jonwil has quit [Remote host closed the connection]
Wizzup has quit [Quit: reboot]
Wizzup has joined #neo900
nox- has quit [Quit: Leaving]
arcean has quit [Quit: Konversation terminated!]
obsed has quit [Ping timeout: 245 seconds]
modem has quit [Ping timeout: 244 seconds]
<Oksana>
Hello! Is anybody here? It's quiet...
fling has joined #neo900
<Oksana>
Good morning
Oksana has quit [Read error: Connection reset by peer]
silviof has quit [Ping timeout: 272 seconds]
MonkeyofDoom has quit [Ping timeout: 265 seconds]
Svetlana has quit [Remote host closed the connection]
silviof has joined #neo900
Oksana_ has joined #neo900
<Oksana_>
Hello! Anybody knows how to remove the beautiful, large, shiny shield from the SIM card holder, so that restoring SIM contacts to working state would be more feasible? Somebody tried to put a memory card in place of SIM card ;/ The device in question is N950 (N900 is made a bit differently, so it would not have any damage to pins in case of such an attempt)
Oksana_ is now known as Oksana
<Oksana>
Good morning! ^
obsed has joined #neo900
obsed has quit [Ping timeout: 245 seconds]
sixwheeledbeast has quit [Ping timeout: 255 seconds]
sixwheeledbeast has joined #neo900
silviof has quit [Ping timeout: 260 seconds]
sixwheeledbeast has quit [Ping timeout: 258 seconds]
sixwheeledbeast has joined #neo900
sixwheeledbeast has quit [Ping timeout: 255 seconds]
sixwheeledbeast has joined #neo900
obsed has joined #neo900
obsed has quit [Ping timeout: 245 seconds]
Sveta has joined #neo900
obsed has joined #neo900
Sveta is now known as Svetlana
sixwheeledbeast has quit [Ping timeout: 265 seconds]
sixwheeledbeast has joined #neo900
kolp has joined #neo900
arcean has joined #neo900
<Pali>
what will be next??? security bug in 'echo'? or in 'test'? or in 'read'??
<bencoh>
:)
<bencoh>
you *never* use strings on untrusted files, right? that's obviously out of its scope, no doubt! ;p
mvaenskae has joined #neo900
sixwheeledbeast has quit [Ping timeout: 250 seconds]
sixwheeledbeast has joined #neo900
mvaenskae has quit [Quit: Lost terminal]
modem has joined #neo900
modem has quit [Changing host]
modem has joined #neo900
modem has quit [Ping timeout: 260 seconds]
mvaenskae has joined #neo900
modem has joined #neo900
che1 has joined #neo900
modem has quit [Changing host]
modem has joined #neo900
b1101 has quit [Quit: b1101]
Svetlana has quit [Remote host closed the connection]
jonwil has joined #neo900
sixwheeledbeast has quit [Ping timeout: 260 seconds]
modem has quit [Ping timeout: 244 seconds]
<jonwil>
hi
sixwheeledbeast has joined #neo900
paulk-aldrin has joined #neo900
<obsed>
bencoh: there was a security issue in strings ?
<wpwrak>
hehe, the SIM wars have begun: "When you choose AT&T [...] AT&T dedicates Apple SIM to their network only." http://support.apple.com/kb/HT6499
<obsed>
running strings is quite common on untrusted binaries
<bencoh>
obsed: yeah, hence my sarcasm :)
<obsed>
yep ;)
che1 has quit [Ping timeout: 244 seconds]
<jonwil>
Thank god I live in Australia where its possible to get internet and mobile phone from companies that aren't scumbags (referencing the AT&T comment earlier)
<DocScrutinizer05>
Pali: wazzup?
mvaenskae has quit [Ping timeout: 245 seconds]
<DocScrutinizer05>
ooh, strings has a vuln. OMFG!!! ;-P
Kabouik has joined #neo900
<DocScrutinizer05>
bencoh: right
<DocScrutinizer05>
running strings(1) on a unknown input in a root environment makes no sense whatsoever
<DocScrutinizer05>
actually I can't think of any script using strings
<kerio>
DocScrutinizer05: otoh, running strings MUST NOT run arbitrary code
<DocScrutinizer05>
sure it's a bug
<DocScrutinizer05>
no doubt
<DocScrutinizer05>
just I won't have sleepless nights from that one
<bencoh>
tell that to reverse engineers ;)
<bencoh>
(you do RE often as well dont you ?)
