2:0710 (rev 02)
0001:03:00.0 Ethernet controller: Emulex Corporation OneConnect \
NIC (Lancer) (rev 10)
0001:03:00.0 0200: 10df:e220 (rev 10)
The patch adds mechanism to emulate EEH recovery (for hot reset
on parent PCI bus) on 3 gates to fix the issue: open/release one
adapter of the PE, enable EEH functionality on one adapter of the
PE.
Reported-by: Murilo Fossa Vicentini <muvic@br.ibm.com>
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The patch uses eeh_unfreeze_pe() to replace the logic clearing
frozen IO and DMA, in order to simplify the code.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
pr_warn() is equal to pr_warning(), but the former is a bit more
formal according to commit fc62f2f ("kernel.h: add pr_warn for
symmetry to dev_warn, netdev_warn").
The patch replaces pr_warning() with pr_warn().
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
On PowerNV platform, EEH errors are reported by IO accessors or poller
driven by interrupt. After the PE is isolated, we won't produce EEH
event for the PE. The current implementation has possibility of EEH
event lost in this way:
The interrupt handler queues one "special" event, which drives the poller.
EEH thread doesn't pick the special event yet. IO accessors kicks in, the
frozen PE is marked as "isolated" and EEH event is queued to the list.
EEH thread runs because of special event and purge all existing EEH events.
However, we never produce an other EEH event for the frozen PE. Eventually,
the PE is marked as "isolated" and we don't have EEH event to recover it.
The patch fixes the issue to keep EEH events for PEs that have been
marked as "isolated" with the help of additional "force" help to
eeh_remove_event().
Reported-by: Rolf Brudeseth <rolfb@us.ibm.com>
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Since commit cb523e09 ("powerpc/eeh: Avoid I/O access during PE
reset"), the PE is kept as frozen state on hardware level until
the PE reset is done completely. After that, we explicitly clear
the frozen state of the affected PE. However, there might have
frozen child PEs of the affected PE and we also need clear their
frozen state as well. Otherwise, the recovery is going to fail.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
When PCI_ERS_RESULT_CAN_RECOVER returned from device drivers, the
EEH core should enable I/O and DMA for the affected PE. However,
it was missed to have DMA enabled in eeh_handle_normal_event().
Besides, the frozen state of the affected PE should be cleared
after successful recovery, but we didn't.
The patch fixes both of the issues as above.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
The issue was detected in a bit complicated test case where
we have multiple hierarchical PEs shown as following figure:
+-----------------+
| PE#3 p2p#0 |
| p2p#1 |
+-----------------+
|
+-----------------+
| PE#4 pdev#0 |
| pdev#1 |
+-----------------+
PE#4 (have 2 PCI devices) is the child of PE#3, which has 2 p2p
bridges. We accidentally had less-known scenario: PE#4 was removed
permanently from the system because of permanent failure (e.g.
exceeding the max allowd failure times in last hour), then we detects
EEH errors on PE#3 and tried to recover it. However, eeh_dev instances
for pdev#0/1 were not detached from PE#4, which was still connected to
PE#3. All of that was because of the fact that we rely on count-based
pcibios_release_device(), which isn't reliable enough. When doing
recovery for PE#3, we still apply hotplug on PE#4 and pdev#0/1, which
are not valid any more. Eventually, we run into kernel crash.
The patch fixes above issue from two aspects. For unplug, we simply
skip those permanently removed PE, whose state is (EEH_PE_STATE_ISOLATED
&& !EEH_PE_STATE_RECOVERING) and its frozen count should be greater
than EEH_MAX_ALLOWED_FREEZES. For plug, we marked all permanently
removed EEH devices with EEH_DEV_REMOVED and return 0xFF's on read
its PCI config so that PCI core will omit them.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
We have suffered recrusive frozen PE a lot, which was caused
by IO accesses during the PE reset. Ben came up with the good
idea to keep frozen PE until recovery (BAR restore) gets done.
With that, IO accesses during PE reset are dropped by hardware
and wouldn't incur the recrusive frozen PE any more.
The patch implements the idea. We don't clear the frozen state
until PE reset is done completely. During the period, the EEH
core expects unfrozen state from backend to keep going. So we
have to reuse EEH_PE_RESET flag, which has been set during PE
reset, to return normal state from backend. The side effect is
we have to clear frozen state for towice (PE reset and clear it
explicitly), but that's harmless.
We have some limitations on pHyp. pHyp doesn't allow to enable
IO or DMA for unfrozen PE. So we don't enable them on unfrozen PE
in eeh_pci_enable(). We have to enable IO before grabbing logs on
pHyp. Otherwise, 0xFF's is always returned from PCI config space.
Also, we had wrong return value from eeh_pci_enable() for
EEH_OPT_THAW_DMA case. The patch fixes it too.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
We've observed multiple PE reset failures because of PCI-CFG
access during that period. Potentially, some device drivers
can't support EEH very well and they can't put the device to
motionless state before PE reset. So those device drivers might
produce PCI-CFG accesses during PE reset. Also, we could have
PCI-CFG access from user space (e.g. "lspci"). Since access to
frozen PE should return 0xFF's, we can block PCI-CFG access
during the period of PE reset so that we won't get recrusive EEH
errors.
The patch adds flag EEH_PE_RESET, which is kept during PE reset.
The PowerNV/pSeries PCI-CFG accessors reuse the flag to block
PCI-CFG accordingly.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
The PE state (for eeh_pe instance) EEH_PE_PHB_DEAD is duplicate to
EEH_PE_ISOLATED. Originally, those PHBs (PHB PE) with EEH_PE_PHB_DEAD
would be removed from the system. However, it's safe to replace
that with EEH_PE_ISOLATED.
The patch also clear EEH_PE_RECOVERING after fenced PHB has been handled,
either failure or success. It makes the PHB PE state consistent with:
PHB functions normally NONE
PHB has been removed EEH_PE_ISOLATED
PHB fenced, recovery in progress EEH_PE_ISOLATED | RECOVERING
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Commit b8a9a11b9 (powerpc: eeh: Kill another abuse of irq_desc) is
missing some brackets .....
It's not a good idea to write patches in grumpy mode and then forget
to at least compile test them or rely on the few eyeballs discussing
that patch to spot it.....
Reported-by: fengguang.wu@intel.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: ppc <linuxppc-dev@lists.ozlabs.org>
commit 91150af3a (powerpc/eeh: Fix unbalanced enable for IRQ) is
another brilliant example of trainwreck engineering.
The patch "fixes" the issue of an unbalanced call to irq_enable()
which causes a prominent warning by checking the disabled state of the
interrupt line and call conditionally into the core code.
This is wrong in two aspects:
1) The warning is there to tell users, that they need to fix their
asymetric enable/disable patterns by finding the root cause and
solving it there.
It's definitely not meant to work around it by conditionally
calling into the core code depending on the random state of the irq
line.
Asymetric irq_disable/enable calls are a clear sign of wrong usage
of the interfaces which have to be cured at the root and not by
somehow hacking around it.
2) The abuse of core internal data structure instead of using the
proper interfaces for retrieving the information for the 'hack
around'
irq_desc is core internal and it's clear enough stated.
Replace at least the irq_desc abuse with the proper functions and add
a big fat comment why this is absurd and completely wrong.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: ppc <linuxppc-dev@lists.ozlabs.org>
Link: http://lkml.kernel.org/r/20140223212736.562906212@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Commit f5c57710dd ("powerpc/eeh: Use
partial hotplug for EEH unaware drivers") introduces eeh_rmv_device,
which may grab a reference to a driver, but not release it.
That prevents a driver from being removed after it has gone through EEH
recovery.
This patch drops the reference if it was taken.
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@linux.vnet.ibm.com>
Acked-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Pull powerpc updates from Ben Herrenschmidt:
"So here's my next branch for powerpc. A bit late as I was on vacation
last week. It's mostly the same stuff that was in next already, I
just added two patches today which are the wiring up of lockref for
powerpc, which for some reason fell through the cracks last time and
is trivial.
The highlights are, in addition to a bunch of bug fixes:
- Reworked Machine Check handling on kernels running without a
hypervisor (or acting as a hypervisor). Provides hooks to handle
some errors in real mode such as TLB errors, handle SLB errors,
etc...
- Support for retrieving memory error information from the service
processor on IBM servers running without a hypervisor and routing
them to the memory poison infrastructure.
- _PAGE_NUMA support on server processors
- 32-bit BookE relocatable kernel support
- FSL e6500 hardware tablewalk support
- A bunch of new/revived board support
- FSL e6500 deeper idle states and altivec powerdown support
You'll notice a generic mm change here, it has been acked by the
relevant authorities and is a pre-req for our _PAGE_NUMA support"
* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (121 commits)
powerpc: Implement arch_spin_is_locked() using arch_spin_value_unlocked()
powerpc: Add support for the optimised lockref implementation
powerpc/powernv: Call OPAL sync before kexec'ing
powerpc/eeh: Escalate error on non-existing PE
powerpc/eeh: Handle multiple EEH errors
powerpc: Fix transactional FP/VMX/VSX unavailable handlers
powerpc: Don't corrupt transactional state when using FP/VMX in kernel
powerpc: Reclaim two unused thread_info flag bits
powerpc: Fix races with irq_work
Move precessing of MCE queued event out from syscall exit path.
pseries/cpuidle: Remove redundant call to ppc64_runlatch_off() in cpu idle routines
powerpc: Make add_system_ram_resources() __init
powerpc: add SATA_MV to ppc64_defconfig
powerpc/powernv: Increase candidate fw image size
powerpc: Add debug checks to catch invalid cpu-to-node mappings
powerpc: Fix the setup of CPU-to-Node mappings during CPU online
powerpc/iommu: Don't detach device without IOMMU group
powerpc/eeh: Hotplug improvement
powerpc/eeh: Call opal_pci_reinit() on powernv for restoring config space
powerpc/eeh: Add restore_config operation
...
