| Message ID | 20220304172852.274126-2-benjamin.tissoires@redhat.com |
|---|---|
| State | Superseded |
| Headers | show |
| Series | Introduce eBPF support for HID devices | expand |
On Fri, Mar 4, 2022 at 9:30 AM Benjamin Tissoires <benjamin.tissoires@redhat.com> wrote: > > LIRC_MODE2 does not really need net_admin capability, but only sys_admin. > > Extract a new helper for it, it will be also used for the HID bpf > implementation. > > Cc: Sean Young <sean@mess.org> > Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> > > --- > > new in v2 > --- > kernel/bpf/syscall.c | 14 +++++++++++++- > 1 file changed, 13 insertions(+), 1 deletion(-) > > diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c > index db402ebc5570..cc570891322b 100644 > --- a/kernel/bpf/syscall.c > +++ b/kernel/bpf/syscall.c > @@ -2165,7 +2165,6 @@ static bool is_net_admin_prog_type(enum bpf_prog_type prog_type) > case BPF_PROG_TYPE_LWT_SEG6LOCAL: > case BPF_PROG_TYPE_SK_SKB: > case BPF_PROG_TYPE_SK_MSG: > - case BPF_PROG_TYPE_LIRC_MODE2: > case BPF_PROG_TYPE_FLOW_DISSECTOR: > case BPF_PROG_TYPE_CGROUP_DEVICE: > case BPF_PROG_TYPE_CGROUP_SOCK: > @@ -2202,6 +2201,17 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) > } > } > > +static bool is_sys_admin_prog_type(enum bpf_prog_type prog_type) > +{ > + switch (prog_type) { > + case BPF_PROG_TYPE_LIRC_MODE2: > + case BPF_PROG_TYPE_EXT: /* extends any prog */ > + return true; > + default: > + return false; > + } > +} I am not sure whether we should do this. This is a behavior change, that may break some user space. Also, BPF_PROG_TYPE_EXT is checked in is_perfmon_prog_type(), and this change will make that case useless. Thanks, Song [...]
On Sat, Mar 5, 2022 at 12:12 AM Song Liu <song@kernel.org> wrote: > > On Fri, Mar 4, 2022 at 9:30 AM Benjamin Tissoires > <benjamin.tissoires@redhat.com> wrote: > > > > LIRC_MODE2 does not really need net_admin capability, but only sys_admin. > > > > Extract a new helper for it, it will be also used for the HID bpf > > implementation. > > > > Cc: Sean Young <sean@mess.org> > > Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> > > > > --- > > > > new in v2 > > --- > > kernel/bpf/syscall.c | 14 +++++++++++++- > > 1 file changed, 13 insertions(+), 1 deletion(-) > > > > diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c > > index db402ebc5570..cc570891322b 100644 > > --- a/kernel/bpf/syscall.c > > +++ b/kernel/bpf/syscall.c > > @@ -2165,7 +2165,6 @@ static bool is_net_admin_prog_type(enum bpf_prog_type prog_type) > > case BPF_PROG_TYPE_LWT_SEG6LOCAL: > > case BPF_PROG_TYPE_SK_SKB: > > case BPF_PROG_TYPE_SK_MSG: > > - case BPF_PROG_TYPE_LIRC_MODE2: > > case BPF_PROG_TYPE_FLOW_DISSECTOR: > > case BPF_PROG_TYPE_CGROUP_DEVICE: > > case BPF_PROG_TYPE_CGROUP_SOCK: > > @@ -2202,6 +2201,17 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) > > } > > } > > > > +static bool is_sys_admin_prog_type(enum bpf_prog_type prog_type) > > +{ > > + switch (prog_type) { > > + case BPF_PROG_TYPE_LIRC_MODE2: > > + case BPF_PROG_TYPE_EXT: /* extends any prog */ > > + return true; > > + default: > > + return false; > > + } > > +} > > I am not sure whether we should do this. This is a behavior change, that may > break some user space. Also, BPF_PROG_TYPE_EXT is checked in > is_perfmon_prog_type(), and this change will make that case useless. Sure, I can drop it from v3 and make this function appear for HID only. Regarding BPF_PROG_TYPE_EXT, it was already in both is_net_admin_prog_type() and is_perfmon_prog_type(), so I duplicated it here, but I agree, given that it's already in the first function there, CPA_SYS_ADMIN is already checked. Cheers, Benjamin > > Thanks, > Song > > [...] >