<DocScrutinizer05>
not really, on a level you imply here
<DocScrutinizer05>
I'm doing hw RE usually
sixwheeledbeast has quit [Ping timeout: 245 seconds]
<DocScrutinizer05>
then improving and augmenting and fixing bugs on that level
<DocScrutinizer05>
or find workarounds, see H-E-N
<DocScrutinizer05>
H-E-N been one dang piece of RE work. took almost a year
<DocScrutinizer05>
fallout results been BME replacement and bq27200.sh
<DocScrutinizer05>
incl several batteye monitor apps that based on it
<DocScrutinizer05>
btw you might think nothing happens when e.g. Oksana comes in here complaining "so quiet!". Actually there's almost always happening a lot of stuff behind the scenes. That's not because of any "sekrits" but simply boring work getting done in a closed channel of sorts
<DocScrutinizer05>
every now and then we do it in this channel, but usually not
sixwheeledbeast has joined #neo900
jonwil has quit [Remote host closed the connection]
modem has joined #neo900
modem has joined #neo900
mvaenskae has joined #neo900
sixwheeledbeast has quit [Ping timeout: 245 seconds]
<DocScrutinizer05>
[2014-10-27 Mon 13:54:08] <DocScrutinizer05> well, at least one of the reasons
<DocScrutinizer05>
[2014-10-27 Mon 13:56:07] <wpwrak> (BSI) you mean battery disconnect ? that would haev to come from the charger. the BSI pin doesn't seems to be very well defined. (i.e., can be many things)
<DocScrutinizer05>
[2014-10-27 Mon 13:56:42] <DocScrutinizer05> can be many things, but all have a certain pullup/pulldown
<DocScrutinizer05>
[2014-10-27 Mon 13:57:34] <DocScrutinizer05> BB5 modem shuts down immediately from BSI disconnect. Also APE does in N900
<DocScrutinizer05>
[2014-10-27 Mon 13:58:25] <DocScrutinizer05> for BL-5J battery BSI is rather well defined as "70..100kR to GND"
<DocScrutinizer05>
[2014-10-27 Mon 13:59:16] <DocScrutinizer05> of course would be funny when somebody actually uses a HDQ smartbat and first bit getting sent causes IRQ to tear down whole system ;-P
<DocScrutinizer05>
[2014-10-27 Mon 13:59:18] <wpwrak> yes, if things are static, then we could generate an interrupt, why not. the line goes to the CPU. we could add a 1 MOhm pull-up
<DocScrutinizer05>
[2014-10-27 Mon 13:59:37] <wpwrak> yeah, if communicating, you have to first disarm the interrupt :)
<DocScrutinizer05>
[2014-10-27 Mon 14:00:22] <wpwrak> and if you're really unlucky, the removal will happen during commuication. when then you won't find out before a protocol timeout. so better don't probe the bat too often :)
<wpwrak>
btw, if we have a very fast shutdown-on-battery-removal, then the ninja battery swap won't work anymore
<DocScrutinizer05>
yes, that's why fastswap needs to get 'announced' anyway. I also planned to charge special buffer capacitor prior to ninja swap
<DocScrutinizer05>
and signal a "OK, ready" via the 3color LED "privacy"
<wpwrak>
the coolness factor is dropping
<wpwrak>
a cute hack that needs to much work is like a joke that takes an hour of explaining ...
<DocScrutinizer05>
something along the line of "when battery lid opened while device active, prepare for uSd swap or ninja batt swap. When lid opened while device locked, do suspend to disk"
<DocScrutinizer05>
for suspend-to-disk the LED signal is Pink->off. For ninja swap it's Red->Green
<wpwrak>
how long does suspend to disk usually take ?
<DocScrutinizer05>
dunno
<DocScrutinizer05>
never did it
<wpwrak>
and how much would the system draw in suspend to ram ?
<DocScrutinizer05>
new feature with kernel 3.??
