#qi-hardware IRC log for Monday, 2010-12-06

xiangfuexit01:57
kyakxiangfu: hey02:52
kyakxiangfu: btw, mocp is working just fine for me, with the Makefile in git (only removed the BROKEN tag)02:53
kyakit complains about terminal size, however.. But doesn't segfault02:53
xiangfukyak: oh.. hmm.. I only test in my laptop. which the .config is different. I will try in buildhost.02:54
xiangfukyak: I will try to find out why.02:54
kyakxiangfu: i didn't try on buildhost, too..02:55
kyakxiangfu: seems that stardict builds its own version of libsigc, and links against it. And it causes some problems for me. I tryed adding libsigcxx to stadict's dependecies, this should help (i hope)06:55
kyakin the full build, it uses libsigc from openwrt06:56
kyakalso, when building stardict in minimal build, i foudn that it lacks for libstdcpp dependency06:57
xiangfukyak: oh. good job :)06:57
kyakbut.. this can be confirmed only in the evening :)06:58
xiangfuI just send out some email to u-boot, send ben nanonote patches to upstream :)06:58
kyakoh great! is our u-boot very far away from upstream?06:58
xiangfukyak: yes. it's eating time for me. I will be back in ~ 1 hour. talk to you later. sorry.06:59
kyakbon appetit!06:59
xiangfukyak: Ingenic's u-boot very far from upstream. ours much much better. :)07:00
kyakwould be nice to get rid of u-boot some day :) and use grub07:00
lekernelwhy?07:01
kyakbecause uboot needs specially built uImage and uboot doesn't provide a great flexibility (menus, editing kernel cmdline etc) , like grub does?07:04
qi-bot[commit] Andres Calderon: Has been moved the labels that overlap. http://qi-hw.com/p/xue/fb92de307:06
virickyak: what dictionaries you use in stardict? You seem very interest in it07:11
viricinterested07:11
kyakviric: russian<->english and english <-> german07:12
kyaki mostly use sdcv however (console version of stardict)07:12
kyakstardict fails (or failed?) to start on my minimal build07:12
kyaki'm pretty much sure it will start now...07:13
viricok07:25
viricI use, for similar situations, 'dict'.07:25
xiangfukyak: we can set kernel command in ben nanonote.07:26
virickyak: http://vicerveza.homeunix.net/~viric/akcentiga/index.html.en07:26
xiangfu1. run /usr/bin/fw_setenv_default  which write the default u-boot envs to nand.07:26
xiangfu2. using fw_setenv bootargs .... for set the cmdline.07:27
xiangfukyak: I think we should create a package like "nandnote-full-system-files", put all gmenu2x, ... all files not rewrite openwrt base file to this package.07:30
kyakviric: as i remember, "dict" is gtk1, and gtk1 can't work with framebuffer (needs X)07:30
kyakxiangfu: yeah, we can use the fw_setenv_default for that, but with grub it's just easy as pressing 'e' during boot. and the menu?07:31
virickyak: 'dict' is a server program07:32
virickyak: and it comes with a console client07:32
xiangfukyak: hmm.. then we must write keyboard and lcd driver for grub.07:32
kyakxiangfu: i think it is better to move such files to relevant packages. The files that are not a part of some pacakge, can be left as they are? (like /etc/inputrc)07:32
virickyak: you can install other clients too.  It works through serving in tcp/localhost.07:32
xiangfukyak: compare to grub. maybe kboot is much better.07:32
kyakviric:  ah ok.. need to have a look07:33
kyakxiangfu: maybe! i never really tried anything except for grub/lilo07:34
kyakviric: put a word "60;N78" in your accentizer, it won't work ;)07:35
virickyak: eh... :)07:35
kyakxiangfu: hm, i didn't think it was so hard.. a special driver for keyboard/lcd for grub?07:35
kyakviric: how does it work anyway?07:36
virickyak: http://vicerveza.homeunix.net/~viric//cgi-bin/Dict dictionaries07:36
kyakparsing some dictionary?07:36
virickyak: yes07:36
