DM365/mt9p031 Media framework question

Jeremy Larsen

Other Parts Discussed in Thread: ASH, TVP7002

I am using a DM365 on a custom board. I am using a 2.6.37 kernel from arago.org. I am trying to convert an older application to use the new media framework. I am getting 0 entities enumerated and I suspect it is my filesystem (maybe missing /dev nodes). Where is the best place I can find and example of a root file system?

Here is the bootup and some ouput from a sample app called mt9p031_ccdc_prv_rsz_file:

U-Boot 2009.08 (May 15 2012 - 10:49:54)

I2C:   ready
DRAM: 128 MB
Using default environment

In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot: 0
## Booting kernel from Legacy Image at 80700000 ...
   Image Name:   Linux-2.6.37-ntouch_rcu
   Created:      2012-05-29 22:58:07 UTC
   Image Type:   ARM Linux Kernel Image (uncompressed)
   Data Size:    8077424 Bytes = 7.7 MB
   Load Address: 80008000
   Entry Point: 80008000
   Verifying Checksum ... OK
   Loading Kernel Image ... OK
OK

Starting kernel ...

Uncompressing Linux... done, booting the kernel.
Linux version 2.6.37-ntouch_rcu (jlarsen@salmon) (gcc version 4.5.1 (Sourcery G++ Lite 2010.09-50) ) #38 PREEMPT Tue May 29 16:58:01 MDT 2012
CPU: ARM926EJ-S [41069265] revision 5 (ARMv5TEJ), cr=00053177
CPU: VIVT data cache, VIVT instruction cache
Machine: DaVinci DM365 NTOUCH RCU
Memory policy: ECC disabled, Data cache writeback
DaVinci dm36x_rev1.2 variant 0x8
Built 1 zonelists in Zone order, mobility grouping on. Total pages: 16256
Kernel command line: console=ttyS0,115200n8 rw mem=64M vpfe-capture.interface=1 vpfe-capture.bufsize=4147200
PID hash table entries: 256 (order: -2, 1024 bytes)
Dentry cache hash table entries: 8192 (order: 3, 32768 bytes)
Inode-cache hash table entries: 4096 (order: 2, 16384 bytes)
Memory: 64MB = 64MB total
Memory: 44928k/44928k available, 20608k reserved, 0K highmem
Virtual kernel memory layout:
    vector : 0xffff0000 - 0xffff1000   (   4 kB)
    fixmap : 0xfff00000 - 0xfffe0000   ( 896 kB)
    DMA     : 0xff000000 - 0xffe00000   ( 14 MB)
    vmalloc : 0xc4800000 - 0xfea00000   ( 930 MB)
    lowmem : 0xc0000000 - 0xc4000000   ( 64 MB)
    modules : 0xbf000000 - 0xc0000000   ( 16 MB)
      .init : 0xc0008000 - 0xc0025000   ( 116 kB)
      .text : 0xc0025000 - 0xc036e930   (3367 kB)
      .data : 0xc0370000 - 0xc039c0a0   ( 177 kB)
SLUB: Genslabs=13, HWalign=32, Order=0-3, MinObjects=0, CPUs=1, Nodes=1
Preemptable hierarchical RCU implementation.
        RCU-based detection of stalled CPUs is disabled.
        Verbose stalled-CPUs detection is disabled.
NR_IRQS:245
Console: colour dummy device 80x30
Calibrating delay loop... 149.50 BogoMIPS (lpj=747520)
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 512
CPU: Testing write buffer coherency: ok
DaVinci: 8 gpio irqs
NET: Registered protocol family 16
bio: create slab <bio-0> at 0
SCSI subsystem initialized
vpss vpss: dm365_vpss vpss probed
vpss vpss: dm365_vpss vpss probe success
dm365_afew_hw_init
DM365 IPIPE initialized in Continuous mode
Advanced Linux Sound Architecture Driver Version 1.0.23.
Switching to clocksource timer0_1
musb-hdrc: version 6.0, peripheral, debug=0
NET: Registered protocol family 2
IP route cache hash table entries: 1024 (order: 0, 4096 bytes)
TCP established hash table entries: 2048 (order: 2, 16384 bytes)
TCP bind hash table entries: 2048 (order: 1, 8192 bytes)
TCP: Hash tables configured (established 2048 bind 2048)
TCP reno registered
UDP hash table entries: 256 (order: 0, 4096 bytes)
UDP-Lite hash table entries: 256 (order: 0, 4096 bytes)
NET: Registered protocol family 1
RPC: Registered udp transport module.
RPC: Registered tcp transport module.
RPC: Registered tcp NFSv4.1 backchannel transport module.
msgmni has been set to 87
io scheduler noop registered (default)
Serial: 8250/16550 driver, 2 ports, IRQ sharing disabled
serial8250.0: ttyS0 at MMIO 0x1c20000 (irq = 40) is a 16550A
console [ttyS0] enabled
serial8250.0: ttyS1 at MMIO 0x1d06000 (irq = 41) is a 16550A
brd: module loaded
davinci_mdio davinci_mdio.0: davinci mdio revision 1.4
davinci_mdio davinci_mdio.0: detected phy mask fffffffc
davinci_mdio.0: probed
davinci_mdio davinci_mdio.0: phy[0]: device 0:00, driver unknown
davinci_mdio davinci_mdio.0: phy[1]: device 0:01, driver unknown
console [netcon0] enabled
netconsole: network logging started
i2c /dev entries driver
Linux media interface: v0.10
lirc_dev: IR Remote Control driver registered, major 253
IR NEC protocol handler initialized
IR RC5(x) protocol handler initialized
IR RC6 protocol handler initialized
IR JVC protocol handler initialized
IR Sony protocol handler initialized
IR RC5 (streamzap) protocol handler initialized
IR LIRC bridge handler initialized
Linux video capture interface: v2.00
vpfe_init
vpfe-capture: vpss clock vpss_master enabled
dm365_isif is registered with vpfe.
imp serializer initialized
AEW Driver initialized
AF Driver initialized
vpfe-capture vpfe-capture: v4l2 device registered
mt9p031 1-0048: Detected a MT9P031 chip ID 1801
mt9p031: mt9p031 sensor driver registered !!
vpfe-capture vpfe-capture: v4l2 sub device mt9p031 registered
DM365 IPIPEIF probed
ALSA device list:
No soundcards found.
TCP cubic registered
NET: Registered protocol family 17
davinci_emac_probe: using random MAC addr: 62:21:d2:e3:67:cc
Freeing init memory: 116K
Enabling hot-plug          : [SUCCESS]
[SUCCESS]
mach: module license 'Proprietary' taints kernel.
Disabling lock debugging due to kernel taint
Built May 15 2012 at 10:52:34
*** ntouch_rcu_release.sh script ***
CMEMK module: built on May 15 2012 at 10:52:14
Reference Linux version 2.6.37
File /home/jlarsen/bsp_rcu_new_kernel/davinci1/open/linuxutils_2_26_03_06/packages/ti/sdo/linuxutils/cmem/src/module/cmemk.c
allocated heap buffer 0xc5000000 of size 0x30f8000
CMEM Range Overlaps Kernel Physical - allowing overlap
CMEM phys_start (0x1000) overlaps kernel (0x80000000 -> 0x84000000)
cmemk initialized
IRQK module: built on May 15 2012 at 10:52:15
Reference Linux version 2.6.37
File /home/jlarsen/bsp_rcu_new_kernel/davinci1/open/linuxutils_2_26_03_06/packages/ti/sdo/linuxutils/irq/src/module/irqk.c
irqk initialized
EDMAK module: built on May 15 2012 at 10:52:14
Reference Linux version 2.6.37
File /home/jlarsen/bsp_rcu_new_kernel/davinci1/open/linuxutils_2_26_03_06/packages/ti/sdo/linuxutils/edma/src/module/edmak.c
cat9532: Using cat9532 16-bit LED driver at slave address 0x60
NET: Registered protocol family 10
modprobe: 'ext/rcu2r.ko': No such device
*** STARTING APP ****
app exited..rebooting RCU

login[995]: root login on 'ttyS0'

BusyBox v1.15.3 (2012-05-15 09:58:41 MDT) built-in shell (ash)
Enter 'help' for a list of built-in commands.

******* .profile script *******
# ./mt9p031_ccdc_prv_rsz_file
calling cmem utilities for allocating frame buffIRQK Error: ioctl(): bad command type 0x4d (should be 0xfc)
ers
Allocating capture buffers :buf size = 692224
Got 0x403b4000 from CMEM, phy = 0x87eae000
Got 0x404f4000 IRQK Error: ioctl(): bad command type 0x4d (should be 0xfc)
from CMEM, phy =IRQK Error: Unknown ioctl received: -1070314236.
0x87e05000
Got 0x40690000 from CMEM, phy = 0x87d5c000
**********************************************
4.enumerating media entities
total number of entities: 0
**********************************************
5.enumerating links/pads for entities
**********************************************
6. ENABLEing link [tvp7002]----------->[ccdc]
failed to enable link between tvp7002 and ccdc
closing all the file descriptors. . .
closed media device
Segmentation fault
#
#

over 13 years ago

0 Prabhakar Lad over 13 years ago

Guru 11260 points

Hi Jeremy,

You can find the the root file system in arago releases http://arago-project.org/files/releases/

Are the devices nodes and subdevices created for media controller framework ?

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

I am trying to get media controller framework working. Your question actually helps me clarify mine. How do I create the nodes and subdevices for media controller framework?

0 Jeremy Larsen over 13 years ago in reply to Jeremy Larsen

Prodigy 250 points

To clarify:

I am using v2.6.37_DAVINCIPSP_03.21.00.04 on a custom board using DM365. I am trying to get media controller functionality working. I would like a base file system for this setup.

As a side note I do have openembedded installed and can make the dm365-evm recipes and look at those images but these use the 2.6.32-psp03.01.01.39 kernel and the media controller code is not in that version.

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi Jeremy,

Which Branch are you working on ? The following branch

http://arago-project.org/git/projects/?p=linux-davinci.git;a=shortlog;h=refs/heads/dvsdk

from arago has capture driver based on media controller framework for dm365/dm368.

It has been tested, And this is the http://arago-project.org/git/projects/?p=examples-davinci.git;a=shortlog;h=refs/heads/422-420SP

examples branch from arago for this framework. From the top most what I can see you are still using the old driver rite?

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

I am not using old drivers. What part of the above output would lead you to believe I am using old drivers?

I am working from the branches mentioned above. It seems that /dev/media0 is not being setup correctly. I can make it with mknod but it still does not work.

The above branches do not build a root file system. Could you point me in the right direction to create a base root file system that will work with media controller framework?

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi,

You can download the file system from http://arago-project.org/files/releases/davinci-psp_03.01.01.39/images/dm365-evm/

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

This filesystem seems to be based of a 2.6.32 Kernel. The media controller is not in this version of the kernel.

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi,

Media controller framework is tested on this filesystem too.

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

Thank you. I was able to get the example app working.

Does the media controller framework support "resizer b" or in other words can I base my app on the examples and get a second re sized image form the resizer?

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi Jeremy,

Yes it does support resizer b, but this code is not present in arago branches, you can use this branch

http://git.linuxtv.org/mhadli/v4l-dvb-davinci_devices.git/shortlog/refs/heads/mc_opensource_full

which has this support.

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

Is there any example code that uses resizer b? The example code from arago has resizer b setup commented out and it is not clear how the resizer b output would be obtained.

How does the above code relate to the arago code?

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi Jeremy,

Is there any example code that uses resizer b? The example code from arago has resizer b setup commented out and it is not clear how the resizer b output would be obtained.

How does the above code relate to the arago code

You can refer this example http://git.linuxtv.org/mhadli/v4l-dvb-davinci_devices.git/blob/196d693662b4aca3943757ff01dc7def7cd21c94:/mc/dm365/resizer_b_ss.c

for resizer-b.

This branch is a development branch, once all the features are added and tested

it will be pushed on to arago branches.

Thx,

--Prabhakar Lad

------------------------------------------------------------------------------

Please click the Verify Answer button on this post if it answers your question.
-------------------------------------------------------------------------------

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

Thank you so much for your help Prabhakar. Is there anymore media controller sample code that uses this kernel, preferably code that uses ccdc->prev->resizeAandB in continuous mode?

Jeremy Larsen

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi,

http://git.linuxtv.org/mhadli/v4l-dvb-davinci_devices.git/tree/196d693662b4aca3943757ff01dc7def7cd21c94:/mc/dm365

you can find the examples over here.

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

I am using the kernel at http://git.linuxtv.org/mhadli/v4l-dvb-davinci_devices.git branch mc_opensource_full for resizerB support. This kernel also has differences in drivers/usb from arago. Which usb code should I use? Which one is later?

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi,

You can take only video related from http://git.linuxtv.org/mhadli/v4l-dvb-davinci_devices.git this

git tree rest all stuff you need to take from the Arago itself.

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

Is there any reason that I cant get the resizer to output semi-planar format? I am using continuous mode. [MT9P031]--->[CCDC]--->[PRV]--->[RSZ]--->encode (need nv12)

0 Jeremy Larsen over 13 years ago in reply to Jeremy Larsen

Prodigy 250 points

I am now using video stuff from http://git.linuxtv.org/mhadli/v4l-dvb-davinci_devices.git branch mc_opensource_full and the rest of the kernel is arago linux-davinci. I am getting resizer B output but I cannot seem to make the output from A or B be in the semi-planar format (NV12). Does this code support this format in continuous mode?

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi,

Yes it does support NV12 format in continious mode too.

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

Thanks for your response. None of the examples produce NV12 in continuous mode. I have tried to do it myself. No matter what format I try and set on the resizer sink or source pad I get the same format which looks like UYVY8. I can see where it is supported in the kernel code but this code only gets run at the very start of my program somehow. Is does not get run when I set the resizer up. Is there any example or description on how to get NV12, in continuous mode, while using both resizers?

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi,

Please refer this post http://e2e.ti.com/support/dsp/davinci_digital_media_processors/f/100/p/190773/693090.aspx#693090

I have attached application over there for NV12 in continuous mode.

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

Thanks I have used that application to create mine. I have another question. When I try to flip either vertical or horizontal in the resizer the frame freezes. I get subsequent frames but they rotate through about three frames then start again with the same three frames. Is this functionality supported?

Jeremy Larsen

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi,

When I try to flip either vertical or horizontal in the resizer the frame freezes. I get subsequent frames but they rotate through about three frames then start again with the same three frames. Is this functionality supported?

From the outset looks like its supported. never tested it though.

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

Prabhakar Lad,

I am trying to upgrade from a 2.6.30 kernel, with early external video drivers (previewer, resizer, h3a,aew from TI), to a 2.6.37 kernel. These old video drivers are working fine with resizer B and flip horizontal and vertical working. The reason I want to upgrade is not for better video functionality. The reason for the upgrade is for better USB support. The USB in 2.6.30 is unstable and we have tried to find the problem at great lengths and failed. I would like to know what your suggestion would be to get the USB code working properly. I have tried updating video to using media controller framework and using kernel 2.6.37, but cant get video working correctly. Flip does not work and I cant get the semi-planar format correct.

I have also tried just using the old (2.6.30) video code in the new kernel but there are so many dependencies between the drivers/media/video/davinci folder and the arch/arm/mach-davinci folder. This makes it hard to use the old video code, plus there have been changes to v4l2.

Is there an easy way to get all the benefits of the new kernel but still use my old external video drivers?

0 Prabhakar Lad over 13 years ago in reply to Jeremy Larsen

Guru 11260 points

Hi,

I have tried updating video to using media controller framework and using kernel 2.6.37, but cant get video working correctly. Flip does not work and I cant get the semi-planar format correct.

Cant get video what is that its failing ? Cant get semi planar format correct

that is working if I am not wrong. I had posted in my earlier post, that app

is tested and is working.

Thx,

--Prabhakar Lad

0 Jeremy Larsen over 13 years ago in reply to Prabhakar Lad

Prodigy 250 points

I cannot get the semi-planar format correct. I need this format because the image is then encoded. It produces an image but the format is not correct. I tried continuous mode. Would I have better luck in single shot mode?

Also, if I flip horizontal or vertical (in the resizer) the video stream continues but just cycles through the same three frames.

My current plan is to us the 2.6.37 kernel but completely replace the video code with the 2.6.30 video code. This becomes tricky because of the dm365.c file in /arch/arm/mach-davinci is dependent on the new video code.

Jeremy

Processors

Processors forum

DM365/mt9p031 Media framework question