When building Kaleidoscope, the compiled object files are linked together into a
static archive. This static archive has a very simple structure, and only stores
filenames, not paths, not even relative ones. As such, we can't have files with
the same name, because they will conflict, and one will override the other.
To avoid this situation, this script will find all cpp source files (we don't
need to care about header-only things, those do not result in an object file),
and will comb through them to find conflicting filenames.
If a conflict is found, it will print all files that share the name, and will
exit with an error at the end. It does not exit at the first duplicate, but will
find and print all of them.
If no conflict is found, the script just prints its status message and exits
with zero.
This addresses the bulk of #850.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
for some reason, this line shows up on stderr anytime I (or the
find-device-port script) run `system_profiler SPUSBDataType`
2020-02-24 23:19:14.656 system_profiler[23932:4118769] SPUSBDevice:
IOCreatePlugInInterfaceForService failed 0xe00002be
I got annoyed enough to suppress it, so I could see the other output
that I actually care about.
This adds some quotes to various paths used in Kaleidoscope's build system.
This fixes builds on msys2 that failed due to whitespaces in system paths.
Signed-off-by: Florian Fleissner <florian.fleissner@inpartik.de>
Many build systems allow C and C++ compilers to be specified
via environment variables C and CXX.
This commit enables this for kaleidoscope-builder.
Furtheron, on unixoid systems virtual builds are possible through a
command line similar to
CXX=<path to C++ compiler> C=<path to C compiler> ARCH=virtual make
Signed-off-by: Florian Fleissner <florian.fleissner@inpartik.de>
This commit induces the following changes:
bin/kaleidoscope-builder
* ccache dummy compiler and executables are now generated in a way
that allows using the same CCACHE_WRAPPER_PATH for virtual and
non-virtual builds
* virtual builds are now triggered by either specifying the full FQBN with x86
as architecture or by defining ARCH=x86
etc/kaleidoscope-builder.conf
* COMPILER_PREFIX and COMPILER_PATH are now determined (if not predefined)
based on ARCH or FQBN
Signed-off-by: Florian Fleissner <florian.fleissner@inpartik.de>
1. Better matches other _PATH env variables
2. Since it's something that might reasonably get set in the
environment, prefix it with KALEIDOSCOPE
Signed-off-by: Jesse Vincent <jesse@keyboard.io>
This is particularly important for in-kaleidoscope-tree artifacts like
the examples.
If we build the examples and move the compiled artifacts into the
Kaleidoscope tree, even under examples, it busts arduino's core
compilation cache, adding an extra 10s to the next compilation.
Signed-off-by: Jesse Vincent <jesse@keyboard.io>
We only cared because we were trying to get a nice version number
That doesn't matter as much as being able to build at all
Signed-off-by: Jesse Vincent <jesse@keyboard.io>
Apparently, the shellcheck on Travis does not like comments after a disable
command, while the one I used locally didn't have a problem with it. Lets make
the one on Travis happy.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
When calling avr-size, we want to expand ${AVR_SIZE_FLAGS}, not pass it as a
single argument, hence, we do not need to - and do not want to! - quote it.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
* Uses usbconfig to determine the Model 01's USB modem port.
* Works around incompatibe avrsize flags.
You need to be able to run usbconfig to flash the firmware from the
buildtools. This can be accomplished with appropriate groups and devfs
rules.
Requires gmake, perl, avrdude, and (probably) arduino18 from ports.
The version of avrdude in ports uses an avr-size command that doesn't
understand the -C or --mcu flags. From what I can tell, these flags
are uneccessary, as the size computed with them is the same as what
you get from adding up the appropriate segments from the standard
output of avr-size without any flags. However, since the size is only
informative, I've opted to simply check to see if the command
succeeded, and if not, output a string saying it could not be.
It would probably be better to:
* Determine appropriate flags based on build tools, or,
* Just not use the flags at all, and grab the .text, etc., segment
sizes from the standard output and add them up via `dc` for
display.
I've been using this toolchain to build successfully on FreeBSD 12 for
the Model 01 without issue. It should work with earlier versions of
FreeBSD as well.
Signed-off-by: Brian Cully <bjc@kublai.com>
For better organisation, we're putting examples in subdirectories of
`examples/`. The builder should support that.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
If there's a config file in the sketch dir, load BOARD & FQBN from it, so we can
compile for different boards from within the same repo.
Fixes#425.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
After talking with Jesse, this changes the license to GPLv3 (only), where
appropriate, and adds copyright headers to all files that were missing them.
Signed-off-by: Gergely Nagy <algernon@madhouse-project.org>
Use ./_build/ for the build path by default, and do not delete it at the end of
compilation, only when doing a clean.
Fixes#315.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Instead of hardcoding the max program size, pick it out of boards.txt, like we
pick the device VID and PID.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Based on #306, with slightly improved text. Thanks to Ross Donaldson
(@Gastove) for the original pull request!
Closes#306.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
There are a number of false-positives, where ShellCheck warns about behaviour we
do want, or are otherwise intentional.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
If the user has missed the step about setting up their account
with the right group membership, they would get a cryptic failure
from stty, so catch this and explain the problem.
This should never happen, but could if something goes badly wrong
in the device detection code, e.g. someone changing it in a way
which caused extra output on STDOUT.
Further testing suggests that macOS unconditionally replaces the final character with a
`1` (I tested with `kbio02` as the device short name, which became the "Serial Number"
`CDkbio02`, and it got changed to `CDkbio01` in the filename). This change will try
looking for any file that matches the serial number with the final character replaced with
"1" (not just if that charater is "E"). I also added some comments that might help update
the script in the future if the OS changes behaviour.
Several people have reported difficulty flashing firmware on macOS High Sierra because the
device port filename doesn't match the serial number from system_profiler. In particular,
system_profiler would return a string ending in `E` whereas the device filename would have
a `1`. This change adds a check for that filename explictily.
I also corrected the location_id fallback (the substring should have been just 3
characters long, not 4), and it works properly on my system if I make the device shortname
7 characters long, and the filename reverts to using the location id instead.
Last, I added one more check, simply listing the filenames, and searching for a match for
the string `kbio01`, which should be present (although in one case, it wasn't).
It's easier for most people to interpret numbers in decimal than hexadecimal; adding this
option to avr-nm prints out the sizes in the size_map in a more intuitive format.
On some locales awk will generate commas instead of dots in the percentage. This will mess up the sed command. Instead of forcing awk to always generate dots we can simply cange the sed command to treat commas as regular text.
Currently, if kaleidoscope-builder fails to find a sketch file, the
resulting error message is very confusing and unhelpful. This commit
makes it more straightforward and helpful.