Race conditions are theoretically possible between the PCI device addition
and removal in the PPC64 PCI error recovery driver and the generic PCI bus
rescan and device removal that can be triggered via sysfs.
To avoid those race conditions make PPC64 PCI error recovery driver use
global PCI rescan-remove locking around PCI device addition and removal.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
For one PCI error relevant OPAL event, we possibly have multiple
EEH errors for that. For example, multiple frozen PEs detected on
different PHBs. Unfortunately, we didn't cover the case. The patch
enumarates the return value from eeh_ops::next_error() and change
eeh_handle_special_event() and eeh_ops::next_error() to handle all
existing EEH errors.
As Ben pointed out, we needn't list_for_each_entry_safe() since we
are not deleting any PHB from the hose_list and the EEH serialized
lock should be held while purging EEH events. The patch covers those
suggestions as well.
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
When EEH error comes to one specific PCI device before its driver
is loaded, we will apply hotplug to recover the error. During the
plug time, the PCI device will be probed and its driver is loaded.
Then we wrongly calls to the error handlers if the driver supports
EEH explicitly.
The patch intends to fix by introducing flag EEH_DEV_NO_HANDLER and
set it before we remove the PCI device. In turn, we can avoid wrongly
calls the error handlers of the PCI device after its driver loaded.
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
In order to support concurrent adapter firmware download
to SR-IOV adapters on pSeries, each VF will see an EEH event
where the slot will remain in the unavailable state for
the duration of the adapter firmware update, which can take
as long as 5 minutes. Extend the EEH recovery timeout to
account for this.
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Acked-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
When EEH error happens to one specific PE, some devices with drivers
supporting EEH won't except hotplug on the device. However, there
might have other deivces without driver, or with driver without EEH
support. For the case, we need do partial hotplug in order to make
sure that the PE becomes absolutely quite during reset. Otherise,
the PE reset might fail and leads to failure of error recovery.
The current code doesn't handle that 'mixed' case properly, it either
uses the error callbacks to the drivers, or tries hotplug, but doesn't
handle a PE (EEH domain) composed of a combination of the two.
The patch intends to support so-called "partial" hotplug for EEH:
Before we do reset, we stop and remove those PCI devices without
EEH sensitive driver. The corresponding EEH devices are not detached
from its PE, but with special flag. After the reset is done, those
EEH devices with the special flag will be scanned one by one.
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
When we do normal hotplug, the PE (shadow EEH structure) shouldn't be
kept around.
However, we need to keep it if the hotplug an artifial one caused by
EEH errors recovery.
Since we remove EEH device through the PCI hook pcibios_release_device(),
the flag "purge_pe" passed to various functions is meaningless. So the patch
removes the meaningless flag and introduce new flag "EEH_PE_KEEP"
to save the PE while doing hotplug during EEH error recovery.
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
We needn't the the whole backtrace other than one-line message in
the error reporting interrupt handler. For errors triggered by
access PCI config space or MMIO, we replace "WARN(1, ...)" with
pr_err() and dump_stack(). The patch also adds more output messages
to indicate what EEH core is doing. Besides, some printk() are
replaced with pr_warning().
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
On PowerNV platform, the EEH event caused by interrupt won't have
binding PE. The patch enables EEH core to handle the special event.
To avoid the current logic we have, The eeh_handle_event() is renamed
to eeh_handle_normal_event(), and the eeh_handle_special_event() is
introduced. The function eeh_handle_event() dispatches to above two
functions according to the input parameter. Besides, new backend
"next_error" added to eeh_ops and it's expected to have following
return values:
4 - Dead IOC 3 - Dead PHB
2 - Fenced PHB 1 - Frozen PE
0 - No error found
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
We're not expecting that one specific PE got frozen for over 5
times in last hour. Otherwise, the PE will be removed from the
system upon newly coming EEH errors. The patch introduces time
stamp to trace the first error on specific PE in last hour and
function to update that accordingly. Besides, the time stamp
is recovered during PE hotplug path as we did for frozen count.
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Ellerman