<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">Oh my…<br>
      <br>
      Can it be the first 2 MB are taboo for core 1 and the second 2 MB
      for core 0?<br>
      <br>
      Anyway, we won't need those 2 MB for now, and hopefully this will
      be fixed by Espressif soon.<br>
      <br>
      So I'd say yes, commit.<br>
      <br>
      Regards,<br>
      Michael<br>
      <br>
      <br>
      Am 30.01.19 um 12:26 schrieb Mark Webb-Johnson:<br>
    </div>
    <blockquote type="cite"
      cite="mid:2F06DE6F-A324-4137-90C7-B7450C61A0BA@webb-johnson.net">
      <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
      I tried changing CONFIG_SPIRAM_SIZE in sdkconfig directly, but it
      just reset back to the default 4MB.
      <div class=""><br class="">
      </div>
      <div class="">So changed esp-idf itself, components/esp32/Kconfig,
        and added a prompt for that config option. I could then change
        it in menuconfig and the change persisted.</div>
      <div class=""><br class="">
      </div>
      <div class="">I then tested:</div>
      <div class=""><br class="">
      </div>
      <blockquote style="margin: 0 0 0 40px; border: none; padding:
        0px;" class="">
        <div class="">
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (0) cpu_start: App
                cpu up.</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (715) spiram: SPI
                SRAM memory test OK</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (716) heap_init:
                Initializing. RAM available for dynamic allocation:</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (716) heap_init: At
                3FFAE6E0 len 00001920 (6 KiB): DRAM</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (723) heap_init: At
                3FFBA920 len 000256E0 (149 KiB): DRAM</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (728) heap_init: At
                3FFE0440 len 00003BC0 (14 KiB): D/IRAM</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (735) heap_init: At
                3FFE4350 len 0001BCB0 (111 KiB): D/IRAM</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (742) heap_init: At
                400953CC len 0000AC34 (43 KiB): IRAM</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (747) cpu_start: Pro
                cpu start user code</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (752) spiram: Adding
                pool of 2048K of external SPI memory to heap allocator</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (158) cpu_start:
                Starting scheduler on PRO CPU.</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (0) cpu_start:
                Starting scheduler on APP CPU.</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">This is ESP32 chip
                with 2 CPU cores, WiFi/BT/BLE, silicon revision 1, 16MB
                external flash</span></font></div>
        </div>
        <div class=""><font class="" face="Andale Mono"><span
              style="font-size: 14px;" class="">…</span></font></div>
        <div class="">
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">I (9566) test_task:
                testing mode 1 loopcnt 1000</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">E (9576) test_task:
                loop #1: corruption detected at pos=497 size=498</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">4348864OVMS WIFI BLE
                BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class=""><br class="">
              </span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">#1: size=498, data
                @0x3f901ca0</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">4348864OVMS WIFI BLE
                BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class=""><br class="">
              </span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">E (9806) test_task:
                loop #5: corruption detected at pos=497 size=498</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">4348864OVMS WIFI BLE
                BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class=""><br class="">
              </span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">E (10036) test_task:
                loop #9: corruption detected at pos=829 size=830</span></font></div>
          <div class=""><font class="" face="Andale Mono"><span
                style="font-size: 14px;" class="">4348864OVMS WIFI BLE
                BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23)4348864OVMS WIFI BLE BT cores=2
                rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                15:08:23</span></font></div>
        </div>
      </blockquote>
      <div class="">
        <div><br class="">
        </div>
        <div>which looks like a FAIL to me?</div>
        <div><br class="">
        </div>
        <div>So, then I tried something a little funky:</div>
        <div><br class="">
        </div>
        <div>
          <ol class="MailOutline">
            <li class="">Go back to normal CONFIG_SPIRAM_SIZE 4MB<br
                class="">
              <br class="">
            </li>
            <li class="">Add this to main/test_memory_corruption.cpp
              app_main():<br class="">
              <br class="">
              <font class="" face="Andale Mono"><span style="font-size:
                  14px;" class="">void* big2mb =
                  heap_caps_malloc(1024*1024*2, MALLOC_CAP_SPIRAM);<br
                    class="">
                  printf("Big 2MB spi memory allocation is at
                  %p\n",big2mb);</span></font><br class="">
              <br class="">
            </li>
            <li class="">Rebuild and run...</li>
          </ol>
        </div>
      </div>
      <div><br class="">
      </div>
      <div>Here is what I get:</div>
      <div><br class="">
      </div>
      <blockquote style="margin: 0 0 0 40px; border: none; padding:
        0px;" class="">
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (0) cpu_start: App
                  cpu up.</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1151) spiram: SPI
                  SRAM memory test OK</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1152) heap_init:
                  Initializing. RAM available for dynamic allocation:</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1152) heap_init:
                  At 3FFAE6E0 len 00001920 (6 KiB): DRAM</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1159) heap_init:
                  At 3FFBA920 len 000256E0 (149 KiB): DRAM</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1164) heap_init:
                  At 3FFE0440 len 00003BC0 (14 KiB): D/IRAM</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1171) heap_init:
                  At 3FFE4350 len 0001BCB0 (111 KiB): D/IRAM</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1179) heap_init:
                  At 400953CC len 0000AC34 (43 KiB): IRAM</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1184) cpu_start:
                  Pro cpu start user code</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (1188) spiram:
                  Adding pool of 4096K of external SPI memory to heap
                  allocator</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (206) cpu_start:
                  Starting scheduler on PRO CPU.</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (0) cpu_start:
                  Starting scheduler on APP CPU.</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">...</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">This is ESP32 chip
                  with 2 CPU cores, WiFi/BT/BLE, silicon revision 1,
                  16MB external flash</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">Big 2MB spi memory
                  allocation is at 0x3f800028</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">...</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (291) wifi
                  station: ESP_WIFI_MODE_APSTA</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (291) wifi: wifi
                  driver task: 3ffc25c4, prio:23, stack:3584, core=0</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">...</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (3281) wifi
                  station: retry to connect to the AP</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (3281) wifi
                  station: connect to the AP fail</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class=""><br class="">
                </span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (6111) wifi
                  station: retry to connect to the AP</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (6111) wifi
                  station: connect to the AP fail</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class=""><br class="">
                </span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (8461)
                  mem_stresser: started on core 0 mem @0x3fa01b98</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (8461)
                  mem_stresser: started on core 1 mem @0x3fa81bac</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (8951) wifi
                  station: retry to connect to the AP</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (8951) wifi
                  station: connect to the AP fail</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class=""><br class="">
                </span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (9461) test_task:
                  running on core 1</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class=""><br class="">
                </span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (9561) test_task:
                  testing mode 1 loopcnt 1000</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">#1: size=1039, data
                  @0x3fb01ca4</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">4348864OVMS WIFI BLE
                  BT cores=2
                  rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                  (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                  15:08:23)4348864OVMS WIFI BLE BT cores=2
                  rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                  (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                  15:08:23)4348864OVMS WIFI BLE BT cores=2
                  rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                  (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                  15:08:23)4348864OVMS WIFI BLE BT cores=2
                  rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                  (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                  15:08:23)4348864OVMS WIFI BLE BT cores=2
                  rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                  (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                  15:08:23)4348864OVMS WIFI BLE BT cores=2
                  rev=ESP32/1WLAN-214677-52wifi103.2.001-22-g14d16854-dirty/factory/edge
                  (build idf v3.2-dev-1419-g5afc1b0cb-dirty Jan 27 2019
                  15:08:23)4348864OVMS WIFI BLE BT cores=2 rev=ESP32/1</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class=""><br class="">
                </span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (11781) wifi
                  station: retry to connect to the AP</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (11781) wifi
                  station: connect to the AP fail</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class=""><br class="">
                </span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (14621) wifi
                  station: retry to connect to the AP</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (14621) wifi
                  station: connect to the AP fail</span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class=""><br class="">
                </span></font></div>
          </div>
        </div>
        <div class="">
          <div>
            <div><font class="" face="Andale Mono"><span
                  style="font-size: 14px;" class="">I (17451) wifi
                  station: connect to the AP fail</span></font></div>
          </div>
        </div>
      </blockquote>
      <div class="">
        <div>
          <div><br class="">
          </div>
          <div>I left it running for a few minutes, but no faults shown.
