July 2018 - Linux-stable-mirror

[PATCH 4.4.y 077/101] seccomp: Add filter flag to opt-out of SSB mitigation

by Srivatsa S. Bhat

From: Kees Cook <keescook(a)chromium.org> commit 00a02d0c502a06d15e07b857f8ff921e3e402675 upstream If a seccomp user is not interested in Speculative Store Bypass mitigation by default, it can set the new SECCOMP_FILTER_FLAG_SPEC_ALLOW flag when adding filters. Signed-off-by: Kees Cook <keescook(a)chromium.org> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- include/linux/seccomp.h | 3 + include/uapi/linux/seccomp.h | 4 + kernel/seccomp.c | 19 ++++-- tools/testing/selftests/seccomp/seccomp_bpf.c | 78 +++++++++++++++++++++++++ 4 files changed, 93 insertions(+), 11 deletions(-) diff --git a/include/linux/seccomp.h b/include/linux/seccomp.h index 2296e6b..5a53d34 100644 --- a/include/linux/seccomp.h +++ b/include/linux/seccomp.h @@ -3,7 +3,8 @@ #include <uapi/linux/seccomp.h> -#define SECCOMP_FILTER_FLAG_MASK (SECCOMP_FILTER_FLAG_TSYNC) +#define SECCOMP_FILTER_FLAG_MASK (SECCOMP_FILTER_FLAG_TSYNC | \ + SECCOMP_FILTER_FLAG_SPEC_ALLOW) #ifdef CONFIG_SECCOMP diff --git a/include/uapi/linux/seccomp.h b/include/uapi/linux/seccomp.h index 0f238a4..e4acb61 100644 --- a/include/uapi/linux/seccomp.h +++ b/include/uapi/linux/seccomp.h @@ -15,7 +15,9 @@ #define SECCOMP_SET_MODE_FILTER 1 /* Valid flags for SECCOMP_SET_MODE_FILTER */ -#define SECCOMP_FILTER_FLAG_TSYNC 1 +#define SECCOMP_FILTER_FLAG_TSYNC (1UL << 0) +/* In v4.14+ SECCOMP_FILTER_FLAG_LOG is (1UL << 1) */ +#define SECCOMP_FILTER_FLAG_SPEC_ALLOW (1UL << 2) /* * All BPF programs must return a 32-bit value. diff --git a/kernel/seccomp.c b/kernel/seccomp.c index f33539f..4bb8a5a 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -230,7 +230,8 @@ static inline void spec_mitigate(struct task_struct *task, } static inline void seccomp_assign_mode(struct task_struct *task, - unsigned long seccomp_mode) + unsigned long seccomp_mode, + unsigned long flags) { assert_spin_locked(&task->sighand->siglock); @@ -240,8 +241,9 @@ static inline void seccomp_assign_mode(struct task_struct *task, * filter) is set. */ smp_mb__before_atomic(); - /* Assume seccomp processes want speculation flaw mitigation. */ - spec_mitigate(task, PR_SPEC_STORE_BYPASS); + /* Assume default seccomp processes want spec flaw mitigation. */ + if ((flags & SECCOMP_FILTER_FLAG_SPEC_ALLOW) == 0) + spec_mitigate(task, PR_SPEC_STORE_BYPASS); set_tsk_thread_flag(task, TIF_SECCOMP); } @@ -309,7 +311,7 @@ static inline pid_t seccomp_can_sync_threads(void) * without dropping the locks. * */ -static inline void seccomp_sync_threads(void) +static inline void seccomp_sync_threads(unsigned long flags) { struct task_struct *thread, *caller; @@ -350,7 +352,8 @@ static inline void seccomp_sync_threads(void) * allow one thread to transition the other. */ if (thread->seccomp.mode == SECCOMP_MODE_DISABLED) - seccomp_assign_mode(thread, SECCOMP_MODE_FILTER); + seccomp_assign_mode(thread, SECCOMP_MODE_FILTER, + flags); } } @@ -469,7 +472,7 @@ static long seccomp_attach_filter(unsigned int flags, /* Now that the new filter is in place, synchronize to all threads. */ if (flags & SECCOMP_FILTER_FLAG_TSYNC) - seccomp_sync_threads(); + seccomp_sync_threads(flags); return 0; } @@ -764,7 +767,7 @@ static long seccomp_set_mode_strict(void) #ifdef TIF_NOTSC disable_TSC(); #endif - seccomp_assign_mode(current, seccomp_mode); + seccomp_assign_mode(current, seccomp_mode, 0); ret = 0; out: @@ -822,7 +825,7 @@ static long seccomp_set_mode_filter(unsigned int flags, /* Do not free the successfully attached filter. */ prepared = NULL; - seccomp_assign_mode(current, seccomp_mode); + seccomp_assign_mode(current, seccomp_mode, flags); out: spin_unlock_irq(&current->sighand->siglock); if (flags & SECCOMP_FILTER_FLAG_TSYNC) diff --git a/tools/testing/selftests/seccomp/seccomp_bpf.c b/tools/testing/selftests/seccomp/seccomp_bpf.c index 29487e0..b3f3454 100644 --- a/tools/testing/selftests/seccomp/seccomp_bpf.c +++ b/tools/testing/selftests/seccomp/seccomp_bpf.c @@ -1477,7 +1477,11 @@ TEST_F(TRACE_syscall, syscall_dropped) #endif #ifndef SECCOMP_FILTER_FLAG_TSYNC -#define SECCOMP_FILTER_FLAG_TSYNC 1 +#define SECCOMP_FILTER_FLAG_TSYNC (1UL << 0) +#endif + +#ifndef SECCOMP_FILTER_FLAG_SPEC_ALLOW +#define SECCOMP_FILTER_FLAG_SPEC_ALLOW (1UL << 2) #endif #ifndef seccomp @@ -1576,6 +1580,78 @@ TEST(seccomp_syscall_mode_lock) } } +/* + * Test detection of known and unknown filter flags. Userspace needs to be able + * to check if a filter flag is supported by the current kernel and a good way + * of doing that is by attempting to enter filter mode, with the flag bit in + * question set, and a NULL pointer for the _args_ parameter. EFAULT indicates + * that the flag is valid and EINVAL indicates that the flag is invalid. + */ +TEST(detect_seccomp_filter_flags) +{ + unsigned int flags[] = { SECCOMP_FILTER_FLAG_TSYNC, + SECCOMP_FILTER_FLAG_SPEC_ALLOW }; + unsigned int flag, all_flags; + int i; + long ret; + + /* Test detection of known-good filter flags */ + for (i = 0, all_flags = 0; i < ARRAY_SIZE(flags); i++) { + int bits = 0; + + flag = flags[i]; + /* Make sure the flag is a single bit! */ + while (flag) { + if (flag & 0x1) + bits ++; + flag >>= 1; + } + ASSERT_EQ(1, bits); + flag = flags[i]; + + ret = seccomp(SECCOMP_SET_MODE_FILTER, flag, NULL); + ASSERT_NE(ENOSYS, errno) { + TH_LOG("Kernel does not support seccomp syscall!"); + } + EXPECT_EQ(-1, ret); + EXPECT_EQ(EFAULT, errno) { + TH_LOG("Failed to detect that a known-good filter flag (0x%X) is supported!", + flag); + } + + all_flags |= flag; + } + + /* Test detection of all known-good filter flags */ + ret = seccomp(SECCOMP_SET_MODE_FILTER, all_flags, NULL); + EXPECT_EQ(-1, ret); + EXPECT_EQ(EFAULT, errno) { + TH_LOG("Failed to detect that all known-good filter flags (0x%X) are supported!", + all_flags); + } + + /* Test detection of an unknown filter flag */ + flag = -1; + ret = seccomp(SECCOMP_SET_MODE_FILTER, flag, NULL); + EXPECT_EQ(-1, ret); + EXPECT_EQ(EINVAL, errno) { + TH_LOG("Failed to detect that an unknown filter flag (0x%X) is unsupported!", + flag); + } + + /* + * Test detection of an unknown filter flag that may simply need to be + * added to this test + */ + flag = flags[ARRAY_SIZE(flags) - 1] << 1; + ret = seccomp(SECCOMP_SET_MODE_FILTER, flag, NULL); + EXPECT_EQ(-1, ret); + EXPECT_EQ(EINVAL, errno) { + TH_LOG("Failed to detect that an unknown filter flag (0x%X) is unsupported! Does a new flag need to be added to this test?", + flag); + } +} + TEST(TSYNC_first) { struct sock_filter filter[] = {

