commit 734ad0af3609464f8f93e00b6c0de1e112f44559 upstream.
If an interrupt occurs in queued_spin_lock_slowpath() after we increment
qnodesp->count and before node->lock is initialized, another CPU might
see stale lock values in get_tail_qnode(). If the stale lock value happens
to match the lock on that CPU, then we write to the "next" pointer of
the wrong qnode. This causes a deadlock as the former CPU, once it becomes
the head of the MCS queue, will spin indefinitely until it's "next" pointer
is set by its successor in the queue.
Running stress-ng on a 16 core (16EC/16VP) shared LPAR, results in
occasional lockups similar to the following:
$ stress-ng --all 128 --vm-bytes 80% --aggressive \
--maximize --oomable --verify --syslog \
--metrics --times --timeout 5m
watchdog: CPU 15 Hard LOCKUP
......
NIP [c0000000000b78f4] queued_spin_lock_slowpath+0x1184/0x1490
LR [c000000001037c5c] _raw_spin_lock+0x6c/0x90
Call Trace:
0xc000002cfffa3bf0 (unreliable)
_raw_spin_lock+0x6c/0x90
raw_spin_rq_lock_nested.part.135+0x4c/0xd0
sched_ttwu_pending+0x60/0x1f0
__flush_smp_call_function_queue+0x1dc/0x670
smp_ipi_demux_relaxed+0xa4/0x100
xive_muxed_ipi_action+0x20/0x40
__handle_irq_event_percpu+0x80/0x240
handle_irq_event_percpu+0x2c/0x80
handle_percpu_irq+0x84/0xd0
generic_handle_irq+0x54/0x80
__do_irq+0xac/0x210
__do_IRQ+0x74/0xd0
0x0
do_IRQ+0x8c/0x170
hardware_interrupt_common_virt+0x29c/0x2a0
--- interrupt: 500 at queued_spin_lock_slowpath+0x4b8/0x1490
......
NIP [c0000000000b6c28] queued_spin_lock_slowpath+0x4b8/0x1490
LR [c000000001037c5c] _raw_spin_lock+0x6c/0x90
--- interrupt: 500
0xc0000029c1a41d00 (unreliable)
_raw_spin_lock+0x6c/0x90
futex_wake+0x100/0x260
do_futex+0x21c/0x2a0
sys_futex+0x98/0x270
system_call_exception+0x14c/0x2f0
system_call_vectored_common+0x15c/0x2ec
The following code flow illustrates how the deadlock occurs.
For the sake of brevity, assume that both locks (A and B) are
contended and we call the queued_spin_lock_slowpath() function.
CPU0 CPU1
---- ----
spin_lock_irqsave(A) |
spin_unlock_irqrestore(A) |
spin_lock(B) |
| |
▼ |
id = qnodesp->count++; |
(Note that nodes[0].lock == A) |
| |
▼ |
Interrupt |
(happens before "nodes[0].lock = B") |
| |
▼ |
spin_lock_irqsave(A) |
| |
▼ |
id = qnodesp->count++ |
nodes[1].lock = A |
| |
▼ |
Tail of MCS queue |
| spin_lock_irqsave(A)
▼ |
Head of MCS queue ▼
| CPU0 is previous tail
▼ |
Spin indefinitely ▼
(until "nodes[1].next != NULL") prev = get_tail_qnode(A, CPU0)
|
▼
prev == &qnodes[CPU0].nodes[0]
(as qnodes[CPU0].nodes[0].lock == A)
|
▼
WRITE_ONCE(prev->next, node)
|
▼
Spin indefinitely
(until nodes[0].locked == 1)
Thanks to Saket Kumar Bhaskar for help with recreating the issue
Fixes: 84990b1695 ("powerpc/qspinlock: add mcs queueing for contended waiters")
Cc: stable@vger.kernel.org # v6.2+
Reported-by: Geetika Moolchandani <geetika@linux.ibm.com>
Reported-by: Vaishnavi Bhat <vaish123@in.ibm.com>
Reported-by: Jijo Varghese <vargjijo@in.ibm.com>
Signed-off-by: Nysal Jan K.A. <nysal@linux.ibm.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20240829022830.1164355-1-nysal@linux.ibm.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 35f20786c481d5ced9283ff42de5c69b65e5ed13 ]
arch/powerpc/lib/xor_vmx.o is built with '-msoft-float' (from the main
powerpc Makefile) and '-maltivec' (from its CFLAGS), which causes an
error when building with clang after a recent change in main:
error: option '-msoft-float' cannot be specified with '-maltivec'
make[6]: *** [scripts/Makefile.build:243: arch/powerpc/lib/xor_vmx.o] Error 1
Explicitly add '-mhard-float' before '-maltivec' in xor_vmx.o's CFLAGS
to override the previous inclusion of '-msoft-float' (as the last option
wins), which matches how other areas of the kernel use '-maltivec', such
as AMDGPU.
