Nabble - HW+ / DXR3 device driver for Linux

First release-candidate for 0.17.2 -- please test!

2008-10-11T06:42:03Z

Hi,

Linux kernel 2.6.27 was released yesterday, and as with most linux
kernel releases nowadays, it breaks the API.
Hance, I just packed the first release candidate for 0.17.2. There isn't
much news in this release, except the support for linux 2.6.27, and much
cleanup by Christian Gmeiner.
As usual you can get it from
http://dxr3.sourceforge.net/download/em8300-0.17.2-rc1.tar.gz

Many thanks to Christian Gmeiner who did most the work.

All comments are welcome. Please give it a try and report.

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Any progress on 2.6.27?

2008-10-11T01:34:59Z

Adam wrote:
> Is there any progress with linux 2.6.27 support?
>
> I'll be thankful for information.

Works great for me. You have to use Christians repo though:
http://freehg.org/u/austriancoder/em8300-cgmeiner/

- jan

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Any progress on 2.6.27?

2008-10-10T18:12:41Z

Hello,

Is there any progress with linux 2.6.27 support?

I'll be thankful for information.

--
Regards,
Adam

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-10-07T12:28:30Z

* Nicolas Boullis <nboullis@...>:

| Hmmm... it looks like SF's service is not very reliable... :-(
| As for the "Sender:" header, as far as I know, it is not required; why
| would you expect one?

I'm sorti most of the incoming mailing list messages based on the
Sender headers. Who removed it from the previous setup?

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-10-06T16:37:16Z

On Sun, Oct 05, 2008 at 07:20:39PM +0200, Gabor Z. Papp wrote:

> * "Gabor Z. Papp" <gzp@...>:
>
> | | > And where disappeared the "Sender:" header?
>
> | | Uh!? What do you mean?
>
> | I have no Sender header in the list mails.
>
> And according to the Received lines this mail slept a bit at SF.net
>
> Received: from localhost ([127.0.0.1] helo=sfs-ml-1.v29.ch3.sourceforge.com)
> by 235xhf1.ch3.sourceforge.com with esmtp (Exim 4.69)
> (envelope-from <dxr3-devel-bounces@...>)
> id 1Km9R6-0003BQ-Uw; Sat, 04 Oct 2008 15:52:52 +0000
> Received: from sfi-mx-3.v28.ch3.sourceforge.com ([172.29.28.123] helo=mx.sourceforge.net)
> by 235xhf1.ch3.sourceforge.com with esmtp (Exim 4.69)
> (envelope-from <gzp@...>) id 1Kj93w-0006h4-Q1
> for dxr3-devel@...; Fri, 26 Sep 2008 08:52:32 +0000
>
> Whats happening there?

Hmmm... it looks like SF's service is not very reliable... :-(
As for the "Sender:" header, as far as I know, it is not required; why
would you expect one?

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: autoconf problem

2008-10-06T08:47:21Z

This was spot on and I was able to do the make and make install without problem.

----- Original Message ----
From: Brendon Higgins <blhiggins@...>
To: dxr3-devel@...; Clendon Gibson <bsandyman@...>
Sent: Sunday, October 5, 2008 8:26:16 PM
Subject: Re: [Dxr3-devel] autoconf problem

Hi,

Clendon Gibson wrote (2008-10-05 2:06 pm):
> gcc: '-V' option must have argument

That part seems to be okay (that is, it's a non-fatal failure). The real
problem is this:
/usr/bin/ld: crt1.o: No such file: No such file or directory
collect2: ld returned 1 exit status

Seems you're missing the libc6-dev package, which contains the crt1.o file.
There are couple of other -dev packages you'll also need. Install the
libgtk2.0-dev and it ought to pull in libc6-dev as well as some others
required.

In case you didn't realise, the dxr3 drivers do have precompiled packages in
the contrib section ("em8300", et al.), though in Etch they're version
0.16.0, so perhaps a little dated.

Good luck!

Peace,
Brendon

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: autoconf problem

2008-10-05T18:26:16Z

Hi,

Clendon Gibson wrote (2008-10-05 2:06 pm):
> gcc: '-V' option must have argument

That part seems to be okay (that is, it's a non-fatal failure). The real
problem is this:
/usr/bin/ld: crt1.o: No such file: No such file or directory
collect2: ld returned 1 exit status

Seems you're missing the libc6-dev package, which contains the crt1.o file.
There are couple of other -dev packages you'll also need. Install the
libgtk2.0-dev and it ought to pull in libc6-dev as well as some others
required.

In case you didn't realise, the dxr3 drivers do have precompiled packages in
the contrib section ("em8300", et al.), though in Etch they're version
0.16.0, so perhaps a little dated.

Good luck!

Peace,
Brendon

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-10-05T10:20:39Z

* "Gabor Z. Papp" <gzp@...>:

| | > And where disappeared the "Sender:" header?

| | Uh!? What do you mean?

| I have no Sender header in the list mails.

And according to the Received lines this mail slept a bit at SF.net

Received: from localhost ([127.0.0.1] helo=sfs-ml-1.v29.ch3.sourceforge.com)
by 235xhf1.ch3.sourceforge.com with esmtp (Exim 4.69)
(envelope-from <dxr3-devel-bounces@...>)
id 1Km9R6-0003BQ-Uw; Sat, 04 Oct 2008 15:52:52 +0000
Received: from sfi-mx-3.v28.ch3.sourceforge.com ([172.29.28.123] helo=mx.sourceforge.net)
by 235xhf1.ch3.sourceforge.com with esmtp (Exim 4.69)
(envelope-from <gzp@...>) id 1Kj93w-0006h4-Q1
for dxr3-devel@...; Fri, 26 Sep 2008 08:52:32 +0000

Whats happening there?

