Load-balancing improvements:
============================
- Improve NUMA balancing on AMD Zen systems for affine workloads.
- Improve the handling of reduced-capacity CPUs in load-balancing.
- Energy Model improvements: fix & refine all the energy fairness metrics (PELT),
and remove the conservative threshold requiring 6% energy savings to
migrate a task. Doing this improves power efficiency for most workloads,
and also increases the reliability of energy-efficiency scheduling.
- Optimize/tweak select_idle_cpu() to spend (much) less time searching
for an idle CPU on overloaded systems. There's reports of several
milliseconds spent there on large systems with large workloads ...
[ Since the search logic changed, there might be behavioral side effects. ]
- Improve NUMA imbalance behavior. On certain systems
with spare capacity, initial placement of tasks is non-deterministic,
and such an artificial placement imbalance can persist for a long time,
hurting (and sometimes helping) performance.
The fix is to make fork-time task placement consistent with runtime
NUMA balancing placement.
Note that some performance regressions were reported against this,
caused by workloads that are not memory bandwith limited, which benefit
from the artificial locality of the placement bug(s). Mel Gorman's
conclusion, with which we concur, was that consistency is better than
random workload benefits from non-deterministic bugs:
"Given there is no crystal ball and it's a tradeoff, I think it's
better to be consistent and use similar logic at both fork time
and runtime even if it doesn't have universal benefit."
- Improve core scheduling by fixing a bug in sched_core_update_cookie() that
caused unnecessary forced idling.
- Improve wakeup-balancing by allowing same-LLC wakeup of idle CPUs for newly
woken tasks.
- Fix a newidle balancing bug that introduced unnecessary wakeup latencies.
ABI improvements/fixes:
=======================
- Do not check capabilities and do not issue capability check denial messages
when a scheduler syscall doesn't require privileges. (Such as increasing niceness.)
- Add forced-idle accounting to cgroups too.
- Fix/improve the RSEQ ABI to not just silently accept unknown flags.
(No existing tooling is known to have learned to rely on the previous behavior.)
- Depreciate the (unused) RSEQ_CS_FLAG_NO_RESTART_ON_* flags.
Optimizations:
==============
- Optimize & simplify leaf_cfs_rq_list()
- Micro-optimize set_nr_{and_not,if}_polling() via try_cmpxchg().
Misc fixes & cleanups:
======================
- Fix the RSEQ self-tests on RISC-V and Glibc 2.35 systems.
- Fix a full-NOHZ bug that can in some cases result in the tick not being
re-enabled when the last SCHED_RT task is gone from a runqueue but there's
still SCHED_OTHER tasks around.
- Various PREEMPT_RT related fixes.
- Misc cleanups & smaller fixes.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
-----BEGIN PGP SIGNATURE-----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=xr2b
-----END PGP SIGNATURE-----
Merge tag 'sched-core-2022-08-01' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler updates from Ingo Molnar:
"Load-balancing improvements:
- Improve NUMA balancing on AMD Zen systems for affine workloads.
- Improve the handling of reduced-capacity CPUs in load-balancing.
- Energy Model improvements: fix & refine all the energy fairness
metrics (PELT), and remove the conservative threshold requiring 6%
energy savings to migrate a task. Doing this improves power
efficiency for most workloads, and also increases the reliability
of energy-efficiency scheduling.
- Optimize/tweak select_idle_cpu() to spend (much) less time
searching for an idle CPU on overloaded systems. There's reports of
several milliseconds spent there on large systems with large
workloads ...
[ Since the search logic changed, there might be behavioral side
effects. ]
- Improve NUMA imbalance behavior. On certain systems with spare
capacity, initial placement of tasks is non-deterministic, and such
an artificial placement imbalance can persist for a long time,
hurting (and sometimes helping) performance.
The fix is to make fork-time task placement consistent with runtime
NUMA balancing placement.
Note that some performance regressions were reported against this,
caused by workloads that are not memory bandwith limited, which
benefit from the artificial locality of the placement bug(s). Mel
Gorman's conclusion, with which we concur, was that consistency is
better than random workload benefits from non-deterministic bugs:
"Given there is no crystal ball and it's a tradeoff, I think
it's better to be consistent and use similar logic at both fork
time and runtime even if it doesn't have universal benefit."
