[Bluez-users] Timeout message from kernel

[Bluez-users] Timeout message from kernel

[Jeff Sutherland]

Some newbie bluetooth questions: I'm running the latest bluez-utils on a 
2.4.24 uClinx kernel on a Moto ColdFire 5282 processor.  It's sort of 
working, but having a bit of trouble trying to connect between a brain box 
serial dongle on the ColdFire board and a brainbox usb dongle on a pc.  
'hcitool info from the pc yields:

[root@Arwen bluetooth]# hcitool info 00:0A:4F:00:13:5F
Requesting information ...
        BD Address:  00:0A:4F:00:13:5F
        Device Name: BlueZ (0)
        LMP Version: 1.1 (0x1) LMP Subversion: 0x175
        Manufacturer: Cambridge Silicon Radio (10)
        Features: 0xff 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
                <3-slot packets> <5-slot packets> <encryption> <slot offset>
                <timing accuracy> <role switch> <hold mode> <sniff mode>
                <park state> <RSSI> <channel quality> <SCO link>
                <HV2 packets> <HV3 packets> <u-law log> <A-law log>
                <CVSD> <paging scheme> <power control> <transparent SCO>

so it looks like the ColdFire dongle is alive and responding to the pc.  A 
'hcitool scan' from the ColdFire board yields:

/> hcitool scan
Scanning ...
bcsp_timed_event: Timeout, retransmitting 1 pkts
bcsp_timed_event: Timeout, retransmitting 1 pkts
bcsp_timed_event: Timeout, retransmitting 1 pkts
        00:0A:4F:00:09:5E       BlueZ (0)
/>

and hcitool info on the ColdFire board yields:

/> hcitool info 00:0A:4F:00:09:5E
Requesting information ...
bcsp_timed_event: Timeout, retransmitting 1 pkts
Can't create connection: Input/output error
pid 52: failed 256
/>

Anyone have an idea of what the kernel is complaining about?  My thoughts are 
there might still be some serial port issues on the ColdFire board...

-Jeff
-- 
Secure your email with Gnu Privacy Guard and Mozilla, the open source
cross platform solution for Windows, Linux, and Mac OS X.
See http://enigmail.mozdev.org/



-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users

Re: [Bluez-users] Timeout message from kernel

[Marcel Holtmann]

Hi Jeff,

> Some newbie bluetooth questions: I'm running the latest bluez-utils on a 
> 2.4.24 uClinx kernel on a Moto ColdFire 5282 processor.  It's sort of 
> working, but having a bit of trouble trying to connect between a brain box 
> serial dongle on the ColdFire board and a brainbox usb dongle on a pc.  
> 'hcitool info from the pc yields:
> 
> [root@Arwen bluetooth]# hcitool info 00:0A:4F:00:13:5F
> Requesting information ...
>         BD Address:  00:0A:4F:00:13:5F
>         Device Name: BlueZ (0)
>         LMP Version: 1.1 (0x1) LMP Subversion: 0x175
>         Manufacturer: Cambridge Silicon Radio (10)
>         Features: 0xff 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
>                 <3-slot packets> <5-slot packets> <encryption> <slot offset>
>                 <timing accuracy> <role switch> <hold mode> <sniff mode>
>                 <park state> <RSSI> <channel quality> <SCO link>
>                 <HV2 packets> <HV3 packets> <u-law log> <A-law log>
>                 <CVSD> <paging scheme> <power control> <transparent SCO>
> 
> so it looks like the ColdFire dongle is alive and responding to the pc.  A 
> 'hcitool scan' from the ColdFire board yields:
> 
> /> hcitool scan
> Scanning ...
> bcsp_timed_event: Timeout, retransmitting 1 pkts
> bcsp_timed_event: Timeout, retransmitting 1 pkts
> bcsp_timed_event: Timeout, retransmitting 1 pkts

you can ignore these messages, because they are no errors.

> and hcitool info on the ColdFire board yields:
> 
> /> hcitool info 00:0A:4F:00:09:5E
> Requesting information ...
> bcsp_timed_event: Timeout, retransmitting 1 pkts
> Can't create connection: Input/output error

This can mean many things. Try using l2ping or l2test. Can you run
hcidump on the ColdFire board?

Regards

Marcel




-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users

Re: [Bluez-users] Timeout message from kernel

[Jeff Sutherland]

On Thursday 12 February 2004 19:14, Marcel Holtmann wrote:
> Hi Jeff,
<snip>

> > /> hcitool info 00:0A:4F:00:09:5E
> > Requesting information ...
> > bcsp_timed_event: Timeout, retransmitting 1 pkts
> > Can't create connection: Input/output error
>
> This can mean many things. Try using l2ping or l2test. Can you run
> hcidump on the ColdFire board?

l2ping from either end works.  I didn't cross-compile hcidump for the ColdFire 
board, need to do some makefile wizardry in the parser subdir first.  l2test 
on both ends just sits there waiting for something, no doubt I'm not invoking 
it properly.  hcidump on the pc end shows the ping packets arriving from 
ColdFire.  I guess my connection is basically working.  On the ColdFire 
board:
/bin> hciconfig
hci0:   Type: UART
        BD Address: 00:0A:4F:00:13:5F ACL MTU: 192:8  SCO MTU: 64:8
        UP RUNNING PSCAN ISCAN
        RX bytes:289 acl:0 sco:0 events:14 errors:0
        TX bytes:554 acl:0 sco:0 commands:14 errors:0

But invoking hcitool still yields:
/bin> hcitool info 00:0A:4F:00:09:5E
Requesting information ...
bcsp_timed_event: Timeout, retransmitting 1 pkts
Can't create connection: Input/output error
pid 54: failed 256

bcsp_timed_event: is a kernel error msg.  I should probably try to set up an 
l2test session properly next.  Any advice?

Thanks,
-Jeff
-- 
Secure your email with Gnu Privacy Guard and Mozilla, the open source
cross platform solution for Windows, Linux, and Mac OS X.
See http://enigmail.mozdev.org/

Re: [Bluez-users] Timeout message from kernel

[Marcel Holtmann]

Hi Jeff,

> l2ping from either end works.  I didn't cross-compile hcidump for the ColdFire 
> board, need to do some makefile wizardry in the parser subdir first.  l2test 
> on both ends just sits there waiting for something, no doubt I'm not invoking 
> it properly.  hcidump on the pc end shows the ping packets arriving from 
> ColdFire.  I guess my connection is basically working.  On the ColdFire 
> board:
> /bin> hciconfig
> hci0:   Type: UART
>         BD Address: 00:0A:4F:00:13:5F ACL MTU: 192:8  SCO MTU: 64:8
>         UP RUNNING PSCAN ISCAN
>         RX bytes:289 acl:0 sco:0 events:14 errors:0
>         TX bytes:554 acl:0 sco:0 commands:14 errors:0
> 
> But invoking hcitool still yields:
> /bin> hcitool info 00:0A:4F:00:09:5E
> Requesting information ...
> bcsp_timed_event: Timeout, retransmitting 1 pkts
> Can't create connection: Input/output error
> pid 54: failed 256
> 
> bcsp_timed_event: is a kernel error msg.  I should probably try to set up an 
> l2test session properly next.  Any advice?

with the next update of the BCSP driver I will turn this into a debug
message.

Using l2test is easy and with -h you see all the options ;)

	One side:	l2test -r	/* listen and receive */
	Other side:	l2test -s	/* connect and send */

Regards

Marcel




-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users

Re: [Bluez-users] Timeout message from kernel

[Jeff Sutherland]

On Friday 13 February 2004 09:22, Marcel Holtmann wrote:
> > /bin> hcitool info 00:0A:4F:00:09:5E
> > Requesting information ...
> > bcsp_timed_event: Timeout, retransmitting 1 pkts
> > Can't create connection: Input/output error
> > pid 54: failed 256
> >
> > bcsp_timed_event: is a kernel error msg.  I should probably try to set up
> > an l2test session properly next.  Any advice?
>
> with the next update of the BCSP driver I will turn this into a debug
> message.
>
> Using l2test is easy and with -h you see all the options ;)
>
> 	One side:	l2test -r	/* listen and receive */
> 	Other side:	l2test -s	/* connect and send */

Hmmm...
[root@Arwen test]# ./l2test -w
l2test[29184]: Waiting for connection on psm 10 ...