kyakoh ok07:36
virickyak: did you notice that you can click on words?07:37
viric(in the 'accentizer')07:37
kyakuep07:37
viricI think it's a great tool... but I failed to spread it.07:37
xiangfukyak: I am not sure about grub. I just want say. if we can write those driver for grub, then we can write them for u-boot. if you have a keyboard in u-boot. then you can easy edit just as grub :)07:37
kyakthe tooltip is a little bit over the word though07:37
kyakin firefox07:37
virickyak: I only learnt the very very minimal javascript+css for that ;)07:37
viricI use the akcentiga almost always I need to read anything in Russian07:38
viric(I wrote it some years ago)07:38
kyakhmm, i didn't realize it is hard to put accents for foreigners.. but i guess, yes, it is hard.. there are not general rules for that, a lot of excpetions :)07:39
xiangfukyak: since the gmenu2x store in projects.qi-hardware.com, maybe we just put all files/usr/share/gmenu2x/* to gmenu2x.git07:40
kyakxiangfu: having menu in u-boot would be cool ;)07:40
virickyak: My sixth sense for guessing the accent usually puts the accent in a totally wrong position ;)07:40
virickyak: and it's very hard to find texts with accents marked.07:41
kyakxiangfu: do you think maybe it is better to have those files in gmenu2x in openwrt-packages?07:41
xiangfukyak: my plan on u-boot is first send to upstream  then add new feature, so we stop add new thing to u-boot. (I am a little slow on send ben nanonote to upstream :-(07:41
kyakviric: believe me, there are a lot of native speakers who are putting wrong accents07:42
viricI imagine. But not on almost every word I imagine. :)07:42
kyakusually other people laugh in their faces07:42
kyakno, not like that ;)07:42
virichaving a quick dictionary like that of clicking the words also helps a lot.07:43
kyakxiangfu: this is a good plan!07:43
viricAnd it's no client-server interaction.. the HTML has all the dictionary definitions. You can save the html and have the click-code working07:43
kyakviric: http://slovari.yandex.ru/~:=838/!;>20@L%20C40@5=89/07:45
kyakA;>20@L C40@5=89 )07:45
kyakxiangfu: i think putting gmenu2x files in gmenu2x.git will lead to mixing between source code and configs... I think configs need to be provided by gmenu2x in openwrt-packages07:48
kyakconfigs+icons07:48
xiangfukyak: yes. sound good, let's do it.07:51
kyakviric: dict homepage - is it http://www.dict.org/?07:52
kyakxiangfu: btw, you've shown a nice way of postinstall handling (in nightsky). There are still files in /root that need to be moved to relevant packages (like netsurf and mplayer)07:54
kyaki don't know what to do with /root/.vimrc though :)07:54
kyakvim is from openwrt07:54
xiangfukyak: yes I will work on the move those files to relevant packages.07:56
kyakor /root/.tclshrc07:56
kyakok, great!07:57
xiangfukyak: I also not sure about .vimrc .tclshrc . let's keep it in files/ for now.  the plan is stick upstream as much as possible.07:57
kyakyes, they live in file/ just fine for now..07:58
kyakhmm, i just noticed there is a qstardict08:00
kyaktaking into account that cyrillic input is working in Qt, might be a better choice than stardict08:01
kyaki wonder what's working faster08:01
virickyak: yes08:02
qi-bot[commit] Xiangfu Liu: mv all gmenu2x stuff to it's package http://qi-hw.com/p/openwrt-xburst/32e461108:29
qi-bot[commit] Xiangfu Liu: move gmenu2x from openwrt-xburst.git to it's package http://qi-hw.com/p/openwrt-packages/e517e9708:29
qi-bot[commit] Xiangfu Liu: remove temp file http://qi-hw.com/p/openwrt-packages/ce1b4fa08:35
qi-bot[commit] Xiangfu Liu: add icon for snownews http://qi-hw.com/p/openwrt-packages/af5690908:35
virickyak: Qt on directfb?09:05
viricxiangfu, larsc: yesterday I had a problem on 2.6.3509:05
viricLeave a cpu intensive program run for more than 10 minutes09:05
viric(maybe others can do the test)09:06
viricAnd check its 'user time' and 'cpu time' (/proc/PID/stat some numbers there)09:07
viricAfter more or less 11 minutes, the user-time counter stops, and all gets added into the system-time counter.09:08
kyakviric: yes, Qt can run on top of directfb09:10
viricA usual cpu intensive program you can run is "openssl speed"09:10
viricIt checks the 'user-time' counter to display the cpu time passed09:10
viricand after 11 minutes of running, it starts saying 0.0;09:11
viricmonitoring with strace, I see that what accumulated as user time, it goes to system (kernel) time and the user-time counter stops.09:11
viricDifferent runs make the stopping moment vary. But it's always between 10 and 11 minutes, for what I tried. I tried with gzip and openssl speed.09:12
viricCan someone put his nanonote with the heavy job of 'openssl speed'?09:12
viricIs there a report of what is ready for the nanonote in 2.6.36?09:14
viricmaybe noone here has a running nanonote? :)09:19
viricxiangfu testing 'nerase', kyak with qemu, ... :)09:19
kyaknanonote? i heard someinth about it..09:21
kyak:)09:21
bartbesviric: running? :O09:21
bartbes:P09:21
virichehe09:21
viricI mean with an OS there and able to run programs09:21
viricI'll try to build a 2.6.36 kernel09:22
viricfor i09:22
bartbesmaybe I do..09:22
virict09:22
viricbartbes: could you make it run for 11 minutes (at least) "openssl speed"?09:22
bartbesif you keep the time, I'll surely forget09:22
viricwell09:22
bartbesno openssl installed?09:22
viricyou can make it run until it finishes :) It may take around 20 minutes.09:22
viricah you don't have openssl?09:23
viricbad.09:23
viricsooo09:23
kyakopenssl binary is not installed, yes09:23
viricanother thing. Try to make it run:   gzip </dev/zero > /dev/null &09:23
virickyak: I cannot reproduce the trouble in 2.6.36 malta, btw.09:23
bartbesviric: alright, it's doing that09:24
viricbartbes: then, note the gzip PID, and paste from time to time /proc/PID/stat here.09:24
bartbesand now we wait09:24
viric(a line with numbers)09:24
bartbesugh, time for ssh, so I can copy-paste09:25
viricah09:25
viricbartbes: you can run:   cat /proc/PID/stat | cut -d ' ' -f 12,1309:26
viricthat will choose the numbers I want to run09:26
viricI want to *see*09:26
bartbesviric: 3 009:26
viricouch. maybe not those then09:27
bartbes663 (gzip) R 436 663 436 1025 663 4194304 168 0 3 0 17827 917 0 0 20 0 1 0 8550 1638400 127 2147483647 4194304 4653656 2142728432 2142727632 4417872 0 0 0 0 0 0 0 18 0 0 0 0 0 009:27
virictry 14,1509:27
viricaha, 14,1509:27
viric17827 = user jiffies = 178.27 seconds of user-mode time09:28
viric917 = system jiffies = 9.17 seconds of kernel-mode time09:28
viricSo what I see here is that, around 64000 jiffies (640 seconds), the user-mode time jiffie counter STOPS, and then all the time goes to the system jiffies (kernel mode time)09:28
bartbes27413 138809:29
viriclooks fine.09:29
viriclet's wait for the 640 seconds09:29
viricbartbes: what kernel you run btw?09:29
bartbesehm..09:30
viric(thank you very much for the test)09:30
viricuname -a09:30
bartbesroot@BenNanoNote:~# uname -a09:30
bartbesLinux BenNanoNote 2.6.32.10 #1 PREEMPT Tue Jun 15 17:53:33 CEST 2010 mips GNU/Linux09:30
bartbesyeah did that already09:30
viricok09:30
viricgood test then.09:30
bartbes36223 183409:30
viricthat looks fine. user time increasing a lot, system time increasing little.09:30
bartbesso ehm09:31
kyakbartbes: you are good to update your system :) pretty old summer image09:31
bartbesapparently you know09:31
bartbeswhat are those time values supposed to be anyway?09:31
bartbeskyak: well did SDL get fixed?09:31
viricbartbes: user time in centiseconds, and kernel mode time in centiseconds09:31
kyakbartbes: hm, did it get broken?09:31
bartbeswell yeah, but is the meaning of it?09:31
bartbeskyak: it did, sometime09:31
viricbartbes: How much time the process is spending CPU time in user-mode, and how much time the process is spending CPU time in kernel-mode (inside system calls)09:31
kyakbartbes: should be fine in the last testing image09:32
bartbeskyak: no stables coming up?09:32
bartbes46725 233909:32
viric(I'm building 2.6.36 to test if I also see the problem therE)09:32
kyakbartbes: http://en.qi-hardware.com/wiki/Official_Software_Image#Image_2010-11-1709:33
kyaki think it's official;09:33
bartbesgforth added, yay09:33
kyakbeware though that the rootfs is 512 Mb now. you shoud probably save your data from datafs09:34
bartbesI don't think I have much data anyway09:34
bartbesthe thing I'm worried about is confs ;)09:35
kyakoh. do youreally have a lot of modifications?09:35
viricwait - are you overwriting the whole rootfs to update your openwrt in the nanonote?09:36
bartbesisn't that.. necessary?09:36
bartbes69752 346109:36
viricbartbes: 69752! I could not reach this. So it seems you are not seeing the problem I was seeing09:36
viricbartbes: maybe it was something particular of the 2.6.35 I was running then.09:36
bartbesI will keep it running for a little bit09:36
viricbartbes: thank you109:37
viric!09:37
viricSo, for openwrt... I have no idea, I never used openwrt on the nanonote.09:37
viricBut I expected it had some update system...09:37
bartbesI also have jlime on my sd card09:38
bartbes82731 406209:38
bartbeslooks like it's working properly09:38
kyakopkg can update. however now i think reflashing a newer image is only supported09:38
viricbartbes: yes. Fine. thank you very much :)09:39
kyaki think it's considerably faster, too09:39
bartbesroot@BenNanoNote:~# kill 66309:39
Last message repeated 1 time(s).09:39
bartbesash: cannot kill pid 663: No such process09:39
bartbeswee09:39
viricbartbes: I'll check on 2.6.36. Maybe the bug was introduced at some time before 2.6.32 and 2.6.3509:39
bartbespeace and quiet returns09:40
viric:)09:40
bartbesI.. forgot that shutting down from ssh leaves my ssh client wondering what the hell happened...09:40
viric~.09:40
viricThat kills the openssh client09:41
bartbesthe server, not the client09:41
viricah09:41
bartbesso I'll have to wait until it times out, or kill it09:41
virichow you know it's the server?09:41
bartbesbecause.. the server runson on the ben?09:41
viricIsn't it shut down?09:42
bartbesand I just shut down the ben?09:42
bartbeswell exactly09:42
bartbesso the server is down09:42
bartbesthe client doesn't know that yet09:42
virictype ~.09:42
viricthat kills the client :)09:42
bartbesno it doesn't09:42
viricwhat ssh client you useƧ'09:42
viric?09:42
viricopenssh client is killed by 'tilde dot'09:42
viricsorry09:42
bartbesopenssh09:42
viricenter-tilde-dot09:42
bartbesyeah, but it needs to be on its own line09:42
bartbesI think that's where it fails09:42
bartbesand I may or may not have overridden that key combo09:43
viricmaybe the stop sequence on the ben is broken; it shuts down before the ssh server can properly finish.09:43
viricbartbes: why would you override enter-tilde-dot? :)09:43
bartbesoh hey, I finally got it to work09:43
bartbesor it timed out..09:43
bartbesviric: I could have changed the key09:43
viricyou should not forget the 'enter'09:43
bartbesI had to do alt-gr+shift+tilde then dot09:43
kyakhm, never heard of that magic ~. :)09:44
viricenter-tilde-?  explains other combinations09:44
kyaki had to kill the process manually09:44
viricit's similar to the telnet ^]09:44
bartbeskyak: and the enter is there because it needs to be on its own line09:44