Cc: stable@vger.kernel.org
Closes: https://github.com/ClangBuiltLinux/linux/issues/1986
Link: 4792f912b2
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20240127-ppc-xor_vmx-drop-msoft-float-v1-1-f24140e81376@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8f9abaa6d7de0a70fc68acaedce290c1f96e2e59 ]
Some of the fp/vmx code in sstep.c assume a certain maximum size for the
instructions being emulated. The size of those operations however is
determined separately in analyse_instr().
Add a check to validate the assumption on the maximum size of the
operations, so as to prevent any unintended kernel stack corruption.
Signed-off-by: Naveen N Rao <naveen@kernel.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Build-tested-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20231123071705.397625-1-naveen@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1b1e38002648819c04773647d5242990e2824264 ]
crtsavres.o is linked to modules. However, as explained in commit
d0e628cd81 ("kbuild: doc: clarify the difference between extra-y
and always-y"), 'make modules' does not build extra-y.
For example, the following command fails:
$ make ARCH=powerpc LLVM=1 KBUILD_MODPOST_WARN=1 mrproper ps3_defconfig modules
[snip]
LD [M] arch/powerpc/platforms/cell/spufs/spufs.ko
ld.lld: error: cannot open arch/powerpc/lib/crtsavres.o: No such file or directory
make[3]: *** [scripts/Makefile.modfinal:56: arch/powerpc/platforms/cell/spufs/spufs.ko] Error 1
make[2]: *** [Makefile:1844: modules] Error 2
make[1]: *** [/home/masahiro/workspace/linux-kbuild/Makefile:350: __build_one_by_one] Error 2
make: *** [Makefile:234: __sub-make] Error 2
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Fixes: baa25b571a ("powerpc/64: Do not link crtsavres.o in vmlinux")
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20231120232332.4100288-1-masahiroy@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
yield_cpu is a sample of a preempted lock holder that gets propagated
back through the queue. Queued waiters use this to yield to the
preempted lock holder without continually sampling the lock word (which
would defeat the purpose of MCS queueing by bouncing the cache line).
The problem is that yield_cpu can become stale. It can take some time to
be passed down the chain, and if any queued waiter gets preempted then
it will cease to propagate the yield_cpu to later waiters.
This can result in yielding to a CPU that no longer holds the lock,
which is bad, but particularly if it is currently in H_CEDE (idle),
then it appears to be preempted and some hypervisors (PowerVM) can
cause very long H_CONFER latencies waiting for H_CEDE wakeup. This
results in latency spikes and hard lockups on oversubscribed
partitions with lock contention.
This is a minimal fix. Before yielding to yield_cpu, sample the lock
word to confirm yield_cpu is still the owner, and bail out of it is not.
Thanks to a bunch of people who reported this and tracked down the
exact problem using tracepoints and dispatch trace logs.
Fixes: 28db61e207 ("powerpc/qspinlock: allow propagation of yield CPU down the queue")
Cc: stable@vger.kernel.org # v6.2+
Reported-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Reported-by: Laurent Dufour <ldufour@linux.ibm.com>
Reported-by: Shrikanth Hegde <sshegde@linux.vnet.ibm.com>
Debugged-by: "Nysal Jan K.A" <nysal@linux.ibm.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Tested-by: Shrikanth Hegde <sshegde@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20231016124305.139923-2-npiggin@gmail.com
The only callers of zalloc_maybe_bootmem() are PCI setup routines. These
used to be called early during boot before slab setup, and also during
runtime due to hotplug.
