Hi,
I have this hardware combination : the new Logitech MX900, which is
talking bluetooth to its hub. This hub goes to my laptop using USB.
I'm using gentoo linux, with kernel 2.6.0test11. I have patched my
XFree86 (xfree-4.3.99.16.ebuild) with
http://www.yhbt.net/normalperson/files/linux/LogitechMX_kernel_and_X_patch/
to make it understand the PS2++ protocol, and the mouse somewhat works.
The setup is working fine, even if I didn't finish the setup completely
(all buttons aren't working yet).
Now, having a internal bluetooth module within my Dell Inspiron 8600, I
would like to connect the mouse directly, without going through the hub
and USB.
When I use the basic tools (hcitool scan, hcitool info), I get this info
block from the mx900 :
==============
Requesting information ...
BD Address: 00:07:61:08:7E:2E
Device Name: Logitech MX900 Mouse
LMP Version: 1.1 (0x1) LMP Subversion: 0x350
Manufacturer: Cambridge Silicon Radio (10)
Features: 0xfc 0xff 0x0f 0x00
<encryption> <slot offset> <timing accuracy> <role switch> <hold mode>
<sniff mode> <park mode> <RSSI> <channel quality> <SCO link> <HV2
packets> <HV3 packets> <u-law log> <A-law log> <CVSD> <paging scheme>
<power control> <transparent SCO>
==============
I now understand that I need some piece of software to get the mouse
events from the bluez stack and pass them to a device that X would
understand.
I tried to compile bthid from Peter Klauser. bthid itself compile, but
fake.o choke on line 87 of fake.c :
====================
make -f make.fake fake.o
make[1]: Entering directory `/data/util/bthid-0.8'
gcc -pipe -Iinclude/asm-i386/mach-default -DMODULE -DMODVERSIONS
-include /usr/src/linux-2.6.0-test11-gentoo/include/config/modversions.h
-D__KERNEL__ -DMAJORNUM=250 -I/usr/src/linux-2.6.0-test11-gentoo/include
-fno-strict-aliasing -fno-common -c -o fake.o fake.c
fake.c:86: parse error before "devfs_handle"
fake.c:86: warning: data definition has no type or storage class
make[1]: *** [fake.o] Error 1
make[1]: Leaving directory `/data/util/bthid-0.8'
make: *** [fake.o] Error 2
====================
hints welcome. I can test whatewer piece of code you throw at me. I'm
however not knowledeable enough to code it myself.
Charles
--
Charles Bueche <[email protected]>
sand, snow, wave, wind and net -surfer
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi again,
responding to my own email after studying fake.c : I got it to compile
on 2.6, but this code isn't able to run as module on 2.6 kernels, which
changed the module format.
Peter : do you plan a 2.6 version anytime soon ?
TIA,
Charles
On Sun, 2003-11-30 at 15:12, Charles Bueche wrote:
> Hi,
>
> I have this hardware combination : the new Logitech MX900, which is
> talking bluetooth to its hub. This hub goes to my laptop using USB.
>
> I'm using gentoo linux, with kernel 2.6.0test11. I have patched my
> XFree86 (xfree-4.3.99.16.ebuild) with
> http://www.yhbt.net/normalperson/files/linux/LogitechMX_kernel_and_X_patch/
> to make it understand the PS2++ protocol, and the mouse somewhat works.
>
> The setup is working fine, even if I didn't finish the setup completely
> (all buttons aren't working yet).
>
> Now, having a internal bluetooth module within my Dell Inspiron 8600, I
> would like to connect the mouse directly, without going through the hub
> and USB.
>
> When I use the basic tools (hcitool scan, hcitool info), I get this info
> block from the mx900 :
>
> ==============
> Requesting information ...
> BD Address: 00:07:61:08:7E:2E
> Device Name: Logitech MX900 Mouse
> LMP Version: 1.1 (0x1) LMP Subversion: 0x350
> Manufacturer: Cambridge Silicon Radio (10)
> Features: 0xfc 0xff 0x0f 0x00
> <encryption> <slot offset> <timing accuracy> <role switch> <hold mode>
> <sniff mode> <park mode> <RSSI> <channel quality> <SCO link> <HV2
> packets> <HV3 packets> <u-law log> <A-law log> <CVSD> <paging scheme>
> <power control> <transparent SCO>
> ==============
>
> I now understand that I need some piece of software to get the mouse
> events from the bluez stack and pass them to a device that X would
> understand.
>
> I tried to compile bthid from Peter Klauser. bthid itself compile, but
> fake.o choke on line 87 of fake.c :
>
> ====================
> make -f make.fake fake.o
> make[1]: Entering directory `/data/util/bthid-0.8'
> gcc -pipe -Iinclude/asm-i386/mach-default -DMODULE -DMODVERSIONS
> -include /usr/src/linux-2.6.0-test11-gentoo/include/config/modversions.h
> -D__KERNEL__ -DMAJORNUM=250 -I/usr/src/linux-2.6.0-test11-gentoo/include
> -fno-strict-aliasing -fno-common -c -o fake.o fake.c
> fake.c:86: parse error before "devfs_handle"
> fake.c:86: warning: data definition has no type or storage class
> make[1]: *** [fake.o] Error 1
> make[1]: Leaving directory `/data/util/bthid-0.8'
> make: *** [fake.o] Error 2
> ====================
>
> hints welcome. I can test whatewer piece of code you throw at me. I'm
> however not knowledeable enough to code it myself.
