From 13fa2c65c9a8c2cd5f2a9799891582c40b6f5cfa Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 14 Jul 2018 02:35:28 -0700 Subject: prctl: Add speculation control prctls commit b617cfc858161140d69cc0b5cc211996b557a1c7 upstream Add two new prctls to control aspects of speculation related vulnerabilites and their mitigations to provide finer grained control over performance impacting mitigations. PR_GET_SPECULATION_CTRL returns the state of the speculation misfeature which is selected with arg2 of prctl(2). The return value uses bit 0-2 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 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. PR_SET_SPECULATION_CTRL allows to control the speculation misfeature, which is selected by arg2 of prctl(2) per task. arg3 is used to hand in the control value, i.e. either PR_SPEC_ENABLE or PR_SPEC_DISABLE. The common return values are: EINVAL prctl is not implemented by the architecture or the unused prctl() arguments are not 0 ENODEV arg2 is selecting a not supported speculation misfeature PR_SET_SPECULATION_CTRL has these additional return values: ERANGE arg3 is incorrect, i.e. it's not either PR_SPEC_ENABLE or PR_SPEC_DISABLE ENXIO prctl control of the selected speculation misfeature is disabled The first supported controlable speculation misfeature is PR_SPEC_STORE_BYPASS. Add the define so this can be shared between architectures. Based on an initial patch from Tim Chen and mostly rewritten. Signed-off-by: Thomas Gleixner Reviewed-by: Ingo Molnar Reviewed-by: Konrad Rzeszutek Wilk Signed-off-by: David Woodhouse Signed-off-by: Greg Kroah-Hartman Signed-off-by: Srivatsa S. Bhat Reviewed-by: Matt Helsley (VMware) Reviewed-by: Alexey Makhalov Reviewed-by: Bo Gan Signed-off-by: Greg Kroah-Hartman --- kernel/sys.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) (limited to 'kernel') diff --git a/kernel/sys.c b/kernel/sys.c index 6624919ef0e7..d80c33f9aff7 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -2075,6 +2075,16 @@ 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) +{ + return -EINVAL; +} + +int __weak arch_prctl_spec_ctrl_set(unsigned long which, unsigned long ctrl) +{ + return -EINVAL; +} + SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3, unsigned long, arg4, unsigned long, arg5) { @@ -2269,6 +2279,16 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3, case PR_GET_FP_MODE: error = GET_FP_MODE(me); break; + case PR_GET_SPECULATION_CTRL: + if (arg3 || arg4 || arg5) + return -EINVAL; + error = arch_prctl_spec_ctrl_get(arg2); + break; + case PR_SET_SPECULATION_CTRL: + if (arg4 || arg5) + return -EINVAL; + error = arch_prctl_spec_ctrl_set(arg2, arg3); + break; default: error = -EINVAL; break; -- cgit v1.2.3 From b6f4a6285d7979b45d629e65c880279930b98ef1 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Sat, 14 Jul 2018 02:36:17 -0700 Subject: nospec: Allow getting/setting on non-current task 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 Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse Signed-off-by: Greg Kroah-Hartman Signed-off-by: Srivatsa S. Bhat Reviewed-by: Matt Helsley (VMware) Reviewed-by: Alexey Makhalov Reviewed-by: Bo Gan Signed-off-by: Greg Kroah-Hartman --- kernel/sys.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/sys.c b/kernel/sys.c index d80c33f9aff7..f718742e55e6 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; -- cgit v1.2.3 From 0b1174054e0f4afd999c56ddecbbfb18f598f099 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Sat, 14 Jul 2018 02:36:33 -0700 Subject: seccomp: Enable speculation flaw mitigations 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 Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse Signed-off-by: Greg Kroah-Hartman Signed-off-by: Srivatsa S. Bhat Reviewed-by: Matt Helsley (VMware) Reviewed-by: Alexey Makhalov Reviewed-by: Bo Gan Signed-off-by: Greg Kroah-Hartman --- kernel/seccomp.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) (limited to 'kernel') diff --git a/kernel/seccomp.c b/kernel/seccomp.c index efd384f3f852..bfb1ee845ba6 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -16,6 +16,8 @@ #include #include #include +#include +#include #include #include #include @@ -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); } -- cgit v1.2.3 From a08c3f484c34df1e3bec3c47818d570483bf67fa Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 14 Jul 2018 02:36:49 -0700 Subject: seccomp: Use PR_SPEC_FORCE_DISABLE commit b849a812f7eb92e96d1c8239b06581b2cfd8b275 upstream Use PR_SPEC_FORCE_DISABLE in seccomp() because seccomp does not allow to widen restrictions. Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse Signed-off-by: Greg Kroah-Hartman Signed-off-by: Srivatsa S. Bhat Reviewed-by: Matt Helsley (VMware) Reviewed-by: Alexey Makhalov Reviewed-by: Bo Gan Signed-off-by: Greg Kroah-Hartman --- kernel/seccomp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/seccomp.c b/kernel/seccomp.c index bfb1ee845ba6..f33539f31bc0 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, -- cgit v1.2.3 From c463c0f037f2d83aea54415ed7c61deb0b90333b Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Sat, 14 Jul 2018 02:36:57 -0700 Subject: seccomp: Add filter flag to opt-out of SSB mitigation 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 Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse Signed-off-by: Greg Kroah-Hartman Signed-off-by: Srivatsa S. Bhat Reviewed-by: Matt Helsley (VMware) Reviewed-by: Alexey Makhalov Reviewed-by: Bo Gan Signed-off-by: Greg Kroah-Hartman --- kernel/seccomp.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/seccomp.c b/kernel/seccomp.c index f33539f31bc0..4bb8a5a5d68a 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(¤t->sighand->siglock); if (flags & SECCOMP_FILTER_FLAG_TSYNC) -- cgit v1.2.3 From 9237a1b0828962191107e702cf56c88db9f9d455 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 14 Jul 2018 02:37:05 -0700 Subject: seccomp: Move speculation migitation control to arch code commit 8bf37d8c067bb7eb8e7c381bdadf9bd89182b6bc upstream The migitation control is simpler to implement in architecture code as it avoids the extra function call to check the mode. Aside of that having an explicit seccomp enabled mode in the architecture mitigations would require even more workarounds. Move it into architecture code and provide a weak function in the seccomp code. Remove the 'which' argument as this allows the architecture to decide which mitigations are relevant for seccomp. Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse Signed-off-by: Greg Kroah-Hartman Signed-off-by: Srivatsa S. Bhat Reviewed-by: Matt Helsley (VMware) Reviewed-by: Alexey Makhalov Reviewed-by: Bo Gan Signed-off-by: Greg Kroah-Hartman --- kernel/seccomp.c | 15 ++------------- 1 file changed, 2 insertions(+), 13 deletions(-) (limited to 'kernel') diff --git a/kernel/seccomp.c b/kernel/seccomp.c index 4bb8a5a5d68a..9a9203b15cde 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -216,18 +216,7 @@ 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_FORCE_DISABLE); -} +void __weak arch_seccomp_spec_mitigate(struct task_struct *task) { } static inline void seccomp_assign_mode(struct task_struct *task, unsigned long seccomp_mode, @@ -243,7 +232,7 @@ static inline void seccomp_assign_mode(struct task_struct *task, smp_mb__before_atomic(); /* Assume default seccomp processes want spec flaw mitigation. */ if ((flags & SECCOMP_FILTER_FLAG_SPEC_ALLOW) == 0) - spec_mitigate(task, PR_SPEC_STORE_BYPASS); + arch_seccomp_spec_mitigate(task); set_tsk_thread_flag(task, TIF_SECCOMP); } -- cgit v1.2.3