But commit 5537fcb319 ("powerpc/pci: Add ppc_md.discover_phbs()")
moved the boot-time calls later, after slab setup, meaning there's no
longer any need for zalloc_maybe_bootmem(), kzalloc() can be used in all
cases.
Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230823055430.752550-1-mpe@ellerman.id.au
Commit ddb5cdbafa ("kbuild: generate KSYMTAB entries by modpost")
deprecated <asm/export.h>, which is now a wrapper of <linux/export.h>.
Replace #include <asm/export.h> with #include <linux/export.h>.
After all the <asm/export.h> lines are converted, <asm/export.h> and
<asm-generic/export.h> will be removed.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
[mpe: Fixup selftests that stub asm/export.h]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230806150954.394189-2-masahiroy@kernel.org
On powerpc32, features fixup is performed very early and that's too
early to read the cmdline and take into account 'nosmap' parameter.
On the other hand, no userspace access is performed that early and
KUAP feature fixup can be performed later.
Add a function to update mmu features. The function is passed a
mask with the features that can be updated.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/31b27ee2c9d338f4f82cd8cd69d6bff979495290.1689091022.git.christophe.leroy@csgroup.eu
Commit e441273947 ("Documentation: raise minimum supported version of
binutils to 2.25") allows us to remove the checks for old binutils.
There is no more user for ld-ifversion. Remove it as well.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230119082250.151485-1-masahiroy@kernel.org
Annotate the release barrier and memory clobber (in effect, producing a
compiler barrier) in the publish_tail_cpu call. These barriers have the
effect of ensuring that qnode attributes are all written to prior to
publish the node to the waitqueue.
Even while the initial write to the 'locked' attribute is guaranteed to
terminate prior to the node being visible, KCSAN still complains that
the write is reorderable by the compiler. Issue a kcsan_release() to
inform KCSAN of the release barrier contained in publish_tail_cpu().
Signed-off-by: Rohan McLure <rmclure@linux.ibm.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230510033117.1395895-3-rmclure@linux.ibm.com
The powerpc implementation of qspinlocks will both poll and spin on the
bitlock guarding a qnode. Mark these accesses with READ_ONCE to convey
to KCSAN that polling is intentional here.
Signed-off-by: Rohan McLure <rmclure@linux.ibm.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230510033117.1395895-2-rmclure@linux.ibm.com
PC-Relative or PCREL addressing is an extension to the ELF ABI which
uses Power ISA v3.1 PC-relative instructions to calculate addresses,
rather than the traditional TOC scheme.
Add an option to build vmlinux using pcrel addressing. Modules continue
to use TOC addressing.
- TOC address helpers and r2 are poisoned with -1 when running vmlinux.
r2 could be used for something useful once things are ironed out.
- Assembly must call C functions with @notoc annotation, or the linker
complains aobut a missing nop after the call. This is done with the
CFUNC macro introduced earlier.
- Boot: with the exception of prom_init, the execution branches to the
kernel virtual address early in boot, before any addresses are
generated, which ensures 34-bit pcrel addressing does not miss the
high PAGE_OFFSET bits. TOC relative addressing has a similar
requirement. prom_init does not go to the virtual address and its
addresses should not carry over to the post-prom kernel.
- Ftrace trampolines are converted from TOC addressing to pcrel
addressing, including module ftrace trampolines that currently use the
kernel TOC to find ftrace target functions.
- BPF function prologue and function calling generation are converted
from TOC to pcrel.
- copypage_64.S has an interesting problem, prefixed instructions have
alignment restrictions so the linker can add padding, which makes the
assembler treat the difference between two local labels as
non-constant even if alignment is arranged so padding is not required.
This may need toolchain help to solve nicely, for now move the prefix
instruction out of the alternate patch section to work around it.
This reduces kernel text size by about 6%.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230408021752.862660-6-npiggin@gmail.com
This macro is to be used in assembly where C functions are called.
pcrel addressing mode requires branches to functions with a
localentry value of 1 to have either a trailing nop or @notoc.
This macro permits the latter without changing callers.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Add dummy definitions to fix selftests build]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230408021752.862660-5-npiggin@gmail.com
Commit 21b56c8477 ("iov_iter: get rid of separate bvec and xarray
callbacks") removed the calls to memcpy_page_flushcache().
