Avoids confusion when printing Oops message like below
Faulting instruction address: 0xc00000000008bdb4
Oops: Kernel access of bad area, sig: 11 [#1]
LE PAGE_SIZE=64K MMU=Radix MMU=Hash SMP NR_CPUS=2048 NUMA PowerNV
This was because we never clear the MMU_FTR_HPTE_TABLE feature flag
even if we run with radix translation. It was discussed that we should
look at this feature flag as an indication of the capability to run
hash translation and we should not clear the flag even if we run in
radix translation. All the code paths check for radix_enabled() check and
if found true consider we are running with radix translation. Follow the
same sequence for finding the MMU translation string to be used in Oops
message.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Acked-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190711145814.17970-1-aneesh.kumar@linux.ibm.com
Use DEFINE_DEBUGFS_ATTRIBUTE rather than DEFINE_SIMPLE_ATTRIBUTE
for debugfs files.
Semantic patch information:
Rationale: DEFINE_SIMPLE_ATTRIBUTE + debugfs_create_file()
imposes some significant overhead as compared to
DEFINE_DEBUGFS_ATTRIBUTE + debugfs_create_file_unsafe().
Generated by: scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/1543498518-107601-1-git-send-email-yuehaibing@huawei.com
On powerpc, watchpoint match range is double-word granular. On a
watchpoint hit, DAR is set to the first byte of overlap between actual
access and watched range. And thus it's quite possible that DAR does
not point inside user specified range. Ex, say user creates a
watchpoint with address range 0x1004 to 0x1007. So hw would be
configured to watch from 0x1000 to 0x1007. If there is a 4 byte access
from 0x1002 to 0x1005, DAR will point to 0x1002 and thus interrupt
handler considers it as extraneous, but it's actually not, because
part of the access belongs to what user has asked.
Instead of blindly ignoring the exception, get actual address range by
analysing an instruction, and ignore only if actual range does not
overlap with user specified range.
Note: The behavior is unchanged for 8xx.
Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20191017093204.7511-5-ravi.bangoria@linux.ibm.com
ptrace_set_debugreg() does not consider new length while overwriting
the watchpoint. Fix that. ppc_set_hwdebug() aligns watchpoint address
to doubleword boundary but does not change the length. If address
range is crossing doubleword boundary and length is less then 8, we
will lose samples from second doubleword. So fix that as well.
Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20191017093204.7511-4-ravi.bangoria@linux.ibm.com
Watchpoint match range is always doubleword(8 bytes) aligned on
powerpc. If the given range is crossing doubleword boundary, we need
to increase the length such that next doubleword also get
covered. Ex,
address len = 6 bytes
|=========.
|------------v--|------v--------|
| | | | | | | | | | | | | | | | |
|---------------|---------------|
<---8 bytes--->
In such case, current code configures hw as:
start_addr = address & ~HW_BREAKPOINT_ALIGN
len = 8 bytes
And thus read/write in last 4 bytes of the given range is ignored.
Fix this by including next doubleword in the length.
Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20191017093204.7511-3-ravi.bangoria@linux.ibm.com
The issue was showing "Mitigation" message via sysfs whatever the
state of "RFI Flush", but it should show "Vulnerable" when it is
disabled.
If you have "L1D private" feature enabled and not "RFI Flush" you are
vulnerable to meltdown attacks.
"RFI Flush" is the key feature to mitigate the meltdown whatever the
"L1D private" state.
SEC_FTR_L1D_THREAD_PRIV is a feature for Power9 only.
So the message should be as the truth table shows:
CPU | L1D private | RFI Flush | sysfs
----|-------------|-----------|-------------------------------------
P9 | False | False | Vulnerable
P9 | False | True | Mitigation: RFI Flush
P9 | True | False | Vulnerable: L1D private per thread
P9 | True | True | Mitigation: RFI Flush, L1D private per thread
P8 | False | False | Vulnerable
P8 | False | True | Mitigation: RFI Flush
Output before this fix:
# cat /sys/devices/system/cpu/vulnerabilities/meltdown
Mitigation: RFI Flush, L1D private per thread
# echo 0 > /sys/kernel/debug/powerpc/rfi_flush
# cat /sys/devices/system/cpu/vulnerabilities/meltdown
Mitigation: L1D private per thread
Output after fix:
# cat /sys/devices/system/cpu/vulnerabilities/meltdown
Mitigation: RFI Flush, L1D private per thread
# echo 0 > /sys/kernel/debug/powerpc/rfi_flush
# cat /sys/devices/system/cpu/vulnerabilities/meltdown
Vulnerable: L1D private per thread
Signed-off-by: Gustavo L. F. Walbon <gwalbon@linux.ibm.com>
Signed-off-by: Mauro S. M. Rodrigues <maurosr@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190502210907.42375-1-gwalbon@linux.ibm.com
Merge the secureboot support, as well as the IMA changes needed to
support it.
From Nayna's cover letter:
In order to verify the OS kernel on PowerNV systems, secure boot
requires X.509 certificates trusted by the platform. These are
stored in secure variables controlled by OPAL, called OPAL secure
variables. In order to enable users to manage the keys, the secure
variables need to be exposed to userspace.
OPAL provides the runtime services for the kernel to be able to
access the secure variables. This patchset defines the kernel
interface for the OPAL APIs. These APIs are used by the hooks, which
load these variables to the keyring and expose them to the userspace
for reading/writing.
Overall, this patchset adds the following support:
* expose secure variables to the kernel via OPAL Runtime API interface
* expose secure variables to the userspace via kernel sysfs interface
* load kernel verification and revocation keys to .platform and
.blacklist keyring respectively.
The secure variables can be read/written using simple linux
utilities cat/hexdump.
For example:
Path to the secure variables is: /sys/firmware/secvar/vars
Each secure variable is listed as directory.
$ ls -l
total 0
drwxr-xr-x. 2 root root 0 Aug 20 21:20 db
drwxr-xr-x. 2 root root 0 Aug 20 21:20 KEK
drwxr-xr-x. 2 root root 0 Aug 20 21:20 PK
The attributes of each of the secure variables are (for example: PK):
$ ls -l
total 0
-r--r--r--. 1 root root 4096 Oct 1 15:10 data
-r--r--r--. 1 root root 65536 Oct 1 15:10 size
--w-------. 1 root root 4096 Oct 1 15:12 update
The "data" is used to read the existing variable value using
hexdump. The data is stored in ESL format. The "update" is used to
write a new value using cat. The update is to be submitted as AUTH
file.
PowerNV secure variables, which store the keys used for OS kernel
verification, are managed by the firmware. These secure variables need to
be accessed by the userspace for addition/deletion of the certificates.
This patch adds the sysfs interface to expose secure variables for PowerNV
secureboot. The users shall use this interface for manipulating
the keys stored in the secure variables.
Signed-off-by: Nayna Jain <nayna@linux.ibm.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Eric Richter <erichte@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/1573441836-3632-3-git-send-email-nayna@linux.ibm.com
The X.509 certificates trusted by the platform and required to secure
boot the OS kernel are wrapped in secure variables, which are
controlled by OPAL.
This patch adds firmware/kernel interface to read and write OPAL
secure variables based on the unique key.
This support can be enabled using CONFIG_OPAL_SECVAR.
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Nayna Jain <nayna@linux.ibm.com>
Signed-off-by: Eric Richter <erichte@linux.ibm.com>
[mpe: Make secvar_ops __ro_after_init, only build opal-secvar.c if PPC_SECURE_BOOT=y]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/1573441836-3632-2-git-send-email-nayna@linux.ibm.com
The arch specific kernel module policy rule requires kernel modules to
be signed, either as an IMA signature, stored as an xattr, or as an
appended signature. As a result, kernel modules appended signatures
could be enforced without "sig_enforce" being set or reflected in
/sys/module/module/parameters/sig_enforce. This patch sets
"sig_enforce".
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/1572492694-6520-10-git-send-email-zohar@linux.ibm.com
While secure boot permits only properly verified signed kernels to be
booted, trusted boot calculates the file hash of the kernel image and
stores the measurement prior to boot, that can be subsequently
compared against good known values via attestation services.
This patch reads the trusted boot state of a PowerNV system. The state
is used to conditionally enable additional measurement rules in the
IMA arch-specific policies.
Signed-off-by: Nayna Jain <nayna@linux.ibm.com>
Signed-off-by: Eric Richter <erichte@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/e9eeee6b-b9bf-1e41-2954-61dbd6fbfbcf@linux.ibm.com
PowerNV systems use a Linux-based bootloader, which rely on the IMA
subsystem to enforce different secure boot modes. Since the
verification policy may differ based on the secure boot mode of the
system, the policies must be defined at runtime.
This patch implements arch-specific support to define IMA policy rules
based on the runtime secure boot mode of the system.
This patch provides arch-specific IMA policies if PPC_SECURE_BOOT
config is enabled.
Signed-off-by: Nayna Jain <nayna@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/1572492694-6520-3-git-send-email-zohar@linux.ibm.com
Similar to commit 22e9c88d48
("powerpc/64: reuse PPC32 static inline flush_dcache_range()")
this patch converts the following ASM symbols to C:
flush_icache_range()
__flush_dcache_icache()
__flush_dcache_icache_phys()
This was done as we discovered a long-standing bug where the length of the
range was truncated due to using a 32 bit shift instead of a 64 bit one.
By converting these functions to C, it becomes easier to maintain.
flush_dcache_icache_phys() retains a critical assembler section as we must
ensure there are no memory accesses while the data MMU is disabled
(authored by Christophe Leroy). Since this has no external callers, it has
also been made static, allowing the compiler to inline it within
flush_dcache_icache_page().
Signed-off-by: Alastair D'Silva <alastair@d-silva.org>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
[mpe: Minor fixups, don't export __flush_dcache_icache()]
Link: https://lore.kernel.org/r/20191104023305.9581-5-alastair@au1.ibm.com
Bring powerpc in line with other architectures that support extending or
overriding the bootloader provided command line.
The current behaviour is most like CMDLINE_FROM_BOOTLOADER where the
bootloader command line is preferred but the kernel config can provide a
fallback so CMDLINE_FROM_BOOTLOADER is the default. CMDLINE_EXTEND can
be used to append the CMDLINE from the kernel config to the one provided
by the bootloader.
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190801225006.21952-1-chris.packham@alliedtelesis.co.nz
Otherwise the build fails because prom_init is calling symbols it's
not allowed to, eg:
Error: External symbol 'ftrace_likely_update' referenced from prom_init.c
make[3]: *** [arch/powerpc/kernel/Makefile:197: arch/powerpc/kernel/prom_init_check] Error 1
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20191106051129.7626-1-mpe@ellerman.id.au
"powerpc_security_features" is "unsigned long", i.e. 32-bit or 64-bit,
depending on the platform (PPC_FSL_BOOK3E or PPC_BOOK3S_64). Hence
casting its address to "u64 *", and calling debugfs_create_x64() is
wrong, and leaks 32-bit of nearby data to userspace on 32-bit platforms.
While all currently defined SEC_FTR_* security feature flags fit in
32-bit, they all have "ULL" suffixes to make them 64-bit constants.
Hence fix the leak by changing the type of "powerpc_security_features"
(and the parameter types of its accessors) to "u64". This also allows
to drop the cast.
Fixes: 398af57112 ("powerpc/security: Show powerpc_security_features in debugfs")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20191021142309.28105-1-geert+renesas@glider.be
Some PowerPC CPUs are vulnerable to L1TF to the same extent as to
Meltdown. It is also mitigated by flushing the L1D on privilege
transition.
Currently the sysfs gives a false negative on L1TF on CPUs that I
verified to be vulnerable, a Power9 Talos II Boston 004e 1202, PowerNV
T2P9D01.
Signed-off-by: Anthony Steinhauser <asteinhauser@google.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
[mpe: Just have cpu_show_l1tf() call cpu_show_meltdown() directly]
Link: https://lore.kernel.org/r/20191029190759.84821-1-asteinhauser@google.com
As part of the uapi we export a lot of PT_xx defines for each register
in struct pt_regs. These are expressed as an index from gpr[0], in
units of unsigned long.
Currently there's nothing tying the values of those defines to the
actual layout of the struct.
But we *don't* want to change the uapi defines to derive the PT_xx
values based on the layout of the struct, those values are ABI and
must never change.
Instead we want to do the reverse, make sure that the layout of the
struct never changes vs the PT_xx defines. So add build time checks of
that.
This probably seems paranoid, but at least once in the past someone
has sent a patch that would have broken the ABI if it hadn't been
spotted. Although it probably would have been detected via testing,
it's preferable to just quash any issues at the source.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20191030111231.22720-1-mpe@ellerman.id.au
`pt_regs_check` is a dummy function, its purpose is to break the build
if struct pt_regs and struct user_pt_regs don't match.