>
> Charles
--
Charles Bueche <[email protected]>
sand, snow, wave, wind and net -surfer
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Charles,
> > this shouldn't be needed, because the mouse reconnects if you start
> > moving it. The only thing you need is a HID server process on your host.
>
> I have to test more. What is your command line for this ? I use
> /data/util/bthid-0.8-2.6/bthid -m -d -p 600 00:07:61:08:7E:2E
I used Peter's code only once, but I remember that you must use -l to
fork a server process.
> You do a very good job in that limited time, given all these beginners
> like me asking basic stuff on mailing lists :-)
Thank you.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi,
On Sun, 2003-12-14 at 15:22, Marcel Holtmann wrote:
> Hi Charles,
>
> > Now, my mouse isn't jumpy at all anymore. The only glitch left is that I
> > have to press its "connect" button just after starting bthid if I want a
> > link. Is there a way to get this automated ?
>
> this shouldn't be needed, because the mouse reconnects if you start
> moving it. The only thing you need is a HID server process on your host.
I have to test more. What is your command line for this ? I use
/data/util/bthid-0.8-2.6/bthid -m -d -p 600 00:07:61:08:7E:2E
> > I don't like to criticize about missing docs, because I have the same
> > problem at my main contract as Solaris admin (too many things to do to
> > write proper docs). However, bluez suffer a bit from that problem. They
> > are loads of small HOWTOs, but I miss a sort of white paper to explain
> > bluetooth in general. Any pointer is welcome.
>
> If you understand german you are in the lucky position that I have
> written some articles for german magazines and most of them are also
> available online.
I can read it, I will look at your articles, thx for the hint. I have
already seen a nice presentation which helped to understand the
bluetooth layers. A long time ago, I developped low-level code in Z80
assembler and C for embedded systems, so I'm not completely lost :-)
> The missing of a general white paper is a good point. I have thought so
> many times about writing the one and only introduction into Linux and
> Bluetooth, but always the same things stopped me from doing this. The
> first is that no company pays me for working on BlueZ and so we come to
> the second, that my time is very limited.
You do a very good job in that limited time, given all these beginners
like me asking basic stuff on mailing lists :-)
Charles
--
Charles Bueche <[email protected]>
sand, snow, wave, wind and net -surfer
Hi Charles,
> Now, my mouse isn't jumpy at all anymore. The only glitch left is that I
> have to press its "connect" button just after starting bthid if I want a
> link. Is there a way to get this automated ?
this shouldn't be needed, because the mouse reconnects if you start
moving it. The only thing you need is a HID server process on your host.
> I don't like to criticize about missing docs, because I have the same
> problem at my main contract as Solaris admin (too many things to do to
> write proper docs). However, bluez suffer a bit from that problem. They
> are loads of small HOWTOs, but I miss a sort of white paper to explain
> bluetooth in general. Any pointer is welcome.
If you understand german you are in the lucky position that I have
written some articles for german magazines and most of them are also
available online.
The missing of a general white paper is a good point. I have thought so
many times about writing the one and only introduction into Linux and
Bluetooth, but always the same things stopped me from doing this. The
first is that no company pays me for working on BlueZ and so we come to
the second, that my time is very limited.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Marcel and Olivier,
it works. Shame on me. Your remark about my empty link policy googled me
to another hcid.conf, and I suddenly saw mine, which was wrongly
configured, and worst, *hcid wasn't started*.
I'm running gentoo, and for some reason, there are no startup script for
bluetooth, so I was starting it by hand (hciconfig hci0 up).
Now, my mouse isn't jumpy at all anymore. The only glitch left is that I
have to press its "connect" button just after starting bthid if I want a
link. Is there a way to get this automated ?
I don't like to criticize about missing docs, because I have the same
problem at my main contract as Solaris admin (too many things to do to
write proper docs). However, bluez suffer a bit from that problem. They
are loads of small HOWTOs, but I miss a sort of white paper to explain
bluetooth in general. Any pointer is welcome.
Meanwhile, I'm still ready to test new versions of bthid or its future
replacement. Thanks a lot for your help. As soon as my last prob is
worked out (auto-connect of MX900), I'll finish my
"yet-another-bluez-mini-HOWTO" :-)
Charles
On Sat, 2003-12-13 at 20:04, Marcel Holtmann wrote:
> Hi Charles,
>
> > it's a CSR, it's included in my Dell Inspiron 8600. The I8600 is recent,
> > but of course it tells nothing about the included bluetooth module. Here
> > are the details :
> > ---
> > # hciconfig -a
> > hci0: Type: USB
> > BD Address: 00:10:C6:22:54:AF ACL MTU: 192:8 SCO MTU: 64:8
> > UP RUNNING PSCAN ISCAN
> > RX bytes:88278 acl:5562 sco:0 events:305 errors:0
> > TX bytes:3779 acl:170 sco:0 commands:54 errors:0
> > Features: 0xff 0xff 0x0f 0x00
> > Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
> > Link policy:
> > Link mode: SLAVE ACCEPT
> > Name: 'laptopcb (0)'
> > Class: 0x000000
> > Service Classes: Unspecified
> > Device Class: Miscellaneous,
> > HCI Ver: 1.1 (0x1) HCI Rev: 0x235 LMP Ver: 1.1 (0x1) LMP Subver:
> > 0x235
> > Manufacturer: Cambridge Silicon Radio (10)
> >
> > ---
> > I don't know how to find the firmware version. I have a dual boot with
> > XP, so I could look there if there is now way in Linux.
