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] service: Enhance
__connman_service_iterate_services() (Daniel Wagner)
2. Re: [PATCH 4/4] wps: add new WPS API for technology (Slava Monich)
3. Re: [PATCH 4/4] wps: add new WPS API for technology
(Daniel Wagner)
4. Re: [PATCH] gdhcp: Fix use of dhcp_client after free
(Daniel Wagner)
5. Re: [PATCH] gdhcp: Retry to get an IPv4ll ip even after
MAX_CONFLICTS (Daniel Wagner)
6. Re: [PATCH 10/10] vpn: Disconnect VPN provider when transport
is gone (Daniel Wagner)
7. Re: [PATCH 00/10] Disconnect VPN provider when transport is
gone (Daniel Wagner)
8. Re: [PATCH 4/4] wps: add new WPS API for technology (Slava Monich)
9. [PATCH] build: Remove multiply defined backtrace_sources
(Slava Monich)
10. Re: [PATCH 4/4] wps: add new WPS API for technology
(Daniel Wagner)
----------------------------------------------------------------------
Message: 1
Date: Tue, 18 Sep 2018 09:12:36 +0200
From: Daniel Wagner <[email protected]>
To: Slava Monich <[email protected]>, [email protected]
Subject: Re: [PATCH] service: Enhance
__connman_service_iterate_services()
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
On 09/17/2018 06:39 PM, Slava Monich wrote:
> On 14/09/18 16:11, Slava Monich wrote:
>> Let the callback to break the loop by returning non-zero value.
>> ---
>> src/connman.h | 3 ++-
>> src/service.c | 10 ++++------
>> 2 files changed, 6 insertions(+), 7 deletions(-)
>
> Please ignore this patch, I'm planning to submit a patch set which will
> supersede this one.
>
> Sorry about the noise.
No worries. I am slow as a slug anyway
------------------------------
Message: 2
Date: Tue, 18 Sep 2018 10:15:07 +0300
From: Slava Monich <[email protected]>
To: [email protected]
Subject: Re: [PATCH 4/4] wps: add new WPS API for technology
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
On 31/08/18 14:17, [email protected] wrote:
> This implementation is follow the the original patch.
> https://lists.01.org/pipermail/connman/2018-January/022367.html
>
> Signed-off-by: n-itaya <[email protected]>
> ---
> include/technology.h | 20 +++-
> src/technology.c | 233 +++++++++++++++++++++++++++++++++++++++++++
> 2 files changed, 252 insertions(+), 1 deletion(-)
>
> diff --git a/include/technology.h b/include/technology.h
> index 97db6607..febf6685 100644
> --- a/include/technology.h
> +++ b/include/technology.h
> @@ -34,10 +34,19 @@ extern "C" {
> * @short_description: Functions for handling technology details
> */
>
> +enum connman_technology_wps_mode {
> + CONNMAN_TECHNOLOGY_WPS_STA_MODE = 0,
> + CONNMAN_TECHNOLOGY_WPS_AP_MODE = 1,
> +};
> +
> struct connman_technology;
>
> int connman_technology_tethering_notify(struct connman_technology
> *technology,
> - bool enabled);
> + bool enabled);
> +
> +void connman_technology_wps_state_change_notify
> +(struct connman_technology *technology, const char *state);
> +
I'm wondering what connman "technology" is in the first place. connman
has wifi technology, cellular, bluetooth etc. There is even gps
technology. Which suggests that "technology" is something which all of
these have in common.
Now what would WPS stuff mean in context of, say, cellular, ethernet,
gps or vpn technology? I am thoroughly confused.
Cheers,
-Slava
------------------------------
Message: 3
Date: Tue, 18 Sep 2018 09:22:09 +0200
From: Daniel Wagner <[email protected]>
To: Slava Monich <[email protected]>, [email protected]
Subject: Re: [PATCH 4/4] wps: add new WPS API for technology
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
>> int connman_technology_tethering_notify(struct connman_technology
>> *technology,
>> - bool enabled);
>> + bool enabled);
>> +
>> +void connman_technology_wps_state_change_notify
>> +(struct connman_technology *technology, const char *state);
>> +
>
> I'm wondering what connman "technology" is in the first place. connman
> has wifi technology, cellular, bluetooth etc. There is even gps
> technology. Which suggests that "technology" is something which all of
> these have in common.
The technology API is a kind of abstraction for a device. As you know
there is no API for an individual device. Technology groups several
devices of the same kind together.
> Now what would WPS stuff mean in context of, say, cellular, ethernet,
> gps or vpn technology? I am thoroughly confused.
Yes, good point.