- Improve core scheduling by fixing a bug in
sched_core_update_cookie() that caused unnecessary forced idling.
- Improve wakeup-balancing by allowing same-LLC wakeup of idle CPUs
for newly woken tasks.
- Fix a newidle balancing bug that introduced unnecessary wakeup
latencies.
ABI improvements/fixes:
- Do not check capabilities and do not issue capability check denial
messages when a scheduler syscall doesn't require privileges. (Such
as increasing niceness.)
- Add forced-idle accounting to cgroups too.
- Fix/improve the RSEQ ABI to not just silently accept unknown flags.
(No existing tooling is known to have learned to rely on the
previous behavior.)
- Depreciate the (unused) RSEQ_CS_FLAG_NO_RESTART_ON_* flags.
Optimizations:
- Optimize & simplify leaf_cfs_rq_list()
- Micro-optimize set_nr_{and_not,if}_polling() via try_cmpxchg().
Misc fixes & cleanups:
- Fix the RSEQ self-tests on RISC-V and Glibc 2.35 systems.
- Fix a full-NOHZ bug that can in some cases result in the tick not
being re-enabled when the last SCHED_RT task is gone from a
runqueue but there's still SCHED_OTHER tasks around.
- Various PREEMPT_RT related fixes.
- Misc cleanups & smaller fixes"
* tag 'sched-core-2022-08-01' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (32 commits)
rseq: Kill process when unknown flags are encountered in ABI structures
rseq: Deprecate RSEQ_CS_FLAG_NO_RESTART_ON_* flags
sched/core: Fix the bug that task won't enqueue into core tree when update cookie
nohz/full, sched/rt: Fix missed tick-reenabling bug in dequeue_task_rt()
sched/core: Always flush pending blk_plug
sched/fair: fix case with reduced capacity CPU
sched/core: Use try_cmpxchg in set_nr_{and_not,if}_polling
sched/core: add forced idle accounting for cgroups
sched/fair: Remove the energy margin in feec()
sched/fair: Remove task_util from effective utilization in feec()
sched/fair: Use the same cpumask per-PD throughout find_energy_efficient_cpu()
sched/fair: Rename select_idle_mask to select_rq_mask
sched, drivers: Remove max param from effective_cpu_util()/sched_cpu_util()
sched/fair: Decay task PELT values during wakeup migration
sched/fair: Provide u64 read for 32-bits arch helper
sched/fair: Introduce SIS_UTIL to search idle CPU based on sum of util_avg
sched: only perform capability check on privileged operation
sched: Remove unused function group_first_cpu()
sched/fair: Remove redundant word " *"
selftests/rseq: check if libc rseq support is registered
...
kvm_hypercall has to place the hypercall number in rax.
Trace events show that kvm_pv_test doesn't work properly:
kvm_pv_test-53132: kvm_hypercall: nr 0x0 a0 0x0 a1 0x0 a2 0x0 a3 0x0
kvm_pv_test-53132: kvm_hypercall: nr 0x0 a0 0x0 a1 0x0 a2 0x0 a3 0x0
kvm_pv_test-53132: kvm_hypercall: nr 0x0 a0 0x0 a1 0x0 a2 0x0 a3 0x0
With this change, it starts working as expected:
kvm_pv_test-54285: kvm_hypercall: nr 0x5 a0 0x0 a1 0x0 a2 0x0 a3 0x0
kvm_pv_test-54285: kvm_hypercall: nr 0xa a0 0x0 a1 0x0 a2 0x0 a3 0x0
kvm_pv_test-54285: kvm_hypercall: nr 0xb a0 0x0 a1 0x0 a2 0x0 a3 0x0
Signed-off-by: Andrei Vagin <avagin@google.com>
Message-Id: <20220722230241.1944655-5-avagin@google.com>
Fixes: ac4a4d6de2 ("selftests: kvm: test enforcement of paravirtual cpuid features")
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
GCC 12 thinks that `actual` might be used uninitialised. It's not, the
use is guarded by `bad_mmcr2` which is only set to true at the same
point where `actual` is initialised.
cycles_with_mmcr2_test.c: In function ‘cycles_with_mmcr2’:
cycles_with_mmcr2_test.c:81:17: error: ‘actual’ may be used uninitialized [-Werror=maybe-uninitialized]
81 | printf("Bad MMCR2 value seen is 0x%lx\n", actual);
Silence the warning by initialising `actual` to zero.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220801113746.802046-1-mpe@ellerman.id.au
The only sensible exponent for a boolean stat is 0. Add a test assertion
requiring all boolean statistics to have an exponent of 0.