>
> this is a HCI 16.5 with 128 bit encryption and it should work fine. The
> problematic part is maybe that you have no link policy set. The minimum
> options should be rswitch and sniff. Please use
>
> http://www.holtmann.org/linux/bluetooth/btdfu-0.2.tar.gz
>
> to extract your firmware and send it to me.
>
> Regards
>
> Marcel
--
Charles Bueche <[email protected]>
sand, snow, wave, wind and net -surfer
Hi Charles,
> it's a CSR, it's included in my Dell Inspiron 8600. The I8600 is recent,
> but of course it tells nothing about the included bluetooth module. Here
> are the details :
> ---
> # hciconfig -a
> hci0: Type: USB
> BD Address: 00:10:C6:22:54:AF ACL MTU: 192:8 SCO MTU: 64:8
> UP RUNNING PSCAN ISCAN
> RX bytes:88278 acl:5562 sco:0 events:305 errors:0
> TX bytes:3779 acl:170 sco:0 commands:54 errors:0
> Features: 0xff 0xff 0x0f 0x00
> Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
> Link policy:
> Link mode: SLAVE ACCEPT
> Name: 'laptopcb (0)'
> Class: 0x000000
> Service Classes: Unspecified
> Device Class: Miscellaneous,
> HCI Ver: 1.1 (0x1) HCI Rev: 0x235 LMP Ver: 1.1 (0x1) LMP Subver:
> 0x235
> Manufacturer: Cambridge Silicon Radio (10)
>
> ---
> I don't know how to find the firmware version. I have a dual boot with
> XP, so I could look there if there is now way in Linux.
this is a HCI 16.5 with 128 bit encryption and it should work fine. The
problematic part is maybe that you have no link policy set. The minimum
options should be rswitch and sniff. Please use
http://www.holtmann.org/linux/bluetooth/btdfu-0.2.tar.gz
to extract your firmware and send it to me.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Marcel,
it's a CSR, it's included in my Dell Inspiron 8600. The I8600 is recent,
but of course it tells nothing about the included bluetooth module. Here
are the details :
---
# hciconfig -a
hci0: Type: USB
BD Address: 00:10:C6:22:54:AF ACL MTU: 192:8 SCO MTU: 64:8
UP RUNNING PSCAN ISCAN
RX bytes:88278 acl:5562 sco:0 events:305 errors:0
TX bytes:3779 acl:170 sco:0 commands:54 errors:0
Features: 0xff 0xff 0x0f 0x00
Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
Link policy:
Link mode: SLAVE ACCEPT
Name: 'laptopcb (0)'
Class: 0x000000
Service Classes: Unspecified
Device Class: Miscellaneous,
HCI Ver: 1.1 (0x1) HCI Rev: 0x235 LMP Ver: 1.1 (0x1) LMP Subver:
0x235
Manufacturer: Cambridge Silicon Radio (10)
---
# hciconfig hci0 features
hci0: Type: USB
BD Address: 00:10:C6:22:54:AF ACL MTU: 192:8 SCO MTU: 64:8
Features: 0xff 0xff 0x0f 0x00
<3-slot packets> <5-slot packets> <encryption> <slot
offset>
<timing accuracy> <role switch> <hold mode> <sniff mode>
<park mode> <RSSI> <channel quality> <SCO link>
<HV2 packets> <HV3 packets> <u-law log> <A-law log>
<CVSD> <paging scheme> <power control> <transparent SCO>
---
# hciconfig hci0 version
hci0: Type: USB
BD Address: 00:10:C6:22:54:AF ACL MTU: 192:8 SCO MTU: 64:8
HCI Ver: 1.1 (0x1) HCI Rev: 0x235 LMP Ver: 1.1 (0x1) LMP Subver:
0x235
Manufacturer: Cambridge Silicon Radio (10)
---
# hciconfig hci0 revision
hci0: Type: USB
BD Address: 00:10:C6:22:54:AF ACL MTU: 192:8 SCO MTU: 64:8
Unknown firmware
I don't know how to find the firmware version. I have a dual boot with
XP, so I could look there if there is now way in Linux.
Charles
On Sat, 2003-12-13 at 17:39, Marcel Holtmann wrote:
> Hi Charles,
>
> > Thanks for your answer. I don't run GPM, so the problem isn't from
> > there. But you gave me another idea : I loaded my modules, started the
> > patched bthid, exited X, and tried gpm in console. I got the same
> > problem !!! The mouse is jumpy as well. I think I can now exclude Xfree
> > from the potential problem.