This function has no functionnal purpose, and will get eliminated at
link time or after init depending on CONFIG_LD_DEAD_CODE_DATA_ELIMINATION
This commit adds a prototype to fix warning at W=1:
arch/powerpc/kernel/ptrace.c:3339:13: error: no previous prototype for ‘pt_regs_check’ [-Werror=missing-prototypes]
Suggested-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Mathieu Malaterre <malat@debian.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20181208154624.6504-1-malat@debian.org
The ultravisor will do an integrity check of the kernel image but we
relocated it so the check will fail. Restore the original image by
relocating it back to the kernel virtual base address.
This works because during build vmlinux is linked with an expected
virtual runtime address of KERNELBASE.
Fixes: 6a9c930bd7 ("powerpc/prom_init: Add the ESM call to prom_init")
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Tested-by: Michael Anderson <andmike@linux.ibm.com>
[mpe: Add IS_ENABLED() to fix the CONFIG_RELOCATABLE=n build]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190911163433.12822-1-bauerman@linux.ibm.com
accumulate_stolen_time() is called prior to interrupt state being
reconciled, which can trip the warning in arch_local_irq_restore():
WARNING: CPU: 5 PID: 1017 at arch/powerpc/kernel/irq.c:258 .arch_local_irq_restore+0x9c/0x130
...
NIP .arch_local_irq_restore+0x9c/0x130
LR .rb_start_commit+0x38/0x80
Call Trace:
.ring_buffer_lock_reserve+0xe4/0x620
.trace_function+0x44/0x210
.function_trace_call+0x148/0x170
.ftrace_ops_no_ops+0x180/0x1d0
ftrace_call+0x4/0x8
.accumulate_stolen_time+0x1c/0xb0
decrementer_common+0x124/0x160
For now just mark it as notrace. We may change the ordering to call it
after interrupt state has been reconciled, but that is a larger
change.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20191024055932.27940-1-mpe@ellerman.id.au
At the beginning of setup_64.c, it has,
#ifdef DEBUG
#define DBG(fmt...) udbg_printf(fmt)
#else
#define DBG(fmt...)
#endif
where DBG() could be compiled away, and generate warnings,
arch/powerpc/kernel/setup_64.c: In function 'initialize_cache_info':
arch/powerpc/kernel/setup_64.c:579:49: warning: suggest braces around
empty body in an 'if' statement [-Wempty-body]
DBG("Argh, can't find dcache properties !\n");
^
arch/powerpc/kernel/setup_64.c:582:49: warning: suggest braces around
empty body in an 'if' statement [-Wempty-body]
DBG("Argh, can't find icache properties !\n");
Fix it by using the suggestions from Michael:
"Neither of those sites should use DBG(), that's not really early
boot code, they should just use pr_warn().
And the other uses of DBG() in initialize_cache_info() should just
be removed.
In smp_release_cpus() the entry/exit DBG's should just be removed,
and the spinning_secondaries line should just be pr_debug().
That would just leave the two calls in early_setup(). If we taught
udbg_printf() to return early when udbg_putc is NULL, then we could
just call udbg_printf() unconditionally and get rid of the DBG macro
entirely."
Suggested-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Qian Cai <cai@lca.pw>
[mpe: Split udbg change out into previous patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/1563215552-8166-1-git-send-email-cai@lca.pw
Make udbg_printf() check if udbg_putc is set, and if not just return.
This makes it safe to call udbg_printf() anytime, even when a udbg
backend has not been registered, which means we can avoid some ifdefs
at call sites.
Signed-off-by: Qian Cai <cai@lca.pw>
[mpe: Split out of larger patch, write change log]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
On POWER9, under some circumstances, a broadcast TLB invalidation will
fail to invalidate the ERAT cache on some threads when there are
parallel mtpidr/mtlpidr happening on other threads of the same core.
This can cause stores to continue to go to a page after it's unmapped.
The workaround is to force an ERAT flush using PID=0 or LPID=0 tlbie
flush. This additional TLB flush will cause the ERAT cache
invalidation. Since we are using PID=0 or LPID=0, we don't get
filtered out by the TLB snoop filtering logic.