Signed-off-by: Oliver Upton <oupton@google.com>
Reviewed-by: Andrew Jones <andrew.jones@linux.dev>
Message-Id: <20220719143134.3246798-4-oliver.upton@linux.dev>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
As it turns out, tests sometimes fail. When that is the case, packing
the test assertion with as much relevant information helps track down
the problem more quickly.
Sharpen up the stat descriptor assertions in kvm_binary_stats_test to
more precisely describe the reason for the test assertion and which
stat is to blame.
Signed-off-by: Oliver Upton <oupton@google.com>
Reviewed-by: Andrew Jones <andrew.jones@linux.dev>
Message-Id: <20220719143134.3246798-3-oliver.upton@linux.dev>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
In order to provide more useful test assertions that describe the broken
stats descriptor, perform sanity check on the stat name before any other
descriptor field. While at it, avoid dereferencing the name field if the
sanity check fails as it is more likely to contain garbage.
Signed-off-by: Oliver Upton <oupton@google.com>
Reviewed-by: Andrew Jones <andrew.jones@linux.dev>
Message-Id: <20220719143134.3246798-2-oliver.upton@linux.dev>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM/arm64 updates for 5.20:
- Unwinder implementations for both nVHE modes (classic and
protected), complete with an overflow stack
- Rework of the sysreg access from userspace, with a complete
rewrite of the vgic-v3 view to allign with the rest of the
infrastructure
- Disagregation of the vcpu flags in separate sets to better track
their use model.
- A fix for the GICv2-on-v3 selftest
- A small set of cosmetic fixes
KVM/s390, KVM/x86 and common infrastructure changes for 5.20
x86:
* Permit guests to ignore single-bit ECC errors
* Fix races in gfn->pfn cache refresh; do not pin pages tracked by the cache
* Intel IPI virtualization
* Allow getting/setting pending triple fault with KVM_GET/SET_VCPU_EVENTS
* PEBS virtualization
* Simplify PMU emulation by just using PERF_TYPE_RAW events
* More accurate event reinjection on SVM (avoid retrying instructions)
* Allow getting/setting the state of the speaker port data bit
* Refuse starting the kvm-intel module if VM-Entry/VM-Exit controls are inconsistent
* "Notify" VM exit (detect microarchitectural hangs) for Intel
* Cleanups for MCE MSR emulation
s390:
* add an interface to provide a hypervisor dump for secure guests
* improve selftests to use TAP interface
* enable interpretive execution of zPCI instructions (for PCI passthrough)
* First part of deferred teardown
* CPU Topology
* PV attestation
* Minor fixes
Generic:
* new selftests API using struct kvm_vcpu instead of a (vm, id) tuple
x86:
* Use try_cmpxchg64 instead of cmpxchg64
* Bugfixes
* Ignore benign host accesses to PMU MSRs when PMU is disabled
* Allow disabling KVM's "MONITOR/MWAIT are NOPs!" behavior
* x86/MMU: Allow NX huge pages to be disabled on a per-vm basis
* Port eager page splitting to shadow MMU as well
* Enable CMCI capability by default and handle injected UCNA errors
* Expose pid of vcpu threads in debugfs
* x2AVIC support for AMD
* cleanup PIO emulation
* Fixes for LLDT/LTR emulation
* Don't require refcounted "struct page" to create huge SPTEs
x86 cleanups:
* Use separate namespaces for guest PTEs and shadow PTEs bitmasks
* PIO emulation
* Reorganize rmap API, mostly around rmap destruction
* Do not workaround very old KVM bugs for L0 that runs with nesting enabled
* new selftests API for CPUID
The VERBOSE option in Makefile has been moved, but there still have the
description left in README. For now, we use `-v` options when running
memblock test to print information, so using the new to replace the
obsolete items.
Signed-off-by: Shaoqin Huang <shaoqin.huang@intel.com>
Acked-by: Rebecca Mckeever <remckee0@gmail.com>
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Link: https://lore.kernel.org/r/20220729161125.190845-1-shaoqin.huang@intel.com
Andrii Nakryiko says:
====================
bpf-next 2022-07-29
We've added 22 non-merge commits during the last 4 day(s) which contain
a total of 27 files changed, 763 insertions(+), 120 deletions(-).
