regards to the data used by the ROM.
Static alloaction sections should end at the begining of the ROM data.
The rest of memory (End of ROM data --> End of DRAM) is added to the
heap.
Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
This adds the inital wiring for i2c bus support in the sim target
and for Linux host adds the lower half that uses the i2c chardev.
Signed-off-by: Brennan Ashton <bashton@brennanashton.com>
Squashed commit of the following:
sim audio: call alsa to playback/capture data
sim/audio: correct the format capability
sim/audio: add pause/resume support
sim/audio: add auto stop when meet AUDIO_APB_FINAL
sim/audio: fix abort when set small buffer_size
sim/audio: move sim_audio.c to sim_alsa.c
Change-Id: I8e00ece79159e844ca17fd4c363480b985ee0490
Signed-off-by: ligd <liguiding1@xiaomi.com>
In the current compilation environment, the recursive assignment(=) for compile
flags will be delayed until every file is actually need to be compile.
For example:
--------------------------------------------------------------------------------
arch/arm/src/Makefile:
INCLUDES += ${shell $(INCDIR) "$(CC)" $(ARCH_SRCDIR)$(DELIM)chip}
INCLUDES += ${shell $(INCDIR) "$(CC)" $(ARCH_SRCDIR)$(DELIM)common}
INCLUDES += ${shell $(INCDIR) "$(CC)" $(ARCH_SRCDIR)$(DELIM)$(ARCH_SUBDIR)}
INCLUDES += ${shell $(INCDIR) "$(CC)" $(TOPDIR)$(DELIM)sched}
CPPFLAGS += $(INCLUDES) $(EXTRAFLAGS)
CFLAGS += $(INCLUDES) $(EXTRAFLAGS)
CXXFLAGS += $(INCLUDES) $(EXTRAFLAGS)
AFLAGS += $(INCLUDES) $(EXTRAFLAGS)
--------------------------------------------------------------------------------
All compilation options will be included recursively,
which will be delayed until the compilation options are actually used:
tools/Config.mk:
--------------------------------------------------------------------------------
define COMPILE
@echo "CC: $1"
$(Q) $(CC) -c $(CFLAGS) $($(strip $1)_CFLAGS) $1 -o $2
endef
--------------------------------------------------------------------------------
All compile flags to be reexecuted $(INCDIR) as long as one file needs to be compiled,
but in fact, the compilation options have not changed in the current directory.
So the we recommand to change the syntax of assignment
From
Recursive (=)
To
Simple (:=)
In this way, we can ensure that all compilation options are expanded only once and reducing repeated works.
Signed-off-by: chao.an <anchao@xiaomi.com>
## Summary of changes
Disable CONFIG_NSH_ARCHINIT to prevent the boardctl() logic to call
multiple times the oneshot_register() function. The boardctl logic is
already called from nxwm_main.
Signed-off-by: Sebastian Ene <sene@apache.org>
This separation allows to interact with the watchdog from OS code,
for example initiating the watchdog very early on boot. Moreover,
these changes make the lower-half driver support an already running
watchdog, which may happen if there's a bootloader which already
started it.
Follow the POSIX description.
SIGTSTP should be sent when the Ctrl-Z characters is encountered, not SIGSTP.
Testing:
Built with hifive1-revb:nsh (CONFIG_SERIAL_TERMIOS=y, CONFIG_SIG_DEFAULT=y and CONFIG_TTY_SIGTSTP=y)
Add new option -A is added to tools/testbuild.sh that will take the
created build executable and store it in a folder for the config
that generated it under $ARTIFACTDIR which can be set via an
environment variable or defaulted to $(TOPDIR)/buildartifacts
This is also helpful for local testing because you can now run
tools/testbuild.sh -A sim.dat and have all of the simulation
targets generated without having to rebuild along the way.
In the GitHub Actions workflow the artifacs are uploaded
two two bundles one for macOS and one for Linux
Signed-off-by: Brennan Ashton <bashton@brennanashton.com>
the mkspk tool will automaticaly be built if CONFIG_CXD56_BINARY
is set.
