Send connman mailing list submissions to
[email protected]
To subscribe or unsubscribe via the World Wide Web, visit
https://lists.01.org/mailman/listinfo/connman
or, via email, send a message with subject or body 'help' to
[email protected]
You can reach the person managing the list at
[email protected]
When replying, please edit your Subject line so it is more specific
than "Re: Contents of connman digest..."
Today's Topics:
1. Re: [PATCH] technology: Allow raw key for tethering (i.e. 64
bytes in hexadecimal representation) (Patrik Flykt)
2. Re: [PATCH] technology: Allow raw key for tethering (i.e. 64
bytes in hexadecimal representation) (Jose Blanquicet)
3. dhcp failure and link local (Prasant J)
4. [PATCH v2] technology: Allow raw key for tethering (i.e. 64
bytes in hexadecimal representation) (Jose Blanquicet)
5. RE: [PATCH v2] nat: Remember previous IPv4 forwarding value
(Blanquicet-Melendez Jose (MM))
6. Re: dhcp failure and link local (Patrik Flykt)
7. Re: [PATCH v2] technology: Allow raw key for tethering (i.e.
64 bytes in hexadecimal representation) (Patrik Flykt)
8. [PATCH] nat: Open IPv4 forwarding file with proper mode
(Patrik Flykt)
----------------------------------------------------------------------
Message: 1
Date: Fri, 15 Apr 2016 09:28:58 +0300
From: Patrik Flykt <[email protected]>
To: Jose Blanquicet <[email protected]>, [email protected]
Subject: Re: [PATCH] technology: Allow raw key for tethering (i.e. 64
bytes in hexadecimal representation)
Message-ID: <[email protected]>
Content-Type: text/plain; charset="UTF-8"
On Thu, 2016-04-14 at 11:18 +0200, Jose Blanquicet wrote:
> Although the gsupplicant component allows to use a raw key when a new?
> network is being created through the function add_network_security_psk,?
> the technology component does not allow it because in order to set the?
> property TetheringPassphrase the string's length must be within the?
> range [8, 63], otherwise it will be taken as wrong value without check?
> if it is a raw key.
>
> ---
> ?src/technology.c | 10 +++++++++-
> ?1 file changed, 9 insertions(+), 1 deletion(-)
>
> diff --git a/src/technology.c b/src/technology.c
> index 1891d06..ffc34ca 100644
> --- a/src/technology.c
> +++ b/src/technology.c
> @@ -917,13 +917,21 @@ static DBusMessage *set_property(DBusConnection *conn,
> ? }
> ? } else if (g_str_equal(name, "TetheringPassphrase")) {
> ? const char *str;
> + int i;
> + size_t len;
>
> ? dbus_message_iter_get_basic(&value, &str);
> + len = strlen(str);
> ?
> ? if (technology->type != CONNMAN_SERVICE_TYPE_WIFI)
> ? return __connman_error_not_supported(msg);
> ?
> - if (strlen(str) < 8 || strlen(str) > 63)
> + if (len == 64) {
> + for (i = 0; i < 64; i++)
> + if (!g_ascii_isxdigit(str[i]))
> + return
> __connman_error_passphrase_required(msg);
> + }
> + else if (len < 8 || len > 63)
> ? return __connman_error_passphrase_required(msg);
> ?
> ? if (g_strcmp0(technology->tethering_passphrase, str) != 0) {
WiFi passphrase/password checking already exists in
service.c,?check_passphrase(). If that function were exported in
src/connman.h for example as __connman_service_check_passphrase(), the
password check can be done centrally in one
place.?CONNMAN_SERVICE_SECURITY_PSK is the only WiFi security type
currently used for tethering.
Cheers,
Patrik
------------------------------
Message: 2
Date: Fri, 15 Apr 2016 09:28:40 +0200
From: Jose Blanquicet <[email protected]>
To: Patrik Flykt <[email protected]>
Cc: [email protected]
Subject: Re: [PATCH] technology: Allow raw key for tethering (i.e. 64
bytes in hexadecimal representation)
Message-ID:
<CAFC8iJJxfU+wpF2DTr1W=t4iten571094lnsbziwreagyxa...@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8
Hi,
I agree, you are right. I'll send a second version of the patch
following your suggestions.