The main changes are:
1) Fixes to allow setting any source IP with bpf_skb_set_tunnel_key() helper,
from Paul Chaignon.
2) Fix for bpf_xdp_pointer() helper when doing sanity checking, from Joanne Koong.
3) Fix for XDP frame length calculation, from Lorenzo Bianconi.
4) Libbpf BPF_KSYSCALL docs improvements and fixes to selftests to accommodate
s390x quirks with socketcall(), from Ilya Leoshkevich.
5) Allow/denylist and CI configs additions to selftests/bpf to improve BPF CI,
from Daniel Müller.
6) BPF trampoline + ftrace follow up fixes, from Song Liu and Xu Kuohai.
7) Fix allocation warnings in netdevsim, from Jakub Kicinski.
8) bpf_obj_get_opts() libbpf API allowing to provide file flags, from Joe Burton.
9) vsnprintf usage fix in bpf_snprintf_btf(), from Fedor Tokarev.
10) Various small fixes and clean ups, from Daniel Müller, Rongguang Wei,
Jörn-Thorben Hinz, Yang Li.
* https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (22 commits)
bpf: Remove unneeded semicolon
libbpf: Add bpf_obj_get_opts()
netdevsim: Avoid allocation warnings triggered from user space
bpf: Fix NULL pointer dereference when registering bpf trampoline
bpf: Fix test_progs -j error with fentry/fexit tests
selftests/bpf: Bump internal send_signal/send_signal_tracepoint timeout
bpftool: Don't try to return value from void function in skeleton
bpftool: Replace sizeof(arr)/sizeof(arr[0]) with ARRAY_SIZE macro
bpf: btf: Fix vsnprintf return value check
libbpf: Support PPC in arch_specific_syscall_pfx
selftests/bpf: Adjust vmtest.sh to use local kernel configuration
selftests/bpf: Copy over libbpf configs
selftests/bpf: Sort configuration
selftests/bpf: Attach to socketcall() in test_probe_user
libbpf: Extend BPF_KSYSCALL documentation
bpf, devmap: Compute proper xdp_frame len redirecting frames
bpf: Fix bpf_xdp_pointer return pointer
selftests/bpf: Don't assign outer source IP to host
bpf: Set flow flag to allow any source IP in bpf_tunnel_key
geneve: Use ip_tunnel_key flow flags in route lookups
...
====================
Link: https://lore.kernel.org/r/20220729230948.1313527-1-andrii@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add a simple test case for when hmm_range_fault() is called with the
HMM_PFN_REQ_FAULT flag and a device private PTE is found for a device
other than the hmm_range::dev_private_owner. This should cause the page
to be faulted back to system memory from the other device and the PFN
returned in the output array.
Also, remove a piece of code that unnecessarily unmaps part of the buffer.
Link: https://lkml.kernel.org/r/20220727000837.4128709-3-rcampbell@nvidia.com
Link: https://lkml.kernel.org/r/20220725183615.4118795-3-rcampbell@nvidia.com
Signed-off-by: Ralph Campbell <rcampbell@nvidia.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Cc: Felix Kuehling <felix.kuehling@amd.com>
Cc: Philip Yang <Philip.Yang@amd.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20220725142048.30450-4-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Nadav Amit <nadav.amit@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add two soft-dirty test cases for mprotect() on both anon or file.
Link: https://lkml.kernel.org/r/20220725142048.30450-3-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Nadav Amit <nadav.amit@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Initialize "length" to zero by default.
Link: https://lkml.kernel.org/r/YtZzjvHXVXMXxpXO@kili
Fixes: ff712a627f ("selftests/vm: cleanup hugetlb file after mremap test")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Mina Almasry <almasrymina@google.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
This code just reads from memory without caring about the data itself.
However static checkers complain that "tmp" is never properly initialized.
Initialize it to zero and change the name to "dummy" to show that we
don't care about the value stored in it.
