| 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: <Zp50gJvEN0NOsPpm@lizhi-Precision-Tower-5810>
Date: Mon, 22 Jul 2024 11:02:24 -0400
From: Frank Li <Frank.li@....com>
To: Carlos Song <carlos.song@....com>
Cc: "miquel.raynal@...tlin.com" <miquel.raynal@...tlin.com>,
"conor.culhane@...vaco.com" <conor.culhane@...vaco.com>,
"alexandre.belloni@...tlin.com" <alexandre.belloni@...tlin.com>,
"linux-i3c@...ts.infradead.org" <linux-i3c@...ts.infradead.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"imx@...ts.linux.dev" <imx@...ts.linux.dev>,
dl-linux-imx <linux-imx@....com>
Subject: Re: [PATCH v2] i3c: master: svc: adjust the first broadcast speed
On Sat, Jul 20, 2024 at 04:42:18AM +0000, Carlos Song wrote:
>
>
> > -----Original Message-----
> > From: Frank Li <frank.li@....com>
> > Sent: Saturday, July 20, 2024 2:26 AM
> > To: Carlos Song <carlos.song@....com>
> > Cc: miquel.raynal@...tlin.com; conor.culhane@...vaco.com;
> > alexandre.belloni@...tlin.com; linux-i3c@...ts.infradead.org;
> > linux-kernel@...r.kernel.org; imx@...ts.linux.dev; dl-linux-imx
> > <linux-imx@....com>
> > Subject: Re: [PATCH v2] i3c: master: svc: adjust the first broadcast speed
> >
> > On Fri, Jul 19, 2024 at 04:37:01PM +0000, Carlos Song wrote:
> > >
> > >
> > > > -----Original Message-----
> > > > From: Frank Li <frank.li@....com>
> > > > Sent: Friday, July 19, 2024 11:01 PM
> > > > To: Carlos Song <carlos.song@....com>
> > > > Cc: miquel.raynal@...tlin.com; conor.culhane@...vaco.com;
> > > > alexandre.belloni@...tlin.com; linux-i3c@...ts.infradead.org;
> > > > linux-kernel@...r.kernel.org; imx@...ts.linux.dev; dl-linux-imx
> > > > <linux-imx@....com>
> > > > Subject: Re: [PATCH v2] i3c: master: svc: adjust the first broadcast
> > > > speed
> > > >
> > > > On Fri, Jul 19, 2024 at 04:03:51PM +0800, carlos.song@....com wrote:
> > > > > From: Carlos Song <carlos.song@....com>
> > > > >
> > > > > According to the i3c spec 6.2 Timing Specification, the first
> > > > > broadcast open drain timing should be adjust to High Period of SCL
> > > > > Clock is 200ns at least. I3C device working as a i2c device will
> > > > > see the broadcast to close its Spike Filter to change to i3c mode.
> > > > > After that I3C open drain SCL high level should be adjust to 32ns~45ns.
> > > > >
> > > > > Signed-off-by: Carlos Song <carlos.song@....com>
> > > > > ---
> > > > > Change for V2:
> > > > > - use slow_speed instead of first_broadcast
> > > > > - add default_speed variable in svc_i3c_xfer to avoid set default
> > > > > speed every time
> > > > > - change start_xfer if else for easy understand
> > > > > ---
> > > > > drivers/i3c/master/svc-i3c-master.c | 55
> > > > > +++++++++++++++++++++++++++++
> > > > > 1 file changed, 55 insertions(+)
> > > > >
> > > > > diff --git a/drivers/i3c/master/svc-i3c-master.c
> > > > > b/drivers/i3c/master/svc-i3c-master.c
> > > > > index 78116530f431..7cd3a9a4d7dd 100644
> > > > > --- a/drivers/i3c/master/svc-i3c-master.c
> > > > > +++ b/drivers/i3c/master/svc-i3c-master.c
> > > > > @@ -142,6 +142,7 @@ struct svc_i3c_cmd {
> > > > > unsigned int actual_len;
> > > > > struct i3c_priv_xfer *xfer;
> > > > > bool continued;
> > > > > + bool slow_address;
> > > > > };
> > > > >
> > > > > struct svc_i3c_xfer {
> > > > > @@ -214,6 +215,11 @@ struct svc_i3c_master {
> > > > > } ibi;
> > > > > struct mutex lock;
> > > > > int enabled_events;
> > > > > +
> > > > > + unsigned long fclk_rate;
> > > > > + u32 mctrl_config;
> > > > > + bool slow_speed;
> > > > > + bool default_speed;
> > > >
> > > > I think you needn't two varible 'slow_speed' and 'default_speed'.