We need to still follow this up with another tlbie to take care of
store vs tlbie ordering issue explained in commit:
a5d4b5891c ("powerpc/mm: Fixup tlbie vs store ordering issue on
POWER9"). The presence of ERAT cache implies we can still get new
stores and they may miss store queue marking flush.
Cc: stable@vger.kernel.org
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190924035254.24612-3-aneesh.kumar@linux.ibm.com
The store ordering vs tlbie issue mentioned in commit
a5d4b5891c ("powerpc/mm: Fixup tlbie vs store ordering issue on
POWER9") is fixed for Nimbus 2.3 and Cumulus 1.3 revisions. We don't
need to apply the fixup if we are running on them
We can only do this on PowerNV. On pseries guest with KVM we still
don't support redoing the feature fixup after migration. So we should
be enabling all the workarounds needed, because whe can possibly
migrate between DD 2.3 and DD 2.2
Fixes: a5d4b5891c ("powerpc/mm: Fixup tlbie vs store ordering issue on POWER9")
Cc: stable@vger.kernel.org # v4.16+
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190924035254.24612-1-aneesh.kumar@linux.ibm.com
On a 2-socket Power9 system with 32 cores/128 threads (SMT4) and 1TB
of memory running the following guest configs:
guest A:
- 224GB of memory
- 56 VCPUs (sockets=1,cores=28,threads=2), where:
VCPUs 0-1 are pinned to CPUs 0-3,
VCPUs 2-3 are pinned to CPUs 4-7,
...
VCPUs 54-55 are pinned to CPUs 108-111
guest B:
- 4GB of memory
- 4 VCPUs (sockets=1,cores=4,threads=1)
with the following workloads (with KSM and THP enabled in all):
guest A:
stress --cpu 40 --io 20 --vm 20 --vm-bytes 512M
guest B:
stress --cpu 4 --io 4 --vm 4 --vm-bytes 512M
host:
stress --cpu 4 --io 4 --vm 2 --vm-bytes 256M
the below soft-lockup traces were observed after an hour or so and
persisted until the host was reset (this was found to be reliably
reproducible for this configuration, for kernels 4.15, 4.18, 5.0,
and 5.3-rc5):
[ 1253.183290] rcu: INFO: rcu_sched self-detected stall on CPU
[ 1253.183319] rcu: 124-....: (5250 ticks this GP) idle=10a/1/0x4000000000000002 softirq=5408/5408 fqs=1941
[ 1256.287426] watchdog: BUG: soft lockup - CPU#105 stuck for 23s! [CPU 52/KVM:19709]
[ 1264.075773] watchdog: BUG: soft lockup - CPU#24 stuck for 23s! [worker:19913]
[ 1264.079769] watchdog: BUG: soft lockup - CPU#31 stuck for 23s! [worker:20331]
[ 1264.095770] watchdog: BUG: soft lockup - CPU#45 stuck for 23s! [worker:20338]
[ 1264.131773] watchdog: BUG: soft lockup - CPU#64 stuck for 23s! [avocado:19525]
[ 1280.408480] watchdog: BUG: soft lockup - CPU#124 stuck for 22s! [ksmd:791]
[ 1316.198012] rcu: INFO: rcu_sched self-detected stall on CPU
[ 1316.198032] rcu: 124-....: (21003 ticks this GP) idle=10a/1/0x4000000000000002 softirq=5408/5408 fqs=8243
[ 1340.411024] watchdog: BUG: soft lockup - CPU#124 stuck for 22s! [ksmd:791]
[ 1379.212609] rcu: INFO: rcu_sched self-detected stall on CPU
[ 1379.212629] rcu: 124-....: (36756 ticks this GP) idle=10a/1/0x4000000000000002 softirq=5408/5408 fqs=14714
[ 1404.413615] watchdog: BUG: soft lockup - CPU#124 stuck for 22s! [ksmd:791]
[ 1442.227095] rcu: INFO: rcu_sched self-detected stall on CPU
[ 1442.227115] rcu: 124-....: (52509 ticks this GP) idle=10a/1/0x4000000000000002 softirq=5408/5408 fqs=21403
[ 1455.111787] INFO: task worker:19907 blocked for more than 120 seconds.
[ 1455.111822] Tainted: G L 5.3.0-rc5-mdr-vanilla+ #1
[ 1455.111833] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1455.111884] INFO: task worker:19908 blocked for more than 120 seconds.
