The comment behind the ASM MultiU16X8toH16 was misleading.
It actually computes ((a<<8)*b)>>16, or (a*b)>>8.
Correct the comment and C reference implementation accordingly.
In PR #3093 the progmem array FW_VERSION_NR was introduced to store the
version components, however the code didn't read it properly using the
pgm_read_* functions, making version comparisons fail.
Fix the existing/unused is_provided_version_newer() and reuse it in
show_upgrade_dialog_if_version_newer().
Similarly also read/update correctly the version in the eeprom.
- Move D23 into it's own function inside Dcodes
- Correctly include a break in the switch statement
- Show the dumper status (enabled/disabled) after toggling
- Allow to generate an immediate dump via g-code using D23 E for
symmetry with D20 E
When building with GCC 4.9.2 (bundled with PF-build-env-1.0.6.*), -Os
and LTO enabled, PID_autotune gets automatically inlined into
process_commands().
Sadly, due to the massive size of process_commands(), it results in
codegen bug doing a partial stack overwrite in process_commands()
itself, manifesting as random behavior depending on the timing of
interrupts and the codepath taken inside the merged function.
Mark the function as noinline and add a note about the affected compiler
version in order to be checked again in the future.
Instead of using a mixture of int8_t, unsigned char and (incorrectly)
int, use uint8_t consistently for indexing the current block.
This improves the performance of the wait loop in plan_buffer_line,
which currently expands all comparisons to a word for no reason.
This also extends the theoretical limit to 128 entries.
Add some static assertions to ensure BLOCK_BUFFER_SIZE is correct.
This avoids the "busy" output interleaving with regular output in very
rare scenarios.
We should focus in finding which calls are not using manage_inactivity()
properly instead of working it around.
Code running in the temperature ISR needs to be fully reentrant, which
is hard to track down.
Move autoreporting to the main processing loop. This can make the
autoreporting slower or pause at times, but removes the reentrant
restriction, which allows us to use printf_P.
This PROGMEM is currently ignored by gcc, but even if it wasn't it
wouldn't be correct since the following code is expecting to read "item"
without fetching the array itself from PROGMEM.
The longpress function is currently called within the temperature ISR,
which is bogus. Calling the longpress function at the wrong moment
can corrupt the menu buffers.
Move the call to the main loop by changing the logic slightly:
- still sample the lcd buttons inside the temperature ISR, which keeps
scrollong/pressing responsive, but...
- set a flag to indicate that longpress has been triggered instead of
calling the function directly
- call the function on the next manage_inactivity run
Combined with #3180 this removes _most_ unsafe operations out of the ISR
which can happen during a normal run (max/mintemp warnings are still an
exception).
Instead of having to guess the PC where the SP was sampled, always take
both. This allows "seamless" stack decoding for both serial and xflash
dumps, since we don't have to guess which function generated the dump.
Make the core functions (doing the sampling) be ``noinline`` as well,
so that they always have valid frame.
Save SP which is closest to the crash location, which simplifies
debugging. For serial_dump, write SP just before the dump.
For xfdump, save SP in the dump header.
This makes xfdump_dump and xfdump_full_dump_and_reset() equivalent for
stack debugging.
Now that the stack_error function is truly minimal,
we can check for stack errors much more frequently.
Also move away stack_error from ultralcd to Marlin_main.
Rename EEPROM_CRASH_ACKNOWLEDGED to EEPROM_FW_CRASH_FLAG.
Use EEPROM_FW_CRASH_FLAG to always set the last crash reason, which
simplifies handling between the online/offline variants.
Make stack_error safe, by setting the flag and restarting immediately,
so that the error can be shown after restart.
When XFLASH is not available, allow users to request _online_ crash
dumps by using D23 (since these require active user cooperation).
Once enabled, instead of just rebooting, dump memory directly to
the serial.
As similarly done with EMERGENCY_DUMP, we have two features that can be
enabled:
EMERGENCY_SERIAL_DUMP: enables dumping on crash after being requested
MENU_SERIAL_DUMP: allow triggering the same manually through the support
menu.
