Recent changes to bugs

Lib-Compile fails under Fedora 13

2010-10-27T18:16:36Z

I downloaded the source files for nut and btnode. The building steps work till I try to make the library-files under lib/ by executing "make clean install" with an error in one of the source files:

avr-gcc -c -mmcu=atmega128 -Os -fno-delete-null-pointer-checks -Wall -Werror -Wstrict-prototypes -Wa,-ahlms=lanc111.lst -D__HARVARD_ARCH__ -DETHERNUT1 -I../include avr/dev/lanc111.c -o avr/dev/lanc111.o
cc1: warnings being treated as errors
avr/dev/lanc111.c: In function 'NicRxLanc':
avr/dev/lanc111.c:1183:9: error: dereferencing type-punned pointer will break strict-aliasing rules
avr/dev/lanc111.c:1183:9: error: dereferencing type-punned pointer will break strict-aliasing rules
make[1]: *** [avr/dev/lanc111.o] Fehler 1
make[1]: Leaving directory `/home/jakob/Desktop/nut/arch'
make: *** [install] Fehler 2

BTNode Strucked -when receives pkts- completing operation

2008-06-24T08:01:16Z

The Mobile sends a request for rssi
Invokes the l2cap Method ready()(This method Determines if there is a packet that can be read via a call to receive() If true, a call to receive() will not block the application.)If there is no packet[The btnode takes sometime to complete the bt_hci_read_rssi ] the mobile sends the request for rssi packet again. Here comes the problem.Before completing bt_hci_read_rssi if it
receives series of packets it gets strucked

This is not happening if the Mobile sends a request for rssi.Invokes call to receive()--This is a blocking operation[Method ready was not used here].It waits till it receive the packet.so never sends the request for
rssi again until it receives the packet.

jaws - tmote tg monitor sets RTS when opening

2008-02-22T12:43:50Z

when the tmote target monitor is used, each time the target monitor is opened, the RTS line is set high for about 10-22ms. this sometimes causes the tmote target to hang. it should be possible to open the target monitor without touching the RTS line unnecessarily.

tg set power does sometimes set false mode (bat)

2008-01-16T19:23:16Z

sometimes, a tg set power on command sets the power to "bat" instead of "on", even when the target has no batteries at all.

Cruisecontrol - TOS Metrics missing for Harvester builds

2007-11-27T17:57:13Z

The TOS2 Metrics page of the Harvester build does not show any metrics but only the menubar on the left.

bootloader - not handle too long wait time to release reset

2007-11-06T13:46:22Z

I get the following error if waiting too long to release the reset button:

avrdude: stk500_getsync(): not in sync: resp=0x15

permasense cc tracker brocken

2007-10-17T08:20:57Z

the memory metrics are broken:

http://tik42x.ee.ethz.ch:8080/cruisecontrol/buildresults/permasense?tab=tos-metrics

and the power tracker needs to be added!

testbeds - cc tinyosmetrics broken

2007-10-05T20:38:52Z

some of the t2 apps do not show any memory information because builds failed and didnt write the logs properly. this is also the case for permasense.

we need to find a good fix for this and recover the plots...

testbeds - power tracker improvements

2007-09-03T16:28:21Z

- install the 2nd power supply

- add a low alarm to the target voltages

- add more temperature probes

- design a switching scheme for the different targets coming from the cruisecontrol builds

server - getrpcs webpage has too small fields

2007-09-03T16:22:16Z

the msg file in the get RPCS mask on the webpage is too smal to show the full data....