> -----Original Message----- > From: [email protected] [mailto:linux-omap- > [email protected]] On Behalf Of Grazvydas Ignotas > Sent: Monday, December 07, 2009 3:57 PM > To: Tomi Valkeinen > Cc: [email protected] > Subject: Re: DSS2 and power supplies > > On Mon, Dec 7, 2009 at 10:44 AM, Tomi Valkeinen > <[email protected]> wrote: > > Hi, > > > > On Fri, 2009-12-04 at 11:53 +0100, ext Grazvydas Ignotas wrote: > >> hi, > >> > >> it seems DSS block in OMAP needs SDI/DSI supplies to operate > properly > >> even with parallel displays, or else it doesn't drive some data > lines > >> resulting in missing/wrong colors. Take a look at figure 15-1 on > OMAP3 > >> TRM (page 2119 on spruf98d), some dss_data lines are going out of > >> SDI/DSI blocks, which are fed by vdds_sdi and vdds_dsi supplies > >> (vdss_sdi in fig 15-1 seems to be a typo). Those 2 are typically > >> connected to VPLL2 on TWL4030/TPS65950 (see fig. 25-2 on p3482), > so > >> that regulator needs to be activated for proper operation (if you > have > >> a parallel display and a board with TWL4030 you can reproduce > this by > >> not enabling VPLL2 in bootloader and kernel; mainline u-boot > enables > >> it). > > > > Hmm, this would explain the color errors. Have you seen this on > some > > other board than TI's SDP? > > > > To me the fig 15-1 does not tell that the powers are needed. The > > DSS_DATA[] pins do not come from SDI/DSI blocks, but from the main > DSS > > block. And the pin muxing component is outside also. This of > course > > doesn't really prove anything. =) > > True, but apparently vdds_dsi and/or vdds_sdi are also used to drive > some data lines then (not supplying them produces blue-ish display > with some parts missing). I'm working with pandora board, and I'm > sure > VPLL2 is not connected to anything else. > [Hiremath, Vaibhav] No, I think the "vdds_dsi" doesn't drive any data in parallel mode of operation. I have also observed similar issue with OMAP3EVM, when you switch the output to DVI interface, you will get bluish image.
The issue came out as VPLL2_DEV_GRP (Bits 7-5) register configuration. When you are working with LCD panel (in my case sharp-ls037v7dw01.c), this file calls regulator_get and regulator_enable functions which configures VPLL2_DEV_GRP to 0x2 [Owns to P1 device group]. But when you switch the output to DVI, it doesn't do any regulator_get/enable call so the VPLL2_DEV_GRP register becomes 0 [Owns to no device group] leading to color loss. I am not regulator framework expert, so could not comment on what is the relation of VPLL2_DEV_GRP with this issue. So what is required here is, panel_generic.c also should enable regulator for "vdvi" I have a patch (tested on OMAP3EVM) which I submit once DSS2 goes in. Thanks, Vaibhav > > > >> Right now DSS2 requests vdds_dsi in DSI driver only, any plans to > >> request vdds_dsi/vdds_sdi for parallel displays too? > > > > I didn't have any plans, but if what you said is true, it sure > needs to > > be added. > > Yeah that would be nice. > > > > > Tomi > > > > Ps. Adding l-o, as others may be interested also. > > > > > > > -- > To unsubscribe from this list: send the line "unsubscribe linux- > omap" in > the body of a message to [email protected] > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
