rseq: Fix: Reject unknown flags on rseq unregister
Subsystems (2)
PATCH 1/3 rseq: Fix: Reject unknown flags on rseq unregister
It is preferrable to reject unknown flags within rseq unregistration rather than to ignore them. It is an oversight caused by the fact that the check for unknown flags is after the rseq unregister flag check.
--- kernel/rseq.c | 2 ++ 1 file changed, 2 insertions(+)
kernel/rseq.c +2 -0
@@ -310,6 +310,8 @@ SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len,int ret;if (flags &RSEQ_FLAG_UNREGISTER ) { +if (flags & ~RSEQ_FLAG_UNREGISTER ) +return -EINVAL ;/* Unregister rseq for current thread. */ if (current->rseq != rseq || !current->rseq)return -EINVAL ;
Signature · 1 line
2.17.1
PATCH 2/3 rseq: Fix: Unregister rseq for CLONE_SETTLS
It has been reported by Google that rseq is not behaving properly with respect to clone when CLONE_VM is used without CLONE_THREAD. It keeps the prior thread's rseq TLS registered when the TLS of the thread has moved, so the kernel deals with the wrong TLS.
The approach of clearing the per task-struct rseq registration on clone with CLONE_THREAD flag is incomplete. It does not cover the use-case of clone with CLONE_VM set, but without CLONE_THREAD.
Looking more closely at each of the clone flags:
- CLONE_THREAD,
- CLONE_VM,
- CLONE_SETTLS.
It appears that the flag we really want to track is CLONE_SETTLS, which moves the location of the TLS for the child, making the rseq registration point to the wrong TLS.
--- include/linux/sched.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
include/linux/sched.h +2 -2
@@ -1919,11 +1919,11 @@ static inline void rseq_migrate(struct task_struct *t)/* * If parent process has a registered restartable sequences area, the -* child inherits. Only applies when forking a process, not a thread. +* child inherits. Unregister rseq for a clone with CLONE_SETTLS set. */ static inline void rseq_fork (struct task_struct *t,unsigned long clone_flags) { -if (clone_flags &CLONE_THREAD ) { +if (clone_flags &CLONE_SETTLS ) { t->rseq =NULL ; t->rseq_sig =0 ; t->rseq_event_mask =0 ;
Signature · 1 line
2.17.1
Re: [PATCH for 5.3 2/3] rseq: Fix: Unregister rseq for CLONE_SETTLS
There is an ongoing discussion on the choice of flag we want to care about here. Therefore, please don't pull this patch until we reach an agreement.
Thanks,
Mathieu
Quoted 54 lines · ----- On Sep 13, 2019, at 11:12 AM, Mathieu Desnoyers…
It has been reported by Google that rseq is not behaving properly with respect to clone when
CLONE_VMis used withoutCLONE_THREAD. It keeps the prior thread's rseq TLS registered when the TLS of the thread has moved, so the kernel deals with the wrong TLS.The approach of clearing the per task-struct rseq registration on clone with
CLONE_THREADflag is incomplete. It does not cover the use-case of clone withCLONE_VMset, but withoutCLONE_THREAD.Looking more closely at each of the clone flags:
-
CLONE_THREAD,
-CLONE_VM,
-CLONE_SETTLS.It appears that the flag we really want to track is
CLONE_SETTLS, which moves the location of the TLS for the child, making the rseq registration point to the wrong TLS.Suggested-by "H . Peter Anvin" <[email protected]>Signed-off-by Mathieu Desnoyers <[email protected]>Cc Thomas Gleixner <[email protected]>Cc Peter Zijlstra (Intel) <[email protected]>Cc "Paul E. McKenney" <[email protected]>Cc Boqun Feng <[email protected]>Cc "H . Peter Anvin" <[email protected]>Cc Paul Turner <[email protected]>Cc Dmitry Vyukov <[email protected]>Cc <[email protected]>---
include/linux/sched.h| 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)include/linux/sched.h +2
-3@@ -1919,11 +1919,11 @@ static inline void rseq_migrate(struct task_struct *t)/* * If parent process has a registered restartable sequences area, the -* child inherits. Only applies when forking a process, not a thread. +* child inherits. Unregister rseq for a clone with CLONE_SETTLS set. */ static inline void rseq_fork (struct task_struct *t,unsigned long clone_flags) { -if (clone_flags &CLONE_THREAD ) { +if (clone_flags &CLONE_SETTLS ) { t->rseq =NULL ; t->rseq_sig =0 ; t->rseq_event_mask =0 ; --2.17.1
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Re: [PATCH for 5.3 2/3] rseq: Fix: Unregister rseq for CLONE_SETTLS
Quoted 3 lines · ----- On Sep 14, 2019, at 10:21 AM, Mathieu Desnoyers…
There is an ongoing discussion on the choice of flag we want to care about here. Therefore, please don't pull this patch until we reach an agreement.