>
> what kind of Bluetooth adapter do you use? I had some problem with CSR
> chips and old firmwares.
>
> > I guess bthid is producing the jumpy mouse. I tried to checkout the CVS
> > version, but I'm lost with autoconf and friends. I guess I will have to
> > wait for Marcel's new snapshot.
>
> Run the bootstrap script first ;)
>
> Regards
>
> Marcel
--
Charles Bueche <[email protected]>
sand, snow, wave, wind and net -surfer
Hi Charles,
> Thanks for your answer. I don't run GPM, so the problem isn't from
> there. But you gave me another idea : I loaded my modules, started the
> patched bthid, exited X, and tried gpm in console. I got the same
> problem !!! The mouse is jumpy as well. I think I can now exclude Xfree
> from the potential problem.
what kind of Bluetooth adapter do you use? I had some problem with CSR
chips and old firmwares.
> I guess bthid is producing the jumpy mouse. I tried to checkout the CVS
> version, but I'm lost with autoconf and friends. I guess I will have to
> wait for Marcel's new snapshot.
Run the bootstrap script first ;)
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Olivier,
Thanks for your answer. I don't run GPM, so the problem isn't from
there. But you gave me another idea : I loaded my modules, started the
patched bthid, exited X, and tried gpm in console. I got the same
problem !!! The mouse is jumpy as well. I think I can now exclude Xfree
from the potential problem.
I guess bthid is producing the jumpy mouse. I tried to checkout the CVS
version, but I'm lost with autoconf and friends. I guess I will have to
wait for Marcel's new snapshot.
What frustrate me is that I got it working twice last week, after
unloading/loading stopping/starting it several times, but I lost the
order and options I had.
Well, we it works, we all have to get a good beer :-)
Charles
On Wed, 2003-12-10 at 09:30, Olivier Bornet wrote:
> Hi Charles,
>
> > I tried your config, the wheel now works OK. Thanks a lot.
>
> Great. :-)
>
> > Another problem appeared after my initial email : the mouse movements
> > are showing sudden accelerations. Makes it difficult to use. I tried
> > several protocols in XF86Config-4, no difference.
>
> I have add this kind of problem with my USB mouse long time ago. This
> was because I have also gpm working (mouse usage in non grphical
> session). I you need to use gpm at the same time of XFree, you need to
> search on the web, as there is some special config to do. A quick search
> on google give lot of tips for this config. For example
> http://linuxfromscratch.org/pipermail/blfs-dev/2003-April/002637.html
>
> > I find this strange, as yesterday evening, the mouse suddenly worked
> > fine... I'm getting confused.
>
> Maybe you have installed gpm ?
>
> > Could you provide your setup ? hciconfig command, loaded modules, bthid
> > command line, etc ?
>
> I think as you have it working (maybe bad, but working), your bluetooth
> config is OK (hciconfig command, loaded modules). For the bthid command
> line, I have modified the /etc/init.d/bluetooth-hid file for having the
> command line:
>
> bthid -m -l -s
>
> > Another question : do I need MOUSEDEV in my kernel ? I have
> >
> > CONFIG_INPUT_MOUSEDEV=y
> > CONFIG_INPUT_MOUSEDEV_PSAUX=y
> > CONFIG_INPUT_MOUSEDEV_SCREEN_X=1920
> > CONFIG_INPUT_MOUSEDEV_SCREEN_Y=1200
>
> I think you need CONFIG_INPUT_MOUSEDEV, but I'm not sure. I have it in
> module. (but as I also have a USB mouse, I really need it).
>
> > should I remove this ?
>
> Because XF86Config-4 use /dev/input/mice, I think you really need it.
>
> Good day.
>
> Olivier
--
Charles Bueche <[email protected]>
sand, snow, wave, wind and net -surfer
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Charles,
> I tried your config, the wheel now works OK. Thanks a lot.
Great. :-)
> Another problem appeared after my initial email : the mouse movements
> are showing sudden accelerations. Makes it difficult to use. I tried
> several protocols in XF86Config-4, no difference.
I have add this kind of problem with my USB mouse long time ago. This
was because I have also gpm working (mouse usage in non grphical
session). I you need to use gpm at the same time of XFree, you need to
search on the web, as there is some special config to do. A quick search
on google give lot of tips for this config. For example
http://linuxfromscratch.org/pipermail/blfs-dev/2003-April/002637.html
> I find this strange, as yesterday evening, the mouse suddenly worked
> fine... I'm getting confused.
Maybe you have installed gpm ?
> Could you provide your setup ? hciconfig command, loaded modules, bthid
> command line, etc ?
I think as you have it working (maybe bad, but working), your bluetooth
config is OK (hciconfig command, loaded modules). For the bthid command
line, I have modified the /etc/init.d/bluetooth-hid file for having the
command line:
bthid -m -l -s
> Another question : do I need MOUSEDEV in my kernel ? I have
>
> CONFIG_INPUT_MOUSEDEV=y
> CONFIG_INPUT_MOUSEDEV_PSAUX=y
> CONFIG_INPUT_MOUSEDEV_SCREEN_X=1920
> CONFIG_INPUT_MOUSEDEV_SCREEN_Y=1200
I think you need CONFIG_INPUT_MOUSEDEV, but I'm not sure. I have it in
module. (but as I also have a USB mouse, I really need it).
