Pull misc x86 fixes from Ingo Molnar:
- Fix hypercall detection on Xen guests
- Extend the AMD microcode loader SHA check to Zen5, to block loading
of any unreleased standalone Zen5 microcode patches
- Add new Intel CPU model number for Bartlett Lake
- Fix the workaround for AMD erratum 1054
- Fix buggy early memory acceptance between SEV-SNP guests and the EFI
stub
* tag 'x86-urgent-2025-04-18' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/boot/sev: Avoid shared GHCB page for early memory acceptance
x86/cpu/amd: Fix workaround for erratum 1054
x86/cpu: Add CPU model number for Bartlett Lake CPUs with Raptor Cove cores
x86/microcode/AMD: Extend the SHA check to Zen5, block loading of any unreleased standalone Zen5 microcode patches
x86/xen: Fix __xen_hypercall_setfunc()
Pull timer fix from Ingo Molnar:
"Fix a lockdep false positive in the i8253 driver"
* tag 'timers-urgent-2025-04-18' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/i8253: Call clockevent_i8253_disable() with interrupts disabled
Pull x86 perf event fixes from Ingo Molnar:
"Miscellaneous fixes and a hardware-enabling change:
- Fix Intel uncore PMU IIO free running counters on SPR, ICX and SNR
systems
- Fix Intel PEBS buffer overflow handling
- Fix skid in Intel PEBS sampling of user-space general purpose
registers
- Enable Panther Lake PMU support - similar to Lunar Lake"
* tag 'perf-urgent-2025-04-18' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/intel: Add Panther Lake support
perf/x86/intel: Allow to update user space GPRs from PEBS records
perf/x86/intel: Don't clear perf metrics overflow bit unconditionally
perf/x86/intel/uncore: Fix the scale of IIO free running counters on SPR
perf/x86/intel/uncore: Fix the scale of IIO free running counters on ICX
perf/x86/intel/uncore: Fix the scale of IIO free running counters on SNR
Erratum 1054 affects AMD Zen processors that are a part of Family 17h
Models 00-2Fh and the workaround is to not set HWCR[IRPerfEn]. However,
when X86_FEATURE_ZEN1 was introduced, the condition to detect unaffected
processors was incorrectly changed in a way that the IRPerfEn bit gets
set only for unaffected Zen 1 processors.
Ensure that HWCR[IRPerfEn] is set for all unaffected processors. This
includes a subset of Zen 1 (Family 17h Models 30h and above) and all
later processors. Also clear X86_FEATURE_IRPERF on affected processors
so that the IRPerfCount register is not used by other entities like the
MSR PMU driver.
Fixes: 232afb5578 ("x86/CPU/AMD: Add X86_FEATURE_ZEN1")
Signed-off-by: Sandipan Das <sandipan.das@amd.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/caa057a9d6f8ad579e2f1abaa71efbd5bd4eaf6d.1744956467.git.sandipan.das@amd.com
Pull xen fix from Juergen Gross:
"Just a single fix for the Xen multicall driver avoiding a percpu
variable referencing initdata by its initializer"
* tag 'for-linus-6.15a-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen: fix multicall debug feature
From PMU's perspective, Panther Lake is similar to the previous
generation Lunar Lake. Both are hybrid platforms, with e-core and
p-core.
The key differences are the ARCH PEBS feature and several new events.
The ARCH PEBS is supported in the following patches.
The new events will be supported later in perf tool.
Share the code path with the Lunar Lake. Only update the name.
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20250415114428.341182-2-dapeng1.mi@linux.intel.com
Currently when a user samples user space GPRs (--user-regs option) with
PEBS, the user space GPRs actually always come from software PMI
instead of from PEBS hardware. This leads to the sampled GPRs to
possibly be inaccurate for single PEBS record case because of the
skid between counter overflow and GPRs sampling on PMI.
For the large PEBS case, it is even worse. If user sets the
exclude_kernel attribute, large PEBS would be used to sample user space
GPRs, but since PEBS GPRs group is not really enabled, it leads to all
samples in the large PEBS record to share the same piece of user space
GPRs, like this reproducer shows:
$ perf record -e branches:pu --user-regs=ip,ax -c 100000 ./foo
$ perf report -D | grep "AX"
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
.... AX 0x000000003a0d4ead
So enable GPRs group for user space GPRs sampling and prioritize reading
GPRs from PEBS. If the PEBS sampled GPRs is not user space GPRs (single
PEBS record case), perf_sample_regs_user() modifies them to user space
GPRs.
[ mingo: Clarified the changelog. ]
Fixes: c22497f583 ("perf/x86/intel: Support adaptive PEBS v4")
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20250415104135.318169-2-dapeng1.mi@linux.intel.com
There was a mistake in the SNR uncore spec. The counter increments for
every 32 bytes of data sent from the IO agent to the SOC, not 4 bytes
which was documented in the spec.
