| Age | Commit message (Collapse) | Author |
|
Fix #189
|
|
|
|
On windows the AVRdude configuration file is now detected properly. The
comments and `README` are updated as well.
Fix #187
|
|
|
|
|
|
Fix #180
Fix #127
|
|
other two BOOTLOADER variables
Fix #178
|
|
Fix #176
|
|
AVRDUDE requires -P usb to be passed as an argument for some ISP
devices, and it is also required in the case where you have multiple
usbtiny or usbasp devices connected at once to select a specific one.
Fix #167
|
|
Fix #175
Fix #174
|
|
Clean up markdown and add code blocks where needed
Fix #171
|
|
Signed-off-by: tinyladi <ladislas@weareleka.com>
|
|
|
|
This results in just a deprecated warning rather than an error
when using libraries (e.g. jeelib) that still use prog_XXX types
This is how the IDE does it, and means that libraries don't have to
add the #define __PROG_TYPES_COMPAT__ line
Reference: http://www.nongnu.org/avr-libc/user-manual/group__avr__pgmspace.html
Fix #169
|
|
The gpio programmer type is used on the raspberry pi to upload using
the gpio spi pins via a sysfs interface, with a modified avrdude
Fix #165
Fix #166
|
|
if bug-reporter confirms this fixes issue #163 then should be ok to merge.
|
|
Added some notes to the SPECfile regarding overriding the paths
to the avr tools - most are in /usr/bin, but gcc/g++ are in an
architecture-dependant ccache directory
Upped version to 1.3.2 devel
|
|
|
|
This target can display all variables that can be overridden
|
|
Huge thanks to @sej7278 for doing this
Fix #75
|
|
so that installations using only the arduino-core packages that
don't have the $(ARDUINO_DIR)/hardware/tools/avr directory can
still use the avr-g++ tools found in the $PATH (/usr/bin)
Previously BUNDLED_AVR_TOOLS_DIR was set to an empty string as
the directory doesn't exist, which meant that "ifdef BUNDLED_AVR_TOOLS_DIR..."
was set rather than skipping to "else SYSTEMPATH_AVR_TOOLS_DIR...."
No user would set BUNDLED_AVR_TOOLS_DIR so the assignment operator
:= should be used not ?=
Updated the version info/changes in various locations.
|
|
|
|
Fix #94
|
|
|
|
Fix #156
|
|
|
|
Fix #154
|
|
Fix #147
Fix #151
Fix #153
The following are the detailed changes
- Remove ARDMK_PATH, ARDMK_FILE and arduino-mk subdirectory
- Looks for ard-reset-arduino in $PATH or ARDMK_DIR/bin
- Fix git-archive command in RPM SPEC file
- Remove some whitespace
- Remove arduino-mk dir from debian package's arduino-mk.install
- Update docs to reflect the above changes
- Bump Up version to 1.2.0
|
|
|
|
|
|
Fix #131
|
|
Changing the optimization level from -Os to something else breaks the
SoftwareSerial library. This patch allow optimization flags to be set on a
per target basis.
|
|
|
|
|
|
|
|
When using ICSP programmers e.g. usbasp, you can burn sketches directly
to the chip without having to burn a bootloader, however you do need to
set fuses e.g. if you're changing speed/BOD.
Updated help text to include "make set_fuses" and also changed
"make burn_bootloader" help text which wasn't entirely accurate.
Fix #141
|
|
This way we don't assume the make file is 'Makefile' and it will
work even if it is named as 'makefile' or 'makefile-uno'
Fix #130
|
|
Fix #135
|
|
|
|
rationale:
http://unix.stackexchange.com/questions/65803/why-is-printf-better-than-echo
fix #129
|
|
Fix #120
|
|
Code to burn fuses is moved from ispload target to this new target, so
that fuses are burned only once when needed.
Fix #85
|
|
Arduino IDE also reads this from boards.txt file
Fix #125
|
|
Fix #124
|
|
When usbasp is used as programmer, there is no need to specify the port
to avrdude.
Fix #123
|
|
No change in functionality
|
|
This is needed in Mega boards, because of a bug in Mega bootloader. See
https://github.com/sudar/Arduino-Makefile/issues/114#issuecomment-25011005
for detailed explanation.
Fix #114
|
|
|
|
If these executables are not overridden in the child makefile, then the
default values are not set properly in the master makefile.
This complicates the solution for #119 :(
This reverts commit 66eec82a3af8ee5aa16e28230346e4a1f166da2b.
|
|
Lot of changes went it for this release and the code is stable enough
for 1.0.0
|