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 0/6] VPN provider fixes and VPNC VPN agent support
(Jussi Laakkonen)
2. Re: [PATCH] network: Fix disconnect_callback crash on error
(Yasser)
----------------------------------------------------------------------
Message: 1
Date: Thu, 6 Jun 2019 13:31:23 +0300
From: Jussi Laakkonen <[email protected]>
To: Daniel Wagner <[email protected]>
Cc: [email protected]
Subject: Re: [PATCH 0/6] VPN provider fixes and VPNC VPN agent support
Message-ID: <[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
Hi Daniel,
On 6/5/19 3:33 PM, Daniel Wagner wrote:
> Hi Jussi,
>
> On Tue, May 28, 2019 at 03:47:28PM +0300, Jussi Laakkonen wrote:
>> This set of patches improves and fixes some issues with vpn-provider.c
>> and implements VPN agent support for VPNC.
>>
>> vpn-provider.c changes:
>> * Support plugin specific data in provider. Plugins may set pointer to
>> data for latter use, e.g., in case notify needs additional data.
>> * Support retrieval of setting string immutable status. The main
>> values within provider have the same status as the provider as other
>> strings have individual statuses. Some VPN plugins need to
>> temporarily cache credentials, but immutable ones should not be
>> cleared.
>> * Allow only DBUS_TYPE_STRING values in set_property()
>>
>> VPNC changes:
>> * Support VPN agent for setting authentication credentials. VPNC now
>> is able to retrieve credentials using agent, if any, and clears them
>> after writing the credentials to VPNC process, so VPN agent is used
>> also with next connect attempt and credentials are not stored in
>> provider.
>> * Detect authentication and connection errors which VPNC outputs to
>> stderr. If these are detected provider is notified with appropriate
>> error code.
>> * Check the D-Bus message arg types in notify. With invalid types
>> crash would be expected.
>
> Many thanks for your excellent work. The commit message are really
> helping understanding what's going on :)
>
You are welcome. For me writing the commit messages forces me to
(re-)think about the implementation as well and thus, helps in finding
possible errors there. So it is a win-win.
> All patches applied. Hopefully I got it right. Scream if I screwed it up!
>
Seems to be correct. :)
BR,
Jussi
------------------------------
Message: 2
Date: Thu, 6 Jun 2019 16:45:20 +0200
From: Yasser <[email protected]>
To: Daniel Wagner <[email protected]>
Cc: [email protected]
Subject: Re: [PATCH] network: Fix disconnect_callback crash on error
Message-ID:
<cakiau+3kbgqnfgpdzswc6djdf2dmptkxugpejmpsfjzzrck...@mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
Hi Daniel,
On Wed, Jun 5, 2019 at 1:41 PM Daniel Wagner <[email protected]> wrote:
>
> Hi Yasser,
>
> On Mon, May 27, 2019 at 11:32:58AM +0200, Yasser wrote:
> > We had a crash in connman_network_set_connected() when
> > disconnect_callback() was called but the value of wifi->network was
> > not checked. This has already been fixed for wifi.c. However
> > connman_network_set_connected() is called by other components where
> > the network parameter is not checked beforehand so we thought it might
> > be a good idea to verify in the function itself.
>
> (commit message should be written differently)
>
> So this code is not needed after all? I am relunctant to add checks
> which are not necessary and will paper over problems.
No, it is not needed. It seemed like a good idea to have consistency in the code
since this function is called in other files (bluetooth.c, ethernet.c, etc)
and there this check is not always done.
>
> >
> > diff --git a/src/network.c b/src/network.c
> > index 56fe24f..0486fad 100644
> > --- a/src/network.c
> > +++ b/src/network.c
> > @@ -1675,6 +1675,9 @@ void connman_network_set_error(struct
> > connman_network *network,
> > int connman_network_set_connected(struct connman_network *network,
> > bool connected)
> > {
> > + if (!network)
> > + return 0;
> > +
> > DBG("network %p connected %d/%d connecting %d associating %d",
> > network, network->connected, connected, network->connecting,
> > network->associating);
>
> BTW, the other patch I just applied was corrupted. This one looks also
> whitespace damaged.
>
Sorry about that, I will run "git diff --check" before sending a patch
next time.
> Thanks,
> Daniel
Thanks for your comments.
Yasser
------------------------------
Subject: Digest Footer
_______________________________________________
connman mailing list
[email protected]
https://lists.01.org/mailman/listinfo/connman
------------------------------
End of connman Digest, Vol 44, Issue 3
**************************************