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 076/101] seccomp: Use PR_SPEC_FORCE_DISABLE

by Srivatsa S. Bhat

From: Thomas Gleixner <tglx(a)linutronix.de> commit b849a812f7eb92e96d1c8239b06581b2cfd8b275 upstream Use PR_SPEC_FORCE_DISABLE in seccomp() because seccomp does not allow to widen restrictions. Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- kernel/seccomp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/seccomp.c b/kernel/seccomp.c index bfb1ee8..f33539f 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -226,7 +226,7 @@ static inline void spec_mitigate(struct task_struct *task, int state = arch_prctl_spec_ctrl_get(task, which); if (state > 0 && (state & PR_SPEC_PRCTL)) - arch_prctl_spec_ctrl_set(task, which, PR_SPEC_DISABLE); + arch_prctl_spec_ctrl_set(task, which, PR_SPEC_FORCE_DISABLE); } static inline void seccomp_assign_mode(struct task_struct *task,

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 075/101] prctl: Add force disable speculation

by Srivatsa S. Bhat

From: Thomas Gleixner <tglx(a)linutronix.de> commit 356e4bfff2c5489e016fdb925adbf12a1e3950ee upstream For certain use cases it is desired to enforce mitigations so they cannot be undone afterwards. That's important for loader stubs which want to prevent a child from disabling the mitigation again. Will also be used for seccomp(). The extra state preserving of the prctl state for SSB is a preparatory step for EBPF dymanic speculation control. Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- Documentation/spec_ctrl.txt | 34 +++++++++++++++++++++------------- arch/x86/kernel/cpu/bugs.c | 35 +++++++++++++++++++++++++---------- fs/proc/array.c | 3 +++ include/linux/sched.h | 9 +++++++++ include/uapi/linux/prctl.h | 1 + 5 files changed, 59 insertions(+), 23 deletions(-) diff --git a/Documentation/spec_ctrl.txt b/Documentation/spec_ctrl.txt index ddbebcd..1b3690d 100644 --- a/Documentation/spec_ctrl.txt +++ b/Documentation/spec_ctrl.txt @@ -25,19 +25,21 @@ 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 0-2 with +which is selected with arg2 of prctl(2). The return value uses bits 0-3 with the following meaning: -==== ================ =================================================== -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 -==== ================ =================================================== +==== ===================== =================================================== +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. +==== ===================== =================================================== If all bits are 0 the CPU is not affected by the speculation misfeature. @@ -47,9 +49,11 @@ misfeature will fail. 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. +in the control value, i.e. either PR_SPEC_ENABLE or PR_SPEC_DISABLE or +PR_SPEC_FORCE_DISABLE. Common error codes ------------------ @@ -70,10 +74,13 @@ Value Meaning 0 Success ERANGE arg3 is incorrect, i.e. it's neither PR_SPEC_ENABLE nor - PR_SPEC_DISABLE + PR_SPEC_DISABLE nor PR_SPEC_FORCE_DISABLE ENXIO Control of the selected speculation misfeature is not possible. See PR_GET_SPECULATION_CTRL. + +EPERM Speculation was disabled with PR_SPEC_FORCE_DISABLE and caller + tried to enable it again. ======= ================================================================= Speculation misfeature controls @@ -84,3 +91,4 @@ Speculation misfeature controls * 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); diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c index 64b54a4..d6897ca 100644 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -531,21 +531,37 @@ static void ssb_select_mitigation() static int ssb_prctl_set(struct task_struct *task, unsigned long ctrl) { - bool rds = !!test_tsk_thread_flag(task, TIF_RDS); + bool update; if (ssb_mode != SPEC_STORE_BYPASS_PRCTL) return -ENXIO; - if (ctrl == PR_SPEC_ENABLE) - clear_tsk_thread_flag(task, TIF_RDS); - else - set_tsk_thread_flag(task, TIF_RDS); + switch (ctrl) { + case PR_SPEC_ENABLE: + /* If speculation is force disabled, enable is not allowed */ + if (task_spec_ssb_force_disable(task)) + return -EPERM; + task_clear_spec_ssb_disable(task); + update = test_and_clear_tsk_thread_flag(task, TIF_RDS); + break; + case PR_SPEC_DISABLE: + task_set_spec_ssb_disable(task); + update = !test_and_set_tsk_thread_flag(task, TIF_RDS); + break; + case PR_SPEC_FORCE_DISABLE: + task_set_spec_ssb_disable(task); + task_set_spec_ssb_force_disable(task); + update = !test_and_set_tsk_thread_flag(task, TIF_RDS); + break; + default: + return -ERANGE; + } /* * If being set on non-current task, delay setting the CPU * mitigation until it is next scheduled. */ - if (task == current && rds != !!test_tsk_thread_flag(task, TIF_RDS)) + if (task == current && update) speculative_store_bypass_update(); return 0; @@ -557,7 +573,9 @@ static int ssb_prctl_get(struct task_struct *task) case SPEC_STORE_BYPASS_DISABLE: return PR_SPEC_DISABLE; case SPEC_STORE_BYPASS_PRCTL: - if (test_tsk_thread_flag(task, TIF_RDS)) + if (task_spec_ssb_force_disable(task)) + return PR_SPEC_PRCTL | PR_SPEC_FORCE_DISABLE; + if (task_spec_ssb_disable(task)) return PR_SPEC_PRCTL | PR_SPEC_DISABLE; return PR_SPEC_PRCTL | PR_SPEC_ENABLE; default: @@ -570,9 +588,6 @@ static int ssb_prctl_get(struct task_struct *task) int arch_prctl_spec_ctrl_set(struct task_struct *task, unsigned long which, unsigned long ctrl) { - if (ctrl != PR_SPEC_ENABLE && ctrl != PR_SPEC_DISABLE) - return -ERANGE; - switch (which) { case PR_SPEC_STORE_BYPASS: return ssb_prctl_set(task, ctrl); diff --git a/fs/proc/array.c b/fs/proc/array.c index bb48358..3141478 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -341,6 +341,9 @@ static inline void task_seccomp(struct seq_file *m, struct task_struct *p) case PR_SPEC_NOT_AFFECTED: seq_printf(m, "not vulnerable"); break; + case PR_SPEC_PRCTL | PR_SPEC_FORCE_DISABLE: + seq_printf(m, "thread force mitigated"); + break; case PR_SPEC_PRCTL | PR_SPEC_DISABLE: seq_printf(m, "thread mitigated"); break; diff --git a/include/linux/sched.h b/include/linux/sched.h index 90bea39..725498c 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2167,6 +2167,8 @@ static inline void memalloc_noio_restore(unsigned int flags) #define PFA_NO_NEW_PRIVS 0 /* May not gain new privileges. */ #define PFA_SPREAD_PAGE 1 /* Spread page cache over cpuset */ #define PFA_SPREAD_SLAB 2 /* Spread some slab caches over cpuset */ +#define PFA_SPEC_SSB_DISABLE 4 /* Speculative Store Bypass disabled */ +#define PFA_SPEC_SSB_FORCE_DISABLE 5 /* Speculative Store Bypass force disabled*/ #define TASK_PFA_TEST(name, func) \ @@ -2190,6 +2192,13 @@ TASK_PFA_TEST(SPREAD_SLAB, spread_slab) TASK_PFA_SET(SPREAD_SLAB, spread_slab) TASK_PFA_CLEAR(SPREAD_SLAB, spread_slab) +TASK_PFA_TEST(SPEC_SSB_DISABLE, spec_ssb_disable) +TASK_PFA_SET(SPEC_SSB_DISABLE, spec_ssb_disable) +TASK_PFA_CLEAR(SPEC_SSB_DISABLE, spec_ssb_disable) + +TASK_PFA_TEST(SPEC_SSB_FORCE_DISABLE, spec_ssb_force_disable) +TASK_PFA_SET(SPEC_SSB_FORCE_DISABLE, spec_ssb_force_disable) + /* * task->jobctl flags */ diff --git a/include/uapi/linux/prctl.h b/include/uapi/linux/prctl.h index 3b316be..64776b7 100644 --- a/include/uapi/linux/prctl.h +++ b/include/uapi/linux/prctl.h @@ -207,5 +207,6 @@ struct prctl_mm_map { # define PR_SPEC_PRCTL (1UL << 0) # define PR_SPEC_ENABLE (1UL << 1) # define PR_SPEC_DISABLE (1UL << 2) +# define PR_SPEC_FORCE_DISABLE (1UL << 3) #endif /* _LINUX_PRCTL_H */

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 074/101] seccomp: Enable speculation flaw mitigations

by Srivatsa S. Bhat

From: Kees Cook <keescook(a)chromium.org> commit 5c3070890d06ff82eecb808d02d2ca39169533ef upstream When speculation flaw mitigations are opt-in (via prctl), using seccomp will automatically opt-in to these protections, since using seccomp indicates at least some level of sandboxing is desired. Signed-off-by: Kees Cook <keescook(a)chromium.org> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- kernel/seccomp.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/kernel/seccomp.c b/kernel/seccomp.c index efd384f..bfb1ee8 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -16,6 +16,8 @@ #include <linux/atomic.h> #include <linux/audit.h> #include <linux/compat.h> +#include <linux/nospec.h> +#include <linux/prctl.h> #include <linux/sched.h> #include <linux/seccomp.h> #include <linux/slab.h> @@ -214,6 +216,19 @@ static inline bool seccomp_may_assign_mode(unsigned long seccomp_mode) return true; } +/* + * If a given speculation mitigation is opt-in (prctl()-controlled), + * select it, by disabling speculation (enabling mitigation). + */ +static inline void spec_mitigate(struct task_struct *task, + unsigned long which) +{ + int state = arch_prctl_spec_ctrl_get(task, which); + + if (state > 0 && (state & PR_SPEC_PRCTL)) + arch_prctl_spec_ctrl_set(task, which, PR_SPEC_DISABLE); +} + static inline void seccomp_assign_mode(struct task_struct *task, unsigned long seccomp_mode) { @@ -225,6 +240,8 @@ static inline void seccomp_assign_mode(struct task_struct *task, * filter) is set. */ smp_mb__before_atomic(); + /* Assume seccomp processes want speculation flaw mitigation. */ + spec_mitigate(task, PR_SPEC_STORE_BYPASS); set_tsk_thread_flag(task, TIF_SECCOMP); }

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 073/101] proc: Provide details on speculation flaw mitigations

by Srivatsa S. Bhat

From: Kees Cook <keescook(a)chromium.org> commit fae1fa0fc6cca8beee3ab8ed71d54f9a78fa3f64 upstream As done with seccomp and no_new_privs, also show speculation flaw mitigation state in /proc/$pid/status. Signed-off-by: Kees Cook <keescook(a)chromium.org> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- fs/proc/array.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/fs/proc/array.c b/fs/proc/array.c index b6c00ce..bb48358 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -79,6 +79,7 @@ #include <linux/delayacct.h> #include <linux/seq_file.h> #include <linux/pid_namespace.h> +#include <linux/prctl.h> #include <linux/ptrace.h> #include <linux/tracehook.h> #include <linux/string_helpers.h> @@ -332,6 +333,28 @@ static inline void task_seccomp(struct seq_file *m, struct task_struct *p) #ifdef CONFIG_SECCOMP seq_printf(m, "Seccomp:\t%d\n", p->seccomp.mode); #endif + seq_printf(m, "\nSpeculation Store Bypass:\t"); + switch (arch_prctl_spec_ctrl_get(p, PR_SPEC_STORE_BYPASS)) { + case -EINVAL: + seq_printf(m, "unknown"); + break; + case PR_SPEC_NOT_AFFECTED: + seq_printf(m, "not vulnerable"); + break; + case PR_SPEC_PRCTL | PR_SPEC_DISABLE: + seq_printf(m, "thread mitigated"); + break; + case PR_SPEC_PRCTL | PR_SPEC_ENABLE: + seq_printf(m, "thread vulnerable"); + break; + case PR_SPEC_DISABLE: + seq_printf(m, "globally mitigated"); + break; + default: + seq_printf(m, "vulnerable"); + break; + } + seq_putc(m, '\n'); } static inline void task_context_switch_counts(struct seq_file *m,

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 072/101] nospec: Allow getting/setting on non-current task

by Srivatsa S. Bhat

From: Kees Cook <keescook(a)chromium.org> commit 7bbf1373e228840bb0295a2ca26d548ef37f448e upstream Adjust arch_prctl_get/set_spec_ctrl() to operate on tasks other than current. This is needed both for /proc/$pid/status queries and for seccomp (since thread-syncing can trigger seccomp in non-current threads). Signed-off-by: Kees Cook <keescook(a)chromium.org> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- arch/x86/kernel/cpu/bugs.c | 27 ++++++++++++++++----------- include/linux/nospec.h | 7 +++++-- kernel/sys.c | 9 +++++---- 3 files changed, 26 insertions(+), 17 deletions(-) diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c index bcfccd3..64b54a4 100644 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -529,31 +529,35 @@ static void ssb_select_mitigation() #undef pr_fmt -static int ssb_prctl_set(unsigned long ctrl) +static int ssb_prctl_set(struct task_struct *task, unsigned long ctrl) { - bool rds = !!test_tsk_thread_flag(current, TIF_RDS); + bool rds = !!test_tsk_thread_flag(task, TIF_RDS); if (ssb_mode != SPEC_STORE_BYPASS_PRCTL) return -ENXIO; if (ctrl == PR_SPEC_ENABLE) - clear_tsk_thread_flag(current, TIF_RDS); + clear_tsk_thread_flag(task, TIF_RDS); else - set_tsk_thread_flag(current, TIF_RDS); + set_tsk_thread_flag(task, TIF_RDS); - if (rds != !!test_tsk_thread_flag(current, TIF_RDS)) + /* + * If being set on non-current task, delay setting the CPU + * mitigation until it is next scheduled. + */ + if (task == current && rds != !!test_tsk_thread_flag(task, TIF_RDS)) speculative_store_bypass_update(); return 0; } -static int ssb_prctl_get(void) +static int ssb_prctl_get(struct task_struct *task) { switch (ssb_mode) { case SPEC_STORE_BYPASS_DISABLE: return PR_SPEC_DISABLE; case SPEC_STORE_BYPASS_PRCTL: - if (test_tsk_thread_flag(current, TIF_RDS)) + if (test_tsk_thread_flag(task, TIF_RDS)) return PR_SPEC_PRCTL | PR_SPEC_DISABLE; return PR_SPEC_PRCTL | PR_SPEC_ENABLE; default: @@ -563,24 +567,25 @@ static int ssb_prctl_get(void) } } -int arch_prctl_spec_ctrl_set(unsigned long which, unsigned long ctrl) +int arch_prctl_spec_ctrl_set(struct task_struct *task, unsigned long which, + unsigned long ctrl) { if (ctrl != PR_SPEC_ENABLE && ctrl != PR_SPEC_DISABLE) return -ERANGE; switch (which) { case PR_SPEC_STORE_BYPASS: - return ssb_prctl_set(ctrl); + return ssb_prctl_set(task, ctrl); default: return -ENODEV; } } -int arch_prctl_spec_ctrl_get(unsigned long which) +int arch_prctl_spec_ctrl_get(struct task_struct *task, unsigned long which) { switch (which) { case PR_SPEC_STORE_BYPASS: - return ssb_prctl_get(); + return ssb_prctl_get(task); default: return -ENODEV; } diff --git a/include/linux/nospec.h b/include/linux/nospec.h index 700bb8a..a908c95 100644 --- a/include/linux/nospec.h +++ b/include/linux/nospec.h @@ -7,6 +7,8 @@ #define _LINUX_NOSPEC_H #include <asm/barrier.h> +struct task_struct; + /** * array_index_mask_nospec() - generate a ~0 mask when index < size, 0 otherwise * @index: array element index @@ -57,7 +59,8 @@ static inline unsigned long array_index_mask_nospec(unsigned long index, }) /* Speculation control prctl */ -int arch_prctl_spec_ctrl_get(unsigned long which); -int arch_prctl_spec_ctrl_set(unsigned long which, unsigned long ctrl); +int arch_prctl_spec_ctrl_get(struct task_struct *task, unsigned long which); +int arch_prctl_spec_ctrl_set(struct task_struct *task, unsigned long which, + unsigned long ctrl); #endif /* _LINUX_NOSPEC_H */ diff --git a/kernel/sys.c b/kernel/sys.c index d80c33f..f718742 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -2075,12 +2075,13 @@ static int prctl_get_tid_address(struct task_struct *me, int __user **tid_addr) } #endif -int __weak arch_prctl_spec_ctrl_get(unsigned long which) +int __weak arch_prctl_spec_ctrl_get(struct task_struct *t, unsigned long which) { return -EINVAL; } -int __weak arch_prctl_spec_ctrl_set(unsigned long which, unsigned long ctrl) +int __weak arch_prctl_spec_ctrl_set(struct task_struct *t, unsigned long which, + unsigned long ctrl) { return -EINVAL; } @@ -2282,12 +2283,12 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3, case PR_GET_SPECULATION_CTRL: if (arg3 || arg4 || arg5) return -EINVAL; - error = arch_prctl_spec_ctrl_get(arg2); + error = arch_prctl_spec_ctrl_get(me, arg2); break; case PR_SET_SPECULATION_CTRL: if (arg4 || arg5) return -EINVAL; - error = arch_prctl_spec_ctrl_set(arg2, arg3); + error = arch_prctl_spec_ctrl_set(me, arg2, arg3); break; default: error = -EINVAL;