The event list has been updated:
"EventName": "UNC_IIO_BANDWIDTH_IN.PART0_FREERUN",
"BriefDescription": "Free running counter that increments for every 32
bytes of data sent from the IO agent to the SOC",
Update the scale of the IIO bandwidth in free running counters as well.
Fixes: 210cc5f9db ("perf/x86/intel/uncore: Add uncore support for Snow Ridge server")
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20250416142426.3933977-1-kan.liang@linux.intel.com
Hypercall detection is failing with xen_hypercall_intel() chosen even on
an AMD processor. Looking at the disassembly, the call to
xen_get_vendor() was removed.
The check for boot_cpu_has(X86_FEATURE_CPUID) was used as a proxy for
the x86_vendor having been set.
When CONFIG_X86_REQUIRED_FEATURE_CPUID=y (the default value), DCE eliminates
the call to xen_get_vendor(). An uninitialized value 0 means
X86_VENDOR_INTEL, so the Intel function is always returned.
Remove the if and always call xen_get_vendor() to avoid this issue.
Fixes: 3d37d9396e ("x86/cpufeatures: Add {REQUIRED,DISABLED} feature configs")
Suggested-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Jason Andryuk <jason.andryuk@amd.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Juergen Gross <jgross@suse.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: "Xin Li (Intel)" <xin@zytor.com>
Link: https://lore.kernel.org/r/20250410193106.16353-1-jason.andryuk@amd.com
Initializing a percpu variable with the address of a struct tagged as
.initdata is breaking the build with CONFIG_SECTION_MISMATCH_WARN_ONLY
not set to "y".
Fix that by using an access function instead returning the .initdata
struct address if the percpu space of the struct hasn't been
allocated yet.
Fixes: 368990a7fe ("xen: fix multicall debug data referencing")
Reported-by: Borislav Petkov <bp@alien8.de>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Acked-by: "Borislav Petkov (AMD)" <bp@alien8.de>
Tested-by: "Borislav Petkov (AMD)" <bp@alien8.de>
Signed-off-by: Juergen Gross <jgross@suse.com>
Message-ID: <20250327190602.26015-1-jgross@suse.com>
There's a lockdep false positive warning related to i8253_lock:
WARNING: HARDIRQ-safe -> HARDIRQ-unsafe lock order detected
...
systemd-sleep/3324 [HC0[0]:SC0[0]:HE0:SE1] is trying to acquire:
ffffffffb2c23398 (i8253_lock){+.+.}-{2:2}, at: pcspkr_event+0x3f/0xe0 [pcspkr]
...
... which became HARDIRQ-irq-unsafe at:
...
lock_acquire+0xd0/0x2f0
_raw_spin_lock+0x30/0x40
clockevent_i8253_disable+0x1c/0x60
pit_timer_init+0x25/0x50
hpet_time_init+0x46/0x50
x86_late_time_init+0x1b/0x40
start_kernel+0x962/0xa00
x86_64_start_reservations+0x24/0x30
x86_64_start_kernel+0xed/0xf0
common_startup_64+0x13e/0x141
...
Lockdep complains due pit_timer_init() using the lock in an IRQ-unsafe
fashion, but it's a false positive, because there is no deadlock
possible at that point due to init ordering: at the point where
pit_timer_init() is called there is no other possible usage of
i8253_lock because the system is still in the very early boot stage
with no interrupts.
But in any case, pit_timer_init() should disable interrupts before
calling clockevent_i8253_disable() out of general principle, and to
keep lockdep working even in this scenario.
Use scoped_guard() for that, as suggested by Thomas Gleixner.
[ mingo: Cleaned up the changelog. ]
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Fernando Fernandez Mancera <ffmancera@riseup.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/Z-uwd4Bnn7FcCShX@gmail.com
Pull misc x86 fixes from Ingo Molnar:
- Fix CPU topology related regression that limited Xen PV guests to a
single CPU
- Fix ancient e820__register_nosave_regions() bugs that were causing
problems with kexec's artificial memory maps
- Fix an S4 hibernation crash caused by two missing ENDBR's that were
mistakenly removed in a recent commit
- Fix a resctrl serialization bug
- Fix early_printk documentation and comments
- Fix RSB bugs, combined with preparatory updates to better match the
code to vendor recommendations.
- Add RSB mitigation document
- Fix/update documentation
- Fix the erratum_1386_microcode[] table to be NULL terminated
* tag 'x86-urgent-2025-04-10' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/ibt: Fix hibernate
x86/cpu: Avoid running off the end of an AMD erratum table
Documentation/x86: Zap the subsection letters
Documentation/x86: Update the naming of CPU features for /proc/cpuinfo
x86/bugs: Add RSB mitigation document
x86/bugs: Don't fill RSB on context switch with eIBRS
x86/bugs: Don't fill RSB on VMEXIT with eIBRS+retpoline
x86/bugs: Fix RSB clearing in indirect_branch_prediction_barrier()
x86/bugs: Use SBPB in write_ibpb() if applicable
x86/bugs: Rename entry_ibpb() to write_ibpb()
x86/early_printk: Use 'mmio32' for consistency, fix comments
x86/resctrl: Fix rdtgroup_mkdir()'s unlocked use of kernfs_node::name
x86/e820: Fix handling of subpage regions when calculating nosave ranges in e820__register_nosave_regions()
x86/acpi: Don't limit CPUs to 1 for Xen PV guests due to disabled ACPI
Pull xen fixes from Juergen Gross:
- A simple fix adding the module description of the Xenbus frontend
module
- A fix correcting the xen-acpi-processor Kconfig dependency for PVH
Dom0 support
- A fix for the Xen balloon driver when running as Xen Dom0 in PVH mode
- A fix for PVH Dom0 in order to avoid problems with CPU idle and
frequency drivers conflicting with Xen
* tag 'for-linus-6.15a-rc2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
x86/xen: disable CPU idle and frequency drivers for PVH dom0
x86/xen: fix balloon target initialization for PVH dom0
xen: Change xen-acpi-processor dom0 dependency
xenbus: add module description
Todd reported, and Len confirmed, that commit 582077c940 ("x86/cfi:
Clean up linkage") broke S4 hiberate on a fair number of machines.
Turns out these machines trip #CP when trying to restore the image.
As it happens, the commit in question removes two ENDBR instructions
in the hibernate code, and clearly got it wrong.
Notably restore_image() does an indirect jump to
relocated_restore_code(), which is a relocated copy of
core_restore_code().
In turn, core_restore_code(), will at the end do an indirect jump to
restore_jump_address (r8), which is pointing at a relocated
restore_registers().
So both sites do indeed need to be ENDBR.
Fixes: 582077c940 ("x86/cfi: Clean up linkage")
Reported-by: Todd Brandt <todd.e.brandt@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Todd Brandt <todd.e.brandt@intel.com>
Tested-by: Len Brown <len.brown@intel.com>
Link: https://bugzilla.kernel.org/show_bug.cgi?id=219998
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=219998
The NULL array terminator at the end of erratum_1386_microcode was
removed during the switch from x86_cpu_desc to x86_cpu_id. This
causes readers to run off the end of the array.
Replace the NULL.
Fixes: f3f3251526 ("x86/cpu: Move AMD erratum 1386 table over to 'x86_cpu_id'")
Reported-by: Jiri Slaby <jirislaby@kernel.org>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
IBPB is expected to clear the RSB. However, if X86_BUG_IBPB_NO_RET is
set, that doesn't happen. Make indirect_branch_prediction_barrier()
take that into account by calling write_ibpb() which clears RSB on
X86_BUG_IBPB_NO_RET:
/* Make sure IBPB clears return stack preductions too. */
FILL_RETURN_BUFFER %rax, RSB_CLEAR_LOOPS, X86_BUG_IBPB_NO_RET
Note that, as of the previous patch, write_ibpb() also reads
'x86_pred_cmd' in order to use SBPB when applicable:
movl _ASM_RIP(x86_pred_cmd), %eax
Therefore that existing behavior in indirect_branch_prediction_barrier()
is not lost.
Fixes: 50e4b3b940 ("x86/entry: Have entry_ibpb() invalidate return predictions")
Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Nikolay Borisov <nik.borisov@suse.com>
Link: https://lore.kernel.org/r/bba68888c511743d4cd65564d1fc41438907523f.1744148254.git.jpoimboe@kernel.org
write_ibpb() does IBPB, which (among other things) flushes branch type
predictions on AMD. If the CPU has SRSO_NO, or if the SRSO mitigation
has been disabled, branch type flushing isn't needed, in which case the
lighter-weight SBPB can be used.
The 'x86_pred_cmd' variable already keeps track of whether IBPB or SBPB
should be used. Use that instead of hardcoding IBPB.
Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/17c5dcd14b29199b75199d67ff7758de9d9a4928.1744148254.git.jpoimboe@kernel.org
First of all, using 'mmio' prevents proper implementation of 8-bit accessors.
Second, it's simply inconsistent with uart8250 set of options. Rename it to
'mmio32'. While at it, remove rather misleading comment in the documentation.
From now on mmio32 is self-explanatory and pciserial supports not only 32-bit
MMIO accessors.
Also, while at it, fix the comment for the "pciserial" case. The comment
seems to be a copy'n'paste error when mentioning "serial" instead of
"pciserial" (with double quotes). Fix this.
With that, move it upper, so we don't calculate 'buf' twice.
