| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
* kvm-arm64/6.6/pmu-fixes:
: .
: Another set of PMU fixes, coutrtesy of Reiji Watanabe.
: From the cover letter:
:
: "This series fixes a couple of PMUver related handling of
: vPMU support.
:
: On systems where the PMUVer is not uniform across all PEs,
: KVM currently does not advertise PMUv3 to the guest,
: even if userspace successfully runs KVM_ARM_VCPU_INIT with
: KVM_ARM_VCPU_PMU_V3."
:
: Additionally, a fix for an obscure counter oversubscription
: issue happening when the hsot profines the guest's EL0.
: .
KVM: arm64: pmu: Guard PMU emulation definitions with CONFIG_KVM
KVM: arm64: pmu: Resync EL0 state on counter rotation
KVM: arm64: PMU: Don't advertise STALL_SLOT_{FRONTEND,BACKEND}
KVM: arm64: PMU: Don't advertise the STALL_SLOT event
KVM: arm64: PMU: Avoid inappropriate use of host's PMUVer
KVM: arm64: PMU: Disallow vPMU on non-uniform PMUVer
Signed-off-by: Marc Zyngier <maz@kernel.org>
|
|
* kvm-arm64/tlbi-range:
: .
: FEAT_TLBIRANGE support, courtesy of Raghavendra Rao Ananta.
: From the cover letter:
:
: "In certain code paths, KVM/ARM currently invalidates the entire VM's
: page-tables instead of just invalidating a necessary range. For example,
: when collapsing a table PTE to a block PTE, instead of iterating over
: each PTE and flushing them, KVM uses 'vmalls12e1is' TLBI operation to
: flush all the entries. This is inefficient since the guest would have
: to refill the TLBs again, even for the addresses that aren't covered
: by the table entry. The performance impact would scale poorly if many
: addresses in the VM is going through this remapping.
:
: For architectures that implement FEAT_TLBIRANGE, KVM can replace such
: inefficient paths by performing the invalidations only on the range of
: addresses that are in scope. This series tries to achieve the same in
: the areas of stage-2 map, unmap and write-protecting the pages."
: .
KVM: arm64: Use TLBI range-based instructions for unmap
KVM: arm64: Invalidate the table entries upon a range
KVM: arm64: Flush only the memslot after write-protect
KVM: arm64: Implement kvm_arch_flush_remote_tlbs_range()
KVM: arm64: Define kvm_tlb_flush_vmid_range()
KVM: arm64: Implement __kvm_tlb_flush_vmid_range()
arm64: tlb: Implement __flush_s2_tlb_range_op()
arm64: tlb: Refactor the core flush algorithm of __flush_tlb_range
KVM: Move kvm_arch_flush_remote_tlbs_memslot() to common code
KVM: Allow range-based TLB invalidation from common code
KVM: Remove CONFIG_HAVE_KVM_ARCH_TLB_FLUSH_ALL
KVM: arm64: Use kvm_arch_flush_remote_tlbs()
KVM: Declare kvm_arch_flush_remote_tlbs() globally
KVM: Rename kvm_arch_flush_remote_tlb() to kvm_arch_flush_remote_tlbs()
Signed-off-by: Marc Zyngier <maz@kernel.org>
|
|
* kvm-arm64/nv-trap-forwarding: (30 commits)
: .
: This implements the so called "trap forwarding" infrastructure, which
: gets used when we take a trap from an L2 guest and that the L1 guest
: wants to see the trap for itself.
: .
KVM: arm64: nv: Add trap description for SPSR_EL2 and ELR_EL2
KVM: arm64: nv: Select XARRAY_MULTI to fix build error
KVM: arm64: nv: Add support for HCRX_EL2
KVM: arm64: Move HCRX_EL2 switch to load/put on VHE systems
KVM: arm64: nv: Expose FGT to nested guests
KVM: arm64: nv: Add switching support for HFGxTR/HDFGxTR
KVM: arm64: nv: Expand ERET trap forwarding to handle FGT
KVM: arm64: nv: Add SVC trap forwarding
KVM: arm64: nv: Add trap forwarding for HDFGxTR_EL2
KVM: arm64: nv: Add trap forwarding for HFGITR_EL2
KVM: arm64: nv: Add trap forwarding for HFGxTR_EL2
KVM: arm64: nv: Add fine grained trap forwarding infrastructure
KVM: arm64: nv: Add trap forwarding for CNTHCTL_EL2
KVM: arm64: nv: Add trap forwarding for MDCR_EL2
KVM: arm64: nv: Expose FEAT_EVT to nested guests
KVM: arm64: nv: Add trap forwarding for HCR_EL2
KVM: arm64: nv: Add trap forwarding infrastructure
KVM: arm64: Restructure FGT register switching
KVM: arm64: nv: Add FGT registers
KVM: arm64: Add missing HCR_EL2 trap bits
...
Signed-off-by: Marc Zyngier <maz@kernel.org>
|
|
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull SCSI fixes from James Bottomley:
"Three small driver fixes and one larger unused function set removal in
the raid class (so no external impact)"
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
scsi: snic: Fix double free in snic_tgt_create()
scsi: core: raid_class: Remove raid_component_add()
scsi: ufs: ufs-qcom: Clear qunipro_g4_sel for HW major version > 5
scsi: ufs: mcq: Fix the search/wrap around logic
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Ingo Molnar:
"Fix an FPU invalidation bug on exec(), and fix a performance
regression due to a missing setting of X86_FEATURE_OSXSAVE"
* tag 'x86-urgent-2023-08-26' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/fpu: Set X86_FEATURE_OSXSAVE feature after enabling OSXSAVE in CR4
x86/fpu: Invalidate FPU state correctly on exec()
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull irq fix from Thomas Gleixner:
"A last minute fix for a regression introduced in the v6.5 merge
window.
The conversion of the software based interrupt resend mechanism to
hlist missed to add a check whether the descriptor is already enqueued
and dropped the interrupt descriptor lookup for nested interrupts.
The missing check whether the descriptor is already queued causes
hlist corruption and can be observed in the wild. The dropped parent
descriptor lookup has not yet caused problems, but it would result in
stale interrupt line in the worst case.
Add the missing enqueued check and bring the descriptor lookup back to
cure this"
* tag 'irq-urgent-2023-08-26' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
genirq: Fix software resend lockup and nested resend
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson
Pull LoongArch fixes from Huacai Chen:
"Fix a ptrace bug, a hw_breakpoint bug, some build errors/warnings and
some trivial cleanups"
* tag 'loongarch-fixes-6.5-2' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson:
LoongArch: Fix hw_breakpoint_control() for watchpoints
LoongArch: Ensure FP/SIMD registers in the core dump file is up to date
LoongArch: Put the body of play_dead() into arch_cpu_idle_dead()
LoongArch: Add identifier names to arguments of die() declaration
LoongArch: Return earlier in die() if notify_die() returns NOTIFY_STOP
LoongArch: Do not kill the task in die() if notify_die() returns NOTIFY_STOP
LoongArch: Remove <asm/export.h>
LoongArch: Replace #include <asm/export.h> with #include <linux/export.h>
LoongArch: Remove unneeded #include <asm/export.h>
LoongArch: Replace -ffreestanding with finer-grained -fno-builtin's
LoongArch: Remove redundant "source drivers/firmware/Kconfig"
|
|
The switch to using hlist for managing software resend of interrupts
broke resend in at least two ways:
First, unconditionally adding interrupt descriptors to the resend list can
corrupt the list when the descriptor in question has already been
added. This causes the resend tasklet to loop indefinitely with interrupts
disabled as was recently reported with the Lenovo ThinkPad X13s after
threaded NAPI was disabled in the ath11k WiFi driver.
This bug is easily fixed by restoring the old semantics of irq_sw_resend()
so that it can be called also for descriptors that have already been marked
for resend.
Second, the offending commit also broke software resend of nested
interrupts by simply discarding the code that made sure that such
interrupts are retriggered using the parent interrupt.
Add back the corresponding code that adds the parent descriptor to the
resend list.
Fixes: bc06a9e08742 ("genirq: Use hlist for managing resend handlers")
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/lkml/20230809073432.4193-1-johan+linaro@kernel.org/
Link: https://lore.kernel.org/r/20230826154004.1417-1-johan+linaro@kernel.org
|
|
In hw_breakpoint_control(), encode_ctrl_reg() has already encoded the
MWPnCFG3_LoadEn/MWPnCFG3_StoreEn bits in info->ctrl. We don't need to
add (1 << MWPnCFG3_LoadEn | 1 << MWPnCFG3_StoreEn) unconditionally.
Otherwise we can't set read watchpoint and write watchpoint separately.
Cc: stable@vger.kernel.org
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
This is a port of commit 379eb01c21795edb4c ("riscv: Ensure the value
of FP registers in the core dump file is up to date").
The values of FP/SIMD registers in the core dump file come from the
thread.fpu. However, kernel saves the FP/SIMD registers only before
scheduling out the process. If no process switch happens during the
exception handling, kernel will not have a chance to save the latest
values of FP/SIMD registers. So it may cause their values in the core
dump file incorrect. To solve this problem, force fpr_get()/simd_get()
to save the FP/SIMD registers into the thread.fpu if the target task
equals the current task.
Cc: stable@vger.kernel.org
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
commit f922c13e778d ("KVM: arm64: Introduce
pkvm_alloc_private_va_range()") and commit 92abe0f81e13 ("KVM: arm64:
Introduce hyp_alloc_private_va_range()") added an alignment for the
start address of any allocation into the nVHE hypervisor private VA
range.
This alignment (order of the size of the allocation) intends to enable
efficient stack verification (if the PAGE_SHIFT bit is zero, the stack
pointer is on the guard page and a stack overflow occurred).
But this is only necessary for stack allocation and can waste a lot of
VA space. So instead make stack-specific functions, handling the guard
page requirements, while other users (e.g. fixmap) will only get page
alignment.
Reviewed-by: Kalesh Singh <kaleshsingh@google.com>
Signed-off-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230811112037.1147863-1-vdonnefort@google.com
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk fixes from Stephen Boyd:
"One clk driver fix and two clk framework fixes:
- Fix an OOB access when devm_get_clk_from_child() is used and
devm_clk_release() casts the void pointer to the wrong type
- Move clk_rate_exclusive_{get,put}() within the correct ifdefs in
clk.h so that the stubs are used when CONFIG_COMMON_CLK=n
- Register the proper clk provider function depending on the value of
#clock-cells in the TI keystone driver"
* tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
clk: Fix slab-out-of-bounds error in devm_clk_release()
clk: Fix undefined reference to `clk_rate_exclusive_{get,put}'
clk: keystone: syscon-clk: Fix audio refclk
|
|
The gcc compiler translates on some architectures the 64-bit
__builtin_clzll() function to a call to the libgcc function __clzdi2(),
which should take a 64-bit parameter on 32- and 64-bit platforms.
But in the current kernel code, the built-in __clzdi2() function is
defined to operate (wrongly) on 32-bit parameters if BITS_PER_LONG ==
32, thus the return values on 32-bit kernels are in the range from
[0..31] instead of the expected [0..63] range.
This patch fixes the in-kernel functions __clzdi2() and __ctzdi2() to
take a 64-bit parameter on 32-bit kernels as well, thus it makes the
functions identical for 32- and 64-bit kernels.
This bug went unnoticed since kernel 3.11 for over 10 years, and here
are some possible reasons for that:
a) Some architectures have assembly instructions to count the bits and
which are used instead of calling __clzdi2(), e.g. on x86 the bsr
instruction and on ppc cntlz is used. On such architectures the
wrong __clzdi2() implementation isn't used and as such the bug has
no effect and won't be noticed.
b) Some architectures link to libgcc.a, and the in-kernel weak
functions get replaced by the correct 64-bit variants from libgcc.a.
c) __builtin_clzll() and __clzdi2() doesn't seem to be used in many
places in the kernel, and most likely only in uncritical functions,
e.g. when printing hex values via seq_put_hex_ll(). The wrong return
value will still print the correct number, but just in a wrong
formatting (e.g. with too many leading zeroes).
d) 32-bit kernels aren't used that much any longer, so they are less
tested.
A trivial testcase to verify if the currently running 32-bit kernel is
affected by the bug is to look at the output of /proc/self/maps:
Here the kernel uses a correct implementation of __clzdi2():
root@debian:~# cat /proc/self/maps
00010000-00019000 r-xp 00000000 08:05 787324 /usr/bin/cat
00019000-0001a000 rwxp 00009000 08:05 787324 /usr/bin/cat
0001a000-0003b000 rwxp 00000000 00:00 0 [heap]
f7551000-f770d000 r-xp 00000000 08:05 794765 /usr/lib/hppa-linux-gnu/libc.so.6
...
and this kernel uses the broken implementation of __clzdi2():
root@debian:~# cat /proc/self/maps
0000000010000-0000000019000 r-xp 00000000 000000008:000000005 787324 /usr/bin/cat
0000000019000-000000001a000 rwxp 000000009000 000000008:000000005 787324 /usr/bin/cat
000000001a000-000000003b000 rwxp 00000000 00:00 0 [heap]
00000000f73d1000-00000000f758d000 r-xp 00000000 000000008:000000005 794765 /usr/lib/hppa-linux-gnu/libc.so.6
...
Signed-off-by: Helge Deller <deller@gmx.de>
Fixes: 4df87bb7b6a22 ("lib: add weak clz/ctz functions")
Cc: Chanho Min <chanho.min@lge.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: stable@vger.kernel.org # v3.11+
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Pull misc fixes from Andrew Morton:
"18 hotfixes. 13 are cc:stable and the remainder pertain to post-6.4
issues or aren't considered suitable for a -stable backport"
* tag 'mm-hotfixes-stable-2023-08-25-11-07' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
shmem: fix smaps BUG sleeping while atomic
selftests: cachestat: catch failing fsync test on tmpfs
selftests: cachestat: test for cachestat availability
maple_tree: disable mas_wr_append() when other readers are possible
madvise:madvise_free_pte_range(): don't use mapcount() against large folio for sharing check
madvise:madvise_free_huge_pmd(): don't use mapcount() against large folio for sharing check
madvise:madvise_cold_or_pageout_pte_range(): don't use mapcount() against large folio for sharing check
mm: multi-gen LRU: don't spin during memcg release
mm: memory-failure: fix unexpected return value in soft_offline_page()
radix tree: remove unused variable
mm: add a call to flush_cache_vmap() in vmap_pfn()
selftests/mm: FOLL_LONGTERM need to be updated to 0x100
nilfs2: fix general protection fault in nilfs_lookup_dirty_data_buffers()
mm/gup: handle cont-PTE hugetlb pages correctly in gup_must_unshare() via GUP-fast
selftests: cgroup: fix test_kmem_basic less than error
mm: enable page walking API to lock vmas during the walk
smaps: use vm_normal_page_pmd() instead of follow_trans_huge_pmd()
mm/gup: reintroduce FOLL_NUMA as FOLL_HONOR_NUMA_FAULT
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux
Pull RISC-V fixes from Palmer Dabbelt:
"This is obviously not ideal, particularly for something this late in
the cycle.
Unfortunately we found some uABI issues in the vector support while
reviewing the GDB port, which has triggered a revert -- probably a
good sign we should have reviewed GDB before merging this, I guess I
just dropped the ball because I was so worried about the context
extension and libc suff I forgot. Hence the late revert.
There's some risk here as we're still exposing the vector context for
signal handlers, but changing that would have meant reverting all of
the vector support. The issues we've found so far have been fixed
already and they weren't absolute showstoppers, so we're essentially
just playing it safe by holding ptrace support for another release (or
until we get through a proper userspace code review).
Summary:
- The vector ucontext extension has been extended with vlenb
- The vector registers ELF core dump note type has been changed to
avoid aliasing with the CSR type used in embedded systems
- Support for accessing vector registers via ptrace() has been
reverted
- Another build fix for the ISA spec changes around Zifencei/Zicsr
that manifests on some systems built with binutils-2.37 and
gcc-11.2"
* tag 'riscv-for-linus-6.5-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
riscv: Fix build errors using binutils2.37 toolchains
RISC-V: vector: export VLENB csr in __sc_riscv_v_state
RISC-V: Remove ptrace support for vectors
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux
Pull gpio fixes from Bartosz Golaszewski:
- fix an irq mapping leak in gpio-sim
- associate the GPIO device's software node with the irq domain in
gpio-sim
* tag 'gpio-fixes-for-v6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:
gpio: sim: pass the GPIO device's software node to irq domain
gpio: sim: dispose of irq mappings before destroying the irq_sim domain
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl
Pull pin control fixes from Linus Walleij:
"Here are some Renesas and AMD driver fixes, the AMD fix affects
important laptops in the wild so this one is pretty important. It
seems a bit tough to get this right.
- Fix DT parsing and related locking in the Renesas driver.
- Fix wakeup IRQs in the AMD driver once again. Really tricky this
one"
* tag 'pinctrl-v6.5-4' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
pinctrl: amd: Mask wake bits on probe again
pinctrl: renesas: rza2: Add lock around pinctrl_generic{{add,remove}_group,{add,remove}_function}
pinctrl: renesas: rzv2m: Fix NULL pointer dereference in rzv2m_dt_subnode_to_map()
pinctrl: renesas: rzg2l: Fix NULL pointer dereference in rzg2l_dt_subnode_to_map()
|
|
Delete KVM's printk about KVM_SET_TSS_ADDR not being called. When the
printk was added by commit 776e58ea3d37 ("KVM: unbreak userspace that does
not sets tss address"), KVM also stuffed a "hopefully safe" value, i.e.
the message wasn't purely informational. For reasons unknown, ostensibly
to try and help people running outdated qemu-kvm versions, the message got
left behind when KVM's stuffing was removed by commit 4918c6ca6838
("KVM: VMX: Require KVM_SET_TSS_ADDR being called prior to running a VCPU").
Today, the message is completely nonsensical, as it has been over a decade
since KVM supported userspace running a Real Mode guest, on a CPU without
unrestricted guest support, without doing KVM_SET_TSS_ADDR before KVM_RUN.
I.e. KVM's ABI has required KVM_SET_TSS_ADDR for 10+ years.
To make matters worse, the message is prone to false positives as it
triggers when simply *creating* a vCPU due to RESET putting vCPUs into
Real Mode, even when the user has no intention of ever *running* the vCPU
in a Real Mode. E.g. KVM selftests stuff 64-bit mode and never touch Real
Mode, but trigger the message even though they run just fine without
doing KVM_SET_TSS_ADDR. Creating "dummy" vCPUs, e.g. to probe features,
can also trigger the message. In both scenarios, the message confuses
users and falsely implies that they've done something wrong.
Reported-by: Thorsten Glaser <t.glaser@tarent.de>
Closes: https://lkml.kernel.org/r/f1afa6c0-cde2-ab8b-ea71-bfa62a45b956%40tarent.de
Link: https://lore.kernel.org/r/20230815174215.433222-1-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Give KVM x86 the same treatment as all other KVM architectures, and
officially take ownership of x86 specific KVM selftests (changes have
been routed through kvm and/or kvm-x86 for quite some time).
Cc: kvm@vger.kernel.org
Link: https://lore.kernel.org/r/20230817234114.1420092-1-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Explicitly set the exception vector to #UD when potentially injecting an
exception in sync_regs_test's subtests that try to detect TOCTOU bugs
in KVM's handling of exceptions injected by userspace. A side effect of
the original KVM bug was that KVM would clear the vector, but relying on
KVM to clear the vector (i.e. make it #DE) makes it less likely that the
test would ever find *new* KVM bugs, e.g. because only the first iteration
would run with a legal vector to start.
Explicitly inject #UD for race_events_inj_pen() as well, e.g. so that it
doesn't inherit the illegal 255 vector from race_events_exc(), which
currently runs first.
Link: https://lore.kernel.org/r/20230817233430.1416463-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Reload known good vCPU state if the vCPU triple faults in any of the
race_sync_regs() subtests, e.g. if KVM successfully injects an exception
(the vCPU isn't configured to handle exceptions). On Intel, the VMCS
is preserved even after shutdown, but AMD's APM states that the VMCB is
undefined after a shutdown and so KVM synthesizes an INIT to sanitize
vCPU/VMCB state, e.g. to guard against running with a garbage VMCB.
The synthetic INIT results in the vCPU never exiting to userspace, as it
gets put into Real Mode at the reset vector, which is full of zeros (as is
GPA 0 and beyond), and so executes ADD for a very, very long time.
Fixes: 60c4063b4752 ("KVM: selftests: Extend x86's sync_regs_test to check for event vector races")
Cc: Michal Luczaj <mhal@rbox.co>
Link: https://lore.kernel.org/r/20230817233430.1416463-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Disallow SEV (and beyond) if nrips is disabled via module param, as KVM
can't read guest memory to partially emulate and skip an instruction. All
CPUs that support SEV support NRIPS, i.e. this is purely stopping the user
from shooting themselves in the foot.
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Link: https://lore.kernel.org/r/20230825013621.2845700-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Don't inject a #UD if KVM attempts to "emulate" to skip an instruction
for an SEV guest, and instead resume the guest and hope that it can make
forward progress. When commit 04c40f344def ("KVM: SVM: Inject #UD on
attempted emulation for SEV guest w/o insn buffer") added the completely
arbitrary #UD behavior, there were no known scenarios where a well-behaved
guest would induce a VM-Exit that triggered emulation, i.e. it was thought
that injecting #UD would be helpful.
However, now that KVM (correctly) attempts to re-inject INT3/INTO, e.g. if
a #NPF is encountered when attempting to deliver the INT3/INTO, an SEV
guest can trigger emulation without a buffer, through no fault of its own.
