PATCHES: device definitions, 624f, 627f

[Comer352l]

Hi,

I'm sending you 3 patches for testing, review and sign-off:

PATCH 1: Updates device definitions
- adds the following device definitions:
0c45:624c (SN9C201 + MI1320)
0c45:6253 (SN9C201 + OV9650)
0c45:6262 (SN9C201 + OM6802)
0c45:628c (SN9C202 + MI1320)
0c45:62a2 (SN9C202 + OM6802)
0c45:62b3 (SN9C202 + OV9655)
0c45:62be (SN9c202 + OV7663)
045e:00f4 (SN9C202 + OV9650) Microsoft LifeCam VX-6000 (clone of
0c45:628f)
145f:013d (SN9C201 + OV7660) Trust WB-3600R (clone of 0c45:627b)
04f2:a128 (SN9C202 + OV7663) Chicony Panda 7 (clone of 0c45:62be +
EEPROM)
All VID:PID-combinations and their bridge-sensor-combinations are
confirmed by .inf-files of MS Windows drivers.
The "clone"-devices are internally handled as Microdia-devices, but
there is an individual dmesg-output.

PATCH 2: Several fixes for device 624f
- fixes regression of manual exposure control
- fixes 2 incorrect numbers of bytes for I2C-writes in
microdia_624f_start_stream()
- fixes write to wrong (bridge-) register address in
microdia_624f_start_stream()

PATCH 3: adds support for device 627f
- adds initialisation-function for device 627f
- merges 627f-specific reads/writes to function microdia_624f_start_stream()
- enables device 627f

As you can see, they should be tested especially by 624f and 627f-owners !

Frank

[Comer352l]

This time with patches attached...

[Vasily Khoruzhick]

On 21 June 2008 16:13:04 Comer352l wrote:
> This time with patches attached...
>

> --~--~---------~--~----~------------~-------~--~----~
> Lets make microdia webcams plug'n play, (currently plug'n pray)
> To post to this group, send email to micr...@googlegroups.com
> Visit us online https://groups.google.com/group/microdia
> -~----------~----~----~----~------~----~------~--~---

First patch breaks build on my machine:

/home/anarsoul/work/microdia/microdia: struct usb_device_id is 20 bytes. The
last of 38 is:
0x03 0x00 0xf2 0x04 0x28 0xa1 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x00 0x00 0x00 0x00 0x00
FATAL: /home/anarsoul/work/microdia/microdia: struct usb_device_id is not
terminated with a NULL entry!

you've forgot to terminate microdia_table in microdia-usb.c with empty entry.

other 2 patches looks OK to me. You can push them ;)

Some comments on master branch:

fps is VERY low on my 624f when I use v4l2 with master branch (its OK with
i420 and v4l). Also it uses rgb-decoder by default (why? we can get rgb only
by converting i420 to rgb, output stream format is mixed i420, so it's better
to use i420 by default), and I can't switch it by specifying format=i420 in
mplayer.

Also RGB-order is wrong (I've got wrong colors)

Regards
Vasily

[Bilbo75]

Hi Comer,I have a microsoft VX-6000.
Are you sure that this model of webcam works like a 628f?
Until now I've always manually patched the entry for 624f !
I've tried anyway your patches,the camera is detected but it doesn't
work at all.
Also JoJo told me that the camera looks like a 624f (SN9C201)

Andrea

[Comer352l]

Vasily Khoruzhick schrieb:

> On 21 June 2008 16:13:04 Comer352l wrote:
>
>> This time with patches attached...
>>
>> >>
>

> First patch breaks build on my machine:
>
> /home/anarsoul/work/microdia/microdia: struct usb_device_id is 20 bytes. The
> last of 38 is:
> 0x03 0x00 0xf2 0x04 0x28 0xa1 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> 0x00 0x00 0x00 0x00 0x00
> FATAL: /home/anarsoul/work/microdia/microdia: struct usb_device_id is not
> terminated with a NULL entry!
>
> you've forgot to terminate microdia_table in microdia-usb.c with empty entry.
>
> other 2 patches looks OK to me. You can push them ;)
>
> Some comments on master branch:
>
> fps is VERY low on my 624f when I use v4l2 with master branch (its OK with
> i420 and v4l). Also it uses rgb-decoder by default (why? we can get rgb only
> by converting i420 to rgb, output stream format is mixed i420, so it's better
> to use i420 by default), and I can't switch it by specifying format=i420 in
> mplayer.
>
> Also RGB-order is wrong (I've got wrong colors)
>
> Regards
> Vasily
>

Fixed. Thanks for your fast feedback !
I420 should be the standard palette, but I get the same results: it
doesn't work.

Frank

[Vasily Khoruzhick]

On 22 June 2008 11:38:54 Comer352l wrote:
>
> Fixed. Thanks for your fast feedback !

OK, one GO from me :)

> I420 should be the standard palette, but I get the same results: it
> doesn't work.
>
> Frank

why? It works OK in v4l2-only branch. And it worked OK earlier :)

[Comer352l]

Hi Andrea,

first of all: the patches do NOT add support for your device yet.
Yes, the Microsoft VX-6000 is definitely a 628f (SN9C202 + OV9650).
624f uses the same sensor but bridge SN9C201, which has no audio-processor.
The 202-devices seem to work with the corresponding functions of their
201-brothers, but we need to find out the differences and therefore we
need logs. As soon as we have a complete and valid set of logs, 628f
will be supported.