Create a gap between the BSS and the stack guard.
Set this gap (STACK_GUARD_MARGIN) to 32 bytes in all variants.
The gap serves two purposes:
- Detect a stack overflow earlier (falsely triggering in overtight
situations is OK!), so that we can hopefully avoid smashing
the heap and have a clean view during the dump.
- Reserve spack space itself for the stack dumping machinery,
which is going to grow the stack even further.
Remove get_stack_guard_test_value() which was unused.
As suggested by @3d-gussner, announce to the host that a dump is
available for retrieval using an action "dump_available".
Any kind of dump is announced (even if manually triggered).
To avoid reading from xflash twice, remove some duplication and return
the crash reason directly in xfdump_check_state().
If MENU_DUMP is enabled, a new entry at the end of the "Support" menu is
added that allows to dump memory for offline use.
This allows to trigger a memory dump at any moment during regular usage
(either idling or printing) and to recover the dump later even after a
hardware reset.
If EMERGENCY_DUMP is defined, crash and dump using the new xflash dump
functionality instead of just continuing with an error message.
When an emergency crash is stored, the first restart after a crash
displays a message that debug data is available and to contact support
to submit the crash for analysis.
This requires expanding the dcode_core address type to 32bit type,
thus enlarges the D2/D3 implementation as a result.
Still allow to save all the original space if D6 is disabled, for now.
Allow to read up to 0x21ff, which is the last byte of SRAM. Set default
starting address to 0x200, which is the first true byte.
0x0-200 is mapped to register/io space and could cause issues when read
from bi-directional registers.
Farmers want to abuse a bug from the previous firmware releases
- they need to see the filename on the status screen instead of "Wait for user..."
So we won't update the message in farm mode...
Check for lcd_draw_update to see if the lcd has been altered outside the
function and redraw the screen if full.
This fixes scenarios such as #3129 where the "Card removed" message or
the SD menu is entered from outside the function's control.
This requires checking/resetting bFilamentWaitingFlag carefully to avoid
re-entering this function _twice_ (raise_z_above will run the main
loop until complete).
Scale extruder motor current linearly with speed.
49% less heating when running at low speed and standstill, 4% more torque at maximum extrusion rate (15mm^3/s), 15% more torque in high speed movements (un/retractions).
StealthChop mode is used for low speeds (below 900mm/min)
spreadCycle is used above. Transition speed is well above maximum extrusion rate of 15mm^3/s (275mm/min) so mode transition is not expected to be visible on printed surface.
StealthChop is expected to improve printed surface quality (less artifacts).
Warning you can burn extruder motor if it is not the same impedance as original Prusa i3 Extruder stepper motor. There is no current feedback in low speed so lower impedance motor can be burned by over current.
Even there is no direct current feedback, there is no risk for original motor thermal runaway, as motor resistance increases with temperature, current decreases.
Standstill peak phase current is expected to be 500 mA and linearly increase with speed to 970 mA at 900mm/min where spreadCycle constant current regulation takes over and keeps peak current at 805 mA to maximum speed possible.
As motor heating increases with current squared, lowering low speed current from 700mA to 500mA decreases heating 49% in thate mode, where motor spends most of the time.
Enable E-motor cool mode in farm mode only (and experimental menu) - the experimental menu is visible AND the EEPROM_ECOOL variable has a value of the universal answer to all problems of the universe - i.e. two conditions must be met at the start of the FW to enable the E-cool mode. If the user enables the experimental menu, sets the E-cool mode and disables the menu afterwards, on the next start of the FW the E-cool mode will be DISABLED. This is still subject to discussion how much obscure (security through obscurity) we'd like this option to have .
Additional stuff:
* Add serial debug msg to verify if E-cool mode is on
* Avoid access to E-cool mode switch on machines without TMC2130
* Do not allow only M907 E in case of E-cool mode+warn the user on the serial line that the command was skipped
Co-authored-by: D.R.racer <drracer@drracer.eu>
gcode_M600_filament_change_z_shift is almost useless, since it performs
what is already been done internally by filament_unload().
However it *does* cause the carriage to raise earlier during unload
compared to making the user wait after "press for unload".
Change it so the calculated Z height matches MIN_Z_FOR_SWAP.
This change restores the minimum extruder height for filament purge
during M600 from the current 50mm back to 27mm from FW 3.9.
We do this by introducing a new option for unload_filament() to indicate
that the unload is part of an automatic swap, and in such cases avoid
raising more than absolutely necessary (this will _also_ come in handy
to avoid the extra purge in PR #2318 during M600).
A new define MIN_Z_FOR_SWAP is introduced for this purpose.
MIN_Z_FOR_UNLOAD is still used for manual lcd unload and for M702 and
hasn't been changed.
If the printer was already being pre-heated but didn't reach the target
temperature yet and a new filament is being inserted, the LCD used to
display a "Preheating to load" message to block the loading until the
extruder is hot.
This message is currently missing, and the ">Cancel" option doesn't
display immediately either, depending on the extruder height.
This PR fixes this behavior, which was broken during an earlier update.
We now force-update LCD updates during the first (and _only_) time the
screen is setup, and push all messages _before_ the carriage is
eventually raised, so that ">Cancel" is shown immediately as well.
mesh_bed_leveling_flag and homing_flag should be set only when the
planner is empty as a consequence of #3100, since any operation executed
*while* the mbl/homing flag is set is considered part of the compound
instruction that generated it (G80, G28, G30, G76).
Failure in doing so can result in instructions just prior being lost,
since they're incorrectly discarded (assumed to be redone).
G1 ; lost ..
G80 ; .. if PP happens here
PR #2967 altered the way ``sdpos_atomic`` was set, causing issues in the
crashdetect/powerpanic recovery offset if the instruction being
recovered happens to contain a comment.
Previously ``sdpos`` was assumed to be a single byte prior to the last
read character. sdpos+1 would thus position the index to the next
instruction. With gcode-filtering in place, sdpos is left just before
the comment, while the actual read position is at the newline. This
causes to parser to resume in the middle of the comment.
Change the value returned by cardreader::get_sdpos() to always return
the last read position, as everybody expects (!!).
This avoids the +1, and correctly sets the resume position to the next
valid instruction without overhead.
Fix cmd overflow introduced in 186ce0f.
With a precision of %f being .6 and assuming a maximum (theoretical)
acceleration of 99999 we need 47 bytes including terminating byte.
Round it to 48
This is handled in the same way crash detection is handled: homing/mbl
invalidates saved_target _and_ current position.
Fixes PP recovery during MBL and homing.
At a higher level, crash detection will repeat the last gcode command.
Some commands such as G28 or G80 need to be repeated in full and require
special handling.
In such cases, do not store the saved target coordinate (which is fake
in such cases) and invalidate the saved position.
This fixes the last coordinate of the first G1 move after recovery.
We also avoid moving from the origin when the saved position is
invalidated, which is not strictly necessary, but saves some time.
Crash detection is handled at a higher level (which automatically
repeats the current command), however we still need to abort the current
command correctly.
Handle XY crashes by checking the planner status after st_synchronize().
No actual changes done in the function in this, besides break->return.
In G28, simply call the new function instead of using goto, making the
code more readable.
Also remove the senseless comments in G28 about command queuing (dating
back to when G80 was queued instead of being executed).
As done when initializing the printer from a cold start, we need to
clamp the starting position to software endstops before setting the
planner position since 0,0 is frequently out-of-bounds.
This avoids an useless move during recovery that can cause a crash:
- Initial X is set to be 0
- G1 performed by homing will clamp X>=0, resulting in a positive shift
- If X is already at max X (extruder being parked due to PP), this will
slam at X+, causing an immediate crash.
