arm, arm64: check for number of arguments in syscall_get/set_arguments()
Subsystems (4)
A message that is not in the archive (sent off-list or not yet received): [email protected]
v2 cover arm, arm64: check for number of arguments in syscall_get/set_arguments()
# added Roland <mcgrathr@chromium> in CC for asm-generic/syscall.h
I modified my patch based on Will's comments.
Speaking of arm, i + n is checked against SYSCALL_MAX_ARGS (== 7), but it seems that the kernel code assumes that the current max number of arguments would be 6. For example, SYSCALL_DEFINEn are defined from 0 to 6. Am i wrong?
Changes:
* added a patch to arm as well
* added a patch to asm-generic/syscall.h; corrected the description on syscall_get_arguments() and syscall_put_arguments()
* removed portion of commit message which referred to asm-generic/syscall.h in arm64
AKASHI Takahiro (3):
arm64: check for number of arguments in syscall_get/set_arguments()
arm: check for number of arguments in syscall_get/set_arguments()
asm-generic: syscall_get/set_arguments accept zero for number of
arguments
arch/arm/include/asm/syscall.h | 6 ++++++ arch/arm64/include/asm/syscall.h | 6 ++++++ include/asm-generic/syscall.h | 4 ++-- 3 files changed, 14 insertions(+), 2 deletions(-)
Signature · 1 line
1.7.9.5
PATCH v2 1/3 arm64: check for number of arguments in syscall_get/set_arguments()
In ftrace_syscall_enter(),
syscall_get_arguments(..., 0, n, ...)
if (i == 0) { <handle orig_x0> ...; n--;}
memcpy(..., n * sizeof(args[0]));If 'number of arguments(n)' is zero and 'argument index(i)' is also zero in syscall_get_arguments(), none of arguments should be copied by memcpy(). Otherwise 'n--' can be a big positive number and unexpected amount of data will be copied. Tracing system calls which take no argument, say sync(void), may hit this case and eventually make the system corrupted.
This patch fixes the issue both in syscall_get_arguments() and syscall_set_arguments().
--- arch/arm64/include/asm/syscall.h | 6 ++++++ 1 file changed, 6 insertions(+)
arch/arm64/include/asm/syscall.h +6 -0
@@ -63,6 +63,9 @@ static inline void syscall_get_arguments(struct task_struct *task, unsigned int i, unsigned int n, unsigned long *args) { + if (n == 0) + return; +if (i + n >SYSCALL_MAX_ARGS ) {unsigned long *args_bad = args +SYSCALL_MAX_ARGS - i;unsigned int n_bad = n + i -SYSCALL_MAX_ARGS ;
@@ -86,6 +89,9 @@ static inline void syscall_set_arguments(struct task_struct *task, unsigned int i, unsigned int n, const unsigned long *args) { + if (n == 0) + return; +if (i + n >SYSCALL_MAX_ARGS ) {pr_warning ("%s called with max args %d, handling only %d\n" , __func__, i + n,SYSCALL_MAX_ARGS );
Signature · 1 line
1.7.9.5
PATCH v2 1/3 arm64: check for number of arguments in syscall_get/set_arguments()
Quoted 12 lines · AKASHI Takahiro
In
ftrace_syscall_enter(),syscall_get_arguments(..., 0, n, ...) if (i == 0) { <handle orig_x0> ...; n--;} memcpy(..., n * sizeof(args[0]));If 'number of arguments(n)' is zero and 'argument index(i)' is also zero in
syscall_get_arguments(), none of arguments should be copied bymemcpy(). Otherwise 'n--' can be a big positive number and unexpected amount of data will be copied. Tracing system calls which take no argument, say sync(void), may hit this case and eventually make the system corrupted.
This patch fixes the issue both insyscall_get_arguments()andsyscall_set_arguments().
Quoted 1 line
Signed-off-by AKASHI Takahiro <[email protected]>
Cheers,
Will
PATCH v2 2/3 arm: check for number of arguments in syscall_get/set_arguments()
In reply to AKASHI TakahiroIn ftrace_syscall_enter(),
syscall_get_arguments(..., 0, n, ...)
if (i == 0) { <handle ORIG_r0> ...; n--;}
memcpy(..., n * sizeof(args[0]));If 'number of arguments(n)' is zero and 'argument index(i)' is also zero in syscall_get_arguments(), none of arguments should be copied by memcpy(). Otherwise 'n--' can be a big positive number and unexpected amount of data will be copied. Tracing system calls which take no argument, say sync(void), may hit this case and eventually make the system corrupted.
This patch fixes the issue both in syscall_get_arguments() and syscall_set_arguments().
--- arch/arm/include/asm/syscall.h | 6 ++++++ 1 file changed, 6 insertions(+)
arch/arm/include/asm/syscall.h +6 -0
@@ -57,6 +57,9 @@ static inline void syscall_get_arguments(struct task_struct *task, unsigned int i, unsigned int n, unsigned long *args) { + if (n == 0) + return; +if (i + n >SYSCALL_MAX_ARGS ) {unsigned long *args_bad = args +SYSCALL_MAX_ARGS - i;unsigned int n_bad = n + i -SYSCALL_MAX_ARGS ;
@@ -81,6 +84,9 @@ static inline void syscall_set_arguments(struct task_struct *task, unsigned int i, unsigned int n, const unsigned long *args) { + if (n == 0) + return; +if (i + n >SYSCALL_MAX_ARGS ) {pr_warning ("%s called with max args %d, handling only %d\n" , __func__, i + n,SYSCALL_MAX_ARGS );
Signature · 1 line
1.7.9.5
PATCH v2 2/3 arm: check for number of arguments in syscall_get/set_arguments()
Quoted 32 lines · AKASHI Takahiro
In
ftrace_syscall_enter(),syscall_get_arguments(..., 0, n, ...) if (i == 0) { <handle ORIG_r0> ...; n--;} memcpy(..., n * sizeof(args[0]));If 'number of arguments(n)' is zero and 'argument index(i)' is also zero in
syscall_get_arguments(), none of arguments should be copied bymemcpy(). Otherwise 'n--' can be a big positive number and unexpected amount of data will be copied. Tracing system calls which take no argument, say sync(void), may hit this case and eventually make the system corrupted.