Fixes: 3181424aea ("x86/early_printk: Add support for MMIO-based UARTs")
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Denis Mukhin <dmukhin@ford.com>
Link: https://lore.kernel.org/r/20250407172214.792745-1-andriy.shevchenko@linux.intel.com
Since
741c10b096 ("kernfs: Use RCU to access kernfs_node::name.")
a helper rdt_kn_name() that checks that rdtgroup_mutex is held has been used
for all accesses to the kernfs node name.
rdtgroup_mkdir() uses the name to determine if a valid monitor group is being
created by checking the parent name is "mon_groups". This is done without
holding rdtgroup_mutex, and now triggers the following warning:
| WARNING: suspicious RCU usage
| 6.15.0-rc1 #4465 Tainted: G E
| -----------------------------
| arch/x86/kernel/cpu/resctrl/internal.h:408 suspicious rcu_dereference_check() usage!
[...]
| Call Trace:
| <TASK>
| dump_stack_lvl
| lockdep_rcu_suspicious.cold
| is_mon_groups
| rdtgroup_mkdir
| kernfs_iop_mkdir
| vfs_mkdir
| do_mkdirat
| __x64_sys_mkdir
| do_syscall_64
| entry_SYSCALL_64_after_hwframe
Creating a control or monitor group calls mkdir_rdt_prepare(), which uses
rdtgroup_kn_lock_live() to take the rdtgroup_mutex.
To avoid taking and dropping the lock, move the check for the monitor group
name and position into mkdir_rdt_prepare() so that it occurs under
rdtgroup_mutex. Hoist is_mon_groups() earlier in the file.
[ bp: Massage. ]
Fixes: 741c10b096 ("kernfs: Use RCU to access kernfs_node::name.")
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20250407124637.2433230-1-james.morse@arm.com
Pull kvm fixes from Paolo Bonzini:
"ARM:
- Rework heuristics for resolving the fault IPA (HPFAR_EL2 v. re-walk
stage-1 page tables) to align with the architecture. This avoids
possibly taking an SEA at EL2 on the page table walk or using an
architecturally UNKNOWN fault IPA
- Use acquire/release semantics in the KVM FF-A proxy to avoid
reading a stale value for the FF-A version
- Fix KVM guest driver to match PV CPUID hypercall ABI
- Use Inner Shareable Normal Write-Back mappings at stage-1 in KVM
selftests, which is the only memory type for which atomic
instructions are architecturally guaranteed to work
s390:
- Don't use %pK for debug printing and tracepoints
x86:
- Use a separate subclass when acquiring KVM's per-CPU posted
interrupts wakeup lock in the scheduled out path, i.e. when adding
a vCPU on the list of vCPUs to wake, to workaround a false positive
deadlock. The schedule out code runs with a scheduler lock that the
wakeup handler takes in the opposite order; but it does so with
IRQs disabled and cannot run concurrently with a wakeup
- Explicitly zero-initialize on-stack CPUID unions
- Allow building irqbypass.ko as as module when kvm.ko is a module
- Wrap relatively expensive sanity check with KVM_PROVE_MMU
- Acquire SRCU in KVM_GET_MP_STATE to protect guest memory accesses
selftests:
- Add more scenarios to the MONITOR/MWAIT test
- Add option to rseq test to override /dev/cpu_dma_latency
- Bring list of exit reasons up to date
- Cleanup Makefile to list once tests that are valid on all
architectures
Other:
- Documentation fixes"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (26 commits)
KVM: arm64: Use acquire/release to communicate FF-A version negotiation
KVM: arm64: selftests: Explicitly set the page attrs to Inner-Shareable
KVM: arm64: selftests: Introduce and use hardware-definition macros
KVM: VMX: Use separate subclasses for PI wakeup lock to squash false positive
KVM: VMX: Assert that IRQs are disabled when putting vCPU on PI wakeup list
KVM: x86: Explicitly zero-initialize on-stack CPUID unions
KVM: Allow building irqbypass.ko as as module when kvm.ko is a module
KVM: x86/mmu: Wrap sanity check on number of TDP MMU pages with KVM_PROVE_MMU
KVM: selftests: Add option to rseq test to override /dev/cpu_dma_latency
KVM: x86: Acquire SRCU in KVM_GET_MP_STATE to protect guest memory accesses
Documentation: kvm: remove KVM_CAP_MIPS_TE
Documentation: kvm: organize capabilities in the right section
Documentation: kvm: fix some definition lists
Documentation: kvm: drop "Capability" heading from capabilities
Documentation: kvm: give correct name for KVM_CAP_SPAPR_MULTITCE
Documentation: KVM: KVM_GET_SUPPORTED_CPUID now exposes TSC_DEADLINE
selftests: kvm: list once tests that are valid on all architectures
selftests: kvm: bring list of exit reasons up to date
selftests: kvm: revamp MONITOR/MWAIT tests
KVM: arm64: Don't translate FAR if invalid/unsafe
...
