<
roh>
hm 12 cases done. 8 todo ;)
<
lekernel>
on the flash reset? and did it help?
<
adamw_>
this is to let PROGRAM_B suffered from un-known voltage level while the tiny duration comes from ON to OFF. :-)
<
adamw_>
i tested 50 times without un-boot. :-)
<
lekernel>
mh, ok, sounds good... maybe try a little more, the problem is really rare sometimes
<
adamw_>
but the flash reset, I am thinking also connect reset ic's output connected directly to RP# of flash to let them work in parallel!
<
lekernel>
can't we just use a capacitor instead of this reset ic?
<
adamw_>
NO...cannot.
<
adamw_>
i measured many scope, I can show you later to give a very clear explaination. hehe. :)
<
adamw_>
because we also can not realize how ender user 'fast' cycling.
<
lekernel>
yeah, if you use the open drain version of the reset ic, it should work in parallel
<
adamw_>
means capacitor (== constant)Â Â this won't meet end user's unexpectedly action (variable.) :-)
<
roh>
this reset ic shouldnt be expensive.. how much is it?
<
adamw_>
that's why I asked the question on list to know if s/w or some where will pull down RP#.
<
lekernel>
yes, the fpga needs to do that
<
adamw_>
so yes, I need to solder manually the output to RP# tomorrow. :-)
<
lekernel>
so the main problem was on PROGRAM_B?
<
adamw_>
right, exactly!
<
lekernel>
mh, I don't get it
<
adamw_>
its pin3 VDD connected to 3V3, pin2 -> GND and pin1 Vout -> PROGRAM_B;
<
adamw_>
well...no worries.
<
lekernel>
so you're delaying the FPGA configuration?
<
adamw_>
yeah...just delay 20us to 're-start' fpga's configuration.
<
adamw_>
roh, US 0.092/each, ;-)
<
adamw_>
I'll upload more waveforms to explain.
<
adamw_>
you can check ug380.pdf page 39 to know more first. :-)
<
antgreen>
lekernel: I just hit the approval button and it should get processed
<
antgreen>
( I think it's a manual step by somebody )
<
mwalle>
larsc: were modules supported with theobromas original linux port? (is CONFIG_MODULES defined?)
<
larsc>
mwalle: no idea, i've never tried it
<
mwalle>
larsc: do you have a running image? could you try the command 'mount -t a a /dir'? /dir has to be existing
<
larsc>
'no such device'
<
lekernel>
antgreen: thanks for the gcc write access
<
kristianpaul>
is wondering if lekernel became now a gcc matainer (at least for rtems?..)
<
kristianpaul>
mantainer*
<
lekernel>
i'll maintain the lm32 port yes
<
lekernel>
someone has to do it, right? :)
<
kristianpaul>
thanks nice, as my bet now is for rtems,no mather if we get linux later,i just realized real time os are ideal for SDR
<
mwalle>
lekernel: cool so i can send you some patches soon :)
<
lekernel>
yup, please do
<
Fallenou>
wooow lekernel you have commit access :)
<
Fallenou>
very nice !
<
Fallenou>
now it's you that will be insulted on the bug tracker :D
<
mwalle>
hehe, yeah there are some ICEs :)