<div dir="ltr">Sorry Michael,<div><br></div><div>Haven't really had the time yet (medical emergency last weekend, on call with work this weekend... the Telco industry never sleeps :-) )</div><div><br></div><div>(The hiccup happened again on thursday, phone number became "+3365" instead of "+33651886877")</div><div><br></div><div>I'll be jobless after december 13th (no problem, got a year and a half of paid vacation to get paid ;-) ) so I'll have a little bit of time to flash the code in before visiting family.</div><div><br></div><div>JaXX./.</div><div><br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Sun, Dec 11, 2016 at 8:57 AM, Michael Balzer <span dir="ltr"><<a href="mailto:dexter@expeedo.de" target="_blank">dexter@expeedo.de</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
  
    
  
  <div text="#000000" bgcolor="#FFFFFF">
    Got the buffer overrun effect yesterday with the new release.<br>
    <br>
    As I now...<br>
    <br>
    <blockquote type="cite">
      <pre>- Keep volatile features over reset</pre>
    </blockquote>
    <br>
    ...it now also shows up with random data in features 0-7. Annoying.<br>
    <br>
    Julien, Pierre, any crash debug data collected yet?<br>
    <br>
    Regards,<br>
    Michael<div><div class="h5"><br>
    <br>
    <br>
    <div class="m_-5713992841334186999moz-cite-prefix">Am 29.11.2016 um 22:13 schrieb Julien
      Banchet:<br>
    </div>
    <blockquote type="cite">
      <div dir="ltr">
        <div>
          <div>
            <div>Thank you Michael,<br>
            </div>
            Will flash it this weekend!<br>
            <br>
          </div>
          MfG,<br>
        </div>
        JaXX./.<br>
      </div>
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On Tue, Nov 29, 2016 at 10:10 PM,
          Michael Balzer <span dir="ltr"><<a href="mailto:dexter@expeedo.de" target="_blank">dexter@expeedo.de</a>></span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div text="#000000" bgcolor="#FFFFFF"> Julien, Pierre,<br>
              <br>
              here's a firmware build with crash debug enabled:<br>
              <a class="m_-5713992841334186999m_-7214223298656431776moz-txt-link-freetext" href="https://dexters-web.de/f/tw-beta/OVMS-Twizy-3.8.2-crashdebug.zip" target="_blank">https://dexters-web.de/f/tw-be<wbr>ta/OVMS-Twizy-3.8.2-crashdebug<wbr>.zip</a><br>
              <br>
              It excludes the diag mode (serial interface command shell)
              but is otherwise complete.<br>
              <br>
              The crash debug logs to the server (when it is/becomes
              available), historical record type "*-OVM-DebugCrash".
              These records expire after 30 days, so you can collect
              some days.<br>
              <br>
              The crash count and data from the most recent crash will
              also be shown by the SMS/text command "DIAG".<br>
              <br>
              The crash records contain a crash count, crash reasons
              (flags from RCON) and the last checkpoint seen before the
              crash.<br>
              <br>
              Regards,<br>
              Michael
              <div>
                <div class="m_-5713992841334186999h5"><br>
                  <br>
                  <br>
                  <div class="m_-5713992841334186999m_-7214223298656431776moz-cite-prefix">Am
                    28.11.2016 um 22:37 schrieb Julien Banchet:<br>
                  </div>
                  <blockquote type="cite">
                    <div dir="ltr">
                      <div>
                        <div>
                          <div>Sure Michael,<br>
                            <br>
                          </div>
                          I've never got around compiling it right, and
                          I'm out of a Windows since, wow, I don't even
                          know if grub would boot the partition
                          anymore... If you have it under hand (i hope),
                          I'd be happy to give it a shot, it would also
                          be the occasion for me next weekend to pull a
                          ribbon cable out once and for all (or find a
                          clever way to keep it hidden though
                          programmable)<br>
                          <br>
                        </div>
                        How does crash recording work ? over Diag ? (PS:
                        50M/month data if needed)<br>
                        <br>
                      </div>
                      Julien<br>
                      ./.<br>
                    </div>
                    <div class="gmail_extra"><br>
                      <div class="gmail_quote">On Mon, Nov 28, 2016 at
                        10:16 PM, Michael Balzer <span dir="ltr"><<a href="mailto:dexter@expeedo.de" target="_blank">dexter@expeedo.de</a>></span>
                        wrote:<br>
                        <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                          <div text="#000000" bgcolor="#FFFFFF"> There's
