And the syslog messages, edited to fit.  Note that the messages ran on
until the kernel was halted.

-----Original Message-----
From: Quentin Conner [mailto:[EMAIL PROTECTED] 
Sent: Wednesday, January 05, 2005 5:54 PM
To: '[email protected]'
Subject: RE: [ivtv-devel] Fedora Core 3



The Xorg stdout/stderr.

-----Original Message-----
From: Quentin Conner [mailto:[EMAIL PROTECTED] 
Sent: Wednesday, January 05, 2005 5:42 PM
To: '[email protected]'
Subject: RE: [ivtv-devel] Fedora Core 3



Okay.  I will send the /var/log/messages and Xorg output under separate
cover (problems with 20K limit).  Please have a look and let me know
what you think.

I've src on hand if you want to send me any instrumented code.  I'm
grateful for the help!

-Quentin

-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of John Harvey
Sent: Wednesday, January 05, 2005 3:14 PM
To: [email protected]
Subject: RE: [ivtv-devel] Fedora Core 3

I don't believe it's the x-driver. It is more likely to be the ivtv osd
driver.

The top bottom half thing comes from the x driver splitting large
updates into 2. So the 1st one has worked and the second on failed for
some reason. Turn full ivtv logging on and send me your messages file at
the point it fails if you like. 

The most likely cause is that there is a loop where the osd code tries
to gain access to the dma engine and waits until it can get access. This
can block indefinitely if something goes wrong and the x server is
blocked in the ioctl waiting for it to happen. 
Maybe we should return from this loop after a short time (1 minute or
so) with a failure since something has gone horribly wrong when that
happens.

There is debugging for this happening something like Reg Status .......
But the decoder debugging needs to be on to show it.

John

> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:ivtv-devel-
> [EMAIL PROTECTED] On Behalf Of Quentin Conner
> Sent: 05 January 2005 20:50
> To: [email protected]
> Subject: RE: [ivtv-devel] Fedora Core 3
> 
> 
> I'm currently working on FC3 as well and successful with MythTV other
> than the OSD Framebuffer feature on the PVR-350.
> 
> This PVR-350 IVTVDEV driver with Xorg doesn't work for me on 2.6.9
> with Xorg 6.8.1.  After starting/stopping X on my CLE266 chipset (KDE 
> via
> gdm) a few of times the IVTVDEV FB driver *may* work with X restarted
> against the PVR-350 FB.  When the IVTVDEV FB doesn't work I see the
> composite display wipe the top half of the display to black;  the
bottom
> half retains it's original pattern;  the X cursor displays in the
> middle;  keyboard is locked up.
> 
> Since I am only a couple of weeks into this exercise I can't offer
> much more at this time other than I am thinking about trying another 
> distribution (suspect kernel version and/or Xorg distribution).  I've 
> tried the 2.6.10 FC4 development kernel as well as played with my own 
> FC3 based kernels (stripped down to eliminate other modules).
> 
> If anyone has suggestions how to debug the ivtvdev driver (need to
> find the source to start with), how to use the fbdev FB X driver, or 
> knows which of the ivtvdev_drv.o Framebuffer object files from badzzzz

> or Chris Kennedy 0.2 or 0.3 genre are supposed to work with Xorg and 
> 2.6.9., please let me know.
> 
> Thanks,
> 
> Quentin
> 
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of D.
> Hageman
> Sent: Wednesday, December 29, 2004 5:19 PM
> To: [email protected]
> Subject: [ivtv-devel] Fedora Core 3
> 
> 
> I upgraded my PVR box to Core 3 recently and I am having issues with a
> couple of items.
> 
> 1) The driver series greater then 0.3.1r give me errors such as:
>       ivtv: OSD: DMA xfer from 0xb7efb7d8 of 655360 bytes failed with
> (-512) offset = 0x000b17d0, total 1382352
> I noticed that in 0.3.1s a patch was added to fix DMA issues in
> kernels above 2.6.8.  Apparently the stock fedora kernels don't have 
> whatever issue was being addressed.
> 
> 2) The only drivers I have been able to get work with MPlayer's ivtv
> patch is 0.3.x series.  All other drivers seem to complain about 
> invalid arguments to mmap'ing the device and then I get errors about 
> "Error with
> 
> sending DMA to ivtv".  This has been mentioned several times before on
> this list without resolution.
> 
> 3) The mythtv backend (0.16 and cvs) will segfault (without core)
> after an hour of recording (two 30 minute programs).  I don't think 
> this is entirely a fault of mythtv on Core 3.  I have tried debugging 
> the backend a
> couple of times unsuccessfully, but when I have ... I have caused the
> ivtv
> driver to start complaining about ivtv: "ENC IRQ OVERFLOW: #4007
> Stealing
> a Buffer, 320 currently allocated"  I think some condition that ivtv
is
> creating under Core 3 is unexpected in the mythbackend.  If anyone is
> running mythtv and core 3, please let me know what kernel, driver
> version,
> etc you are using.  I will continue to investigate on this end.
> 
> 
> //========================================================\\
> ||  D. Hageman                    <[email protected]>  ||
> \\========================================================//
> 
> 
> -------------------------------------------------------
> The SF.Net email is sponsored by: Beat the post-holiday blues Get a
> FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun 
> and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
> _______________________________________________
> ivtv-devel mailing list
> [email protected] 
> https://lists.sourceforge.net/lists/listinfo/ivtv-devel
> 
> 
> 
> -------------------------------------------------------
> The SF.Net email is sponsored by: Beat the post-holiday blues Get a
> FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun 
> and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
> _______________________________________________
> ivtv-devel mailing list
> [email protected] 
> https://lists.sourceforge.net/lists/listinfo/ivtv-devel



-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE
limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and
FREE -- well, almost....http://www.thinkgeek.com/sfshirt
_______________________________________________
ivtv-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ivtv-devel

Attachment: msgs.gz
Description: GNU Zip compressed data

Reply via email to