Nabble - freebsd-bluetooth

Re: obexapp patches for type-punning

2009-12-01T15:25:55Z

On Tue, Dec 1, 2009 at 11:31 AM, Iain Hibbert <plunky@...> wrote:
> Hi,
>
> while on the subject of obexapp, I have a patch to remove some annoying
> 'type punning' compilation errors that gcc spits out that I'm not totally
> sure are valid, but see below anyway..

that is why i had -fno-strict-aliasing in makefile :)

> I don't understand the OBEX protocol or libopenobex enough to know if it
> is possible, but the last case also removes a potential NULL dereference -
> if the OBEX_ObjectGetNonHdrData() fails to recover a proper data, an error
> is logged but it carries on and hdr->flags might cause a segfault?

yes, i think those are good. combined patch went out in the previous email.

thanks,
max
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: patch to make obexapp work with BSD iconv(3)

2009-12-01T15:24:18Z

On Tue, Dec 1, 2009 at 11:15 AM, Iain Hibbert <plunky@...> wrote:

> On Tue, 1 Dec 2009, Matthias Drochner wrote:
>
>> iconv_open(3), this is more robust than trying to interpret
>> the locale string
>
> btw nl_langinfo(3) is defined to always return a valid pointer, so the
>
> if (locale == NULL)
> return (-1);
>
> is not required..
>
> also, the locale variable can be private to obexapp_util_locale_init() as
> it is not used elsewhere

ok, so, combined patch would look something like the attached.

could you please give it a try and see if it still works?

thanks,
max

Index: Makefile
===================================================================
RCS file: /usr/local/cvs/ports/obexapp/Makefile,v
retrieving revision 1.11
diff -u -r1.11 Makefile
--- Makefile 23 Apr 2007 18:29:18 -0000 1.11
+++ Makefile 1 Dec 2009 20:00:19 -0000
@@ -14,8 +14,7 @@
DEBUG_FLAGS= -g

CFLAGS+= -I/usr/local/include \
- -I/usr/local/include/openobex \
- -fno-strict-aliasing
+ -I/usr/local/include/openobex

DPADD= ${LIBBLUETOOTH} ${LIBSDP} ${LIBBSDXML} ${LIBREADLINE}
LDADD= -L/usr/lib -L/usr/local/lib \
Index: client.c
===================================================================
RCS file: /usr/local/cvs/ports/obexapp/client.c,v
retrieving revision 1.18
diff -u -r1.18 client.c
--- client.c 9 Apr 2009 23:16:31 -0000 1.18
+++ client.c 1 Dec 2009 19:40:32 -0000
@@ -1219,9 +1219,8 @@
int obex_rsp)
{
context_p context = (context_p) OBEX_GetUserData(handle);
- obex_connect_hdr_t *hdr = NULL;
obex_headerdata_t hv;
- uint8_t hi;
+ uint8_t hi, *data = NULL;
uint32_t hlen;

log_debug("%s(): Connect completed, response %#x", __func__, obex_rsp);
@@ -1232,10 +1231,12 @@
if (obex_rsp != OBEX_RSP_SUCCESS)
return (obex_rsp);

- if (OBEX_ObjectGetNonHdrData(object, (uint8_t **) &hdr) == sizeof(*hdr))
+ if (OBEX_ObjectGetNonHdrData(object, &data) == sizeof(obex_connect_hdr_t))
log_debug("%s(): OBEX connect header: " \
- "version=%#x, flags=%#x, mtu=%d", __func__,
- hdr->version, hdr->flags, ntohs(hdr->mtu));
+ "version=%#x, flags=%#x, mtu=%d", __func__,
+ ((obex_connect_hdr_t *) data)->version,
+ ((obex_connect_hdr_t *) data)->flags,
+ ntohs(((obex_connect_hdr_t *) data)->mtu));
else
log_err("%s(): Invalid OBEX connect header?!", __func__);

Index: server.c
===================================================================
RCS file: /usr/local/cvs/ports/obexapp/server.c,v
retrieving revision 1.12
diff -u -r1.12 server.c
--- server.c 20 Aug 2009 21:57:18 -0000 1.12
+++ server.c 1 Dec 2009 20:04:33 -0000
@@ -471,19 +471,20 @@
obexapp_server_request_connect(obex_t *handle, obex_object_t *object,
__unused int obex_rsp)
{
- obex_connect_hdr_t *hdr = NULL;
obex_headerdata_t hv;
- uint8_t hi;
+ uint8_t hi, *data = NULL;
uint32_t hlen;
uint8_t const *target = NULL;
int target_len = 0;

log_debug("%s()", __func__);

- if (OBEX_ObjectGetNonHdrData(object, (uint8_t **) &hdr) == sizeof(*hdr))
+ if (OBEX_ObjectGetNonHdrData(object, &data) == sizeof(obex_connect_hdr_t))
log_debug("%s(): OBEX connect header: version=%#x, " \
- "flags=%#x, mtu=%d", __func__, hdr->version, hdr->flags,
- ntohs(hdr->mtu));
+ "flags=%#x, mtu=%d", __func__,
+ ((obex_connect_hdr_t *) data)->version,
+ ((obex_connect_hdr_t *) data)->flags,
+ ntohs(((obex_connect_hdr_t *) data)->mtu));
else
log_err("%s(): Invalid OBEX connect header?!", __func__);

@@ -1086,9 +1087,8 @@
__unused int obex_rsp)
{
context_p context = (context_p) OBEX_GetUserData(handle);
- obex_setpath_hdr_t *hdr = NULL;
obex_headerdata_t hv;
- uint8_t hi;
+ uint8_t hi, flags = 0xff, *data = NULL;
uint32_t hlen;
int got_name = 0;

@@ -1096,10 +1096,12 @@

context->file[0] = '\0';

- if (OBEX_ObjectGetNonHdrData(object, (uint8_t **) &hdr) == sizeof(*hdr))
- log_debug("%s(): OBEX setpath header: flags=%#x, constants=%d",
- __func__, hdr->flags, hdr->constants);
- else
+ if (OBEX_ObjectGetNonHdrData(object, &data) == sizeof(obex_setpath_hdr_t)) {
+ flags = ((obex_setpath_hdr_t *) data)->flags;
+
+ log_debug("%s(): OBEX setpath header: flags=%#x, constants=%d",
+ __func__, flags, ((obex_setpath_hdr_t *) data)->constants);
+ } else
log_err("%s(): Invalid OBEX setpath header?!", __func__);

while (OBEX_ObjectGetNextHeader(handle, object, &hi, &hv, &hlen)) {
@@ -1151,9 +1153,9 @@
* directory) means "cd ..". Everything else is forbidden.
*/

- if (hdr->flags != 0x3) {
+ if (flags != 0x3) {
log_err("%s(): Invalid flags for 'cd ..', flags=%#x",
- __func__, hdr->flags);
+ __func__, flags);

return (OBEXAPP_PACK_RSP_CODES(OBEX_RSP_FORBIDDEN,
OBEX_RSP_FORBIDDEN));
@@ -1169,9 +1171,9 @@
* 'cd /'. Everything else is forbidden
*/

- if (hdr->flags != 0x2) {
+ if (flags != 0x2) {
log_err("%s(): Invalid flags for 'cd /', flags=%#x",
- __func__, hdr->flags);
+ __func__, flags);

return (OBEXAPP_PACK_RSP_CODES(OBEX_RSP_FORBIDDEN,
OBEX_RSP_FORBIDDEN));
@@ -1180,7 +1182,7 @@
strlcpy(context->file, context->root, PATH_MAX);
}

- if (hdr->flags == 0) {
+ if (flags == 0) {
if (mkdir(context->file, 0755) < 0 && errno != EEXIST) {
log_err("%s(): mkdir(%s) failed. %s (%d)",
__func__, context->file,
Index: util.c
===================================================================
RCS file: /usr/local/cvs/ports/obexapp/util.c,v
retrieving revision 1.15
diff -u -r1.15 util.c
--- util.c 20 Aug 2009 21:57:18 -0000 1.15
+++ util.c 1 Dec 2009 19:58:53 -0000
@@ -34,6 +34,7 @@
#include <ctype.h>
#include <errno.h>
#include <iconv.h>
+#include <langinfo.h>
#include <libgen.h>
#include <limits.h>
#include <locale.h>
@@ -56,7 +57,6 @@
* Init locale
*/

-static char const *locale = NULL;
static int32_t utf16 = 0;
static iconv_t utf82locale = (iconv_t) -1;
static iconv_t locale2utf8 = (iconv_t) -1;
@@ -66,25 +66,10 @@
int
obexapp_util_locale_init(void)
{
- int one = 1;
+ char const *locale;

- locale = setlocale(LC_CTYPE, "");
- if (locale == NULL)
- return (-1);
-
- if (strcmp(locale, "ASCII") == 0 ||
- strcmp(locale, "C") == 0 ||
- strcmp(locale, "POSIX") == 0 ||
- strcmp(locale, "US-ASCII") == 0)
- locale = "en_US.US-ASCII";
-
- locale = strchr(locale, '.');
- if (locale == NULL)
- return (-1);
-
- locale ++;
- if (locale[0] == '\0')
- return (-1);
+ setlocale(LC_CTYPE, "");
+ locale = nl_langinfo(CODESET);

utf16 = (strstr(locale, "UTF-16") != NULL)? 1 : 0;

@@ -94,7 +79,6 @@
return (-1);

iconv(utf82locale, NULL, NULL, NULL, NULL);
- iconvctl(utf82locale, ICONV_SET_DISCARD_ILSEQ, &one);

/* current locale -> UTF-8 */
locale2utf8 = iconv_open("UTF-8", locale);
@@ -104,7 +88,6 @@
}

iconv(locale2utf8, NULL, NULL, NULL, NULL);
- iconvctl(locale2utf8, ICONV_SET_DISCARD_ILSEQ, &one);

/* UTF-16BE -> current locale */
utf16be2locale = iconv_open(locale, "UTF-16BE");
@@ -114,7 +97,6 @@
}

iconv(utf16be2locale, NULL, NULL, NULL, NULL);
- iconvctl(utf16be2locale, ICONV_SET_DISCARD_ILSEQ, &one);

/* current locale -> UTF-16BE */
locale2utf16be = iconv_open("UTF-16BE", locale);
@@ -124,7 +106,6 @@
}

iconv(locale2utf16be, NULL, NULL, NULL, NULL);
- iconvctl(locale2utf16be, ICONV_SET_DISCARD_ILSEQ, &one);

return (0);
}

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

obexapp patches for type-punning

2009-12-01T11:30:56Z

Hi,

while on the subject of obexapp, I have a patch to remove some annoying
'type punning' compilation errors that gcc spits out that I'm not totally
sure are valid, but see below anyway..

I don't understand the OBEX protocol or libopenobex enough to know if it
is possible, but the last case also removes a potential NULL dereference -
if the OBEX_ObjectGetNonHdrData() fails to recover a proper data, an error
is logged but it carries on and hdr->flags might cause a segfault?

iain

--- ~client.c 2009-04-10 00:16:31.000000000 +0100
+++ client.c 2009-12-01 16:44:49.000000000 +0000
@@ -1219,10 +1219,10 @@ obexapp_client_request_connect_done(obex
int obex_rsp)
{
context_p context = (context_p) OBEX_GetUserData(handle);
- obex_connect_hdr_t *hdr = NULL;
obex_headerdata_t hv;
uint8_t hi;
uint32_t hlen;
+ uint8_t *data = NULL;

log_debug("%s(): Connect completed, response %#x", __func__, obex_rsp);

@@ -1232,10 +1232,12 @@ obexapp_client_request_connect_done(obex
if (obex_rsp != OBEX_RSP_SUCCESS)
return (obex_rsp);

- if (OBEX_ObjectGetNonHdrData(object, (uint8_t **) &hdr) == sizeof(*hdr))
+ if (OBEX_ObjectGetNonHdrData(object, &data) == sizeof(obex_connect_hdr_t))
log_debug("%s(): OBEX connect header: " \
"version=%#x, flags=%#x, mtu=%d", __func__,
- hdr->version, hdr->flags, ntohs(hdr->mtu));
+ ((obex_connect_hdr_t *)data)->version,
+ ((obex_connect_hdr_t *)data)->flags,
+ ntohs(((obex_connect_hdr_t *)data)->mtu));
else
log_err("%s(): Invalid OBEX connect header?!", __func__);

--- ~server.c 2009-08-20 22:57:18.000000000 +0100
+++ server.c 2009-12-01 16:57:08.000000000 +0000
@@ -471,19 +471,21 @@ static int
obexapp_server_request_connect(obex_t *handle, obex_object_t *object,
__unused int obex_rsp)
{
- obex_connect_hdr_t *hdr = NULL;
obex_headerdata_t hv;
uint8_t hi;
uint32_t hlen;
uint8_t const *target = NULL;
int target_len = 0;
+ uint8_t *data = NULL;

log_debug("%s()", __func__);

- if (OBEX_ObjectGetNonHdrData(object, (uint8_t **) &hdr) == sizeof(*hdr))
+ if (OBEX_ObjectGetNonHdrData(object, &data) == sizeof(obex_connect_hdr_t))
log_debug("%s(): OBEX connect header: version=%#x, " \
- "flags=%#x, mtu=%d", __func__, hdr->version, hdr->flags,
- ntohs(hdr->mtu));
+ "flags=%#x, mtu=%d", __func__,
+ ((obex_connect_hdr_t *)data)->version,
+ ((obex_connect_hdr_t *)data)->flags,
+ ntohs(((obex_connect_hdr_t *)data)->mtu));
else
log_err("%s(): Invalid OBEX connect header?!", __func__);

@@ -1086,20 +1088,22 @@ obexapp_server_request_setpath(obex_t *h
__unused int obex_rsp)
{
context_p context = (context_p) OBEX_GetUserData(handle);
- obex_setpath_hdr_t *hdr = NULL;
obex_headerdata_t hv;
uint8_t hi;
uint32_t hlen;
int got_name = 0;
+ uint8_t *data = NULL;
+ uint8_t flags = 0;

log_debug("%s()", __func__);

context->file[0] = '\0';

- if (OBEX_ObjectGetNonHdrData(object, (uint8_t **) &hdr) == sizeof(*hdr))
+ if (OBEX_ObjectGetNonHdrData(object, &data) == sizeof(obex_setpath_hdr_t)) {
+ flags = ((obex_setpath_hdr_t *)data)->flags;
log_debug("%s(): OBEX setpath header: flags=%#x, constants=%d",
- __func__, hdr->flags, hdr->constants);
- else
+ __func__, flags, ((obex_setpath_hdr_t *)data)->constants);
+ } else
log_err("%s(): Invalid OBEX setpath header?!", __func__);

while (OBEX_ObjectGetNextHeader(handle, object, &hi, &hv, &hlen)) {
@@ -1145,15 +1149,14 @@ obexapp_server_request_setpath(obex_t *h
}

if (!got_name) {
-
/*
* No name and flags == 0x3 (back up one level + don't create
* directory) means "cd ..". Everything else is forbidden.
*/

- if (hdr->flags != 0x3) {
+ if (flags != 0x3) {
log_err("%s(): Invalid flags for 'cd ..', flags=%#x",
- __func__, hdr->flags);
+ __func__, flags);

return (OBEXAPP_PACK_RSP_CODES(OBEX_RSP_FORBIDDEN,
OBEX_RSP_FORBIDDEN));
@@ -1169,9 +1172,9 @@ obexapp_server_request_setpath(obex_t *h
* 'cd /'. Everything else is forbidden
*/

- if (hdr->flags != 0x2) {
+ if (flags != 0x2) {
log_err("%s(): Invalid flags for 'cd /', flags=%#x",
- __func__, hdr->flags);
+ __func__, flags);

return (OBEXAPP_PACK_RSP_CODES(OBEX_RSP_FORBIDDEN,
OBEX_RSP_FORBIDDEN));
@@ -1180,7 +1183,7 @@ obexapp_server_request_setpath(obex_t *h
strlcpy(context->file, context->root, PATH_MAX);
}

- if (hdr->flags == 0) {
+ if (flags == 0) {
if (mkdir(context->file, 0755) < 0 && errno != EEXIST) {
log_err("%s(): mkdir(%s) failed. %s (%d)",
__func__, context->file,

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: patch to make obexapp work with BSD iconv(3)

2009-12-01T11:15:40Z

On Tue, 1 Dec 2009, Matthias Drochner wrote:

> iconv_open(3), this is more robust than trying to interpret
> the locale string

btw nl_langinfo(3) is defined to always return a valid pointer, so the

if (locale == NULL)
return (-1);

is not required..

also, the locale variable can be private to obexapp_util_locale_init() as
it is not used elsewhere

regards,
iain

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: patch to make obexapp work with BSD iconv(3)

2009-12-01T10:28:40Z

Matthias,

> Here is a patch which removes the GNU iconv dependency from obexapp.
> I've successfully tested it by sending vcards with strange names;
> the resulting files were identical.

thanks for the patch!

could you please tell me which bsd flavor you tested this on?

> The relevant changes are:
> -just use the result of nl_langinfo(CODESET) as input of
> iconv_open(3), this is more robust than trying to interpret
> the locale string
> -remove iconvctl(ICONV_SET_DISCARD_ILSEQ) -- BSD doesn't
> have this and it should be unneeded because it only
> compensates for some non-standard behaviour of GNU iconv.
> (The bsdxml->expat part is unrelated.)

i dont really have any objections to those, just wanted to get some
clarification.

thanks
max
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: patch to make obexapp work with BSD iconv(3)

2009-12-01T04:52:39Z

On Tue, 1 Dec 2009 12:50:54 +0100
Matthias Drochner <m.drochner@...> wrote:
> Here is a patch

Oops, this time for real.
best regards
Matthias

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

patch-ac (2K) Download Attachment

patch to make obexapp work with BSD iconv(3)

2009-12-01T03:50:54Z

Here is a patch which removes the GNU iconv dependency from obexapp.
I've successfully tested it by sending vcards with strange names;
the resulting files were identical.
The relevant changes are:
-just use the result of nl_langinfo(CODESET) as input of
iconv_open(3), this is more robust than trying to interpret
the locale string
-remove iconvctl(ICONV_SET_DISCARD_ILSEQ) -- BSD doesn't
have this and it should be unneeded because it only
compensates for some non-standard behaviour of GNU iconv.
(The bsdxml->expat part is unrelated.)

best regards
Matthias
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: Problems getting the bluetooth stack to start

2009-11-23T01:38:58Z

On Sun, Nov 22, 2009 at 07:23:35PM -0800, Maksim Yevmenkin wrote:

> this will not work. as per l2control(8) man page

oops! sorry for the noise.

> well, it could not open /dev/ptyp6? is it busy? did you try different terminal?

yep, got i/o error with them all.

> l2control(8) is not an issue you. rfcomm_sppd(1) had problem with
> opening master pty.

I'm going to update and buildworld again, and remake all my devices. I
saw some odd ownership for /dev/ttyp6 and friends.

cheers

--
John - comp dot john at googlemail dot com
OpenBSD firewall | FreeBSD desktop | Ubuntu Karmic laptop
GPG: 0xF08A33C5
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: Problems getting the bluetooth stack to start

2009-11-23T01:26:59Z

On Mon, Nov 23, 2009 at 08:40:05AM +0000, Iain Hibbert wrote:
> On Mon, 23 Nov 2009, John wrote:
>
> > Then I pulled the dongle from the socket on the back and
> > plugged it into one at the front, and this showed more promise.
>
> perhaps the one at the back is EHCI but the one at the front is not (or,
> full-speed and high-speed or whatever they call it) since you had to
> disable ehci to get it working..?

No, I recompiled without EHCI on the basis that there were issues with
it enabled (I did a lot of googling before posting here, and it was
advice though I think it pertained to netbsd and some other specific
advice) but at that stage I was grasping at straws ;)
--
John - comp dot john at googlemail dot com
OpenBSD firewall | FreeBSD desktop | Ubuntu Karmic laptop
GPG: 0xF08A33C5
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: Problems getting the bluetooth stack to start

2009-11-23T00:39:59Z

On Mon, 23 Nov 2009, John wrote:

> Then I pulled the dongle from the socket on the back and
> plugged it into one at the front, and this showed more promise.

perhaps the one at the back is EHCI but the one at the front is not (or,
full-speed and high-speed or whatever they call it) since you had to
disable ehci to get it working..?

iain

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: Problems getting the bluetooth stack to start

2009-11-22T19:23:35Z

On Sun, Nov 22, 2009 at 5:49 PM, John <comp.john@...> wrote:

[...]

everything up to this point looks good

> What I can't work out now is why i can't see the channels:
>
> potato# l2ping -a 00:0a:d9:3e:70:24
> 0 bytes from 00:0a:d9:3e:70:24 seq_no=0 time=500.567 ms result=0
> 0 bytes from 00:0a:d9:3e:70:24 seq_no=1 time=31.890 ms result=0
> 0 bytes from 00:0a:d9:3e:70:24 seq_no=2 time=31.895 ms result=0
> 0 bytes from 00:0a:d9:3e:70:24 seq_no=3 time=31.897 ms result=0
> 0 bytes from 00:0a:d9:3e:70:24 seq_no=4 time=32.900 ms result=0
> 0 bytes from 00:0a:d9:3e:70:24 seq_no=5 time=30.906 ms result=0

this looks fine

> potato# l2control -a 00:0a:d9:3e:70:24 Read_Channel_List
> l2control: Could not bind socket, bdaddr=00:0a:d9:3e:70:24: Network is
> down

this will not work. as per l2control(8) man page

"
The l2control utility connects to the local device with the specified
BD_ADDR and attempts to send the specified command.
"

i.e. l2control(8) will display _local_ l2cap channels (basically l2cap
connections). you dont have any l2cap channels (i.e. connections).

i'm guessing you want to see the services offered on your t610 phone.
you need to use sdpcontrol(8) for this.

> I have paired the handset with the device and it shows the fqdn of the
> freebsd machine. If I do sdpcontrol -a 00:0a:d9:3e:70:24 browse I get
> pages of output.

yes, this will tell you service names and service parameters

> But! If I do rfcomm_sppd -a 00:0a:d9:3e:70:24 -t /dev/ttyp6 I see this
> in messages:
>
> Nov 23 01:38:24 potato rfcomm_sppd: Could not open(/dev/ptyp6).
> Input/output error

well, it could not open /dev/ptyp6? is it busy? did you try different terminal?

> any ideas? I think the l2control error and the rfcomm one may be
> related, but I haven't a clue how to resolve the issue.

l2control(8) is not an issue you. rfcomm_sppd(1) had problem with
opening master pty.

thanks
max
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: Problems getting the bluetooth stack to start

2009-11-22T17:49:16Z

Hi, thanks for replying.

On Sun, Nov 22, 2009 at 09:45:25AM +0000, Iain Hibbert wrote:

> On Sat, 21 Nov 2009, John wrote:
>
> > Firstly, the small size of the device means that it is very awkward to
> > pull the dongle out once it's pushed in. When restarting the computer,
> > is it sufficient to reboot(8) then power the machine off, wait a few
> > seconds then power it back up? I mean does this properly reset the
> > dongle, or does it truly have to be pulled out and pushed back in for
> > the system to re-read it?
>
> power down should be sufficient to reset the device entirely, internal
> state is rarely kept (sometimes there is nvram for link keys but thats
> all)

Yeah I thought it should be, but it seems that is not the case on my
hardware. (gigabyte motherboard & case, AMD Athalon 64 x2 6000+). Whilst
trying to fix things, I brought the OS up-to-date with 7-STABLE (I was
running 7.2-R p4) and went through mergemaster the long and hard way,
making sure all the bluetooth-related stuff was the newest. Rebooted,
still no joy. Then I pulled the dongle from the socket on the back and
plugged it into one at the front, and this showed more promise. ubt0
didn't appear in messages before, now it did. Seems it needs to be
unplugged then plugged back in. Additionally, I had to go through the
routine of stopping the bluetooth stack, starting it again, getting the
"Unable to setup Bluetooth stack for device ubt0" error, then starting
it again and getting no error.

Now I get:

potato# kldstat
Id Refs Address Size Name
1 14 0xffffffff80100000 bf99e0 kernel
2 1 0xffffffff80cfa000 7118 geom_concat.ko
3 1 0xffffffff80d02000 9808 ng_ubt.ko
4 6 0xffffffff80d0c000 14e98 netgraph.ko
5 1 0xffffffff80e22000 972 pflog.ko
6 1 0xffffffff80e23000 2aefc pf.ko
7 1 0xffffffff80e4e000 12c9e ng_btsocket.ko
8 3 0xffffffff80e61000 89a ng_bluetooth.ko
9 1 0xffffffff80e62000 aa7a fuse.ko
10 1 0xffffffff80e6d000 827 accf_http.ko
11 1 0xffffffff80e6e000 8b1a ng_hci.ko
12 1 0xffffffff80e77000 b498 ng_l2cap.ko
13 1 0xffffffff80e83000 19ba ng_socket.ko

and

potato# ngctl li
There are 7 total nodes:
Name: btsock_l2c_raw Type: btsock_l2c_raw ID: 00000002 Num hooks: 1
Name: ngctl23568 Type: socket ID: 00000063 Num hooks: 0
Name: ubt0hci Type: hci ID: 00000059 Num hooks: 3
Name: btsock_hci_raw Type: btsock_hci_raw ID: 00000001 Num hooks: 1
Name: ubt0l2cap Type: l2cap ID: 0000005d Num hooks: 3
Name: btsock_l2c Type: btsock_l2c ID: 00000003 Num hooks: 1
Name: ubt0 Type: ubt ID: 00000004 Num hooks: 1

so that looks OK

potato# hccontrol -n ubt0hci inquiry
Inquiry result, num_responses=1
Inquiry result #0
BD_ADDR: 00:0a:d9:3e:70:24
Page Scan Rep. Mode: 0x1
Page Scan Period Mode: 00
Page Scan Mode: 00
Class: 52:02:04
Clock offset: 0x273b
Inquiry complete.
Status: No error [00]

potato# hccontrol -n ubt0hci remote_name_request 00:0a:d9:3e:70:24
BD_ADDR: 00:0a:d9:3e:70:24
Name: T610

so, in the parlance of the literature, the phone is my server and this
is the server's MAC address: 00:0a:d9:3e:70:24

What I can't work out now is why i can't see the channels:

potato# l2ping -a 00:0a:d9:3e:70:24
0 bytes from 00:0a:d9:3e:70:24 seq_no=0 time=500.567 ms result=0
0 bytes from 00:0a:d9:3e:70:24 seq_no=1 time=31.890 ms result=0
0 bytes from 00:0a:d9:3e:70:24 seq_no=2 time=31.895 ms result=0
0 bytes from 00:0a:d9:3e:70:24 seq_no=3 time=31.897 ms result=0
0 bytes from 00:0a:d9:3e:70:24 seq_no=4 time=32.900 ms result=0
0 bytes from 00:0a:d9:3e:70:24 seq_no=5 time=30.906 ms result=0
^C
potato# l2control -a 00:0a:d9:3e:70:24 Read_Channel_List
l2control: Could not bind socket, bdaddr=00:0a:d9:3e:70:24: Network is
down

I have paired the handset with the device and it shows the fqdn of the
freebsd machine. If I do sdpcontrol -a 00:0a:d9:3e:70:24 browse I get
pages of output.

But! If I do rfcomm_sppd -a 00:0a:d9:3e:70:24 -t /dev/ttyp6 I see this
in messages:

Nov 23 01:38:24 potato rfcomm_sppd: Could not open(/dev/ptyp6).
Input/output error

any ideas? I think the l2control error and the rfcomm one may be
related, but I haven't a clue how to resolve the issue.
--
John - comp dot john at googlemail dot com
OpenBSD firewall | FreeBSD desktop | Ubuntu Karmic laptop
GPG: 0xF08A33C5
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: iPhone and bcm-2045b

2009-11-22T10:52:55Z

On Mon, 16 Nov 2009, Iain Hibbert wrote:

> A friend of mine just got an iPhone 3G, he was here the other day and I
> tried to probe the device using my laptop (Thinkpad T60 with NetBSD) which
> has bluetooth built in (a Broadcom BCM2045B connected via internal USB)
>
> I managed to discover it ok, but when I opened a connection to the device,
> my controller detached from the USB bus!

Just as a followup to this, my CSR dongle managed to connect just fine so
I guess the Broadcom controller has a bug causing it to crash. Output of
my SDP query tool below. (searched for L2CAP uuid)

The iPhone bluetooth capability is kind of limited for a phone though,
with no Object Push or File Transfer. The NAP service is advertised but
returns "Not Allowed" to connection setup attempts (supposedly its
disabled by the network operator).

iain

ServiceRecordHandle: 0x4f491116
ServiceClassIDList:
Network Access Point
ServiceRecordState: 0x00000000
ProtocolDescriptorList:
L2CAP (PSM 0x000f)
BNEP (v1.0; IPv4, ARP, 0x8100, IPv6)
BrowseGroupList:
Public Browse Root
LanguageBaseAttributeIDList:
en.UTF-8 base 0x0100
fr.UTF-8 base 0x0110
de.UTF-8 base 0x0120
ja.UTF-8 base 0x0130
ServiceAvailability: 255/255
BluetoothProfileDescriptorList:
Network Access Point, v1.0
ServiceName: "PAN Network Access Profile"
ServiceDescription: "Network Access Point"
SecurityDescription: Service-level Security
NetAccessType: 3G Cellular
MaxNetAccessRate: 256000

ServiceRecordHandle: 0x4f49a620
ServiceClassIDList:
00000000-deca-fade-deca-deafdecacafe
ServiceRecordState: 0x00000000
ProtocolDescriptorList:
L2CAP
RFCOMM (channel 2)
BrowseGroupList:
Public Browse Root
LanguageBaseAttributeIDList:
en.UTF-8 base 0x0100
fr.UTF-8 base 0x0110
de.UTF-8 base 0x0120
ja.UTF-8 base 0x0130
ServiceAvailability: 255/255
BluetoothProfileDescriptorList:
Serial Port, v1.0
ServiceName: "Wireless iAP"

ServiceRecordHandle: 0x4f49110c
ServiceClassIDList:
A/V Remote Control Target
ServiceRecordState: 0x00000000
ProtocolDescriptorList:
L2CAP (PSM 0x0017)
AVCTP (v1.2)
BrowseGroupList:
Public Browse Root
LanguageBaseAttributeIDList:
en.UTF-8 base 0x0100
fr.UTF-8 base 0x0110
de.UTF-8 base 0x0120
ja.UTF-8 base 0x0130
ServiceAvailability: 255/255
BluetoothProfileDescriptorList:
A/V Remote Control, v1.0
ServiceName: "AVRCP Device"
ServiceDescription: "Remote Control Device"
SupportedFeatures:
Category 1

ServiceRecordHandle: 0x4f49110a
ServiceClassIDList:
Audio Source
ServiceRecordState: 0x00000000
ProtocolDescriptorList:
L2CAP (PSM 0x0019)
AVDTP (v1.2)
BrowseGroupList:
Public Browse Root
LanguageBaseAttributeIDList:
en.UTF-8 base 0x0100
fr.UTF-8 base 0x0110
de.UTF-8 base 0x0120
ja.UTF-8 base 0x0130
ServiceAvailability: 255/255
BluetoothProfileDescriptorList:
Advanced Audio Distribution, v1.2
ServiceName: "Audio Source"
SupportedFeatures:
Player

ServiceRecordHandle: 0x4f49112f
ServiceClassIDList:
Phonebook Access Server
ServiceRecordState: 0x00000000
ProtocolDescriptorList:
L2CAP
RFCOMM (channel 13)
OBEX
BrowseGroupList:
Public Browse Root
LanguageBaseAttributeIDList:
en.UTF-8 base 0x0100
fr.UTF-8 base 0x0110
de.UTF-8 base 0x0120
ja.UTF-8 base 0x0130
ServiceAvailability: 255/255
BluetoothProfileDescriptorList:
Phonebook Access, v1.0
ServiceName: "Phonebook"
SupportedRepositories:
Local Phonebook

ServiceRecordHandle: 0x4f491104
ServiceClassIDList:
IrMC Sync
ServiceRecordState: 0x00000000
ProtocolDescriptorList:
L2CAP
RFCOMM (channel 1)
OBEX
BrowseGroupList:
Public Browse Root
LanguageBaseAttributeIDList:
en.UTF-8 base 0x0100
fr.UTF-8 base 0x0110
de.UTF-8 base 0x0120
ja.UTF-8 base 0x0130
ServiceAvailability: 255/255
BluetoothProfileDescriptorList:
IrMC Sync, v1.0
ServiceName: "IrMC Sync"
SupportedDataStoresList:
Phonebook

ServiceRecordHandle: 0x4f49111f
ServiceClassIDList:
Handsfree Audio Gateway
Generic Audio
ServiceRecordState: 0x00000000
ProtocolDescriptorList:
L2CAP
RFCOMM (channel 8)
BrowseGroupList:
Public Browse Root
LanguageBaseAttributeIDList:
en.UTF-8 base 0x0100
fr.UTF-8 base 0x0110
de.UTF-8 base 0x0120
ja.UTF-8 base 0x0130
ServiceAvailability: 255/255
BluetoothProfileDescriptorList:
Handsfree, v1.5
ServiceName: "Handsfree Gateway"
Network: Ability to reject a call
SupportedFeatures:
3 Way Calling
Echo Cancellation/Noise Reduction

ServiceRecordHandle: 0x00000000
ServiceClassIDList:
Service Discovery Server
ServiceRecordState: 0x00000000
ProtocolDescriptorList:
L2CAP
SDP
LanguageBaseAttributeIDList:
en.UTF-8 base 0x0100
fr.UTF-8 base 0x0110
de.UTF-8 base 0x0120
ja.UTF-8 base 0x0130
ServiceAvailability: 255/255
VersionNumberList: v1.0
ServiceDatabaseState: 0x00000008
AttributeID 0x0300:
seq8(56)
str8(3) "2.1"
str8(49) "Copyright (c) 2007 Apple Inc. All rights rese ...
AttributeID 0x0308:
url8(13) "www%2eapple%2ecom"

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: Problems getting the bluetooth stack to start

2009-11-22T01:45:16Z

On Sat, 21 Nov 2009, John wrote:

> Firstly, the small size of the device means that it is very awkward to
> pull the dongle out once it's pushed in. When restarting the computer,
> is it sufficient to reboot(8) then power the machine off, wait a few
> seconds then power it back up? I mean does this properly reset the
> dongle, or does it truly have to be pulled out and pushed back in for
> the system to re-read it?

power down should be sufficient to reset the device entirely, internal
state is rarely kept (sometimes there is nvram for link keys but thats
all)

it seems (to me) that your modules setup is ok, since the modules are
loaded when you want them to be

> potato# /etc/rc.d/bluetooth start ubt0
> /etc/rc.d/bluetooth: ERROR: Unable to setup Bluetooth stack for device
> ubt0

Not sure if I can really be of much assistance, does hccontrol show that
the node is up and running?

Also, before you do this can you run hcidump as root user? It might show
something happening (or not)

Then, you could enable some kind of debugging in the startup script (use
sh -x at least) to see where it was failing

iain

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Problems getting the bluetooth stack to start

2009-11-21T10:43:36Z

Hello list,

my system:
7.2-RELEASE-p4 FreeBSD 7.2-RELEASE-p4 #1: Sat Nov 21 16:55:14 GMT 2009
root@potato:/usr/src/sys/amd64/compile/POTATO amd64

The kernel is generic apart from ehci being commented out. Sources
cvsupped earlier today from cvsup.uk.freebsd.org

I'm trying to get my sony erikson T610 to work with a nexus bluemicro
pro dongle on 7.2-R. The galling thing is, I got it to work once, but it
was late and I forgot how. I can't get the bluetooth stack to start, and
I'm following the instructins as per the handbook

Firstly, the small size of the device means that it is very awkward to
pull the dongle out once it's pushed in. When restarting the computer,
is it sufficient to reboot(8) then power the machine off, wait a few
seconds then power it back up? I mean does this properly reset the
dongle, or does it truly have to be pulled out and pushed back in for
the system to re-read it?

I've tried various suggestions while googling trying to find the answer.
One suggestion (with a different device) was to disable ehci and
recompile the kernel, which I've done. No difference. Another was to
stop the bluetooth script before starting it:

potato# /etc/rc.d/bluetooth start ubt0
/etc/rc.d/bluetooth: ERROR: Unable to setup Bluetooth stack for device
ubt0
potato# /etc/rc.d/bluetooth stop ubt0
potato# /etc/rc.d/bluetooth start ubt0
/etc/rc.d/bluetooth: ERROR: Unable to setup Bluetooth stack for device
ubt0

I find that when the system boots up, only one bluetooth-related kernel module is loaded:

potato# kldstat
Id Refs Address Size Name
1 9 0xffffffff80100000 768b98 kernel
2 1 0xffffffff80869000 7128 geom_concat.ko
3 1 0xffffffff80871000 9808 ng_ubt.ko
4 2 0xffffffff8087b000 14e08 netgraph.ko
5 1 0xffffffff80a22000 978 pflog.ko
6 1 0xffffffff80a23000 2ae8c pf.ko
7 1 0xffffffff80a4e000 aa7a fuse.ko
8 1 0xffffffff80a59000 82e accf_http.ko

potato# ngctl li
There are 1 total nodes:
Name: ngctl1196 Type: socket ID: 00000001 Num hooks:0

When I try to start the stack, this happens:

potato# /etc/rc.d/bluetooth start ubt0
/etc/rc.d/bluetooth: ERROR: Unable to setup Bluetooth stack for device
ubt0
potato# /etc/rc.d/bluetooth stop ubt0
potato# /etc/rc.d/bluetooth start ubt0
/etc/rc.d/bluetooth: ERROR: Unable to setup Bluetooth stack for device
ubt0

let's do a kldstat again:

potato# kldstat
Id Refs Address Size Name
1 14 0xffffffff80100000 768b98 kernel
2 1 0xffffffff80869000 7128 geom_concat.ko
3 1 0xffffffff80871000 9808 ng_ubt.ko
4 6 0xffffffff8087b000 14e08 netgraph.ko
5 1 0xffffffff80a22000 978 pflog.ko
6 1 0xffffffff80a23000 2ae8c pf.ko
7 1 0xffffffff80a4e000 aa7a fuse.ko
8 1 0xffffffff80a59000 82e accf_http.ko
9 1 0xffffffff80a5a000 19ba ng_socket.ko
10 4 0xffffffff80a5c000 89a ng_bluetooth.ko
11 1 0xffffffff80a5d000 8b1a ng_hci.ko
12 1 0xffffffff80a66000 b498 ng_l2cap.ko
13 1 0xffffffff80a72000 12c9e ng_btsocket.ko

loads of stuff now.

potato# ngctl li
There are 4 total nodes:
Name: btsock_l2c_raw Type: btsock_l2c_raw ID: 00000003 Num hooks: 0
Name: btsock_hci_raw Type: btsock_hci_raw ID: 00000002 Num hooks: 0
Name: btsock_l2c Type: btsock_l2c ID: 00000004 Num hooks: 0
Name: ngctl1781 Type: socket ID: 00000012 Num hooks: 0

but still I cannot get the stack to work! I have also tried commenting
out the ubt* stuff in devd.conf and rebooting, no difference. The little
light on the dongle flashes merrily, but if the stack won't start I
won't be able to do anything.

Can anyone help please?

--
John - comp dot john at googlemail dot com
OpenBSD firewall | FreeBSD desktop | Ubuntu Karmic laptop
GPG: 0xF08A33C5
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: Visibility

2009-11-16T12:33:07Z

Helge Oldach wrote:

> Guido Falsi wrote on Mon, 16 Nov 2009 17:50:07 +0100 (CET):
>> Today I was playing around a little with bluetooth and I had the fancy
>> to just make my FreeBSD system not visible via bluetooth by default.
>>
>> I could not find any configuration to do this, nor any instruction in
>> the manual.
>
> Is this what you want?
>
> toor@sep / # cat /etc/bluetooth/ubt0.conf
> discoverable="NO"
> toor@sep / #

Thanks! This is exactly what I was looking for. Checking the defaults
directory did not occur to me :)

Thanks also to Iain, His suggestion showed me how to enable visibility
on request, which can be quite useful!

Thanks again, sorry for the obvious question.

--
Guido Falsi <mad@...>
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

iPhone and bcm-2045b

2009-11-16T09:37:26Z

Hi,
A friend of mine just got an iPhone 3G, he was here the other day and I
tried to probe the device using my laptop (Thinkpad T60 with NetBSD) which
has bluetooth built in (a Broadcom BCM2045B connected via internal USB)

I managed to discover it ok, but when I opened a connection to the device,
my controller detached from the USB bus!

I can't see that its NetBSD specifically (at least, not the Bluetooth
part) but I don't have any other OS installed to test. Next time he comes
around I'll sniff with hcidump to see how far the connection gets, and
also try it with an external 2.1 CSR dongle that I have.

In the meantime, I wonder if anybody else has had any issues connecting to
an iPhone 3G?

iain

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: Visibility

2009-11-16T09:28:01Z

On Mon, 16 Nov 2009, Guido Falsi wrote:

> Today I was playing around a little with bluetooth and I had the fancy
> to just make my FreeBSD system not visible via bluetooth by default.
>
> I could not find any configuration to do this, nor any instruction in
> the manual.
>
> Is there a simple way to do this? is this supported(no problem in
> writing small scripts or code snippets...) or simply impossibile without
> major tinkering with the stack?
>
> Thanks in advance, this is just an idea I had, nothing urgent!

You need to have "Inquiry Scan" off. You can switch this manually with
hccontrol(8) but I don't know how to make it so that it isn't turned on,
perhaps using bluetooth_device_discoverable=NO in your /etc/rc.conf?

iain

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Visibility

2009-11-16T08:50:07Z

Today I was playing around a little with bluetooth and I had the fancy
to just make my FreeBSD system not visible via bluetooth by default.

I could not find any configuration to do this, nor any instruction in
the manual.

Is there a simple way to do this? is this supported(no problem in
writing small scripts or code snippets...) or simply impossibile without
major tinkering with the stack?

Thanks in advance, this is just an idea I had, nothing urgent!

--
Guido Falsi <mad@...>
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: bluetooth.h, c++ include error

2009-10-26T10:43:35Z

On Tue, Oct 6, 2009 at 8:48 AM, Masoom Shaikh <masoom.shaikh@...> wrote:

>
>
> On Mon, Oct 5, 2009 at 5:26 PM, Iain Hibbert <plunky@...> wrote:
>>
>> On Mon, 5 Oct 2009, Masoom Shaikh wrote:
>>
>> > what i am requesting is just variable name change ;-) my tree is at
>> > revision 197571, thus the patch i am presenting is also generated
>> > against it reject this if you want patch against latest revision
>> >
>> > renamed second argument of bt_devfilter() from 'new' to 'dfnew'. df for
>> > device filter to maintain uniformity renamed third argument from 'old'
>> > to 'dfold'.
>>
>> style(9) mentions that for kernel include files exposed to userland
>> (technically, this is always in userland but perhaps the same rules apply)
>> the prototype tag arguments should be made safe, eg starting with a _
>> perhaps for much this reason..
>>
>> so, my suggestion would be to use _new and _old etc..
>>
>> iain
>>
>> (or, just don't use them, they are not especially required :)
>>
> so this shall work
> i just renamed bt_devfilter() variable names, because it was the one which
> caused trouble
> will someone commit this ?

i just committed fix for this.

thanks
max
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: bluetooth.h, c++ include error

2009-10-06T08:48:02Z

On Mon, Oct 5, 2009 at 5:26 PM, Iain Hibbert <plunky@...> wrote:

> On Mon, 5 Oct 2009, Masoom Shaikh wrote:
>
> > what i am requesting is just variable name change ;-) my tree is at
> > revision 197571, thus the patch i am presenting is also generated
> > against it reject this if you want patch against latest revision
> >
> > renamed second argument of bt_devfilter() from 'new' to 'dfnew'. df for
> > device filter to maintain uniformity renamed third argument from 'old'
> > to 'dfold'.
>
> style(9) mentions that for kernel include files exposed to userland
> (technically, this is always in userland but perhaps the same rules apply)
> the prototype tag arguments should be made safe, eg starting with a _
> perhaps for much this reason..
>
> so, my suggestion would be to use _new and _old etc..
>
> iain
>
> (or, just don't use them, they are not especially required :)
>
>
>
>
> so this shall work

i just renamed bt_devfilter() variable names, because it was the one which
caused trouble
will someone commit this ?

Index: bluetooth.h
===================================================================
--- bluetooth.h (revision 197804)
+++ bluetooth.h (working copy)
@@ -163,8 +163,8 @@
int bt_devsend (int s, uint16_t opcode, void *param, size_t plen);
ssize_t bt_devrecv (int s, void *buf, size_t size, time_t to);
int bt_devreq (int s, struct bt_devreq *r, time_t to);
-int bt_devfilter(int s, struct bt_devfilter const *new,
- struct bt_devfilter *old);
+int bt_devfilter(int s, struct bt_devfilter const *_new,
+ struct bt_devfilter *_old);
void bt_devfilter_pkt_set(struct bt_devfilter *filter, uint8_t
type);
void bt_devfilter_pkt_clr(struct bt_devfilter *filter, uint8_t
type);
int bt_devfilter_pkt_tst(struct bt_devfilter const *filter, uint8_t
type);
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: bluetooth.h, c++ include error

2009-10-05T10:26:37Z

On Mon, 5 Oct 2009, Masoom Shaikh wrote:

> what i am requesting is just variable name change ;-) my tree is at
> revision 197571, thus the patch i am presenting is also generated
> against it reject this if you want patch against latest revision
>
> renamed second argument of bt_devfilter() from 'new' to 'dfnew'. df for
> device filter to maintain uniformity renamed third argument from 'old'
> to 'dfold'.

style(9) mentions that for kernel include files exposed to userland
(technically, this is always in userland but perhaps the same rules apply)
the prototype tag arguments should be made safe, eg starting with a _
perhaps for much this reason..

so, my suggestion would be to use _new and _old etc..

iain

(or, just don't use them, they are not especially required :)

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: bluetooth.h, c++ include error

2009-10-05T09:58:10Z

On Sun, Oct 4, 2009 at 6:26 PM, Maksim Yevmenkin <maksim.yevmenkin@...
> wrote:

> On Sun, Oct 4, 2009 at 10:42 AM, Masoom Shaikh <masoom.shaikh@...>
> wrote:
> > posting here after @current and @questions ignored
> >
> > Hello,
> >
> > while going through bluetooth.h, I observed
> >
> > int bt_devfilter(int s, struct bt_devfilter const *new, struct
> bt_devfilter
> > *old);
> >
> > this line appears in bluetooth.h @ line # 166, rv197571
> >
> > see ? the variable named 'new' will cause trouble to c++ source files
> > i honestly believe this is not intentional, c++ programmers will have to
> > perform some acrobats to get past this
>
>
> patches are welcome
>
> thanks,
> max
>

what i am requesting is just variable name change ;-)
my tree is at revision 197571, thus the patch i am presenting is also
generated against it
reject this if you want patch against latest revision

renamed second argument of bt_devfilter() from 'new' to 'dfnew'. df for
device filter
to maintain uniformity renamed third argument from 'old' to 'dfold'.

the patch follows

Index: bluetooth.h
===================================================================
--- bluetooth.h (revision 197571)
+++ bluetooth.h (working copy)
@@ -28,7 +28,7 @@
* SUCH DAMAGE.
*
* $Id: bluetooth.h,v 1.5 2003/09/14 23:28:42 max Exp $
- * $FreeBSD$
+ * $FreeBSD: stable/8/lib/libbluetooth/bluetooth.h 191388 2009-04-22
15:50:03Z emax $
*/

#ifndef _BLUETOOTH_H_
@@ -163,8 +163,8 @@
int bt_devsend (int s, uint16_t opcode, void *param, size_t plen);
ssize_t bt_devrecv (int s, void *buf, size_t size, time_t to);
int bt_devreq (int s, struct bt_devreq *r, time_t to);
-int bt_devfilter(int s, struct bt_devfilter const *new,
- struct bt_devfilter *old);
+int bt_devfilter(int s, struct bt_devfilter const *dfnew,
+ struct bt_devfilter *dfold);
void bt_devfilter_pkt_set(struct bt_devfilter *filter, uint8_t
type);
void bt_devfilter_pkt_clr(struct bt_devfilter *filter, uint8_t
type);
int bt_devfilter_pkt_tst(struct bt_devfilter const *filter, uint8_t
type);
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: bluetooth.h, c++ include error

2009-10-04T11:26:56Z

On Sun, Oct 4, 2009 at 10:42 AM, Masoom Shaikh <masoom.shaikh@...> wrote:

> posting here after @current and @questions ignored
>
> Hello,
>
> while going through bluetooth.h, I observed
>
> int bt_devfilter(int s, struct bt_devfilter const *new, struct bt_devfilter
> *old);
>
> this line appears in bluetooth.h @ line # 166, rv197571
>
> see ? the variable named 'new' will cause trouble to c++ source files
> i honestly believe this is not intentional, c++ programmers will have to
> perform some acrobats to get past this

patches are welcome

thanks,
max
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

bluetooth.h, c++ include error

2009-10-04T10:42:07Z

posting here after @current and @questions ignored

Hello,

while going through bluetooth.h, I observed

int bt_devfilter(int s, struct bt_devfilter const *new, struct bt_devfilter
*old);

this line appears in bluetooth.h @ line # 166, rv197571

see ? the variable named 'new' will cause trouble to c++ source files
i honestly believe this is not intentional, c++ programmers will have to
perform some acrobats to get past this

thanks,
Masoom Shaikh
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: remote_name_request, using libbluetooth

2009-10-04T07:39:42Z

On Sun, Oct 4, 2009 at 7:57 AM, Iain Hibbert <plunky@...> wrote:

> On Sun, 4 Oct 2009, Masoom Shaikh wrote:
>
> > > On Sat, 3 Oct 2009, Masoom Shaikh wrote:
> > >
> > > > int status = bt_devreq( s, &request, 0);
> >
> > yes, i did forgot to say where it failed, devices are inquired properly
> no
> > issues there
> > but bt_devreq() fails with status == -1
>
> ah, I see.. you should put a non-zero timeout value there (waiting for 0
> seconds will not leave enough time for the response :)
>
> if you use eg
>
> if (status == -1)
> err(1, "bt_devreq (RemoteNameRequest)")
>
> then you would see an "Operation timed out" message.. I should think 5
> seconds would be reasonable..
>
> regards,
> iain
>
>
> u got the bug by neck, and i would have slit its neck by the time this mail

is in your inbox ;-)
that was the exact error, timeout value of 5 didn't work for me, thus am
using 60, and it works
thanks Iain
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: remote_name_request, using libbluetooth

2009-10-03T21:56:59Z

On Sat, Oct 3, 2009 at 8:13 PM, Iain Hibbert <plunky@...> wrote:

> On Sat, 3 Oct 2009, Masoom Shaikh wrote:
>
> > Hi,
> >
> > today i spent ages hacking hccontrol to do something similar it does when
> > invoked as below
> >
> > hccontrol remote_name_request <BD_ADDR>
> >
> > somehow i couldn't succeed ;-(
>
> you didn't say exactly the failure you have reached, did you use hcidump
> to check the actions?
>
> > what really i want is to search for devices and request their names
> > i have a ruby script which does the same by using hccontrol and
> manipulating
> > it console output
> > but am interested in C version.
> >
> > pasted below is full source.
> >
> > #include <stdio.h>
> > #include <stdlib.h>
> > #include <errno.h>
> > #include <string.h>
> > #include <bluetooth.h>
> >
> > /*
> > * removes duplicate entries from result and returns the new size
> > * free()'s the original array, result is calloc()ed
> > * TODO: implement in a better way
> > */
> > int do_uniq( const int size, struct bt_devinquiry** result)
> > {
> > struct bt_devinquiry* newResult = (struct bt_devinquiry*)calloc(
> size,
> > sizeof(struct bt_devinquiry));
> >
> > struct bt_devinquiry* srcCurr = *result;
> > struct bt_devinquiry* dstCurr = newResult;
> >
> > int count = 0;
> > int index = 0;
> > for ( ; index < size; ++index)
> > {
> > int j = 0;
> > int found = 0;
> > while ( j < count)
> > {
> > if ( bdaddr_same( &( newResult[j++].bdaddr), &(
> > srcCurr->bdaddr)))
> > {
> > found = 1;
> > break;
> > }
> > }
> >
> > if ( !found)
> > {
> > *dstCurr = *srcCurr;
> > ++dstCurr;
> > ++count;
> > }
> > ++srcCurr;
> > }
> > free(*result);
> > *result = newResult;
> > return count;
> > }
> >
> > int main( int argc, char* argv[])
> > {
> > /* search devices */
> > struct bt_devinquiry* result = 0;
> > int num = bt_devinquiry( 0, 0, 0, &result);
> > if ( num <= 0)
> > {
> > if ( h_errno)
>
> what is h_errno? I think these checks are uninvolved with the code at
> hand..
>
> > herror( "no devices found");
> > else
> > printf( "no devices found\n");
> > return num;
> > }
> > /* remove duplicate entries */
> > num = do_uniq( num, &result);
> > printf( "%d device(s) found\n", num);
>
> also, I don't think the do_uniq() step should be necessary -- we did
> discuss it at least and in the version i wrote for NetBSD it doesn't
> return duplicate results
>
> > /* try to query device's name */
> > int s = bt_devopen( "ubt0hci");
> > if ( s == -1)
> > {
> > if ( h_errno)
> > herror( "bt_devopen error\n");
> > else
> > printf( "bt_devopen error\n");
> > return -1;
> > }
> > int i = 0;
> > for ( ; i < num; ++i)
> > {
> > struct bt_devreq request;
> > memset( &request, 0, sizeof(request));
> > request.opcode = NG_HCI_OPCODE(NG_HCI_OGF_LINK_CONTROL,
> > NG_HCI_OCF_REMOTE_NAME_REQ);
> > request.event = NG_HCI_EVENT_REMOTE_NAME_REQ_COMPL;
> >
> > ng_hci_remote_name_req_cp cp;
> > memset(&cp, 0, sizeof(cp));
> > bdaddr_copy( &cp.bdaddr, &result->bdaddr);
> > cp.page_scan_rep_mode = NG_HCI_SCAN_REP_MODE0;
> > cp.page_scan_mode = NG_HCI_MANDATORY_PAGE_SCAN_MODE;
> > request.cparam = (void*)&cp;
> > request.clen = sizeof(cp);
> >
> > char buffer[512];
> > memset( buffer, 0, 512);
> > request.rparam = (void*)buffer;
> > request.rlen = 512;
> >
> > int status = bt_devreq( s, &request, 0);
> > if ( status == 0)
> > {
> > ng_hci_event_pkt_t *e = (ng_hci_event_pkt_t*)buffer;
> > ng_hci_remote_name_req_compl_ep *ep =
> > (ng_hci_remote_name_req_compl_ep*)(e + 1);
> > printf( "status: %d\n", ep->status);
> > printf( "name: %s\n", ep->name);
> > }
>
> I think your problem might be here. The event header is not returned, only
> the remote_name_req_compl_ep event packet.. (I would have used that
> directly in the rparam field rather than a separate buffer -- bt_devreq
> will not overflow the given space)
>
> > else if (status == -1)
> > {
> > if ( h_errno)
> > herror( "bt_devreq error\n");
> > else
> > printf( "bt_devreq error\n");
> > }
> > else
> > {
> > printf("bt_devreq unknown return value\n");
> > }
> > }
> > bt_devclose(s);
> > return 0;
> > }
>
> iain
>
>
> thanks for replying

yes, i did forgot to say where it failed, devices are inquired properly no
issues there
but bt_devreq() fails with status == -1

comment taken
i will remove the separate buffer, thank you again

here is the hcidump output

HCIDump - HCI packet analyzer ver 1.5
device: any snap_len: 65535 filter: 0xffffffffffffffff
< HCI Command: Inquiry(0x01|0x0001) plen 5
> HCI Event: Command Status(0x0f) plen 4
> HCI Event: Inquiry Result(0x02) plen 15
> HCI Event: Inquiry Result(0x02) plen 15
> HCI Event: Inquiry Complete(0x01) plen 1
< HCI Command: Remote Name Request(0x01|0x0019) plen 10
> HCI Event: Command Status(0x0f) plen 4
> HCI Event: Remote Name Req Complete(0x07) plen 255

i guess the last line indicates that we does receive device name but some
how bt_devreq() returns fails
i will poke around bt_devreq() for clues
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

remote_name_request, using libbluetooth

2009-10-03T07:06:10Z

Hi,

today i spent ages hacking hccontrol to do something similar it does when
invoked as below

hccontrol remote_name_request <BD_ADDR>

somehow i couldn't succeed ;-(
what really i want is to search for devices and request their names
i have a ruby script which does the same by using hccontrol and manipulating
it console output
but am interested in C version.

pasted below is full source.

#include <stdio.h>
#include <stdlib.h>
#include <errno.h>
#include <string.h>
#include <bluetooth.h>

/*
* removes duplicate entries from result and returns the new size
* free()'s the original array, result is calloc()ed
* TODO: implement in a better way
*/
int do_uniq( const int size, struct bt_devinquiry** result)
{
struct bt_devinquiry* newResult = (struct bt_devinquiry*)calloc( size,
sizeof(struct bt_devinquiry));

struct bt_devinquiry* srcCurr = *result;
struct bt_devinquiry* dstCurr = newResult;

int count = 0;
int index = 0;
for ( ; index < size; ++index)
{
int j = 0;
int found = 0;
while ( j < count)
{
if ( bdaddr_same( &( newResult[j++].bdaddr), &(
srcCurr->bdaddr)))
{
found = 1;
break;
}
}

if ( !found)
{
*dstCurr = *srcCurr;
++dstCurr;
++count;
}
++srcCurr;
}
free(*result);
*result = newResult;
return count;
}

int main( int argc, char* argv[])
{
/* search devices */
struct bt_devinquiry* result = 0;
int num = bt_devinquiry( 0, 0, 0, &result);
if ( num <= 0)
{
if ( h_errno)
herror( "no devices found");
else
printf( "no devices found\n");
return num;
}
/* remove duplicate entries */
num = do_uniq( num, &result);
printf( "%d device(s) found\n", num);

/* try to query device's name */
int s = bt_devopen( "ubt0hci");
if ( s == -1)
{
if ( h_errno)
herror( "bt_devopen error\n");
else
printf( "bt_devopen error\n");
return -1;
}
int i = 0;
for ( ; i < num; ++i)
{
struct bt_devreq request;
memset( &request, 0, sizeof(request));
request.opcode = NG_HCI_OPCODE(NG_HCI_OGF_LINK_CONTROL,
NG_HCI_OCF_REMOTE_NAME_REQ);
request.event = NG_HCI_EVENT_REMOTE_NAME_REQ_COMPL;

ng_hci_remote_name_req_cp cp;
memset(&cp, 0, sizeof(cp));
bdaddr_copy( &cp.bdaddr, &result->bdaddr);
cp.page_scan_rep_mode = NG_HCI_SCAN_REP_MODE0;
cp.page_scan_mode = NG_HCI_MANDATORY_PAGE_SCAN_MODE;
request.cparam = (void*)&cp;
request.clen = sizeof(cp);

char buffer[512];
memset( buffer, 0, 512);
request.rparam = (void*)buffer;
request.rlen = 512;

int status = bt_devreq( s, &request, 0);
if ( status == 0)
{
ng_hci_event_pkt_t *e = (ng_hci_event_pkt_t*)buffer;
ng_hci_remote_name_req_compl_ep *ep =
(ng_hci_remote_name_req_compl_ep*)(e + 1);
printf( "status: %d\n", ep->status);
printf( "name: %s\n", ep->name);
}
else if (status == -1)
{
if ( h_errno)
herror( "bt_devreq error\n");
else
printf( "bt_devreq error\n");
}
else
{
printf("bt_devreq unknown return value\n");
}
}
bt_devclose(s);
return 0;
}
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Invitation to connect on LinkedIn

2009-09-02T15:07:37Z

LinkedIn
------------

Hello, I am the host of the Power Networking Events!!!I'd like to add you to my professional network on LinkedIn.

- Matthew

Accept Matthew McCarthy's invite:
https://www.linkedin.com/e/isd/714708318/x_AxcQiz/

------
(c) 2009, LinkedIn Corporation

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

obexapp 1.4.12

2009-08-20T15:06:07Z

hello,

obexapp 1.4.12 can be downloaded from

http://www.geocities.com/m_evmenkin/obexapp-1.4.12.tar.gz

the biggest change is support for virtual per-device root folders
(mostly contributed by mi@) also few patches from other people. sorry
i can not recall of them at the moment, but they all get credit for
their work.

thanks,
max
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Order To New Zealand

2009-07-18T14:20:51Z

<<< No Message Collected >>>
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Atheros wireless device driver developer

2009-07-14T12:40:52Z

Hello Everyone,

I have an opportunity for a device driver developer with a heavy wireless
Atheros background.

Please let me know if anyone is interested or might know someone.

Please see job description below.

Thank you very much....

-Dave

Senior Software Engineer with WLAN Device Driver Development Experience

The ideal candidate will have several years of communication experience as
well as experience programming low level hardware drivers. Requirements:

* BSCS or BSEE or relevant experience

* 5+ years of experience in development of WLAN device drivers

* Fluency in coding and debugging C

* Experience with Atheros drivers a plus

* Expertise in one or more of these protocols:

§ 802.11

§ ATM

§ Sonet/SDH

§ NDIS

§ Bluetooth

§ Ethernet, GBit Ethernet

* Experience with one or more of the following operating systems:

§ MS Windows, WinCE

§ Linux

§ Embedded RTOS

David A. Horvay
Sr. Account Executive
Technology Solutions Division
MRINetwork Ultimate Placements, LLC
One Park Centre Drive, Suite 305A
TF:877-334-0285 ext. 202
dhorvay@...
<blocked::blocked::http://www.linkedin.com/in/davidhorvay>
http://www.linkedin.com/in/davidhorvay
www.MRINetwork.com <blocked::blocked::http://www.mrinetwork.com/>
<blocked::blocked::http://www.mrinetwork.com/>
BUILDING THE HEART OF BUSINESS (TM)

Please understand my mission at MRI Ultimate Placements is to partner with
those select clients where there is a philosophical fit. My goal has never
been to be all things to all people. As a client-focused search consultant
I evaluate each potential assignment based on alignment with my area of
expertise and the timing and urgency of each search.

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: RFC: obexapp - virtual root folder for each device

2009-07-14T09:26:22Z

2009/7/13 Mikhail T. <mi+thun@...>:
> Maksim, I just tried building the most recent obexapp from port (1.4.11)
> and was surprised to find out, that the feature, which we spent so much
> effort rehashing back in April, is still not there... What's holding it
> up? Thanks! Yours,

thanks for the ping. yes, terribly sorry, i'm completely swamped with
$realjob :( i'll try to get to it as soon as i can.

max
_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: RFC: obexapp - virtual root folder for each device

2009-07-13T21:30:54Z

Maksim, I just tried building the most recent obexapp from port (1.4.11)
and was surprised to find out, that the feature, which we spent so much
effort rehashing back in April, is still not there... What's holding it
up? Thanks! Yours,

-mi

_______________________________________________
freebsd-bluetooth@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-bluetooth
To unsubscribe, send any mail to "freebsd-bluetooth-unsubscribe@..."

Re: btpand uses a lot of CPU

2009-07-06T04:32:44Z

On Sat, 4 Jul 2009, Alexander Motin wrote:
> Daniel O'Connor wrote:
> > I have btpand working OK, however I notice that it uses a lot of
> > CPU (20-70% of a Pentium-M 1.4GHz) even when idle.
>
> What system version do you use? There was a bug causing high CPU
> usage fixed few months ago.

Ah yes, it is a bit stale, I updated event.c and now it's all good.

Thanks!

--
Daniel O'Connor software and network engineer
for Genesis Software - http://www.gsoft.com.au
"The nice thing about standards is that there
are so many of them to choose from."
-- Andrew Tanenbaum
GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C

signature.asc (195 bytes) Download Attachment