Test:
nuttx.spk was succesfully generated
Signed-off-by: Alin Jerpelea <alin.jerpelea@sony.com>
### Summary of Changes ###
Running the NuttX simulation 'as fast as possible' breaks the features
that depend on timing: eg. the Bluetooth stack. Enabling this option by
default SIM_WALLTIME=y will introduce delays and will tick the simulation at
a real pace.
Refresh defconfigs for sim targets and add depends on SIM_WALLTIME for
SIM_HCISOCKET.
Signed-off-by: Sebastian Ene <sene@apache.org>
There is a good case on sim platform:
When we input some cmd and click enter key to start application in terminal,
this context will change to application from IDLE loop. Althrough entey key '\r'
has been received to recv buffer and complete post semaphore of reader, but
pollnotify may not be called because context change. So when application run
poll function, because no events happend and poll enter wait, context will
again change to IDLE loop, this pollnotify of IDLE loop will run to send poll
events, poll function of applicaton will wake up. It's wrong!
Change-Id: I812a889f2e90781a9c3cb4b0251cccc4d32bebd1
Signed-off-by: dongjiuzhu <dongjiuzhu1@xiaomi.com>
Fixes an issue I saw when trying libcxx.
esptool.py has a limit of 16 sections.
It complains like:
A fatal error occurred: Invalid segment count 23 (max 16).
Usually this indicates a linker script problem.
Usage:
nsh> ifconfig
eth0 Link encap:Ethernet HWaddr 52:13:FF:FF:FF:FF at UP
inet addr:10.0.0.2 DRaddr:10.0.0.1 Mask:255.255.255.0
nsh> renew eth0 && ifconfig
eth0 Link encap:Ethernet HWaddr 52:13:FF:FF:FF:FF at UP
inet addr:192.168.1.42 DRaddr:192.168.1.254 Mask:255.255.255.0
CONFIG_NETDB_DNSCLIENT is needed for
apps/netutils/netinit/netinit.c:659: \
undefined reference to `netlib_set_ipv4dnsaddr'
Change-Id: I54b8495637baf0f3f28abf67d4d4ce861ef49ced
Bug: https://github.com/apache/incubator-nuttx/issues/1988
Forwarded: https://github.com/apache/incubator-nuttx/pulls/rzr
Signed-off-by: Philippe Coval <rzr@users.sf.net>
It was tested using:
nsh> ifconfig
eth0 Link encap:Ethernet HWaddr 52:13:FF:FF:FF:FF at DOWN
inet addr:10.0.0.2 DRaddr:10.0.0.1 Mask:255.255.255.0
All flags are needed to prevent build failures like:
chip/kinetis_enet.c:181:4:\
error: #error "Unrecognized or missing PHY selection"
netlib_setifstatus.c: In function 'netlib_ifup':
netlib_setifstatus.c:83:27: \
error: 'NETLIB_SOCK_TYPE' undeclared (first use in this function);
Change-Id: I9496528ae09f6244d73ab77dbcd307ced95f46fd
Bug: https://github.com/apache/incubator-nuttx/pull/1988
Forwarded: https://github.com/apache/incubator-nuttx/pull/2066
Signed-off-by: Philippe Coval <rzr@users.sf.net>
This will suppress a warning.
For some reason if declared weak,
it's not properly resolved at runtime,
this will need more investigations, any clues ?
Change-Id: I3cc3b7d145c7b32db20f5f2cbf4314ac20dba588
Forwarded: https://github.com/apache/incubator-nuttx/pulls/rzr
Signed-off-by: Philippe Coval <rzr@users.sf.net>
Also enable ifup, use USB console and create raw image.
Windows is not mandatory, I use linux as host
so defconfig can rely on defaults.