Link: https://lkml.kernel.org/r/YtZ8mKJmktA2GaHB@kili
Fixes: c4b6cb8840 ("selftests/vm: add hugetlb madvise MADV_DONTNEED MADV_REMOVE test")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Souptick Joarder (HPE) <jrdr.linux@gmail.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The test va_128TBswitch.c exercises a feature only supported on PPC and
x86_64, but it's run on other 64-bit archs as well. Before this patch,
the test did nothing and returned 0 for KSFT_PASS. This patch makes it
return the KSFT codes from kselftest.h, including KSFT_SKIP when
appropriate.
Verified on arm64 and x86_64.
Link: https://lkml.kernel.org/r/20220704123813.427625-1-adam@wowsignal.io
Signed-off-by: Adam Sindelar <adam@wowsignal.io>
Cc: David Vernet <void@manifault.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
mrelease_test should return KSFT_SKIP when process_mrelease is not
defined, but due to a perror call consuming the errno, it returns
KSFT_FAIL.
This patch decides the exit code before calling perror.
[adam@wowsignal.io: fix remaining instances of errno mishandling]
Link: https://lkml.kernel.org/r/20220706141602.10159-1-adam@wowsignal.io
Link: https://lkml.kernel.org/r/20220704173351.19595-1-adam@wowsignal.io
Fixes: 33776141b8 ("selftests: vm: add process_mrelease tests")
Signed-off-by: Adam Sindelar <adam@wowsignal.io>
Reviewed-by: David Vernet <void@manifault.com>
Reviewed-by: Suren Baghdasaryan <surenb@google.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The send_signal/send_signal_tracepoint is pretty flaky, with at least
one failure in every ten runs on a few attempts I've tried it:
> test_send_signal_common:PASS:pipe_c2p 0 nsec
> test_send_signal_common:PASS:pipe_p2c 0 nsec
> test_send_signal_common:PASS:fork 0 nsec
> test_send_signal_common:PASS:skel_open_and_load 0 nsec
> test_send_signal_common:PASS:skel_attach 0 nsec
> test_send_signal_common:PASS:pipe_read 0 nsec
> test_send_signal_common:PASS:pipe_write 0 nsec
> test_send_signal_common:PASS:reading pipe 0 nsec
> test_send_signal_common:PASS:reading pipe error: size 0 0 nsec
> test_send_signal_common:FAIL:incorrect result unexpected incorrect result: actual 48 != expected 50
> test_send_signal_common:PASS:pipe_write 0 nsec
> #139/1 send_signal/send_signal_tracepoint:FAIL
The reason does not appear to be a correctness issue in the strict
sense. Rather, we merely do not receive the signal we are waiting for
within the provided timeout.
Let's bump the timeout by a factor of ten. With that change I have not
been able to reproduce the failure in 150+ iterations. I am also sneaking
in a small simplification to the test_progs test selection logic.
Signed-off-by: Daniel Müller <deso@posteo.net>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20220727182955.4044988-1-deso@posteo.net
Add IPv4 and IPv6 test cases that ensure that we are not leaking a
reference on the nexthop device when we are unable to delete its
associated route.
Without the fix in a previous patch ("netdevsim: fib: Fix reference
count leak on route deletion failure") both test cases get stuck,
waiting for the reference to be released from the dummy device [1][2].
[1]
unregister_netdevice: waiting for dummy1 to become free. Usage count = 5
leaked reference.
fib_check_nh+0x275/0x620
fib_create_info+0x237c/0x4d30
fib_table_insert+0x1dd/0x1d20
inet_rtm_newroute+0x11b/0x200
rtnetlink_rcv_msg+0x43b/0xd20
netlink_rcv_skb+0x15e/0x430
netlink_unicast+0x53b/0x800
netlink_sendmsg+0x945/0xe40
____sys_sendmsg+0x747/0x960
___sys_sendmsg+0x11d/0x190
__sys_sendmsg+0x118/0x1e0
do_syscall_64+0x34/0x80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
[2]
unregister_netdevice: waiting for dummy1 to become free. Usage count = 5
leaked reference.
fib6_nh_init+0xc46/0x1ca0
ip6_route_info_create+0x1167/0x19a0
ip6_route_add+0x27/0x150
inet6_rtm_newroute+0x161/0x170
rtnetlink_rcv_msg+0x43b/0xd20
netlink_rcv_skb+0x15e/0x430
netlink_unicast+0x53b/0x800
netlink_sendmsg+0x945/0xe40
____sys_sendmsg+0x747/0x960
___sys_sendmsg+0x11d/0x190
__sys_sendmsg+0x118/0x1e0
do_syscall_64+0x34/0x80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This selftest is designed for testing the H.L2Encaps.Red behavior. It
instantiates a virtual network composed of several nodes: hosts and SRv6
routers. Each node is realized using a network namespace that is
properly interconnected to others through veth pairs.