On Sat, Mar 05, 2022 at 11:07:04AM +0100, Benjamin Tissoires wrote: > On Sat, Mar 5, 2022 at 12:12 AM Song Liu <song@kernel.org> wrote: > > > > On Fri, Mar 4, 2022 at 9:30 AM Benjamin Tissoires > > <benjamin.tissoires@redhat.com> wrote: > > > > > > LIRC_MODE2 does not really need net_admin capability, but only sys_admin. > > > > > > Extract a new helper for it, it will be also used for the HID bpf > > > implementation. > > > > > > Cc: Sean Young <sean@mess.org> > > > Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> > > > > > > --- > > > > > > new in v2 > > > --- > > > kernel/bpf/syscall.c | 14 +++++++++++++- > > > 1 file changed, 13 insertions(+), 1 deletion(-) > > > > > > diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c > > > index db402ebc5570..cc570891322b 100644 > > > --- a/kernel/bpf/syscall.c > > > +++ b/kernel/bpf/syscall.c > > > @@ -2165,7 +2165,6 @@ static bool is_net_admin_prog_type(enum bpf_prog_type prog_type) > > > case BPF_PROG_TYPE_LWT_SEG6LOCAL: > > > case BPF_PROG_TYPE_SK_SKB: > > > case BPF_PROG_TYPE_SK_MSG: > > > - case BPF_PROG_TYPE_LIRC_MODE2: > > > case BPF_PROG_TYPE_FLOW_DISSECTOR: > > > case BPF_PROG_TYPE_CGROUP_DEVICE: > > > case BPF_PROG_TYPE_CGROUP_SOCK: > > > @@ -2202,6 +2201,17 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) > > > } > > > } > > > > > > +static bool is_sys_admin_prog_type(enum bpf_prog_type prog_type) > > > +{ > > > + switch (prog_type) { > > > + case BPF_PROG_TYPE_LIRC_MODE2: > > > + case BPF_PROG_TYPE_EXT: /* extends any prog */ > > > + return true; > > > + default: > > > + return false; > > > + } > > > +} > > > > I am not sure whether we should do this. This is a behavior change, that may > > break some user space. Also, BPF_PROG_TYPE_EXT is checked in > > is_perfmon_prog_type(), and this change will make that case useless. > > Sure, I can drop it from v3 and make this function appear for HID only. For BPF_PROG_TYPE_LIRC_MODE2, I don't think this change will break userspace. This is called from ir-keytable(1) which is called from udev. It should have all the necessary permissions. In addition, the vast majority IR decoders are non-bpf. bpf ir decoders have very few users at the moment. I am working on completely new userspace tooling which will make extensive use of bpf ir decoding with full lircd and IRP compatibility, but this is not finished yet (see https://github.com/seanyoung/cir). Thanks Sean
On Sat, Mar 5, 2022 at 2:07 AM Benjamin Tissoires <benjamin.tissoires@redhat.com> wrote: > > On Sat, Mar 5, 2022 at 12:12 AM Song Liu <song@kernel.org> wrote: > > > > On Fri, Mar 4, 2022 at 9:30 AM Benjamin Tissoires > > <benjamin.tissoires@redhat.com> wrote: > > > > > > LIRC_MODE2 does not really need net_admin capability, but only sys_admin. > > > > > > Extract a new helper for it, it will be also used for the HID bpf > > > implementation. > > > > > > Cc: Sean Young <sean@mess.org> > > > Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> > > > > > > --- > > > > > > new in v2 > > > --- > > > kernel/bpf/syscall.c | 14 +++++++++++++- > > > 1 file changed, 13 insertions(+), 1 deletion(-) > > > > > > diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c > > > index db402ebc5570..cc570891322b 100644 > > > --- a/kernel/bpf/syscall.c > > > +++ b/kernel/bpf/syscall.c > > > @@ -2165,7 +2165,6 @@ static bool is_net_admin_prog_type(enum bpf_prog_type prog_type) > > > case BPF_PROG_TYPE_LWT_SEG6LOCAL: > > > case BPF_PROG_TYPE_SK_SKB: > > > case BPF_PROG_TYPE_SK_MSG: > > > - case BPF_PROG_TYPE_LIRC_MODE2: > > > case BPF_PROG_TYPE_FLOW_DISSECTOR: > > > case BPF_PROG_TYPE_CGROUP_DEVICE: > > > case BPF_PROG_TYPE_CGROUP_SOCK: > > > @@ -2202,6 +2201,17 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) > > > } > > > } > > > > > > +static bool is_sys_admin_prog_type(enum bpf_prog_type prog_type) > > > +{ > > > + switch (prog_type) { > > > + case BPF_PROG_TYPE_LIRC_MODE2: > > > + case BPF_PROG_TYPE_EXT: /* extends any prog */ > > > + return true; > > > + default: > > > + return false; > > > + } > > > +} > > > > I am not sure whether we should do this. This is a behavior change, that may > > break some user space. Also, BPF_PROG_TYPE_EXT is checked in > > is_perfmon_prog_type(), and this change will make that case useless. > > Sure, I can drop it from v3 and make this function appear for HID only. > > Regarding BPF_PROG_TYPE_EXT, it was already in both > is_net_admin_prog_type() and is_perfmon_prog_type(), so I duplicated > it here, but I agree, given that it's already in the first function > there, CPA_SYS_ADMIN is already checked. I think with current code, a user with CAP_BPF, CAP_NET_ADMIN, and CAP_PERFMON (but not CAP_SYS_ADMIN) can load programs of type BPF_PROG_TYPE_EXT. But after the patch, the same user will not be able to do it. Did I misread it? It is not a common case though.
