== DF-4958 build (guest /root/df4958, DragonFly 6.5-DEVELOPMENT, cc 8.3) ==
[captured live; guest later reset with-src (required by the DF-4957 panic
 on the same boot), so guest-side object files are gone - rebuild via
 build.sh after re-transfer]

== building trig.ko (stock /usr/src/sys/dev/misc/led/led.c) ==
cc -O2 -pipe -D_KERNEL [kmod flags] -c trig_leak.c
cc -O2 -pipe -D_KERNEL [kmod flags] -c /usr/src/sys/dev/misc/led/led.c
cc -Wl,--build-id=sha1 -nostdlib -Wl,--hash-style=sysv -r -Wl,-d -o trig.ko trig_leak.o led.o
-rw-r--r-- 1 root 1000 14832 Sep 25 10:49 trig.ko

== building trig_fixed.ko (led.c + fix.diff) ==
Hmm...  Looks like a unified diff to me...
Patching file led_fixed.c using Plan A...
Hunk #1 succeeded at 201.
done
199:			sbuf_delete(*sb);
204:		sbuf_delete(*sb);     <-- the fix lands in led_parse's empty/error branch
252:		sbuf_delete(sb);
277:		sbuf_delete(sb);
338:		sbuf_delete(sc->spec);
cc -O2 -pipe -D_KERNEL [kmod flags] -c led_fixed.c
cc -Wl,--build-id=sha1 -nostdlib -Wl,--hash-style=sysv -r -Wl,-d -o trig_fixed.ko trig_leak.o led_fixed.o
-rw-r--r-- 1 root 1000 14824 Sep 25 10:50 trig_fixed.ko

== building wr ==
-rwxr-xr-x 1 root 1000 8328 Sep 25 10:50 wr
BUILD_OK