From: Daniel Wagner <wagi@monom.org>
To: Jupiter <jupiter.hce@gmail.com>
Cc: connman <connman@lists.linux.dev>
Subject: Re: Connman stability degraded upgrading new versions
Date: Mon, 20 Sep 2021 09:12:51 +0200 [thread overview]
Message-ID: <20210920071251.qwix47owh7ybajdc@beryllium.lan> (raw)
In-Reply-To: <CAA=hcWQTDvN-vh+UgXx8jpV--NPBrXWWsd1O6Sg=FVufPapDWw@mail.gmail.com>
Hi Jupiter
On Tue, Sep 14, 2021 at 04:58:15PM +1000, Jupiter wrote:
> When I used connman v1.35, it was very stable, it could switch to 4G
> LTE automatically and reliably when the WiFi was not available.
>
> After upgrading to version 1.37, that network automatic switch is
> gone, I have to restart connman to make it work.
Right, did anything else change in your setup? 1.37 is pretty outdated,
it was released early in 2019.
> After trying to upgrade to version 1.39 in the Hardknott branch, I
> cannot not even get WiFi and 4G LTE work, the WiFi network was
> assigned with a fake IP address 10.x.x.x, it could not get the 4G LTE
> interface.
An 10.x.x.x is not something ConnMan would do. It only auto assigns link
local addresses if the DHCP does not work. This IP seems to be handed
out by a DHCP server.
> I have to abandon the upgrading. I was told that connman
> 1.39 did not work well with Hardknott libraries, I have no idea who is
> managing the connman release on OE Yocto build.
I don't know either. This is the first time I hear of it.
> What has been changed in connman new versions?
Between 1.35 and 1.37 there a lot of changes. I really urge you to
update the latest version of ConnMan. I don't have time to debug older
versions of ConnMan.
> I must miss something,
> I was told that connman uses WPA supplicant for WiFi, I have used WPA
> supplicant v2.6 for the connman v1.37, could it be WPA_supplicant
> problem, or Yocto build problem or connman problem?
wpa_supplicant might be a problem but as we had a lot of changes between
the two version I really don't know.
> My apology, I don't have good debug skills to debug it, but if you
> could instruct me debug technique I'll provide more testing and debug
> information.
I see two option what you could try to do:
- Try to bissect the offending commit. This is pretty simple to do but
takes awhile.
- The first step is to enable the logs and try to match this with the
source code. This is hard and takes a long time.
Hope this helps,
Daniel
prev parent reply other threads:[~2021-09-20 7:13 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-14 6:58 Connman stability degraded upgrading new versions Jupiter
2021-09-20 7:12 ` Daniel Wagner [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210920071251.qwix47owh7ybajdc@beryllium.lan \
--to=wagi@monom.org \
--cc=connman@lists.linux.dev \
--cc=jupiter.hce@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).