> should I remove this ?
Because XF86Config-4 use /dev/input/mice, I think you really need it.
Good day.
Olivier
--
Olivier Bornet http://www.smartdata.ch/
[email protected] SMARTDATA SA
GPG key ID: C53D9218 CH Martigny/Lausanne
Hi Olivier,
I tried your config, the wheel now works OK. Thanks a lot.
Another problem appeared after my initial email : the mouse movements
are showing sudden accelerations. Makes it difficult to use. I tried
several protocols in XF86Config-4, no difference.
I re-emerged Xfree without the patch, the problem persist.
I find this strange, as yesterday evening, the mouse suddenly worked
fine... I'm getting confused.
Could you provide your setup ? hciconfig command, loaded modules, bthid
command line, etc ?
Another question : do I need MOUSEDEV in my kernel ? I have
CONFIG_INPUT_MOUSEDEV=y
CONFIG_INPUT_MOUSEDEV_PSAUX=y
CONFIG_INPUT_MOUSEDEV_SCREEN_X=1920
CONFIG_INPUT_MOUSEDEV_SCREEN_Y=1200
should I remove this ?
TIA,
Charles
On Tue, 2003-12-09 at 23:42, Olivier Bornet wrote:
> Hi Charles,
>
> > 1. XFree patch
> >
> > I patched my XFree (gentoo xfree-4.3.99.16) with this patch
> > (http://www.yhbt.net/normalperson/files/linux/LogitechMX_kernel_and_X_patch/4.3.0-XFree86_logitech_ps2.patch),
> > to make it understand PS/2+ protocol. Do I need this for the MX900 ?
>
> I don't have applied this patch, so I think this is not needed. (but I
> don't know anything about this patch.
>
> > 2. bthid patches
> >
> > - started with the bthid 0.8 stock source
> > - applied Edd Dumbill's uinput patch
> > - applied Olivier Bornet's patch-bthid-0.6-for-lt.txt
> > - modified out.c to use /dev/misc/uinput (Marcel says devfs produce
> > this, I don't see exactly what I do wrong or what I should do to get
> > /dev/input/uinput).
>
> OK. Almost the same for me, with a 2.6.0-test11 kernel.
>
> > 3. XF86Config-4
> >
> > --------
> > Section "InputDevice"
> > Identifier "LogitechUSBMouse"
> > Driver "mouse"
> > Option "Protocol" "ExplorerPS/2"
> > Option "Device" "/dev/input/mice"
> > Option "ZAxisMapping" "7 8"
> > Option "Resolution" "800"
> > Option "SmartScroll" "1"
> > Option "Buttons" "8"
> > Option "Emulate3Buttons" "false"
> > EndSection
> > ----------------
>
> Usualy, the ZAxisMapping in on the buttons 4 and 5. Here is my
> XF86Config-4 related section:
>
> ----------
> Section "InputDevice"
> Identifier "USB Mouse"
> Driver "mouse"
> Option "SendCoreEvents" "true"
> Option "Device" "/dev/input/mice"
> Option "Protocol" "ImPS/2"
> Option "Buttons" "5"
> Option "Emulate3Buttons" "true"
> Option "ZAxisMapping" "4 5"
> EndSection
> ----------
>
> With this config, the mouse wheel is working ok.
>
> Hope this help.
>
> Olivier
--
Charles Bueche <[email protected]>
sand, snow, wave, wind and net -surfer
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Charles,
> - modified out.c to use /dev/misc/uinput (Marcel says devfs produce
> this, I don't see exactly what I do wrong or what I should do to get
> /dev/input/uinput).
I don't use devfs and this is what I have in /dev
crw-rw-rw- 1 root root 10, 223 Aug 31 22:05 /dev/input/uinput
You can use mknod for creating it. Maybe your devfs problem is a
permission problem.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Charles,
> 1. XFree patch
>
> I patched my XFree (gentoo xfree-4.3.99.16) with this patch
> (http://www.yhbt.net/normalperson/files/linux/LogitechMX_kernel_and_X_patch/4.3.0-XFree86_logitech_ps2.patch),
> to make it understand PS/2+ protocol. Do I need this for the MX900 ?
I don't have applied this patch, so I think this is not needed. (but I
don't know anything about this patch.
> 2. bthid patches
>
> - started with the bthid 0.8 stock source
> - applied Edd Dumbill's uinput patch
> - applied Olivier Bornet's patch-bthid-0.6-for-lt.txt
> - modified out.c to use /dev/misc/uinput (Marcel says devfs produce
> this, I don't see exactly what I do wrong or what I should do to get
> /dev/input/uinput).
OK. Almost the same for me, with a 2.6.0-test11 kernel.