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

autoconf problem

2008-10-04T21:06:06Z

Hi folks,

I am trying to install the dxr3 driver on a home computer with a Hollywood RealMagic card. The machine is a 800 MHZ Celeron with Debian Etch on it. I have the 2.6.18-6 kernel.

I attempted to install the development tools to compile the module but I suspect I am missing something. When I do "./configure" in the top level is complains with the following messages.

checking for C compiler default output file name...
configure: error: C compiler cannot create executables
See `config.log' for more details.

I think this is the relevant portion, but the whole thing is attached.

checking for gcc... gccconfigure:2130: gcc -V >&5
gcc: '-V' option must have argument
configure:2133: $? = 1
configure:2156: checking for C compiler default output file name
configure:2183: gcc conftest.c >&5
/usr/bin/ld: crt1.o: No such file: No such file or directory
collect2: ld returned 1 exit status

I figure that I am missing some portion of the development system, but it is not obvious to me which. Any help is appreciated. Thanks!

P.S. I apologize if this is the typical noob question that gets asked 20 times a day. I really did look at the website FAQ and the e-mail archives for this one and did not see it.
[config.log]

This file contains any messages produced by compilers while running configure, to aid debugging if configure makes a mistake. It was created by EM8300 configure 0.17.1-rc1, which was generated by GNU Autoconf 2.61. Invocation command line was $ ./configure ## --------- ## ## Platform. ## ## --------- ## hostname = Holly uname -m = i686 uname -r = 2.6.18-6-686 uname -s = Linux uname -v = #1 SMP Mon Aug 18 08:42:39 UTC 2008 /usr/bin/uname -p = unknown /bin/uname -X = unknown /bin/arch = i686 /usr/bin/arch -k = unknown /usr/convex/getsysinfo = unknown /usr/bin/hostinfo = unknown /bin/machine = unknown /usr/bin/oslevel = unknown /bin/universe = unknown PATH: /usr/local/bin PATH: /usr/bin PATH: /bin PATH: /usr/bin/X11 PATH: /usr/games ## ----------- ## ## Core tests. ## ## ----------- ## configure:1838: checking for gcc configure:1854: found /usr/bin/gcc configure:1865: result: gcc configure:2103: checking for C compiler version configure:2110: gcc --version >&5 gcc (GCC) 4.1.2 20061115 (prerelease) (Debian 4.1.1-21) Copyright (C) 2006 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. configure:2113: $? = 0 configure:2120: gcc -v >&5 Using built-in specs. Target: i486-linux-gnu Configured with: ../src/configure -v --enable-languages=c,c++,fortran,objc,obj-c++,treelang --prefix=/usr --enable-shared --with-system-zlib --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --enable-nls --program-suffix=-4.1 --enable-__cxa_atexit --enable-clocale=gnu --enable-libstdcxx-debug --enable-mpfr --with-tune=i686 --enable-checking=release i486-linux-gnu Thread model: posix gcc version 4.1.2 20061115 (prerelease) (Debian 4.1.1-21) configure:2123: $? = 0 configure:2130: gcc -V >&5 gcc: '-V' option must have argument configure:2133: $? = 1 configure:2156: checking for C compiler default output file name configure:2183: gcc conftest.c >&5 /usr/bin/ld: crt1.o: No such file: No such file or directory collect2: ld returned 1 exit status configure:2186: $? = 1 configure:2224: result: configure: failed program was: | /* confdefs.h. */ | #define PACKAGE_NAME "EM8300" | #define PACKAGE_TARNAME "em8300" | #define PACKAGE_VERSION "0.17.1-rc1" | #define PACKAGE_STRING "EM8300 0.17.1-rc1" | #define PACKAGE_BUGREPORT "" | /* end confdefs.h. */ | | int | main () | { | | ; | return 0; | } configure:2231: error: C compiler cannot create executables See `config.log' for more details. ## ---------------- ## ## Cache variables. ## ## ---------------- ## ac_cv_env_CC_set= ac_cv_env_CC_value= ac_cv_env_CFLAGS_set= ac_cv_env_CFLAGS_value= ac_cv_env_CPPFLAGS_set= ac_cv_env_CPPFLAGS_value= ac_cv_env_CPP_set= ac_cv_env_CPP_value= ac_cv_env_LDFLAGS_set= ac_cv_env_LDFLAGS_value= ac_cv_env_LIBS_set= ac_cv_env_LIBS_value= ac_cv_env_build_alias_set= ac_cv_env_build_alias_value= ac_cv_env_host_alias_set= ac_cv_env_host_alias_value= ac_cv_env_target_alias_set= ac_cv_env_target_alias_value= ac_cv_prog_ac_ct_CC=gcc ## ----------------- ## ## Output variables. ## ## ----------------- ## ACLOCAL='' ALSA_DIR='' AMDEPBACKSLASH='' AMDEP_FALSE='' AMDEP_TRUE='' AMTAR='' AUTOCONF='' AUTOHEADER='' AUTOMAKE='' AWK='' CC='gcc' CCDEPMODE='' CFLAGS='' CPP='' CPPFLAGS='' CYGPATH_W='' DEBUG_CFLAGS='' DEFS='' DEPDIR='' ECHO_C='' ECHO_N='-n' ECHO_T='' EGREP='' EXEEXT='' FIRMWARE_DIR='' GLOBAL_CFLAGS='' GREP='' GTK_CFLAGS='' GTK_LIBS='' HAVE_GTK_FALSE='' HAVE_GTK_TRUE='' INSTALL_DATA='' INSTALL_PROGRAM='' INSTALL_SCRIPT='' INSTALL_STRIP_PROGRAM='' LDFLAGS='' LIBOBJS='' LIBS='' LTLIBOBJS='' MAINT='' MAINTAINER_MODE_FALSE='' MAINTAINER_MODE_TRUE='' MAKEINFO='' OBJEXT='' PACKAGE='' PACKAGE_BUGREPORT='' PACKAGE_NAME='EM8300' PACKAGE_STRING='EM8300 0.17.1-rc1' PACKAGE_TARNAME='em8300' PACKAGE_VERSION='0.17.1-rc1' PATH_SEPARATOR=':' PKG_CONFIG='' SET_MAKE='' SHELL='/bin/sh' STRIP='' VERSION='' ac_ct_CC='gcc' am__fastdepCC_FALSE='' am__fastdepCC_TRUE='' am__include='' am__leading_dot='' am__quote='' am__tar='' am__untar='' bindir='${exec_prefix}/bin' build='' build_alias='' build_cpu='' build_os='' build_vendor='' datadir='${datarootdir}' datarootdir='${prefix}/share' docdir='${datarootdir}/doc/${PACKAGE_TARNAME}' dvidir='${docdir}' exec_prefix='NONE' host='' host_alias='' host_cpu='' host_os='' host_vendor='' htmldir='${docdir}' includedir='${prefix}/include' infodir='${datarootdir}/info' install_sh='' libdir='${exec_prefix}/lib' libexecdir='${exec_prefix}/libexec' localedir='${datarootdir}/locale' localstatedir='${prefix}/var' mandir='${datarootdir}/man' mkdir_p='' oldincludedir='/usr/include' pdfdir='${docdir}' prefix='NONE' program_transform_name='s,x,x,' psdir='${docdir}' sbindir='${exec_prefix}/sbin' sharedstatedir='${prefix}/com' sysconfdir='${prefix}/etc' target_alias='' ## ----------- ## ## confdefs.h. ## ## ----------- ## #define PACKAGE_NAME "EM8300" #define PACKAGE_TARNAME "em8300" #define PACKAGE_VERSION "0.17.1-rc1" #define PACKAGE_STRING "EM8300 0.17.1-rc1" #define PACKAGE_BUGREPORT "" configure: exit 77