kyakhm09:44
kyakneed to try it09:45
viricown line?09:45
viricwhat own line? :)09:45
kyakhow openssh client distinguish - maybe i want to send ~ to the server?09:45
virickyak: enter-tilde-tilde sends enter-tilde09:45
kyakno, i mean i want to type ~09:46
kyaklike cd ~09:46
bartbesviric: you can't do:09:46
bartbesa~.09:46
bartbesbut you can do:09:46
viricit does not have any problem. You can type ~ whenever you want. You only have to care if you want to type it AFTER *enter*09:46
bartbeslol, fail09:46
bartbes~.09:46
viricand in that case, type it twice.09:46
bartbesI just disconnected trying to type that09:46
bartbes:P09:46
kyakheh09:46
viricThere is not any magic. The parser takes the sequence "enter-tilde", and that's all.09:47
kyakok, great thing to try when i get to openssh09:47
viricit's not about any line. :)09:47
bartbesviric: not true09:47
bartbesit searches for ~ at the start of the line buffer09:47
viricI'm at an irssi session09:47
bartbesor so some docs said09:47
viricI might have echo disabled09:47
viricI might have the terminal in raw mode  (so no line buffer)09:47
viric(like what I have now at the irssi session :)09:48
bartbesviric: yeah yeah09:48
bartbeswhatever09:48
bartbesit just needed to be at the start of the line09:48
virichaha09:48
bartbeswhich corresponds to \n~09:48
viricno no09:48
viricit's Enter-Tilde09:48
viricnot \n Tilde09:48
viricIn xterm, vt100 and other terminals, the enter key in raw mode provides a scancode of \r, and not \n.09:48
bartbesviric: that was to signify a newline!09:49
bartbes:@09:49
viricAnd mine was to signify a carriage return ;)09:49
bartbes!$@# those scancodes09:49
bartbes(censorship provided by Quality Boats, get your censorship today!)09:50
virichehehe09:50
viricutime 31038, stime 25. let's see how far it goes.09:50
bartbesuntil it overflows ;)09:52
viricWill anything be faster in 2.6.36? Who knows09:53
bartbesI wonder09:53
bartbescan those counters overflow?09:53
bartbes(well, why wouldn't they?)09:53
viricfew may be worried on that :)09:54
bartbesoh it won't cause anything bad09:55
bartbesjust wondering if it can09:55
viricI don't think anyone implemented infinite-precision arithmetic for those.09:56
viricutime 6003709:56
viricutime 6412509:56
viric6611009:56
viricgreat great09:56
viricI consider it fixed on 2.6.3609:56
bartbesyeah thought so09:57
bartbesI mean, why would you bother?09:58
viricwell, all benchmark reports from openssl get broken09:58
viricas it uses the 'user time' as measure.09:58
bartbesmeh09:59
bartbes:P09:59
viricupdating openwrt through reflash looks annoying10:02
xiangfuviric: kyak have added flashing progress. which is better now :), but I would like advice flash from sd card. :)10:29
viricprogress where?10:30
viricin usbboot?10:30
xiangfuviric: in reflash_ben.sh10:33
viricahh.10:33
qi-bot[commit] Xiangfu Liu: add nanonote-example-files pakcage to config.full_system http://qi-hw.com/p/openwrt-xburst/225b32c10:36
lekernelhttp://www.getupanddiy.com/projects/thin-film-sputtering-machine/index.php11:30
wpwrak_lekernel: working on your fab ? ;-)11:36
wolfspraulgreat article11:38
lekernelwpwrak_: preparing said work11:38
wolfspraulI saw a machine once at a lense maker, I think it was some sort of thin film as well, not sure11:38
lekernelwpwrak_: right now I have friends who got hands on vacuum pumps (2 primary pumps, 1 mercury (!) diffusion pump, 2 turbomolecular pumps and even one ion pump)11:39
lekerneli'm going to check that out mid January11:39
lekernelI also found an old evaporator from the 70s that no one uses and I can play around with... problem it weights hundreds of kg and is 1500km from where I live11:40
wolfspraulwhat do you plan to build?11:40
