[meta-freescale] Issues with sync-boundary-changes.patch on imx6qsabrelite
Eric Nelson
eric.nelson at boundarydevices.com
Thu Feb 28 10:16:31 PST 2013
On 02/28/2013 08:34 AM, Andrei Gherzan wrote:
>
> On Thu, Feb 28, 2013 at 5:30 PM, Eric Nelson
> <eric.nelson at boundarydevices.com
> <mailto:eric.nelson at boundarydevices.com>> wrote:
>
> Hi Otavio and Andrei,
>
> On 02/28/2013 08:01 AM, Otavio Salvador wrote:
>
> On Thu, Feb 28, 2013 at 11:42 AM, Andrei Gherzan
> <andrei at gherzan.ro <mailto:andrei at gherzan.ro>> wrote:
>
> While using imx6, I faced 2 issues after updating the kernel
> to ver. / rel.
> 1.1.0:
>
> 1. If i boot the board with hdmi monitor connected the boot
> process hangs
> with:
>
> ---
> MIPI DSI driver module loaded
> mxc_sdc_fb mxc_sdc_fb.0: register mxc display driver hdmi
> mxc_hdmi mxc_hdmi: Detected HDMI controller 0x13:0xa:0xa0:0xc1
> fbcvt: 1920x1080 at 60: CVT Name - 2.073M9
> imx-ipuv3 imx-ipuv3.0: IPU DMFC DP HIGH RESOLUTION: 1(0,1),
> 5B(2~5), 5F(6,7)
> Console: switching to colour frame buffer device 240x67
> mxc_sdc_fb mxc_sdc_fb.1: register mxc display driver lcd
> mxc_sdc_fb mxc_sdc_fb.2: register mxc display driver ldb
> _regulator_get: get() with no identifier
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 0(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 1(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 2(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 3(VIC 1):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 4(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 5(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 6(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 7(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 8(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 9(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 10(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 11(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 12(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 13(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 14(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 15(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 16(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 17(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 18(VIC 0):
> mxc_hdmi mxc_hdmi: mxc_hdmi_edid_rebuild___modelist: Added
> mode 19(VIC 0):
> fbcvt: 1920x1080 at 60: CVT Name - 2.073M9
> mxc_sdc_fb mxc_sdc_fb.3: register mxc display driver ldb
>
> This all looks valid, though things shouldn't hang here.
> ---
>
> Additional infos:
>
> $ cat /sys/devices/platform/mxc_sdc___fb.0/graphics/fb0/modes
> V:640x480p-60
> U:640x480p-60
> Was this from a boot without the HDMI monitor connected? The mxc_hdmi
> driver should list all of the display resolutions supported by the
> monitor as reported by EDID.
>
>
> Nope. This is from a boot with HDMI monitor connected.
>
> Kernel command line: console=ttymxc1,115200
> console=ttymxc1,115200
> root=/dev/sda2 rootwait
> video=mxcfb0:dev=hdmi,__1920x1080M at 60,if=RGB24
>
When trying to reproduce things, I noted a
-- You have two 'console=ttymxc1,115200' clauses here
-- You appear to be using SATA (/dev/sda2). For SD card boot,
this should be /dev/mmcblk0p2 with a Yocto-standard image
-- There's some extraneous stuff (__) in your video= clause
I just booted with a slightly modified kernel command-line to try
and match yours:
Kernel command line: console=ttymxc1,115200
video=mxcfb0:dev=hdmi,1920x1080M at 60,if=RGB24 fbmem=48M vmalloc=400
root=/dev/mmcblk0p1 rootwait
What I found is that things booted, and I got a login prompt on the
serial console, but my display isn't functioning properly.
>
> You'll need to add a an 'fbmem=' clause to the kernel command line.
> For 1080P, I think this should be ~48M and you may need a 'vmalloc'
> clause:
> root=/dev/sda2 rootwait console=ttymxc1,115200
> video=video=mxcfb0:dev=hdmi,__1920x1080M at 60,if=RGB24 fbmem=48M
> vmalloc=400
>
> 2. The first client connected to X server won't show
> anything on screen. But
> absolutely anything -without seeing any issues /errors /
> warnings.
> Restarting application everything works OK. And until reboot
> no related
> issues. After a reboot, the same behavior.
>
> So I've been struggling to get these fixed and search for
> the source of the
> problem. I was pretty sure that was something related to
> kernel so I tried
> some bisects with no luck. After some digging a realized the
> issue was
> something inside the sync-boundary-changes.patch. Without
> this patch 1) and
> 2) are fixed.
>
> I will take a look as we know the source of the problem now.
> But maybe
> somebody who already knows this patch can work with me in
> parallel.
>
>
> Eric, is it a known issue?
>
>
> I'll try to reproduce when I get into the office.
>
>
> Thanks.
>
>
> --
> *Andrei Gherzan*
> m: +40.744.478.414 | f: +40.31.816.28.12
More information about the meta-freescale
mailing list