[ 1455.111905] Tainted: G L 5.3.0-rc5-mdr-vanilla+ #1
[ 1455.111925] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1455.111966] INFO: task worker:20328 blocked for more than 120 seconds.
[ 1455.111986] Tainted: G L 5.3.0-rc5-mdr-vanilla+ #1
[ 1455.111998] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1455.112048] INFO: task worker:20330 blocked for more than 120 seconds.
[ 1455.112068] Tainted: G L 5.3.0-rc5-mdr-vanilla+ #1
[ 1455.112097] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1455.112138] INFO: task worker:20332 blocked for more than 120 seconds.
[ 1455.112159] Tainted: G L 5.3.0-rc5-mdr-vanilla+ #1
[ 1455.112179] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1455.112210] INFO: task worker:20333 blocked for more than 120 seconds.
[ 1455.112231] Tainted: G L 5.3.0-rc5-mdr-vanilla+ #1
[ 1455.112242] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1455.112282] INFO: task worker:20335 blocked for more than 120 seconds.
[ 1455.112303] Tainted: G L 5.3.0-rc5-mdr-vanilla+ #1
[ 1455.112332] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1455.112372] INFO: task worker:20336 blocked for more than 120 seconds.
[ 1455.112392] Tainted: G L 5.3.0-rc5-mdr-vanilla+ #1
CPUs 45, 24, and 124 are stuck on spin locks, likely held by
CPUs 105 and 31.
CPUs 105 and 31 are stuck in smp_call_function_many(), waiting on
target CPU 42. For instance:
# CPU 105 registers (via xmon)
R00 = c00000000020b20c R16 = 00007d1bcd800000
R01 = c00000363eaa7970 R17 = 0000000000000001
R02 = c0000000019b3a00 R18 = 000000000000006b
R03 = 000000000000002a R19 = 00007d537d7aecf0
R04 = 000000000000002a R20 = 60000000000000e0
R05 = 000000000000002a R21 = 0801000000000080
R06 = c0002073fb0caa08 R22 = 0000000000000d60
R07 = c0000000019ddd78 R23 = 0000000000000001
R08 = 000000000000002a R24 = c00000000147a700
R09 = 0000000000000001 R25 = c0002073fb0ca908
R10 = c000008ffeb4e660 R26 = 0000000000000000
R11 = c0002073fb0ca900 R27 = c0000000019e2464
R12 = c000000000050790 R28 = c0000000000812b0
R13 = c000207fff623e00 R29 = c0002073fb0ca808
R14 = 00007d1bbee00000 R30 = c0002073fb0ca800
R15 = 00007d1bcd600000 R31 = 0000000000000800
pc = c00000000020b260 smp_call_function_many+0x3d0/0x460
cfar= c00000000020b270 smp_call_function_many+0x3e0/0x460
lr = c00000000020b20c smp_call_function_many+0x37c/0x460
msr = 900000010288b033 cr = 44024824
ctr = c000000000050790 xer = 0000000000000000 trap = 100
CPU 42 is running normally, doing VCPU work:
# CPU 42 stack trace (via xmon)
[link register ] c00800001be17188 kvmppc_book3s_radix_page_fault+0x90/0x2b0 [kvm_hv]
[c000008ed3343820] c000008ed3343850 (unreliable)
[c000008ed33438d0] c00800001be11b6c kvmppc_book3s_hv_page_fault+0x264/0xe30 [kvm_hv]
[c000008ed33439d0] c00800001be0d7b4 kvmppc_vcpu_run_hv+0x8dc/0xb50 [kvm_hv]
[c000008ed3343ae0] c00800001c10891c kvmppc_vcpu_run+0x34/0x48 [kvm]
[c000008ed3343b00] c00800001c10475c kvm_arch_vcpu_ioctl_run+0x244/0x420 [kvm]
[c000008ed3343b90] c00800001c0f5a78 kvm_vcpu_ioctl+0x470/0x7c8 [kvm]
[c000008ed3343d00] c000000000475450 do_vfs_ioctl+0xe0/0xc70
[c000008ed3343db0] c0000000004760e4 ksys_ioctl+0x104/0x120
[c000008ed3343e00] c000000000476128 sys_ioctl+0x28/0x80
[c000008ed3343e20] c00000000000b388 system_call+0x5c/0x70
--- Exception: c00 (System Call) at 00007d545cfd7694
SP (7d53ff7edf50) is in userspace
It was subsequently found that ipi_message[PPC_MSG_CALL_FUNCTION]
was set for CPU 42 by at least 1 of the CPUs waiting in
smp_call_function_many(), but somehow the corresponding
call_single_queue entries were never processed by CPU 42, causing the
callers to spin in csd_lock_wait() indefinitely.
Nick Piggin suggested something similar to the following sequence as
a possible explanation (interleaving of CALL_FUNCTION/RESCHEDULE
IPI messages seems to be most common, but any mix of CALL_FUNCTION and
!CALL_FUNCTION messages could trigger it):
CPU
X: smp_muxed_ipi_set_message():
X: smp_mb()
X: message[RESCHEDULE] = 1
X: doorbell_global_ipi(42):
X: kvmppc_set_host_ipi(42, 1)
X: ppc_msgsnd_sync()/smp_mb()
X: ppc_msgsnd() -> 42
42: doorbell_exception(): // from CPU X
42: ppc_msgsync()
105: smp_muxed_ipi_set_message():
105: smb_mb()
// STORE DEFERRED DUE TO RE-ORDERING
--105: message[CALL_FUNCTION] = 1
| 105: doorbell_global_ipi(42):
| 105: kvmppc_set_host_ipi(42, 1)
| 42: kvmppc_set_host_ipi(42, 0)
| 42: smp_ipi_demux_relaxed()
| 42: // returns to executing guest
| // RE-ORDERED STORE COMPLETES
->105: message[CALL_FUNCTION] = 1
105: ppc_msgsnd_sync()/smp_mb()
105: ppc_msgsnd() -> 42
42: local_paca->kvm_hstate.host_ipi == 0 // IPI ignored
105: // hangs waiting on 42 to process messages/call_single_queue
This can be prevented with an smp_mb() at the beginning of
kvmppc_set_host_ipi(), such that stores to message[<type>] (or other
state indicated by the host_ipi flag) are ordered vs. the store to
to host_ipi.
However, doing so might still allow for the following scenario (not
yet observed):
CPU
X: smp_muxed_ipi_set_message():
X: smp_mb()
X: message[RESCHEDULE] = 1
X: doorbell_global_ipi(42):
X: kvmppc_set_host_ipi(42, 1)
X: ppc_msgsnd_sync()/smp_mb()
X: ppc_msgsnd() -> 42
42: doorbell_exception(): // from CPU X
42: ppc_msgsync()
// STORE DEFERRED DUE TO RE-ORDERING
-- 42: kvmppc_set_host_ipi(42, 0)
| 42: smp_ipi_demux_relaxed()
| 105: smp_muxed_ipi_set_message():
| 105: smb_mb()
| 105: message[CALL_FUNCTION] = 1
| 105: doorbell_global_ipi(42):
| 105: kvmppc_set_host_ipi(42, 1)
| // RE-ORDERED STORE COMPLETES
-> 42: kvmppc_set_host_ipi(42, 0)
42: // returns to executing guest
105: ppc_msgsnd_sync()/smp_mb()
105: ppc_msgsnd() -> 42
42: local_paca->kvm_hstate.host_ipi == 0 // IPI ignored
105: // hangs waiting on 42 to process messages/call_single_queue
Fixing this scenario would require an smp_mb() *after* clearing
host_ipi flag in kvmppc_set_host_ipi() to order the store vs.
subsequent processing of IPI messages.
To handle both cases, this patch splits kvmppc_set_host_ipi() into
separate set/clear functions, where we execute smp_mb() prior to
setting host_ipi flag, and after clearing host_ipi flag. These
functions pair with each other to synchronize the sender and receiver
sides.
With that change in place the above workload ran for 20 hours without
triggering any lock-ups.
Fixes: 755563bc79 ("powerpc/powernv: Fixes for hypervisor doorbell handling") # v4.0
Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Acked-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190911223155.16045-1-mdroth@linux.vnet.ibm.com
Currently the reserved bits of the Processor Compatibility
Register (PCR) are cleared as per the Programming Note in Section
1.3.3 of version 3.0B of the Power ISA. This causes all new
architecture features to be made available when running on newer
processors with new architecture features added to the PCR as bits
must be set to disable a given feature.
