I don't suspect there will be major changes in the ir-driver for 1.03 final, so here is a new version for 1.03rc1.
Since the IR-driver has changed, all IR-codes have changed also. I've setup a table for the kenwood and rio remote :

                 Kenwood old  Kenwood new     Rio new

-----------------------------------------------------

IR_0             0x0000B900 | 0x00B94600 | 0x0020DF0D

IR_1             0x0000B901 | 0x00B94601 | 0x0020DF00

IR_2             0x0000B902 | 0x00B94602 | 0x0020DF01

IR_3             0x0000B903 | 0x00B94603 | 0x0020DF02

IR_4             0x0000B904 | 0x00B94604 | 0x0020DF04

IR_5             0x0000B905 | 0x00B94605 | 0x0020DF05

IR_6             0x0000B906 | 0x00B94606 | 0x0020DF06

IR_7             0x0000B907 | 0x00B94607 | 0x0020DF08

IR_8             0x0000B908 | 0x00B94608 | 0x0020DF09

IR_9             0x0000B909 | 0x00B94609 | 0x0020DF0A

IR_TRACK_MINUS   0x0000B90A | 0x00B9460A | 0x0020DF10

IR_TRACK_PLUS    0x0000B90B | 0x00B9460B | 0x0020DF11

IR_AM_INFO       0x0000B90C | 0x00B9460C | 0x0020DF14

IR_FM_VISUAL     0x0000B90D | 0x00B9460D | 0x0020DF15

IR_PROG_HUSH     0x0000B90E | 0x00B9460E | 0x0020DF16

IR_DIRECT_SEARCH 0x0000B90F | 0x00B9460F | 0x0020DF0E

IR_VOL_PLUS      0x0000B914 | 0x00B94614 | 0x0020DF13

IR_VOL_MINUS     0x0000B915 | 0x00B94615 | 0x0020DF17

IR_STAR_CANCEL   0x0000B91B | 0x00B9461B | 0x0020DF0C

IR_TUNER_SRC     0x0000B91C | 0x00B9461C | 0x0020DF03

IR_TAPE_TUNER    0x0000B91D | 0x00B9461D | 0x0020DF07

IR_CD_MODE       0x0000B91E | 0x00B9461E | 0x0020DF0B

IR_CDMDCH_SOUND  0x0000B91F | 0x00B9461F | 0x0020DF0F

IR_DNPP_MENU     0x0000B95E | 0x00B9465E | 0x0020DF12

If someone would like to add the new codes for the pioneer remotes, I can create some kernels with correct mappings.
You do not need trace_ir for this; I've added a way to display them on serial when the player reads the codes.
type the following in the shell to enable verbose IR-reporting :

# echo '#1' > /proc/empeg_ir

to disable it :

# echo '#0' > /proc/empeg_ir

Before any codes are shown, there must first be a listener on /dev/ir. just quit the shell to restart the player-app to accomplish this.
You can also type :

# cat /dev/ir > /dev/null

IR-code tables should also be catted to /proc/empeg_ir instead of /dev/ir
You can show the currently loaded translation table by typing :

# cat /proc/empeg_ir

displayserver hack is also in there. works the same as usual...

kernel can be found
here

Frank van Gestel

thanks frank. So have you setup the default mapping as you had it previously? was that the problem, that the ir stuff had changed?

thanks for your time and help, i am an idiot in unix..so you lost me a bit, especially since i am new to the empeg too.

oopsie, forbidden message when trying to get the new kernel. :(

no not yet. I don't know the new codes for the pioneer remote. I'm waiting for someone to install the kernel on a developer image and read the codes from the remote... Once people send me the codes I will recompile and put two new images on the site for download.
let's try linking again : http://empeg.dyndns.org/empeg/ir_irans_kernel/1.03rc1
or like this : ir_trans_kernel

Frank van Gestel

getting there now.

so i need to install the image, then follow the directions for creating the button mapping, if i am using the 77 model. Gotcha, i will try to stumble through. I can try to send you the codes, i will have a look and see.

You mentioned display server is also patched into this conrrect? so i can do it all from my main pc.