The test considers SRv6 routers implementing a L2 VPN leveraged by hosts
for communicating with each other. Such routers make use of the SRv6
H.L2Encaps.Red behavior for applying SRv6 policies to L2 traffic coming
from hosts.
The correct execution of the behavior is verified through reachability
tests carried out between hosts belonging to the same VPN.
Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Signed-off-by: David S. Miller <davem@davemloft.net>
This selftest is designed for testing the H.Encaps.Red behavior. It
instantiates a virtual network composed of several nodes: hosts and SRv6
routers. Each node is realized using a network namespace that is
properly interconnected to others through veth pairs.
The test considers SRv6 routers implementing L3 VPNs leveraged by hosts
for communicating with each other. Such routers make use of the SRv6
H.Encaps.Red behavior for applying SRv6 policies to L3 traffic coming
from hosts.
The correct execution of the behavior is verified through reachability
tests carried out between hosts belonging to the same VPN.
Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Signed-off-by: David S. Miller <davem@davemloft.net>
Do 'make -C tools/testing/memblock', get the following errors:
memblock.o: In function `memblock_find_in_range.constprop.9':
memblock.c:(.text+0x4651): undefined reference to `pr_warn_ratelimited'
memblock.o: In function `memblock_mark_mirror':
memblock.c:(.text+0x7171): undefined reference to `mirrored_kernelcore'
Fixes: 902c2d9158 ("memblock: Disable mirror feature if kernelcore is not specified")
Fixes: 14d9a675fd ("mm: Ratelimited mirrored memory related warning messages")
Signed-off-by: Liu Xinpeng <liuxp11@chinatelecom.cn>
Tested-by: Ma Wupeng <mawupeng1@huawei.com>
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Link: https://lore.kernel.org/r/1658916453-26312-1-git-send-email-liuxp11@chinatelecom.cn
Add a Makefile which takes care of installing the selftests in
tools/testing/selftests/drivers/net/dsa. This can be used to install all
DSA specific selftests and forwarding.config using the same approach as
for the selftests in tools/testing/selftests/net/forwarding.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20220727191642.480279-1-martin.blumenstingl@googlemail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add a handful of memory randomizations and precise length checks.
Nothing is really broken here, I did this to increase confidence
when debugging. It does fix a GCC warning, tho. Apparently GCC
recognizes that memory needs to be initialized for send() but
does not recognize that for write().
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Verify that KVM allows toggling VMX MSR bits to be "more" restrictive,
and also allows restoring each MSR to KVM's original, less restrictive
value.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220607213604.3346000-16-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add a command line option to dirty_log_perf_test to run vCPUs for the
entire duration of disabling dirty logging. By default, the test stops
running runs vCPUs before disabling dirty logging, which is faster but
less interesting as it doesn't stress KVM's handling of contention
between page faults and the zapping of collapsible SPTEs. Enabling the
flag also lets the user verify that KVM is indeed rebuilding zapped SPTEs
as huge pages by checking KVM's pages_{1g,2m,4k} stats. Without vCPUs to
fault in the zapped SPTEs, the stats will show that KVM is zapping pages,
but they never show whether or not KVM actually allows huge pages to be
recreated.
Note! Enabling the flag can _significantly_ increase runtime, especially
if the thread that's disabling dirty logging doesn't have a dedicated
pCPU, e.g. if all pCPUs are used to run vCPUs.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220715232107.3775620-5-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The ISA states: "when ACC[i] contains defined data, the contents of VSRs
4×i to 4×i+3 are undefined until either a VSX Move From ACC instruction
is used to copy the contents of ACC[i] to VSRs 4×i to 4×i+3 or some other
instruction directly writes to one of these VSRs." We aren't doing this.
This test only works on Power10 because the hardware implementation
happens to map ACC0 to VSRs 0-3, but will fail on any other implementation
that doesn't do this. So add xxmfacc between writing to the accumulator
and accessing the VSRs.