When running as a PVH dom0 the ACPI tables exposed to Linux are (mostly)
the native ones, thus exposing the C and P states, that can lead to
attachment of CPU idle and frequency drivers. However the entity in
control of the CPU C and P states is Xen, as dom0 doesn't have a full view
of the system load, neither has all CPUs assigned and identity pinned.
Like it's done for classic PV guests, prevent Linux from using idle or
frequency state drivers when running as a PVH dom0.
On an AMD EPYC 7543P system without this fix a Linux PVH dom0 will keep the
host CPUs spinning at 100% even when dom0 is completely idle, as it's
attempting to use the acpi_idle driver.
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Jason Andryuk <jason.andryuk@amd.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Message-ID: <20250407101842.67228-1-roger.pau@citrix.com>
Objtool uses an arbitrary rule for INSN_SYSCALL and INSN_SYSRET that
almost works by accident: if it's in a function, control flow continues
after the instruction, otherwise it terminates.
That behavior should instead be based on the semantics of the underlying
instruction. Change INSN_SYSCALL to always preserve control flow and
INSN_SYSRET to always terminate it.
The changed semantic for INSN_SYSCALL requires a tweak to the
!CONFIG_IA32_EMULATION version of xen_entry_SYSCALL_compat(). In Xen,
SYSCALL is a hypercall which usually returns. But in this case it's a
hypercall to IRET which doesn't return. Add UD2 to tell objtool to
terminate control flow, and to prevent undefined behavior at runtime.
Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Juergen Gross <jgross@suse.com> # for the Xen part
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/r/19453dfe9a0431b7f016e9dc16d031cad3812a50.1744095216.git.jpoimboe@kernel.org
While debugging kexec/hibernation hangs and crashes, it turned out that
the current implementation of e820__register_nosave_regions() suffers from
multiple serious issues:
- The end of last region is tracked by PFN, causing it to find holes
that aren't there if two consecutive subpage regions are present
- The nosave PFN ranges derived from holes are rounded out (instead of
rounded in) which makes it inconsistent with how explicitly reserved
regions are handled
Fix this by:
- Treating reserved regions as if they were holes, to ensure consistent
handling (rounding out nosave PFN ranges is more correct as the
kernel does not use partial pages)
- Tracking the end of the last RAM region by address instead of pages
to detect holes more precisely
These bugs appear to have been introduced about ~18 years ago with the very
first version of e820_mark_nosave_regions(), and its flawed assumptions were
carried forward uninterrupted through various waves of rewrites and renames.
[ mingo: Added Git archeology details, for kicks and giggles. ]
Fixes: e8eff5ac29 ("[PATCH] Make swsusp avoid memory holes and reserved memory regions on x86_64")
Reported-by: Roberto Ricci <io@r-ricci.it>
Tested-by: Roberto Ricci <io@r-ricci.it>
Signed-off-by: Myrrh Periwinkle <myrrhperiwinkle@qtmlabs.xyz>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: David Woodhouse <dwmw@amazon.co.uk>
Cc: Len Brown <len.brown@intel.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20250406-fix-e820-nosave-v3-1-f3787bc1ee1d@qtmlabs.xyz
Closes: https://lore.kernel.org/all/Z4WFjBVHpndct7br@desktop0a/
Xen disables ACPI for PV guests in DomU, which causes acpi_mps_check() to
return 1 when CONFIG_X86_MPPARSE is not set. As a result, the local APIC is
disabled and the guest is later limited to a single vCPU, despite being
configured with more.
This regression was introduced in version 6.9 in commit 7c0edad364
("x86/cpu/topology: Rework possible CPU management"), which added an
early check that limits CPUs to 1 if apic_is_disabled.
Update the acpi_mps_check() logic to return 0 early when running as a Xen
PV guest in DomU, preventing APIC from being disabled in this specific case
and restoring correct multi-vCPU behaviour.
Fixes: 7c0edad364 ("x86/cpu/topology: Rework possible CPU management")
Signed-off-by: Petr Vaněk <arkamar@atlas.cz>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250407132445.6732-2-arkamar@atlas.cz
PVH dom0 re-uses logic from PV dom0, in which RAM ranges not assigned to
dom0 are re-used as scratch memory to map foreign and grant pages. Such
logic relies on reporting those unpopulated ranges as RAM to Linux, and
mark them as reserved. This way Linux creates the underlying page
structures required for metadata management.
Such approach works fine on PV because the initial balloon target is
calculated using specific Xen data, that doesn't take into account the
memory type changes described above. However on HVM and PVH the initial
balloon target is calculated using get_num_physpages(), and that function
does take into account the unpopulated RAM regions used as scratch space
for remote domain mappings.
This leads to PVH dom0 having an incorrect initial balloon target, which
causes malfunction (excessive memory freeing) of the balloon driver if the
dom0 memory target is later adjusted from the toolstack.
