Also fixes warnings (and bugs) in F4 STM32_USB_OTG_Driver code, allowing -Werror to be enabled for all flight code.
Fixes all other compiler warnings that would otherwise cause the flight code to not compile with -Werror enabled.
Along the way, this also adds some uses of isnan() to various places rather than questionable tests for x != x and
x == x to check for NaNs.
+review OPReview
It seems we must ship GCS with all Qt libs like we do for Windows.
On Ubuntu x64 the GCS built with this crashes using system libs.
But since it's possible to build GCS using natively installed
Qt SDK, it is not a problem for now. Something is better than
nothing, and two ways of building are better then just one.
+review OPReview-441
This is a part of OP-726, OP-727 tasks to normalize source code.
Use the following to normalise your local repository to be able to merge:
git rm --cached -r .
git diff --cached --name-only -z | xargs -0 git add
git commit -m "Normalise line endings"
git ls-files -z | xargs -0 rm
git checkout .
This changeset:
- moves all ground targets one level up, under build directory. The
build/ground was created as a workaround, now unnecessary;
- fixes QtCreator builds, they are separated from command line builds;
- moves GCS autogenerated files into openpilotgcs-synthetics directory.
The resulting build subdirectory now looks like:
build
openpilotgcs <- Qt-Creator build directory
openpilotgcs-synthetics <- version-info and opfw_resource
openpilotgcs_debug
openpilotgcs_release <- Makefile build directory
uavobject-synthetics
uavobjgenerator
NOTE: you should update the shadow build path in QtCreator to build,
not build/ground as before.
+review OPReview
The problem is that we have modules defined as:
Airspeed
or
Airspeed/revolution
or
Extensions/MagBaro
and, probably
Extensions/MagBaro/revolution
can be expected too.
It is impossible to get module name from such mixed cases.
So the better solution does not work for MagBaro, reverted.
This reverts commit 5d760d20fa.
+review OPReview-422
This is the first cleanup pass through makefiles and pios.
Probably it is difficult to track changes due to the nature of them.
I would recommend to look at resulting files and compiled code instead.
NOTE: original branch was rebased and lot of conflicts were fixed on
the way. So do not expect that every commit in this series will be
buildable (unlike original branch). Only final result was tested.
The main goal was to remove as much duplication of code (and copy/paste
errors) as possible, moving common parts out of Makefiles. It still is
not perfect, and mostly no code changes made - Makefiles and #ifdefs only.
But please while testing make sure that all code works as before, and no
modules/options are missed by accident.
Brief list of changes:
- Moved common parts of Makefiles into the set of *.mk files.
- Changed method of passing common vars from top Makefile to lower ones.
- Some pios cleanup, mostly #ifdefs, and all pios_config.h files.
- Many obsolete files removed (for instance, AHRS files, op_config.h).
- Many obsolete or unused macros removed or fixed/renamed (ALL_DIGNOSTICS).
- Unified pios_config.h template. Please don't remove lines for board
configs, only comment/uncomment them. Adding new PIOS options, please
propagate them to all board files keeping the same order.
- Some formatting, spacing, indentation (no line endings change yet).
- Some cosmetic fixes (no more C:\X\Y\filename.c printings on Windows).
- Added some library.mk files to move libs into AR achives later.
- EntireFlash target now uses cross-platform python script to generate bin
files. So it works on all supported platforms: Linux, OSX, Windows.
- Top level packaging is completely rewritten. Now it is a part of top
Makefile. As such, all dependencies are checked and accounted, no
more 'make -j' problems should occur.
- Default GCS_BUILD_CONF is release now, may be changed if necessary
using 'make GCS_BUILD_CONF=debug gcs'.
- GCS build paths are separated into debug and release, so no more obj
file clashes. Packaging system supports only release builds.
- New target is introduced: 'clean_package'. Now 'make package' does not
clean build directory. Use clean_package instead for distributable builds.
- Targets like 'all', 'opfw_resource', etc now will print extra contex
in parallel builds too.
- If any of 'package', 'clean_package', 'opfw_resource' targets are given
on command line, GCS build will depend on the resource, so all fw_*.opfw
targets will be built and embedded into GCS. By default GCS does not
depend on resource, and will be built w/o firmware (unless the resource
files already exist and the Qt resource file is generated).
- fw_simposix (ELF executable) is now packaged for linux. Run'n'play!
- Make help is refined and is now up to date.
Still broken:
- UnitTests, should be fixed
- SimPosix: buildable, but should be reworked.
Next planned passes to do:
- toolchain bootstrapping and packaging (including windows - WIP)
- CMSIS/StdPeriph lib cleanup
- more PIOS cleanup
- move libs into AR archives to save build time
- sim targets refactir and cleanup
- move android-related directories under <top>/android
- unit test targets fix
- source code line ending changes (there are many different, were not changed)
- coding style
Merging this, please use --no-ff git option to make it the real commit point
Conflicts:
A lot of... :-)
They are not a PiOS, but 3rd party libraries. They even should not
be placed to the same (PiOS) library directory due to the printf symbol
clashes (if all *.c from the folder are built). So they are moved, and
each target can include that one which is necessary/desired.
The difference was in dfs_sdcard.c, line 107:
-if((status = PIOS_SDCARD_SectorRead(sector, buffer)) < 0) {
+if((status = PIOS_SDCARD_SectorRead(sector, buffer)) != 0) {
Currenly unused, kept as is.
Now PipXtreme uses the same apps-defs.mk file as CopterControl.
Next steps are F4 boards.
Conflicts:
flight/targets/PipXtreme/Makefile
flight/targets/PipXtreme/System/inc/pios_config.h
These names are automatically generated by make using the list
of mandatory (always started) modules. Wrong names mean they can
be disabled, fixed now.
Conflicts:
flight/Modules/Radio/radio.c
Note that $(QUOTE) variable is uppercased. Windows make does not
export lowercased variables with export keyword (even if must do).
Conflicts:
make/firmware-defs.mk
Make targets changed to:
make ut_<test>
make ut_<test>_tap
make ut_<test>_run
make ut_<test>_clean
make all_ut
make all_ut_tap
make all_ut_run
make all_ut_clean
Conflicts:
Makefile
This template will be used to comment auto builds created with Bamboo.
Currently it is emulated by Bamboo build agents using build scripts,
but should be generated using Makefile. As soon as majority of branches
will be updated to this next, those scripts will be removed and this
template used.
Bamboo does not preserve git repository origin when making builds, and
it is shown later in About dialog as "Built from None". This patch fixes
that providing more accurate source description.
${REVISION} - revision info string (tag or branch:hash date time)
${LABEL} - package label string (tag or date-hash)
${DAY} - day of last commit
${MONTH} - month of last commit
${YEAR} - year of last commit
Conflicts:
package/Makefile
${REVISION} - revision info string (tag or branch:hash date time)
${LABEL} - package label string (tag or date-hash)
${DAY} - day of last commit
${MONTH} - month of last commit
${YEAR} - year of last commit
Line endings in the UAVO XML files differ from file to file
and this results in different SHA1 sums being computed on
Windows vs. Linux/Mac.
This normalizes the line endings to LF (python native) while
processing the XML files.
In python, os.walk() is not sorted. This can result in the
hash being computed over the files in different orders on
different OSes, or even after touching a directory.
This ensures that the files are considered in lexical order
so that the hash is stable.
Removes hard-coding of JTAG interface config in the
<board>_program make macros.
This allows the use of STLINKv2 for F4 boards while
continuing to use the FOSS JTAG revB on F1 boards.
When building the various all_* targets, it was hard to tell which
board/build-type that each line of output applied to. Now, the
all_* target types will include something like:
CC [fw|cc ] flight/PiOS/STM32F10x/pios_gpio.c
which includes the necessary additional context.
This will help with identifying the context for warnings and errors
when building a group of targets.
Conflicts:
Makefile
When building the various all_* targets, it was hard to tell which
board/build-type that each line of output applied to. Now, the
all_* target types will include something like:
CC [fw|cc ] flight/PiOS/STM32F10x/pios_gpio.c
which includes the necessary additional context.
This will help with identifying the context for warnings and errors
when building a group of targets.
This makes the BL and FW images distinct devices with unique
serial numbers.
Windows (and maybe Mac) remembers the device descriptors and
the associated drivers based on this serial number. Providing
unique serial numbers for the BL and FW images allows us to
provide different sets of descriptors for the BL and FW images
without confusing these OSes.
BL version number is now also bumped to reflect the new
serial number generation algorithm.
The F4 CPUs have 16KB x 4, 64KB x 1 and 128KB x 7 sectors
of flash. The FW image currently ends in the middle of a
128KB sector. FW_BANK is now increased to end on a sector
boundary.
The code in RCC_GetClocksFreq() uses HSE_VALUE to determine
the tick rate of the SYSCLK. This in turn is used by the
code in PIOS_DELAY_* to compute delays.
ST Library defaults this to 25MHz for the F4. Our board
actually has an 8MHz oscillator so delays were way off.
Mostly affects the bootloaders since most FW code uses RTOS
delays rather than busy-waiting using PIOS_DELAY_*.
Removes hard-coding of JTAG interface config in the
<board>_program make macros.
This allows the use of STLINKv2 for F4 boards while
continuing to use the FOSS JTAG revB on F1 boards.
It is now easier to trigger warm restarts of a board
via jtag.
Examples:
* make fw_coptercontrol_reset
* make fw_coptercontrol_safeboot
NOTE: These targets are making chip-specific assumptions
so they have to be rewritten to support the F2/F4
boards.
Macros for JTAG program and wipe for each target are now
provided in firmware-defs.mk.
The _wipe target for each firmware and bootloader image will
erase either the bootloader (bl_*_wipe) or firmware (fw_*_wipe)
bank.
- New macros for fw, bl and bu rules in top-level make
- Per-board info factored into make/board/*/board-info.mk
- Per-board info now shared btw. fw, bl and blupd for each board
- BOARD_TYPE, BOARD_REVISION, BOOTLOADER_VERSION, HW_TYPE
- MCU, CHIP, BOARD, MODEL, MODEL_SUFFIX
- START_OF_BL_CODE, START_OF_FW_CODE
- blupd_* goals renamed to bu_*
- all_blupd goal renamed to all_bu
- firmware goals renamed to fw_*, board name goals are preserved
- bu_*_program now writes updater to correct address for all boards
- BL updater firmware builds now produce .opf format including
version info blob.
- BL updater firmware name now includes board name.
- INS makefile brought up to date w.r.t. linker scripts
Since msysGit lacks the install command, simple cp -f replacement
is provided. Don't forget to copy it into the msysGit bin subdirectory.
Introduced CLEAN_BUILD=NO make option to test packaging w/o cleaning.
Recommended for test builds only, not for production run.
make -j works great after PyMite dependency fix by commit
0005729056
So comment is updated as well.
USE_BOOTLOADER now deprecated and removed.
Scripts are moved into the same subdirectories like in msysGit.
Added make.sh script (for advanced usage scenario).
Few misprints are fixed.
This change is made up of a number of tightly coupled
changes:
- Deprecate the use of the USE_BOOTLOADER command-line
option. It is now hard-coded in each Makefile.
Overriding it on the command line is not allowed.
- Split apart the memory declaration and the section
declaration in all linker files (*_memory.ld and
*_sections.ld).
- Describe the split between bootloader and app sections
of flash in each board's _memory.ld file.
- Change program target to selectively erase flash so
that the installed bootloader is preserved across even
JTAG programming operations.
- All elf files are built with debug symbols and are not
stripped. This should help debugging with gdb. The
images programmed on the boards are all .bin files now
which do not include symbols.
The .bin.o rule places the contents of a raw .bin file
into an .o file wrapped within fixed symbols for start
and end. This can be used to embed a binary file inside
of an executable.
The symbols for the embedded binary blob are:
_binary_start
_binary_end
_binary_size
NOTE: The way the .bin.o rule is currently written, you
can only embed one binary blob in an executable since the
symbol names will collide if you add multiple blobs. This
limitation is easily removed later if necessary.
firmware include and delete all the extra foss-jtag config files. There is now
a legacy file for the revA board a second for AHRS that changes the port
This significantly reduces the amount of duplication
across the various firmware makefiles.
The new firmware-defs.mk file should contain only
macros/declarations that will apply to all firmware
makefiles.
git-svn-id: svn://svn.openpilot.org/OpenPilot/trunk@2854 ebee16cc-31ac-478f-84a7-5cbb03baadba