> > > > default_speed should always !slow_speed
> > > >
> > > > Frank
> > > >
> > >
> > > Hi, Frank
> > >
> > > In fact, I am struggling for using just one variable: slow speed. Adding a
> > variable "default_speed "was also a have-to move.
> > >
> > > If I use "if else" in xfer for easy understand, it means I only can change the
> > MCTRL register value one time in every xfer. So in the first xfer, I must change
> > slow_speed to false, then next xfer cmd->slow_address will be false. So in next
> > xfer, I can set initial configuration back to the controller.
> > > But the question is I have to set it every time, so I add the extra variable to
> > avoid writel in every xfer.
> > > It looks bad... Sorry for my poor coding.
> > >
> > > If only one variable " slow_speed " is used , I think "if else" and "writel one
> > time", only one can be used. Maybe there is a better code method but I don't
> > get it?
> >
> > If I understand correct.
> >
> > svc_i3c_master_set_slow_address_speed()
> > {
> > ...
> > master->slow_speed = true;
> > }
> >
> > svc_i3c_master_set_default_speed()
> > {
> > if (master->slow_speed) {
> > writel();
> > master->slow_speed false;
> > }
> > }
> >
> > if (cmd->slow_address)
> > svc_i3c_master_set_slow_address_speed(master);
> > else
> > svc_i3c_master_set_default_speed(master);
> >
>
> Above logic will never enter the svc_i3c_master_set_default_speed(master);
> Because this :
> svc_i3c_master_send_bdcast_ccc_cmd(){
> if (master->slow_speed)
I think, it should be master->first_cmd, which reflect the real means.
Then it means use two variables.
If it is i3c standard, I prefer it should do by framework to avoid every
driver add similar logic.
May Alexandre Belloni can provide comemnts about this.
static int i3c_master_bus_init(struct i3c_master_controller *master)
{
...
master->ops->set_speed(low);
ret = i3c_master_rstdaa_locked(master, I3C_BROADCAST_ADDR);
master->ops->set_speed(normal);
...
}
> cmd->slow_address = true;
> else
> cmd->slow_address = false;
> }
>
> Then svc_i3c_master_start_xfer_locked(){
> if (cmd->slow_address)
> svc_i3c_master_set_slow_address_speed(master);
> else
> svc_i3c_master_set_default_speed(master);
> }
>
> In svc_i3c_master_send_bdcast_ccc_cmd, we always need slow_speed to dicide cmd->slow_address is true or false. Then in svc_i3c_master_start_xfer_locked,
> We use cmd->slow_address to chose set slow_speed or default speed.
> when use "if else", it means we put slow_speed true->false into next xfer.
>
> When first xfer finish, slow_speed is still true. So in next svc_i3c_master_send_bdcast_ccc_cmd:
> then cmd->slow_address=ture, so in next xfer it will continue enter the "if" branch not the "else" branch:
>
> if (cmd->slow_address)
> svc_i3c_master_set_slow_address_speed(master);
> else
> svc_i3c_master_set_default_speed(master);
>
> so whatever we need change master->slow_speed = false in the first xfer, don't put it into next xfer.
> This is the reason I always do taht in my V1 V2 patch.
>
> Carlos
>
> > When slow_address is ture, always set slow_speed = ture when slow_address is
> > false, call to set_sefault_speed, if previous
> > slow_speed is true, then change to default speed, slow_speed will be false.
> > when next time call to set_default_speed() do nothing.
> >
> > Frank
> > >
> > > Carlos
> > > > > };
> > > > >
> > > > > /**
> > > > > @@ -531,6 +537,43 @@ static irqreturn_t
> > > > > svc_i3c_master_irq_handler(int
> > > > irq, void *dev_id)
> > > > > return IRQ_HANDLED;
> > > > > }
> > > > >
> > > > > +static void svc_i3c_master_set_slow_address_speed(struct
> > > > > +svc_i3c_master *master) {
> > > > > + struct i3c_bus *bus = i3c_master_get_bus(&master->base);
> > > > > + u32 ppbaud, odbaud, odhpp, mconfig;
> > > > > +
> > > > > + master->mctrl_config = readl(master->regs + SVC_I3C_MCONFIG);
> > > > > + mconfig = master->mctrl_config;
> > > > > +
> > > > > + /*
> > > > > + * Set the I3C OPEN-DRAIN mode to the FM speed of 50%
> > > > duty-cycle(400K/2500ns),
> > > > > + * so that the first broadcast is visible to all devices on the i3c bus.