Fix this by using xen_released_pages to account for any pages that are part
of the memory map, but are already unpopulated when the balloon driver is
initialized. This accounts for any regions used for scratch remote
mappings. Note on x86 xen_released_pages definition is moved to
enlighten.c so it's uniformly available for all Xen-enabled builds.
Take the opportunity to unify PV with PVH/HVM guests regarding the usage of
get_num_physpages(), as that avoids having to add different logic for PV vs
PVH in both balloon_add_regions() and arch_xen_unpopulated_init().
Much like a6aa4eb994, the code in this changeset should have been part of
38620fc4e8.
Fixes: a6aa4eb994 ('xen/x86: add extra pages to unpopulated-alloc if available')
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Cc: stable@vger.kernel.org
Signed-off-by: Juergen Gross <jgross@suse.com>
Message-ID: <20250407082838.65495-1-roger.pau@citrix.com>
Pull timer cleanups from Thomas Gleixner:
"A set of final cleanups for the timer subsystem:
- Convert all del_timer[_sync]() instances over to the new
timer_delete[_sync]() API and remove the legacy wrappers.
Conversion was done with coccinelle plus some manual fixups as
coccinelle chokes on scoped_guard().
- The final cleanup of the hrtimer_init() to hrtimer_setup()
conversion.
This has been delayed to the end of the merge window, so that all
patches which have been merged through other trees are in mainline
and all new users are catched.
Doing this right before rc1 ensures that new code which is merged post
rc1 is not introducing new instances of the original functionality"
* tag 'timers-cleanups-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
tracing/timers: Rename the hrtimer_init event to hrtimer_setup
hrtimers: Rename debug_init_on_stack() to debug_setup_on_stack()
hrtimers: Rename debug_init() to debug_setup()
hrtimers: Rename __hrtimer_init_sleeper() to __hrtimer_setup_sleeper()
hrtimers: Remove unnecessary NULL check in hrtimer_start_range_ns()
hrtimers: Make callback function pointer private
hrtimers: Merge __hrtimer_init() into __hrtimer_setup()
hrtimers: Switch to use __htimer_setup()
hrtimers: Delete hrtimer_init()
treewide: Convert new and leftover hrtimer_init() users
treewide: Switch/rename to timer_delete[_sync]()
Pull more irq updates from Thomas Gleixner:
"A set of updates for the interrupt subsystem:
- A treewide cleanup for the irq_domain code, which makes the naming
consistent and gets rid of the original oddity of naming domains
'host'.
This is a trivial mechanical change and is done late to ensure that
all instances have been catched and new code merged post rc1 wont
reintroduce new instances.
- A trivial consistency fix in the migration code
The recent introduction of irq_force_complete_move() in the core
code, causes a problem for the nostalgia crowd who maintains ia64
out of tree.
The code assumes that hierarchical interrupt domains are enabled
and dereferences irq_data::parent_data unconditionally. That works
in mainline because both architectures which enable that code have
hierarchical domains enabled. Though it breaks the ia64 build,
which enables the functionality, but does not have hierarchical
domains.
While it's not really a problem for mainline today, this
unconditional dereference is inconsistent and trivially fixable by
using the existing helper function irqd_get_parent_data(), which
has the appropriate #ifdeffery in place"
* tag 'irq-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
genirq/migration: Use irqd_get_parent_data() in irq_force_complete_move()
irqdomain: Stop using 'host' for domain
irqdomain: Rename irq_get_default_host() to irq_get_default_domain()
irqdomain: Rename irq_set_default_host() to irq_set_default_domain()
Pull Kbuild updates from Masahiro Yamada:
- Improve performance in gendwarfksyms
- Remove deprecated EXTRA_*FLAGS and KBUILD_ENABLE_EXTRA_GCC_CHECKS
- Support CONFIG_HEADERS_INSTALL for ARCH=um
- Use more relative paths to sources files for better reproducibility
- Support the loong64 Debian architecture
- Add Kbuild bash completion
- Introduce intermediate vmlinux.unstripped for architectures that need
static relocations to be stripped from the final vmlinux
- Fix versioning in Debian packages for -rc releases
- Treat missing MODULE_DESCRIPTION() as an error
- Convert Nios2 Makefiles to use the generic rule for built-in DTB
- Add debuginfo support to the RPM package
* tag 'kbuild-v6.15' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild: (40 commits)
kbuild: rpm-pkg: build a debuginfo RPM
kconfig: merge_config: use an empty file as initfile
nios2: migrate to the generic rule for built-in DTB
rust: kbuild: skip `--remap-path-prefix` for `rustdoc`
kbuild: pacman-pkg: hardcode module installation path
kbuild: deb-pkg: don't set KBUILD_BUILD_VERSION unconditionally
modpost: require a MODULE_DESCRIPTION()
kbuild: make all file references relative to source root
x86: drop unnecessary prefix map configuration
kbuild: deb-pkg: add comment about future removal of KDEB_COMPRESS
kbuild: Add a help message for "headers"
kbuild: deb-pkg: remove "version" variable in mkdebian
kbuild: deb-pkg: fix versioning for -rc releases
Documentation/kbuild: Fix indentation in modules.rst example
x86: Get rid of Makefile.postlink
kbuild: Create intermediate vmlinux build with relocations preserved
kbuild: Introduce Kconfig symbol for linking vmlinux with relocations
kbuild: link-vmlinux.sh: Make output file name configurable
kbuild: do not generate .tmp_vmlinux*.map when CONFIG_VMLINUX_MAP=y
Revert "kheaders: Ignore silly-rename files"
...
timer_delete[_sync]() replaces del_timer[_sync](). Convert the whole tree
over and remove the historical wrapper inlines.