Resuming the guest and retrying the INT3/INTO is architecturally wrong,
e.g. the vCPU will incorrectly re-hit code #DBs, but for SEV guests there
is literally no other option that has a chance of making forward progress.
Drop the #UD injection for all "skip" emulation, not just those related to
INT3/INTO, even though that means that the guest will likely end up in an
infinite loop instead of getting a #UD (the vCPU may also crash, e.g. if
KVM emulated everything about an instruction except for advancing RIP).
There's no evidence that suggests that an unexpected #UD is actually
better than hanging the vCPU, e.g. a soft-hung vCPU can still respond to
IRQs and NMIs to generate a backtrace.
Reported-by: Wu Zongyo <wuzongyo@mail.ustc.edu.cn>
Closes: https://lore.kernel.org/all/8eb933fd-2cf3-d7a9-32fe-2a1d82eac42a@mail.ustc.edu.cn
Fixes: 6ef88d6e36c2 ("KVM: SVM: Re-inject INT3/INTO instead of retrying the instruction")
Cc: stable@vger.kernel.org
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Link: https://lore.kernel.org/r/20230825013621.2845700-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Skip initializing the VMSA physical address in the VMCB if the VMSA is
NULL, which occurs during intrahost migration as KVM initializes the VMCB
before copying over state from the source to the destination (including
the VMSA and its physical address).
In normal builds, __pa() is just math, so the bug isn't fatal, but with
CONFIG_DEBUG_VIRTUAL=y, the validity of the virtual address is verified
and passing in NULL will make the kernel unhappy.
Fixes: 6defa24d3b12 ("KVM: SEV: Init target VMCBs in sev_migrate_from")
Cc: stable@vger.kernel.org
Cc: Peter Gonda <pgonda@google.com>
Reviewed-by: Peter Gonda <pgonda@google.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta@amd.com>
Link: https://lore.kernel.org/r/20230825022357.2852133-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Fix a goof where KVM tries to grab source vCPUs from the destination VM
when doing intrahost migration. Grabbing the wrong vCPU not only hoses
the guest, it also crashes the host due to the VMSA pointer being left
NULL.
BUG: unable to handle page fault for address: ffffe38687000000
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
PGD 0 P4D 0
Oops: 0000 [#1] SMP NOPTI
CPU: 39 PID: 17143 Comm: sev_migrate_tes Tainted: GO 6.5.0-smp--fff2e47e6c3b-next #151
Hardware name: Google, Inc. Arcadia_IT_80/Arcadia_IT_80, BIOS 34.28.0 07/10/2023
RIP: 0010:__free_pages+0x15/0xd0
RSP: 0018:ffff923fcf6e3c78 EFLAGS: 00010246
RAX: 0000000000000000 RBX: ffffe38687000000 RCX: 0000000000000100
RDX: 0000000000000100 RSI: 0000000000000000 RDI: ffffe38687000000
RBP: ffff923fcf6e3c88 R08: ffff923fcafb0000 R09: 0000000000000000
R10: 0000000000000000 R11: ffffffff83619b90 R12: ffff923fa9540000
R13: 0000000000080007 R14: ffff923f6d35d000 R15: 0000000000000000
FS: 0000000000000000(0000) GS:ffff929d0d7c0000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffffe38687000000 CR3: 0000005224c34005 CR4: 0000000000770ee0
PKRU: 55555554
Call Trace:
<TASK>
sev_free_vcpu+0xcb/0x110 [kvm_amd]
svm_vcpu_free+0x75/0xf0 [kvm_amd]
kvm_arch_vcpu_destroy+0x36/0x140 [kvm]
kvm_destroy_vcpus+0x67/0x100 [kvm]
kvm_arch_destroy_vm+0x161/0x1d0 [kvm]
kvm_put_kvm+0x276/0x560 [kvm]
kvm_vm_release+0x25/0x30 [kvm]
__fput+0x106/0x280
____fput+0x12/0x20
task_work_run+0x86/0xb0
do_exit+0x2e3/0x9c0
do_group_exit+0xb1/0xc0
__x64_sys_exit_group+0x1b/0x20
do_syscall_64+0x41/0x90
entry_SYSCALL_64_after_hwframe+0x63/0xcd
</TASK>
CR2: ffffe38687000000
Fixes: 6defa24d3b12 ("KVM: SEV: Init target VMCBs in sev_migrate_from")
Cc: stable@vger.kernel.org
Cc: Peter Gonda <pgonda@google.com>
Reviewed-by: Peter Gonda <pgonda@google.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta@amd.com>
Link: https://lore.kernel.org/r/20230825022357.2852133-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"Hopefully the last bits for 6.5. It's slightly higher LOCs than
wished, but it doesn't look scary.
The biggest change is MAINTAINERS update for TI; it's good to have the
update before the final release, so that people can contact to the
right persons for bug reports (which shouldn't happen of course!)