                            no memory management in the module, there's
                            too few RAM available to do dynamic
                            allocations.<br>
                            <br>
                            All buffers are static, the error can be
                            some pointer corruption or missing length
                            check / end of string check.<br>
                            <br>
                            A stack overflow could also be possible, the
                            software stack is 256 bytes (1 page). As
                            there's not enough ROM left to support the
                            crash debug framework anymore in the Twizy
                            build, I cannot tell if that happens again.<br>
                            <br>
                            Julien, you could try a reduced build with
                            crash debug recording enabled...<br>
                            <br>
                            Regards,<br>
                            Michael
                            <div>
                              <div class="m_-5713992841334186999m_-7214223298656431776h5"><br>
                                <br>
                                <br>
                                <div class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-cite-prefix">Am
                                  28.11.2016 um 22:00 schrieb Greg D.:<br>
                                </div>
                                <blockquote type="cite"> Ah!  That
                                  sounds much more likely.<br>
                                  <br>
                                  Is there an error (timeout) path that
                                  might be freeing the buffer when it's
                                  still in use?<br>
                                  <br>
                                  Greg.<br>
                                  <br>
                                  <br>
                                  <div class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-cite-prefix">Michael
                                    Balzer wrote:<br>
                                  </div>
                                  <blockquote type="cite"> Hi Greg,
                                    welcome :)<br>
                                    <br>
                                    Juliens parameter #0 has been
                                    overwritten with "Topping o", which
                                    is part of the charge status message
                                    ("Topping off").<br>
                                    <br>
                                    So it's much more likely we've got a
                                    buffer corruption bug that's
                                    triggered by (maybe) bad
                                    connectivity situations.<br>
                                    <br>
                                    Regards,<br>
                                    Michael<br>
                                    <br>
                                    <br>
                                    <div class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-cite-prefix">Am
                                      28.11.2016 um 06:38 schrieb Greg
                                      D.:<br>
                                    </div>
                                    <blockquote type="cite"> Hi all,<br>
                                      <br>
                                      New here, so this may be totally
                                      off the wall.<br>
                                      <br>
                                      I notice that the first example,
                                      with a "5" going to a "k", it may
                                      be more easily explained as a bit
                                      <b><i>shift</i></b> instead of a
                                      set of flips.  Is there anywhere
                                      in the system where the bytes are
                                      clocked serially between devices,
                                      where there may be some marginal
                                      timing?<br>
                                      <br>
                                      Just a thought.<br>
                                      <br>
                                      Greg.<br>
                                      <br>
                                      <br>
                                      <div class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-cite-prefix">Mark
                                        Webb-Johnson wrote:<br>
                                      </div>
                                      <blockquote type="cite">
                                        <pre>$ perl -e 'printf "%08b\n",ord("5");'
00110101
$ perl -e 'printf "%08b\n",ord("k");'
01101011
$ perl -e 'printf "%08b\n",ord("8");'
00111000
$ perl -e 'printf "%08b\n",ord("Z");'
01011010