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-09-26T01:52:14Z

* Nicolas Boullis <nboullis@...>:

| > And where disappeared the "Sender:" header?

| Uh!? What do you mean?

I have no Sender header in the list mails.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-25T16:43:32Z

Hi,

On Thu, Sep 25, 2008 at 08:22:49AM +0000, Christian Gmeiner wrote:
>
> how should we proceed? Fix all issues (remove wrong fw loading stuff,
> do the i2c stuff in an
> other way) and then you pull or shall I "reset" my repository and do
> better.... or you want
> to merge some of my patches?

Sorry for the long silence.
I just merged most of your changes.
As for 455d3227a288 ("rework firmware loading"), we already talked about
it.
As for ef34058b80f3 ("add a useful error message"), you are right that
this apparently is not a normal end condition (what I thought). But I
think it's quite useless to test this since there are *many* other
"strange" conditions where we could think the firmware is corrupted.
I also changed the description ec05b2763b8d ("fix compile error" -> "fix
compile warning").
And since I skipped the first change, I transplanted the remaining ones,
so their IDs changed.

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-09-25T16:31:12Z

Hi,

On Wed, Sep 24, 2008 at 08:09:22AM +0200, Gabor Z. Papp wrote:
>
> And where disappeared the "Sender:" header?

Uh!? What do you mean?

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-25T01:22:49Z

Hi Nicolas,

how should we proceed? Fix all issues (remove wrong fw loading stuff,
do the i2c stuff in an
other way) and then you pull or shall I "reset" my repository and do
better.... or you want
to merge some of my patches?

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-09-23T23:09:22Z

* Nicolas Boullis <nboullis@...>:

| > Interesing... but I can not find this message at
| > http://sourceforge.net/mailarchive/forum.php?forum_name=dxr3-devel

| I'm not that surprised; I have often seen the archives lagging behind.
| By the way, as for the messages I could not send, I fugured out it was a
| problem in the configuration of my e-mail server.

And where disappeared the "Sender:" header?

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-09-23T14:40:23Z

Hi,

On Tue, Sep 23, 2008 at 06:32:20AM +0000, Christian Gmeiner wrote:
> Interesing... but I can not find this message at
> http://sourceforge.net/mailarchive/forum.php?forum_name=dxr3-devel

I'm not that surprised; I have often seen the archives lagging behind.
By the way, as for the messages I could not send, I fugured out it was a
problem in the configuration of my e-mail server.

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-09-22T23:32:20Z

Interesing... but I can not find this message at
http://sourceforge.net/mailarchive/forum.php?forum_name=dxr3-devel

2008/9/22 Nicolas Boullis <nboullis@...>:

> On Mon, Sep 22, 2008 at 05:44:36AM +0000, Christian Gmeiner wrote:
>> Sorry for spam.... I only want to check if ml is still up and working.
>
> Hmmm... So it works for you... Does it work for me, now?
>
>
> Nicolas
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Dxr3-devel mailing list
> Dxr3-devel@...
> https://lists.sourceforge.net/lists/listinfo/dxr3-devel
>

--
Christian Gmeiner, B.Sc.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-22T14:35:18Z

Hi,

On Sat, Sep 20, 2008 at 03:35:43PM +0000, Christian Gmeiner wrote:

>
> Oh.. and the main repository seems to be broken:
>
> Traceback (most recent call last):
> File "/home/groups/d/dx/dxr3/cgi-bin/hgwebdir.cgi", line 21, in ?
> from mercurial.hgweb.hgwebdir_mod import hgwebdir
> File "/home/groups/d/dx/dxr3/lib/hg-0.9.4/mercurial/hgweb/__init__.py",
> line 9, in ?
> import hgweb_mod, hgwebdir_mod
> File "/home/groups/d/dx/dxr3/lib/hg-0.9.4/mercurial/hgweb/hgweb_mod.py",
> line 9, in ?
> import os, mimetypes, re, zlib, mimetools, cStringIO, sys
> File "/usr/lib64/python2.4/mimetools.py", line 6, in ?
> import tempfile
> File "/usr/lib64/python2.4/tempfile.py", line 33, in ?
> from random import Random as _Random
> File "/usr/lib64/python2.4/random.py", line 47, in ?
> from binascii import hexlify as _hexlify
> ImportError: /usr/lib64/python2.4/lib-dynload/binascii.so: failed to
> map segment from shared object: Cannot allocate memory

Hmmm, that's bad. I hope SourceForge did not break it on purpose.
Anyway, until I can fix it, I just set up a new public copy of my main
hg repository:
http://freehg.org/u/nboullis/em8300/

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Just a Test

2008-09-22T14:13:59Z

On Mon, Sep 22, 2008 at 05:44:36AM +0000, Christian Gmeiner wrote:
> Sorry for spam.... I only want to check if ml is still up and working.

Hmmm... So it works for you... Does it work for me, now?

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Just a Test

2008-09-21T22:44:36Z

Sorry for spam.... I only want to check if ml is still up and working.

--
Christian Gmeiner, B.Sc.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-20T08:35:43Z

Hi,

2008/9/19 Nicolas Boullis <nboullis@...>:

> Hi,
>
> On Thu, Sep 18, 2008 at 06:03:41AM +0000, Christian Gmeiner wrote:
>>
>> I think you have misunderstood me. Image following situation:
>> Kernel 2.6.26 with hg em8300 drivers. You have forgotten to add the
>> firmware to the correct place. You load your driver, and start vdr or mplayer.
>> Now... nothing is working, but in your log you find lot of lines saying your
>> firmware is missing and vdr/mplayer tries to write data or is working
>> with ioctls.
>>
>> And this is the point of this rework. If we are on a CONFIG_FW_LOADER system,
>> require the fw at module loading time, and not if we do an ioctl or
>> something else which
>> triggers the loading.
>> As you can see in the diff, I have added no code which was not used
>> before. I have moved the
>>
>> case _IOC_NR(EM8300_IOCTL_WRITEREG):
>> em8300_require_ucode(em);
>>
>> if (!em->ucodeloaded) {
>> return -ENOTTY;
>> }
>>
>> and all others, only into one em8300_probe function block.
>>
>> em8300_require_ucode(em);
>> if (!em->ucodeloaded) {
>> printk(KERN_ERR "em8300: Failed to load firmware\n");
>> result = -ENODEV;
>> goto firmware_error;
>> }
>>
>>
>> Keep in mind, that this does _only_ affect CONFIG_FW_LOADER based systems.
>> Hope you get the point now... if you do not trust me, try it by your
>> own and rename
>> microcode file and see what is happen.
>
> Sorry, but I still disagree quite strongly.
>
> Here is what happens, before your change, on a system without
> CONFIG_FW_LOADER:
> 1. the driver is loaded and the card is half-enabled;
> 2. the firmware is loaded with the em8300setup userspace application
> (or any other userspace application, as I think the dxr3 plugin for
> vdr loads it as well) as the corresponding ioctl is allowed even
> without a firmware;
> 3. the card is fully-enabled and everything's fine.
>
> Now, with your patch, on such a system, em8300_require_ucode is no-op.
> Then em8300_probe fails because nothing sets em->ucodeloaded, and the
> user is given no chance to load the firmware.
>
> Unless you prove me wrong, I won't accept such a change.

Ahhh... now I got it... shame on me.

>
> I can see several solutions:
> 1. Officially drop support for systems without CONFIG_FW_LOADER, but do
> it cleanly. I guess that means that the build should fail on such a
> system. This means drop support for pre-2.4.23 kernels. (This is not
> my prefered solution.)
> 2. Only require the firmware loading in em8300_probe on systems with
> CONFIG_FW_LOADER, but do not break anything on systems without it.
> 3. Leave things unchanged.
>

I would go for 3... I am fine with it.

>
>> > - As for ef34058b80f3 ("add a useful error message"), *if I remember
>> > correctly* flags==0 marks the end of the microcode and is expected,
>> > so I guess the error message is wrong.
>>
>> I have created a little test program to see what is going on at fw
>> loading, and as far as I
>> can see, flags == 0 does _not_ mark the end of the microcode. See
>> attachment and do
>> some own tests.
>
> OK, I'll check again and more thoroughly.
>
>
>> > - As for 09b3e93eac9f ("rework device memory allocation"), how do you
>> > get those figures?
>>
>> size em8300.ko
>
> OK, thanks.
>
>
>> >> Oh.. an I have a mistake in on commit summary: fix compile error
>> >> should be fix compile warning
>> >
>> > Can't you fix it?
>>
>> I will have a look at the mercurial docs..
>
> You can export a changeset, modify it, and reimport it. Note that it
> means it makes a new branch. Then do the same export/import game for the
> following changes (or use transplant).
>

Thanks, will try it.

>
>> > By the way, I think you have removed some changesets on freehg.org. How
>> > did you do this?
>>
>> I am not sure... but after each round of merge I reset my repository
>> and start at the same
>> level as the main repository.
>
> Hmmm... I did not know it was possible to reset a repository on
> freehg.org; how do you do that?
>

Its not like a reset, but is more like.. delete repository... create a
new one with
the same name and push your em8300 repository to it.

>> > 2) It is still somewhat broken because the EM8300 chip sometimes
>> > switches itself to 16:9 using the "old" way. I have not yet found a
>> > way to prevent it.
>>
>> What is the "old" way?
>
> Set (or clear) bit 4 of DICOM_TvOut to enable/disable "hardware"
> letterbox.
> When the EM8300 chip (or its firmware) sees an aspect ratio change, it
> enables or disables "hardware" letterbox mode.
> The new code makes no use of this feature; "hardware" letterbox mode is
> supposed to be always disabled, while letterbox and pillarbox are
> implemented in software by playing with
> DICOM_{Frame,Visible}{Top,Bottom,Left,Right}.
> Unfortunately, the hardware sometimes enables "hardware" letterbox mode
> anyway, which adds to the scaling implemented in software, and leads to
> en image vertically compressed.
>
> Note that even if you don't have an adv7170-based card, it might still
> be worth testing if you have a 16:9 TV set (to have pillarbox for 4:3
> broadcasts). But you won't of course have WSS.
>

I only own a 4:3 TV set... :(

Oh.. and the main repository seems to be broken:

Traceback (most recent call last):
File "/home/groups/d/dx/dxr3/cgi-bin/hgwebdir.cgi", line 21, in ?
from mercurial.hgweb.hgwebdir_mod import hgwebdir
File "/home/groups/d/dx/dxr3/lib/hg-0.9.4/mercurial/hgweb/__init__.py",
line 9, in ?
import hgweb_mod, hgwebdir_mod
File "/home/groups/d/dx/dxr3/lib/hg-0.9.4/mercurial/hgweb/hgweb_mod.py",
line 9, in ?
import os, mimetypes, re, zlib, mimetools, cStringIO, sys
File "/usr/lib64/python2.4/mimetools.py", line 6, in ?
import tempfile
File "/usr/lib64/python2.4/tempfile.py", line 33, in ?
from random import Random as _Random
File "/usr/lib64/python2.4/random.py", line 47, in ?
from binascii import hexlify as _hexlify
ImportError: /usr/lib64/python2.4/lib-dynload/binascii.so: failed to
map segment from shared object: Cannot allocate memory

cheers
--
Christian Gmeiner, B.Sc.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-18T17:23:02Z

Hi,

On Thu, Sep 18, 2008 at 06:03:41AM +0000, Christian Gmeiner wrote:

>
> I think you have misunderstood me. Image following situation:
> Kernel 2.6.26 with hg em8300 drivers. You have forgotten to add the
> firmware to the correct place. You load your driver, and start vdr or mplayer.
> Now... nothing is working, but in your log you find lot of lines saying your
> firmware is missing and vdr/mplayer tries to write data or is working
> with ioctls.
>
> And this is the point of this rework. If we are on a CONFIG_FW_LOADER system,
> require the fw at module loading time, and not if we do an ioctl or
> something else which
> triggers the loading.
> As you can see in the diff, I have added no code which was not used
> before. I have moved the
>
> case _IOC_NR(EM8300_IOCTL_WRITEREG):
> em8300_require_ucode(em);
>
> if (!em->ucodeloaded) {
> return -ENOTTY;
> }
>
> and all others, only into one em8300_probe function block.
>
> em8300_require_ucode(em);
> if (!em->ucodeloaded) {
> printk(KERN_ERR "em8300: Failed to load firmware\n");
> result = -ENODEV;
> goto firmware_error;
> }
>
>
> Keep in mind, that this does _only_ affect CONFIG_FW_LOADER based systems.
> Hope you get the point now... if you do not trust me, try it by your
> own and rename
> microcode file and see what is happen.

Sorry, but I still disagree quite strongly.

Here is what happens, before your change, on a system without
CONFIG_FW_LOADER:
1. the driver is loaded and the card is half-enabled;
2. the firmware is loaded with the em8300setup userspace application
(or any other userspace application, as I think the dxr3 plugin for
vdr loads it as well) as the corresponding ioctl is allowed even
without a firmware;
3. the card is fully-enabled and everything's fine.

Now, with your patch, on such a system, em8300_require_ucode is no-op.
Then em8300_probe fails because nothing sets em->ucodeloaded, and the
user is given no chance to load the firmware.

Unless you prove me wrong, I won't accept such a change.

I can see several solutions:
1. Officially drop support for systems without CONFIG_FW_LOADER, but do
it cleanly. I guess that means that the build should fail on such a
system. This means drop support for pre-2.4.23 kernels. (This is not
my prefered solution.)
2. Only require the firmware loading in em8300_probe on systems with
CONFIG_FW_LOADER, but do not break anything on systems without it.
3. Leave things unchanged.

As for mplayer (and possibly others): why should a sane application
insist if it fails to open /dev/em8300_{ma,mv,sp}? If it does, it's not
the driver's fault!

> > - As for ef34058b80f3 ("add a useful error message"), *if I remember
> > correctly* flags==0 marks the end of the microcode and is expected,
> > so I guess the error message is wrong.
>
> I have created a little test program to see what is going on at fw
> loading, and as far as I
> can see, flags == 0 does _not_ mark the end of the microcode. See
> attachment and do
> some own tests.

OK, I'll check again and more thoroughly.

> > - As for 09b3e93eac9f ("rework device memory allocation"), how do you
> > get those figures?
>
> size em8300.ko

OK, thanks.

> >> Oh.. an I have a mistake in on commit summary: fix compile error
> >> should be fix compile warning
> >
> > Can't you fix it?
>
> I will have a look at the mercurial docs..

You can export a changeset, modify it, and reimport it. Note that it
means it makes a new branch. Then do the same export/import game for the
following changes (or use transplant).