> 3. XF86Config-4
>
> --------
> Section "InputDevice"
> Identifier "LogitechUSBMouse"
> Driver "mouse"
> Option "Protocol" "ExplorerPS/2"
> Option "Device" "/dev/input/mice"
> Option "ZAxisMapping" "7 8"
> Option "Resolution" "800"
> Option "SmartScroll" "1"
> Option "Buttons" "8"
> Option "Emulate3Buttons" "false"
> EndSection
> ----------------
Usualy, the ZAxisMapping in on the buttons 4 and 5. Here is my
XF86Config-4 related section:
----------
Section "InputDevice"
Identifier "USB Mouse"
Driver "mouse"
Option "SendCoreEvents" "true"
Option "Device" "/dev/input/mice"
Option "Protocol" "ImPS/2"
Option "Buttons" "5"
Option "Emulate3Buttons" "true"
Option "ZAxisMapping" "4 5"
EndSection
----------
With this config, the mouse wheel is working ok.
Hope this help.
Olivier
--
Olivier Bornet http://www.smartdata.ch/
[email protected] SMARTDATA SA
GPG key ID: C53D9218 CH Martigny/Lausanne
Hi,
I have still some small glitches with my MX900. I have done the
following procedure :
1. XFree patch
I patched my XFree (gentoo xfree-4.3.99.16) with this patch
(http://www.yhbt.net/normalperson/files/linux/LogitechMX_kernel_and_X_patch/4.3.0-XFree86_logitech_ps2.patch), to make it understand PS/2+ protocol. Do I need this for the MX900 ?
2. bthid patches
- started with the bthid 0.8 stock source
- applied Edd Dumbill's uinput patch
- applied Olivier Bornet's patch-bthid-0.6-for-lt.txt
- modified out.c to use /dev/misc/uinput (Marcel says devfs produce
this, I don't see exactly what I do wrong or what I should do to get
/dev/input/uinput).
3. XF86Config-4
--------
Section "InputDevice"
Identifier "LogitechUSBMouse"
Driver "mouse"
Option "Protocol" "ExplorerPS/2"
Option "Device" "/dev/input/mice"
Option "ZAxisMapping" "7 8"
Option "Resolution" "800"
Option "SmartScroll" "1"
Option "Buttons" "8"
Option "Emulate3Buttons" "false"
EndSection
----------------
works :
- mouse movements : clean, precise, perfect
- buttons : 1,2,3 (2 = wheel-press)
problems :
- wheel doesn't work, even if I see button-7 and button-8 in xev output when I move the wheel.
do I use the correct protocol ?
Hints welcome...
TIA,
Charles
Hi Olivier,
> > Do you know how to access the display of the diNovo mediapad?
>
> Unfortunately, no. Has far as I known, this is not part of the HID
> profile. This is a "special" modified version by Logitech. I already
> have asked for this, without success at this time.
maybe someone with very good HID knowledge can decode the HID descriptor
and find an answer. Or we must use a protocol analyser to sniff the HIDP
packets.
> Also, for switching from USB to BT the "Bluetooth HUB", I will look with
> Logitech if we can have the specification for doing this switch. Hope I
> will have more success.
I have some response from CSR according the switch from HID to HCI mode.
This fits perfect for the EPoX dongle, but the Logitech dongle is quite
more complex. It presents a USB hub, two USB HID devices and four extra
HID devices with Windows XP.
However I actually don't how to implement the switch in Linux by now,
but I will take a look at it.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Marcel,
> Do you know how to access the display of the diNovo mediapad?
Unfortunately, no. Has far as I known, this is not part of the HID
profile. This is a "special" modified version by Logitech. I already
have asked for this, without success at this time.
Also, for switching from USB to BT the "Bluetooth HUB", I will look with
Logitech if we can have the specification for doing this switch. Hope I
will have more success.
Good day.
Olivier
--
Olivier Bornet http://www.smartdata.ch/
[email protected] SMARTDATA SA
GPG key ID: C53D9218 CH Martigny/Lausanne
Hi Olivier,
> Good. Hope this work with your implementation too.
I have tested it with the Microsoft, the Logitech and the EPoX mouse.
All of them are working fine now. I can't tell you much for the keyboard
layout patch, because my Microsoft keyboard has an UK layout and the
EPoX has an US layout. Only the diNovo keyboard uses a German layout and
this is working fine for me.
> I was posting only to bluz-users. Now, I will port to bluez-devel if I
> have other patches.
Do you know how to access the display of the diNovo mediapad?
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Marcel,
> I have now applied both patches to my implementation.
Good. Hope this work with your implementation too.
> > Here is another try (on bluez-devel this time). Thanks to let me known
> > what is bad with my patches. I'm sure if you will be put them in bthid,
> > people using Logitech deviceces will be happy. And also people using
> > non-us keyboard.
>
> You should always post them to the mailing list.
I was posting only to bluz-users. Now, I will port to bluez-devel if I
have other patches.
Good day.
Olivier
--
Olivier Bornet http://www.smartdata.ch/
[email protected] SMARTDATA SA
GPG key ID: C53D9218 CH Martigny/Lausanne
Hi Olivier,
> October 13, I have send a patch to Peter for the mouse button problems.
> See the two first email attached. No answer at all from Peter...