wolfspraul(or manufacture)11:40
lekernelmess around with semiconductors, OLED displays, etc.11:40
lekernelnothing commercial, just a fun hobby11:41
wolfspraulgreat11:42
wolfspraulwe find a commercial application :-)11:42
lekernelhm, will be hard imo :)11:43
wpwrak_lekernel: well, we need a better lcm ... ;-)11:45
lekernelLCM?11:46
lekerneldisplay11:46
lekernel?11:46
lekernelwell manufacturing them myself won't probably be economically viable11:46
wpwrak_yes. lcd module11:46
wpwrak_naw, invent the ground-breaking technology. let other figure out how to produce it :)11:47
wolfspraulroh: you there?11:53
rohwolfspraul hey11:53
wolfspraulah hi!11:53
wolfspraulI posted a few things for you the other day, about the m1 case11:54
wolfspraul2 boards are on the way to Sebastien, hopefully they arrive tomorrow or the next few days11:54
rohnice11:54
wolfsprauldid you read what I wrote? was about tolerances I think11:54
wolfspraulworst seems to be the DC jack11:55
wolfspraulI didn't know how bad this is until I saw it now :-)11:55
wolfspraulso that one needs a lot of tolerance left and right11:55
roheh.. nope.. email?11:55
wolfspraulin the next run we may want to tighten that a bit11:55
wolfspraulprobably here in the channel11:55
wolfspraulok let's just redo it11:55
wolfspraulyou asked about tolerances11:55
wolfspraulthe connectors to watch out for are: 1) DC jack 2) line-in and line-out 3) vga maybe a little11:56
wolfspraulin that order11:56
wolfspraulDC jack needs at least 1mm left and right tolerance11:56
rohok in case of doubt i can always add some more space11:56
roh1mm is _huge_11:56
wolfspraulif you want to try, you can unsolder the DC jack on the sample you have, then you see how much wiggle room you have11:56
wolfspraulyes11:57
wolfspraulI know11:57
wolfspraulthere was a back and forth about vendors, and the result is that the holes in the pcb are quite big11:57
wolfspraulunsolder your DC jack on the sample, then you see it yourself11:57
wolfspraulso DC jack is the worst11:57
wolfspraulafter that - line-in and line-out11:57
rohwell.. its a dc jack.. i think the contacts are quite normal (the holes)11:57
wolfspraulyes 'quite'11:58
wolfspraulbut when you go into the details and tolerances well there are still differences11:58
wolfsprauljust unsolder and see yourself :-)11:58
rohlets just try some boards in plural and see how it comes out11:58
wolfspraulyou can try the same for line-in and line-out11:58
wolfspraulok up to you11:58
wolfspraulunsoldering the DC jack will give you a quick idea11:58
wolfspraulhow is it going with the case?12:00
rohwaiting for the rc2 board12:01
rohgot my spacers12:01
qi-bot[commit] kyak: stardict: add missing dependencies http://qi-hw.com/p/openwrt-packages/6c23e1312:15
qi-bot[commit] Werner Almesberger: ecn/ecn0007.txt: Revised via spacing (for the antenna) http://qi-hw.com/p/ben-wpan/fbd3e5316:59
kristianpauloh well..17:55
kristianpauli hope i wasnt too hard on those mails17:56
wpwrak_kristianpaul: naw. you're right. carlos should be in a very good position to post about "interesting" things. so if he doesn't do it, he shouldn't complain that others are filling the void18:09
kristianpaulwpwrak_: you'were righ about your comments yday18:49
kristianpaultesting now, i got the SPI to parallel now, but still missing SYN part i hope get it soon18:49
wpwrak_kewl. can you use DMA with that parallel interface ? i haven't examined the fancier GPIO features yet, but I would suspect that you can't.18:58
kristianpaulhmm close but now i'm missing 4 nibbles per convertion :S19:49
kristianpaulhttp://kristianpaul.org/~paul/tmp/spytoprallel.png19:49
kristianpaul(fianlly got an esy way of use scp and not deal with www permisions :))19:49
wpwrak_kristianpaul: downloads.qi-hardware.com works quite nicely, too ;-)19:50
