linux-kernel - Re: [PATCH v2 1/2] extcon: Add extcon_set_cable_line_state() to inform the additional state of external connectors

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:	Thu, 28 May 2015 12:37:25 +0300
From:	"Ivan T. Ivanov" <iivanov@...sol.com>
To:	cw00.choi@...sung.com
Cc:	Roger Quadros <rogerq@...com>,
	linux-kernel <linux-kernel@...r.kernel.org>,
	Robert Baldyga <r.baldyga@...sung.com>,
	Peter Chen <peter.chen@...escale.com>,
	Kishon Vijay Abraham I <kishon@...com>,
	Felipe Balbi <balbi@...com>,
	"myungjoo.ham@...sung.com" <myungjoo.ham@...sung.com>
Subject: Re: [PATCH v2 1/2] extcon: Add extcon_set_cable_line_state() to
 inform the additional state of external connectors

Hi, 

On Thu, 2015-05-28 at 18:17 +0900, Chanwoo Choi wrote:
> 
> 
> 2015년 5월 28일 목요일, Ivan T. Ivanov<iivanov@...sol.com>님이 작성한 메시지:
> > Hi Chanwoo,
> > 
> > On Thu, 2015-05-28 at 00:06 +0900, Chanwoo Choi wrote:
> > > On Wed, May 27, 2015 at 11:38 PM, Roger Quadros <rogerq@...com> wrote:
> > > > Chanwoo,
> > > >
> > > > On 27/05/15 15:15, Chanwoo Choi wrote:
> > > > > This patch adds the extcon_set_cable_line_state() function to inform
> > > > > the additional state of each external connector and 'enum
> > > > > extcon_line_state'
> > > > > enumeration which include the specific states of each external connector.
> > > > >
> > > > > The each external connector might need the different line state. So,
> > > > > current
> > > > > 'extcon_line_state' enumeration contains the specific state for USB as
> > > > > following:
> > > > >
> > > > > - Following the state mean the state of both ID and VBUS line for USB:
> > > > > enum extcon_line_state {
> > > > >         EXTCON_USB_ID_LOW       = BIT(1),       /* ID line is low. */
> > > > >         EXTCON_USB_ID_HIGH      = BIT(2),       /* ID line is high. */
> > > >
> > > >
> > > > ID_LOW and ID_HIGH cannot happen simultaneously so they can use the same bit
> > > > position.
> > > > if the bit is set means it is high, if it is cleared means it is low.
> > >
> > > These are the notifier event. So, extcon_line_state have to include
> > > each event for both low or high state of ID.
> > > because the extcon consumer driver using the
> > > extcon_register_notifier() need the each event to distinguish the type
> > > of event.
> > >
> > > > >         EXTCON_USB_VBUS_LOW     = BIT(3),       /* VBUS line is low. */
> > > > >         EXTCON_USB_VBUS_HIGH    = BIT(4),       /* VBUS line is high. */
> > > >
> > > >
> > > > same here.
> > >
> > > ditto.
> > >
> > > > enum doesn't look like the right type for extcon_line_state as it is more of
> > > > a bitmask.
> > >
> > > Why? I prefer to use the enum if there are no problem.
> > 
> If you insist your opinion, you shoud suggest the reason. Could you tell me the reason?

It looks unreadable. There is no need to mix interface API (enum extcon_line_state)
with implementation details. You can use plain enum at interface level and hide
shifting inside implementation.

> 
> 
> > > > >
> > > > > +enum extcon_line_state {
> > > > > +       /* Following two definition are used for whether external
> > > > > connectors
> > > > > +        * is attached or detached. */
> > > > > +       EXTCON_DETACHED         = 0x0,
> > > > > +       EXTCON_ATTACHED         = 0x1,
> > > > > +
> > > > > +       /* Following states are only used for EXTCON_USB. */
> > > > > +       EXTCON_USB_ID_LOW       = BIT(1),       /* ID line is low. */
> > > > > +       EXTCON_USB_ID_HIGH      = BIT(2),       /* ID line is high. */
> > > > > +       EXTCON_USB_VBUS_LOW     = BIT(3),       /* VBUS line is low. */
> > > > > +       EXTCON_USB_VBUS_HIGH    = BIT(4),       /* VBUS line is high. */
> > > > > +};
> > > >
> > > >
> > > > Why do you use enum? How about the following bit definitions for line state.
> > > >
> > > > #define EXTCON_ATTACHED_DETACHED_BIT    BIT(0)
> > > > #define EXTCON_USB_ID_BIT               BIT(1)
> > > > #define EXTCON_USB_VBUS_BIT             BIT(2)
> > > > ...
> > > >
> > > > code must check if appropriate bit is set or cleared to get the high/low
> > > > state.
> > >
> > > I answer about it on upper.
> > >
> > 
> > Funny. We can use the one bit for DETACHED/ATTACHED, but we
> hmm. Did you read my answer about low/high?
> 
> But, I'm considering again about using the detached/attached value.

Ok, but consumers will receive one event at time, right?

Regards,
Ivan

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/