> > By the way, I think you have removed some changesets on freehg.org. How
> > did you do this?
>
> I am not sure... but after each round of merge I reset my repository
> and start at the same
> level as the main repository.

Hmmm... I did not know it was possible to reset a repository on
freehg.org; how do you do that?

> > 2) It is still somewhat broken because the EM8300 chip sometimes
> > switches itself to 16:9 using the "old" way. I have not yet found a
> > way to prevent it.
>
> What is the "old" way?

Set (or clear) bit 4 of DICOM_TvOut to enable/disable "hardware"
letterbox.
When the EM8300 chip (or its firmware) sees an aspect ratio change, it
enables or disables "hardware" letterbox mode.
The new code makes no use of this feature; "hardware" letterbox mode is
supposed to be always disabled, while letterbox and pillarbox are
implemented in software by playing with
DICOM_{Frame,Visible}{Top,Bottom,Left,Right}.
Unfortunately, the hardware sometimes enables "hardware" letterbox mode
anyway, which adds to the scaling implemented in software, and leads to
en image vertically compressed.

Note that even if you don't have an adv7170-based card, it might still
be worth testing if you have a 16:9 TV set (to have pillarbox for 4:3
broadcasts). But you won't of course have WSS.

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-18T04:41:57Z

* Nicolas Boullis <nboullis@...>:

| > | It'd be ok for me, if needed, to drop support for old 2.4 and old 2.6
| > | kernels.
| >
| > Please no, at least for 2.4.

| Sorry, I meant something like "dropping support for pre-2.4.26 2.4
| kernels and for pre-2.6.18 2.6 kernels", certainly not dropping support
| for all 2.4 kernels. Would you object this?

I'm using always the latest 2.4

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-17T23:03:41Z

Hi,

2008/9/17 Nicolas Boullis <nboullis@...>:

> Hi,
>
> On Wed, Sep 17, 2008 at 09:02:03AM +0000, Christian Gmeiner wrote:
>> Hello,
>>
>> >
>> > I did it and merged your other patches.
>> > (Edit: I've just seen you have new ones...)
>> >
>>
>> fine to see... yes there are some new patches which want to get reviewed :)
>> Here the interesting facts:
>> * reworked firmware loading
>> * rework device memory allocation
>> * some other cleanups
>
> Some comments already, after a quick look:
>
> - I think 455d3227a288 ("rework firmware loading") is plain wrong. On
> systems without CONFIG_FW_LOADER, em8300_require_ucode is no-op, so I
> think the driver would fail to setup the EM8300-based cards. I'd
> rather not stop supporting systems without CONFIG_FW_LOADER (which
> includes all pre-2.4.23 kernels).
> Moreover, if mplayer is broken, we don't have to fix it here.

I think you have misunderstood me. Image following situation:
Kernel 2.6.26 with hg em8300 drivers. You have forgotten to add the
firmware to the correct place. You load your driver, and start vdr or mplayer.
Now... nothing is working, but in your log you find lot of lines saying your
firmware is missing and vdr/mplayer tries to write data or is working
with ioctls.

And this is the point of this rework. If we are on a CONFIG_FW_LOADER system,
require the fw at module loading time, and not if we do an ioctl or
something else which
triggers the loading.
As you can see in the diff, I have added no code which was not used
before. I have moved the

case _IOC_NR(EM8300_IOCTL_WRITEREG):
em8300_require_ucode(em);

if (!em->ucodeloaded) {
return -ENOTTY;
}

and all others, only into one em8300_probe function block.

em8300_require_ucode(em);
if (!em->ucodeloaded) {
printk(KERN_ERR "em8300: Failed to load firmware\n");
result = -ENODEV;
goto firmware_error;
}

Keep in mind, that this does _only_ affect CONFIG_FW_LOADER based systems.
Hope you get the point now... if you do not trust me, try it by your
own and rename
microcode file and see what is happen.

>
> - As for ef34058b80f3 ("add a useful error message"), *if I remember
> correctly* flags==0 marks the end of the microcode and is expected,
> so I guess the error message is wrong.

I have created a little test program to see what is going on at fw
loading, and as far as I
can see, flags == 0 does _not_ mark the end of the microcode. See
attachment and do
some own tests.

> - As for 6059be5ad914 ("use a template for i2c_algo_bit_data"), I think
> em8300_i2c_algo_template should be made const. I guess I would also
> use simple assignment of the structure rather than memcpy (easier to
> read).
ack

> - As for ec05b2763b8d ("fix compile error"), I think I'd use
> KERN_WARNING rather than KERN_ERR.
ack

> - As for 09b3e93eac9f ("rework device memory allocation"), how do you
> get those figures?
>

size em8300.ko

>
>> Oh.. an I have a mistake in on commit summary: fix compile error
>> should be fix compile warning
>
> Can't you fix it?

I will have a look at the mercurial docs..

> By the way, I think you have removed some changesets on freehg.org. How
> did you do this?

I am not sure... but after each round of merge I reset my repository
and start at the same
level as the main repository.

>
>> Btw, whats about your http://freehg.org/u/nboullis/em8300-scale-n-crop/ ?
>
> 1) I've had very little feedback of people with the corresponding
> hardware (ADV7170-based card + TV set that understands WSS).

me has only adv7175 based cards..

> 2) It is still somewhat broken because the EM8300 chip sometimes
> switches itself to 16:9 using the "old" way. I have not yet found a
> way to prevent it.

