diff options
| author | Linus Torvalds <torvalds@linux-foundation.org> | 2026-09-05 11:59:05 -0700 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2026-09-05 11:59:05 -0700 |
| commit | 9f0346dcbea363787186c94ef94dd01aaa215afa (patch) | |
| tree | b81c168f4d0dbe4c7f4750f4598d48a6deb72094 /Documentation/userspace-api/spec_ctrl.rst | |
| download | linux-stable-9f0346dcbea363787186c94ef94dd01aaa215afa.tar.gz linux-stable-9f0346dcbea363787186c94ef94dd01aaa215afa.zip | |
Merge tag 'driver-core-7.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/driver-core/driver-coregrafted
Pull driver core fixes from Danilo Krummrich:
- Fix kernfs listxattr() not returning security xattr names (e.g.
SELinux labels) when the kernfs node has no allocated kernfs_iattrs
- Fix silent truncation of IRQ vector indices in the Rust PCI
abstractions
- Don't select OF from DRIVER_PE_KUNIT_TEST; skip the test when OF is
disabled instead of silently enabling extra kernel functionality
- Russ Weight is retiring from kernel development; update the Firmware
Loader sysfs contact to the driver-core mailing list, add a CREDITS
entry for Firmware Upload, and update MAINTAINERS accordingly
* tag 'driver-core-7.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/driver-core/driver-core:
MAINTAINERS: Remove Russ Weight from Firmware Loader
CREDITS: Add CREDITS entry for Firmware Upload
firmware_loader: Change contact for sysfs nodes
rust: pci: reject IRQ vector indices that do not fit in u32
kernfs: preserve security xattrs without allocating iattrs
drivers: base: test: DRIVER_PE_KUNIT_TEST should not select OF
Diffstat (limited to 'Documentation/userspace-api/spec_ctrl.rst')
| -rw-r--r-- | Documentation/userspace-api/spec_ctrl.rst | 124 |
1 files changed, 124 insertions, 0 deletions
diff --git a/Documentation/userspace-api/spec_ctrl.rst b/Documentation/userspace-api/spec_ctrl.rst new file mode 100644 index 000000000..61fe020b2 --- /dev/null +++ b/Documentation/userspace-api/spec_ctrl.rst @@ -0,0 +1,124 @@ +=================== +Speculation Control +=================== + +Quite some CPUs have speculation-related misfeatures which are in +fact vulnerabilities causing data leaks in various forms even across +privilege domains. + +The kernel provides mitigation for such vulnerabilities in various +forms. Some of these mitigations are compile-time configurable and some +can be supplied on the kernel command line. + +There is also a class of mitigations which are very expensive, but they can +be restricted to a certain set of processes or tasks in controlled +environments. The mechanism to control these mitigations is via +:manpage:`prctl(2)`. + +There are two prctl options which are related to this: + + * PR_GET_SPECULATION_CTRL + + * PR_SET_SPECULATION_CTRL + +PR_GET_SPECULATION_CTRL +----------------------- + +PR_GET_SPECULATION_CTRL returns the state of the speculation misfeature +which is selected with arg2 of prctl(2). The return value uses bits 0-3 with +the following meaning (with the caveat that PR_SPEC_L1D_FLUSH has less obvious +semantics, see documentation for that specific control below): + +==== ====================== ================================================== +Bit Define Description +==== ====================== ================================================== +0 PR_SPEC_PRCTL Mitigation can be controlled per task by + PR_SET_SPECULATION_CTRL. +1 PR_SPEC_ENABLE The speculation feature is enabled, mitigation is + disabled. +2 PR_SPEC_DISABLE The speculation feature is disabled, mitigation is + enabled. +3 PR_SPEC_FORCE_DISABLE Same as PR_SPEC_DISABLE, but cannot be undone. A + subsequent prctl(..., PR_SPEC_ENABLE) will fail. +4 PR_SPEC_DISABLE_NOEXEC Same as PR_SPEC_DISABLE, but the state will be + cleared on :manpage:`execve(2)`. +==== ====================== ================================================== + +If all bits are 0 the CPU is not affected by the speculation misfeature. + +If PR_SPEC_PRCTL is set, then the per-task control of the mitigation is +available. If not set, prctl(PR_SET_SPECULATION_CTRL) for the speculation +misfeature will fail. + +.. _set_spec_ctrl: + +PR_SET_SPECULATION_CTRL +----------------------- + +PR_SET_SPECULATION_CTRL allows to control the speculation misfeature, which +is selected by arg2 of :manpage:`prctl(2)` per task. arg3 is used to hand +in the control value, i.e. either PR_SPEC_ENABLE or PR_SPEC_DISABLE or +PR_SPEC_FORCE_DISABLE. + +Common error codes +------------------ +======= ================================================================= +Value Meaning +======= ================================================================= +EINVAL The prctl is not implemented by the architecture or unused + prctl(2) arguments are not 0. + +ENODEV arg2 is selecting a not supported speculation misfeature. +======= ================================================================= + +PR_SET_SPECULATION_CTRL error codes +----------------------------------- +======= ================================================================= +Value Meaning +======= ================================================================= +0 Success + +ERANGE arg3 is incorrect, i.e. it's neither PR_SPEC_ENABLE nor + PR_SPEC_DISABLE nor PR_SPEC_FORCE_DISABLE. + +ENXIO For PR_SPEC_STORE_BYPASS: control of the selected speculation misfeature + is not possible via prctl, because of the system's boot configuration. + +EPERM Speculation was disabled with PR_SPEC_FORCE_DISABLE and caller tried to + enable it again. + +EPERM For PR_SPEC_L1D_FLUSH and PR_SPEC_INDIRECT_BRANCH: control of the + mitigation is not possible because of the system's boot configuration. + +======= ================================================================= + +Speculation misfeature controls +------------------------------- +- PR_SPEC_STORE_BYPASS: Speculative Store Bypass + + Invocations: + * prctl(PR_GET_SPECULATION_CTRL, PR_SPEC_STORE_BYPASS, 0, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_STORE_BYPASS, PR_SPEC_ENABLE, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_STORE_BYPASS, PR_SPEC_DISABLE, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_STORE_BYPASS, PR_SPEC_FORCE_DISABLE, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_STORE_BYPASS, PR_SPEC_DISABLE_NOEXEC, 0, 0); + +- PR_SPEC_INDIR_BRANCH: Indirect Branch Speculation in User Processes + (Mitigate Spectre V2 style attacks against user processes) + + Invocations: + * prctl(PR_GET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, 0, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_ENABLE, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_DISABLE, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_FORCE_DISABLE, 0, 0); + +- PR_SPEC_L1D_FLUSH: Flush L1D Cache on context switch out of the task + (works only when tasks run on non SMT cores) + +For this control, PR_SPEC_ENABLE means that the **mitigation** is enabled (L1D +is flushed), PR_SPEC_DISABLE means it is disabled. + + Invocations: + * prctl(PR_GET_SPECULATION_CTRL, PR_SPEC_L1D_FLUSH, 0, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_L1D_FLUSH, PR_SPEC_ENABLE, 0, 0); + * prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_L1D_FLUSH, PR_SPEC_DISABLE, 0, 0); |
