Rasclock (PCF2127 ) Hardware Clock FreeBSD 12.0

[Ian Lepore]

On Wed, 2019-07-17 at 18:45 +0300, Stefan Parvu wrote:
> fixed the svn. thanks for tips.
> 
> and applied the patched, make and make install worked fine:
> 
> # patch < /home/krmx/nxprtc.patch
> Hmm...  Looks like a unified diff to me...
> The text leading up to this was:
> --------------------------
> > --- sys/dev/iicbus/nxprtc.c	2019/05/23 18:24:27	348184
> > +++ sys/dev/iicbus/nxprtc.c	2019/07/15 21:47:40	350016
> 
> --------------------------
> Patching file sys/dev/iicbus/nxprtc.c using Plan A...
> Hunk #1 succeeded at 33.
> Hunk #2 succeeded at 42.
> Hunk #3 succeeded at 243.
> Hunk #4 succeeded at 312.
> Hunk #5 succeeded at 478.
> Hunk #6 succeeded at 534.
> Hunk #7 succeeded at 805.
> done
> 
> # cd /usr/src/sys/modules/i2c/nxprtc
> # make
> machine -> /usr/src/sys/arm64/include
> awk -f /usr/src/sys/tools/makeobjops.awk /usr/src/sys/kern/bus_if.m -h
> awk -f /usr/src/sys/tools/makeobjops.awk /usr/src/sys/kern/clock_if.m -h
> awk -f /usr/src/sys/tools/makeobjops.awk /usr/src/sys/kern/device_if.m -h
> awk -f /usr/src/sys/tools/makeobjops.awk /usr/src/sys/dev/iicbus/iicbus_if.m -h
> awk -f /usr/src/sys/tools/makeobjops.awk /usr/src/sys/dev/ofw/ofw_bus_if.m -h
> :> opt_platform.h
> cc  -O2 -pipe  -fno-strict-aliasing -Werror -D_KERNEL -DKLD_MODULE -nostdinc   -I. -I/usr/src/sys -I/usr/src/sys/contrib/ck/include -fno-common  -fPIC   -MD  -MF.depend.nxprtc.o -MTnxprtc.o -mgeneral-regs-only -ffixed-x18 -ffreestanding -fwrapv -fstack-protector -Wall -Wredundant-decls -Wnested-externs -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wcast-qual -Wundef -Wno-pointer-sign -D__printf__=__freebsd_kprintf__ -Wmissing-include-dirs -fdiagnostics-show-option -Wno-unknown-pragmas -Wno-error-tautological-compare -Wno-error-empty-body -Wno-error-parentheses-equality -Wno-error-unused-function -Wno-error-pointer-sign -Wno-error-shift-negative-value -Wno-address-of-packed-member    -std=iso9899:1999 -c /usr/src/sys/dev/iicbus/nxprtc.c -o nxprtc.o
> ld -m aarch64elf -d -warn-common --build-id=sha1 -r -d -o nxprtc.kld nxprtc.o
> :> export_syms
> awk -f /usr/src/sys/conf/kmod_syms.awk nxprtc.kld  export_syms | xargs -J% objcopy % nxprtc.kld
> ld -m aarch64elf -Bshareable -znotext -d -warn-common --build-id=sha1  -o nxprtc.ko nxprtc.kld
> objcopy --strip-debug nxprtc.ko
> 
> # make install
> install -T release -o root -g wheel -m 555   nxprtc.ko /boot/modules/
> kldxref /boot/modules
> 
> 
> Bingo. Let me try now the new ko module file on another system. Will report back the findings.
> thanks all for help.

What worries me about that output is this line:

> :> opt_platform.h

that looks like it created a completely empty opt_platform.h, but that
file must contain #define FDT 1 for the nxprtc module to build with fdt
support.

That's why I have been lifelong-skeptical about the whole concept of
trying to build a single module separate from the kernel.  Maybe it
mostly works for amd64, but I've never thought it was a safe thing to
do on arm systems.

You might be able to work around it by hand-editing that file and then
rebuilding the module.  The file will be in

 /usr/obj/arm64.aarch64/usr/src/sys/GENERIC/modules/i2c/nxprtc/opt_platform.h

-- Ian