> This is against version 0.6 of bthid, but this is a very simple patch,
> so it must be easy to put in on the last release of bthid. The major
> problem is that the MX900 send all the button status when there is a
> change. And because if a button is not recognized by bthid, it was
> assigning it by default to the button 1. Has the MX900 has more buttons
> than bthid think. So, button 1 is always not pressed... (see attached
> emails for more info).
>
> October 18, I have send a patch for non-US keyboard. This was used by
> another guy on bluez-users. Again, no answer from Peter... See 3rd email
> attached to this one.
>
> October 28, I have send the patches on the bluez-users list. No answer
> from Peter... (see mailing list archive to find my message)
I have now applied both patches to my implementation.
> Here is another try (on bluez-devel this time). Thanks to let me known
> what is bad with my patches. I'm sure if you will be put them in bthid,
> people using Logitech deviceces will be happy. And also people using
> non-us keyboard.
You should always post them to the mailing list.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hello,
> > After this, I get the mouse events in X. The movements are OK, but the
> > button 1 is not doing its job.
> no. I have the same problem with the Logitech MX900 mouse. It must be a
> bug in the HID parser or in the HID descriptor of the device.
Since about 2 month now, I'm using bthid from Peter with a diNovo
desktop. This include the MX900 mouse. All is working OK. But...
But it requiere some patching...
October 13, I have send a patch to Peter for the mouse button problems.
See the two first email attached. No answer at all from Peter...
This is against version 0.6 of bthid, but this is a very simple patch,
so it must be easy to put in on the last release of bthid. The major
problem is that the MX900 send all the button status when there is a
change. And because if a button is not recognized by bthid, it was
assigning it by default to the button 1. Has the MX900 has more buttons
than bthid think. So, button 1 is always not pressed... (see attached
emails for more info).
October 18, I have send a patch for non-US keyboard. This was used by
another guy on bluez-users. Again, no answer from Peter... See 3rd email
attached to this one.
October 28, I have send the patches on the bluez-users list. No answer
from Peter... (see mailing list archive to find my message)
Here is another try (on bluez-devel this time). Thanks to let me known
what is bad with my patches. I'm sure if you will be put them in bthid,
people using Logitech deviceces will be happy. And also people using
non-us keyboard.
Good day.
Olivier
--
Olivier Bornet http://www.smartdata.ch/
[email protected] SMARTDATA SA
GPG key ID: C53D9218 CH Martigny/Lausanne
Hi Charles,
> I tried your patch on my linux-2.6.0-test11-gentoo Dell I8600 with my
> Logitech MX900 mouse. It applies OK, but (don't know if it's gentoo
> specific), my uinput is at /dev/misc/uinput, not /dev/input/uinput. So I
> had to fix the path in out.c.
this is because of devfs.
> After this, I get the mouse events in X. The movements are OK, but the
> button 1 is not doing its job. If I cat < /dev/input/mice, I see
> something happen when I click it, so it's obviously not the button
> hardware.
>
> I have compiled my X with a patch to enable the PS/2++ protocol. The
> patch is from Eric Wong, can be downloaded here :
> http://www.yhbt.net/normalperson/files/linux/LogitechMX_kernel_and_X_patch/4.3.0-XFree86_logitech_ps2.patch
>
> could this patch be the reason for my button 1 problem ?
no. I have the same problem with the Logitech MX900 mouse. It must be a
bug in the HID parser or in the HID descriptor of the device.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi,
[first, thanks a lot for helping me. I plan a mini-HOWTO ln a web page
when I'm done so other can get the same setup working.]
I tried your patch on my linux-2.6.0-test11-gentoo Dell I8600 with my
Logitech MX900 mouse. It applies OK, but (don't know if it's gentoo
specific), my uinput is at /dev/misc/uinput, not /dev/input/uinput. So I
had to fix the path in out.c.
After this, I get the mouse events in X. The movements are OK, but the
button 1 is not doing its job. If I cat < /dev/input/mice, I see
something happen when I click it, so it's obviously not the button
hardware.
I have compiled my X with a patch to enable the PS/2++ protocol. The
patch is from Eric Wong, can be downloaded here :
http://www.yhbt.net/normalperson/files/linux/LogitechMX_kernel_and_X_patch/4.3.0-XFree86_logitech_ps2.patch
could this patch be the reason for my button 1 problem ?
Hints welcome.
Charles
On Tue, 2003-12-02 at 17:01, Edd Dumbill wrote:
> On Tue, 2003-12-02 at 15:55, Edd Dumbill wrote:
> > Hmm, I'd thought I'd sent this to the list, but obviously didn't.
> > Here's a patch for Peter's bthid stuff to work with the uinput driver in
> > Linux 2.6. Use at your own risk. I had it going OK with an Apple
> > keyboard and MS mouse.
>
> And if I'd attached the patch first time, it would have been truly
> impressive.
>
> Here we are.