What is the "old" way?

greets
--
Christian Gmeiner, B.Sc.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

main.cpp (3K) Download Attachment

From forum: dxr3-devel

Re: Some development questions

2008-09-17T14:56:50Z

Hi,

On Wed, Sep 17, 2008 at 09:02:03AM +0000, Christian Gmeiner wrote:

> Hello,
>
> >
> > I did it and merged your other patches.
> > (Edit: I've just seen you have new ones...)
> >
>
> fine to see... yes there are some new patches which want to get reviewed :)
> Here the interesting facts:
> * reworked firmware loading
> * rework device memory allocation
> * some other cleanups

Some comments already, after a quick look:

- I think 455d3227a288 ("rework firmware loading") is plain wrong. On
systems without CONFIG_FW_LOADER, em8300_require_ucode is no-op, so I
think the driver would fail to setup the EM8300-based cards. I'd
rather not stop supporting systems without CONFIG_FW_LOADER (which
includes all pre-2.4.23 kernels).
Moreover, if mplayer is broken, we don't have to fix it here.

- As for ef34058b80f3 ("add a useful error message"), *if I remember
correctly* flags==0 marks the end of the microcode and is expected,
so I guess the error message is wrong.

- As for 6059be5ad914 ("use a template for i2c_algo_bit_data"), I think
em8300_i2c_algo_template should be made const. I guess I would also
use simple assignment of the structure rather than memcpy (easier to
read).

- As for ec05b2763b8d ("fix compile error"), I think I'd use
KERN_WARNING rather than KERN_ERR.

- As for 09b3e93eac9f ("rework device memory allocation"), how do you
get those figures?

> Oh.. an I have a mistake in on commit summary: fix compile error
> should be fix compile warning

Can't you fix it?
By the way, I think you have removed some changesets on freehg.org. How
did you do this?

> Btw, whats about your http://freehg.org/u/nboullis/em8300-scale-n-crop/ ?

1) I've had very little feedback of people with the corresponding
hardware (ADV7170-based card + TV set that understands WSS).

2) It is still somewhat broken because the EM8300 chip sometimes
switches itself to 16:9 using the "old" way. I have not yet found a
way to prevent it.

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-17T14:32:42Z

On Wed, Sep 17, 2008 at 09:54:14AM +0200, Gabor Z. Papp wrote:
>
> | It'd be ok for me, if needed, to drop support for old 2.4 and old 2.6
> | kernels.
>
> Please no, at least for 2.4.

Sorry, I meant something like "dropping support for pre-2.4.26 2.4
kernels and for pre-2.6.18 2.6 kernels", certainly not dropping support
for all 2.4 kernels. Would you object this?

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-17T02:02:03Z

Hello,

>
> I did it and merged your other patches.
> (Edit: I've just seen you have new ones...)
>

fine to see... yes there are some new patches which want to get reviewed :)
Here the interesting facts:
* reworked firmware loading
* rework device memory allocation
* some other cleanups

Oh.. an I have a mistake in on commit summary: fix compile error
should be fix compile warning

>
>> By the way... which kernel version do we support?
>
> As far as I am concerned, I'm willing to support all 2.4 and 2.6
> kernels.
> I know most developers don't care about 2.4 any more, but:
> * I know at least one person who uses the em8300 drivers with a 2.4
> kernel;
> * I hate the way the 2.6 kernel is managed (unstable API, invasive
> changes, ...), so I understand why one may prefer to keep the
> rock-solid 2.4 branch.
>
> It'd be ok for me, if needed, to drop support for old 2.4 and old 2.6
> kernels.
>

I will drop support for 2.4 and old 2.6 if v4l people - looking at
Hans - have done some
needed changes. And then I will start a new repository for it.

Btw, whats about your http://freehg.org/u/nboullis/em8300-scale-n-crop/ ?

have a nice day
--
Christian Gmeiner, B.Sc.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-17T01:54:45Z

Hi

2008/9/17 Gabor Z. Papp <gzp@...>:

> * Nicolas Boullis <nboullis@...>:
>
> | I know most developers don't care about 2.4 any more, but:
> | * I know at least one person who uses the em8300 drivers with a 2.4
> | kernel;
>
> /me yelling :-)
>
> Thanks Nicolas.
>
> | * I hate the way the 2.6 kernel is managed (unstable API, invasive
> | changes, ...), so I understand why one may prefer to keep the
> | rock-solid 2.4 branch.
>
> My dxr3 system works like a charm since many years with 2.4, using P3
> and sata disks, lirc (note the current lirc madness on 2.6...)
> 2.6.25.x releases were the first ones that ran mostly stable on the
> same hardware where 2.4 not fails. So probably I could switch, but why...

Never change a running system :)

>
> | It'd be ok for me, if needed, to drop support for old 2.4 and old 2.6
> | kernels.
>
> Please no, at least for 2.4. If you guys send me patches for testing
> or ask me to test the svn/cvs/anything, I would report
> success/problems on 2.4. I'm silently following the current
> development here and using always the latest release/rc.

That is fine to hear that someone is testing the driver with a 2.4 kernel. I
want to say that in the main tree, I will not remove 2.4 and old 2.6 stuff,
but around the end of this year I will start a repository for kernel inclusion.
So in this new repository 2.4 and old 2.6 support will gets removed, but I
will announce this step here at the ml.

have a nice day
--
Christian Gmeiner, B.Sc.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-17T00:54:14Z

* Nicolas Boullis <nboullis@...>:

| I know most developers don't care about 2.4 any more, but:
| * I know at least one person who uses the em8300 drivers with a 2.4
| kernel;

/me yelling :-)

Thanks Nicolas.

| * I hate the way the 2.6 kernel is managed (unstable API, invasive
| changes, ...), so I understand why one may prefer to keep the
| rock-solid 2.4 branch.

