Hi, Great work Hans, I sure hope this fixes the lingering stability problems for ivtv.
Did a quick test: Seems to record fine here with mythtv 0.20 on a PVR-350, and the remote works too (I'm not using the 350's tvout anymore because I got a nvidia card, and with 'nvidia-settings -a /tvoverscan=19' it has a perfectly fine svideo-out and much less hassle with getting xorg drivers, worrying about the additional activity on the '350 crashing my box, etc). I'll keep using it and report problems if I find&identify any. My main question is: Why are you using a 376836 byte firmware and not one of 262144 bytes? There are really only 8 nonzero bytes in the rather large part after the first 256KB, do you think they mean anything useful? I'm asking, because I'm pondering about submitting a patch for cx88-blackbird in v4l to accept the 376836-byte version of the firmware, because both drivers look for the firmware with the same filename, so there now is a conflict between the drivers... But I'm not sure if I should make it upload all 376836 bytes or just the first 256K. Jelle. Hans Verkuil wrote: > Hi all, > > I've released ivtv driver 0.10.0 for kernels 2.6.18 and up. I have also > updated the encoder firmware to the latest version (2.06.039). > > You can get the new firmware here: > > http://dl.ivtvdriver.org/ivtv/firmware/firmware.tar.gz > > And the new driver here: > > http://dl.ivtvdriver.org/ivtv/archive/0.10.x/ivtv-0.10.0.tar.gz > > The new firmware is required by the new driver. > > This driver replaces the ivtv-0.8.x and ivtv-0.9.x releases. These > driver versions will no longer be maintained. > > This driver is the result of almost 6 months of work and contains a > major internal redesign and clean up. All new development will > concentrate on this version. Older drivers are effectively discontinued > (they are as stable as they are ever going to get anyway). > > The README of the 0.10.0 driver contains some important notes that I > have copied below: > > > NOTE: this is the first release of the 0.10.x series. The core of the > driver has been redesigned completely because the older drivers had > design flaws that made it impossible to fix certain race conditions. I > also took the opportunity to clean up the code (an almost 25% reduction > in code size!). > > While this driver has been tested pretty well, I cannot guarantee that > there are absolutely no bugs remaining. > > So if you are using ivtv in a production environment, then you may want > to be careful introducing this new version. I'm phasing out support for > the older drivers, so upgrading is strongly recommended. > > > NOTE: This driver will work with kernels 2.6.18, 2.6.19 and 2.6.20. For > the latter kernel you will need to apply the cx25840.diff patch that is > in the 'misc' directory (only if your card has the cx2584x chip, it is > not needed otherwise). The 2.6.20 kernel caused a cx2584x bug to > surface that prevents loading of the cx2584x firmware. > > Hopefully this will be fixed in the next 2.6.20.x dot release. > > > NOTE: ivtv-fb now requires that you enable the following kernel config > options: > Go to 'Device drivers -> Graphics support'. Enable 'Support for frame > buffer devices'. Enable 'Trident support' (the Trident module itself is > not required). > > To get working console output, keep going to 'Console display driver > support', and enable 'Framebuffer Console support'. Enable 'Select > compiled-in fonts' & once that's done, you should have a list of fonts. > Choose one. With the default OSD size, 'VGA 8x16' gives 80x30(PAL) > 80x25(NTSC). > > > NOTE: users of MythTV who use the PVR350 for mpeg decoding need to > specify the option osd_compat=1 when loading the ivtv-fb module. This > is not required when using the default Xv output. > > > NOTE: users of ivtv X driver may need to specify the option osd_compat=1 > when loading the ivtv-fb module. You will only need it if the display > becomes scrambled in X. > > > Enjoy! > > Hans > > _______________________________________________ > ivtv-devel mailing list > [email protected] > http://ivtvdriver.org/mailman/listinfo/ivtv-devel > _______________________________________________ ivtv-devel mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-devel