Following discussion with Neel Natu (Google) and Paul Turner (Google), I plan to modify this patch, and unregister RSEQ on clone CLONE_VM for the following reasons:
1) CLONE_THREAD requires CLONE_SIGHAND, which requires CLONE_VM to be set. Therefore, just checking for CLONE_VM covers all CLONE_THREAD uses,
2) There is the possibility of an unlikely scenario where CLONE_SETTLS is used without CLONE_VM. In order to be an issue, it would require that the rseq TLS is in a shared memory area.
I do not plan on adding CLONE_SETTLS to the set of clone flags which unregister RSEQ, because it would require that we also unregister RSEQ on set_thread_area(2) and arch_prctl(2) ARCH_SET_FS for completeness. So rather than doing a partial solution, it appears better to let user-space explicitly perform rseq unregistration across clone if needed in scenarios where CLONE_VM is not set.
Thoughts ?
Thanks,
Mathieu
Quoted 67 lines
Thanks,
Mathieu
----- On Sep 13, 2019, at 11:12 AM, Mathieu Desnoyers
Quoted 53 lines · [email protected] wrote:
It has been reported by Google that rseq is not behaving properly with respect to clone when
CLONE_VMis used withoutCLONE_THREAD. It keeps the prior thread's rseq TLS registered when the TLS of the thread has moved, so the kernel deals with the wrong TLS.The approach of clearing the per task-struct rseq registration on clone with
CLONE_THREADflag is incomplete. It does not cover the use-case of clone withCLONE_VMset, but withoutCLONE_THREAD.Looking more closely at each of the clone flags:
-
CLONE_THREAD,
-CLONE_VM,
-CLONE_SETTLS.It appears that the flag we really want to track is
CLONE_SETTLS, which moves the location of the TLS for the child, making the rseq registration point to the wrong TLS.Suggested-by "H . Peter Anvin" <[email protected]>Signed-off-by Mathieu Desnoyers <[email protected]>Cc Thomas Gleixner <[email protected]>Cc Peter Zijlstra (Intel) <[email protected]>Cc "Paul E. McKenney" <[email protected]>Cc Boqun Feng <[email protected]>Cc "H . Peter Anvin" <[email protected]>Cc Paul Turner <[email protected]>Cc Dmitry Vyukov <[email protected]>Cc <[email protected]>---
include/linux/sched.h| 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)include/linux/sched.h +2
-3@@ -1919,11 +1919,11 @@ static inline void rseq_migrate(struct task_struct *t)/* * If parent process has a registered restartable sequences area, the -* child inherits. Only applies when forking a process, not a thread. +* child inherits. Unregister rseq for a clone with CLONE_SETTLS set. */ static inline void rseq_fork (struct task_struct *t,unsigned long clone_flags) { -if (clone_flags &CLONE_THREAD ) { +if (clone_flags &CLONE_SETTLS ) { t->rseq =NULL ; t->rseq_sig =0 ; t->rseq_event_mask =0 ; --Quoted 1 line
2.17.1
--
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
PATCH 3/3 rseq/selftests: Fix: Namespace gettid() for compatibility with glibc 2.30
In reply to Mathieu Desnoyersglibc 2.30 introduces gettid() in public headers, which clashes with the internal static definition within rseq selftests.
Rename gettid() to rseq_gettid() to eliminate this symbol name clash.
--- tools/testing/selftests/rseq/param_test.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-)
tools/testing/selftests/rseq/param_test.c +10 -8
@@ -15,7 +15,7 @@#include <errno.h> #include <stddef.h> -static inline pid_t gettid (void ) +static inline pid_t rseq_gettid (void ) {return syscall (__NR_gettid); }
@@ -373,11 +373,12 @@ void *test_percpu_spinlock_thread(void *arg)rseq_percpu_unlock (&data->lock, cpu);#ifndef BENCHMARK if (i !=0 && !(i % (reps /10 ))) -printf_verbose ("tid %d: count %lld\n" , (int )gettid (), i); +printf_verbose ("tid %d: count %lld\n" , + (int )rseq_gettid (), i);#endif }printf_verbose ("tid %d: number of rseq abort: %d, signals delivered: %u\n" , - (int )gettid (), nr_abort, signals_delivered); + (int )rseq_gettid (), nr_abort, signals_delivered);if (!opt_disable_rseq && thread_data->reg &&rseq_unregister_current_thread ())abort ();
@@ -454,11 +455,12 @@ void *test_percpu_inc_thread(void *arg) }while (rseq_unlikely (ret));#ifndef BENCHMARK if (i !=0 && !(i % (reps /10 ))) -printf_verbose ("tid %d: count %lld\n" , (int )gettid (), i); +printf_verbose ("tid %d: count %lld\n" , + (int )rseq_gettid (), i);#endif }printf_verbose ("tid %d: number of rseq abort: %d, signals delivered: %u\n" , - (int )gettid (), nr_abort, signals_delivered); + (int )rseq_gettid (), nr_abort, signals_delivered);if (!opt_disable_rseq && thread_data->reg &&rseq_unregister_current_thread ())abort ();
@@ -605,7 +607,7 @@ void *test_percpu_list_thread(void *arg) }printf_verbose ("tid %d: number of rseq abort: %d, signals delivered: %u\n" , - (int )gettid (), nr_abort, signals_delivered); + (int )rseq_gettid (), nr_abort, signals_delivered);if (!opt_disable_rseq &&rseq_unregister_current_thread ())abort ();
@@ -796,7 +798,7 @@ void *test_percpu_buffer_thread(void *arg) }printf_verbose ("tid %d: number of rseq abort: %d, signals delivered: %u\n" , - (int )gettid (), nr_abort, signals_delivered); + (int )rseq_gettid (), nr_abort, signals_delivered);if (!opt_disable_rseq &&rseq_unregister_current_thread ())abort ();
@@ -1011,7 +1013,7 @@ void *test_percpu_memcpy_buffer_thread(void *arg) }printf_verbose ("tid %d: number of rseq abort: %d, signals delivered: %u\n" , - (int )gettid (), nr_abort, signals_delivered); + (int )rseq_gettid (), nr_abort, signals_delivered);if (!opt_disable_rseq &&rseq_unregister_current_thread ())abort ();
Signature · 1 line
2.17.1
A message that is not in the archive (sent off-list or not yet received): [email protected]
Re: [PATCH for 5.3 3/3] rseq/selftests: Fix: Namespace gettid() for compatibility with glibc 2.30
Quoted 18 lines · ----- On Sep 14, 2019, at 3:47 PM, Sasha Levin [email protected] wrote:
Hi,
[This is an automated email]
This commit has been processed because it contains a -stable tag. The stable tag indicates that it's relevant for the following trees: all
The bot has tested the following trees: v5.2.14, v4.19.72, v4.14.143, v4.9.192, v4.4.192.
v5.2.14: Build OK!
v4.19.72: Build OK!
v4.14.143: Failed to apply! Possible dependencies:c960e9909d33 ("rseq/selftests: Provide parametrized tests")v4.9.192: Failed to apply! Possible dependencies:
c960e9909d33 ("rseq/selftests: Provide parametrized tests")
Quoted 7 lines
v4.4.192: Failed to apply! Possible dependencies:
c960e9909d33 ("rseq/selftests: Provide parametrized tests")NOTE: The patch will not be queued to stable trees until it is upstream.
How should we proceed with this patch?
rseq was merged into 4.18, so none of those patches are needed prior to that kernel version.
This applies to all 3 rseq patches.
Thanks,
Mathieu
Quoted 4 lines
--
Thanks,
Sasha
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com