My dxr3 system works like a charm since many years with 2.4, using P3
and sata disks, lirc (note the current lirc madness on 2.6...)
2.6.25.x releases were the first ones that ran mostly stable on the
same hardware where 2.4 not fails. So probably I could switch, but why...

| It'd be ok for me, if needed, to drop support for old 2.4 and old 2.6
| kernels.

Please no, at least for 2.4. If you guys send me patches for testing
or ask me to test the svn/cvs/anything, I would report
success/problems on 2.4. I'm silently following the current
development here and using always the latest release/rc.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-16T14:50:18Z

Hi,

On Mon, Sep 15, 2008 at 06:00:14AM +0000, Christian Gmeiner wrote:

>
> If you want I will explain it to you, but I think my linux driver
> programming book is too old :( I
> had a quick look at drivers/pci/pci.c and found out, that all those
> check, if our device supports
> busmastering are not needed, as pci_set_master does all the hard work.
> So.. I have ordered
> a fresh book.
>
> So please change the whole block to:
>
>
> static int em8300_pci_setup(struct pci_dev *dev)
> {
> struct em8300_s *em = pci_get_drvdata(dev);
> unsigned char revision;
> int rc = 0;
> u16 cmd;
>
> rc = pci_enable_device(dev);
> if (rc < 0) {
> printk(KERN_ERR "em8300: Unable to enable PCI device\n");
> return rc;
> }
>
> pci_set_master(dev);
>
> em->adr = pci_resource_start(dev, 0);
> em->memsize = pci_resource_len(dev, 0);
>
> pci_read_config_byte(dev, PCI_CLASS_REVISION, &revision);
> em->pci_revision = revision;
>
> return 0;
> }

I did it and merged your other patches.
(Edit: I've just seen you have new ones...)

> I am sorry for this trouble, but I have learned much from it and hope
> such situations will
> never happen again, as I will have a look at the sources of the kernel
> if I am not sure.

It's no problem.

> By the way... which kernel version do we support?

As far as I am concerned, I'm willing to support all 2.4 and 2.6
kernels.
I know most developers don't care about 2.4 any more, but:
* I know at least one person who uses the em8300 drivers with a 2.4
kernel;
* I hate the way the 2.6 kernel is managed (unstable API, invasive
changes, ...), so I understand why one may prefer to keep the
rock-solid 2.4 branch.

It'd be ok for me, if needed, to drop support for old 2.4 and old 2.6
kernels.

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-14T23:00:14Z

Hi Nicolas,

2008/9/14 Nicolas Boullis <nboullis@...>:

> Hi Christian,
>
> On Fri, Sep 12, 2008 at 02:14:56AM +0200, Nicolas Boullis wrote:
>> Hi,
>>
>> On Wed, Sep 10, 2008 at 08:31:49AM +0000, Christian Gmeiner wrote:
>> >
>> > Fine... more stuff is coming...
>>
>> I just did some more merges.
>> Note that the patches are not rejected, I just have not had time yet to
>> review them (especially
>> http://freehg.org/u/austriancoder/em8300-cgmeiner/rev/452d407a9ddc which
>> is non-trivial).
>
> I just had a more thorough look at 452d407a9ddc and I fal to understand
> why you do all this stuff around the pci_set_master call (check for
> PCI_COMMAND_MASTER both before and after).
> As far as U can see, most drivers that use pci_set_master don't do such
> things (although at least drivers/ide/setup-pci.c and
> drivers/media/video/ivtv/ivtv-driver.c do).
> Do you have any explanation?
>

If you want I will explain it to you, but I think my linux driver
programming book is too old :( I
had a quick look at drivers/pci/pci.c and found out, that all those
check, if our device supports
busmastering are not needed, as pci_set_master does all the hard work.
So.. I have ordered
a fresh book.

So please change the whole block to:

static int em8300_pci_setup(struct pci_dev *dev)
{
struct em8300_s *em = pci_get_drvdata(dev);
unsigned char revision;
int rc = 0;
u16 cmd;

rc = pci_enable_device(dev);
if (rc < 0) {
printk(KERN_ERR "em8300: Unable to enable PCI device\n");
return rc;
}

pci_set_master(dev);

em->adr = pci_resource_start(dev, 0);
em->memsize = pci_resource_len(dev, 0);

pci_read_config_byte(dev, PCI_CLASS_REVISION, &revision);
em->pci_revision = revision;

return 0;
}

I am sorry for this trouble, but I have learned much from it and hope
such situations will
never happen again, as I will have a look at the sources of the kernel
if I am not sure.

By the way... which kernel version do we support?

thanks
--
Christian Gmeiner, B.Sc.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel

Re: Some development questions

2008-09-14T14:17:05Z

Hi Christian,

On Fri, Sep 12, 2008 at 02:14:56AM +0200, Nicolas Boullis wrote:

> Hi,
>
> On Wed, Sep 10, 2008 at 08:31:49AM +0000, Christian Gmeiner wrote:
> >
> > Fine... more stuff is coming...
>
> I just did some more merges.
> Note that the patches are not rejected, I just have not had time yet to
> review them (especially
> http://freehg.org/u/austriancoder/em8300-cgmeiner/rev/452d407a9ddc which
> is non-trivial).

I just had a more thorough look at 452d407a9ddc and I fal to understand
why you do all this stuff around the pci_set_master call (check for
PCI_COMMAND_MASTER both before and after).
As far as U can see, most drivers that use pci_set_master don't do such
things (although at least drivers/ide/setup-pci.c and
drivers/media/video/ivtv/ivtv-driver.c do).
Do you have any explanation?

Cheers,

Nicolas

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Dxr3-devel mailing list
Dxr3-devel@...
https://lists.sourceforge.net/lists/listinfo/dxr3-devel

From forum: dxr3-devel