linux-user: rt_sigprocmask, check read perms first
PATCH 1/1 linux-user: rt_sigprocmask, check read perms first
From: Shu-Chun Weng <[email protected]>
Linux kernel does it this way (checks read permission before validating how) and the latest version of ABSL's AddressIsReadable() depends on this behavior.
--- linux-user/syscall.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
linux-user/syscall.c +5 -5
@@ -9491,6 +9491,11 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1, }if (arg2) { +if (!(p =lock_user (VERIFY_READ , arg2,sizeof (target_sigset_t),1 ))) +return -TARGET_EFAULT ; +target_to_host_sigset (&set, p); +unlock_user (p, arg2,0 ); + set_ptr = &set;switch (how) {case TARGET_SIG_BLOCK : how =SIG_BLOCK ;
@@ -9504,11 +9509,6 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,default :return -TARGET_EINVAL ; } -if (!(p =lock_user (VERIFY_READ , arg2,sizeof (target_sigset_t),1 ))) -return -TARGET_EFAULT ; -target_to_host_sigset (&set, p); -unlock_user (p, arg2,0 ); - set_ptr = &set; }else { how =0 ; set_ptr =NULL ;
Signature · 1 line
2.34.1.448.ga2b2bfdf31-goog
Le 06/01/2022 à 23:00, Patrick Venture a écrit :
Quoted 31 lines
From: Shu-Chun Weng <[email protected]>
Linux kernel does it this way (checks read permission before validating
how) and the latest version of ABSL'sAddressIsReadable()depends on this behavior.Reviewed-by Patrick Venture <[email protected]>Signed-off-by Shu-Chun Weng <[email protected]>--- linux-user/syscall.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)linux-user/syscall.c +5
-0@@ -9491,6 +9491,11 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1, }if (arg2) { +if (!(p =lock_user (VERIFY_READ , arg2,sizeof (target_sigset_t),1 ))) +return -TARGET_EFAULT ; +target_to_host_sigset (&set, p); +unlock_user (p, arg2,0 ); + set_ptr = &set;switch (how) {case TARGET_SIG_BLOCK : how =SIG_BLOCK ;@@ -9504,11 +9509,6 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,default :
linux-user/syscall.c static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1, · line 9514
Quoted 10 lines
linux-user/syscall.c +0
-5return -TARGET_EINVAL ; } -if (!(p =lock_user (VERIFY_READ , arg2,sizeof (target_sigset_t),1 ))) -return -TARGET_EFAULT ; -target_to_host_sigset (&set, p); -unlock_user (p, arg2,0 ); - set_ptr = &set; }else { how =0 ; set_ptr =NULL ;
I know it's only code move but generally we also update the style to pass scripts/checkpatch.pl successfully.
Could you also update TARGET_NR_sigprocmask in the same way as it seems the kernel behaves like this too in this case?
Thanks,
Laurent
Quoted 49 lines · Laurent Vivier
Quoted 3 lines · Patrick Venture
From: Shu-Chun Weng <[email protected]>
Linux kernel does it this way (checks read permission before validating
how)Quoted 4 lines
and the latest version of ABSL's
AddressIsReadable()depends on this behavior.c.f.
linux-user/syscall.c static abi_long do_syscall1(void *cpu_env, int
Quoted 11 lines
Reviewed-by Patrick Venture <[email protected]>Signed-off-by Shu-Chun Weng <[email protected]>--- linux-user/syscall.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)linux-user/syscall.c +0
-0@@ -9491,6 +9491,11 @@ static abi_long do_syscall1(void *cpu_env, intnum,
abi_longarg1,linux-user/syscall.c static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
Quoted 4 lines
linux-user/syscall.c +1
-0} if (arg2) { + if (!(p = lock_user(VERIFY_READ, arg2,
sizeof(target_sigset_t), 1)))linux-user/syscall.c static abi_long do_syscall1(void *cpu_env, int
Quoted 8 lines
linux-user/syscall.c +4
-0+return -TARGET_EFAULT ; +target_to_host_sigset (&set, p); +unlock_user (p, arg2,0 ); + set_ptr = &set;switch (how) {case TARGET_SIG_BLOCK : how =SIG_BLOCK ;@@ -9504,11 +9509,6 @@ static abi_long do_syscall1(void *cpu_env, intnum,
abi_longarg1,linux-user/syscall.c static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
Quoted 4 lines
linux-user/syscall.c +0
-1default :return -TARGET_EINVAL ; } - if (!(p = lock_user(VERIFY_READ, arg2,
sizeof(target_sigset_t), 1)))linux-user/syscall.c static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1, · line 9514
Quoted 7 lines
linux-user/syscall.c +0
-4-return -TARGET_EFAULT ; -target_to_host_sigset (&set, p); -unlock_user (p, arg2,0 ); - set_ptr = &set; }else { how =0 ; set_ptr =NULL ;
Quoted 4 lines
I know it's only code move but generally we also update the style to pass
scripts/checkpatch.pl
successfully.