> > > > > + * I3C device with 50ns filter will turn off the filter.
> > > > > + */
> > > > > +
> > > > > + ppbaud = FIELD_GET(GENMASK(11, 8), mconfig);
> > > > > + odhpp = 0;
> > > > > + odbaud = DIV_ROUND_UP(master->fclk_rate, bus->scl_rate.i2c * (2
> > > > > ++ 2 *
> > > > ppbaud)) - 1;
> > > > > + mconfig &= ~GENMASK(24, 16);
> > > > > + mconfig |= SVC_I3C_MCONFIG_ODBAUD(odbaud) |
> > > > > +SVC_I3C_MCONFIG_ODHPP(odhpp);
> > > > > +
> > > > > + writel(mconfig, master->regs + SVC_I3C_MCONFIG);
> > > > > + master->slow_speed = false;
> > > > > +}
> > > > > +
> > > > > +static void svc_i3c_master_set_default_speed(struct
> > > > > +svc_i3c_master
> > > > > +*master) {
> > > > > + /*
> > > > > + * The bus mode is already determined when the bus is
> > > > > +initialized, so
> > > > setting initial
> > > > > + * configuration back to the controller. No need to set it in
> > > > > +every transfer,
> > > > just
> > > > > + * restore it once time.
> > > > > + */
> > > > > + if (!master->default_speed) {
> > > > > + writel(master->mctrl_config, master->regs +
> > SVC_I3C_MCONFIG);
> > > > > + master->default_speed = true;
> > > > > + }
> > > > > +}
> > > > > +
> > > > > static int svc_i3c_master_bus_init(struct i3c_master_controller
> > > > > *m) {
> > > > > struct svc_i3c_master *master = to_svc_i3c_master(m); @@ -624,6
> > > > > +667,8 @@ static int svc_i3c_master_bus_init(struct
> > > > > +i3c_master_controller
> > > > *m)
> > > > > SVC_I3C_MCONFIG_I2CBAUD(i2cbaud);
> > > > > writel(reg, master->regs + SVC_I3C_MCONFIG);
> > > > >
> > > > > + master->slow_speed = true;
> > > > > + master->fclk_rate = fclk_rate;
> > > > > /* Master core's registration */
> > > > > ret = i3c_master_get_free_addr(m, 0);
> > > > > if (ret < 0)
> > > > > @@ -1251,6 +1296,11 @@ static void
> > > > svc_i3c_master_start_xfer_locked(struct svc_i3c_master *master)
> > > > > for (i = 0; i < xfer->ncmds; i++) {
> > > > > struct svc_i3c_cmd *cmd = &xfer->cmds[i];
> > > > >
> > > > > + if (cmd->slow_address)
> > > > > + svc_i3c_master_set_slow_address_speed(master);
> > > > > + else
> > > > > + svc_i3c_master_set_default_speed(master);
> > > > > +
> > > > > ret = svc_i3c_master_xfer(master, cmd->rnw, xfer->type,
> > > > > cmd->addr, cmd->in, cmd->out,
> > > > > cmd->len, &cmd->actual_len, @@ -1346,6
> > +1396,11 @@ static
> > > > > int
> > > > svc_i3c_master_send_bdcast_ccc_cmd(struct svc_i3c_master *master,
> > > > > cmd->actual_len = 0;
> > > > > cmd->continued = false;
> > > > >
> > > > > + if (master->slow_speed)
> > > > > + cmd->slow_address = true;
> > > > > + else
> > > > > + cmd->slow_address = false;
> > > > > +
> > > > > mutex_lock(&master->lock);
> > > > > svc_i3c_master_enqueue_xfer(master, xfer);
> > > > > if (!wait_for_completion_timeout(&xfer->comp,
> > > > > msecs_to_jiffies(1000)))
> > > > > --
> > > > > 2.34.1
> > > > >
Powered by blists - more mailing lists