Best regards,
Jose Blanquicet
On Fri, Apr 15, 2016 at 8:28 AM, Patrik Flykt
<[email protected]> wrote:
> On Thu, 2016-04-14 at 11:18 +0200, Jose Blanquicet wrote:
>> Although the gsupplicant component allows to use a raw key when a new
>> network is being created through the function add_network_security_psk,
>> the technology component does not allow it because in order to set the
>> property TetheringPassphrase the string's length must be within the
>> range [8, 63], otherwise it will be taken as wrong value without check
>> if it is a raw key.
>>
>> ---
>> src/technology.c | 10 +++++++++-
>> 1 file changed, 9 insertions(+), 1 deletion(-)
>>
>> diff --git a/src/technology.c b/src/technology.c
>> index 1891d06..ffc34ca 100644
>> --- a/src/technology.c
>> +++ b/src/technology.c
>> @@ -917,13 +917,21 @@ static DBusMessage *set_property(DBusConnection *conn,
>> }
>> } else if (g_str_equal(name, "TetheringPassphrase")) {
>> const char *str;
>> + int i;
>> + size_t len;
>>
>> dbus_message_iter_get_basic(&value, &str);
>> + len = strlen(str);
>>
>> if (technology->type != CONNMAN_SERVICE_TYPE_WIFI)
>> return __connman_error_not_supported(msg);
>>
>> - if (strlen(str) < 8 || strlen(str) > 63)
>> + if (len == 64) {
>> + for (i = 0; i < 64; i++)
>> + if (!g_ascii_isxdigit(str[i]))
>> + return
>> __connman_error_passphrase_required(msg);
>> + }
>> + else if (len < 8 || len > 63)
>> return __connman_error_passphrase_required(msg);
>>
>> if (g_strcmp0(technology->tethering_passphrase, str) != 0) {
>
> WiFi passphrase/password checking already exists in
> service.c, check_passphrase(). If that function were exported in
> src/connman.h for example as __connman_service_check_passphrase(), the
> password check can be done centrally in one
> place. CONNMAN_SERVICE_SECURITY_PSK is the only WiFi security type
> currently used for tethering.
>
>
> Cheers,
>
> Patrik
>
------------------------------
Message: 3
Date: Fri, 15 Apr 2016 13:15:03 +0530
From: Prasant J <[email protected]>
To: [email protected]
Subject: dhcp failure and link local
Message-ID:
<CAA2DH4vmBNJ9=qf3pkdf-rxjcmdntv-mgecny+-2y3p-nz4...@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8
Hi,
I'm working with imx6q based system (using yocto jethro release). It
uses connman version 1.28.
I have observed is that when the dhcp server does not respond, connman
treis 6 times (after every 6 secs) and then falls back to ipv4ll
address.
Then after every 65-66 secs this is repeated (ipv4ll address does not change).
I want my embedded linux system to stop looking out for DHCP server
(stop sending DISCOVERY packets) after ipv4ll address is assigned.
Is it possible? If yes, do I need to change the code or is there a
configuration file?
Any inputs will be of help to me.
Regards, Pj
------------------------------
Message: 4
Date: Fri, 15 Apr 2016 09:47:30 +0200
From: Jose Blanquicet <[email protected]>
To: [email protected]
Subject: [PATCH v2] technology: Allow raw key for tethering (i.e. 64
bytes in hexadecimal representation)
Message-ID: <[email protected]>
Although the gsupplicant component allows to use a raw key when a new
network is being created through the function add_network_security_psk,
the technology component does not allow it because in order to set the
property TetheringPassphrase the string's length must be within the
range [8, 63], otherwise it will be taken as wrong value without check
if it is a raw key.
This patch uses Wi-Fi passphrase/password checking already implemented
in service.c, check_passphrase(), exporting it in src/connman.h. The
connman_service_security argument is hard-coded to
CONNMAN_SERVICE_SECURITY_PSK because it is the only Wi-Fi security type
currently supported for thethering.