/bin> l2test -r
l2test[79]: Waiting for connection on psm 10 ...

and there they sit, nothing happening.  The SCO test works fine going either 
way, however.  It seems that on the ColdFire board most of the hcitool 
commands fail with some kind of 'HCI...timeout' error.

Note that from ColdFire, this fails:
/bin> hcitool cc --role=s 00:0A:4F:00:09:5E
bcsp_timed_event: Timeout, retransmitting 1 pkts
Can't create connection: Input/output error

/bin> hcitool con
Connections:

But from the pc, I get this:
[root@Arwen test]# hcitool cc --role=m 00:0A:4F:00:13:5F
Can't create connection: Connection timed out

But now running 'hcitool con' on Coldfire gives this:
/bin> hcitool con
Connections:
        > ACL 00:0A:4F:00:09:5E handle 40 state 1 lm SLAVE

and from the pc:
[root@Arwen test]# hcitool con
Connections:
        < ACL 00:0A:4F:00:13:5F handle 41 state 1 lm MASTER

So it seems that hcitool <> hcid, or userland <> kernel is messed up on 
ColdFire, is that correct?  Remember I'm still a newbie at this...

Regards,
-Jeff
-- 
Secure your email with Gnu Privacy Guard and Mozilla, the open source
cross platform solution for Windows, Linux, and Mac OS X.
See http://enigmail.mozdev.org/

Re: [Bluez-users] Timeout message from kernel

[Marcel Holtmann]

Hi Jeff,

> > Using l2test is easy and with -h you see all the options ;)
> >
> > 	One side:	l2test -r	/* listen and receive */
> > 	Other side:	l2test -s	/* connect and send */
> 
> Hmmm...
> [root@Arwen test]# ./l2test -w
> l2test[29184]: Waiting for connection on psm 10 ...
> 
> /bin> l2test -r
> l2test[79]: Waiting for connection on psm 10 ...

why don't you try what I have written? And read what "l2test -h" says.

	l2test -r	/* listen and receive */
	l2test -w	/* listen and send */

Two connections that are listening. How should they connect to each
other?

> Note that from ColdFire, this fails:
> /bin> hcitool cc --role=s 00:0A:4F:00:09:5E
> bcsp_timed_event: Timeout, retransmitting 1 pkts
> Can't create connection: Input/output error
> 
> /bin> hcitool con
> Connections:
> 
> But from the pc, I get this:
> [root@Arwen test]# hcitool cc --role=m 00:0A:4F:00:13:5F
> Can't create connection: Connection timed out
> 
> But now running 'hcitool con' on Coldfire gives this:
> /bin> hcitool con
> Connections:
>         > ACL 00:0A:4F:00:09:5E handle 40 state 1 lm SLAVE
> 
> and from the pc:
> [root@Arwen test]# hcitool con
> Connections:
>         < ACL 00:0A:4F:00:13:5F handle 41 state 1 lm MASTER
> 
> So it seems that hcitool <> hcid, or userland <> kernel is messed up on 
> ColdFire, is that correct?  Remember I'm still a newbie at this...

Maybe this is a role switch problem. Show us both outputs from
"hciconfig -a".

Regards

Marcel




-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users

Re: [Bluez-users] Timeout message from kernel

[Jeff Sutherland]

On Friday 13 February 2004 10:28, Marcel Holtmann wrote:
> Hi Jeff,
> why don't you try what I have written? And read what "l2test -h" says.

Ok, Doh, heat of battle kind of thing, just been working on it too long.  
Making ColdFire the l2test listener seems to work ok:
/> l2test -r
l2test[25]: Waiting for connection on psm 10 ...
l2test[25]: Connect from 00:0A:4F:00:09:5E [imtu 672, omtu 672, flush_to 
65535]
                                                                                
l2test[25]: Receiving ...
l2test[25]: 672 bytes in 0.04 sec, 17.70 kB/s
l2test[25]: 672 bytes in 0.01 sec, 50.86 kB/s
l2test[25]: 672 bytes in 0.03 sec, 20.72 kB/s

and so on and so forth.  Making ColdFire the sender however produces this 
output on the pc side:

[root@Arwen test]# ./l2test -r
l2test[30292]: Waiting for connection on psm 10 ...
l2test[30293]: Connect from 00:0A:4F:00:13:5F [imtu 672, omtu 672, flush_to 
65535]
l2test[30293]: Receiving ...
l2test[30293]: 672 bytes in 0.06 sec, 11.28 kB/s
l2test[30293]: seq missmatch: 1 -> 4
l2test[30293]: 672 bytes in 0.09 sec, 7.70 kB/s
l2test[30293]: seq missmatch: 5 -> 8
l2test[30293]: 672 bytes in 0.08 sec, 8.21 kB/s
l2test[30293]: seq missmatch: 9 -> 12
etc., etc.

> Maybe this is a role switch problem. Show us both outputs from
> "hciconfig -a".

ColdFire:
/> hciconfig -a
hci0:   Type: UART
        BD Address: 00:0A:4F:00:13:5F ACL MTU: 192:8  SCO MTU: 64:8
        UP RUNNING PSCAN ISCAN
        RX bytes:306 acl:0 sco:0 events:14 errors:0
        TX bytes:513 acl:0 sco:0 commands:12 errors:0
        Features: 0xff 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
        Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
        Link policy: HOLD SNIFF PARK
        Link mode: SLAVE ACCEPT
        Name: 'BlueZ (0)'
        Class: 0x000100
        Service Classes: Unspecified
        Device Class: Computer, Uncategorized
        HCI Ver: 1.1 (0x1) HCI Rev: 0x175 LMP Ver: 1.1 (0x1) LMP Subver: 0x175
        Manufacturer: Cambridge Silicon Radio (10)

PC:
[root@Arwen test]# hciconfig -a
hci0:   Type: USB
        BD Address: 00:0A:4F:00:09:5E ACL MTU: 192:8  SCO MTU: 64:8
        UP RUNNING PSCAN ISCAN
        RX bytes:1887092 acl:3025 sco:29743 events:1513 errors:0
        TX bytes:309425 acl:1448 sco:1709 commands:109 errors:0
        Features: 0xff 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
        Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
        Link policy: HOLD SNIFF PARK
        Link mode: SLAVE ACCEPT
        Name: 'BlueZ (0)'
        Class: 0x000100
        Service Classes: Unspecified
        Device Class: Computer, Uncategorized
        HCI Ver: 1.1 (0x1) HCI Rev: 0x20d LMP Ver: 1.1 (0x1) LMP Subver: 0x20d
        Manufacturer: Cambridge Silicon Radio (10)

Other than that sequence problem it looks like the link mostly works, still 
issues with most of the commands of hcitool though, on Coldfire.

Thanks,
-Jeff
-- 
Secure your email with Gnu Privacy Guard and Mozilla, the open source
cross platform solution for Windows, Linux, and Mac OS X.
See http://enigmail.mozdev.org/

Re: [Bluez-users] Timeout message from kernel

[Marcel Holtmann]

Hi Jeff,

> Making ColdFire the l2test listener seems to work ok:
> /> l2test -r
> l2test[25]: Waiting for connection on psm 10 ...
> l2test[25]: Connect from 00:0A:4F:00:09:5E [imtu 672, omtu 672, flush_to 
> 65535]
>                                                                                 
> l2test[25]: Receiving ...
> l2test[25]: 672 bytes in 0.04 sec, 17.70 kB/s
> l2test[25]: 672 bytes in 0.01 sec, 50.86 kB/s
> l2test[25]: 672 bytes in 0.03 sec, 20.72 kB/s
> 
> and so on and so forth.  Making ColdFire the sender however produces this 
> output on the pc side:
> 
> [root@Arwen test]# ./l2test -r
> l2test[30292]: Waiting for connection on psm 10 ...
> l2test[30293]: Connect from 00:0A:4F:00:13:5F [imtu 672, omtu 672, flush_to 
> 65535]
> l2test[30293]: Receiving ...
> l2test[30293]: 672 bytes in 0.06 sec, 11.28 kB/s
> l2test[30293]: seq missmatch: 1 -> 4
> l2test[30293]: 672 bytes in 0.09 sec, 7.70 kB/s
> l2test[30293]: seq missmatch: 5 -> 8
> l2test[30293]: 672 bytes in 0.08 sec, 8.21 kB/s
> l2test[30293]: seq missmatch: 9 -> 12
> etc., etc.

should not happen. I don't why it happens to you.

> > Maybe this is a role switch problem. Show us both outputs from
> > "hciconfig -a".
> 
> ColdFire:
> /> hciconfig -a
> hci0:   Type: UART
>         BD Address: 00:0A:4F:00:13:5F ACL MTU: 192:8  SCO MTU: 64:8
>         UP RUNNING PSCAN ISCAN
>         RX bytes:306 acl:0 sco:0 events:14 errors:0
>         TX bytes:513 acl:0 sco:0 commands:12 errors:0
>         Features: 0xff 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
>         Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
>         Link policy: HOLD SNIFF PARK
>         Link mode: SLAVE ACCEPT
>         Name: 'BlueZ (0)'
>         Class: 0x000100
>         Service Classes: Unspecified
>         Device Class: Computer, Uncategorized
>         HCI Ver: 1.1 (0x1) HCI Rev: 0x175 LMP Ver: 1.1 (0x1) LMP Subver: 0x175
>         Manufacturer: Cambridge Silicon Radio (10)
> 
> PC:
> [root@Arwen test]# hciconfig -a
> hci0:   Type: USB
>         BD Address: 00:0A:4F:00:09:5E ACL MTU: 192:8  SCO MTU: 64:8
>         UP RUNNING PSCAN ISCAN
>         RX bytes:1887092 acl:3025 sco:29743 events:1513 errors:0
>         TX bytes:309425 acl:1448 sco:1709 commands:109 errors:0
>         Features: 0xff 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
>         Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
>         Link policy: HOLD SNIFF PARK
>         Link mode: SLAVE ACCEPT
>         Name: 'BlueZ (0)'
>         Class: 0x000100
>         Service Classes: Unspecified
>         Device Class: Computer, Uncategorized
>         HCI Ver: 1.1 (0x1) HCI Rev: 0x20d LMP Ver: 1.1 (0x1) LMP Subver: 0x20d
>         Manufacturer: Cambridge Silicon Radio (10)

Both look fine, but you should enable role switch support in hcid.conf
or set it by hand with hciconfig.

> Other than that sequence problem it looks like the link mostly works, still 
> issues with most of the commands of hcitool though, on Coldfire.

Keep trying and look with hcidump at it, but we can also have a bug in
our tools if they are used on a ColdFire processor.

Regards

Marcel




-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users

Re: [Bluez-users] Timeout message from kernel

[Jeff Sutherland]

On Friday 13 February 2004 18:40, Marcel Holtmann wrote:
> Hi Jeff,

<snip>

> Both look fine, but you should enable role switch support in hcid.conf
> or set it by hand with hciconfig.
>
> > Other than that sequence problem it looks like the link mostly works,
> > still issues with most of the commands of hcitool though, on Coldfire.
>
> Keep trying and look with hcidump at it, but we can also have a bug in
> our tools if they are used on a ColdFire processor.

I've tried the l2test -s from Coldfire several times, doing 'hciconfig hci0 
reset' in between for good measure, always get the same results on the pc 
end.  This looks suspiciously like an endian problem somewhere, but why would 
receive work ok?

2test[30524]: Receiving ...
l2test[30524]: 672 bytes in 0.06 sec, 11.49 kB/s
l2test[30524]: seq missmatch: 1 -> 4
l2test[30524]: 672 bytes in 0.09 sec, 7.57 kB/s
l2test[30524]: seq missmatch: 5 -> 8

Thanks for all your help on this.  I was really hoping once I got the buggered 
mcfserial.c driver working that I would be done...

