Reading errno within pthreads always resulted in 0 breaking various programs
like transmission and collectd. This patch is not 100% thread safe but it brings
back threading into a usable state for now.
git-svn-id: svn://svn.openwrt.org/openwrt/trunk@23905 3c298f89-4303-0410-b956-a3cf2f4a3e73
For kernel 2.6.36 some additional pcmcia backpots would be needed, but
this should not depend on pcmcia.
git-svn-id: svn://svn.openwrt.org/openwrt/trunk@23899 3c298f89-4303-0410-b956-a3cf2f4a3e73
Since b65c7b2c79debcb9017e31913e01eeaa280106fb, the implicit search path
can be disabled by not #defining __LDSO_SEARCH_INTERP_PATH__. This
causes _dl_ldsopath to never be set, so it remains NULL. _dl_ldsopath is
still used when __LDSO_LDD_SUPPORT__ is #defined, to strip the path off
of the beginning of the absolute path to the ld.so interpreter in use
for printing. The _dl_strlen will crash with a NULL argument.
Rather than relying on _dl_ldsopath, this change causes ldd to compute
the interpreter's basename directly.
glibc ld.so seems to print the full path to the interpreter without
any computed basename or =>. I personally prefer glibc's behavior, but
to preserve backwards compatibility with uClibc ld.so, the existing
format with the computed basename, =>, and full path is used here. This
enables simpler (and unchanged) text processing in a pipeline.
Signed-off-by: Mark Mentovai <mark at moxienet.com>
git-svn-id: svn://svn.openwrt.org/openwrt/trunk@23892 3c298f89-4303-0410-b956-a3cf2f4a3e73
I discovered another bug in the ath9k WNDR3700 quirk fix from r23822.
ath9k_hw_apply_gpio_override was testing the original value of gpio_mask
to determine which bits were appropriate, instead of testing the shifted
value. Because bit 0 is always clear in the original value, this
resulted in no calls to set the GPIOs, an improperly-configured radio,
and reduced RSSI on other systems listening to the radio.
Signed-off-by: Mark Mentovai <mark@moxienet.com>
git-svn-id: svn://svn.openwrt.org/openwrt/trunk@23890 3c298f89-4303-0410-b956-a3cf2f4a3e73
For WNDR3700, use green for 1Gb/s and amber for 100Mb/s and 10Mb/s on
the four LAN port switch LEDs. The LEDs will be lit to indicate a link
and will blink to indicate activity. This matches the stock firmware
and the description printed on the label on the unit's underside.
Signed-off-by: Mark Mentovai <mark@moxienet.com>
git-svn-id: svn://svn.openwrt.org/openwrt/trunk@23879 3c298f89-4303-0410-b956-a3cf2f4a3e73
Make the WiFi LEDs blink for activity on the WNDR3700, matching the
stock firmware and user expectations. The green 2.4GHz and blue 5GHz
LEDs will illuminate with the radio on, and will blink to indicate
transmission and reception.
Signed-off-by: Mark Mentovai <mark@moxienet.com>
git-svn-id: svn://svn.openwrt.org/openwrt/trunk@23877 3c298f89-4303-0410-b956-a3cf2f4a3e73
The WNDR3700's art/caldata partition contains three MAC addresses at
offsets 0, 6, and 12, used for eth0/wlan0 (shared), eth1, and wlan1,
respectively. The existing code configures eth1's MAC address by adding
1 to eth0's instead of using the value stored in the EEPROM. This patch
causes the stored value to be used directly. (On my WNDR3700s, the three
stored addresses are all sequential anyway.)
Signed-off-by: Mark Mentovai <mark@moxienet.com>
git-svn-id: svn://svn.openwrt.org/openwrt/trunk@23876 3c298f89-4303-0410-b956-a3cf2f4a3e73
* adds a rewrite of the tapi drivers + sip app. this is the result of lars' gsoc 2010 project, Thanks !
git-svn-id: svn://svn.openwrt.org/openwrt/trunk@23840 3c298f89-4303-0410-b956-a3cf2f4a3e73