---
src/connman.h | 2 ++
src/service.c | 4 ++--
src/technology.c | 6 ++++--
3 files changed, 8 insertions(+), 4 deletions(-)
diff --git a/src/connman.h b/src/connman.h
index c74ab91..e849ed8 100644
--- a/src/connman.h
+++ b/src/connman.h
@@ -781,6 +781,8 @@ void __connman_service_set_agent_identity(struct
connman_service *service,
int __connman_service_set_passphrase(struct connman_service *service,
const char *passphrase);
const char *__connman_service_get_passphrase(struct connman_service *service);
+int __connman_service_check_passphrase(enum connman_service_security security,
+ const char *passphrase);
int __connman_service_reset_ipconfig(struct connman_service *service,
enum connman_ipconfig_type type, DBusMessageIter *array,
enum connman_service_state *new_state);
diff --git a/src/service.c b/src/service.c
index a54e4da..768426b 100644
--- a/src/service.c
+++ b/src/service.c
@@ -2831,7 +2831,7 @@ void __connman_service_set_agent_identity(struct
connman_service *service,
service->agent_identity);
}
-static int check_passphrase(enum connman_service_security security,
+int __connman_service_check_passphrase(enum connman_service_security security,
const char *passphrase)
{
guint i;
@@ -2898,7 +2898,7 @@ int __connman_service_set_passphrase(struct
connman_service *service,
service->security != CONNMAN_SERVICE_SECURITY_8021X)
return -EINVAL;
- err = check_passphrase(service->security, passphrase);
+ err = __connman_service_check_passphrase(service->security, passphrase);
if (err < 0)
return err;
diff --git a/src/technology.c b/src/technology.c
index 1891d06..442557b 100644
--- a/src/technology.c
+++ b/src/technology.c
@@ -841,7 +841,7 @@ static DBusMessage *set_property(DBusConnection *conn,
struct connman_technology *technology = data;
DBusMessageIter iter, value;
const char *name;
- int type;
+ int type, err;
DBG("conn %p", conn);
@@ -923,7 +923,9 @@ static DBusMessage *set_property(DBusConnection *conn,
if (technology->type != CONNMAN_SERVICE_TYPE_WIFI)
return __connman_error_not_supported(msg);
- if (strlen(str) < 8 || strlen(str) > 63)
+ err =
__connman_service_check_passphrase(CONNMAN_SERVICE_SECURITY_PSK,
+ str);
+ if (err < 0)
return __connman_error_passphrase_required(msg);
if (g_strcmp0(technology->tethering_passphrase, str) != 0) {
--
1.9.1
------------------------------
Message: 5
Date: Fri, 15 Apr 2016 08:53:41 +0000
From: "Blanquicet-Melendez Jose (MM)"
<[email protected]>
To: Patrik Flykt <[email protected]>,
"[email protected]" <[email protected]>
Subject: RE: [PATCH v2] nat: Remember previous IPv4 forwarding value
Message-ID: <[email protected]>
Content-Type: text/plain; charset="iso-8859-1"
Hi,
After this patch was applied I have not been able to enable tethering. Through
debugging I found that function __connman_nat_enable returns the error: "Cannot
enable NAT -9/Bad file descriptor", this error is gotten when parameter
ip_forward is trying to be modified by editing the file with write() fucntion.
Trying to find the reason of the error I added the O_RDWR mode in open() to
check if it was a permissions problem but it did not work and it continue
giving the same error. Next, I tried to come back to fopen() instead of use
open() and it worked.
What do you think is happening? What is wrong when file
"/proc/sys/net/ipv4/ip_forward" is handled with the system calls
open/read/write instead of standard functions fopen/fprintf/fscanf?
Best regards,
Jose Blanquicet
-----Original Message-----
From: connman [mailto:[email protected]] On Behalf Of Patrik Flykt
Sent: marted? 12 aprile 2016 15:02
To: [email protected]
Subject: [PATCH v2] nat: Remember previous IPv4 forwarding value
When NAT is enabled, store the previous IPv4 forwarding setting so that it can
be restored to its former value when disabling NAT.