            Modes 1 through 4 all clean.</div>
          <div><br class="">
          </div>
          <div>So, to take the funkiness to the next level, I added the
            same to ovms_main, right at the start (actually in the log
            level setter, as that seemed to be the first thing that
            initialises). Somebody else gets in for 28bytes of SPIRAM
            (or maybe that is a structure header), but anyway...</div>
          <div><br class="">
          </div>
        </div>
      </div>
      <blockquote style="margin: 0 0 0 40px; border: none; padding:
        0px;" class="">
        <div class="">
          <div>
            <div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1742) spiram:
                    SPI SRAM memory test OK</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1743)
                    heap_init: Initializing. RAM available for dynamic
                    allocation:</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1743)
                    heap_init: At 3FFAFF10 len 000000F0 (0 KiB): DRAM</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1750)
                    heap_init: At 3FFCEDE8 len 00011218 (68 KiB): DRAM</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1755)
                    heap_init: At 3FFE0440 len 00003BC0 (14 KiB): D/IRAM</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1762)
                    heap_init: At 3FFE4350 len 0001BCB0 (111 KiB):
                    D/IRAM</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1768)
                    heap_init: At 4009F530 len 00000AD0 (2 KiB): IRAM</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1775)
                    cpu_start: Pro cpu start user code</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (1779) spiram:
                    Adding pool of 4096K of external SPI memory to heap
                    allocator</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (125) ovms_main:
                    Set default logging level for * to INFO</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">I (126) ovms_main:
                    Pre-allocated 2MB of SPIRAM at 0x3f800028
                    (workaround ESP32 bug)</span></font></div>
              <div><font class="" face="Andale Mono"><span
                    style="font-size: 14px;" class="">...</span></font></div>
              <div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class="">I (17783)
                      webserver: Launching Web Server</span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class="">I (17823) ssh:
                      Launching SSH Server</span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class="">OVMS# test
                      string 500 2</span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class="">#1: stdlen =
                      1027</span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class="">1985504OVMS WIFI
                      BLE BT cores=2
                      rev=ESP32/168HIGHWAYS-29wifi163.2.001-22-g14d16854-dirty/factory/edge
                      (build idf v3.1-dev-2835-g151269458-dirty Jan 30
                      2019 19:17:20)012.61985504OVMS WIFI BLE BT cores=2
rev=ESP32/168HIGHWAYS-29wifi163.2.001-22-g14d16854-dirty/factory/edge
                      (build idf v3.1-dev-2835-g151269458-dirty Jan 30
                      2019 19:17:20)012.61985504OVMS WIFI BLE BT cores=2
rev=ESP32/168HIGHWAYS-29wifi163.2.001-22-g14d16854-dirty/factory/edge
                      (build idf v3.1-dev-2835-g151269458-dirty Jan 30
                      2019 19:17:20)012.61985504OVMS WIFI BLE BT cores=2
rev=ESP32/168HIGHWAYS-29wifi163.2.001-22-g14d16854-dirty/factory/edge
                      (build idf v3.1-dev-2835-g151269458-dirty Jan 30
                      2019 19:17:20)012.61985504OVMS WIFI BLE BT cores=2
rev=ESP32/168HIGHWAYS-29wifi163.2.001-22-g14d16854-dirty/factory/edge
                      (build idf v3.1-dev-2835-g151269458-dirty Jan 30
                      2019 19:17:20)012.61985504OVMS WIFI BLE BT cores=2
rev=ESP32/168HIGHWAYS-29wifi163.2.001-22-g14d16854-dirty/factory/edge
                      (build idf v3.1-dev-2835-g151269458-dirty Jan 30
                      2019 19:17:20)012.61985504</span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class=""><br class="">
                    </span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class="">finished</span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class="">OVMS#</span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class=""><br class="">
                    </span></font></div>
                <div><font class="" face="Andale Mono"><span
                      style="font-size: 14px;" class="">OVMS# module
                      memory<br class="">
                      Free 8-bit 76160/230212, 32-bit 812/2720, SPIRAM
                      1886108/4194252<br class="">
                      --Task--     Total DRAM D/IRAM   IRAM SPIRAM   +/-
                      DRAM D/IRAM   IRAM SPIRAM<br class="">
                      main*              8368      0      0   8456    
                       +8368     +0     +0  +8456<br class="">
                      no task*           5376      0      0      0    
                       +5376     +0     +0     +0<br class="">
                      esp_timer         34808      0    6442123100    
                      +34808     +0   +644+2123100<br class="">
                      OVMS Events       36352  41040      0  35120    
                      +36352 +41040     +0 +35120<br class="">
                      OVMS DukTape        188      0      0  34668      
                      +188     +0     +0 +34668<br class="">
                      OVMS CanRx         3124      0      0  26168    
                       +3124     +0     +0 +26168<br class="">
                      ipc0               7776      0      0      0    
                       +7776     +0     +0     +0<br class="">
                      ipc1                 12      0      0      0      
                       +12     +0     +0     +0<br class="">