>
> -- Edd
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Stephen,
> If this posting is anything to go by, 2.4.24 is the very last chance for
> getting this into the 2.4 series, after that it's bugfixes only:
>
> http://www.ussg.iu.edu/hypermail/linux/kernel/0312.0/0052.html
I read his posting and as you see, he already listed the input userlevel
driver support. I hope that he will merge it very soon ;)
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Marcel,
If this posting is anything to go by, 2.4.24 is the very last chance for
getting this into the 2.4 series, after that it's bugfixes only:
http://www.ussg.iu.edu/hypermail/linux/kernel/0312.0/0052.html
On Tue, 2003-12-02 at 17:39, Marcel Holtmann wrote:
> Hi Edd,
>
> > Hmm, I'd thought I'd sent this to the list, but obviously didn't.
>
> yes you sent it to list. It still resides in my Bluetooth mailbox and it
> is working fine. I think Marcelo will take my backport of the uinput
> driver into 2.4.24 and so uinput should be the way to go, even if this
> means that we have to implement our own HID parser in userspace.
>
> Regards
>
> Marcel
>
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: SF.net Giveback Program.
> Does SourceForge.net help you be more productive? Does it
> help you create better code? SHARE THE LOVE, and help us help
> YOU! Click Here: http://sourceforge.net/donate/
> _______________________________________________
> Bluez-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/bluez-devel
--
Stephen Crane, Rococo Software Ltd. http://www.rococosoft.com
[email protected] +353-1-6601315 (ext 209)
Hi Peter,
> > yes you sent it to list. It still resides in my Bluetooth mailbox and
> > it
> > is working fine. I think Marcelo will take my backport of the uinput
> > driver into 2.4.24 and so uinput should be the way to go, even if this
> > means that we have to implement our own HID parser in userspace.
>
> What's wrong with the HID parser in bthid today?
my problem is that it depends on static variables and I don't see any
method to create HID reports from input events that the kernel send to
us and we have to forward to the HID device. For exmaple the keyboard
leds.
I have rewritten my core HID layer which is now only one process with an
event loop. It handles all connection establishment and termination fine
without any problems or stalled ACL links. It also gets the vendor and
product ID's from the PNP information. Some minor things have to be
fixed first, but I think that I will commit it to CVS very soon. If this
code is available I hope we can work together on a full featured HID
parser implementation and I think we can use your current code as a
starting point.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
On 2003.12.02 11:39 Marcel Holtmann wrote:
> Hi Edd,
>
> > Hmm, I'd thought I'd sent this to the list, but obviously didn't.
>
> yes you sent it to list. It still resides in my Bluetooth mailbox and
> it
> is working fine. I think Marcelo will take my backport of the uinput
> driver into 2.4.24 and so uinput should be the way to go, even if this
> means that we have to implement our own HID parser in userspace.
What's wrong with the HID parser in bthid today?
Hi Edd,
> Hmm, I'd thought I'd sent this to the list, but obviously didn't.
yes you sent it to list. It still resides in my Bluetooth mailbox and it
is working fine. I think Marcelo will take my backport of the uinput
driver into 2.4.24 and so uinput should be the way to go, even if this
means that we have to implement our own HID parser in userspace.
Regards
Marcel
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
On Tue, 2003-12-02 at 15:55, Edd Dumbill wrote:
> Hmm, I'd thought I'd sent this to the list, but obviously didn't.
> Here's a patch for Peter's bthid stuff to work with the uinput driver in
> Linux 2.6. Use at your own risk. I had it going OK with an Apple
> keyboard and MS mouse.
And if I'd attached the patch first time, it would have been truly
impressive.
Here we are.
-- Edd
Hmm, I'd thought I'd sent this to the list, but obviously didn't.
Here's a patch for Peter's bthid stuff to work with the uinput driver in
Linux 2.6. Use at your own risk. I had it going OK with an Apple
keyboard and MS mouse.
-- Edd
On Mon, 2003-12-01 at 15:38, Peter Klausler wrote:
> On 2003.11.30 12:51 Charles Bueche wrote:
> > Hi again,
> >
> > responding to my own email after studying fake.c : I got it to compile
> > on 2.6, but this code isn't able to run as module on 2.6 kernels,
> > which
> > changed the module format.
> >
> > Peter : do you plan a 2.6 version anytime soon ?
>
> For 2.6 the fake.o driver shouldn't be needed; I'm told that
> there's a legitimate driver in 2.6 that can be used to
> inject input events from user-land. But I've been way
> too busy porting 2.6 to a new architecture to spend much
> time playing with it on i386, so I can't tell you how hard
> it would be to adjust bthid to use the new driver.
--
Edd Dumbill -- more from me at http://usefulinc.com/edd/blog/
Managing Editor, XML.com, XMLhack.com; Chair, XML Europe 2003
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
On 2003.11.30 12:51 Charles Bueche wrote:
> Hi again,
>
> responding to my own email after studying fake.c : I got it to compile
> on 2.6, but this code isn't able to run as module on 2.6 kernels,
> which
> changed the module format.
>
> Peter : do you plan a 2.6 version anytime soon ?
For 2.6 the fake.o driver shouldn't be needed; I'm told that
there's a legitimate driver in 2.6 that can be used to
inject input events from user-land. But I've been way
too busy porting 2.6 to a new architecture to spend much
time playing with it on i386, so I can't tell you how hard
it would be to adjust bthid to use the new driver.
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel