| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <ZUECsLFCzroIccgw@pluto>
Date: Tue, 31 Oct 2023 13:35:44 +0000
From: Cristian Marussi <cristian.marussi@....com>
To: Peng Fan <peng.fan@....com>
Cc: "Peng Fan (OSS)" <peng.fan@....nxp.com>,
"sudeep.holla@....com" <sudeep.holla@....com>,
"linux-arm-kernel@...ts.infradead.org"
<linux-arm-kernel@...ts.infradead.org>,
"linux-clk@...r.kernel.org" <linux-clk@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Ranjani Vaidyanathan <ranjani.vaidyanathan@....com>,
Glen G Wienecke <glen.wienecke@....com>,
"nitin.garg@....com" <nitin.garg@....com>,
Chuck Cannon <chuck.cannon@....com>,
"Souvik.Chakravarty@....com" <Souvik.Chakravarty@....com>
Subject: Re: [PATCH V1] firmware: arm_scmi: clock: support clock denied flags
On Tue, Oct 31, 2023 at 01:33:17PM +0000, Peng Fan wrote:
> > Subject: Re: [PATCH V1] firmware: arm_scmi: clock: support clock denied
> > flags
> >
> > On Tue, Oct 31, 2023 at 08:27:34PM +0800, Peng Fan (OSS) wrote:
> > > From: Peng Fan <peng.fan@....com>
> > >
> > > The firmware may export flags to indicate whether the clock is allowed
> > > to set rate, set parent, enable/disable from the Agent.
> > >
> > > If Agent is not allowed to enable/disable, directly return success.
> >
> > ... even though, this success on deny built-in policy could be
> > controversial..
> Yeah.
>
> .Have you tried the different approach of not registerig the
> > enable/prepare callbacks in clk-scmi instead ?
>
> Not yet, I could give a look. But not have much time
> in the following week.
>
>
> > From the Clk framework code it seems could be handled gracefully.
> > (I have not tried though...)
>
> If there is no enable/disable/prepare/unprepared hooks, the enable
> counter or prepare counter would be used.
>
Yes this is what I could understand.
Thanks,
Cristian
> Thanks,
> Peng.
> >
> > Thanks,
> > Cristian
> >
> > > If Agent is not allowed to set rate/parent, directly return -EACCES to
> > > avoid SCMI RPC calls.
> > >
> > > Signed-off-by: Peng Fan <peng.fan@....com>
> > > ---
> > >
> > > V1:
> > > drop the changes in clock.c, add an attribute entry in clock info
> > > which may be easy for extending new flag.
> > >
> > > SPEC still not have such support, this is for discussion
> > >
> > > drivers/firmware/arm_scmi/clock.c | 19 +++++++++++++++++++
> > > include/linux/scmi_protocol.h | 1 +
> > > 2 files changed, 20 insertions(+)
> > >
> > > diff --git a/drivers/firmware/arm_scmi/clock.c
> > > b/drivers/firmware/arm_scmi/clock.c
> > > index 42b81c181d68..fad4329a21fc 100644
> > > --- a/drivers/firmware/arm_scmi/clock.c
> > > +++ b/drivers/firmware/arm_scmi/clock.c
> > > @@ -46,6 +46,9 @@ struct scmi_msg_resp_clock_attributes {
> > > #define SUPPORTS_RATE_CHANGE_REQUESTED_NOTIF(x) ((x) &
> > BIT(30))
> > > #define SUPPORTS_EXTENDED_NAMES(x) ((x) & BIT(29))
> > > #define SUPPORTS_PARENT_CLOCK(x) ((x) & BIT(28))
> > > +#define SCMI_CLOCK_SET_ENABLE_DENIED BIT(15)
> > > +#define SCMI_CLOCK_SET_RATE_DENIED BIT(14)
> > > +#define SCMI_CLOCK_SET_PARENT_DENIED BIT(13)
> > > u8 name[SCMI_SHORT_NAME_MAX_SIZE];
> > > __le32 clock_enable_latency;
> > > };
> > > @@ -307,6 +310,7 @@ static int scmi_clock_attributes_get(const struct
> > scmi_protocol_handle *ph,
> > > if (PROTOCOL_REV_MAJOR(version) >= 0x2)
> > > latency = le32_to_cpu(attr->clock_enable_latency);
> > > clk->enable_latency = latency ? : U32_MAX;
> > > + clk->attributes = attributes;
> > > }
> > >
> > > ph->xops->xfer_put(ph, t);
> > > @@ -499,6 +503,10 @@ static int scmi_clock_rate_set(const struct
> > scmi_protocol_handle *ph,
> > > struct scmi_xfer *t;
> > > struct scmi_clock_set_rate *cfg;
> > > struct clock_info *ci = ph->get_priv(ph);
> > > + struct scmi_clock_info *clk = ci->clk + clk_id;
> > > +
> > > + if (clk->attributes & SCMI_CLOCK_SET_RATE_DENIED)
> > > + return -EACCES;
> > >
> > > ret = ph->xops->xfer_get_init(ph, CLOCK_RATE_SET, sizeof(*cfg), 0,
> > &t);
> > > if (ret)
> > > @@ -585,6 +593,9 @@ scmi_clock_set_parent(const struct
> > scmi_protocol_handle *ph, u32 clk_id,
> > > if (parent_id >= clk->num_parents)
> > > return -EINVAL;
> > >
> > > + if (clk->attributes & SCMI_CLOCK_SET_PARENT_DENIED)
> > > + return -EACCES;
> > > +
> > > ret = ph->xops->xfer_get_init(ph, CLOCK_PARENT_SET,
> > > sizeof(*cfg), 0, &t);
> > > if (ret)
> > > @@ -668,6 +679,10 @@ static int scmi_clock_enable(const struct
> > scmi_protocol_handle *ph, u32 clk_id,
> > > bool atomic)
> > > {
> > > struct clock_info *ci = ph->get_priv(ph);
> > > + struct scmi_clock_info *clk = ci->clk + clk_id;
> > > +
> > > + if (clk->attributes & SCMI_CLOCK_SET_ENABLE_DENIED)
> > > + return 0;
> > >
> > > return ci->clock_config_set(ph, clk_id, CLK_STATE_ENABLE,
> > > NULL_OEM_TYPE, 0, atomic);
> > > @@ -677,6 +692,10 @@ static int scmi_clock_disable(const struct
> > scmi_protocol_handle *ph, u32 clk_id,
> > > bool atomic)
> > > {
> > > struct clock_info *ci = ph->get_priv(ph);
> > > + struct scmi_clock_info *clk = ci->clk + clk_id;
> > > +
> > > + if (clk->attributes & SCMI_CLOCK_SET_ENABLE_DENIED)
> > > + return 0;
> > >
> > > return ci->clock_config_set(ph, clk_id, CLK_STATE_DISABLE,
> > > NULL_OEM_TYPE, 0, atomic);
> > > diff --git a/include/linux/scmi_protocol.h
> > > b/include/linux/scmi_protocol.h index f2f05fb42d28..ddf5363c8cfd
> > > 100644
> > > --- a/include/linux/scmi_protocol.h
> > > +++ b/include/linux/scmi_protocol.h
> > > @@ -58,6 +58,7 @@ struct scmi_clock_info {
> > > u64 step_size;
> > > } range;
> > > };
> > > + u32 attributes;
> > > int num_parents;
> > > u32 *parents;
> > > };
> > > --
> > > 2.37.1
> > >
Powered by blists - more mailing lists