                      Tmr Svc               0     88      0      0      
                        +0    +88     +0     +0<br class="">
                      tiT                 232    468      0    664      
                      +232   +468     +0   +664<br class="">
                      wifi                600   8896      0   3324      
                      +600  +8896     +0  +3324<br class="">
                      OVMS Console          0    200   1180  26332      
                        +0   +200  +1180 +26332<br class="">
                      mdns                  0    104      0      4      
                        +0   +104     +0     +4<br class="">
                      OVMS NetMan           0    216      0     84      
                        +0   +216     +0    +84</span></font></div>
              </div>
            </div>
          </div>
        </div>
      </blockquote>
      <div class="">
        <div><br class="">
        </div>
        <div>I think that wins as the most ugly kludge to date.</div>
        <div><br class="">
        </div>
        <div>Shall I commit?</div>
        <div><br class="">
        </div>
        <div>Regards, Mark.</div>
        <div><br class="">
          <blockquote type="cite" class="">
            <div class="">On 30 Jan 2019, at 5:00 PM, Mark Webb-Johnson
              <<a href="mailto:mark@webb-johnson.net" class=""
                moz-do-not-send="true">mark@webb-johnson.net</a>>
              wrote:</div>
            <br class="Apple-interchange-newline">
            <div class="">
              <meta http-equiv="Content-Type" content="text/html;
                charset=UTF-8" class="">
              <div style="word-wrap: break-word; -webkit-nbsp-mode:
                space; line-break: after-white-space;" class=""><br
                  class="">
                <div class=""><br class="">
                  <blockquote type="cite" class="">
                    <div class="">On 30 Jan 2019, at 4:41 PM, Michael
                      Balzer <<a href="mailto:dexter@expeedo.de"
                        class="" moz-do-not-send="true">dexter@expeedo.de</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="">
                        <div class="moz-cite-prefix">Mark,<br class="">
                          <br class="">
                          issue #2892 mentions using only the lower 2 MB
                          of SPI RAM as a workaround. Where are our
                          allocations placed by the allocator? Does it
                          fill from the middle or end? If not we don't
                          use the upper half yet.<br class="">
                        </div>
                      </div>
                    </div>
                  </blockquote>
                  <div class=""><br class="">
                  </div>
                  I saw that, and had a look. I don’t see anything in
                  makeconfig to limit to 2MB (not 4MB).</div>
                <div class=""><br class="">
                </div>
                <div class="">I do see a CONFIG_SPIRAM_SIZE=4194304 in
                  sdkconfig, but not sure how it gets there. Perhaps we
                  can just change there to 2MB and try?</div>
                <div class=""><br class="">
                </div>
                <div class="">I did see this in the documentation:</div>
                <div class=""><br class="">
                </div>
                <blockquote style="margin: 0 0 0 40px; border: none;
                  padding: 0px;" class="">
                  <div class=""><span style="caret-color: rgb(64, 64,
                      64); color: rgb(64, 64, 64); font-family: Lato,
                      proxima-nova, "Helvetica Neue", Arial,
                      sans-serif; font-size: 16px; background-color:
                      rgb(252, 252, 252);" class="">During ESP-IDF
                      startup, external RAM is mapped into the data
                      address space starting at at address 0x3F800000
                      (byte-accessible). The length of this region is
                      the same as the SPIRAM size (up to the limit of
                      4MiB).</span></div>
                </blockquote>
                <div class=""><br class="">
                </div>
                <div class="">So, maybe we can look at the address our
                  allocations come from to see their offset from
                  0x3F800000? If it is coming from the top 2MB, then
                  perhaps we start with a big 2MB allocation that we
                  never use?</div>
                <div class=""> <br class="">
                  <blockquote type="cite" class="">
                    <div class="">
                      <div text="#000000" bgcolor="#FFFFFF" class="">
                        <div class="moz-cite-prefix">A possible
                          workaround for the most apparent issues with
                          this (string assembly) could be to use char
                          buffers instead of std::string. I was thinking
                          about doing that at least for the websocket
                          stream.<br class="">
                        </div>
                      </div>
                    </div>
                  </blockquote>
                  <div class=""><br class="">
                  </div>
                  <div class="">It seems strange that we are only seeing
                    it with std::string. Maybe that just stresses the
                    system more, or does a lot of reallocations?</div>
                  <br class="">
                  <blockquote type="cite" class="">
                    <div class="">
                      <div text="#000000" bgcolor="#FFFFFF" class="">
                        <div class="moz-cite-prefix"> But if this really
                          is a hardware issue it can affect all objects
                          in SPI RAM, appending to std::string then only
                          triggers this more often.<br class="">
                          <br class="">