This patch fixes the issue both insyscall_get_arguments()andsyscall_set_arguments().Signed-off-by AKASHI Takahiro <[email protected]>--- arch/arm/include/asm/syscall.h | 6 ++++++ 1 file changed, 6 insertions(+)arch/arm/include/asm/syscall.h +3
-0@@ -57,6 +57,9 @@ static inline void syscall_get_arguments(struct task_struct *task, unsigned int i, unsigned int n, unsigned long *args) { + if (n == 0) + return; +if (i + n >SYSCALL_MAX_ARGS ) {unsigned long *args_bad = args +SYSCALL_MAX_ARGS - i;unsigned int n_bad = n + i -SYSCALL_MAX_ARGS ;
arch/arm/include/asm/syscall.h static inline void syscall_set_arguments(struct task_struct *task, · line 92
Quoted 9 lines
arch/arm/include/asm/syscall.h +3
-0@@ -81,6 +84,9 @@ static inline void syscall_set_arguments(struct task_struct *task, unsigned int i, unsigned int n, const unsigned long *args) { + if (n == 0) + return; +if (i + n >SYSCALL_MAX_ARGS ) {pr_warning ("%s called with max args %d, handling only %d\n" , __func__, i + n,SYSCALL_MAX_ARGS );
Please can you put this into the patch system, with a CC stable tag?
Will
PATCH v2 3/3 asm-generic: syscall_get/set_arguments accept zero for number of arguments
In reply to AKASHI TakahiroThere is a possibility of calling syscall_get_arguments() with zero as 'number of arguments(n)' in ftrace_syscall_enter().
The comments in syscall_get_argumets() as well as syscall_set_arugments() are corrected to clarify that.
--- include/asm-generic/syscall.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
include/asm-generic/syscall.h +2 -2
@@ -106,7 +106,7 @@ void syscall_set_return_value(struct task_struct *task, struct pt_regs *regs,* @task: task of interest, must be blocked * @regs: task_pt_regs() of @task * @i: argument index [0,5] -* @n: number of arguments; n+i must be [1,6]. +* @n: number of arguments; n+i must be [0,6]. * @args: array filled with argument values * * Fetches @n arguments to the system call starting with the @i'th argument
@@ -126,7 +126,7 @@ void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs,* @task: task of interest, must be in system call entry tracing * @regs: task_pt_regs() of @task * @i: argument index [0,5] -* @n: number of arguments; n+i must be [1,6]. +* @n: number of arguments; n+i must be [0,6]. * @args: array of argument values to store * * Changes @n arguments to the system call starting with the @i'th argument.
Signature · 1 line
1.7.9.5
Re: [PATCH v2 3/3] asm-generic: syscall_get/set_arguments accept zero for number of arguments
[extending CC list to add the ftrace guys]
Quoted 4 lines · AKASHI Takahiro
There is a possibility of calling
syscall_get_arguments()with zero as 'number of arguments(n)' inftrace_syscall_enter().
The comments insyscall_get_argumets()as well assyscall_set_arugments()are corrected to clarify that.
Probably also worth mentioning that all architectures deal with n + i == 0 (after your first two patches fixing up arm and arm64, anyway).
Quoted 19 lines
Signed-off-by AKASHI Takahiro <[email protected]>--- include/asm-generic/syscall.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)include/asm-generic/syscall.h +1
-1@@ -106,7 +106,7 @@ void syscall_set_return_value(struct task_struct *task, struct pt_regs *regs,* @task: task of interest, must be blocked * @regs: task_pt_regs() of @task * @i: argument index [0,5] -* @n: number of arguments; n+i must be [1,6]. +* @n: number of arguments; n+i must be [0,6]. * @args: array filled with argument values * * Fetches @n arguments to the system call starting with the @i'th argument
include/asm-generic/syscall.h void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, · line 132
Quoted 8 lines
include/asm-generic/syscall.h +1
-1@@ -126,7 +126,7 @@ void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs,* @task: task of interest, must be in system call entry tracing * @regs: task_pt_regs() of @task * @i: argument index [0,5] -* @n: number of arguments; n+i must be [1,6]. +* @n: number of arguments; n+i must be [0,6]. * @args: array of argument values to store * * Changes @n arguments to the system call starting with the @i'th argument.
Will
v2 cover arm, arm64: check for number of arguments in syscall_get/set_arguments()
In reply to AKASHI TakahiroQuoted 7 lines · AKASHI Takahiro
# added Roland <mcgrathr@chromium> in CC for asm-generic/syscall.hI modified my patch based on Will's comments.
Speaking of arm, i + n is checked againstSYSCALL_MAX_ARGS(== 7), but it seems that the kernel code assumes that the current max number of arguments would be 6. For example,SYSCALL_DEFINEnare defined from 0 to 6. Am i wrong?
We do have a 7 argument syscall on OABI ARM. See the original thread:
http://lists.infradead.org/pipermail/linux-arm-kernel/2012-February/085509.html
Will