Currently the btrfs module (and other fs modules) is loaded before other modules to provide early extroot compatibility. This will fail for btrfs however, because the dependencies crypto-core and crypto-hash are missing. This patch makes them available for early loading too.
Signed-off-by: Jan Willies <jan@willies.info>
SVN-Revision: 30542
Add a new wireless config parameter short_preamble=0|1 to enable
usage of short preambles. Default is to only allow long preambles
as before.
Even if short_preamble is set to 1 hostapd will take care that
short preambles are disabled as soon as a STA associates that
cannot handle short preambles.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
SVN-Revision: 30539
This patch adds support for a kernel module package for SLIP. We needed it forusing tunslip6 from Contiki OS for 6LoWPAN communications.
Signed-off-by: Markus Becker <mab@comnets.uni-bremen.de>
SVN-Revision: 30538
Update e2fsprogs to the last upstream version 1.42.
blkid works fine without 100_add_missing_libpthread_for_blkid.patch.
Signed-off-by: Luka Perkov <openwrt@lukaperkov.net>
SVN-Revision: 30535
We want the `_start's symbols to be at the very beginning of the `stage2'
binary's `.text' section. However, nothing enforces this requirement. With
recent binutils (ie. post 2.19.1), this results in a corrupt binary, as ld(1)
puts another symbol (eg. `journal_init') at the beginning of the section.
Fix this by providing a stripped down version of the default linker script
where we enforces the position of `_start' in the `.text' section.
Patch by Arnaud Lacombe (#10444)
SVN-Revision: 30484
When using the Broadcom SDK the SSB bus is emulated as an PCI bus so
the PCI bus number of the first real pci bus is increased by one. The
variable names in the nvram are created with that structure in mind. To
fix this we have ti increases the pci bus number by one. This was also
done for ssb some time ago.
This is based on a patch by nlh.
This closes#10917
SVN-Revision: 30422
Hi!
On Mon, Nov 14, 2011 at 03:01:34PM +0100, Daniel Golle wrote:
> The driver works quite nice and stable for me using a RaLink Rt5370 USB device.
>
> Signed-off-by: Daniel Golle <dgolle@allnet.de>
>
> Index: package/mac80211/Makefile
> ===================================================================
> --- package/mac80211/Makefile (revision 29114)
> +++ package/mac80211/Makefile (working copy)
> @@ -1147,6 +1147,7 @@
> CONFIG_RT2800_LIB=$(if $(CONFIG_PACKAGE_kmod-rt2800-lib),m) \
> CONFIG_RT2800PCI=$(if $(CONFIG_PACKAGE_kmod-rt2800-pci),m) \
> CONFIG_RT2800USB=$(if $(CONFIG_PACKAGE_kmod-rt2800-usb),m) \
> + CONFIG_RT2800USB_RT53XX=$(if $(CONFIG_PACKAGE_kmod-rt2800-usb),y) \
> CONFIG_RTL8180=$(if $(CONFIG_PACKAGE_kmod-rtl8180),m) \
> CONFIG_RTL8187=$(if $(CONFIG_PACKAGE_kmod-rtl8187),m) \
> CONFIG_RTL8192CE= \
Aparently this was wrong, CONFIG_RT2800USB_RT53XX is supposed to live in
BUILDARGS rather than MAKE_OPTS to actually get support for the Rt5370.
Thanks to actmnophn for the hint!
This reverts changeset 29116 and adds it to the right section in the Makefile.
Signed-off-by: Daniel Golle <dgolle@allnet.de>
SVN-Revision: 29906