                          Another workaround could be to run everything
                          on core 0.<br class="">
                        </div>
                      </div>
                    </div>
                  </blockquote>
                  <div class=""><br class="">
                  </div>
                  Urgh.</div>
                <div class=""><br class="">
                  <blockquote type="cite" class="">
                    <div class="">
                      <div text="#000000" bgcolor="#FFFFFF" class="">
                        <div class="moz-cite-prefix">Regards,<br
                            class="">
                          Michael<br class="">
                          <br class="">
                          <br class="">
                          Am 30.01.19 um 03:33 schrieb Mark
                          Webb-Johnson:<br class="">
                        </div>
                        <blockquote type="cite"
                          cite="mid:4737F304-A344-4BD8-9B0C-95366DB7A40F@webb-johnson.net"
                          class="">
                          <meta http-equiv="Content-Type"
                            content="text/html; charset=UTF-8" class="">
                          Michael,
                          <div class=""><br class="">
                          </div>
                          <div class="">Espressif’s response (and
                            linking to that other issue) sounds like
                            others are seeing this.</div>
                          <div class=""><br class="">
                          </div>
                          <div class="">Lousy timing (with Chinese New
                            Year next week), so don’t expect anything
                            quick from Espressif. I guess we’ll just
                            have to live with it until they can find a
                            workaround. It sounds like the issue is at
                            the hardware level and a compiler patch will
                            be needed.</div>
                          <div class=""><br class="">
                          </div>
                          <div class="">Regards, Mark.<br class="">
                            <div class=""><br class="">
                              <blockquote type="cite" class="">
                                <div class="">On 30 Jan 2019, at 3:07
                                  AM, Michael Balzer <<a
                                    href="mailto:dexter@expeedo.de"
                                    class="" moz-do-not-send="true">dexter@expeedo.de</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=""> → <a
                                      class="moz-txt-link-freetext"
                                      href="https://github.com/espressif/esp-idf/issues/3006"
                                      moz-do-not-send="true">https://github.com/espressif/esp-idf/issues/3006</a><br
                                      class="">
                                    <br class="">
                                    Regards,<br class="">
                                    Michael<br class="">
                                    <br class="">
                                    <br class="">
                                    <div class="moz-cite-prefix">Am
                                      28.01.19 um 20:46 schrieb Michael
                                      Balzer:<br class="">
                                    </div>
                                    <blockquote type="cite"
                                      cite="mid:b193b172-9a83-6df5-7b61-9e0cbfa46281@expeedo.de"
                                      class="">
                                      <meta http-equiv="Content-Type"
                                        content="text/html;
                                        charset=UTF-8" class="">
                                      To clarify: the bug is most likely
                                      not restricted to the case of
                                      building a message in a buffer. It
                                      can possibly cause corruptions in
                                      any RAM section, so can well be
                                      responsible for many/most of the
                                      unidentified crashes and
                                      stack/heap corruptions we're
                                      experiencing.<br class="">
                                      <br class="">
                                      Regards,<br class="">
                                      Michael<br class="">
                                      <br class="">
                                      <br class="">
                                      <div class="moz-cite-prefix">Am
                                        28.01.19 um 20:43 schrieb
                                        Michael Balzer:<br class="">
                                      </div>
                                      <blockquote type="cite"
                                        cite="mid:06d0ea65-f4d5-e2c5-9ca1-b0ad63bfa3b1@expeedo.de"
                                        class="">
                                        <meta http-equiv="Content-Type"
                                          content="text/html;
                                          charset=UTF-8" class="">
                                        For those not following the
                                        github discussion: I'm pretty
                                        sure I've nailed the bug down.<br
                                          class="">
                                        <br class="">
                                        I have reproduced the bug in a
                                        simple test project and intend
                                        to raise an issue with Espressif
                                        on this.<br class="">
                                        <br class="">