Fixes: 3527e1ab9a ("selftests/powerpc: Add matrix multiply assist (MMA) test")
Signed-off-by: Rashmica Gupta <rashmica@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220617043935.428083-1-rashmica@linux.ibm.com
clang has -Wconstant-conversion by default, and the constant 0xAAAAAAAAA
(9 As) being converted to an int, which is generally 32 bits, results
in the compile warning:
clang -Wl,-no-as-needed -Wall -isystem ../../../../usr/include/ -lpthread seccomp_bpf.c -lcap -o seccomp_bpf
seccomp_bpf.c:812:67: warning: implicit conversion from 'long' to 'int' changes value from 45812984490 to -1431655766 [-Wconstant-conversion]
int kill = kill_how == KILL_PROCESS ? SECCOMP_RET_KILL_PROCESS : 0xAAAAAAAAA;
~~~~ ^~~~~~~~~~~
1 warning generated.
-1431655766 is the expected truncation, 0xAAAAAAAA (8 As), so use
this directly in the code to avoid the warning.
Fixes: 3932fcecd9 ("selftests/seccomp: Add test for unknown SECCOMP_RET kill behavior")
Signed-off-by: YiFei Zhu <zhuyifei@google.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220526223407.1686936-1-zhuyifei@google.com
So far the vmtest.sh script, which can be used as a convenient way to
run bpf selftests, has obtained the kernel config safe to use for
testing from the libbpf/libbpf GitHub repository [0].
Given that we now have included this configuration into this very
repository, we can just consume it from here as well, eliminating the
necessity of remote accesses.
With this change we adjust the logic in the script to use the
configuration from below tools/testing/selftests/bpf/configs/ instead
of pulling it over the network.
[0] https://github.com/libbpf/libbpf
Signed-off-by: Daniel Müller <deso@posteo.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Mykola Lysenko <mykolal@fb.com>
Link: https://lore.kernel.org/bpf/20220727001156.3553701-4-deso@posteo.net
This change integrates libbpf maintained configurations and black/white
lists [0] into the repository, co-located with the BPF selftests themselves.
We minimize the kernel configurations to keep future updates as small as
possible [1].
Furthermore, we make both kernel configurations build on top of the existing
configuration tools/testing/selftests/bpf/config (to be concatenated before
build). Lastly, we replaced the terms blacklist & whitelist with denylist and
allowlist, respectively.
[0] 20f0330235/travis-ci/vmtest/configs
[1] https://lore.kernel.org/bpf/20220712212124.3180314-1-deso@posteo.net/T/#m30a53648352ed494e556ac003042a9ad0a8f98c6
Signed-off-by: Daniel Müller <deso@posteo.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Mykola Lysenko <mykolal@fb.com>
Link: https://lore.kernel.org/bpf/20220727001156.3553701-3-deso@posteo.net
This change makes sure to sort the existing minimal kernel configuration
containing options required for running BPF selftests alphabetically.
Doing so will make it easier to diff it against other configurations,
which in turn helps with maintaining disjunct config files that build on
top of each other. It also helped identify the CONFIG_IPV6_GRE being set
twice and removes one of the occurrences.
Lastly, we change NET_CLS_BPF from 'm' to 'y'. Having this option as 'm'
will cause failures of the btf_skc_cls_ingress selftest.
Signed-off-by: Daniel Müller <deso@posteo.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Mykola Lysenko <mykolal@fb.com>
Link: https://lore.kernel.org/bpf/20220727001156.3553701-2-deso@posteo.net
Restore the +x bit to va_128TBswitch.sh, which got dropped from the
previous patch, somehow.
Link: https://lkml.kernel.org/r/20220708090646.34927-1-adam@wowsignal.io
Fixes: 1afd01d43efc3 ("selftests/vm: Only run 128TBswitch with 5-level paging")
Signed-off-by: Adam Sindelar <adam@wowsignal.io>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Once line card is activated, check the FW version and PSID are exposed.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Once line card is provisioned, check if HW revision and INI version
are exposed on associated nested auxiliary device.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
test_probe_user fails on architectures where libc uses
socketcall(SYS_CONNECT) instead of connect(). Fix by attaching
to socketcall as well.
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20220726134008.256968-3-iii@linux.ibm.com
The previous commit fixed a bug in the bpf_skb_set_tunnel_key helper to
avoid dropping packets whose outer source IP address isn't assigned to a
host interface. This commit changes the corresponding selftest to not
assign the outer source IP address to an interface.