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 071/101] x86/speculation: Add prctl for Speculative Store Bypass mitigation

by Srivatsa S. Bhat

From: Thomas Gleixner <tglx(a)linutronix.de> commit a73ec77ee17ec556fe7f165d00314cb7c047b1ac upstream Add prctl based control for Speculative Store Bypass mitigation and make it the default mitigation for Intel and AMD. Andi Kleen provided the following rationale (slightly redacted): There are multiple levels of impact of Speculative Store Bypass: 1) JITed sandbox. It cannot invoke system calls, but can do PRIME+PROBE and may have call interfaces to other code 2) Native code process. No protection inside the process at this level. 3) Kernel. 4) Between processes. The prctl tries to protect against case (1) doing attacks. If the untrusted code can do random system calls then control is already lost in a much worse way. So there needs to be system call protection in some way (using a JIT not allowing them or seccomp). Or rather if the process can subvert its environment somehow to do the prctl it can already execute arbitrary code, which is much worse than SSB. To put it differently, the point of the prctl is to not allow JITed code to read data it shouldn't read from its JITed sandbox. If it already has escaped its sandbox then it can already read everything it wants in its address space, and do much worse. The ability to control Speculative Store Bypass allows to enable the protection selectively without affecting overall system performance. Based on an initial patch from Tim Chen. Completely rewritten. Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk(a)oracle.com> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- Documentation/kernel-parameters.txt | 6 ++ arch/x86/include/asm/nospec-branch.h | 1 arch/x86/kernel/cpu/bugs.c | 83 ++++++++++++++++++++++++++++++---- 3 files changed, 79 insertions(+), 11 deletions(-) diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index dc138b8..80202de 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -3651,7 +3651,11 @@ bytes respectively. Such letter suffixes can also be entirely omitted. off - Unconditionally enable Speculative Store Bypass auto - Kernel detects whether the CPU model contains an implementation of Speculative Store Bypass and - picks the most appropriate mitigation + picks the most appropriate mitigation. + prctl - Control Speculative Store Bypass per thread + via prctl. Speculative Store Bypass is enabled + for a process by default. The state of the control + is inherited on fork. Not specifying this option is equivalent to spec_store_bypass_disable=auto. diff --git a/arch/x86/include/asm/nospec-branch.h b/arch/x86/include/asm/nospec-branch.h index 47c454c..155d955 100644 --- a/arch/x86/include/asm/nospec-branch.h +++ b/arch/x86/include/asm/nospec-branch.h @@ -187,6 +187,7 @@ extern u64 x86_spec_ctrl_get_default(void); enum ssb_mitigation { SPEC_STORE_BYPASS_NONE, SPEC_STORE_BYPASS_DISABLE, + SPEC_STORE_BYPASS_PRCTL, }; extern char __indirect_thunk_start[]; diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c index 0f8303e..bcfccd3 100644 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -11,6 +11,8 @@ #include <linux/utsname.h> #include <linux/cpu.h> #include <linux/module.h> +#include <linux/nospec.h> +#include <linux/prctl.h> #include <asm/spec-ctrl.h> #include <asm/cmdline.h> @@ -411,20 +413,23 @@ enum ssb_mitigation_cmd { SPEC_STORE_BYPASS_CMD_NONE, SPEC_STORE_BYPASS_CMD_AUTO, SPEC_STORE_BYPASS_CMD_ON, + SPEC_STORE_BYPASS_CMD_PRCTL, }; static const char *ssb_strings[] = { [SPEC_STORE_BYPASS_NONE] = "Vulnerable", - [SPEC_STORE_BYPASS_DISABLE] = "Mitigation: Speculative Store Bypass disabled" + [SPEC_STORE_BYPASS_DISABLE] = "Mitigation: Speculative Store Bypass disabled", + [SPEC_STORE_BYPASS_PRCTL] = "Mitigation: Speculative Store Bypass disabled via prctl" }; static const struct { const char *option; enum ssb_mitigation_cmd cmd; } ssb_mitigation_options[] = { - { "auto", SPEC_STORE_BYPASS_CMD_AUTO }, /* Platform decides */ - { "on", SPEC_STORE_BYPASS_CMD_ON }, /* Disable Speculative Store Bypass */ - { "off", SPEC_STORE_BYPASS_CMD_NONE }, /* Don't touch Speculative Store Bypass */ + { "auto", SPEC_STORE_BYPASS_CMD_AUTO }, /* Platform decides */ + { "on", SPEC_STORE_BYPASS_CMD_ON }, /* Disable Speculative Store Bypass */ + { "off", SPEC_STORE_BYPASS_CMD_NONE }, /* Don't touch Speculative Store Bypass */ + { "prctl", SPEC_STORE_BYPASS_CMD_PRCTL }, /* Disable Speculative Store Bypass via prctl */ }; static enum ssb_mitigation_cmd __init ssb_parse_cmdline(void) @@ -474,14 +479,15 @@ static enum ssb_mitigation_cmd __init __ssb_select_mitigation(void) switch (cmd) { case SPEC_STORE_BYPASS_CMD_AUTO: - /* - * AMD platforms by default don't need SSB mitigation. - */ - if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) - break; + /* Choose prctl as the default mode */ + mode = SPEC_STORE_BYPASS_PRCTL; + break; case SPEC_STORE_BYPASS_CMD_ON: mode = SPEC_STORE_BYPASS_DISABLE; break; + case SPEC_STORE_BYPASS_CMD_PRCTL: + mode = SPEC_STORE_BYPASS_PRCTL; + break; case SPEC_STORE_BYPASS_CMD_NONE: break; } @@ -492,7 +498,7 @@ static enum ssb_mitigation_cmd __init __ssb_select_mitigation(void) * - X86_FEATURE_RDS - CPU is able to turn off speculative store bypass * - X86_FEATURE_SPEC_STORE_BYPASS_DISABLE - engage the mitigation */ - if (mode != SPEC_STORE_BYPASS_NONE) { + if (mode == SPEC_STORE_BYPASS_DISABLE) { setup_force_cpu_cap(X86_FEATURE_SPEC_STORE_BYPASS_DISABLE); /* * Intel uses the SPEC CTRL MSR Bit(2) for this, while AMD uses @@ -523,6 +529,63 @@ static void ssb_select_mitigation() #undef pr_fmt +static int ssb_prctl_set(unsigned long ctrl) +{ + bool rds = !!test_tsk_thread_flag(current, TIF_RDS); + + if (ssb_mode != SPEC_STORE_BYPASS_PRCTL) + return -ENXIO; + + if (ctrl == PR_SPEC_ENABLE) + clear_tsk_thread_flag(current, TIF_RDS); + else + set_tsk_thread_flag(current, TIF_RDS); + + if (rds != !!test_tsk_thread_flag(current, TIF_RDS)) + speculative_store_bypass_update(); + + return 0; +} + +static int ssb_prctl_get(void) +{ + switch (ssb_mode) { + case SPEC_STORE_BYPASS_DISABLE: + return PR_SPEC_DISABLE; + case SPEC_STORE_BYPASS_PRCTL: + if (test_tsk_thread_flag(current, TIF_RDS)) + return PR_SPEC_PRCTL | PR_SPEC_DISABLE; + return PR_SPEC_PRCTL | PR_SPEC_ENABLE; + default: + if (boot_cpu_has_bug(X86_BUG_SPEC_STORE_BYPASS)) + return PR_SPEC_ENABLE; + return PR_SPEC_NOT_AFFECTED; + } +} + +int arch_prctl_spec_ctrl_set(unsigned long which, unsigned long ctrl) +{ + if (ctrl != PR_SPEC_ENABLE && ctrl != PR_SPEC_DISABLE) + return -ERANGE; + + switch (which) { + case PR_SPEC_STORE_BYPASS: + return ssb_prctl_set(ctrl); + default: + return -ENODEV; + } +} + +int arch_prctl_spec_ctrl_get(unsigned long which) +{ + switch (which) { + case PR_SPEC_STORE_BYPASS: + return ssb_prctl_get(); + default: + return -ENODEV; + } +} + void x86_spec_ctrl_setup_ap(void) { if (boot_cpu_has(X86_FEATURE_IBRS))

