Hi,
On Fri, Aug 03, 2012 at 12:45:53PM +0300, Aaro Koskinen wrote:
> Hi,
>
> On Fri, Aug 03, 2012 at 11:05:27AM +0300, Tomi Valkeinen wrote:
> > On Thu, 2012-08-02 at 17:54 -0600, Grant Likely wrote:
> > > On Thu, Aug 2, 2012 at 5:08 PM, Aaro Koskinen <aaro.koskinen@xxxxxx> wrote:
> > > > This reverts commit 9fdca9dfe093c76fe1ac1a09888ba9679d46996a.
> > > >
> > > > Changing omap2_mcspi_init() from subsys_initcall to device_initcall broke
> > > > the display initialization on N900 when all the drivers are compiled
> > > > built-in. Display subsystem drivers need a certain initialization order
> > > > and having all of them initialize with device_initcall seems to be too
> > > > fragile. Without this revert the display init fails and the boot hangs
> > > > with the following messages:
> > > >
> > > > [ 1.260955] acx565akm spi1.2: invalid display ID
> > > > [ 1.265899] panel-acx565akm display0: acx_panel_probe panel detect error
> > > > [ 1.273071] omapdss CORE error: driver probe failed: -19
> > >
> > > The dependencies are all messed up. The reverted commit is part of
> > > fixing that and I don't really want to go backwards on it. How many
> > > drivers are failing? Can you try modifying the failure path of those
> > > drivers to return -EPROBE_DEFER and see if that helps?
> >
> > The description of 9fdca9dfe093c76fe1ac1a09888ba9679d46996a seems to be
> > overly positive: "this will delete a few lines of code, no functional
> > changes".
> >
> > But yes, the probe dependencies with omap's display are quite fragile.
> > The probe order should be something like (from first to last):
> >
> > - omapdss, i2c, spi, etc.
>
> I looked more into this, and it seems this particular problem is related
> to probe order within SPI (and a special case with chip selects). The
> panel driver on N900 is spi1.2. It's listed in the board file before
> spi1.0 (touch controller). When the panel driver is probed, SPI transfers
> appear to work (no errors) but all returned data zero:
>
> [ 1.371704] omap2_mcspi omap2_mcspi.1: registered master spi1
> [ 1.378143] spi spi1.2: setup: speed 6000000, sample leading edge, clk normal
> [ 1.385864] spi spi1.2: setup mode 0, 8 bits/w, 6000000 Hz max --> 0
> [ 1.393005] acx565akm spi1.2: acx565akm_spi_probe
> [ 1.398345] panel-acx565akm display0: acx_panel_probe
> [ 1.416931] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.425415] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.433868] acx565akm spi1.2: LCD panel not enabled by bootloader (status 0x0000)
> [ 1.441894] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.450347] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.458801] acx565akm spi1.2: MIPI display ID: 000000
> [ 1.464202] acx565akm spi1.2: invalid display ID
> [ 1.469085] panel-acx565akm display0: acx_panel_probe panel detect error
> [ 1.476196] omapdss CORE error: driver probe failed: -19
> [ 1.482269] omap2_mcspi omap2_mcspi.1: registered child spi1.2
> [ 1.488525] spi spi1.0: setup: speed 6000000, sample leading edge, clk normal
> [ 1.496124] spi spi1.0: setup mode 0, 8 bits/w, 6000000 Hz max --> 0
> [ 1.503265] omap2_mcspi omap2_mcspi.1: registered child spi1.0
>
> If I modify the board file so that child spi1.0 is registered before
> spi1.2, then it starts to work properly:
that sounds quite odd... Shubhro, have you seen this before on any of
our platforms ?
> [ 1.371978] omap2_mcspi omap2_mcspi.1: registered master spi1
> [ 1.378417] spi spi1.0: setup: speed 6000000, sample leading edge, clk normal
> [ 1.386108] spi spi1.0: setup mode 0, 8 bits/w, 6000000 Hz max --> 0
> [ 1.393249] omap2_mcspi omap2_mcspi.1: registered child spi1.0
> [ 1.399505] spi spi1.2: setup: speed 6000000, sample leading edge, clk normal
> [ 1.407104] spi spi1.2: setup mode 0, 8 bits/w, 6000000 Hz max --> 0
> [ 1.414245] acx565akm spi1.2: acx565akm_spi_probe
> [ 1.419586] panel-acx565akm display0: acx_panel_probe
> [ 1.440429] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.448883] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.457305] acx565akm spi1.2: LCD panel enabled by bootloader (status 0x80730417)
> [ 1.465301] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.473724] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.482147] acx565akm spi1.2: MIPI display ID: 108a77
> [ 1.487548] acx565akm spi1.2: omapfb: acx565akm rev 8a LCD detected
> [ 1.494689] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.503112] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.511566] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.520019] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.528442] acx565akm spi1.2: acx565akm_bl_update_status
> [ 1.534149] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.542572] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.550964] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.559417] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.567840] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.576293] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.584716] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.593109] acx565akm spi1.2: setup: speed 6000000, sample trailing edge, clk inverted
> [ 1.601867] omap2_mcspi omap2_mcspi.1: registered child spi1.2
>
> So an altenative hack to fix this is to modify the board file
> (board-rx51-peripherals.c):
>
> /* list all spi devices here */
> enum {
> RX51_SPI_WL1251,
> - RX51_SPI_MIPID, /* LCD panel */
> RX51_SPI_TSC2005, /* Touch Controller */
> + RX51_SPI_MIPID, /* LCD panel */
> };
>
> I guess the proper fix would be to modify SPI core so that it first does
> spi_setup for all the children/chip selects, before calling any of the
> probe functions of SPI devices? (Initializing the controller driver at
> subsys_initcall is one way to accomplish this.)
--
balbi
Attachment:
signature.asc
Description: Digital signature