The rest are all device-specific fixes and quirks, most for various
ASoC platforms"
* tag 'sound-6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
ASoC: amd: yc: Fix a non-functional mic on Lenovo 82SJ
ALSA: ymfpci: Fix the missing snd_card_free() call at probe error
ASoC: cs35l41: Correct amp_gain_tlv values
ASoC: amd: yc: Add VivoBook Pro 15 to quirks list for acp6x
ASoC: tas2781: fixed register access error when switching to other chips
ASoC: cs35l56: Add an ACPI match table
ASoC: cs35l56: Read firmware uuid from a device property instead of _SUB
ASoC: SOF: ipc4-pcm: fix possible null pointer deference
MAINTAINERS: Add entries for TEXAS INSTRUMENTS ASoC DRIVERS
|
|
The initial aim is to silence the following objtool warning:
arch/loongarch/kernel/process.o: warning: objtool: arch_cpu_idle_dead() falls through to next function start_thread()
According to tools/objtool/Documentation/objtool.txt, this is because
the last instruction of arch_cpu_idle_dead() is a call to a noreturn
function play_dead(). In order to silence the warning, one simple way
is to add the noreturn function play_dead() to objtool's hard-coded
global_noreturns array, that is to say, just put "NORETURN(play_dead)"
into tools/objtool/noreturns.h, it works well.
But I noticed that play_dead() is only defined once and only called by
arch_cpu_idle_dead(), so put the body of play_dead() into the caller
arch_cpu_idle_dead(), then remove the noreturn function play_dead() is
an alternative way which can reduce the overhead of the function call
at the same time.
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
Add identifier names to arguments of die() declaration in ptrace.h
to fix the following checkpatch warnings:
WARNING: function definition argument 'const char *' should also have an identifier name
WARNING: function definition argument 'struct pt_regs *' should also have an identifier name
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
After the call to oops_exit(), it should not panic or execute
the crash kernel if the oops is to be suppressed.
Suggested-by: Maciej W. Rozycki <macro@orcam.me.uk>
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
If notify_die() returns NOTIFY_STOP, honor the return value from the
handler chain invocation in die() and return without killing the task
as, through a debugger, the fault may have been fixed. It makes sense
even if ignoring the event will make the system unstable: by allowing
access through a debugger it has been compromised already anyway. It
makes our port consistent with x86, arm64, riscv and csky.
Commit 20c0d2d44029 ("[PATCH] i386: pass proper trap numbers to die
chain handlers") may be the earliest of similar changes.
Link: https://lore.kernel.org/r/43DDF02E.76F0.0078.0@novell.com/
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
All *.S files under arch/loongarch/ have been converted to include
<linux/export.h> instead of <asm/export.h>.
Remove <asm/export.h>.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
Commit ddb5cdbafaaad ("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>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
There is no EXPORT_SYMBOL() line there, hence #include <asm/export.h>
is unneeded.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
As explained by Nick in the original issue: the kernel usually does a
good job of providing library helpers that have similar semantics as
their ordinary userspace libc equivalents, but -ffreestanding disables
such libcall optimization and other related features in the compiler,
which can lead to unexpected things such as CONFIG_FORTIFY_SOURCE not
working (!).
However, due to the desire for better control over unaligned accesses
with respect to CONFIG_ARCH_STRICT_ALIGN, and also for avoiding the
GCC bug https://gcc.gnu.org/PR109465, we do want to still disable
optimizations for the memory libcalls (memcpy, memmove and memset for
now). Use finer-grained -fno-builtin-* toggles to achieve this without
losing source fortification and other libcall optimizations.
Closes: https://github.com/ClangBuiltLinux/linux/issues/1897
Reported-by: Nathan Chancellor <nathan@kernel.org>
Suggested-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: WANG Xuerui <git@xen0n.name>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
In drivers/Kconfig, drivers/firmware/Kconfig is sourced for all ports so
there is no need to source it in the port-specific Kconfig file. And
sourcing it here also caused the "Firmware Drivers" menu appeared two
times: one in the "Device Drivers" menu, another in the toplevel menu.
This is really puzzling so remove it.
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Xi Ruoyao <xry111@xry111.site>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
|
|
Pull drm fixes from Dave Airlie:
"A bit bigger than I'd care for, but it's mostly a single vmwgfx fix
and a fix for an i915 hotplug probing. Otherwise misc i915, bridge,
panfrost and dma-buf fixes.
core:
- add a HPD poll helper
i915:
- fix regression in i915 polling
- fix docs build warning
- fix DG2 idle power consumption
bridge:
- samsung-dsim: init fix
panfrost:
- fix speed binning issue
dma-buf:
- fix recursive lock in fence signal
vmwgfx:
- fix shader stage validation
- fix NULL ptr derefs in gem put"
* tag 'drm-fixes-2023-08-25' of git://anongit.freedesktop.org/drm/drm:
drm/i915: Fix HPD polling, reenabling the output poll work as needed
drm: Add an HPD poll helper to reschedule the poll work
drm/vmwgfx: Fix possible invalid drm gem put calls
drm/vmwgfx: Fix shader stage validation
dma-buf/sw_sync: Avoid recursive lock during fence signal
drm/i915: fix Sphinx indentation warning
drm/i915/dgfx: Enable d3cold at s2idle
drm/display/dp: Fix the DP DSC Receiver cap size
drm/panfrost: Skip speed binning on EOPNOTSUPP
drm: bridge: samsung-dsim: Fix init during host transfer
|
|
https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus
ASoC: Quirk for v6.5
One additional fix for v6.5, an additional quirk. As with the other
fixes this could wait for the merge window.
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace
Pull tracing fixes from Steven Rostedt:
- Fix ring buffer being permanently disabled due to missed
record_disabled()
Changing the trace cpu mask will disable the ring buffers for the
CPUs no longer in the mask. But it fails to update the snapshot
buffer. If a snapshot takes place, the accounting for the ring buffer
being disabled is corrupted and this can lead to the ring buffer
being permanently disabled.
- Add test case for snapshot and cpu mask working together
- Fix memleak by the function graph tracer not getting closed properly.
The iterator is used to read the ring buffer. When it opens, it calls
the open function of a tracer, and when it is closed, it calls the
close iteration. While a trace is being read, it is still possible to
change the tracer.
If this happens between the function graph tracer and the wakeup
tracer (which uses function graph tracing), the tracers are not
closed properly during when the iterator sees the switch, and the
wakeup function did not initialize its private pointer to NULL, which
is used to know if the function graph tracer was the last tracer. It
could be fooled in thinking it is, but then on exit it does not call
the close function of the function graph tracer to clean up its data.
- Fix synthetic events on big endian machines, by introducing a union
that does the conversions properly.
- Fix synthetic events from printing out the number of elements in the
stacktrace when it shouldn't.
- Fix synthetic events stacktrace to not print a bogus value at the
end.
- Introduce a pipe_cpumask that prevents the trace_pipe files from
being opened by more than one task (file descriptor).
There was a race found where if splice is called, the iter->ent could
become stale and events could be missed. There's no point reading a
producer/consumer file by more than one task as they will corrupt
each other anyway. Add a cpumask that keeps track of the per_cpu
trace_pipe files as well as the global trace_pipe file that prevents
more than one open of a trace_pipe file that represents the same ring
buffer. This prevents the race from happening.
- Fix ftrace samples for arm64 to work with older compilers.
* tag 'trace-v6.5-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
samples: ftrace: Replace bti assembly with hint for older compiler
tracing: Introduce pipe_cpumask to avoid race on trace_pipes
tracing: Fix memleak due to race between current_tracer and trace
tracing/synthetic: Allocate one additional element for size
tracing/synthetic: Skip first entry for stack traces
tracing/synthetic: Use union instead of casts
selftests/ftrace: Add a basic testcase for snapshot
tracing: Fix cpu buffers unavailable due to 'record_disabled' missed
|
|
Commit 41320b18a0e0 ("scsi: snic: Fix possible memory leak if device_add()
fails") fixed the memory leak caused by dev_set_name() when device_add()
failed. However, it did not consider that 'tgt' has already been released
when put_device(&tgt->dev) is called. Remove kfree(tgt) in the error path
to avoid double free of 'tgt' and move put_device(&tgt->dev) after the
removed kfree(tgt) to avoid a use-after-free.
Fixes: 41320b18a0e0 ("scsi: snic: Fix possible memory leak if device_add() fails")
Signed-off-by: Zhu Wang <wangzhu9@huawei.com>
Link: https://lore.kernel.org/r/20230819083941.164365-1-wangzhu9@huawei.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media fix from Mauro Carvalho Chehab:
"Fix a potential array out-of-bounds in the mediatek vcodec driver"
* tag 'media/v6.5-4' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
media: vcodec: Fix potential array out-of-bounds in encoder queue_setup
|
|
The raid_component_add() function was added to the kernel tree via patch
"[SCSI] embryonic RAID class" (2005). Remove this function since it never
has had any callers in the Linux kernel. And also raid_component_release()
is only used in raid_component_add(), so it is also removed.
Signed-off-by: Zhu Wang <wangzhu9@huawei.com>
Link: https://lore.kernel.org/r/20230822015254.184270-1-wangzhu9@huawei.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Fixes: 04b5b5cb0136 ("scsi: core: Fix possible memory leak if device_add() fails")
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
|
|
git://anongit.freedesktop.org/drm/drm-intel into drm-fixes
- Fix power consumption at s2idle on DG2 (Anshuman)
- Fix documentation build warning (Jani)
- Fix Display HPD (Imre)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/ZOdPRFSJpo0ErPX/@intel.com
|
|
smaps_pte_hole_lookup() is calling shmem_partial_swap_usage() with page
table lock held: but shmem_partial_swap_usage() does cond_resched_rcu() if
need_resched(): "BUG: sleeping function called from invalid context".
Since shmem_partial_swap_usage() is designed to count across a range, but
smaps_pte_hole_lookup() only calls it for a single page slot, just break
out of the loop on the last or only page, before checking need_resched().
Link: https://lkml.kernel.org/r/6fe3b3ec-abdf-332f-5c23-6a3b3a3b11a9@google.com
Fixes: 230100321518 ("mm/smaps: simplify shmem handling of pte holes")
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Peter Xu <peterx@redhat.com>
Cc: <stable@vger.kernel.org> [5.16+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
The cachestat kselftest runs a test on a normal file, which is created
temporarily in the current directory. Among the tests it runs there is a
call to fsync(), which is expected to clean all dirty pages used by the
file.
However the tmpfs filesystem implements fsync() as noop_fsync(), so the
call will not even attempt to clean anything when this test file happens
to live on a tmpfs instance. This happens in an initramfs, or when the
current directory is in /dev/shm or sometimes /tmp.
To avoid this test failing wrongly, use statfs() to check which filesystem
the test file lives on. If that is "tmpfs", we skip the fsync() test.
Since the fsync test is only one part of the "normal file" test, we now
execute this twice, skipping the fsync part on the first call. This way
only the second test, including the fsync part, would be skipped.
Link: https://lkml.kernel.org/r/20230821160534.3414911-3-andre.przywara@arm.com
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Patch series "selftests: cachestat: fix run on older kernels", v2.
I ran all kernel selftests on some test machine, and stumbled upon
cachestat failing (among others). These patches fix the run on older
kernels and when the current directory is on a tmpfs instance.
This patch (of 2):
As cachestat is a new syscall, it won't be available on older kernels, for
instance those running on a development machine. At the moment the test
reports all tests as "not ok" in this case.
Test for the cachestat syscall availability first, before doing further
tests, and bail out early with a TAP SKIP comment.
This also uses the opportunity to add the proper TAP headers, and add one
check for proper error handling (illegal file descriptor).
Link: https://lkml.kernel.org/r/20230821160534.3414911-1-andre.przywara@arm.com
Link: https://lkml.kernel.org/r/20230821160534.3414911-2-andre.przywara@arm.com
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Nhat Pham <nphamcs@gmail.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
The current implementation of append may cause duplicate data and/or
incorrect ranges to be returned to a reader during an update. Although
this has not been reported or seen, disable the append write operation
while the tree is in rcu mode out of an abundance of caution.
During the analysis of the mas_next_slot() the following was
artificially created by separating the writer and reader code:
Writer: reader:
mas_wr_append
set end pivot
updates end metata
Detects write to last slot
last slot write is to start of slot
store current contents in slot
overwrite old end pivot
mas_next_slot():
read end metadata
read old end pivot
return with incorrect range
store new value
Alternatively:
Writer: reader:
mas_wr_append
set end pivot
updates end metata
Detects write to last slot
last lost write to end of slot
store value
mas_next_slot():
read end metadata
read old end pivot
read new end pivot
return with incorrect range
set old end pivot
There may be other accesses that are not safe since we are now updating
both metadata and pointers, so disabling append if there could be rcu
readers is the safest action.
Link: https://lkml.kernel.org/r/20230819004356.1454718-2-Liam.Howlett@oracle.com
Fixes: 54a611b60590 ("Maple Tree: add new data structure")
Signed-off-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
for sharing check
Commit 98b211d6415f ("madvise: convert madvise_free_pte_range() to use a
folio") replaced the page_mapcount() with folio_mapcount() to check
whether the folio is shared by other mapping.
It's not correct for large folios. folio_mapcount() returns the total
mapcount of large folio which is not suitable to detect whether the folio
is shared.
Use folio_estimated_sharers() which returns a estimated number of shares.
That means it's not 100% correct. It should be OK for madvise case here.
User-visible effects is that the THP is skipped when user call madvise.
But the correct behavior is THP should be split and processed then.
NOTE: this change is a temporary fix to reduce the user-visible effects
before the long term fix from David is ready.
Link: https://lkml.kernel.org/r/20230808020917.2230692-4-fengwei.yin@intel.com
Fixes: 98b211d6415f ("madvise: convert madvise_free_pte_range() to use a folio")
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Reviewed-by: Yu Zhao <yuzhao@google.com>
Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
for sharing check
Commit fc986a38b670 ("mm: huge_memory: convert madvise_free_huge_pmd to
use a folio") replaced the page_mapcount() with folio_mapcount() to check
whether the folio is shared by other mapping.
It's not correct for large folios. folio_mapcount() returns the total
mapcount of large folio which is not suitable to detect whether the folio
is shared.
Use folio_estimated_sharers() which returns a estimated number of shares.
That means it's not 100% correct. It should be OK for madvise case here.
User-visible effects is that the THP is skipped when user call madvise.
But the correct behavior is THP should be split and processed then.
NOTE: this change is a temporary fix to reduce the user-visible effects
before the long term fix from David is ready.
Link: https://lkml.kernel.org/r/20230808020917.2230692-3-fengwei.yin@intel.com
Fixes: fc986a38b670 ("mm: huge_memory: convert madvise_free_huge_pmd to use a folio")
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Reviewed-by: Yu Zhao <yuzhao@google.com>
Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
large folio for sharing check
Patch series "don't use mapcount() to check large folio sharing", v2.
In madvise_cold_or_pageout_pte_range() and madvise_free_pte_range(),
folio_mapcount() is used to check whether the folio is shared. But it's
not correct as folio_mapcount() returns total mapcount of large folio.
Use folio_estimated_sharers() here as the estimated number is enough.
This patchset will fix the cases:
User space application call madvise() with MADV_FREE, MADV_COLD and
MADV_PAGEOUT for specific address range. There are THP mapped to the
range. Without the patchset, the THP is skipped. With the patch, the
THP will be split and handled accordingly.
David reported the cow self test skip some cases because of MADV_PAGEOUT
skip THP:
https://lore.kernel.org/linux-mm/9e92e42d-488f-47db-ac9d-75b24cd0d037@intel.com/T/#mbf0f2ec7fbe45da47526de1d7036183981691e81
and I confirmed this patchset make it work again.
This patch (of 3):
Commit 07e8c82b5eff ("madvise: convert madvise_cold_or_pageout_pte_range()
to use folios") replaced the page_mapcount() with folio_mapcount() to
check whether the folio is shared by other mapping.
It's not correct for large folio. folio_mapcount() returns the total
mapcount of large folio which is not suitable to detect whether the folio
is shared.
Use folio_estimated_sharers() which returns a estimated number of shares.
That means it's not 100% correct. It should be OK for madvise case here.
User-visible effects is that the THP is skipped when user call madvise.
But the correct behavior is THP should be split and processed then.
NOTE: this change is a temporary fix to reduce the user-visible effects
before the long term fix from David is ready.
Link: https://lkml.kernel.org/r/20230808020917.2230692-1-fengwei.yin@intel.com
Link: https://lkml.kernel.org/r/20230808020917.2230692-2-fengwei.yin@intel.com
Fixes: 07e8c82b5eff ("madvise: convert madvise_cold_or_pageout_pte_range() to use folios")
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Reviewed-by: Yu Zhao <yuzhao@google.com>
Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