Remove the unnecessary memcpy_page_flushcache() call.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20221230-kmap-x86-v1-2-15f1ecccab50@intel.com
Nathan reported that the new per-cpu mm patching oopses if DEBUG_VM is
enabled:
------------[ cut here ]------------
kernel BUG at arch/powerpc/mm/pgtable.c:333!
Oops: Exception in kernel mode, sig: 5 [#1]
LE PAGE_SIZE=64K MMU=Radix SMP NR_CPUS=2048 NUMA PowerNV
Modules linked in:
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 6.1.0-rc2+ #1
Hardware name: IBM PowerNV (emulated by qemu) POWER9 0x4e1200 opal:v7.0 PowerNV
...
NIP assert_pte_locked+0x180/0x1a0
LR assert_pte_locked+0x170/0x1a0
Call Trace:
0x60000000 (unreliable)
patch_instruction+0x618/0x6d0
arch_prepare_kprobe+0xfc/0x2d0
register_kprobe+0x520/0x7c0
arch_init_kprobes+0x28/0x3c
init_kprobes+0x108/0x184
do_one_initcall+0x60/0x2e0
kernel_init_freeable+0x1f0/0x3e0
kernel_init+0x34/0x1d0
ret_from_kernel_thread+0x5c/0x64
It's caused by the assert_spin_locked() failing in assert_pte_locked().
The assert fails because the PTE was unlocked in text_area_cpu_up_mm(),
and never relocked.
The PTE page shouldn't be freed, the patching_mm is only used for
patching on this CPU, only that single PTE is ever mapped, and it's only
unmapped at CPU offline.
In fact assert_pte_locked() has a special case to ignore init_mm
entirely, and the patching_mm is more-or-less like init_mm, so possibly
the check could be skipped for patching_mm too.
But for now be conservative, and use the proper PTE accessors at
patching time, so that the PTE lock is held while the PTE is used. That
also avoids the warning in assert_pte_locked().
With that it's no longer necessary to save the PTE in
cpu_patching_context for the mm_patch_enabled() case.
Fixes: c28c15b6d2 ("powerpc/code-patching: Use temporary mm for Radix MMU")
Reported-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221216125913.990972-1-mpe@ellerman.id.au
Once init section is freed, attempting to patch init code
ends up in the weed.
Commit 51c3c62b58 ("powerpc: Avoid code patching freed init sections")
protected patch_instruction() against that, but it is the responsibility
of the caller to ensure that the patched memory is valid.
All callers have now been verified and fixed so the check
can be removed.
This improves ftrace activation by about 2% on 8xx.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/504310828f473d424e2ed229eff57bf075f52796.1669969781.git.christophe.leroy@csgroup.eu
Once init section is freed, attempting to patch init code
ends up in the weed.
Commit 51c3c62b58 ("powerpc: Avoid code patching freed init sections")
protected patch_instruction() against that, but it is the responsibility
of the caller to ensure that the patched memory is valid.
In the same spirit as jump_label with its jump_label_can_update()
function, add is_fixup_addr_valid() function to skip patching on
freed init section.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/8e9311fc1b057e4e6a2a3a0701ebcc74b787affe.1669969781.git.christophe.leroy@csgroup.eu
No need to have one implementation of patch_instruction() for
CONFIG_STRICT_KERNEL_RWX and one for !CONFIG_STRICT_KERNEL_RWX.
In patch_instruction(), call raw_patch_instruction() when
!CONFIG_STRICT_KERNEL_RWX.
In poking_init(), bail out immediately, it will be equivalent
to the weak default implementation.
Everything else is declared static and will be discarded by
GCC when !CONFIG_STRICT_KERNEL_RWX.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/f67d2a109404d03e8fdf1ea15388c8778337a76b.1669969781.git.christophe.leroy@csgroup.eu
Merge Nick's powerpc qspinlock implementation. From his cover letter:
This replaces the generic queued spinlock code (like s390 does) with our
own implementation.
