The irom0_flash.bin file gets written to offset 0x40000 in flash. Said file
has the following layout
| irom0 | text | data | rodata | chksum |
...so the previous approach of having a _flash_used_end symbol at the end of
the irom0 section no longer gives us an accurate view of how much of the flash
is used.
New driver is a three-way merger between Espressif's esp8266-rtos-sdk example
driver, Espressif's esp32-rtos-sdk not-example driver, and the previous
NodeMCU driver, plus some general clean-ups.
Basic interactivity is now available on the ESP32!
Running without interrupt vectors not considered all that useful...
Disabled flash-size-byte-setting and SPIFFS formatting on ESP32, for now
at least, as both of these appeared to corrupt the flash and prevent
subsequent boots.
Disabled UART init on ESP32 until the driver gets updated.
This now gets us to a banner on ESP32, but not yet a prompt.
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.
The os_timer callback is executed from task rtT, prio 14, so they preempt
the Lua environment whenever they fire. Ideally we should be using the
RTOS timers instead, which run at prio 2 and thus would be more suited
for our uses.
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.
Turns out ets_printf() lied to me. When handed an aligned string in flash it
did 32bit loads on it instead of the expected 8bit loads, so just silencing
the exception was enough to give the appearance of it working.
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
- Stop fighting against the SDK in terms of owning/writing the init_data block.
NodeMCU included a default init_data block because originally the SDK did
not, but by now it's not needed.
- Expose a way to reconfigure the ADC mode from Lua land. With most people
using the cloud builder and not able to change the #define for byte 107
this has been a pain point.
- Less confusion about which init_data has been used. Lua code can now simply
state what mode it wants the ADC to be in, and not worry about the rest of
the init_data complexities such as the init_data changing location due to
flashing with wrong flash_size setting, or doing/not doing a chip-erase
before loading new NodeMCU firmware.
* Change upper limit for timer in `tmr_register` and `tmr_interval` to reflect new limit in SDK 1.5.3.
* Change documentation for `tmr.alarm()`, `tmr.interval()` and `tmr.register()` to match source.
* Improve error reporting to be more descriptive.
* Send a multicast response if it is a QM query
* Use the NSEC record in the style which mDNS demands
* Shuffle the code a bit and allow sending of A response...
* Made the A record work
* Now gets TTLs right and (I think) handles multiple questions
* Fix bug on first write
Pin is 'HIGH' at reset, so we need to pull it down and generate a
reset.
* Move init code to flash section, not needed to be in iram.
* Remove pin choice in API
* Remove lock in ws2812_buffer_write
* Remove naked malloc
* Drop ws2812_writergb
* Drop support of ws2812.buffers to ws2812_writegrb should use ws2812.buffers:write
* Add support for <>3 colors per leds strips (RGBW)
* Remove ICACHE_FLASH_ATTR
* Add static const on _uartData to avoid initialization penalty
Johny Mattsson
Marcel Stör
Marcel Stör
Arnim Läuger
Johny Mattsson
dnc40085
Philip Gladstone
philip
Thomas Soëte
Thomas Mohaupt