Not assigning the source IP to an interface causes two issues in the
existing test:
1. The ARP requests will fail for that IP address so we need to add the
ARP entry manually.
2. The encapsulated ICMP echo reply traffic will not reach the VXLAN
device. It will be dropped by the stack before, because the
outer destination IP is unknown.
To solve 2., we have two choices. Either we perform decapsulation
ourselves in a BPF program attached at veth1 (the base device for the
VXLAN device), or we switch the outer destination address when we
receive the packet at veth1, such that the stack properly demultiplexes
it to the VXLAN device afterward.
This commit implements the second approach, where we switch the outer
destination address from the unassigned IP address to the assigned one,
only for VXLAN traffic ingressing veth1.
Then, at the vxlan device, the BPF program that checks the output of
bpf_skb_get_tunnel_key needs to be updated as the expected local IP
address is now the unassigned one.
Signed-off-by: Paul Chaignon <paul@isovalent.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/4addde76eaf3477a58975bef15ed2788c44e5f55.1658759380.git.paul@isovalent.com
After all the soft validation of the region has completed, convey the
region configuration to hardware while being careful to commit decoders
in specification mandated order. In addition to programming the endpoint
decoder base-address, interleave ways and granularity, the switch
decoder target lists are also established.
While the kernel can enforce spec-mandated commit order, it can not
enforce spec-mandated reset order. For example, the kernel can't stop
someone from removing an endpoint device that is occupying decoderN in a
switch decoder where decoderN+1 is also committed. To reset decoderN,
decoderN+1 must be torn down first. That "tear down the world"
implementation is saved for a follow-on patch.
Callback operations are provided for the 'commit' and 'reset'
operations. While those callbacks may prove useful for CXL accelerators
(Type-2 devices with memory) the primary motivation is to enable a
simple way for cxl_test to intercept those operations.
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/165784338418.1758207.14659830845389904356.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Currently the ptrace-gpr test only tests the GET/SET(FP)REGS ptrace
APIs. But there's an alternate (older) API, called PEEK/POKEUSR.
Add some minimal testing of PEEK/POKEUSR of the FPRs. This is sufficient
to detect the bug that was fixed recently in the 32-bit ptrace FPR
handling.
Depends-on: 8e12784444 ("powerpc/32: Fix overread/overwrite of thread_struct via ptrace")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220627140239.2464900-13-mpe@ellerman.id.au
The ptrace-gpr test uses fixed values to test that registers can be
read/written via ptrace. In particular it sets all GPRs to 1, which
means the test could miss some types of bugs - eg. if the kernel was
only returning the low word.
So generate some random values at startup and use those instead.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220627140239.2464900-12-mpe@ellerman.id.au
The ptrace-gpr test includes some inline asm to load GPR and FPR
registers. It then goes back to C to wait for the parent to trace it and
then checks register contents.
The split between inline asm and C is fragile, it relies on the compiler
not using any non-volatile GPRs after the inline asm block. It also
requires a very large and unwieldy inline asm block.
So convert the logic to set registers, wait, and store registers to a
single asm function, meaning there's no window for the compiler to
intervene.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220627140239.2464900-10-mpe@ellerman.id.au
Some of the ptrace tests check the contents of floating pointer
registers. Currently these use float, which is always 4 bytes, but the
ptrace API supports saving/restoring 8 bytes per register, so switch to
using doubles to exercise the code more fully.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220627140239.2464900-8-mpe@ellerman.id.au
Thare are some asm helpers for creating/popping stack frames in
basic_asm.h. They always save/restore r2 (TOC pointer), but none of the
selftests change r2, so it's unnecessary to save it by default.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220627140239.2464900-5-mpe@ellerman.id.au
Thare are some asm helpers for creating/popping stack frames in
basic_asm.h. They always save/restore CR, but none of the selftests
tests touch non-volatile CR fields, so it's unnecessary to save them by
default.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220627140239.2464900-4-mpe@ellerman.id.au
Currently all ptrace tests are built 64-bit and with TM enabled.
Only the TM tests need TM enabled, so split those out into a separate
variable so that can be specified precisely.
Split the rest of the tests into a variable, and add -m64 to CFLAGS for
those tests, so that in a subsequent patch some tests can be made to
build 32-bit.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220627140239.2464900-3-mpe@ellerman.id.au
