linux-kernel - Re: [PATCH] video: fbdev: mxsfb: fix pixelclock polarity

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

Message-ID: <ccddc4dd5793a718636af4e2e4f009a4@agner.ch>
Date:   Tue, 17 Oct 2017 18:01:58 +0200
From:   Stefan Agner <stefan@...er.ch>
To:     Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>
Cc:     shawnguo@...nel.org, plagnioj@...osoft.com, s.hauer@...gutronix.de,
        Ying.Liu@...escale.com, linux-fbdev@...r.kernel.org,
        linux-kernel@...r.kernel.org, LW@...o-electronics.de,
        fabio.estevam@...escale.com, max.krummenacher@...adex.com,
        mauro.salvini@...group.net
Subject: Re: [PATCH] video: fbdev: mxsfb: fix pixelclock polarity

On 2017-10-17 15:56, Bartlomiej Zolnierkiewicz wrote:
> Hi,
> 
> Stefan/Shawn:
> 
> Should I merge this patch? It seems to be still needed:
> 
> https://www.spinics.net/lists/linux-fbdev/msg23112.html

In my opinion yes, it fixes an issue... 

FWIW, I normally use the DRM driver when using mainline which does not
suffer from this issue, so I did not mind much that this fix never made
it...

--
Stefan

> 
> Best regards,
> --
> Bartlomiej Zolnierkiewicz
> Samsung R&D Institute Poland
> Samsung Electronics
> 
> On Friday, May 06, 2016 01:29:59 PM Stefan Agner wrote:
>> Hi Shawn,
>>
>> On 2016-01-25 17:29, Stefan Agner wrote:
>> > The PIXDATA flags of the display_flags enum are controller centric,
>> > e.g. NEGEDGE means the controller shall drive the data signals on
>> > pixelclocks negative edge. However, the drivers flag is display
>> > centric: Sample the data on negative (falling) edge.
>> >
>> > Therefore, change the if statement to check for the POSEDGE flag
>> > (which is typically not set):
>> > Drive on positive edge => sample on negative edge
>>
>> Any comment on that patch?
>>
>> --
>> Stefan
>>
>>
>> >
>> > Signed-off-by: Stefan Agner <stefan@...er.ch>
>> > ---
>> > Hi all
>> >
>> > Shawn, I would like to have at least your Ack on this before merge.
>> >
>> > It seems that this has been wrong since the driver is able to use
>> > the timings from the device tree, introduced with 669406534b4a
>> > ("video: mxsfb: get display timings from device tree").
>> >
>> > Not sure how many device trees actually specify the wrong pixel clock
>> > polarity due to that. At least the initial flag convertion from the
>> > old platform data structures done with 0d9f8217db15 ("ARM: mxs: move
>> > display timing configurations into device tree") seems to be affected
>> > and would need to be changed accordingly...
>> >
>> > Not sure how we should handle this, maybe just invert all
>> > pixelclk-active properties where the mxsfb driver is in use...?
>> >
>> > --
>> > Stefan
>> >
>> >  drivers/video/fbdev/mxsfb.c | 13 +++++++++++--
>> >  1 file changed, 11 insertions(+), 2 deletions(-)
>> >
>> > diff --git a/drivers/video/fbdev/mxsfb.c b/drivers/video/fbdev/mxsfb.c
>> > index 4e6608c..38898a9 100644
>> > --- a/drivers/video/fbdev/mxsfb.c
>> > +++ b/drivers/video/fbdev/mxsfb.c
>> > @@ -150,7 +150,7 @@
>> >  #define STMLCDIF_24BIT 3 /** pixel data bus to the display is of 24
>> > bit width */
>> >
>> >  #define MXSFB_SYNC_DATA_ENABLE_HIGH_ACT	(1 << 6)
>> > -#define MXSFB_SYNC_DOTCLK_FALLING_ACT	(1 << 7) /* negtive edge sampling */
>> > +#define MXSFB_SYNC_DOTCLK_FALLING_ACT	(1 << 7) /* negative edge sampling */
>> >
>> >  enum mxsfb_devtype {
>> >  	MXSFB_V3,
>> > @@ -788,7 +788,16 @@ static int mxsfb_init_fbinfo_dt(struct mxsfb_info *host,
>> >
>> >  	if (vm.flags & DISPLAY_FLAGS_DE_HIGH)
>> >  		host->sync |= MXSFB_SYNC_DATA_ENABLE_HIGH_ACT;
>> > -	if (vm.flags & DISPLAY_FLAGS_PIXDATA_NEGEDGE)
>> > +
>> > +	/*
>> > +	 * The PIXDATA flags of the display_flags enum are controller
>> > +	 * centric, e.g. NEGEDGE means drive data on negative edge.
>> > +	 * However, the drivers flag is display centric: Sample the
>> > +	 * data on negative (falling) edge. Therefore, check for the
>> > +	 * POSEDGE flag:
>> > +	 * drive on positive edge => sample on negative edge
>> > +	 */
>> > +	if (vm.flags & DISPLAY_FLAGS_PIXDATA_POSEDGE)
>> >  		host->sync |= MXSFB_SYNC_DOTCLK_FALLING_ACT;
>> >
>> >  put_display_node: