<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;" class="">Greg, I merged your request. But it conflicted with the change I had just made (and pushed) a few hours earlier - to fix the same issue.<div class=""><br class=""></div><div class="">Anyway, I cleaned things up a bit, so the init ordering looks better now. Please try.</div><div class=""><br class=""></div><div class="">Regards, Mark.<br class=""><div><br class=""><blockquote type="cite" class=""><div class="">On 24 May 2018, at 4:52 AM, Greg D. <<a href="mailto:gregd2350@gmail.com" class="">gregd2350@gmail.com</a>> wrote:</div><br class="Apple-interchange-newline"><div class="">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" class="">
<div text="#000000" bgcolor="#FFFFFF" class="">
Ah, of course. Depending on timing, I was calling xQueueReceive for
a queue I hadn't created yet, regardless of the contents of said
queue. Listening just directs stuff into the queue.<br class="">
<br class="">
Anyway, pull request created. Let me know if this isn't what you
envisioned for a fix.<br class="">
<br class="">
Greg<br class="">
<br class="">
<br class="">
<div class="moz-cite-prefix">Michael Balzer wrote:<br class="">
</div>
<blockquote type="cite" cite="mid:f46d1554-730e-77ab-243b-a2930ed75532@expeedo.de" class="">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" class="">
Greg,<br class="">
<br class="">
the task is messaged to, not called by the CAN framework, it runs
as soon as it is created.<br class="">
<br class="">
Regards,<br class="">
Michael<br class="">
<br class="">
<br class="">
<div class="moz-cite-prefix">Am 23.05.2018 um 22:16 schrieb Greg
D.:<br class="">
</div>
<blockquote type="cite" cite="mid:85b245c9-3457-327d-667e-c17db79a6d67@gmail.com" class="">
<meta http-equiv="Content-Type" content="text/html;
charset=UTF-8" class="">
Ha, that's interesting. I've pushed a fix to swap the ordering,
but am not quite convinced that this is the issue. The order of
creation shouldn't matter because the OBD2ECU task doesn't
register itself as a listener until after both are done. So why
would it even be called? <br class="">
<br class="">
Greg<br class="">
<br class="">
<br class="">
<div class="moz-cite-prefix">Michael Balzer wrote:<br class="">
</div>
<blockquote type="cite" cite="mid:3805f0c8-f5fd-0cb9-9708-5de311a6b856@expeedo.de" class="">
<meta http-equiv="Content-Type" content="text/html;
charset=UTF-8" class="">
Greg,<br class="">
<br class="">
I just had a look. That crash is simply due to an
initialization order mistake in your constructor, has nothing
to do with the CAN bus: you create the m_rxqueue _after_
starting the OBD2ECU_task, which uses the m_rxqueue:<br class="">
<br class="">
<tt class="">static void OBD2ECU_task(void *pvParameters)</tt><tt class=""><br class="">
</tt><tt class=""> {</tt><tt class=""><br class="">
</tt><tt class=""> obd2ecu *me = (obd2ecu*)pvParameters;</tt><tt class=""><br class="">
</tt><tt class=""><br class="">
</tt><tt class=""> CAN_frame_t frame;</tt><tt class=""><br class="">
</tt><tt class=""> while(1)</tt><tt class=""><br class="">
</tt><tt class=""> {</tt><tt class=""><br class="">
</tt><tt class=""> if (xQueueReceive(me-><b class="">m_rxqueue</b>,
&frame, (portTickType)portMAX_DELAY)==pdTRUE)</tt><tt class=""><br class="">
</tt><tt class=""> {</tt><tt class=""><br class="">
</tt><tt class=""> // Only handle incoming frames on our CAN bus</tt><tt class=""><br class="">
</tt><tt class=""> if (frame.origin == me->m_can)
me->IncomingFrame(&frame);</tt><tt class=""><br class="">
</tt><tt class=""> }</tt><tt class=""><br class="">
</tt><tt class=""> }</tt><tt class=""><br class="">
</tt><tt class=""> }</tt><tt class=""><br class="">
</tt><tt class=""><br class="">
</tt><tt class="">obd2ecu::obd2ecu(const char* name, canbus* can)</tt><tt class=""><br class="">
</tt><tt class=""> : pcp(name)</tt><tt class=""><br class="">
</tt><tt class=""> {</tt><tt class=""><br class="">
</tt><tt class=""> m_can = can;</tt><tt class=""><br class="">
</tt><tt class=""> xTaskCreatePinnedToCore(OBD2ECU_task, "OVMS OBDII
ECU", 6144, (void*)this, 5, &m_task, 1);</tt><tt class=""><br class="">
</tt><tt class=""><br class="">
</tt><b class=""><tt class=""> m_rxqueue =
xQueueCreate(20,sizeof(CAN_frame_t));</tt></b><tt class=""><br class="">
</tt><tt class=""><br class="">
</tt><tt class="">…</tt><tt class=""><br class="">
</tt><br class="">
<br class="">
So the task can reach the xQueueReceive() call before the
queue is actually usable.<br class="">
<br class="">
Regards,<br class="">
Michael<br class="">
<br class="">
<br class="">
<div class="moz-cite-prefix">Am 23.05.2018 um 03:29 schrieb
Greg D.:<br class="">
</div>
<blockquote type="cite" cite="mid:33670a15-78c3-a1bc-95a4-07034c7193fb@gmail.com" class="">
<meta http-equiv="Content-Type" content="text/html;
charset=UTF-8" class="">
Hi Mark,<br class="">
<br class="">
Yeah, big shrug here too. It occurs infrequently, during
what was a reboot anyway, and the system recovered itself,
so this isn't a critical issue unless it keeps happening
(i.e. a crash/reboot loop). I suppose it could also occur
with a frame coming in from a HUD, though that CAN bus is on
a different interface chip so if the cause is hardware
dependent, it might not occur.<br class="">
<br class="">
Perhaps we need to not launch the higher level applications
(e.g. OBD2ECU) until maybe the 5 second mark, when we know
the system is initialized and otherwise pretty stable, but
within that "Early" timeframe? Just a thought,<br class="">
<br class="">
Greg<br class="">
<br class="">
<br class="">
<div class="moz-cite-prefix">Mark Webb-Johnson wrote:<br class="">
</div>
<blockquote type="cite" cite="mid:B296C17D-71F0-4486-B59A-065AA23690FB@webb-johnson.net" class="">
<meta http-equiv="Content-Type" content="text/html;
charset=UTF-8" class="">
Greg,
<div class=""><br class="">
</div>
<div class="">Yes, the MyCan.RegisterListener(m_rxqueue)
will register for all messages on all CAN buses.</div>
<div class=""><br class="">
</div>
<div class="">I am trying to work out a neat way of adding
filters to this (both hardware and software). Some easy
standard way of filtering traffic in a standard way
(like the BFP filters in Unix variants do for ethernet
traffic). That will give us a performance boost for
things like odb2ecu, retools, etc.</div>
<div class=""><br class="">
</div>
<div class="">But you already have a "if (frame.origin ==
me->m_can)”, so not sure how this could be affecting
you?</div>
<div class=""><br class="">
</div>
<div class="">
<blockquote type="cite" class="">What precisely is
"early" in the context of a boot crash? That would
put a bound on the size and location of the timing
window.</blockquote>
</div>
<div class=""><br class="">
</div>
<div class="">10 seconds, I believe. You can see it with
the message "I (10576) housekeeping: System considered
stable (RAM: 8b=99060-99176 32b=20)” about 10 seconds
into the boot.</div>
<div class=""><br class="">
</div>
<div class="">Regards, Mark.<br class="">
<div class=""><br class="">
<blockquote type="cite" class="">
<div class="">On 23 May 2018, at 7:55 AM, Greg D.
<<a href="mailto:gregd2350@gmail.com" class="" moz-do-not-send="true">gregd2350@gmail.com</a>>
wrote:</div>
<br class="Apple-interchange-newline">
<div class="">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" class="">
<div text="#000000" bgcolor="#FFFFFF" class=""> Hi
Michael,<br class="">
<br class="">
Correct, not a config issue. I had assumed it
was a race condition (first time I've seen it),
but didn't realize it involved my code. Yikes!<br class="">
<br class="">
This is the call-back into the OBD2ECU task when
a CAN frame is received in the system.
Apparently everyone looking for a CAN frame gets
called when a frame is received from any bus,
and each process needs to filter out the ones
they own. So, somehow a CAN frame was seen to
have been received before the system was fully
initialized. Perhaps something dropped into the
hardware buffer after the chip was turned on,
but not before the O/S had gotten everything
situated for handling it. I don't currently
have anything connected to the OBDII CAN bus
(CAN3) on my module, so this must have been a
frame from the car itself on CAN 1. CAN 2 is
not used in my car.<br class="">
<br class="">
What precisely is "early" in the context of a
boot crash? That would put a bound on the size
and location of the timing window.<br class="">
<br class="">
Greg<br class="">
<br class="">
<br class="">
<div class="moz-cite-prefix">Michael Balzer
wrote:<br class="">
</div>
<blockquote type="cite" cite="mid:6a8db28c-03ba-2952-d802-d720e6dd2deb@expeedo.de" class="">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" class="">
Greg,<br class="">
<br class="">
the backtrace says the crash happened here:<br class="">
<br class="">
<tt class="">0x4008f6de is in
vTaskEnterCritical
(/Users/mark/esp/esp-idf/components/freertos/include/freertos/portmacro.h:283).<br class="">
278 * Warning: From the ISA docs: in
some (unspecified) cases, the s32c1i
instruction may return the<br class="">
279 * *bitwise inverse* of the old mem
if the mem wasn't written. This doesn't seem
to happen on the<br class="">
280 * ESP32 (portMUX assertions would
fail).<br class="">
281 */<br class="">
282 static inline void
uxPortCompareSet(volatile uint32_t *addr,
uint32_t compare, uint32_t *set) {<br class="">
283 __asm__ __volatile__ (<br class="">
284 "WSR %2,SCOMPARE1 \n"<br class="">
285 "S32C1I %0, %1, 0 \n"<br class="">
286 :"=r"(*set)<br class="">
287 :"r"(addr), "r"(compare),
"0"(*set)<br class="">
0x4008eb10 is in xQueueGenericReceive
(/Users/mark/esp/esp-idf/components/freertos/./queue.c:1455).<br class="">
1450 statements within the function
itself. This is done in the interest<br class="">
1451 of execution time efficiency. */<br class="">
1452 <br class="">
1453 for( ;; )<br class="">
1454 {<br class="">
1455
taskENTER_CRITICAL(&pxQueue->mux);<br class="">
1456 {<br class="">
1457 /* Is there data in the
queue now? To be running the calling task<br class="">
1458 must be the highest
priority task wanting to access the queue.
*/<br class="">
1459 if(
pxQueue->uxMessagesWaiting > (
UBaseType_t ) 0 )<br class="">
<b class="">0x4011f3fd is in
OBD2ECU_task(void*)
(/Users/mark/Documents/ovms/Open-Vehicle-Monitoring-System-3/vehicle/OVMS.V3/components/obd2ecu/src/obd2ecu.cpp:155).</b></tt><tt class=""><br class="">
</tt><tt class="">150 obd2ecu *me =
(obd2ecu*)pvParameters;</tt><tt class=""><br class="">
</tt><tt class="">151 </tt><tt class=""><br class="">
</tt><tt class="">152 CAN_frame_t frame;</tt><tt class=""><br class="">
</tt><tt class="">153 while(1)</tt><tt class=""><br class="">
</tt><tt class="">154 {</tt><tt class=""><br class="">
</tt><b class=""><tt class="">155 if
(xQueueReceive(me->m_rxqueue,
&frame,
(portTickType)portMAX_DELAY)==pdTRUE)</tt></b><tt class=""><br class="">
</tt><tt class="">156 {</tt><tt class=""><br class="">
</tt><tt class="">157 // Only handle
incoming frames on our CAN bus</tt><tt class=""><br class="">
</tt><tt class="">158 if
(frame.origin == me->m_can)
me->IncomingFrame(&frame);</tt><tt class=""><br class="">
</tt><tt class="">159 }</tt><tt class=""><br class="">
</tt><br class="">
<br class="">
So I assume that's the same bug I already
reported in issue #103:<br class="">
<br class="">
<a class="moz-txt-link-freetext" href="https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/issues/103" moz-do-not-send="true">https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/issues/103</a><br class="">
<br class="">
…and as it now happened to you, I assume it's
not configuration related. Can you verify my
assumption?<br class="">
<br class="">
<blockquote type="cite" class="">Looks like an
init race condition, i.e. m_rxqueue used in
undefined state.</blockquote>
<br class="">
Regards,<br class="">
Michael<br class="">
<br class="">
<br class="">
<br class="">
<div class="moz-cite-prefix">Am 22.05.2018 um
19:24 schrieb Greg D.:<br class="">
</div>
<blockquote type="cite" cite="mid:45c73f38-dcea-9bac-7b5a-bedc8e08617d@gmail.com" class="">
<pre class="" wrap="">Just noticed this on my production 3.1 module. Was this an artifact of the
upgrade from .005 to .006?
Last boot was 140983 second(s) ago
Time at boot: 2018-05-20 19:05:59 PDT
This is reset #17 since last power cycle
Detected boot reason: EarlyCrash (12/12)
Crash counters: 1 total, 1 early
Last crash: StoreProhibited exception on core 1
Registers:
PC : 0x4008f6de PS : 0x00060033 A0 : 0x8008eb13 A1 : 0x3ffd6150
A2 : 0xfefeff4e A3 : 0x0000abab A4 : 0xb33fffff A5 : 0x00000001
A6 : 0x00060020 A7 : 0x0000cdcd A8 : 0x0000abab A9 : 0x3ffd6150
A10 : 0x00000003 A11 : 0x00060023 A12 : 0x00060020 A13 : 0x3ffb4188
A14 : 0x000001e8 A15 : 0x00001800 SAR : 0x00000000 EXCCAUSE: 0x0000001d
EXCVADDR: 0xfefeff4e LBEG : 0x00000000 LEND : 0x00000000 LCOUNT : 0x00000000
Backtrace:
0x4008f6de 0x4008eb10 0x4011f3fd 0x40174f95 0x400e73be 0x400e4699 0x40085149
0x40128bc2 0x40128df7 0x401287b4 0x400ea22d 0x400eac7a 0x400ead09 0x400ead19
0x402207d6 0x400e9ece 0x400e9f79 0x400e45f5 0x400e4604 0x400f5c09 0x400f035d
Version: 3.1.006/ota_0/main (build idf v3.1-dev-455-gf4905cdf May 20 2018 20:08:31)
--------------------------------------------------------------------------------
Running partition: ota_0
Boot partition: ota_0
Firmware: 3.1.006/ota_0/main (build idf v3.1-dev-455-gf4905cdf May 20 2018 20:08:31)
</pre>
<!--'"--><br class="">
<fieldset class="mimeAttachmentHeader"></fieldset>
<br class="">
<pre class="" wrap="">_______________________________________________
OvmsDev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.openvehicles.com" moz-do-not-send="true">OvmsDev@lists.openvehicles.com</a>
<a class="moz-txt-link-freetext" href="http://lists.openvehicles.com/mailman/listinfo/ovmsdev" moz-do-not-send="true">http://lists.openvehicles.com/mailman/listinfo/ovmsdev</a>
</pre>
</blockquote>
<br class="">
<pre class="moz-signature" cols="160">--
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
</pre>
<br class="">
<fieldset class="mimeAttachmentHeader"></fieldset>
<br class="">
<pre class="" wrap="">_______________________________________________
OvmsDev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.openvehicles.com" moz-do-not-send="true">OvmsDev@lists.openvehicles.com</a>
<a class="moz-txt-link-freetext" href="http://lists.openvehicles.com/mailman/listinfo/ovmsdev" moz-do-not-send="true">http://lists.openvehicles.com/mailman/listinfo/ovmsdev</a>
</pre>
</blockquote>
<br class="">
</div>
_______________________________________________<br class="">
OvmsDev mailing list<br class="">
<a href="mailto:OvmsDev@lists.openvehicles.com" class="" moz-do-not-send="true">OvmsDev@lists.openvehicles.com</a><br class="">
<a class="moz-txt-link-freetext" href="http://lists.openvehicles.com/mailman/listinfo/ovmsdev" moz-do-not-send="true">http://lists.openvehicles.com/mailman/listinfo/ovmsdev</a><br class="">
</div>
</blockquote>
</div>
<br class="">
</div>
<br class="">
<fieldset class="mimeAttachmentHeader"></fieldset>
<br class="">
<pre wrap="" class="">_______________________________________________
OvmsDev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.openvehicles.com" moz-do-not-send="true">OvmsDev@lists.openvehicles.com</a>
<a class="moz-txt-link-freetext" href="http://lists.openvehicles.com/mailman/listinfo/ovmsdev" moz-do-not-send="true">http://lists.openvehicles.com/mailman/listinfo/ovmsdev</a>
</pre>
</blockquote>
<br class="">
<!--'"--><br class="">
<fieldset class="mimeAttachmentHeader"></fieldset>
<br class="">
<pre wrap="" class="">_______________________________________________
OvmsDev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.openvehicles.com" moz-do-not-send="true">OvmsDev@lists.openvehicles.com</a>
<a class="moz-txt-link-freetext" href="http://lists.openvehicles.com/mailman/listinfo/ovmsdev" moz-do-not-send="true">http://lists.openvehicles.com/mailman/listinfo/ovmsdev</a>
</pre>
</blockquote>
<br class="">
<pre class="moz-signature" cols="160">--
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
</pre>
<br class="">
<fieldset class="mimeAttachmentHeader"></fieldset>
<br class="">
<pre wrap="" class="">_______________________________________________
OvmsDev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.openvehicles.com" moz-do-not-send="true">OvmsDev@lists.openvehicles.com</a>
<a class="moz-txt-link-freetext" href="http://lists.openvehicles.com/mailman/listinfo/ovmsdev" moz-do-not-send="true">http://lists.openvehicles.com/mailman/listinfo/ovmsdev</a>
</pre>
</blockquote>
<br class="">
<!--'"--><br class="">
<fieldset class="mimeAttachmentHeader"></fieldset>
<br class="">
<pre wrap="" class="">_______________________________________________
OvmsDev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.openvehicles.com" moz-do-not-send="true">OvmsDev@lists.openvehicles.com</a>
<a class="moz-txt-link-freetext" href="http://lists.openvehicles.com/mailman/listinfo/ovmsdev" moz-do-not-send="true">http://lists.openvehicles.com/mailman/listinfo/ovmsdev</a>
</pre>
</blockquote>
<br class="">
<pre class="moz-signature" cols="160">--
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
</pre>
<br class="">
<fieldset class="mimeAttachmentHeader"></fieldset>
<br class="">
<pre wrap="" class="">_______________________________________________
OvmsDev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OvmsDev@lists.openvehicles.com">OvmsDev@lists.openvehicles.com</a>
<a class="moz-txt-link-freetext" href="http://lists.openvehicles.com/mailman/listinfo/ovmsdev">http://lists.openvehicles.com/mailman/listinfo/ovmsdev</a>
</pre>
</blockquote>
<br class="">
</div>
_______________________________________________<br class="">OvmsDev mailing list<br class=""><a href="mailto:OvmsDev@lists.openvehicles.com" class="">OvmsDev@lists.openvehicles.com</a><br class="">http://lists.openvehicles.com/mailman/listinfo/ovmsdev<br class=""></div></blockquote></div><br class=""></div></body></html>