This series of commits changes completely the installation and runtime,
moving to UBI.  I used it several times, and there are still minor problems
related to flash access, but it is generally stable.  I plan to find
the flash bug and fix it in a separate commit. This is ready for master,
in my opinion.

Note: a number of intermediate commits won't work, but I chose to split
the changes according to the package (barebox, kernel, userspace, ...)

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The information is not up to date with the new installation.  We are
not using JTAG ourselves these times, so we'd better remove this
chapter, stating it is still in the history of the package, if it is
needed by someone.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The release tar file (wrs-firmware.tar) is not compressed any more, so
fix the name all around, and the extracting command too.

Similarly, the official release is a .tar file, and it is generated by
the "wraprootfs" step, so this removes "--pack" as a build option (it
was a duplicate of wrapping it up anyways).

Release files are already the ones used for installation, so flash-wrs
doesn't need to build the cpio file any more.

The "macigstring" for installation is concatenated to the cpio file by
tellling the usb-loader to copy it to RAM, so no temporary cpio archive
is created any more: release files are always used unchanged.

Wrapping rootfs creates the release tar dire
"--pack" i
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This brings everything into /etc/init.d/wrs-boot-procedure that is
executed as a "sysinit" task, before anything else.  The procedure
runs from the initrd/initramfs and is concerned with the complete
update of the device, if an update is pending.

See documentation (next commit) for details about partitions and boot
procedure.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This also change the flash_wrs script, to match new conventions
(i.e. we use the run-time initramfs for installation, and so on)
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

pass WRS_VERBOSE=y on the kernel command line to make it verbose.
No damaga done, but useful in checking the boot sequence.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This moves wr to usr/wr and copies etc to usr/etc.
Also, it creates the .tar.gz for /usr and an initramfs
that excludes usr.

The current installation still works (it ignores new files,
and can live with wr pointing to usr/wr).
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This changes partitions, mainly adding a kernel-visible mapping of
what we ha in dataflash.  NAND changes are compatible, as only empty
space is used differently.

This also includes a new barebox.bin, that you can just write
to /dev/mtd3 after booting the new kernel.

Documentation is updated to reflect these partitions, but there also
old leftovers that are fixed in this same commit.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

3 are not enough: we need 4. So support up to 16 even though we waste
a few dozen bytes of precious RAM.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

we'd better err out than corrupt memory, if we have more arguments than
expected
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

dirname() and basename() change their argument.  We'd better
strdup(argv[0]) to be able to use it in error messages.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

In our case the name of the switch is its IP address of eth0 interface. Without
this option enabled, only the first octet is reported to remote log server (e.g.
192)

They disappeared quite some time ago: only .tar.gz remains.  Fix
documentation accordingly.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This merges the last commits I made on my laptop last weekq, and I
forgot to push.  Thanks greg for testing wrsw_vlans and thus finding
I miseed some bits.

Conflicts:
	userspace/tools/wrsw_vlans.c

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Some WRS speciments won't work correctly with automatic speed setup.
This patch forces full-speed on the device (instead of the
autodetected high-speed), and thus they work.

Speed is not a problem anyways, because it is just a serial port.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

wr_nic now exports wr_nic_pstats_callback as a pointer. When wr_pstats
is loaded it writes its own pointer in there.

wr_nic can thus ask for statistics and updated net_device_stats with
the proper numbers.   The statistics on snmp reflect this, because
snmpd uses the standard device ioctl commands -- but very few of our
counters actually end up all the way to snmp.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Without this, the switch may be unable to boot NFS-Root over a new system
(e.g. my laptop, currently linux-3.14 with full NFS support.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

open returns -1 on error, not 0

mmap returns -1 on error, not NULL
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>