Generic PV qspinlock code is causing latency / starvation regressions on
large systems that are resulting in hard lockups reported (mostly in
pathoogical cases). The generic qspinlock code has a number of issues
important for powerpc hardware and hypervisors that aren't easily solved
without changing code that would impact other architectures. Follow
s390's lead and implement our own for now.
Issues for powerpc using generic qspinlocks:
- The previous lock value should not be loaded with simple loads, and
need not be passed around from previous loads or cmpxchg results,
because powerpc uses ll/sc-style atomics which can perform more
complex operations that do not require this. powerpc implementations
tend to prefer loads use larx for improved coherency performance.
- The queueing process should absolutely minimise the number of stores
to the lock word to reduce exclusive coherency probes, important for
large system scalability. The pending logic is counter productive
here.
- Non-atomic unlock for paravirt locks is important (atomic
instructions tend to still be more expensive than x86 CPUs).
- Yielding to the lock owner is important in the oversubscribed
paravirt case, which requires storing the owner CPU in the lock
word.
- More control of lock stealing for the paravirt case is important to
keep latency down on large systems.
- The lock acquisition operation should always be made with a special
variant of atomic instructions with the lock hint bit set,
including (especially) in the queueing paths. This is more a matter
of adding more arch lock helpers so not an insurmountable problem
for generic code.
This is a common offset that currently uses the overloaded
STACK_FRAME_OVERHEAD constant. It's easier to read and more
flexible to use a specific regs offset for this.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221127124942.1665522-8-npiggin@gmail.com
With the temp mm context support, there are CPU local variables to hold
the patch address and pte. Use these in the non-temp mm path as well
instead of adding a level of indirection through the text_poke_area
vm_struct and pointer chasing the pte.
As both paths use these fields now, there is no need to let unreferenced
variables be dropped by the compiler, so it is cleaner to merge them
into a single context struct. This has the additional benefit of
removing a redundant CPU local pointer, as only one of cpu_patching_mm /
text_poke_area is ever used, while remaining well-typed. It also groups
each CPU's data into a single cacheline.
Signed-off-by: Benjamin Gray <bgray@linux.ibm.com>
[mpe: Shorten name to 'area' as suggested by Christophe]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221109045112.187069-10-bgray@linux.ibm.com
x86 supports the notion of a temporary mm which restricts access to
temporary PTEs to a single CPU. A temporary mm is useful for situations
where a CPU needs to perform sensitive operations (such as patching a
STRICT_KERNEL_RWX kernel) requiring temporary mappings without exposing
said mappings to other CPUs. Another benefit is that other CPU TLBs do
not need to be flushed when the temporary mm is torn down.
Mappings in the temporary mm can be set in the userspace portion of the
address-space.
Interrupts must be disabled while the temporary mm is in use. HW
breakpoints, which may have been set by userspace as watchpoints on
addresses now within the temporary mm, are saved and disabled when
loading the temporary mm. The HW breakpoints are restored when unloading
the temporary mm. All HW breakpoints are indiscriminately disabled while
the temporary mm is in use - this may include breakpoints set by perf.
Use the `poking_init` init hook to prepare a temporary mm and patching
address. Initialize the temporary mm using mm_alloc(). Choose a
randomized patching address inside the temporary mm userspace address
space. The patching address is randomized between PAGE_SIZE and
DEFAULT_MAP_WINDOW-PAGE_SIZE.
Bits of entropy with 64K page size on BOOK3S_64:
bits of entropy = log2(DEFAULT_MAP_WINDOW_USER64 / PAGE_SIZE)
PAGE_SIZE=64K, DEFAULT_MAP_WINDOW_USER64=128TB
bits of entropy = log2(128TB / 64K)
bits of entropy = 31
The upper limit is DEFAULT_MAP_WINDOW due to how the Book3s64 Hash MMU
operates - by default the space above DEFAULT_MAP_WINDOW is not
available. Currently the Hash MMU does not use a temporary mm so
technically this upper limit isn't necessary; however, a larger
randomization range does not further "harden" this overall approach and
future work may introduce patching with a temporary mm on Hash as well.
Randomization occurs only once during initialization for each CPU as it
comes online.
The patching page is mapped with PAGE_KERNEL to set EAA[0] for the PTE
which ignores the AMR (so no need to unlock/lock KUAP) according to
PowerISA v3.0b Figure 35 on Radix.