Conversion was done with coccinelle plus manual fixups where necessary.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull x86 fixes from Ingo Molnar:
- Fix a performance regression on AMD iGPU and dGPU drivers, related to
the unintended activation of DMA bounce buffers that regressed game
performance if KASLR disturbed things just enough
- Fix a copy_user_generic() performance regression on certain older
non-FSRM/ERMS CPUs
- Fix a Clang build warning due to a semantic merge conflict the Kunit
tree generated with the x86 tree
- Fix FRED related system hang during S4 resume
- Remove an unused API
* tag 'x86-urgent-2025-04-04' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/fred: Fix system hang during S4 resume with FRED enabled
x86/platform/iosf_mbi: Remove unused iosf_mbi_unregister_pmic_bus_access_notifier()
x86/mm/init: Handle the special case of device private pages in add_pages(), to not increase max_pfn and trigger dma_addressing_limited() bounce buffers
x86/tools: Drop duplicate unlikely() definition in insn_decoder_test.c
x86/uaccess: Improve performance by aligning writes to 8 bytes in copy_user_generic(), on non-FSRM/ERMS CPUs
Use a separate subclass when acquiring KVM's per-CPU posted interrupts
wakeup lock in the scheduled out path, i.e. when adding a vCPU on the list
of vCPUs to wake, to workaround a false positive deadlock.
Chain exists of:
&p->pi_lock --> &rq->__lock --> &per_cpu(wakeup_vcpus_on_cpu_lock, cpu)
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(&per_cpu(wakeup_vcpus_on_cpu_lock, cpu));
lock(&rq->__lock);
lock(&per_cpu(wakeup_vcpus_on_cpu_lock, cpu));
lock(&p->pi_lock);
*** DEADLOCK ***
In the wakeup handler, the callchain is *always*:
sysvec_kvm_posted_intr_wakeup_ipi()
|
--> pi_wakeup_handler()
|
--> kvm_vcpu_wake_up()
|
--> try_to_wake_up(),
and the lock order is:
&per_cpu(wakeup_vcpus_on_cpu_lock, cpu) --> &p->pi_lock.
For the schedule out path, the callchain is always (for all intents and
purposes; if the kernel is preemptible, kvm_sched_out() can be called from
something other than schedule(), but the beginning of the callchain will
be the same point in vcpu_block()):
vcpu_block()
|
--> schedule()
|
--> kvm_sched_out()
|
--> vmx_vcpu_put()
|
--> vmx_vcpu_pi_put()
|
--> pi_enable_wakeup_handler()
and the lock order is:
&rq->__lock --> &per_cpu(wakeup_vcpus_on_cpu_lock, cpu)
I.e. lockdep sees AB+BC ordering for schedule out, and CA ordering for
wakeup, and complains about the A=>C versus C=>A inversion. In practice,
deadlock can't occur between schedule out and the wakeup handler as they
are mutually exclusive. The entirely of the schedule out code that runs
with the problematic scheduler locks held, does so with IRQs disabled,
i.e. can't run concurrently with the wakeup handler.
Use a subclass instead disabling lockdep entirely, and tell lockdep that
both subclasses are being acquired when loading a vCPU, as the sched_out
and sched_in paths are NOT mutually exclusive, e.g.
CPU 0 CPU 1
--------------- ---------------
vCPU0 sched_out
vCPU1 sched_in
vCPU1 sched_out vCPU 0 sched_in
where vCPU0's sched_in may race with vCPU1's sched_out, on CPU 0's wakeup
list+lock.
Signed-off-by: Yan Zhao <yan.y.zhao@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-ID: <20250401154727.835231-3-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Assert that IRQs are already disabled when putting a vCPU on a CPU's PI
wakeup list, as opposed to saving/disabling+restoring IRQs. KVM relies on
IRQs being disabled until the vCPU task is fully scheduled out, i.e. until
the scheduler has dropped all of its per-CPU locks (e.g. for the runqueue),
as attempting to wake the task while it's being scheduled out could lead
to deadlock.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Yan Zhao <yan.y.zhao@intel.com>
Message-ID: <20250401154727.835231-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Wrap the TDP MMU page counter in CONFIG_KVM_PROVE_MMU so that the sanity
check is omitted from production builds, and more importantly to remove
the atomic accesses to account pages. A one-off memory leak in production
is relatively uninteresting, and a WARN_ON won't help mitigate a systemic
issue; it's as much about helping triage memory leaks as it is about
detecting them in the first place, and doesn't magically stop the leaks.