Frank

Bilbo75 schrieb:

[Vasily Khoruzhick]

On 22 June 2008 11:38:54 Comer352l wrote:

> Fixed. Thanks for your fast feedback !
> I420 should be the standard palette, but I get the same results: it
> doesn't work.
>
> Frank

There's performance impromevent patch for i420 624* decoder that I've sent
before, now for master branch. Please test it :)

[Bilbo75]

K,I understand.
Thanks again for your answer
I'll try to catch some log for you then.

Ciao.

[Vasily Khoruzhick]

On 22 June 2008 12:11:32 Vasily Khoruzhick wrote:
> There's performance impromevent patch for i420 624* decoder that I've sent
> before, now for master branch. Please test it :)

Also it removes old bayer2anything functions, because sn9c20* format is not
bayer.

[Comer352l]

Vasily Khoruzhick schrieb:

>> I420 should be the standard palette, but I get the same results: it
>> doesn't work.
>>
>> Frank
>>
> why? It works OK in v4l2-only branch. And it worked OK earlier :)
>

Some of the V4l2-ioctls seem to fail (see error messages from MPlayer).
It could have happened with the changes for 6260 that were made in that
part of the code, but that's just a guess at the moment.

Frank

[Comer352l]

Vasily Khoruzhick schrieb:

I can't evaluate the performance progress (because I have a 6270), but
the changes make sense and it doesn't break my cam.
One "GO" from me !

Frank

[Vasily Khoruzhick]

On 22 June 2008 12:35:07 Comer352l wrote:

> I can't evaluate the performance progress (because I have a 6270), but
> the changes make sense and it doesn't break my cam.
> One "GO" from me !

Ok, thanks for response ;)

Vasily

[victorhr]

Hi Comer352l,

First, thank you very very much for these patches, especially the 627f
support ! yay!!!
I have tested the patches in my integrated 0c45:627f cam on a lenovo
3000 n100 laptop, in ubuntu 8.04 standard kernel (2.6.24).

It works with ekiga, mplayer and amsn. The only problem is that the
image is very dark, somewhat more than in windows, but I suppose
that's because of the bad quality of the cam.
Anyway dmesg doesn't report any error messages, only [ 1191.724891]
microdia: Read from EEPROM successful, messages. The dmesg driver
insertion messeges are:
[ 708.037231] Linux video capture interface: v2.00
[ 708.437024] microdia: Microdia USB2.0 webcam driver startup
[ 708.437094] microdia: Microdia USB2.0 Webcam - Product ID 627F.
[ 708.437099] microdia: Release: 0100
[ 708.437102] microdia: Number of interfaces : 1
[ 708.441928] microdia: Microdia USB2.0 Camera is now controlling
video device /dev/video0
[ 708.441997] usbcore: registered new interface driver
usb_microdia_driver
[ 708.442004] microdia: v0.0.0 : Microdia USB Video Camera
Tell me if you need me to test something else.
Thanks again!

Victor

[GWater]

Vasily Khoruzhick schrieb:

All patches work great.
Minor whitespace issues though:

0003-adds-support-for-device-627f.patch:1270: trailing whitespace.
// These 2 regs controls brightness of image
0004-Perfomance-improvements-for-624-i420-decoder.patch:1598: trailing
whitespace.
{ 0 , 0 }, { 1 , 0 }, { 2 , 0 }, { 3 , 0 }, { 4 , 0 }, { 5 , 0 },
{ 6 , 0 }, { 7 , 0 },
0004-Perfomance-improvements-for-624-i420-decoder.patch:1599: trailing
whitespace.
{ 0 , 1 }, { 1 , 1 }, { 2 , 1 }, { 3 , 1 }, { 4 , 1 }, { 5 , 1 },
{ 6 , 1 }, { 7 , 1 },
0004-Perfomance-improvements-for-624-i420-decoder.patch:1600: trailing
whitespace.
{ 0 , 2 }, { 1 , 2 }, { 2 , 2 }, { 3 , 2 }, { 4 , 2 }, { 5 , 2 },
{ 6 , 2 }, { 7 , 2 },
0004-Perfomance-improvements-for-624-i420-decoder.patch:1601: trailing
whitespace.
{ 0 , 3 }, { 1 , 3 }, { 2 , 3 }, { 3 , 3 }, { 4 , 3 }, { 5 , 3 },
{ 6 , 3 }, { 7 , 3 },
warning: squelched 12 whitespace errors
warning: 17 lines add whitespace errors.

BTW I posted a GO for the improved 624x-decoder for v4l2-branch a few
days ago.

So here is one more GO for these four patches.

GWATER

[Vasily Khoruzhick]

On 22 June 2008 14:31:59 GWater wrote:

> BTW I posted a GO for the improved 624x-decoder for v4l2-branch a few
> days ago.
>
> So here is one more GO for these four patches.
>
> GWATER

Frank, push your patches ;)
You can also push my patch (performance improvement patch)

Regards
Vasily

[Comer352l]

Vasily Khoruzhick schrieb:

PUSHED. Please update your local repos.