Based on x86 implementation:
commit 4fc19708b1
("x86/alternatives: Initialize temporary mm for patching")
and:
commit b3fd8e83ad
("x86/alternatives: Use temporary mm for text poking")
From: Benjamin Gray <bgray@linux.ibm.com>
Synchronisation is done according to ISA 3.1B Book 3 Chapter 13
"Synchronization Requirements for Context Alterations". Switching the mm
is a change to the PID, which requires a CSI before and after the change,
and a hwsync between the last instruction that performs address
translation for an associated storage access.
Instruction fetch is an associated storage access, but the instruction
address mappings are not being changed, so it should not matter which
context they use. We must still perform a hwsync to guard arbitrary
prior code that may have accessed a userspace address.
TLB invalidation is local and VA specific. Local because only this core
used the patching mm, and VA specific because we only care that the
writable mapping is purged. Leaving the other mappings intact is more
efficient, especially when performing many code patches in a row (e.g.,
as ftrace would).
Signed-off-by: Christopher M. Riedl <cmr@bluescreens.de>
Signed-off-by: Benjamin Gray <bgray@linux.ibm.com>
[mpe: Use mm_alloc() per 107b6828a7cd ("x86/mm: Use mm_alloc() in poking_init()")]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221109045112.187069-9-bgray@linux.ibm.com
This adds compile-time options that allow the EH lock hint bit to be
enabled or disabled, and adds some new options that may or may not
help matters. To help with experimentation and tuning.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-18-npiggin@gmail.com
Finding the owner or a queued waiter on a lock with a preempted vcpu is
indicative of an oversubscribed guest causing the lock to get into
trouble. Provide some options to detect this situation and have new CPUs
avoid queueing for a longer time (more steal iterations) to minimise the
problems caused by vcpu preemption on the queue.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-17-npiggin@gmail.com
Provide an option that holds off queueing indefinitely while the lock
owner is preempted. This could reduce queueing latencies for very
overcommitted vcpu situations.
This is disabled by default.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-16-npiggin@gmail.com
Allow for a reduction in the number of times a CPU from a different
node than the owner can attempt to steal the lock before queueing.
This could bias the transfer behaviour of the lock across the
machine and reduce NUMA crossings.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-15-npiggin@gmail.com
Use the spin_begin/spin_cpu_relax/spin_end APIs in qspinlock, which helps
to prevent threads issuing a lot of expensive priority nops which may not
have much effect due to immediately executing low then medium priority.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-14-npiggin@gmail.com
This change allows trylock to steal the lock. It also allows the initial
lock attempt to steal the lock rather than bailing out and going to the
slow path.
This gives trylock more strength: without this a continually-contended
lock will never permit a trylock to succeed. With this change, the
trylock has a small but non-zero chance.
It also gives the lock fastpath most of the benefit of passing the
reservation back through to the steal loop in the slow path without the
complexity.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-13-npiggin@gmail.com
After the head of the queue acquires the lock, it releases the
next waiter in the queue to become the new head. Add an option
to prod the new head if its vCPU was preempted. This may only
have an effect if queue waiters are yielding.
Disable this option by default for now, i.e., no logical change.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-12-npiggin@gmail.com
Having all CPUs poll the lock word for the owner CPU that should be
yielded to defeats most of the purpose of using MCS queueing for
scalability. Yet it may be desirable for queued waiters to yield to a
preempted owner.
With this change, queue waiters never sample the owner CPU directly from
the lock word. The queue head (which is spinning on the lock) propagates
the owner CPU back to the next waiter if it finds the owner has been
preempted. That waiter then propagates the owner CPU back to the next
waiter, and so on.
s390 addresses this problem differenty, by having queued waiters sample
the lock word to find the owner at a low frequency. That has the
advantage of being simpler, the advantage of propagation is that the
lock word never has to be accesed by queued waiters, and the transfer of
cache lines to transmit the owner data is only required when lock holder
vCPU preemption occurs.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-11-npiggin@gmail.com
If the head of queue is preventing stealing but it finds the owner vCPU
is preempted, it will yield its cycles to the owner which could cause it
to become preempted. Add an option to re-allow stealers before yielding,
and disallow them again after returning from the yield.
Disable this option by default for now, i.e., no logical change.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-10-npiggin@gmail.com
Queued waiters which are not at the head of the queue don't spin on
the lock word but their qnode lock word, waiting for the previous queued
CPU to release them. Add an option which allows these waiters to yield
to the previous CPU if its vCPU is preempted.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-9-npiggin@gmail.com
Waiters spinning on the lock word should yield to the lock owner if the
vCPU is preempted. This improves performance when the hypervisor has
oversubscribed physical CPUs.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-8-npiggin@gmail.com
Store the owner CPU number in the lock word so it may be yielded to,
as powerpc's paravirtualised simple spinlocks do.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-7-npiggin@gmail.com
Give the queue head the ability to stop stealers. After a number of
spins without successfully acquiring the lock, the queue head sets
this, which halts stealing and will assure it is the next owner.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-6-npiggin@gmail.com
Allow new waiters to "steal" the lock before queueing. That is, to
acquire it while other CPUs have queued.
This particularly helps paravirt performance when physical CPUs are
oversubscribed, by keeping the lock from becoming a strict FIFO and
vCPU preemption causing queue train wrecks.
The new __queued_spin_trylock_steal() function is put in qspinlock.h
to save having to move it, because it will be used there by a later
change.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-5-npiggin@gmail.com
This uses more optimal ll/sc style access patterns (rather than
cmpxchg), and also sets the EH=1 lock hint on those operations
which acquire ownership of the lock.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-4-npiggin@gmail.com
This forms the basis of the qspinlock slow path.
Like generic qspinlocks and unlike the vanilla MCS algorithm, the lock
owner does not participate in the queue, only waiters. The first waiter
spins on the lock word, then when the lock is released it takes
ownership and unqueues the next waiter. This is how qspinlocks can be
implemented with the spinlock API -- lock owners don't need a node, only
waiters do.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221126095932.1234527-2-npiggin@gmail.com
Add a powerpc specific implementation of queued spinlocks. This is the
build framework with a very simple (non-queued) spinlock implementation
to begin with. Later changes add queueing, and other features and
optimisations one-at-a-time. It is done this way to more easily see how
the queued spinlocks are built, and to make performance and correctness
bisects more useful.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Drop paravirt.h & processor.h changes to fix 32-bit build]
[mpe: Fix 32-bit build of qspinlock.o & disallow GENERIC_LOCKBREAK per Nick]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/CONLLQB6DCJU.2ZPOS7T6S5GRR@bobo
BUG_ON() when failing to initialise the code patching window is
unnecessary, and use of BUG_ON is discouraged. We don't set
poking_init_done in this case, so failure to init the boot CPU will
result in a strict RWX error when a following patch_instruction uses
raw_patch_instruction. If it only fails for later CPUs, they won't be
onlined in the first place.
The return value of cpuhp_setup_state() is also >= 0 on success,
so check for < 0.
Signed-off-by: Benjamin Gray <bgray@linux.ibm.com>
Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221109045112.187069-3-bgray@linux.ibm.com
ISA v2.06 (POWER7 and up) as well as e6500 support lbarx and lharx.
Add a compile option that allows code to use it, and add support in
cmpxchg and xchg 8 and 16 bit values without shifting and masking.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220909052312.63916-1-npiggin@gmail.com
schedule must not be explicitly called while KUAP is unlocked, because
the AMR register will not be saved across the context switch on
64s (preemption is allowed because that is driven by interrupts which do
save the AMR).
exit_vmx_usercopy() runs inside an unlocked user access region, and it
calls preempt_enable() which will call schedule() if need_resched() was
set while non-preemptible. This can cause tasks to run unprotected when
the should not, and can cause the user copy to be improperly blocked
when scheduling back to it.
Fix this by avoiding the explicit resched for preempt kernels by
generating an interrupt to reschedule the context if need_resched() got
set.
Reported-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221013151647.1857994-3-npiggin@gmail.com
Use helper macros to access global variables, and place them in .data
sections rather than in .toc. Putting addresses in TOC is not required
because the kernel is linked with a single TOC.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220926034057.2360083-3-npiggin@gmail.com
Nysal Jan K.A.