mirror of
https://github.com/torvalds/linux.git
synced 2026-06-05 13:06:59 +02:00
b6e6cc1f78
The primary bug Alyssa noticed was that with FineIBT enabled function
prologues have a spurious ENDBR instruction:
__cfi_foo:
endbr64
subl $hash, %r10d
jz 1f
ud2
nop
1:
foo:
endbr64 <--- *sadface*
This means that any indirect call that fails to target the __cfi symbol
and instead targets (the regular old) foo+0, will succeed due to that
second ENDBR.
Fixing this lead to the discovery of a single indirect call that was
still doing this: ret_from_fork(), since that's an assembly stub the
compmiler would not generate the proper kCFI indirect call magic and it
would not get patched.
Brian came up with the most comprehensive fix -- convert the thing to C
with only a very thin asm wrapper. This ensures the kernel thread
boostrap is a proper kCFI call.
While discussing all this, Kees noted that kCFI hashes could/should be
poisoned to seal all functions whose address is never taken, further
limiting the valid kCFI targets -- much like we already do for IBT.
So what was a 'simple' observation and fix cascaded into a bunch of
inter-related CFI infrastructure fixes.
-----BEGIN PGP SIGNATURE-----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=RhPI
-----END PGP SIGNATURE-----
Merge tag 'x86_urgent_for_6.5_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 CFI fixes from Peter Zijlstra:
"Fix kCFI/FineIBT weaknesses
The primary bug Alyssa noticed was that with FineIBT enabled function
prologues have a spurious ENDBR instruction:
__cfi_foo:
endbr64
subl $hash, %r10d
jz 1f
ud2
nop
1:
foo:
endbr64 <--- *sadface*
This means that any indirect call that fails to target the __cfi
symbol and instead targets (the regular old) foo+0, will succeed due
to that second ENDBR.
Fixing this led to the discovery of a single indirect call that was
still doing this: ret_from_fork(). Since that's an assembly stub the
compiler would not generate the proper kCFI indirect call magic and it
would not get patched.
Brian came up with the most comprehensive fix -- convert the thing to
C with only a very thin asm wrapper. This ensures the kernel thread
boostrap is a proper kCFI call.
While discussing all this, Kees noted that kCFI hashes could/should be
poisoned to seal all functions whose address is never taken, further
limiting the valid kCFI targets -- much like we already do for IBT.
So what was a 'simple' observation and fix cascaded into a bunch of
inter-related CFI infrastructure fixes"
* tag 'x86_urgent_for_6.5_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/cfi: Only define poison_cfi() if CONFIG_X86_KERNEL_IBT=y
x86/fineibt: Poison ENDBR at +0
x86: Rewrite ret_from_fork() in C
x86/32: Remove schedule_tail_wrapper()
x86/cfi: Extend ENDBR sealing to kCFI
x86/alternative: Rename apply_ibt_endbr()
x86/cfi: Extend {JMP,CAKK}_NOSPEC comment
|
||
|---|---|---|
| .. | ||
| acpi | ||
| apic | ||
| cpu | ||
| fpu | ||
| kprobes | ||
| .gitignore | ||
| alternative.c | ||
| amd_gart_64.c | ||
| amd_nb.c | ||
| aperture_64.c | ||
| apm_32.c | ||
| asm-offsets_32.c | ||
| asm-offsets_64.c | ||
| asm-offsets.c | ||
| audit_64.c | ||
| bootflag.c | ||
| callthunks.c | ||
| cfi.c | ||
| check.c | ||
| cpuid.c | ||
| crash_core_32.c | ||
| crash_core_64.c | ||
| crash_dump_32.c | ||
| crash_dump_64.c | ||
| crash.c | ||
| devicetree.c | ||
| doublefault_32.c | ||
| dumpstack_32.c | ||
| dumpstack_64.c | ||
| dumpstack.c | ||
| e820.c | ||
| early_printk.c | ||
| early-quirks.c | ||
| ebda.c | ||
| eisa.c | ||
| espfix_64.c | ||
| ftrace_32.S | ||
| ftrace_64.S | ||
| ftrace.c | ||
| head_32.S | ||
| head_64.S | ||
| head32.c | ||
| head64.c | ||
| hpet.c | ||
| hw_breakpoint.c | ||
| i8237.c | ||
| i8253.c | ||
| i8259.c | ||
| idt.c | ||
| io_delay.c | ||
| ioport.c | ||
| irq_32.c | ||
| irq_64.c | ||
| irq_work.c | ||
| irq.c | ||
| irqflags.S | ||
| irqinit.c | ||
| itmt.c | ||
| jailhouse.c | ||
| jump_label.c | ||
| kdebugfs.c | ||
| kexec-bzimage64.c | ||
| kgdb.c | ||
| ksysfs.c | ||
| kvm.c | ||
| kvmclock.c | ||
| ldt.c | ||
| machine_kexec_32.c | ||
| machine_kexec_64.c | ||
| Makefile | ||
| mmconf-fam10h_64.c | ||
| module.c | ||
| mpparse.c | ||
| msr.c | ||
| nmi_selftest.c | ||
| nmi.c | ||
| paravirt-spinlocks.c | ||
| paravirt.c | ||
| pci-dma.c | ||
| pcspeaker.c | ||
| perf_regs.c | ||
| platform-quirks.c | ||
| pmem.c | ||
| probe_roms.c | ||
| process_32.c | ||
| process_64.c | ||
| process.c | ||
| process.h | ||
| ptrace.c | ||
| pvclock.c | ||
| quirks.c | ||
| reboot_fixups_32.c | ||
| reboot.c | ||
| relocate_kernel_32.S | ||
| relocate_kernel_64.S | ||
| resource.c | ||
| rethook.c | ||
| rtc.c | ||
| setup_percpu.c | ||
| setup.c | ||
| sev_verify_cbit.S | ||
| sev-shared.c | ||
| sev.c | ||
| signal_32.c | ||
| signal_64.c | ||
| signal.c | ||
| smp.c | ||
| smpboot.c | ||
| stacktrace.c | ||
| static_call.c | ||
| step.c | ||
| sys_ia32.c | ||
| sys_x86_64.c | ||
| tboot.c | ||
| time.c | ||
| tls.c | ||
| tls.h | ||
| topology.c | ||
| trace_clock.c | ||
| trace.c | ||
| tracepoint.c | ||
| traps.c | ||
| tsc_msr.c | ||
| tsc_sync.c | ||
| tsc.c | ||
| umip.c | ||
| unwind_frame.c | ||
| unwind_guess.c | ||
| unwind_orc.c | ||
| uprobes.c | ||
| verify_cpu.S | ||
| vm86_32.c | ||
| vmlinux.lds.S | ||
| vsmp_64.c | ||
| x86_init.c | ||