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 070/101] x86/process: Allow runtime control of Speculative Store Bypass

by Srivatsa S. Bhat

From: Thomas Gleixner <tglx(a)linutronix.de> commit 885f82bfbc6fefb6664ea27965c3ab9ac4194b8c upstream The Speculative Store Bypass vulnerability can be mitigated with the Reduced Data Speculation (RDS) feature. To allow finer grained control of this eventually expensive mitigation a per task mitigation control is required. Add a new TIF_RDS flag and put it into the group of TIF flags which are evaluated for mismatch in switch_to(). If these bits differ in the previous and the next task, then the slow path function __switch_to_xtra() is invoked. Implement the TIF_RDS dependent mitigation control in the slow path. If the prctl for controlling Speculative Store Bypass is disabled or no task uses the prctl then there is no overhead in the switch_to() fast path. Update the KVM related speculation control functions to take TID_RDS into account as well. Based on a patch from Tim Chen. Completely rewritten. Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Reviewed-by: Ingo Molnar <mingo(a)kernel.org> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk(a)oracle.com> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- arch/x86/include/asm/msr-index.h | 3 ++- arch/x86/include/asm/spec-ctrl.h | 17 +++++++++++++++++ arch/x86/include/asm/thread_info.h | 6 ++++-- arch/x86/kernel/cpu/bugs.c | 26 +++++++++++++++++++++----- arch/x86/kernel/process.c | 22 ++++++++++++++++++++++ 5 files changed, 66 insertions(+), 8 deletions(-) diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index 71a2c84..883cf0d 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -35,7 +35,8 @@ #define MSR_IA32_SPEC_CTRL 0x00000048 /* Speculation Control */ #define SPEC_CTRL_IBRS (1 << 0) /* Indirect Branch Restricted Speculation */ #define SPEC_CTRL_STIBP (1 << 1) /* Single Thread Indirect Branch Predictors */ -#define SPEC_CTRL_RDS (1 << 2) /* Reduced Data Speculation */ +#define SPEC_CTRL_RDS_SHIFT 2 /* Reduced Data Speculation bit */ +#define SPEC_CTRL_RDS (1 << SPEC_CTRL_RDS_SHIFT) /* Reduced Data Speculation */ #define MSR_IA32_PRED_CMD 0x00000049 /* Prediction Command */ #define PRED_CMD_IBPB (1 << 0) /* Indirect Branch Prediction Barrier */ diff --git a/arch/x86/include/asm/spec-ctrl.h b/arch/x86/include/asm/spec-ctrl.h index 3ad6442..45ef00a 100644 --- a/arch/x86/include/asm/spec-ctrl.h +++ b/arch/x86/include/asm/spec-ctrl.h @@ -2,6 +2,7 @@ #ifndef _ASM_X86_SPECCTRL_H_ #define _ASM_X86_SPECCTRL_H_ +#include <linux/thread_info.h> #include <asm/nospec-branch.h> /* @@ -18,4 +19,20 @@ extern void x86_spec_ctrl_restore_host(u64); extern u64 x86_amd_ls_cfg_base; extern u64 x86_amd_ls_cfg_rds_mask; +/* The Intel SPEC CTRL MSR base value cache */ +extern u64 x86_spec_ctrl_base; + +static inline u64 rds_tif_to_spec_ctrl(u64 tifn) +{ + BUILD_BUG_ON(TIF_RDS < SPEC_CTRL_RDS_SHIFT); + return (tifn & _TIF_RDS) >> (TIF_RDS - SPEC_CTRL_RDS_SHIFT); +} + +static inline u64 rds_tif_to_amd_ls_cfg(u64 tifn) +{ + return (tifn & _TIF_RDS) ? x86_amd_ls_cfg_rds_mask : 0ULL; +} + +extern void speculative_store_bypass_update(void); + #endif diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index 18c9aaa..36a49b4 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -92,6 +92,7 @@ struct thread_info { #define TIF_SIGPENDING 2 /* signal pending */ #define TIF_NEED_RESCHED 3 /* rescheduling necessary */ #define TIF_SINGLESTEP 4 /* reenable singlestep on user return*/ +#define TIF_RDS 5 /* Reduced data speculation */ #define TIF_SYSCALL_EMU 6 /* syscall emulation active */ #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ #define TIF_SECCOMP 8 /* secure computing */ @@ -114,8 +115,9 @@ struct thread_info { #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) #define _TIF_SIGPENDING (1 << TIF_SIGPENDING) -#define _TIF_SINGLESTEP (1 << TIF_SINGLESTEP) #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) +#define _TIF_SINGLESTEP (1 << TIF_SINGLESTEP) +#define _TIF_RDS (1 << TIF_RDS) #define _TIF_SYSCALL_EMU (1 << TIF_SYSCALL_EMU) #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) #define _TIF_SECCOMP (1 << TIF_SECCOMP) @@ -147,7 +149,7 @@ struct thread_info { /* flags to check in __switch_to() */ #define _TIF_WORK_CTXSW \ - (_TIF_IO_BITMAP|_TIF_NOTSC|_TIF_BLOCKSTEP) + (_TIF_IO_BITMAP|_TIF_NOTSC|_TIF_BLOCKSTEP|_TIF_RDS) #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW) diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c index 47a3cc0..0f8303e 100644 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -32,7 +32,7 @@ static void __init ssb_select_mitigation(void); * Our boot-time value of the SPEC_CTRL MSR. We read it once so that any * writes to SPEC_CTRL contain whatever reserved bits have been set. */ -static u64 x86_spec_ctrl_base; +u64 x86_spec_ctrl_base; /* * The vendor and possibly platform specific bits which can be modified in @@ -139,25 +139,41 @@ EXPORT_SYMBOL_GPL(x86_spec_ctrl_set); u64 x86_spec_ctrl_get_default(void) { - return x86_spec_ctrl_base; + u64 msrval = x86_spec_ctrl_base; + + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) + msrval |= rds_tif_to_spec_ctrl(current_thread_info()->flags); + return msrval; } EXPORT_SYMBOL_GPL(x86_spec_ctrl_get_default); void x86_spec_ctrl_set_guest(u64 guest_spec_ctrl) { + u64 host = x86_spec_ctrl_base; + if (!boot_cpu_has(X86_FEATURE_IBRS)) return; - if (x86_spec_ctrl_base != guest_spec_ctrl) + + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) + host |= rds_tif_to_spec_ctrl(current_thread_info()->flags); + + if (host != guest_spec_ctrl) wrmsrl(MSR_IA32_SPEC_CTRL, guest_spec_ctrl); } EXPORT_SYMBOL_GPL(x86_spec_ctrl_set_guest); void x86_spec_ctrl_restore_host(u64 guest_spec_ctrl) { + u64 host = x86_spec_ctrl_base; + if (!boot_cpu_has(X86_FEATURE_IBRS)) return; - if (x86_spec_ctrl_base != guest_spec_ctrl) - wrmsrl(MSR_IA32_SPEC_CTRL, x86_spec_ctrl_base); + + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) + host |= rds_tif_to_spec_ctrl(current_thread_info()->flags); + + if (host != guest_spec_ctrl) + wrmsrl(MSR_IA32_SPEC_CTRL, host); } EXPORT_SYMBOL_GPL(x86_spec_ctrl_restore_host); diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index d112963..9689e92 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -31,6 +31,7 @@ #include <asm/tlbflush.h> #include <asm/mce.h> #include <asm/vm86.h> +#include <asm/spec-ctrl.h> /* * per-CPU TSS segments. Threads are completely 'soft' on Linux, @@ -198,6 +199,24 @@ static inline void switch_to_bitmap(struct tss_struct *tss, } } +static __always_inline void __speculative_store_bypass_update(unsigned long tifn) +{ + u64 msr; + + if (static_cpu_has(X86_FEATURE_AMD_RDS)) { + msr = x86_amd_ls_cfg_base | rds_tif_to_amd_ls_cfg(tifn); + wrmsrl(MSR_AMD64_LS_CFG, msr); + } else { + msr = x86_spec_ctrl_base | rds_tif_to_spec_ctrl(tifn); + wrmsrl(MSR_IA32_SPEC_CTRL, msr); + } +} + +void speculative_store_bypass_update(void) +{ + __speculative_store_bypass_update(current_thread_info()->flags); +} + void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p, struct tss_struct *tss) { @@ -226,6 +245,9 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p, if ((tifp ^ tifn) & _TIF_NOTSC) cr4_toggle_bits(X86_CR4_TSD); + + if ((tifp ^ tifn) & _TIF_RDS) + __speculative_store_bypass_update(tifn); } /*