---
v2: Fix err = -errno and fix a few conditionals to be more readable
src/nat.c | 40 ++++++++++++++++++++++++++++++----------
1 file changed, 30 insertions(+), 10 deletions(-)
diff --git a/src/nat.c b/src/nat.c
index 063f085..b739e11 100644
--- a/src/nat.c
+++ b/src/nat.c
@@ -25,7 +25,10 @@
#endif
#include <errno.h>
-#include <stdio.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <unistd.h>
#include "connman.h"
@@ -42,20 +45,37 @@ struct connman_nat {
static int enable_ip_forward(bool enable) {
- FILE *f;
+ static char value = 0;
+ int f, err = 0;
- f = fopen("/proc/sys/net/ipv4/ip_forward", "r+");
- if (!f)
+ if ((f = open("/proc/sys/net/ipv4/ip_forward", O_CLOEXEC)) < 0)
return -errno;
- if (enable)
- fprintf(f, "1");
- else
- fprintf(f, "0");
+ if (!value) {
+ if (read(f, &value, sizeof(value)) < 0)
+ value = 0;
+ }
- fclose(f);
+ if (enable) {
+ char allow = '1';
- return 0;
+ if (write (f, &allow, sizeof(allow)) < 0)
+ err = -errno;
+ } else {
+ char deny = '0';
+
+ if (value)
+ deny = value;
+
+ if (write(f, &deny, sizeof(deny)) < 0)
+ err = -errno;
+
+ value = 0;
+ }
+
+ close(f);
+
+ return err;
}
static int enable_nat(struct connman_nat *nat)
--
2.8.0.rc3
_______________________________________________
connman mailing list
[email protected]
https://lists.01.org/mailman/listinfo/connman
------------------------------
Message: 6
Date: Fri, 15 Apr 2016 12:44:35 +0300
From: Patrik Flykt <[email protected]>
To: Prasant J <[email protected]>, [email protected]
Subject: Re: dhcp failure and link local
Message-ID: <[email protected]>
Content-Type: text/plain; charset="UTF-8"
On Fri, 2016-04-15 at 13:15 +0530, Prasant J wrote:
> I want my embedded linux system to stop looking out for DHCP server
> (stop sending DISCOVERY packets) after ipv4ll address is assigned.
> Is it possible? If yes, do I need to change the code or is there a
> configuration file?
Not really. Other users actually run into the opposite problem, which
was that their DHCP server was not immediately reachable, and wanted to
get out of a situation with link-local IPv4 decently quickly.
Cheers,
Patrik
------------------------------
Message: 7
Date: Fri, 15 Apr 2016 12:48:09 +0300
From: Patrik Flykt <[email protected]>
To: Jose Blanquicet <[email protected]>, [email protected]
Subject: Re: [PATCH v2] technology: Allow raw key for tethering (i.e.
64 bytes in hexadecimal representation)
Message-ID: <[email protected]>
Content-Type: text/plain; charset="UTF-8"
On Fri, 2016-04-15 at 09:47 +0200, Jose Blanquicet wrote:
> Although the gsupplicant component allows to use a raw key when a
> new?
> network is being created through the function
> add_network_security_psk,?
> the technology component does not allow it because in order to set
> the?
> property TetheringPassphrase the string's length must be within the?
> range [8, 63], otherwise it will be taken as wrong value without
> check?
> if it is a raw key.
>
> This patch uses Wi-Fi passphrase/password checking already
> implemented?
> in service.c, check_passphrase(), exporting it in src/connman.h. The?
> connman_service_security argument is hard-coded to?
> CONNMAN_SERVICE_SECURITY_PSK because it is the only Wi-Fi security
> type?
> currently supported for thethering.
Applied with whitespace scrubbing end of line?err =
__connman_service_check_passphrase(...
Patrik
------------------------------
Message: 8
Date: Fri, 15 Apr 2016 13:19:53 +0300
From: Patrik Flykt <[email protected]>
To: [email protected]
Subject: [PATCH] nat: Open IPv4 forwarding file with proper mode
Message-ID:
<[email protected]>
The file needs to be read and written to.
---
Yes, something wasn't right. Thanks to Jose for noticing this!
Patrik
src/nat.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/nat.c b/src/nat.c
index b739e11..33ae6df 100644
--- a/src/nat.c
+++ b/src/nat.c
@@ -48,7 +48,7 @@ static int enable_ip_forward(bool enable)
static char value = 0;
int f, err = 0;
- if ((f = open("/proc/sys/net/ipv4/ip_forward", O_CLOEXEC)) < 0)
+ if ((f = open("/proc/sys/net/ipv4/ip_forward", O_CLOEXEC | O_RDWR)) < 0)
return -errno;
if (!value) {
--
2.8.0.rc3
------------------------------
Subject: Digest Footer
_______________________________________________
connman mailing list
[email protected]
https://lists.01.org/mailman/listinfo/connman
------------------------------
End of connman Digest, Vol 6, Issue 11
**************************************