For example to disable new features added as part of Version 2.07 of
the ISA the corresponding bit in the PCR needs to be set.
As new processor features generally require explicit kernel support
they should be disabled until such support is implemented. Therefore
kernels should set all unknown/reserved bits in the PCR such that any
new architecture features which the kernel does not currently know
about get disabled.
An update is planned to the ISA to clarify that the PCR is an
exception to the Programming Note on reserved bits in Section 1.3.3.
Signed-off-by: Alistair Popple <alistair@popple.id.au>
Signed-off-by: Jordan Niethe <jniethe5@gmail.com>
Tested-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190917004605.22471-2-alistair@popple.id.au
Pull powerpc updates from Michael Ellerman:
"This is a bit late, partly due to me travelling, and partly due to a
power outage knocking out some of my test systems *while* I was
travelling.
- Initial support for running on a system with an Ultravisor, which
is software that runs below the hypervisor and protects guests
against some attacks by the hypervisor.
- Support for building the kernel to run as a "Secure Virtual
Machine", ie. as a guest capable of running on a system with an
Ultravisor.
- Some changes to our DMA code on bare metal, to allow devices with
medium sized DMA masks (> 32 && < 59 bits) to use more than 2GB of
DMA space.
- Support for firmware assisted crash dumps on bare metal (powernv).
- Two series fixing bugs in and refactoring our PCI EEH code.
- A large series refactoring our exception entry code to use gas
macros, both to make it more readable and also enable some future
optimisations.
As well as many cleanups and other minor features & fixups.
Thanks to: Adam Zerella, Alexey Kardashevskiy, Alistair Popple, Andrew
Donnellan, Aneesh Kumar K.V, Anju T Sudhakar, Anshuman Khandual,
Balbir Singh, Benjamin Herrenschmidt, Cédric Le Goater, Christophe
JAILLET, Christophe Leroy, Christopher M. Riedl, Christoph Hellwig,
Claudio Carvalho, Daniel Axtens, David Gibson, David Hildenbrand,
Desnes A. Nunes do Rosario, Ganesh Goudar, Gautham R. Shenoy, Greg
Kurz, Guerney Hunt, Gustavo Romero, Halil Pasic, Hari Bathini, Joakim
Tjernlund, Jonathan Neuschafer, Jordan Niethe, Leonardo Bras, Lianbo
Jiang, Madhavan Srinivasan, Mahesh Salgaonkar, Mahesh Salgaonkar,
Masahiro Yamada, Maxiwell S. Garcia, Michael Anderson, Nathan
Chancellor, Nathan Lynch, Naveen N. Rao, Nicholas Piggin, Oliver
O'Halloran, Qian Cai, Ram Pai, Ravi Bangoria, Reza Arbab, Ryan Grimm,
Sam Bobroff, Santosh Sivaraj, Segher Boessenkool, Sukadev Bhattiprolu,
Thiago Bauermann, Thiago Jung Bauermann, Thomas Gleixner, Tom
Lendacky, Vasant Hegde"
* tag 'powerpc-5.4-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: (264 commits)
powerpc/mm/mce: Keep irqs disabled during lockless page table walk
powerpc: Use ftrace_graph_ret_addr() when unwinding
powerpc/ftrace: Enable HAVE_FUNCTION_GRAPH_RET_ADDR_PTR
ftrace: Look up the address of return_to_handler() using helpers
powerpc: dump kernel log before carrying out fadump or kdump
docs: powerpc: Add missing documentation reference
powerpc/xmon: Fix output of XIVE IPI
powerpc/xmon: Improve output of XIVE interrupts
powerpc/mm/radix: remove useless kernel messages
powerpc/fadump: support holes in kernel boot memory area
powerpc/fadump: remove RMA_START and RMA_END macros
powerpc/fadump: update documentation about option to release opalcore
powerpc/fadump: consider f/w load area
powerpc/opalcore: provide an option to invalidate /sys/firmware/opal/core file
powerpc/opalcore: export /sys/firmware/opal/core for analysing opal crashes
powerpc/fadump: update documentation about CONFIG_PRESERVE_FA_DUMP
powerpc/fadump: add support to preserve crash data on FADUMP disabled kernel
powerpc/fadump: improve how crashed kernel's memory is reserved
powerpc/fadump: consider reserved ranges while releasing memory
powerpc/fadump: make crash memory ranges array allocation generic
...