------------------------------
Message: 4
Date: Tue, 18 Sep 2018 09:22:53 +0200
From: Daniel Wagner <[email protected]>
To: Vivien Henriet <[email protected]>
Cc: [email protected]
Subject: Re: [PATCH] gdhcp: Fix use of dhcp_client after free
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
Hi Vivien,
On 09/12/2018 01:02 PM, Vivien Henriet wrote:
> ==6439==ERROR: AddressSanitizer: heap-use-after-free on address 0xb3d031f4 at
> pc 0x0002a86c bp 0xbe897f7c sp 0xbe897f74
> READ of size 4 at 0xb3d031f4 thread T0
> #0 0x2a86b (/usr/sbin/connmand+0x2a86b) ./gdhcp/client.c:1542
> switch_listening_mode
> #1 0x297ff (/usr/sbin/connmand+0x297ff) ./gdhcp/client.c:1392
> ipv4ll_stop
> #2 0x2a0fb (/usr/sbin/connmand+0x2a0fb) ./gdhcp/client.c:1462
> ipv4ll_recv_arp_packet
> #3 0x2f537 (/usr/sbin/connmand+0x2f537) ./gdhcp/client.c:2311
> listener_event
> 0xb3d031f4 is located 52 bytes inside of 320-byte region
> [0xb3d031c0,0xb3d03300)
> freed by thread T0 here:
> #0 0xb6a18483 in free (/usr/lib/libasan.so.3+0xbe483)
> #1 0x36acf (/usr/sbin/connmand+0x36acf) ./gdhcp/client.c:3252
> g_dhcp_client_unref
> #2 0x136d0b (/usr/sbin/connmand+0x136d0b) ./src/dhcp.c:89
> ipv4ll_stop_client
> #3 0x1387bb (/usr/sbin/connmand+0x1387bb) ./src/dhcp.c:295
> ipv4ll_lost_cb
> #4 0x2a0f3 (/usr/sbin/connmand+0x2a0f3) ./gdhcp/client.c:1458
> ipv4ll_recv_arp_packet
> #5 0x2f537 (/usr/sbin/connmand+0x2f537) ./gdhcp/client.c:2311
> listener_event
> #6 0xb68a2a33 in g_main_context_dispatch
> (/usr/lib/libglib-2.0.so.0+0x47a33)
> previously allocated by thread T0 here:
> #0 0xb6a188ef in calloc (/usr/lib/libasan.so.3+0xbe8ef)
> #1 0x27f2f (/usr/sbin/connmand+0x27f2f) ./gdhcp/client.c:1164
> #2 0x137f2b (/usr/sbin/connmand+0x137f2b) ./src/dhcp.c:203
> #3 0x138603 (/usr/sbin/connmand+0x138603) ./src/dhcp.c:271
> #4 0x334c3 (/usr/sbin/connmand+0x334c3) ./gdhcp/client.c:2818
> #5 0x31d0f (/usr/sbin/connmand+0x31d0f) ./gdhcp/client.c:2635
> #6 0xb68a3677 (/usr/lib/libglib-2.0.so.0+0x48677)
Nice catch. I suppose this one was not easy to find.
Patch applies.
Thanks,
Daniel
------------------------------
Message: 5
Date: Tue, 18 Sep 2018 09:24:23 +0200
From: Daniel Wagner <[email protected]>
To: Vivien Henriet <[email protected]>, [email protected]
Subject: Re: [PATCH] gdhcp: Retry to get an IPv4ll ip even after
MAX_CONFLICTS
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
Hi Vivien,
On 09/12/2018 02:15 PM, Vivien Henriet wrote:
> There is no reason to stop retry after MAX_CONFLICTS (10) tries.
> Do so will make the device unable to retrieve an IPv4ll ip on large network.
> The commit will make connman retry forever until it eventually got a non
> conflicting ip, honnoring the RATE_LIMIT_INTERVAL.
Thanks for the commit message. It really helps to understand what was
the motivation for the change. In a couple of weeks I wont remember why
I merged something :)
Patch applied.
Thanks,
Daniel
------------------------------
Message: 6
Date: Tue, 18 Sep 2018 09:31:39 +0200
From: Daniel Wagner <[email protected]>
To: Slava Monich <[email protected]>, [email protected]
Subject: Re: [PATCH 10/10] vpn: Disconnect VPN provider when transport
is gone
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
Hi Slava,
> @@ -558,6 +572,19 @@ static int connect_provider(struct connection_data
> *data, void *user_data,
> cb_data->path = g_strdup(data->path);
> }
>
> + if (transport) {
> + /* This is the service which (most likely) will be used
> + * as a transport for VPN connection */
I've changed this to the other comment style.
/*
*
*/
Patch applied.
Thanks,
Daniel
------------------------------
Message: 7
Date: Tue, 18 Sep 2018 09:32:09 +0200
From: Daniel Wagner <[email protected]>
To: Slava Monich <[email protected]>, [email protected]
Subject: Re: [PATCH 00/10] Disconnect VPN provider when transport is
gone
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
Hi Slava,
On 09/17/2018 06:48 PM, Slava Monich wrote:
> The purpose of this patchset is to make sure that VPN provider
> gets disconnected after switching between WiFi and mobile data.
> See the last patch for details.
All patches applied. Was easy to follow. Excellent work!
Thanks,
Daniel
------------------------------
Message: 8
Date: Tue, 18 Sep 2018 10:55:31 +0300
From: Slava Monich <[email protected]>
To: [email protected]
Cc: Daniel Wagner <[email protected]>
Subject: Re: [PATCH 4/4] wps: add new WPS API for technology
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
On 18/09/18 10:22, Daniel Wagner wrote:
>> I'm wondering what connman "technology" is in the first place. connman
>> has wifi technology, cellular, bluetooth etc. There is even gps
>> technology. Which suggests that "technology" is something which all of
>> these have in common.
>
> The technology API is a kind of abstraction for a device. As you know
> there is no API for an individual device. Technology groups several
> devices of the same kind together.
>
>> Now what would WPS stuff mean in context of, say, cellular, ethernet,
>> gps or vpn technology? I am thoroughly confused.
>
> Yes, good point.
>
Another thing is that if there are two wifi adapters, WPS stuff can be
happening simultaneously on both of them. Correct handling of WPS would
require some sort of "adapter" or "device" abstraction but that's
probably too much to ask.
To avoid throwing technology specific things together into one monstrous
interface (union instead of intersection), an ofono-style approach can
be used. ofono Modem has "Interfaces" property which is a list of other
interfaces that modem supports. connman Technology can have one too.
Wifi Technology would support additional WPS interface and other kinds
of Technology won't. That would make sense to me.
Cheers,
-Slava
------------------------------
Message: 9
Date: Tue, 18 Sep 2018 11:53:59 +0300
From: Slava Monich <[email protected]>
To: [email protected]
Subject: [PATCH] build: Remove multiply defined backtrace_sources
Message-ID: <[email protected]>
Makefile.am:58: warning: backtrace_sources multiply defined in condition
BACKTRACE ...
Makefile.am:37: ... 'backtrace_sources' previously defined here
---
Makefile.am | 4 ----
1 file changed, 4 deletions(-)
diff --git a/Makefile.am b/Makefile.am
index 7f5c5e5..d6dfbf1 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -54,10 +54,6 @@ else
stats_sources = src/nostats.c
endif
-if BACKTRACE
-backtrace_sources = src/backtrace.c
-endif
-
shared_sources = src/shared/util.h src/shared/util.c \
src/shared/netlink.h src/shared/netlink.c \
src/shared/arp.h src/shared/arp.c
--
1.9.1
------------------------------
Message: 10
Date: Tue, 18 Sep 2018 13:35:31 +0200
From: Daniel Wagner <[email protected]>
To: Slava Monich <[email protected]>, [email protected],
"Holtmann, Marcel" <[email protected]>
Subject: Re: [PATCH 4/4] wps: add new WPS API for technology
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8
Hi,
> Another thing is that if there are two wifi adapters, WPS stuff can be
> happening simultaneously on both of them. Correct handling of WPS would
> require some sort of "adapter" or "device" abstraction but that's
> probably too much to ask.
In the early days of ConnMan there was still the D-Bus Device API which
got removed. The main use case for ConnMan was though to be handhelds.
>From what I hear ConnMan is often found in embedded devices with
multiple interfaces. Obviously, for such cases the Technology API is not
up to the task. I suppose if we want to address this, it would be for 2.x
> To avoid throwing technology specific things together into one monstrous
> interface (union instead of intersection), an ofono-style approach can
> be used. ofono Modem has "Interfaces" property which is a list of other
> interfaces that modem supports. connman Technology can have one too.
> Wifi Technology would support additional WPS interface and other kinds
> of Technology won't. That would make sense to me.
So with 'Interfaces' in the Technology API we introduce a device
interface. I don't mind going in this direction. Especially since oFono
has the same pattern for this.
@Marcel: Any thoughs on this idea?
Thanks,
Daniel
------------------------------
Subject: Digest Footer
_______________________________________________
connman mailing list
[email protected]
https://lists.01.org/mailman/listinfo/connman
------------------------------
End of connman Digest, Vol 35, Issue 7
**************************************
