1. 19 Jul, 2019 1 commit
  2. 14 Jul, 2019 1 commit
    • Johny Mattsson's avatar
      Fix net module data loss & RTOS task unsafety (#2829) · f44e6e96
      Johny Mattsson authored
      To avoid races between the lwIP callbacks (lwIP RTOS task) and the Lua
      handlers (LVM RTOS task), the data flow and ownership has been simplified
      and cleaned up.
      
      lwIP callbacks now have no visibility of the userdata struct. They are
      limited to creating small event objects and task_post()ing them over
      to the LVM "thread", passing ownership in doing so. The shared identifier
      then becomes the struct netconn*.
      
      On the LVM side, we keep a linked list of active userdata objects. This
      allows us to retrieve the correct userdata when we get an event with
      a netconn pointer. Because this list is only ever used within the LVM
      task, no locking is necessary.
      
      The old approach of stashing a userdata pointer into the 'socket' field
      on the netconn has been removed entirely, as this was both not
      thread/RTOS-task safe, and also interfered with the IDFs internal use
      of the socket field (even when using only the netconn layer). As an
      added benefit, this removed the need for all the SYS_ARCH_PROTECT()
      locking stuff.
      
      The need to track receive events before the corresponding userdata object
      has been established has been removed by virtue of not reordering the
      "accept" and the "recv" events any more (previously accepts were posted
      with medium priority, while the receives where high priority, leading
      to the observed reordering and associated headaches).
      
      The workaround for IDF issue 784 has been removed as it is now not needed
      and is in fact directly harmful as it results in a double-free. Yay for
      getting rid of old workarounds!
      
      DNS resolution code paths were merged for the two instances of "socket"
      initiated resolves (connect/dns functions).
      
      Also fixed an instance of using a stack variable for receiving the resolved
      IP address, with said variable going out of scope before the DNS resolution
      necessarily completed (hello, memory corruption!).
      
      Where possible, moved to use the Lua allocator rather than plain malloc.
      
      Finally, the NodeMCU task posting mechanism got a polish and an adjustment.
      Given all the Bad(tm) that tends to happen if something fails task posting,
      I went through a couple of iterations on how to avoid that. Alas, the
      preferred solution of blocking non-LVM RTOS tasks until a slot is free
      turned out to not be viable, as this easily resulted in deadlocks with the
      lwIP stack. After much deliberation I settled on increasing the number of
      available queue slots for the task_post() mechanism, but in the interest
      of user control also now made it user configurable via Kconfig.
      f44e6e96
  3. 20 Sep, 2016 1 commit
    • Johny Mattsson's avatar
      Successfully boot barebones NodeMCU on ESP32 (only). · 9bbf8f43
      Johny Mattsson authored
      RTOS driver evicted as it did not play nice with stdio etc.
      
      Implemented a minimal driver to fully support Lua console on UART0. Output
      on UART0 done via stdout (provided by the IDF). Input and setup handled
      via driver_console/console.c. In addition to the direct input function
      console_getc(), the driver also registers in the syscall tables to enable
      regular stdio input functions to work (yay!). The Lua VM is still using the
      direct interface since it's less overhead, but does also work when going
      through stdin/fd 0.
      
      Auto-bauding on the console is not yet functional; revisit when the UART docs
      are available.
      
      Module registration/linking/enabling moved over to be Kconfig based. See
      updates to base_nodemcu/include/module.h and base_nodemcu/Kconfig for
      details.
      
      The sdk-overrides directory/approach is no longer used. The IDF is simply
      too different to the old RTOS SDK - we need to adapt our code directly instead.
      
      Everything in app/ is now unused, and will need to be gradually migrated
      into components/ though it is probably better to migrate straight from the
      latest dev branch.
      9bbf8f43
  4. 16 Sep, 2016 1 commit
  5. 09 Jun, 2016 1 commit
    • Johny Mattsson's avatar
      Make NodeMCU compile and link for ESP32. · 224788b6
      Johny Mattsson authored
      A fair bit of reshuffling with include paths and overrides was necessary, as
      the two RTOS SDKs (ESP8266 and ESP32) don't have the same header structure
      (or even libraries for that matter). Uses the xtensa-esp108-elf toolchain
      to build.
      
