Commit 85d86c8a authored by Oliver O'Halloran's avatar Oliver O'Halloran Committed by Michael Ellerman

selftests/powerpc: Add basic EEH selftest

Use the new eeh_dev_check and eeh_dev_break interfaces to test EEH
recovery.  Historically this has been done manually using platform specific
EEH error injection facilities (e.g. via RTAS). However, documentation on
how to use these facilities is haphazard at best and non-existent at worst
so it's hard to develop a cross-platform test.

The new debugfs interfaces allow the kernel to handle the platform specific
details so we can write a more generic set of sets. This patch adds the
most basic of recovery tests where:

a) Errors are injected and recovered from sequentially,
b) Errors are not injected into PCI-PCI bridges, such as PCIe switches.
c) Errors are only injected into device function zero.
d) No errors are injected into Virtual Functions.

a), b) and c) are largely due to limitations of Linux's EEH support.  EEH
recovery is serialised in the EEH recovery thread which forces a).
Similarly, multi-function PCI devices are almost always grouped into the
same PE so injecting an error on one function exercises the same code
paths. c) is because we currently more or less ignore PCI bridges during
recovery and assume that the recovered topology will be the same as the
original.

d) is due to the limits of the eeh_dev_break interface. With the current
implementation we can't inject an error into a specific VF without
potentially causing additional errors on other VFs. Due to the serialised
recovery process we might end up timing out waiting for another function to
recover before the function of interest is recovered. The platform specific
error injection facilities are finer-grained and allow this capability, but
doing that requires working out how to use those facilities first.

Basicly, it's better than nothing and it's a base to build on.
Signed-off-by: default avatarOliver O'Halloran <oohall@gmail.com>
Signed-off-by: default avatarMichael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190903101605.2890-15-oohall@gmail.com
parent bd6461cc
...@@ -26,6 +26,7 @@ SUB_DIRS = alignment \ ...@@ -26,6 +26,7 @@ SUB_DIRS = alignment \
switch_endian \ switch_endian \
syscalls \ syscalls \
tm \ tm \
eeh \
vphn \ vphn \
math \ math \
ptrace \ ptrace \
......
# SPDX-License-Identifier: GPL-2.0
noarg:
$(MAKE) -C ../
TEST_PROGS := eeh-basic.sh
TEST_FILES := eeh-functions.sh
top_srcdir = ../../../../..
include ../../lib.mk
#!/bin/sh
# SPDX-License-Identifier: GPL-2.0-only
. ./eeh-functions.sh
if ! eeh_supported ; then
echo "EEH not supported on this system, skipping"
exit 0;
fi
if [ ! -e "/sys/kernel/debug/powerpc/eeh_dev_check" ] && \
[ ! -e "/sys/kernel/debug/powerpc/eeh_dev_break" ] ; then
echo "debugfs EEH testing files are missing. Is debugfs mounted?"
exit 1;
fi
pre_lspci=`mktemp`
lspci > $pre_lspci
# Bump the max freeze count to something absurd so we don't
# trip over it while breaking things.
echo 5000 > /sys/kernel/debug/powerpc/eeh_max_freezes
# record the devices that we break in here. Assuming everything
# goes to plan we should get them back once the recover process
# is finished.
devices=""
# Build up a list of candidate devices.
for dev in `ls -1 /sys/bus/pci/devices/ | grep '\.0$'` ; do
# skip bridges since we can't recover them (yet...)
if [ -e "/sys/bus/pci/devices/$dev/pci_bus" ] ; then
echo "$dev, Skipped: bridge"
continue;
fi
# Skip VFs for now since we don't have a reliable way
# to break them.
if [ -e "/sys/bus/pci/devices/$dev/physfn" ] ; then
echo "$dev, Skipped: virtfn"
continue;
fi
# Don't inject errosr into an already-frozen PE. This happens with
# PEs that contain multiple PCI devices (e.g. multi-function cards)
# and injecting new errors during the recovery process will probably
# result in the recovery failing and the device being marked as
# failed.
if ! pe_ok $dev ; then
echo "$dev, Skipped: Bad initial PE state"
continue;
fi
echo "$dev, Added"
# Add to this list of device to check
devices="$devices $dev"
done
dev_count="$(echo $devices | wc -w)"
echo "Found ${dev_count} breakable devices..."
failed=0
for dev in $devices ; do
echo "Breaking $dev..."
if ! pe_ok $dev ; then
echo "Skipping $dev, Initial PE state is not ok"
failed="$((failed + 1))"
continue;
fi
if ! eeh_one_dev $dev ; then
failed="$((failed + 1))"
fi
done
echo "$failed devices failed to recover ($dev_count tested)"
lspci | diff -u $pre_lspci -
rm -f $pre_lspci
exit $failed
#!/bin/sh
# SPDX-License-Identifier: GPL-2.0-only
pe_ok() {
local dev="$1"
local path="/sys/bus/pci/devices/$dev/eeh_pe_state"
if ! [ -e "$path" ] ; then
return 1;
fi
local fw_state="$(cut -d' ' -f1 < $path)"
local sw_state="$(cut -d' ' -f2 < $path)"
# If EEH_PE_ISOLATED or EEH_PE_RECOVERING are set then the PE is in an
# error state or being recovered. Either way, not ok.
if [ "$((sw_state & 0x3))" -ne 0 ] ; then
return 1
fi
# A functioning PE should have the EEH_STATE_MMIO_ACTIVE and
# EEH_STATE_DMA_ACTIVE flags set. For some goddamn stupid reason
# the platform backends set these when the PE is in reset. The
# RECOVERING check above should stop any false positives though.
if [ "$((fw_state & 0x18))" -ne "$((0x18))" ] ; then
return 1
fi
return 0;
}
eeh_supported() {
test -e /proc/powerpc/eeh && \
grep -q 'EEH Subsystem is enabled' /proc/powerpc/eeh
}
eeh_one_dev() {
local dev="$1"
# Using this function from the command line is sometimes useful for
# testing so check that the argument is a well-formed sysfs device
# name.
if ! test -e /sys/bus/pci/devices/$dev/ ; then
echo "Error: '$dev' must be a sysfs device name (DDDD:BB:DD.F)"
return 1;
fi
# Break it
echo $dev >/sys/kernel/debug/powerpc/eeh_dev_break
# Force an EEH device check. If the kernel has already
# noticed the EEH (due to a driver poll or whatever), this
# is a no-op.
echo $dev >/sys/kernel/debug/powerpc/eeh_dev_check
# Enforce a 30s timeout for recovery. Even the IPR, which is infamously
# slow to reset, should recover within 30s.
max_wait=30
for i in `seq 0 ${max_wait}` ; do
if pe_ok $dev ; then
break;
fi
echo "$dev, waited $i/${max_wait}"
sleep 1
done
if ! pe_ok $dev ; then
echo "$dev, Failed to recover!"
return 1;
fi
echo "$dev, Recovered after $i seconds"
return 0;
}
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment