On Sat, Jan 18, 2025 at 04:33:52PM +0000, Jonathan Cameron wrote:
> I briefly applied this, but...
>
> [jic23@jic23-huawei iio]$ make LOCALVERSION= W=1 -j12 C=1
> mkdir -p /home/jic23/src/kernel/iio/tools/objtool && make O=/home/jic23/src/kernel/iio subdir=tools/objtool --no-print-directory -C objtool
> CALL scripts/checksyscalls.sh
> INSTALL libsubcmd_headers
> CC [M] drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.o
> CC [M] drivers/iio/common/cros_ec_sensors/cros_ec_sensors_trace.o
> In file included from drivers/iio/common/cros_ec_sensors/cros_ec_sensors_trace.h:56,
> from drivers/iio/common/cros_ec_sensors/cros_ec_sensors_trace.c:32:
> ./include/trace/define_trace.h:106:42: fatal error: ./cros_ec_sensors_trace.h: No such file or directory
> 106 | #include TRACE_INCLUDE(TRACE_INCLUDE_FILE)
> | ^
>
> Despite dealing with trace points a lot in the day job
> I still find them hard to actually add to a subsystem because
> of fun things like this one.
Weird, I did the same test without seeing the error:
$ make O=build_dir W=1 allmodconfig \
drivers/iio/common/cros_ec_sensors/cros-ec-sensors-core.o
...
CC [M] drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.o
CC [M] drivers/iio/common/cros_ec_sensors/cros_ec_sensors_trace.o
LD [M] drivers/iio/common/cros_ec_sensors/cros-ec-sensors-core.o
