If there is no sprom on an ssb based pci device on the brcm47xx
architecture ssb now asks the architecture code to look into the nvram
to get some sprom data for this device. Now we are able to read out
pci/1/1/ foo or pci/1/3/ foo config options.
This will fix some problems where the wireless devices does not got an
mac address and the following message was show:
ssb: WARNING: Invalid SPROM CRC (corrupt SPROM)
SVN-Revision: 26801
Dual image capable CFEs store an image sequence at the same place as
currently OpenWrt stores the actual rootfs length, so it will get
overwritten when flashing through such a CFE.
To prevent this from happening, move the rootfs length field to the next
four bytes, thus completely using the reserved1 field.
Since the reserved1 field is now completely in use, it does not make sense
to allow it to be set from the imagetag utility, so remove the option.
Signed-off-by: Jonas Gorski <jonas.gorski+openwrt@gmail.com>
SVN-Revision: 26680
I wrote a patch to add 96348SV board (aka MAGIC) support to trunk.
In my case I've found it in a Telsey "Alice W-Gate" router, fairly common
(in the past) in Italy.
I used the board definition (GPIOs, leds, ...) from the Broadcom GPL-2
relased code (here is a mirror: http://beghiero.myftp.org/sorgenti/boardparms.c )
Thanks to KanjiMonster for the support on the IRC channel.
Signed-off-by: Laurento Frittella <laurento.frittella@gmail.com>
SVN-Revision: 25584
The fallback needs to be exported for bcm63xx, too. Runtime tested with a
BCM4322 on a BCM6358.
Signed-off-by: Jonas Gorski <jonas.gorski+openwrt@gmail.com>
SVN-Revision: 25420
Since bcm636x platform embeds two spi master device,
the attached patch removes static bcm_rset usage, replaced by
"bs->regs" field for all I/O operation.
Signed-off-by: Miguel Gaio <miguel.gaio@efixo.com>
SVN-Revision: 24840
Add support for the T-Com Speedport W 500V using a special tagged image as it
identifies itself as a 96348GW, but has some gpios wired differently, and also
has only one ethernet port without a switch attached to it.
Signed-off-by: Jonas Gorski <jonas.gorski+openwrt@gmail.com>
SVN-Revision: 24190
Add a fixup for the kernel to check the bcm tag of the image for a boardid in
the information1 field, and use that instead of the supplied one by cfe. This
is triggered by a + at the start of the information1 field.
Also modify the image generation to generate apropriately tagged images.
Signed-off-by: Jonas Gorski <jonas.gorski+openwrt@gmail.com>
SVN-Revision: 24189
The DSL-2740B/DSL-2741B rev C2 support was missing the board definition,
resulting in a kernel panic on boot.
This patch adds it, and also renames the generated image name to properly
reflect that it is for the revision C2, and for both the DSL-2740B and
DSL-2741B.
Signed-off-by: Jonas Gorski <jonas.gorski+openwrt@gmail.com>
SVN-Revision: 24188
Since some devices have more than one power led, they might be separated by
their color in the name. To also allow them to be on after leaving preinit,
set any status led starting with 'power' to on instead of only those with
the exact name of 'power'.
Signed-off-by: Jonas Gorski <jonas.gorski+openwrt@gmail.com>
SVN-Revision: 24187
Currently eth1 is hardcoded as the expected interface switches in diagmode.
This patch replaces the eth1 with the ifname, and also directly tests if
the interface specified by ifname has a switch attached.
Signed-off-by: Jonas Gorski <jonas.gorski+openwrt@gmail.com>
SVN-Revision: 24186
Port the mechanism for different default configurations from ar71xx and
use the old configuration as the default configuration. This should have
no functional differences to before.
While we are at it, remove the obsolete nat option.
Signed-off-by: Jonas Gorski <jonas.gorski+openwrt@gmail.com>
SVN-Revision: 24185
