Hi Linus, Would you please be kind enough to rip every material done by Trent Piepho out of git2? At least the following commitments please: commit 6bdcc6e6dbab8daffd05e5026486f34ba41a6c72 Author: Trent Piepho Date: Fri Apr 27 12:31:30 2007 -0300 V4L/DVB (5363): Dvb: Remove lgh06xf driver The code of the dvb-pll driver and the lgh06xf driver is nearly identical. The main difference is that the lgh06xf driver would set the AGC TOP value on every tune call. The dvb-pll driver now has the ability to set the AGC TOP when the front-end device is opened, which is a better way to go about it. By using this ability of dvb-pll, the lgh06xf driver is made unnecessary. There is one other difference. dvb-pll will probe for the presence of an I2C pll chip by doing a one byte read, the lgh06xf driver did not do this. In some devices the PLL is not reachable over I2C at the timer the tuner is attached. Some more initialization, such as firmware loading, must take place first. None of the devices using a LG-H06xF should have this problem. Signed-off-by: Trent Piepho Acked-by: Michael Krufky Signed-off-by: Mauro Carvalho Chehab commit 26aed92289ae8caffeedebb3f841a9e0371119a9 Author: Trent Piepho Date: Fri Apr 27 12:31:29 2007 -0300 V4L/DVB (5362): Dvb-pll: add code for doing tuner initialization Some tuners need or benefit from initialization, to change certain settings from their power on default values. Typically, tuners with TUA603x PLLs can benefit from setting the AGC TOP value to something else. This patch includes code to set the AGC TOP to 103 dBuV for the Thomson DTT-761x tuners, which I have experimentally verified gives the best SNR readings, increasing SNR by about 0.19 dB over the default value. Other tuners can make use of this as well. For example, the separate LG TDVS-H06xF driver's only difference from dvb-pll is this same setting of AGC TOP value. Signed-off-by: Trent Piepho Signed-off-by: Mauro Carvalho Chehab commit 9ab1ba38e6a25a480234f196aa26239e28ac2407 Author: Trent Piepho Date: Fri Apr 27 12:31:28 2007 -0300 V4L/DVB (5361): Dvb-pll: Fix Kconfig files and allow dvb-pll to be optional A number of drivers selected DVB_PLL when they did not need it, and some that did need it did not select it. The DVB_PLL option is given a name and help text, so that it will show up in the config menu. DVB_PLL support can be turned on if an out-of-tree driver needs it. The standard dvb fe customization support is added to dvb-pll.h. Since all modules which select DVB_PLL do so unconditionally, it is not possible to turn dvb-pll off when an enabled module selects it, unlike most of the other frontend/tuner drivers. This is because the users of dvb-pll have static references to dvb-pll symbols other than the attach function. If these references are removed, then dvb-pll will be disablable as the other frontend/tuner drivers are. Signed-off-by: Trent Piepho Signed-off-by: Mauro Carvalho Chehab commit 982dd1bdaab39185ccd55cafd3c099b7fee5f2dd Author: Trent Piepho Date: Fri Apr 27 12:31:27 2007 -0300 V4L/DVB (5360): Dvb-pll: Use sizeof() to get name length Better to use sizeof() to get the size of the output buffer for the tuner name, instead of just hard coding 128. Signed-off-by: Trent Piepho Signed-off-by: Mauro Carvalho Chehab Reason for my polite request is simply the following mail: Von: Trent Piepho An: Uwe Bugla Betreff: Re: Your latest work on dvb-pll.c Datum: Sat, 28. Apr 2007 12:35:36 -0700 On Sat, 28 Apr 2007, Uwe Bugla wrote: > > Hi, > > I owe a Pinnacle PCTVSAT. This card does NOT need dvb-pll.c. > > But it unfortunately needs drivers that are only selectable by activating > > VIDEO_V4L1. > > > > But if I select VIDEO_V4L1 the dvb-pll.c module is not deselectable at all. > > It is only deselectable if VIDEO_V4L1_COMPAT is selected. > > > > Can you please give me a hint where some additional code is needed, or can you please write another additional patch to solve that problem, please? > > > > Meanwhile I will try on my own, but it does not seem to be easy to solve this. Currently dvb-pll must be compiled if you have any driver turned on that uses it. There are plans to fix this, but it's not a high priority. This is A LIE! The Truth is: In the current git2 solution you are also condemned to compile it even if you turn on drivers that never in their history ever used it! This is what I call a regression, as I know very exactly that for instance all Pinnacle cards and all TwinHan and clone cards do not need this dvb-pll.c module at all and definitely never needed it, not in their 2.6 kernel appearance, and not in their 2.4 kernel appearance! So this work done by Trent Piepho is a fallback into the year 2005's development state. Utmost reactionary and in NO way tolerable at all! So please rip this stuff out of 2.6.21-git2 until a satisfying solution for everybody is being offered. And do not pay attention neither to Chehab, nor to Krufky! They are just ignorants of the guaranteed poorest sort. They sign and ack crap like this while they keep on ignoring my contributions who do not contain any kind of regression at all! It is a very deep shame what crap is signed by Chehab and acked by Krufky! And thus pulled into mainline! Thanks in advance Yours sincerely Uwe -- "Feel free" - 10 GB Mailbox, 100 FreeSMS/Monat ... Jetzt GMX TopMail testen: http://www.gmx.net/de/go/topmail