Strange. Seems more than a simple bit flip. I haven’t seen this myself, or heard of it much from other users.

Is the Twizy code using EEPROM for anything other than settings storage? I know that there are a very limit number of write cycles into that EEPROM space.

I did some googling, but all I can find is corruption during writing (either power down, or interrupt during write). But, for our case of very very rare EEPROM writes, that shouldn’t be an issue.

Regards, Mark.

</pre>
                                        <blockquote type="cite">
                                          <pre>On 25 Nov 2016, at 5:47 AM, Michael Balzer <a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-rfc2396E" href="mailto:dexter@expeedo.de" target="_blank"><dexter@expeedo.de></a> wrote:

The EEPROM data loss is a known issue, and I'm afraid there's nothing we can do about that. It seems to be a bug in the PIC18.

Regards
Michael

Am 24.11.2016 um 22:10 schrieb Julien Banchet:
</pre>
                                          <blockquote type="cite">
                                            <pre>Hi all,

Thanks to Michaels last message to remind me to ask a question here:

Since a few months, maybe a bit more because it happens only once in a while, I have noticed I wasn't receiving the SMS alerts (feature on SMSIP mode) , and sending a "Stats?" only gave me the nasty access denied answer.

It turns out, using the application, I noticed that the phone number was getting corrupted (afaik, it's the only setting affected)

Normally: "<a href="tel:%2B33651886877" value="+33651886877" target="_blank">+33651886877</a>" it became "+336k1886877" or "+33651886Z77" etc... like a solar flare bit flip !_____________________________<wbr>__________________
</pre>
              </blockquote>
            </blockquote>
            <pre>______________________________<wbr>_________________
OvmsDev mailing list
<a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.teslaclub.hk" target="_blank">OvmsDev@lists.teslaclub.hk</a>
<a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-freetext" href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" target="_blank">http://lists.teslaclub.hk/mail<wbr>man/listinfo/ovmsdev</a>
</pre>
          </blockquote>
          

          

          <fieldset class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798mimeAttachmentHeader"></fieldset>
          

          <pre>______________________________<wbr>_________________
OvmsDev mailing list
<a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.teslaclub.hk" target="_blank">OvmsDev@lists.teslaclub.hk</a>
<a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-freetext" href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" target="_blank">http://lists.teslaclub.hk/mail<wbr>man/listinfo/ovmsdev</a>
</pre>
        </blockquote>
        

        <pre class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-signature" cols="72">-- 
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
</pre>
        

        <fieldset class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798mimeAttachmentHeader"></fieldset>
        

        <pre>______________________________<wbr>_________________
OvmsDev mailing list
<a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.teslaclub.hk" target="_blank">OvmsDev@lists.teslaclub.hk</a>
<a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-freetext" href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" target="_blank">http://lists.teslaclub.hk/mail<wbr>man/listinfo/ovmsdev</a>
</pre>
      </blockquote>
      

      

      <fieldset class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798mimeAttachmentHeader"></fieldset>
      

      <pre>______________________________<wbr>_________________
OvmsDev mailing list
<a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.teslaclub.hk" target="_blank">OvmsDev@lists.teslaclub.hk</a>
<a class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-txt-link-freetext" href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" target="_blank">http://lists.teslaclub.hk/mail<wbr>man/listinfo/ovmsdev</a>
</pre>
    </blockquote>
    

    <pre class="m_-5713992841334186999m_-7214223298656431776m_-2341388285749433798moz-signature" cols="72">-- 
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
</pre>
  </div></div></div>


______________________________<wbr>_________________

OvmsDev mailing list

<a href="mailto:OvmsDev@lists.teslaclub.hk" target="_blank">OvmsDev@lists.teslaclub.hk</a>

<a href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" rel="noreferrer" target="_blank">http://lists.teslaclub.hk/mail<wbr>man/listinfo/ovmsdev</a>


</blockquote></div>
</div>


<fieldset class="m_-5713992841334186999m_-7214223298656431776mimeAttachmentHeader"></fieldset>
<pre>______________________________<wbr>_________________
OvmsDev mailing list
<a class="m_-5713992841334186999m_-7214223298656431776moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.teslaclub.hk" target="_blank">OvmsDev@lists.teslaclub.hk</a>
<a class="m_-5713992841334186999m_-7214223298656431776moz-txt-link-freetext" href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" target="_blank">http://lists.teslaclub.hk/mail<wbr>man/listinfo/ovmsdev</a>
</pre>

</blockquote>
<pre class="m_-5713992841334186999m_-7214223298656431776moz-signature" cols="72">-- 
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
</pre></div></div></div>
______________________________<wbr>_________________

OvmsDev mailing list

<a href="mailto:OvmsDev@lists.teslaclub.hk" target="_blank">OvmsDev@lists.teslaclub.hk</a>

<a href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" rel="noreferrer" target="_blank">http://lists.teslaclub.hk/mail<wbr>man/listinfo/ovmsdev</a>


</blockquote></div>
</div>


<fieldset class="m_-5713992841334186999mimeAttachmentHeader"></fieldset>
<pre>______________________________<wbr>_________________
OvmsDev mailing list
<a class="m_-5713992841334186999moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.teslaclub.hk" target="_blank">OvmsDev@lists.teslaclub.hk</a>
<a class="m_-5713992841334186999moz-txt-link-freetext" href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" target="_blank">http://lists.teslaclub.hk/<wbr>mailman/listinfo/ovmsdev</a>
</pre>

</blockquote>
<pre class="m_-5713992841334186999moz-signature" cols="72">-- 
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
</pre>
</div></div></div><br>______________________________<wbr>_________________<br>
OvmsDev mailing list<br>
<a href="mailto:OvmsDev@lists.teslaclub.hk">OvmsDev@lists.teslaclub.hk</a><br>
<a href="http://lists.teslaclub.hk/mailman/listinfo/ovmsdev" rel="noreferrer" target="_blank">http://lists.teslaclub.hk/<wbr>mailman/listinfo/ovmsdev</a><br>
<br></blockquote></div><br></div>