<DocScrutinizer05>
suspend to ram is useless, no better than 0-clock
sixwheeledbeast has joined #neo900
<DocScrutinizer05>
which is whole point of 0-clock in OMAP
<DocScrutinizer05>
and the reason why I always dispised suspend concept in freerunner
<DocScrutinizer05>
precharge big buffer cap shouldn't take longer than 3 s. Umount swap on uSD can take minutes
<DocScrutinizer05>
(when swap is actually on uSD)
<wpwrak>
ah, minutes is a bit long indeed
<DocScrutinizer05>
N900 umounts uSD hard. Which makes those users quite unhappy who want their swap on uSD, for performance and wear&tear reasons
trx has quit [Disconnected by services]
delphi has joined #neo900
<DocScrutinizer05>
not like uSD had higher performance than eMMC (where swap usually lives) but it has its own bus and controller in uSD
<DocScrutinizer05>
so system can read in new binary text while concurrently swapping out pages to free RAM for it
unclouded has quit [Ping timeout: 265 seconds]
<DocScrutinizer05>
when both /usr/bin and swap is on eMMC, it will collide
<DocScrutinizer05>
which is known to slow down N900 unbbearably
<wpwrak>
hmm, when suspending to disk, you shouldn't really have to read executables
<wpwrak>
just get rid of what's in ram, then shut down
<DocScrutinizer05>
that's completely unrelated
<DocScrutinizer05>
suspend to disk with swap on uSD is actually pretty twisted usecase
<DocScrutinizer05>
since suspend to disk using swap
<DocScrutinizer05>
feeling sick, bbl
<DocScrutinizer05>
aaah I see where I been ambiguous
<DocScrutinizer05>
>>[2014-10-27 Mon 14:08:28] <DocScrutinizer05> for suspend-to-disk the LED signal is Pink->off. For ninja swap it's Red->Green<< Or Red->Yellow for swap-on-uSD which forbids umounting uSD
<DocScrutinizer05>
so actually removing back lid when device in active state: PRIVACE RED: dont touch! -> YELLOW: ninja swap allowed, uSD still mounted, you prolly shouldn't do either of both (yet) -> GREEN: buffercap charged for ninja swap, uSD unmounted so you can swap that too
b1101 has joined #neo900
<DocScrutinizer05>
PRIVACY PINK: device entering suspend-to-disk state -> all OFF: suspended to disk
<DocScrutinizer05>
suspend to disk per default is forbidden when swap is on uSD
<DocScrutinizer05>
2119 you SHOULD NOT change uSD when swap is on it
<DocScrutinizer05>
~2119
<infobot>
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.
<bencoh>
<3
unclouded has joined #neo900
<DocScrutinizer05>
of course we could allow a special "suspend to disk" that swaps all out to uSD and then suspends CPU while keeping PRIVACY LED shining yellow, to signal "do NOT touch the uSD! You however may swap the battery now"
<DocScrutinizer05>
obviously the LED would go dark as soon as you remove battery. And not come up again on inserting new battery, until you "boot2 the device which actually will be resume from disk
<DocScrutinizer05>
disclaimer: all this are just considerations for now! This discalimer needed since some seem to think IRC is a channel for official announcements
<DocScrutinizer05>
[OFFICIAL ANNOUNCEMENT] it actually is! ;-) Under certain circumstances
<DocScrutinizer05>
bbl
<DocScrutinizer05>
wpwrak: I think after a last few edits according to our former chat, the diagram.htm is ready for public
<DocScrutinizer05>
:-)
<DocScrutinizer05>
((<wpwrak> btw, if we have a very fast shutdown-on-battery-removal, then the ninja battery swap won't work anymore)) which is most likely the major reason why it doesn't work on N900 fremantle
<DocScrutinizer05>
not even with USB charger attached (obviously unless you previously did `stop bme`)
mvaenskae has quit [Ping timeout: 265 seconds]
astr has joined #neo900
mvaenskae has joined #neo900
MonkeyofDoom has joined #neo900
che1 has joined #neo900
che1 has quit [Ping timeout: 255 seconds]
silviof has joined #neo900
che1 has joined #neo900
mvaenskae has quit [Quit: leaving]
Pali has quit [Ping timeout: 265 seconds]
Pali has joined #neo900
obsed has quit [Ping timeout: 245 seconds]
mvaenskae has joined #neo900
obsed has joined #neo900
svree4 has quit [Ping timeout: 255 seconds]
modem has quit [Ping timeout: 245 seconds]
<DocScrutinizer05>
haha, never thought that Neo9001 would make such noise. Did I really miss to ever mention the perspectives?