That is a reasonable request, however, can I just send a follow-on patch? I didn't write this one and I honestly don't know much about it, but I don't mind doing the cleanup
Quoted 5 lines
Could you also update
TARGET_NR_sigprocmaskin the same way as it seems the kernel behaves like this
too in this case?
I can take a look. I would prefer then to also prefetch the style fixup in a preceding patch. I don't recall seeing whether qemu supports clang-format.
Quoted 4 lines
Thanks,
Laurent
Patrick
Hi Patrick,
Le 11/01/2022 à 21:14, Patrick Venture a écrit :
Quoted 57 lines · On Sat, Jan 8, 2022 at 10:16 AM Laurent Vivier <[email protected]…
Le 06/01/2022 à 23:00, Patrick Venture a écrit : > From: Shu-Chun Weng <[email protected] <mailto:[email protected]>> > > Linux kernel does it this way (checks read permission before validating `how`) > and the latest version of ABSL's `AddressIsReadable()` depends on this > behavior. > > c.f. https://github.com/torvalds/linux/blob/9539ba4308ad5bdca6cb41c7b73cbb9f796dcdd7/kernel/signal.c#L3147 <https://github.com/torvalds/linux/blob/9539ba4308ad5bdca6cb41c7b73cbb9f796dcdd7/kernel/signal.c#L3147> > Reviewed-by: Patrick Venture <[email protected] <mailto:[email protected]>> > Signed-off-by: Shu-Chun Weng <[email protected] <mailto:[email protected]>> > --- > linux-user/syscall.c | 10 +++++----- > 1 file changed, 5 insertions(+), 5 deletions(-) > > diff --git a/linux-user/syscall.c b/linux-user/syscall.c > index ce9d64896c..3070d31f34 100644 > --- a/linux-user/syscall.c > +++ b/linux-user/syscall.c > @@ -9491,6 +9491,11 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1, > } > > if (arg2) { > + if (!(p = lock_user(VERIFY_READ, arg2, sizeof(target_sigset_t), 1))) > + return -TARGET_EFAULT; > + target_to_host_sigset(&set, p); > + unlock_user(p, arg2, 0); > + set_ptr = &set; > switch(how) { > case TARGET_SIG_BLOCK: > how = SIG_BLOCK; > @@ -9504,11 +9509,6 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1, > default: > return -TARGET_EINVAL; > } > - if (!(p = lock_user(VERIFY_READ, arg2, sizeof(target_sigset_t), 1))) > - return -TARGET_EFAULT; > - target_to_host_sigset(&set, p); > - unlock_user(p, arg2, 0); > - set_ptr = &set; > } else { > how = 0; > set_ptr = NULL; I know it's only code move but generally we also update the style to pass scripts/checkpatch.pl <http://checkpatch.pl> successfully.That is a reasonable request, however, can I just send a follow-on patch? I didn't write this one and I honestly don't know much about it, but I don't mind doing the cleanup
Quoted 7 lines
Could you also update
TARGET_NR_sigprocmaskin the same way as it seems the kernel behaves like this
too in this case?I can take a look. I would prefer then to also prefetch the style fixup in a preceding patch. I don't recall seeing whether qemu supports clang-format.
There is no problem. You can keep this patch unmodified, and add patches to fix the problems.
I only ask to have all the patches in one series.
Thanks,
Laurent
Quoted 83 lines · Laurent Vivier
Hi Patrick,
Quoted 3 lines · Patrick Venture
On Sat, Jan 8, 2022 at 10:16 AM Laurent Vivier <[email protected]
Quoted 5 lines · <mailto:[email protected]>> wrote:
Le 06/01/2022 à 23:00, Patrick Venture a écrit :
> From: Shu-Chun Weng <[email protected] <mailto:[email protected]>> >
> Linux kernel does it this way (checks read permission beforevalidating
how)Quoted 1 line
> and the latest version of ABSL's
AddressIsReadable()depends onthis
Quoted 4 lines
> behavior. > > c.f.Quoted 1 line
<Quoted 2 lines
> Reviewed-by: Patrick Venture <[email protected] <mailto:
Quoted 1 line
> Signed-off-by: Shu-Chun Weng <[email protected] <mailto:
Quoted 9 lines
> --- > linux-user/syscall.c | 10 +++++----- > 1 file changed, 5 insertions(+), 5 deletions(-) > > diff --git a/linux-user/syscall.c b/linux-user/syscall.c > index ce9d64896c..3070d31f34 100644 > --- a/linux-user/syscall.c > +++ b/linux-user/syscall.c > @@ -9491,6 +9491,11 @@ static abi_long do_syscall1(void *cpu_env,int num,
abi_longarg1,Quoted 4 lines
> } > > if (arg2) { > + if (!(p = lock_user(VERIFY_READ, arg2,
sizeof(target_sigset_t), 1)))Quoted 8 lines
> + return -TARGET_EFAULT; > + target_to_host_sigset(&set, p); > + unlock_user(p, arg2, 0); > + set_ptr = &set; > switch(how) { > case TARGET_SIG_BLOCK: > how = SIG_BLOCK; > @@ -9504,11 +9509,6 @@ static abi_long do_syscall1(void *cpu_env,int num,
abi_longarg1,Quoted 4 lines
> default: > return -TARGET_EINVAL; > } > - if (!(p = lock_user(VERIFY_READ, arg2,
sizeof(target_sigset_t), 1)))Quoted 9 lines
> - return -TARGET_EFAULT; > - target_to_host_sigset(&set, p); > - unlock_user(p, arg2, 0); > - set_ptr = &set; > } else { > how = 0; > set_ptr = NULL; I know it's only code move but generally we also update the style topass
scripts/checkpatch.plQuoted 5 lines
<http://checkpatch.pl> successfully.That is a reasonable request, however, can I just send a follow-on
patch? I didn't write this one
Quoted 1 line
and I honestly don't know much about it, but I don't mind doing the
cleanup
Quoted 3 lines
Could you also update
TARGET_NR_sigprocmaskin the same way as itseems the kernel behaves like
Quoted 5 lines
this
too in this case?I can take a look. I would prefer then to also prefetch the style fixup
Quoted 9 lines
in a preceding patch. I
Quoted 2 lines
don't recall seeing whether qemu supports clang-format.
There is no problem. You can keep this patch unmodified, and add patches to fix the problems.
I only ask to have all the patches in one series.
Will take a swing at this for v2.
Quoted 5 lines
Thanks,
Laurent
Quoted 91 lines · Patrick Venture
Quoted 87 lines · Laurent Vivier
Hi Patrick,
Quoted 3 lines · Patrick Venture
On Sat, Jan 8, 2022 at 10:16 AM Laurent Vivier <[email protected]
Quoted 5 lines · <mailto:[email protected]>> wrote:
Le 06/01/2022 à 23:00, Patrick Venture a écrit :
> From: Shu-Chun Weng <[email protected] <mailto:[email protected]>> >
> Linux kernel does it this way (checks read permission beforevalidating
how)Quoted 1 line
> and the latest version of ABSL's
AddressIsReadable()dependson this
Quoted 4 lines
> behavior. > > c.f.Quoted 1 line
<Quoted 2 lines
> Reviewed-by: Patrick Venture <[email protected] <mailto:
Quoted 1 line
> Signed-off-by: Shu-Chun Weng <[email protected] <mailto:
Quoted 9 lines
> --- > linux-user/syscall.c | 10 +++++----- > 1 file changed, 5 insertions(+), 5 deletions(-) > > diff --git a/linux-user/syscall.c b/linux-user/syscall.c > index ce9d64896c..3070d31f34 100644 > --- a/linux-user/syscall.c > +++ b/linux-user/syscall.c > @@ -9491,6 +9491,11 @@ static abi_long do_syscall1(void
*cpu_env, int num,abi_longarg1,Quoted 4 lines
> } > > if (arg2) { > + if (!(p = lock_user(VERIFY_READ, arg2,
sizeof(target_sigset_t), 1)))Quoted 8 lines
> + return -TARGET_EFAULT; > + target_to_host_sigset(&set, p); > + unlock_user(p, arg2, 0); > + set_ptr = &set; > switch(how) { > case TARGET_SIG_BLOCK: > how = SIG_BLOCK; > @@ -9504,11 +9509,6 @@ static abi_long do_syscall1(void
*cpu_env, int num,abi_longarg1,Quoted 4 lines
> default: > return -TARGET_EINVAL; > } > - if (!(p = lock_user(VERIFY_READ, arg2,
sizeof(target_sigset_t), 1)))Quoted 9 lines
> - return -TARGET_EFAULT; > - target_to_host_sigset(&set, p); > - unlock_user(p, arg2, 0); > - set_ptr = &set; > } else { > how = 0; > set_ptr = NULL; I know it's only code move but generally we also update the styleto pass
scripts/checkpatch.plQuoted 5 lines
<http://checkpatch.pl> successfully.That is a reasonable request, however, can I just send a follow-on
patch? I didn't write this one
Quoted 1 line
and I honestly don't know much about it, but I don't mind doing the
cleanup
Quoted 3 lines
Could you also update
TARGET_NR_sigprocmaskin the same way as itseems the kernel behaves like
Quoted 5 lines
this
too in this case?I can take a look. I would prefer then to also prefetch the style
fixup in a preceding patch. I
Quoted 2 lines
don't recall seeing whether qemu supports clang-format.
Quoted 8 lines
Quoted 5 lines
There is no problem. You can keep this patch unmodified, and add patches to fix the problems.
I only ask to have all the patches in one series.
Will take a swing at this for v2.
Laurent,
I spent some time today going over the patches and digging into what they're actually doing and I think I can make this two patches, both with the style changes squashed and will send in a v2 today.
Thanks
Quoted 7 lines
Quoted 7 lines
Thanks,
Laurent