kristianpaulahh well19:51
kristianpaulthats for the dumps for sure :)19:51
wpwrak_the problem may be that you're using values > 3. given that you get a sync every 4 clock cycles, that can't work19:52
wpwrak_so one has to give ...19:53
wpwrak_(well, or at least that's what it was like a while ago)19:53
kristianpauli realy need a verilog boof for chrismass after this19:54
kristianpaulah, wait i can move an if some where19:54
kristianpaulwait19:54
wpwrak_booK ? :)19:54
kristianpaulnah ,)19:54
kristianpauljust "following" the buying spirit19:55
kristianpaulWell i think i may be too stric, cause once i got a  sync signal i can predict next one and so on... well if i really trust hardware...20:00
wpwrak_why not just re-sync ?20:02
wpwrak_or change as follows: have an LSR and an output buffer. shift into LSR, on sync, clock LSR to output buffer.20:04
wpwrak_you need something like this anyway :)20:05
wpwrak_so SYNC is not the beginning of a data packet but the end of the previous one20:05
kristianpaulhmm  i dint consider that one20:06
wpwrak_if there's the wrong number of clocks between syncs, you get garbage. but then, you lose in this case anyway.20:06
wpwrak_if you really want to get fancy, you could have an "error" signal20:06
kristianpaulha you got me20:06
kristianpaulbut thats for later20:07
kristianpaulfor sure20:07
kristianpaulok now it works http://kristianpaul.org/~paul/tmp/spitoparallel.png20:43
kristianpaulbut i *trust* on sync20:43
kristianpaulmay be i can do other modeling to keep eye on sync good behave20:44
wpwrak_you mean you never reset ? :)20:44
kristianpaulyes20:44
kristianpaulno wait20:44
kristianpauli mean yes i do reset20:44
kristianpaullook mcu_rst20:44
wpwrak_so, no trust on sync then ?20:45
wpwrak_naw. i mean "reset the shift"20:45
wpwrak_the full system reset is something else :-)20:45
kristianpaulahh thats a problem20:45
kristianpaul(reset the shift)20:45
wpwrak_well, you don't need to reset the shift. just copy data over to the latch20:45
kristianpaulaja20:46
kristianpaulo just reset count20:46
kristianpaulwait i'll commit20:46
wpwrak_now that you have a shifter, there shouldn't be a counter ;-)20:46
kristianpaulhehe20:47
kristianpaulwell that help me to..20:47
qi-bot[commit] Cristian Paul PeƱaranda Rojas: Code now uses a shift register, and sync once to sync signal http://qi-hw.com/p/ben-gps-sdr/0324a5320:51
kristianpaul"The description style you are using to describe a register or latch is not supported in the current software release."20:54
kristianpauloh here we go20:54
kristianpauli wonder if sofware compilers have those kind of messages :S20:55
wpwrak_"internal compiler error" ? ;-)20:57
kristianpaullol20:57
wpwrak_you indentation style looks very confusing21:00
wpwrak_youR21:00
kristianpauli know21:00
kristianpaulit looks for me too21:00
kristianpaulwjay you suguest?21:00
kristianpaulwhat*21:00
wpwrak_pascal style: if (...)\nbegin\n\t...\n\t...etc...\nend21:01
wpwrak_so the begin/end are at the same level as the if21:01
wpwrak_i hope you realize that the counter now does absolutely nothing :-) it's all input, no output21:03
kristianpaulyes yes21:03
wpwrak_why not   always @ (posedge clk) begin if (sync) out <= shift_reg; shift+ref <= { data,shift_reg[WIDTH-1:1]}; end   ?21:05
wpwrak_(i hope i got the syntax right)21:05
wpwrak_err .. s/shift+ref/shift_reg/21:06
kristianpauli'll comment it for now21:07
kristianpaulbut i may help21:07
kristianpaulis my watchdog21:07
kristianpaulwell watch for now21:07
kristianpauli'm missing the dog21:07
wpwrak_ah, i see. well, a bit of confidence ... :)21:09
kristianpaulhttp://kristianpaul.org/~paul/tmp/wpwrakspitoparallel.png21:18
kristianpaul(confidence) sure you will :)21:18