On Sat, Mar 5, 2022 at 8:58 AM Sean Young <sean@mess.org> wrote: > > On Sat, Mar 05, 2022 at 11:07:04AM +0100, Benjamin Tissoires wrote: > > On Sat, Mar 5, 2022 at 12:12 AM Song Liu <song@kernel.org> wrote: > > > > > > On Fri, Mar 4, 2022 at 9:30 AM Benjamin Tissoires > > > <benjamin.tissoires@redhat.com> wrote: > > > > > > > > LIRC_MODE2 does not really need net_admin capability, but only sys_admin. > > > > > > > > Extract a new helper for it, it will be also used for the HID bpf > > > > implementation. > > > > > > > > Cc: Sean Young <sean@mess.org> > > > > Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> > > > > > > > > --- > > > > > > > > new in v2 > > > > --- > > > > kernel/bpf/syscall.c | 14 +++++++++++++- > > > > 1 file changed, 13 insertions(+), 1 deletion(-) > > > > > > > > diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c > > > > index db402ebc5570..cc570891322b 100644 > > > > --- a/kernel/bpf/syscall.c > > > > +++ b/kernel/bpf/syscall.c > > > > @@ -2165,7 +2165,6 @@ static bool is_net_admin_prog_type(enum bpf_prog_type prog_type) > > > > case BPF_PROG_TYPE_LWT_SEG6LOCAL: > > > > case BPF_PROG_TYPE_SK_SKB: > > > > case BPF_PROG_TYPE_SK_MSG: > > > > - case BPF_PROG_TYPE_LIRC_MODE2: > > > > case BPF_PROG_TYPE_FLOW_DISSECTOR: > > > > case BPF_PROG_TYPE_CGROUP_DEVICE: > > > > case BPF_PROG_TYPE_CGROUP_SOCK: > > > > @@ -2202,6 +2201,17 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) > > > > } > > > > } > > > > > > > > +static bool is_sys_admin_prog_type(enum bpf_prog_type prog_type) > > > > +{ > > > > + switch (prog_type) { > > > > + case BPF_PROG_TYPE_LIRC_MODE2: > > > > + case BPF_PROG_TYPE_EXT: /* extends any prog */ > > > > + return true; > > > > + default: > > > > + return false; > > > > + } > > > > +} > > > > > > I am not sure whether we should do this. This is a behavior change, that may > > > break some user space. Also, BPF_PROG_TYPE_EXT is checked in > > > is_perfmon_prog_type(), and this change will make that case useless. > > > > Sure, I can drop it from v3 and make this function appear for HID only. > > For BPF_PROG_TYPE_LIRC_MODE2, I don't think this change will break userspace. > This is called from ir-keytable(1) which is called from udev. It should have > all the necessary permissions. > > In addition, the vast majority IR decoders are non-bpf. bpf ir decoders have > very few users at the moment. > > I am working on completely new userspace tooling which will make extensive > use of bpf ir decoding with full lircd and IRP compatibility, but this is not > finished yet (see https://github.com/seanyoung/cir). Thanks for these information. I guess change for BPF_PROG_TYPE_LIRC_MODE2 is ok then. Would you mind ack or review this change (either current version or a later version)? Thanks, Song