<fw190>
are there any estimates for neo900 finnal price?
wazrus has joined #neo900
b1101 has quit [Quit: b1101]
wazrus has quit [Quit: leaving]
freemangordon1 has joined #neo900
freemangordon has quit [Read error: Connection reset by peer]
nox- has joined #neo900
<DocScrutinizer05>
700+ EUR for the NeoN bare board. And on top whatever it costs to get the N900 display and mech parts
<DocScrutinizer05>
unless you got a N900 to do a mainboard swap
PeperPots_ has quit [Ping timeout: 272 seconds]
<bencoh>
"NeoN" :]
PeperPots_ has joined #neo900
<DocScrutinizer05>
yeah, obviously STEP2 board will be called ArgoN, or even CryptoN
<bencoh>
oh ... NeoN might actually be a better "marketing" name ;)
<DocScrutinizer05>
NeoNinehundred
<bencoh>
:)
<freemangordon1>
dammit, I am unable to make the last REed function as fast as stock :(. I must be missing something
<DocScrutinizer05>
now I bet you'd love to have an ETM at your disposal
mvaenskae has quit [Quit: Lost terminal]
paulk-aldrin has quit [Ping timeout: 258 seconds]
paulk-aldrin has joined #neo900
freemangordon1 has quit [Quit: Leaving.]
freemangordon has joined #neo900
<freemangordon>
well, it is not *that* bad (~15% slower), but still
<freemangordon>
maybe I should just scratch the intrinsics and do it in ASM
<freemangordon>
but I still can't believe gcc can be worse at scheduling than hand-crafted assembly
<freemangordon>
esp when all other functions are on par or faster
<freemangordon>
gcc produces the same number of NEON instructions as there are in stick code
<freemangordon>
so it *must* be something else, but I just can't find it :(
<DocScrutinizer05>
number of instructions isn't already proportional to time it needs to execute them
<DocScrutinizer05>
optimized utilization of pipes might be one thing, optimized instructions (choosing equivalent but faster ones) another
<DocScrutinizer05>
and actually preparing the whole thing is probably also relevant, though that's wild uneducated guessing now since I have no experience or even knowledge about NEON
<freemangordon>
the count and the instructions themselves are the same
<freemangordon>
and actually I still believe gcc mixes ARM/NEON them better so both instruction pipelines are used
<freemangordon>
so it is a mistery to me
<freemangordon>
maybe my benchmark is flawed
* freemangordon
is afk
<ds2>
NEON is a co-processor
<freemangordon>
ds2: so?
<ds2>
gcc doesn't have good scheduling. for that you want ARM CC
<ds2>
so NEON is more or less independant of the main A8 unless you are doing something that forces a sync
<freemangordon>
ds2: all other functions I used intrinsics in so far are either exactly as fast as stock or faster
<freemangordon>
I don;t think it is bad scheduling, it is rather something related to the cache
<ds2>
ah, we have a terminology difference... I lump cache in with the scheduling stuff
<freemangordon>
hmm, going to try something
<DocScrutinizer05>
http://talk.maemo.org/showpost.php?p=1444641&postcount=190 >>I agree that we need people doing "housekeeping tasks", but I don't think that we need elected mandate for housekeeping. It could "just work" as volunteers task forces - additional benefit is that, without mandate of being elected, none of the volunteer will ever get an idea of having mandate to "rule" maemo, as we have seen with some aspects of Council in the past.<<
<DocScrutinizer05>
SUUUURE go ahead! login for ML maintenance is mailamn@maemo.org, password is '!toidI_letsE', at least until Estel is changing it to what he thinks is better suited for him
<DocScrutinizer05>
then he can resume his habit to send inquiries and orders to hw stores in name of maemo council form council@maemo.org addr
<freemangordon>
ds2: BTW if you have knowledge on that NEON stuff I will appreciate any help I can get :)
fw190 has quit [Ping timeout: 246 seconds]
<ds2>
freemangordon what do you need?
jonwil has joined #neo900
<jonwil>
hi
<Oksana>
:) Good morning!
paulk-aldrin has quit [Ping timeout: 258 seconds]
paulk-aldrin has joined #neo900
<Oksana>
What would you want as next device to be developed after Neo900 : something similar, like N950, or something radically different, like Siemens ME45 ?
<Oksana>
Rugged, shock-proof, water-proof, et-cetera small old phone.
<Oksana>
I was discussing the idea of air-borne phone recently, so that when you are angry at something slow (a person not taking the call) you 'throw' the phone, it flies around, and returns when the call has finally started.
<DocScrutinizer05>
next device is meant to get produced in volumes that allow own case made for it