-Jeff
-- 
Secure your email with Gnu Privacy Guard and Mozilla, the open source
cross platform solution for Windows, Linux, and Mac OS X.
See http://enigmail.mozdev.org/

Re: [Bluez-users] Timeout message from kernel

[Marcel Holtmann]

Hi Jeff,

> I've tried the l2test -s from Coldfire several times, doing 'hciconfig hci0 
> reset' in between for good measure, always get the same results on the pc 
> end.  This looks suspiciously like an endian problem somewhere, but why would 
> receive work ok?

the endian stuff looks fine in the l2test.c source. What endian do the
Coldfire processor use?

Regards

Marcel




-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users

Re: [Bluez-users] Timeout message from kernel

[Jeff Sutherland]

On Friday 13 February 2004 19:06, Marcel Holtmann wrote:
> Hi Jeff,
>
> > I've tried the l2test -s from Coldfire several times, doing 'hciconfig
> > hci0 reset' in between for good measure, always get the same results on
> > the pc end.  This looks suspiciously like an endian problem somewhere,
> > but why would receive work ok?
>
> the endian stuff looks fine in the l2test.c source. What endian do the
> Coldfire processor use?

Big Endian.

Regards,
-Jeff
-- 
Secure your email with Gnu Privacy Guard and Mozilla, the open source
cross platform solution for Windows, Linux, and Mac OS X.
See http://enigmail.mozdev.org/

Re: [Bluez-users] Timeout message from kernel

[Marcel Holtmann]

Hi Jeff,

> > the endian stuff looks fine in the l2test.c source. What endian do the
> > Coldfire processor use?
> 
> Big Endian.

maybe we really have some endian problems here and also in hcitool. The
pkt_type for create_connection is for example uint16 and this explains
why "hcitool info ..." fails. I don't have access to a machine with big
endian, so most of the userspace tools are not fully tested.

Regards

Marcel




-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users

Re: [Bluez-users] Timeout message from kernel

[Jeff Sutherland]

=2D----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Saturday 14 February 2004 03:53, Marcel Holtmann wrote:
> Hi Jeff,
>
> > > the endian stuff looks fine in the l2test.c source. What endian do the
> > > Coldfire processor use?
> >
> > Big Endian.
>
> maybe we really have some endian problems here and also in hcitool. The
> pkt_type for create_connection is for example uint16 and this explains
> why "hcitool info ..." fails. I don't have access to a machine with big
> endian, so most of the userspace tools are not fully tested.

Hello Marcel,
 I'll have access to this machine for a little while, so if you have someth=
ing=20
you'd like to try just send patches this way :-)

Regards,
=2D -Jeff
=2D --=20
Secure your email with Gnu Privacy Guard and Mozilla, the open source
cross platform solution for Windows, Linux, and Mac OS X.
See http://enigmail.mozdev.org/
=2D----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQFALinEvvyQ7TDL7JURAmSeAKCaqzUp4BvjDe2j/wzTmkcQYNeUfgCdFs2c
6++aTGGr+pMUVRcLmnRWG7E=3D
=3DRL2F
=2D----END PGP SIGNATURE-----

Re: [Bluez-users] Timeout message from kernel

[Marcel Holtmann]

[-- Attachment #1: Type: text/plain, Size: 516 bytes --]

Hi Jeff,

> > maybe we really have some endian problems here and also in hcitool. The
> > pkt_type for create_connection is for example uint16 and this explains
> > why "hcitool info ..." fails. I don't have access to a machine with big
> > endian, so most of the userspace tools are not fully tested.
> 
>  I'll have access to this machine for a little while, so if you have something 
> you'd like to try just send patches this way :-)

try this one to solve the problem with "hcitool info ...".

Regards

Marcel


[-- Attachment #2: patch --]
[-- Type: text/x-patch, Size: 919 bytes --]

Index: hcitool.c
===================================================================
RCS file: /cvsroot/bluez/utils/tools/hcitool.c,v
retrieving revision 1.35
diff -u -r1.35 hcitool.c
--- hcitool.c	21 Dec 2003 16:08:25 -0000	1.35
+++ hcitool.c	14 Feb 2004 14:56:52 -0000
@@ -436,7 +436,7 @@
 	bacpy(&cr->bdaddr, &bdaddr);
 	cr->type = ACL_LINK;
 	if (ioctl(dd, HCIGETCONNINFO, (unsigned long) cr) < 0) {
-		if (hci_create_connection(dd, &bdaddr, HCI_DM1 | HCI_DH1, 0, 0, &handle, 25000) < 0) {
+		if (hci_create_connection(dd, &bdaddr, htobs(HCI_DM1 | HCI_DH1), 0, 0, &handle, 25000) < 0) {
 			perror("Can't create connection");
 			close(dd);
 			exit(1);
@@ -653,7 +653,7 @@
 		exit(1);
 	}
 
-	if (hci_create_connection(dd, &bdaddr, ptype, 0, role, &handle, 1000) < 0)
+	if (hci_create_connection(dd, &bdaddr, htobs(ptype), 0, role, &handle, 1000) < 0)
 		perror("Can't create connection");
 	hci_close_dev(dd);
 }

Re: [Bluez-users] Timeout message from kernel

[Jeff Sutherland]

On Saturday 14 February 2004 10:02, Marcel Holtmann wrote:
> Hi Jeff,
>
> > > maybe we really have some endian problems here and also in hcitool. The
> > > pkt_type for create_connection is for example uint16 and this explains
> > > why "hcitool info ..." fails. I don't have access to a machine with big
> > > endian, so most of the userspace tools are not fully tested.
> >
> >  I'll have access to this machine for a little while, so if you have
> > something you'd like to try just send patches this way :-)
>
> try this one to solve the problem with "hcitool info ...".

/> hcitool info 00:0A:4F:00:09:5E
Requesting information ...
bcsp_timed_event: Timeout, retransmitting 1 pkts
        BD Address:  00:0A:4F:00:09:5E
        Device Name: BlueZ (0)
        LMP Version: 1.1 (0x1) LMP Subversion: 0x20d
        Manufacturer: Cambridge Silicon Radio (10)
        Features: 0xff 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
                <3-slot packets> <5-slot packets> <encryption> <slot offset>
                <timing accuracy> <role switch> <hold mode> <sniff mode>
                <park state> <RSSI> <channel quality> <SCO link>
                <HV2 packets> <HV3 packets> <u-law log> <A-law log>
                <CVSD> <paging scheme> <power control> <transparent SCO>
                                                                                
/>

Heh, my hunch was correct for once, looks like it WAS an endian problem :-D

Thanks!
-Jeff

(now to see if I can get a ppp connection going...)
-- 
Secure your email with Gnu Privacy Guard and Mozilla, the open source
cross platform solution for Windows, Linux, and Mac OS X.
See http://enigmail.mozdev.org/

Re: [Bluez-users] Timeout message from kernel

[Marcel Holtmann]

Hi Jeff,

> /> hcitool info 00:0A:4F:00:09:5E
> Requesting information ...
> bcsp_timed_event: Timeout, retransmitting 1 pkts
>         BD Address:  00:0A:4F:00:09:5E
>         Device Name: BlueZ (0)
>         LMP Version: 1.1 (0x1) LMP Subversion: 0x20d
>         Manufacturer: Cambridge Silicon Radio (10)
>         Features: 0xff 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
>                 <3-slot packets> <5-slot packets> <encryption> <slot offset>
>                 <timing accuracy> <role switch> <hold mode> <sniff mode>
>                 <park state> <RSSI> <channel quality> <SCO link>
>                 <HV2 packets> <HV3 packets> <u-law log> <A-law log>
>                 <CVSD> <paging scheme> <power control> <transparent SCO>
>                                                                                 
> />
> 
> Heh, my hunch was correct for once, looks like it WAS an endian problem :-D

the patch is now in CVS. I suspect that the code contains other endian
problems. You always overlook some one. I really think that I need a big
endian machine at home. Are there any sponsors?

Regards

Marcel




-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users