On Fri, Mar 04, 2022 at 06:28:25PM +0100, Benjamin Tissoires wrote: > LIRC_MODE2 does not really need net_admin capability, but only sys_admin. > > Extract a new helper for it, it will be also used for the HID bpf > implementation. > > Cc: Sean Young <sean@mess.org> > Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> For BPF_PROG_TYPE_LIRC_MODE2, I don't think this change will break userspace. This is called from ir-keytable(1) which is called from udev. It should have all the necessary permissions. In addition, the vast majority IR decoders are non-bpf. bpf ir decoders have very few users at the moment. Acked-by: Sean Young <sean@mess.org> Sean > > --- > > new in v2 > --- > kernel/bpf/syscall.c | 14 +++++++++++++- > 1 file changed, 13 insertions(+), 1 deletion(-) > > diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c > index db402ebc5570..cc570891322b 100644 > --- a/kernel/bpf/syscall.c > +++ b/kernel/bpf/syscall.c > @@ -2165,7 +2165,6 @@ static bool is_net_admin_prog_type(enum bpf_prog_type prog_type) > case BPF_PROG_TYPE_LWT_SEG6LOCAL: > case BPF_PROG_TYPE_SK_SKB: > case BPF_PROG_TYPE_SK_MSG: > - case BPF_PROG_TYPE_LIRC_MODE2: > case BPF_PROG_TYPE_FLOW_DISSECTOR: > case BPF_PROG_TYPE_CGROUP_DEVICE: > case BPF_PROG_TYPE_CGROUP_SOCK: > @@ -2202,6 +2201,17 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) > } > } > > +static bool is_sys_admin_prog_type(enum bpf_prog_type prog_type) > +{ > + switch (prog_type) { > + case BPF_PROG_TYPE_LIRC_MODE2: > + case BPF_PROG_TYPE_EXT: /* extends any prog */ > + return true; > + default: > + return false; > + } > +} > + > /* last field in 'union bpf_attr' used by this command */ > #define BPF_PROG_LOAD_LAST_FIELD core_relo_rec_size > > @@ -2252,6 +2262,8 @@ static int bpf_prog_load(union bpf_attr *attr, bpfptr_t uattr) > return -EPERM; > if (is_perfmon_prog_type(type) && !perfmon_capable()) > return -EPERM; > + if (is_sys_admin_prog_type(type) && !capable(CAP_SYS_ADMIN)) > + return -EPERM; > > /* attach_prog_fd/attach_btf_obj_fd can specify fd of either bpf_prog > * or btf, we need to check which one it is > -- > 2.35.1
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index db402ebc5570..cc570891322b 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -2165,7 +2165,6 @@ static bool is_net_admin_prog_type(enum bpf_prog_type prog_type) case BPF_PROG_TYPE_LWT_SEG6LOCAL: case BPF_PROG_TYPE_SK_SKB: case BPF_PROG_TYPE_SK_MSG: - case BPF_PROG_TYPE_LIRC_MODE2: case BPF_PROG_TYPE_FLOW_DISSECTOR: case BPF_PROG_TYPE_CGROUP_DEVICE: case BPF_PROG_TYPE_CGROUP_SOCK: @@ -2202,6 +2201,17 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) } } +static bool is_sys_admin_prog_type(enum bpf_prog_type prog_type) +{ + switch (prog_type) { + case BPF_PROG_TYPE_LIRC_MODE2: + case BPF_PROG_TYPE_EXT: /* extends any prog */ + return true; + default: + return false; + } +} + /* last field in 'union bpf_attr' used by this command */ #define BPF_PROG_LOAD_LAST_FIELD core_relo_rec_size @@ -2252,6 +2262,8 @@ static int bpf_prog_load(union bpf_attr *attr, bpfptr_t uattr) return -EPERM; if (is_perfmon_prog_type(type) && !perfmon_capable()) return -EPERM; + if (is_sys_admin_prog_type(type) && !capable(CAP_SYS_ADMIN)) + return -EPERM; /* attach_prog_fd/attach_btf_obj_fd can specify fd of either bpf_prog * or btf, we need to check which one it is
LIRC_MODE2 does not really need net_admin capability, but only sys_admin. Extract a new helper for it, it will be also used for the HID bpf implementation. Cc: Sean Young <sean@mess.org> Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> --- new in v2 --- kernel/bpf/syscall.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-)