Change-Id: Ic0e5f79f28facb6eff1d9831bcb31e1ff329a5aa
Forwarded: https://github.com/apache/incubator-nuttx/pulls/rzr
Signed-off-by: Philippe Coval <rzr@users.sf.net>
Then it can easly deployed using USB masstorage
(same as on embed)
Forwarded: https://github.com/apache/incubator-nuttx/pulls/rzr
Change-Id: I62381d3f571fbc6e4716fa01d6974954f9193075
Signed-off-by: Philippe Coval <rzr@users.sf.net>
Summary:
- During streaming test, I noticed a dealock when controlling IRQ
- Actually, it will send an IPI when the cpu index is not 0
- However, up_cpu_pause() also sends IPI with critical section
- So the IRQ control must follow the same rule
Impact:
- Affects SMP only
Testing:
- Tested with spresense:wifi_smp
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Both CPUs are soft-reset with a call to board_reset. This is actually a
Core Reset, so both cores and all registers are reset. The only
exception is RTC.
Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
Summary:
- I noticed that stack for the init task was filled almost 100%
- This commit fixes this issue
Impact:
- Affects spresense:wifi configuration only
Testing:
- Tested with renew/mount/wget/nxplayer commands
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
## Summary of Changes
Add a host timer that generates periodic signals and sends SIGALRM to
the process that runs the NuttX simulation. This logic is integrated as
part of the existing NuttX oneshot timer. The host timer installs an
irq handler which is expected to run every CONFIG_USEC_PER_TICK .
Signed-off-by: Sebastian Ene <nuttx@fitbit.com>
Summary:
- This commit adjusts parameters for audio streaming
Impact:
- Only affects spresense:wifi and spresense:wifi_smp
Testing:
- Tested with nxplayer
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
1. There is difference about symbol replace on nuttx-names.in
between MacOS & Linux
2. For MacOS, if open '-fvisibility=hidden' and adjust nuttx-names.in,
it will meet symbol link-back-to-nuttx error.
3. Make the MacOS replace behaviour, same with Linux
Note:
MacOS should install objcopy with command:
$ brew install binutils
$ export PATH=$PATH:/usr/local/opt/binutils/bin
already check in to cibuild.sh
Change-Id: If78b784cc0ecb98cdbf7091de38acef00a8a02f3
Signed-off-by: ligd <liguiding1@xiaomi.com>
and remove the special handling in the stack dump
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Ia1ef9a427bd4c7f6cee9838d0445f29cfaca3998
In some cases, when NuttX configuration changes and this makes the
object list used to build one of the .a libraries change as well,
since the command used to build it is "ar crs" and this simply appends
the list of object files, the library could still include object
files from prior builds. This commit modifies the ARCHIVE macro to
erase the .a file if it already exists.
Since in some cases this behavior was actually expected (object
files from a subdirectory were appended to a library created one
level above) I added a ARCHIVE_ADD which works as ARCHIVE did.
This change should greatly improve behavior of building after
configuration changes.
1.Reduce the default size of task_group_s(~512B each task)
2.Scale better between simple and complex application
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Ia872137504fddcf64d89c48d6f0593d76d582710
This can hidden all nuttx's symbols,
and DO NOT export nuttx's symbols to share libraries.
e.g.
wrong:
NUTTX PC NUTTX
ffmpeg -> asound -> sysconf
right:
NUTTX PC PC
ffmpeg -> asound -> sysconf
Signed-off-by: ligd <liguiding1@xiaomi.com>
boards/arm/stm32/b-g474e-dpow1/scripts/ld.script.dfu:
* New file. Reserve up to 24KiB at the start of FLASH for a
bootloader.
boards/arm/stm32/b-g474e-dpow1/scripts/Make.defs:
* When building with CONFIG_STM32_DFU, use the new ld.script.dfu
to leave room in FLASH for a bootloader.
boards/arm/stm32/b-g474e-dpow1/README.txt:
* "FLASH Bootloader Support": New section.
Summary:
- This commit adds a wifi_smp configuration for spresense
- To avoid deadlock, gs2200m_irq_enable() and gs2200m_irq_disable()
release global spinlock if they are not called on CPU0.
Impact:
- Only affects spresense:wifi_smp
Testing:
- Tested with ostest, renew, smp, telnetd.
- NOTE: tcpserver might be unstable
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Add sotest/sotest32 example configs since they are removed
from module/module32 defconfigs in former commit.
Change-Id: I22808d06ef9e4231979a4467672c86386d216b83
Signed-off-by: liuhaitao <liuhaitao@xiaomi.com>
Sotest and module examples could not co-exist since there is only
one romfs_img array exist meantime. Or one of them would fail to
run.
Change-Id: I73e3b6791678f4e801181552c56af50e6d0e47c3
Signed-off-by: liuhaitao <liuhaitao@xiaomi.com>
The pic32mx_usbpullup() and pic32mz_usbpullup() callbacks were referred to with various incorrect names in comments: pic32mx_pullup(), pic32mz_pullup(), stm32_pullup().
Summary:
- This commit refactors stm32_gs2200m.c
- gs2200m_irq_enable() and gs2200m_irq_disable() are now symmetric.
Impact:
- All use cases which use the gs2200m driver.
Testing:
- Tested with stm32f4discovery:wifi
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Summary:
- This commit refactors cxd56_gs2200m.c
- gs2200m_irq_enable() and gs2200m_irq_disable() are now symmetric.
Impact:
- All use cases which use the gs2200m driver.
Testing:
- Tested with spresense:wifi
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
and enable on sim as a demo. Here is the paper:
ftp://gcc.gnu.org/pub/gcc/summit/2003/Stackguard.pdf
Signed-off-by: qiaowei <qiaowei@xiaomi.com>
Change-Id: I1926936328dad54eee16b322dcaad0b42f9c4a62
sam_ili9488.c:1605:7: error: label 'errout_with_dmadog' used but not defined
bcmf_sdio.c:855:16: error: passing argument 1 of 'wd_start' from incompatible pointer type
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
to save the preserved space(1KB) and also avoid the heap overhead
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: I694073f68e1bd63960cedeea1ddec441437be025
In the commit
612acb9d93
the CONFIG_PREALLOC_MQ_MSGS is set by default to 4 and should be removed from defconfig.
Signed-off-by: Alin Jerpelea <alin.jerpelea@sony.com>
Change the preallocated message and descriptor from 32/24 to 4.
The total size is reduce from 1892 to 532
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: I79d199465daef678986868f773876289859f42fc
boards: metro-m4 Add support for starting the watchdog timer on the metro-m4.
Testing:
- Build check only
Signed-off-by: Leomar Mateus Radke <leomar@falker.com.br>
Summary
- This commit should fix CI errors
Impact
- No impact
Testing
- Build check only
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
since the maximum number of argument pass to wd_start in the whole
code base is 2 and change CONFIG_MAX_WDOGPARMS in some defconfig
from 1 to 2 oherwise pthread_condclockwait will fail
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Ib6cb28b8c0722058849e7be916e164513431d21c
malloc() and free() should never be used within the OS. This will work in the FLAT build because there is only a single heap, but will cause problems in PROTECTED and KERNEL build modes where there are separate heaps for user and kernel memory.
Typically kmm_malloc(), kmm_zalloc(), and kmm_free() should be called within the kernel in those build modes to use the kernel heap.
Memory is never free. Possible memory leak:
./boards/arm/cxd56xx/common/src/cxd56_crashdump.c: pdump = malloc(sizeof(fullcontext_t));
Memory allocated with malloc(), but freed with kmm_free():
./drivers/usbhost/usbhost_composite.c: cfgbuffer = (FAR uint8_t *)malloc(CUSTOM_CONFIG_BUFSIZE);
Memory is never freed in these cases. It is allocated in the driver initialization logic, but there is no corresponding uninitialization logic; memory is not freed on error conditions:
./arch/arm/src/lc823450/lc823450_i2s.c: priv = (struct lc823450_i2s_s *)zalloc(sizeof(struct lc823450_i2s_s));
./arch/arm/src/sam34/sam_spi.c: spics = (struct sam_spics_s *)zalloc(sizeof(struct sam_spics_s));
./arch/arm/src/sama5/sam_spi.c: spics = (struct sam_spics_s *)zalloc(sizeof(struct sam_spics_s));
./arch/arm/src/samv7/sam_spi.c: spics = (struct sam_spics_s *)zalloc(sizeof(struct sam_spics_s));
Memory is allocated with zalloc() but freed on error conditions with kmm_free():
./arch/arm/src/sama5/sam_ssc.c: priv = (struct sam_ssc_s *)zalloc(sizeof(struct sam_ssc_s));
./arch/arm/src/samv7/sam_ssc.c: priv = (struct sam_ssc_s *)zalloc(sizeof(struct sam_ssc_s));
./arch/arm/src/stm32/stm32_i2s.c: priv = (struct stm32_i2s_s *)zalloc(sizeof(struct stm32_i2s_s));
Memory is never freed:
./drivers/spi/spi_bitbang.c: priv = (FAR struct spi_bitbang_s *)zalloc(sizeof(struct spi_bitbang_s));
ESP32 runs code in a SPI Flash, so users can also use it to store
data directly or mount some parts into a filesystem.
The SPI Flash usually use SPI0.
This driver was implemented by Dong Heng dongheng@espressif.com
and modified to fix coding style by Alan Carvalho de Assis.
Provide functions which get and set the 3A (Auto white balance/Auto
exposure/Auto
focus) parameters.
These are achieved by the ioctl request codes VIDIOC_G_EXT_CTRLS and
VIDIOC_S_EXT_CTRLS which have the following settings:
1. ctrl_class = V4L2_CTRL_CLASS_CAMERA
2. id = V4L2_CID_3A_PARAMETER
Also, provide function which get the 3A adjustment status.
This is achieved by the ioctl request codes VIDIOC_G_EXT_CTRLS which has the
following settings:
1. ctrl_class = V4L2_CTRL_CLASS_CAMERA
2. id = V4L2_CID_3A_STATUS
Fix the following three bugs about ON/OFF control of auto white balance and auto
exposure:
- The returned value of VIDIOC_G_CTRL with id = V4L2_CID_AUTO_WHITE_BALANCE is
different from V4L2 specification.
- Auto exposure is locked if auto white balance lock is requested.
- Auto exposure and auto white balance are locked if auto exposure lock is
requested.
Add imageproc_alpha_blend() to blend two images in a specified ratio.
The blended source image is selectable from the following two types:
- Single color image
- YUV422 image
The blend ratio is selectable from the following three types:
- Blend ratio with single value
- Blend ratio with 1bpp
- Blend ratio with 8bpp
Summary:
- Fix style violations in lc823450_syscontrol.h
Impact:
- This commit affects register naming in syscontrol.
- So some files such as lc823450_start.c lc823450_mux.c were also changed.
Testing:
- Build check only
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Make.dep file should be updated by .config changed after first make.
There are 2 cases affected for this problem:
1) Add source files by config symbol
2) Include header files in #ifdef directive
These 2 cases may not be included in Make.dep and this may prevent the
differential build from working correctly.
- Add new functions of GNSS
- Support the lower PWM frequency
- Add CONFIG_CPUFREQ_RELEASE_LOCK
- Add high speed ADC support
- Add HPADC input gain configuration
- Add eMMC device
- Frame buffer support
- Fix SD/GNSS/sensor drivers not worked
- Build errors
- Fix nxstyle issues
- Add board specific logic to altair modem dirver
- Fix issue that SPI4 RX frequency violated AC Spec
- Remove dummy buffer in altair driver
- Fix priority of SPI transfer task is too low
- Remove unnecessary configuration
- Modify timeout value for RX ready
- Fix minor bugs
Kinetis: Reworked USB driver for setup out data phase.
Freedom K28: New config nshsdusb, with RNDIS support
PL2303: Try to avoid clone detection.
General: various nxstyle fixes
General: license changed
.d directories are common for holding config and init scripts.
They are being ignored with the global *.d pattern which is meant to
ignore Make dependency files.
1.It make sense to let Toolchain.defs give the default value
2.The board can still change if the default isn't suitable
3.Avoid the same definition spread more than 200 Make.defs
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Ic2649f1c7689bcf59c105ca8db61cad45b6e0e64
We already have a debug assert and a return error this warning
will fail builds that are not even using this ioctl.
Signed-off-by: Brennan Ashton <bashton@brennanashton.com>
- makes nsh reboot command work
Squashed commit of the following:
commit 5193f6ae9623bfb4d3bed4ecf3d0fb9ae1bfb6e8
Author: Adam Feuer <adam@starcat.io>
Date: Thu Jul 16 16:41:54 2020 -0700
removed conflict tag that was missed
- result of an incorrectly fixed bad merge
commit 731108ea7495655e96e516448887ca8c9ab354d1
Author: Adam Feuer <adam@starcat.io>
Date: Thu Jul 16 16:12:30 2020 -0700
implement system reset to make nsh reboot work
Squashed commit of the following:
commit 245d155cc58d31af412f2b832877736b2088b896
Author: Adam Feuer <adam@starcat.io>
Date: Thu Jul 16 16:10:10 2020 -0700
add Kconfig setting for system reset
commit e7d5def8151821bf359c55c05ba1f59421b2371a
Author: Adam Feuer <adam@starcat.io>
Date: Thu Jul 16 15:51:35 2020 -0700
implement system reset to make nsh reboot work
Make.defs under board folder can still overwrite the default as needed
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: I1c300a8ace4b54d475ef8d398661ed65ca273a2e
The QuickFeather board added as an initial target.
These featrues are minimally implemented:
* Clock Configuration -- All clocking registers are defined and
configuration is used to setup the HSO, M4 Core, and M4 Perif
clocks. Additionally some clock debugging is stubbed for
bringing out clock paths to IO pins.
* UART -- The lowputc as well as the serial driver is implemnted
for the single UART device. Currently the configuration is
hard coded, but uses the proper interfaces to later fill in.
* SysTick -- The system tick timer is implemented and clocking
properly. Tickless mode is not yet implemented.
* Interrupts -- The interrupt system is implemented and verified
using the UART and SysTick systems.
* GPIO -- GPIO and IOMUX systems are defined and implemented.
This is verified using the UART as well as the Arch LED
system. The GPIO interupt system is stubbed out but not
implemented.
* Arch LEDS -- The blue LED as part of the RGB LED is configured
and attached to the Arch LED system. This indicates the device
coming online as well as when a hardfault is triggered.
Applications and Testing:
* There is a nsh configuration implemented that includes debug
features as well as the ostest, getprime, and mem test.
All of these have been run and verified.
Signed-off-by: Brennan Ashton <bashton@brennanashton.com>
so all boards support uClibc++/libc++ automatically
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Ibe6fafbec57f7acb26ea6188a3e9923ea82295c5
otherwise the user is shocked that using cin/cout/cerr will crash the system
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: I453427261f3e2a6e60f7dd2398f7d3bd1043a0d5
Specify -mcmodel=large only on 64-bit case. (x86-64)
The code model is not available for 32-bit.
This fixes the following error:
MODULECC: chardev.c
cc1: error: code model 'large' not supported in the 32 bit mode
Makefile:79: recipe for target 'chardev.o' failed
since cxx initialization is controlled by CONFIG_HAVE_CXXINITIALIZE now
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: I39438dc3006d0a0b810052ecef50cd3c92f09d83
On Linux, we (ab)use the host compiler to compile binaries for NuttX.
Explicitly disable features which might be default on the host while
not available on NuttX.
On macOS, this is not necessary because we use a cross compiler,
which has safer defaults for our purpose, to build binaries to
run on NuttX. But this doesn't hurt either.
Namely, always use -mcmodel=large.
NuttX modules are loaded into the NuttX heap, which can be out of
range for 32-bit relocations generated with -mcmodel=small.
A possible alternative would be to use MAP_32BIT to allocate sim heap.
But I prefer this approach because it's very convenient for me to
be able to share modules between Linux and macOS sim.