      Completely untested beyond linking, as I still can't flash the ESP32 module
      I have :(  I'd be most surprised if it does anything useful at this point
      considering I've spent almost no time on the linker script or UART setup.
      
      Anything using espconn has been ifdef'd out since espconn is not (and
      probably will not be) available. Notably this includes the entire net module
      as well as coap, mqtt and enduser_setup.
      
      Many (most?) hardware bus drivers and related modules are also ifdef'd
      out for now due to hardware differences. Functions surrounding sleep,
      rtc and RF modes have also been hit by the ifdef hammer. Grep'ing for
      __ESP8266__ and/or FIXME is a quick way of finding these places. With
      time I hope all of these will be reinstated.
      224788b6
  6. 31 May, 2016 1 commit
  7. 30 May, 2016 1 commit
    • Johny Mattsson's avatar
      Boot to (nonresponsive) Lua prompt on RTOS! · 56789592
      Johny Mattsson authored
      Uart driver currently disabled as it's not (yet) compatible with RTOS.
      
      Running Lua task with excessive stack to avoid smashing it; need to work out
      what's using so much stack space.
      
      Changed some flash reading functions to not attempt to drop an entire 4k
      flash page onto the stack.
      
      Ensure the task pump doesn't attempt to retrieve from uninitialised queues.
      56789592
  8. 24 May, 2016 2 commits
    • Johny Mattsson's avatar
      Reimplemented the NodeMCU task interface on top of RTOS. · 5c39a1f6
      Johny Mattsson authored
      Not yet hooked up to an RTOS task, however.
      5c39a1f6
    • Johny Mattsson's avatar
      Initial pass at switching to RTOS SDK. · f3e2a3af
      Johny Mattsson authored
      This compiles, links, and starts the RTOS without crashing and burning.
      
      Lua environment does not yet start due to the different task architecture.
      
      Known pain points:
      
        - task implementation needs to be rewritten for RTOS (next up on my TODO)
      
        - secure espconn does not exist, all secure espconn stuff has been #if 0'd
      
        - lwip now built from within the RTOS SDK, but does not appear to include
          MDNS support. Investigation needed.
      
        - there is no access to FRC1 NMI, not sure if we ever actually used that
          however. Also #if 0'd out for now.
      
        - new timing constraints introduced by the RTOS, all use of ets_delay_us()
          and os_delay_us() needs to be reviewed (the tsl2561 driver in particular).
      
        - even more confusion with ets_ vs os_ vs c_ vs non-prefixed versions.
          In the long run everything should be switched to non-prefixed versions.
      
        - system_set_os_print() not available, needs to be reimplemented
      
        - all the RTOS rodata is loaded into RAM, as it apparently uses some
          constants while the flash isn't mapped, so our exception handler can't
          work its magic. This should be narrowed down to the minimum possible
          at some point.
      
        - with each task having its own stack in RTOS, we probably need change
          flash-page buffers from the stack to the heap in a bunch of places.
          A single, shared, page buffer *might* be possible if we limit ourselves
          to running NodeMCU in a single task.
      
        - there's a ton of junk in the sdk-overrides now; over time the core code
          should be updated to not need those shims
      f3e2a3af
  9. 17 Feb, 2016 1 commit
    • TerryE's avatar
      Add New Tasking I/F and rework GPIO, UART, etc to support it · 49733f6f
      TerryE authored
      As with the last commit this rolls up the follwowing, but include the various
      review comments on the PR.
      
      -   **Documentation changes**. I've added the taks FAQ as a stub new Extension
      developer FAQ, and split the old FAQ into a Lua Developer FAQ and a Hardware
      FAQ.
      
      -   **Tasking I/F**.  New `app/task/Makefile`, `app/task/task.c`,
      `app/include/task/task.h` and `app/Makefile` as per previous commit.  Cascade
      changes to `app/driver/uart.c`, `app/include/driver/uart.h`,
      `app/user/user_main.c` and `app/modules/node.c`
      
      -   **GPIO Rework** to `app/modules/gpio.c` and `pin_map.[hc]`, `platform.[hc]`
      in `app/platform`
      
      -   **Other Optimisations** Move the `platform_*_exists()` from
      `app/platform/common.c` to static inline declarations in `platform.h` as
      this generates faster, smaller code. Move lgc.a routines out of iram0.
      49733f6f