                                        <a class="moz-txt-link-freetext"
href="https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/issues/189#issuecomment-457965435"
                                          moz-do-not-send="true">https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/issues/189#issuecomment-457965435</a><br
                                          class="">
                                        <br class="">
                                        If you'd like to test this on
                                        your module, configure the
                                        project to your wifi
                                        credentials, then use "make
                                        flash". As the test project is
                                        small, this normally will not
                                        erase your OVMS config
                                        partition, but a backup is
                                        always recommended.<br class="">
                                        <br class="">
                                        Regards,<br class="">
                                        Michael<br class="">
                                        <br class="">
                                        <br class="">
                                        <div class="moz-cite-prefix">Am
                                          24.01.19 um 21:19 schrieb
                                          Michael Balzer:<br class="">
                                        </div>
                                        <blockquote type="cite"
                                          cite="mid:e4159a1c-b283-25b0-f1b5-ce570b529a1c@expeedo.de"
                                          class="">
                                          <meta
                                            http-equiv="content-type"
                                            content="text/html;
                                            charset=UTF-8" class="">
                                          Everyone please have a look
                                          at…<br class="">
                                          <br class="">
                                          <a
                                            class="moz-txt-link-freetext"
href="https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/issues/189#issuecomment-457334248"
                                            moz-do-not-send="true">https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/issues/189#issuecomment-457334248</a><br
                                            class="">
                                          <br class="">
                                          Please try to reproduce the
                                          bug on your modules.<br
                                            class="">
                                          <br class="">
                                          I'm open for explanations.<br
                                            class="">
                                          <br class="">
                                          I thought this might be some
                                          copy-on-write bug with
                                          std::string, but the gcc 5.x
                                          libstdc++ does no longer use
                                          that implementation (wouldn't
                                          be C++11 compliant as well). I
                                          also tried moving all strings
                                          to temporary buffers, but
                                          modes 5 & 6 eliminated
                                          this explanation as well.<br
                                            class="">
                                          <br class="">
                                          My remaining theories:<br
                                            class="">
                                          <ul class="">
                                            <li class="">A task writing
                                              out of bounds (but only
                                              0-bytes?)</li>
                                            <li class="">A hardware
                                              issue only affecting some
                                              modules<br class="">
                                            </li>
                                          </ul>
                                          A hardware issue only
                                          affecting some percentage of
                                          ESP32 could explain this as
                                          well as the strange heap
                                          corruptions that seem to
                                          affect some modules especially
                                          often.<br class="">
                                          <br class="">
                                          Regards,<br class="">
                                          Michael<br class="">
                                          <br>
                                        </blockquote>
                                      </blockquote>
                                    </blockquote>
                                  </div>
                                </div>
                              </blockquote>
                            </div>
                          </div>
                        </blockquote>
                      </div>
                    </div>
                  </blockquote>
                </div>
              </div>
            </div>
          </blockquote>
        </div>
      </div>
    </blockquote>
    <br>
    <pre class="moz-signature" cols="144">-- 
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
</pre>
  </body>
</html>