I.e. production environments will be quite sad if a severe KVM bug escapes,
regardless of whether or not KVM WARNs.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20250315023448.2358456-1-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Acquire a lock on kvm->srcu when userspace is getting MP state to handle a
rather extreme edge case where "accepting" APIC events, i.e. processing
pending INIT or SIPI, can trigger accesses to guest memory. If the vCPU
is in L2 with INIT *and* a TRIPLE_FAULT request pending, then getting MP
state will trigger a nested VM-Exit by way of ->check_nested_events(), and
emuating the nested VM-Exit can access guest memory.
The splat was originally hit by syzkaller on a Google-internal kernel, and
reproduced on an upstream kernel by hacking the triple_fault_event_test
selftest to stuff a pending INIT, store an MSR on VM-Exit (to generate a
memory access on VMX), and do vcpu_mp_state_get() to trigger the scenario.
=============================
WARNING: suspicious RCU usage
6.14.0-rc3-b112d356288b-vmx/pi_lockdep_false_pos-lock #3 Not tainted
-----------------------------
include/linux/kvm_host.h:1058 suspicious rcu_dereference_check() usage!
other info that might help us debug this:
rcu_scheduler_active = 2, debug_locks = 1
1 lock held by triple_fault_ev/1256:
#0: ffff88810df5a330 (&vcpu->mutex){+.+.}-{4:4}, at: kvm_vcpu_ioctl+0x8b/0x9a0 [kvm]
stack backtrace:
CPU: 11 UID: 1000 PID: 1256 Comm: triple_fault_ev Not tainted 6.14.0-rc3-b112d356288b-vmx #3
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
Call Trace:
<TASK>
dump_stack_lvl+0x7f/0x90
lockdep_rcu_suspicious+0x144/0x190
kvm_vcpu_gfn_to_memslot+0x156/0x180 [kvm]
kvm_vcpu_read_guest+0x3e/0x90 [kvm]
read_and_check_msr_entry+0x2e/0x180 [kvm_intel]
__nested_vmx_vmexit+0x550/0xde0 [kvm_intel]
kvm_check_nested_events+0x1b/0x30 [kvm]
kvm_apic_accept_events+0x33/0x100 [kvm]
kvm_arch_vcpu_ioctl_get_mpstate+0x30/0x1d0 [kvm]
kvm_vcpu_ioctl+0x33e/0x9a0 [kvm]
__x64_sys_ioctl+0x8b/0xb0
do_syscall_64+0x6c/0x170
entry_SYSCALL_64_after_hwframe+0x4b/0x53
</TASK>
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20250401150504.829812-1-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Pull more MM updates from Andrew Morton:
- The series "mm: fixes for fallouts from mem_init() cleanup" from Mike
Rapoport fixes a couple of issues with the just-merged "arch, mm:
reduce code duplication in mem_init()" series
- The series "MAINTAINERS: add my isub-entries to MM part." from Mike
Rapoport does some maintenance on MAINTAINERS
- The series "remove tlb_remove_page_ptdesc()" from Qi Zheng does some
cleanup work to the page mapping code
- The series "mseal system mappings" from Jeff Xu permits sealing of
"system mappings", such as vdso, vvar, vvar_vclock, vectors (arm
compat-mode), sigpage (arm compat-mode)
- Plus the usual shower of singleton patches
* tag 'mm-stable-2025-04-02-22-07' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (31 commits)
mseal sysmap: add arch-support txt
mseal sysmap: enable s390
selftest: test system mappings are sealed
mseal sysmap: update mseal.rst
mseal sysmap: uprobe mapping
mseal sysmap: enable arm64
mseal sysmap: enable x86-64
mseal sysmap: generic vdso vvar mapping
selftests: x86: test_mremap_vdso: skip if vdso is msealed
mseal sysmap: kernel config and header change
mm: pgtable: remove tlb_remove_page_ptdesc()
x86: pgtable: convert to use tlb_remove_ptdesc()
riscv: pgtable: unconditionally use tlb_remove_ptdesc()
mm: pgtable: convert some architectures to use tlb_remove_ptdesc()
mm: pgtable: change pt parameter of tlb_remove_ptdesc() to struct ptdesc*
mm: pgtable: make generic tlb_remove_table() use struct ptdesc
microblaze/mm: put mm_cmdline_setup() in .init.text section
mm/memory_hotplug: fix call folio_test_large with tail page in do_migrate_range
MAINTAINERS: mm: add entry for secretmem
MAINTAINERS: mm: add entry for numa memblocks and numa emulation
...