Sorry, i don't mean to sound like such an idiot, i am trying :)

i have the logging enabled, and i can see the key presses, but how do i get the printout list of the commands as they are sent? is it just a txt file i need to d/l from teh empeg?

the output of the logging should look something like this :

0x00B94600 -> 0x00B94600

0x80B94600 -> 0x80B94600

Above output is shown when pressing zero on the kenwood remote
the first line is the button down code, the second the button up.
what you should do is write down all button down codes and fill them in, in the following table :

                               old code     new code

-------------------------------------------------------

| IR_PIONEER_SR70_ATT       | 0x0000AD0C |            |

| IR_PIONEER_SR70_SRC       | 0x0000AD1A |            |

| IR_PIONEER_SR70_CD        | 0x0000AD1E |            |

| IR_PIONEER_SR70_UP        | 0x0000AD40 |            |

| IR_PIONEER_SR70_DOWN      | 0x0000AD41 |            |

| IR_PIONEER_SR70_LEFT      | 0x0000AD42 |            |

| IR_PIONEER_SR70_RIGHT     | 0x0000AD43 |            |

| IR_PIONEER_SR70_BOTTOM    | 0x0000AD12 |            |

| IR_PIONEER_SR70_VOL_PLUS  | 0x0000AD0A |            |

| IR_PIONEER_SR70_VOL_MINUS | 0x0000AD0B |            |

|---------------------------|------------|------------|

| IR_PIONEER_SR77_ATT       | 0x0000AD0C |            |

| IR_PIONEER_SR77_BAND      | 0x0000AD12 |            |

| IR_PIONEER_SR77_SRC       | 0x0000AD1A |            |

| IR_PIONEER_SR77_UP        | 0x0000AD40 |            |

| IR_PIONEER_SR77_DOWN      | 0x0000AD41 |            |

| IR_PIONEER_SR77_LEFT      | 0x0000AD42 |            |

| IR_PIONEER_SR77_RIGHT     | 0x0000AD43 |            |

| IR_PIONEER_SR77_FUNC      | 0x0000AD19 |            |

| IR_PIONEER_SR77_AUDIO     | 0x0000AD0D |            |

| IR_PIONEER_SR77_VOL_PLUS  | 0x0000AD0A |            |

| IR_PIONEER_SR77_VOL_MINUS | 0x0000AD0B |            |

-------------------------------------------------------

to try your mappings, create a file using notepad on your PC like this :

# required line for IR-translation table

0x[code for volume up]:0x00B94614

0x[code for volume down]:0x00B94615

mount the root disk of the empeg read-write by using command:

# rw

# cd /tmp

use "Send file" in hyperterm to send the file you just created.
use command to make mapping active :

# cat [filename of created file] > /proc/empeg_ir

# ro

type the following to check if the mappings are loaded:

# cat /proc/empeg_ir

this should give the following output :

Valid sequences:      0

Repeated sequences:   0

Unfulfilled repeats:  0

Malformed sequences:  0

Spurious transitions: 0

Missed interrupts:    231

Timings buffer hwm:   64



Infrared Translation Table

 HEX            DEC

---------------------------

0x[code for volume up] -> 0x00B94614

0x[code for volume down] -> 0x00B94615

type exit on the prompt to exit the shell and restart the player.
the volume-up and down buttons of the pioneer remote should work now...
NOTE: You will lose the IR-mappings when you cycle power...


Frank van Gestel

ok, when i enable logging, then what is next? the cat command? when i do that i get no translation table....where am i supposed to view teh output of a keypress? I can see that it is being absorbed, but i cannot see the mapping numbers?

Edited by rockstar on 18/06/01 12:21 PM.

am i supposed to be viewing this through displayserver?

i guess what i am saying is, i follow what to do with the coded.. i just don't know how to get the codes. I have turned on logging, but i am stuck on waht to do from there, which is why i am guessing displayserver?

Edited by rockstar on 18/06/01 12:24 PM.

You've got to use hyperterm; type 'q' to get to the shell. Codes are shown in hyperterm. You do not need displayserver at all.
checklist :
- connect empeg to PC with serial cable
- start hyperterm on PC
- reboot empeg
- after boot-messages in hyperterm press 'q' to get to a shell-prompt
- type :

# echo '#1' > /proc/empeg_ir

# exit

- the player should restart
- press some buttons on the remote or one of the front-panel buttons of the empeg
- the following should be shown in hyperterm when pushing the front-panel buttons 'right' and then 'left':

Shell exit

Starting player

empeg-car 1.03beta6.

2 -> 2

3 -> 3

3 -> 3

4 -> 4

5 -> 5

5 -> 5

the output means :
[code received by IR-driver] -> [code sent to player]

To use the codes in a map file, you should notate them with 0x in front, so 2 would be 0x00000002


Frank van Gestel

ahh, i didn't realize i needed to enable it, restart and then NOT enter shell again, sorry for my idiocy.


Earlier i had an eff up, for some reason my empeg wouldn't get past my boot logo.. i reinstalled the consumer 1.03 rc1, then i reinstalled 1.03 dev rc1

then i reinstalled your dev rc1

thing is, it was strange, it wouldn't let me overwrite teh kernel,, everything showed successful, but it took the use of the consumer 1.03 rc1 to get it back to normal.. main question is, how do i now verify that this is indeed your kernel on here, is it annotated in the boot sequence anywhere.

sorry, i know, another new b ?

Edited by rockstar on 18/06/01 12:52 PM.

12 years later and after feeling like an idiot, here are the new codes.


IR_PIONEER_SR77_ATT 0x0000AD520C
IR_PIONEER_SR77_BAND 0x0000AD5212
IR_PIONEER_SR77_SRC 0x0000AD521A
IR_PIONEER_SR77_UP 0x0000AD5240
IR_PIONEER_SR77_DOWN 0x0000AD5241
IR_PIONEER_SR77_LEFT 0x0000AD5242
IR_PIONEER_SR77_RIGHT 0x0000AD5243
IR_PIONEER_SR77_FUNC 0x0000AF5067
IR_PIONEER_SR77_AUDIO 0x0000AD520D
IR_PIONEER_SR77_VOL_PLUS 0x0000AD520A
IR_PIONEER_SR77_VOL_MINUS 0x0000AD520B

I've put two new kernels on my site. Try them to see if they work...

Frank van Gestel

77 WORKS LIK A CHAMP!!

Thanks for doing that, and for dealing with my lack of knowledge, it seems to be working PERFECTLY, with just a PERFECT delay. You are the man frank!

Hey Frank... Do you have a patch for the 1.03 irtrans with variable repeat timers? The file http://empeg.dyndns.org/empeg/ir_trans_kernel/1.03rc1/irtrans1.03rc1.patch doesn't seem to have this functionality in it. I wrote my own repeat timer modification to the IRtrans in previous versions, but I'm not familiar enough with the new IR kernel code to pull it off, and since you've already figured it out... I know you have the two kernels for the SR70 and SR77 with the repeat timers, but I couldn't locate a patch for it.

Thanks.

-Tony
MkII #554

The timer thing shouldn't be necessary. The rc1 kernel will only send the first button-down code it receives and waits for a button-up code before accepting a new one. This way the repeat-rate is the same as the empeg/rio remotes, which should suffice I guess. If you still want the patch for the timer thing, I can try to supply a patch.

Frank van Gestel

Well, what I'm trying to do is have a different repeat rate for my FF/Rewind keys than the rate for my other keys. The FF/Rew goes in 10 second increments with the new code. I had it working with modifications to your previous IR trans kernels such that searches were in one second increments but were still quick due to a high repeat rate. I know it's picky... If you don't have the timer code lying around anymore then don't worry about it.

-Tony
MkII #554

attached is a file which has the code in it.

Frank van Gestel

Cool, exactly what I was looking for. Thanks!

-Tony
MkII #554

I've just bought a Pioneer remote: CD-SR80. I've looked at all the other messages, assigned the IR codes for the buttons and it works fine ... until I cycle the power and then the remote stops working. How do I get it to keep the codes?

___________________________

Steve - 18GB Mk II Sssmokey

Uh, guess I should add some more information. I'm using the developer 1.03 and have applied the clock, irhack, voladj, 1drive kernel from Drakino's(?) page. I've re-read the posts in this thread and am more than a little confused as to how I get the empeg to use the new codes that I have identified. Do I have to put them in a file that is read at boot time ( I take it /proc/empeg_ir gets refreshed ) or do I need to compile a kernel with them in?

While on the subject, perhaps someone could explain this: I thought if you cat a file it merely displays the contents on screen. What I don't understand is that if I cat my translation file to /proc/empeg_ir then when I cat /proc/empeg_ir I would expect it to only display the contents of this file which should be a replica of the original file as I am replacing the file and not appending to it. However, a cat /proc/empeg_ir reveals extra header information. Where does this come from?

___________________________

Steve - 18GB Mk II Sssmokey

While on the subject, perhaps someone could explain this: I thought if you cat a file it merely displays the contents on screen. What I don't understand is that if I cat my translation file to /proc/empeg_ir then when I cat /proc/empeg_ir I would expect it to only display the contents of this file which should be a replica of the original file as I am replacing the file and not appending to it. However, a cat /proc/empeg_ir reveals extra header information. Where does this come from?

Somebody else will probably be able to provide details, but raughly speaking /proc filesystem is magical: files there are not 'real', but serve as interface to various drivers, 'window' into various kernel tables etc. So, each file has its own particular behaviour and meaning, both when written to and read from.


Dragi "Bonzi" Raos
Zagreb, Croatia
Q#5196, MkII#80000376, 18GB green

While on the subject, perhaps someone could explain this: I thought if you cat a file it merely displays the contents on screen. What I don't
understand is that if I cat my translation file to /proc/empeg_ir then when I cat /proc/empeg_ir I would expect it to only display the contents of this
file which should be a replica of the original file as I am replacing the file and not appending to it. However, a cat /proc/empeg_ir reveals extra
header information. Where does this come from?

Somebody else will probably be able to provide details, but raughly speaking /proc filesystem is magical: files there are not 'real', but serve as
interface to various drivers, 'window' into various kernel tables etc. So, each file has its own particular behaviour and meaning, both when written to
and read from.

That's right.With a default kernel, the /proc/empeg_ir entry is displaying the "header" information. I just added the translation-table at the end and wrote a function that reads from /proc/empeg_ir for input.
To set a custom translation-table, you should add a line in your init script which cats a file containing the translation-table to /proc/empeg_ir :

cat /tmp/ir_transtable > /proc/empeg_ir

Frank van Gestel

[sigh] So, here I am again, cap in hand, asking for help [/sigh]

I tried this 'oh so simple' task of just adding a line to my init file but it was too difficult for me.

I echoed the cat command to append to the init file, which worked but a reboot failed to activate the codes.

As I have DisplayServer installed and I noticed that it calls the original init file I thought that I would try to put the cat command before the DisplayServer calls. I couldn't do this on the empeg so I copied the parts of the file to notepad, inserted my cat command and sent the file to the empeg. Unfortunately, I hadn't noticed the message that popped up when I set the unit to rw that made mention of having to run some 2fsck command. I rebooted my empeg and could only get up to the following prompt:

init-2.03#

Cursing my incompetence I ended up reinstalling a new image and adding the required kernel hacks. To be sure, I entered emplode, deleted a track and synched which took quite a while. I figured it was stupid trying to add the cat command before the original init gets loaded cuz the IR codes probably get overwritten again. So, I'm back to an init file that likes look this:

#!/bin/bash
/bin/mount -n -o nocheck,rw /proc
/bin/mount -n -o nocheck,ro /dev/hda4 /drive0
[ -e /proc/ide/hdb ] && /bin/mount -n -o nocheck,ro /dev/hdc4 /drive1

cd /usr/local/displayserver
./displayserver
exec /sbin/init.before_displayserver
cat /tmp/CDSR80_IR.txt > /proc/empeg_ir

Should I be adding something to this file or to the original init file that gets called from it? Not something I think I really want to do after looking at it.

___________________________

Steve - 18GB Mk II Sssmokey

the cat command should be placed after the mount-commands and before execing the original init.
one way to do it is :

# rw

# echo'#!/bin/bash' > /sbin/init.new

# echo '/bin/mount -n -o nocheck,rw /proc' >> /sbin/init.new 

# echo '/bin/mount -n -o nocheck,ro /dev/hda4 /drive0' >> /sbin/init.new 

# echo '[ -e /proc/ide/hdb ] && /bin/mount -n -o nocheck,ro /dev/hdc4 /drive1 ' >> /sbin/init.new

# echo 'cat /tmp/CDSR80_IR.txt > /proc/empeg_ir ' >> /sbin/init.new

# echo 'cd /usr/local/displayserver' >> /sbin/init.new 

# echo './displayserver' >> /sbin/init.new 

# echo 'exec /sbin/init.before_displayserver' >> /sbin/init.new 

# chmod +x /sbin/init.new

# ro

now test your new init :

# /sbin/init.new

this will give an error message if displayserver is already loaded, but your mappings should be read and the player started.
If you're sure the new init works, stop the player app again and replace the old init with your new init:

# rw

# mv /sbin/init.new /sbin/init

# ro

and reboot your empeg.


Frank van Gestel

Many thanks Frank. That's where I had tried to put the command before everything went horribly wrong. I'll try it again tonight when I get home.

___________________________

Steve - 18GB Mk II Sssmokey

'tis done ... finally! Although, should I worry about the following error which I get when I set my unit to rw:

EXT2-fs warning: mounting unchecked fs, running e2fsck is recommended

I got this last time I tried to edit the init file, but it went away after a re-install of the 1.03 software. I have made some amendments on my playlists and synched but it doesn't go away.

___________________________

Steve - 18GB Mk II Sssmokey

This error means you didn't mount the disks read-only after you were done messing with them. Gotta remember to do that.

Do a search on "e2fsck" to look up how to deal with it.

___________
Tony Fabris

This error means you didn't mount the disks read-only after you were done messing with them. Gotta remember to do that.

I thought that this happened automatically when the power was recycled. Once again I am proved wrong.

Do a search on "e2fsck" to look up how to deal with it.

Yes, I really must get into the habit of doing this first. The trouble is, we all know that a) any new message is flagged automatically to your desktop which you never leave and b) you'll have the answer straight away.


___________________________

Steve - 18GB Mk II Sssmokey

HEY! That's NOT FAIR!!!!

___________
Tony Fabris