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 069/101] x86/process: Optimize TIF_NOTSC switch

by Srivatsa S. Bhat

From: Thomas Gleixner <tglx(a)linutronix.de> commit 5a920155e388ec22a22e0532fb695b9215c9b34d upstream Provide and use a toggle helper instead of doing it with a branch. x86_64: arch/x86/kernel/process.o text data bss dec hex 3008 8577 16 11601 2d51 Before 2976 8577 16 11569 2d31 After i386: arch/x86/kernel/process.o text data bss dec hex 2925 8673 8 11606 2d56 Before 2893 8673 8 11574 2d36 After Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Andy Lutomirski <luto(a)kernel.org> Link: http://lkml.kernel.org/r/20170214081104.9244-4-khuey@kylehuey.com Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- arch/x86/include/asm/tlbflush.h | 10 ++++++++++ arch/x86/kernel/process.c | 22 ++++------------------ 2 files changed, 14 insertions(+), 18 deletions(-) diff --git a/arch/x86/include/asm/tlbflush.h b/arch/x86/include/asm/tlbflush.h index 8ce07db..72cfe3e 100644 --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -111,6 +111,16 @@ static inline void cr4_clear_bits(unsigned long mask) } } +static inline void cr4_toggle_bits(unsigned long mask) +{ + unsigned long cr4; + + cr4 = this_cpu_read(cpu_tlbstate.cr4); + cr4 ^= mask; + this_cpu_write(cpu_tlbstate.cr4, cr4); + __write_cr4(cr4); +} + /* Read the CR4 shadow. */ static inline unsigned long cr4_read_shadow(void) { diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 166aef3..d112963 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -130,11 +130,6 @@ void flush_thread(void) fpu__clear(&tsk->thread.fpu); } -static void hard_disable_TSC(void) -{ - cr4_set_bits(X86_CR4_TSD); -} - void disable_TSC(void) { preempt_disable(); @@ -143,15 +138,10 @@ void disable_TSC(void) * Must flip the CPU state synchronously with * TIF_NOTSC in the current running context. */ - hard_disable_TSC(); + cr4_set_bits(X86_CR4_TSD); preempt_enable(); } -static void hard_enable_TSC(void) -{ - cr4_clear_bits(X86_CR4_TSD); -} - static void enable_TSC(void) { preempt_disable(); @@ -160,7 +150,7 @@ static void enable_TSC(void) * Must flip the CPU state synchronously with * TIF_NOTSC in the current running context. */ - hard_enable_TSC(); + cr4_clear_bits(X86_CR4_TSD); preempt_enable(); } @@ -234,12 +224,8 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p, wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); } - if ((tifp ^ tifn) & _TIF_NOTSC) { - if (tifn & _TIF_NOTSC) - hard_disable_TSC(); - else - hard_enable_TSC(); - } + if ((tifp ^ tifn) & _TIF_NOTSC) + cr4_toggle_bits(X86_CR4_TSD); } /*

6 years, 7 months

1
0
0 0

[PATCH 4.4.y 068/101] x86/process: Correct and optimize TIF_BLOCKSTEP switch

by Srivatsa S. Bhat

From: Kyle Huey <me(a)kylehuey.com> commit b9894a2f5bd18b1691cb6872c9afe32b148d0132 upstream The debug control MSR is "highly magical" as the blockstep bit can be cleared by hardware under not well documented circumstances. So a task switch relying on the bit set by the previous task (according to the previous tasks thread flags) can trip over this and not update the flag for the next task. To fix this its required to handle DEBUGCTLMSR_BTF when either the previous or the next or both tasks have the TIF_BLOCKSTEP flag set. While at it avoid branching within the TIF_BLOCKSTEP case and evaluating boot_cpu_data twice in kernels without CONFIG_X86_DEBUGCTLMSR. x86_64: arch/x86/kernel/process.o text data bss dec hex 3024 8577 16 11617 2d61 Before 3008 8577 16 11601 2d51 After i386: No change [ tglx: Made the shift value explicit, use a local variable to make the code readable and massaged changelog] Originally-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: Kyle Huey <khuey(a)kylehuey.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Andy Lutomirski <luto(a)kernel.org> Link: http://lkml.kernel.org/r/20170214081104.9244-3-khuey@kylehuey.com Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: David Woodhouse <dwmw(a)amazon.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Srivatsa S. Bhat <srivatsa(a)csail.mit.edu> Reviewed-by: Matt Helsley (VMware) <matt.helsley(a)gmail.com> Reviewed-by: Alexey Makhalov <amakhalov(a)vmware.com> Reviewed-by: Bo Gan <ganb(a)vmware.com> --- arch/x86/include/asm/msr-index.h | 1 + arch/x86/kernel/process.c | 12 +++++++----- 2 files changed, 8 insertions(+), 5 deletions(-) diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index a29edb7..71a2c84 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -150,6 +150,7 @@ /* DEBUGCTLMSR bits (others vary by model): */ #define DEBUGCTLMSR_LBR (1UL << 0) /* last branch recording */ +#define DEBUGCTLMSR_BTF_SHIFT 1 #define DEBUGCTLMSR_BTF (1UL << 1) /* single-step on branches */ #define DEBUGCTLMSR_TR (1UL << 6) #define DEBUGCTLMSR_BTS (1UL << 7) diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index cc0f288..166aef3 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -223,13 +223,15 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p, propagate_user_return_notify(prev_p, next_p); - if ((tifp ^ tifn) & _TIF_BLOCKSTEP) { - unsigned long debugctl = get_debugctlmsr(); + if ((tifp & _TIF_BLOCKSTEP || tifn & _TIF_BLOCKSTEP) && + arch_has_block_step()) { + unsigned long debugctl, msk; + rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); debugctl &= ~DEBUGCTLMSR_BTF; - if (tifn & _TIF_BLOCKSTEP) - debugctl |= DEBUGCTLMSR_BTF; - update_debugctlmsr(debugctl); + msk = tifn & _TIF_BLOCKSTEP; + debugctl |= (msk >> TIF_BLOCKSTEP) << DEBUGCTLMSR_BTF_SHIFT; + wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); } if ((tifp ^ tifn) & _TIF_NOTSC) {

6 years, 7 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror July 2018