wpwrak_that doesn't look too bad :)21:18
kristianpaulassign out = shift_reg;21:22
kristianpaulthat fits better21:22
kristianpaulwich is actually the snippet i founded http://hdlsnippets.com/verilog_parameterized_serial_in_parallel_out_shift_register21:22
kristianpaula bit close21:23
kristianpauli'm just afraid it start to shift data even when not sync signal is there :/21:24
wpwrak_yes, it will shift without sync. but that's an abnormal situation anyway.21:26
wpwrak_so what every you do it wrong :)21:26
wpwrak_(assign) hmm, "procedural continuous assignment". seems to have the same effect as a normal assignment, no ?21:27
wpwrak_(in this case)21:27
kristianpaulorder vs parallel21:29
kristianpaulhttp://www.asic-world.com/verilog/vbehave1.html#Blocking_and_Nonblocking_assignment21:29
wpwrak_ah ... got it. but that's = vs. <=, not assign vs. <=/=21:30
wpwrak_so ...21:30
wpwrak_... out = shift_reg; shift_reg = { data,shift_reg[WIDTH-1:1]}; ...21:30
kristianpaulah yes21:31
kristianpaulsorry21:31
kristianpaulassign is used for wiring to real world actually21:32
kristianpaulthats why i now21:32
kristianpauls/why/what21:32
kristianpauls/now/know21:32
kristianpaulargg typo21:32
wpwrak_isn't the "real world" part something that's in the declaration ? (not sure)21:33
wpwrak_i have a book on verilog, and it says that "assign" basically overrides assignments with <= and =21:33
wpwrak_the example being a reset signal: while reset is asserted, you want to force the output to a given value, even if there's a clock or any other change of input21:34
kristianpaulhmm you got me on that, i'm learning verilog btw ;)21:34
wpwrak_so when reset gets asserted, you "assign" the reset value. when reset gets deasserted, you "deassign" it. if anything else tries to change the value with = or <= while "assign"ed, nothing happens21:35
wpwrak_or at least that's how i understand it :)21:35
wpwrak_don't worry, i don't even know verilog. i just have a book :)21:36
kristianpaulThe assign and deassign procedural assignment statements allow continuous assignments to be placed onto registers for controlled periods of time. The assign procedural statement overrides procedural assignments to a register. The deassign procedural statement ends a continuous assignment to a register.21:40
kristianpaulspace.gif21:40
kristianpaulops21:40
kristianpaulI need read mre21:41
kristianpaulhmmm i may want to do a 16bit register at once in order to save time21:46
wpwrak_sounds like the great return of the counter :)21:47
wpwrak_16 bit shifter, count to 4 syncs, then out = shift_reg21:48
wpwrak_or use a multiplexer. not sure which is more efficient.21:48
kristianpaulneither me, actually thats  a question for xilinx eng21:49
wpwrak_don't you get some chip resource usage info when you do the synthesis ?21:50
kristianpaulsure21:50
kristianpauli can do becnhmark later21:50
kristianpaulalso i get timings21:50
wpwrak_then you can just implement both and see which comes back smaller21:50
kristianpaulsure21:50
kristianpaulokay 16 bits looks good also to take the whole A port21:52
kristianpaulbut  thats not valid for a ben.. if i may want try on it later..21:52
wpwrak_for then ben, you could implement SDIO ! ;-)21:52
kristianpaulthats 8 bits?21:53
kristianpaulor 4..21:53
kristianpauli always ask the same :/21:53
kristianpaullet see21:53
wpwrak_4 bits + clock + command21:55
kristianpaulah ok i'll stick on 4 and let that work (16bit) to sofware21:56
wpwrak_for sdio, things get a little more complicated ... e.g., the clock comes from the host21:56
kristianpaulhmm21:57
wpwrak_so you may need a buffer for the whole block21:57
kristianpaulyeap21:57
kristianpaulwhats the clock rate?21:57
wpwrak_but that's something you don't have to worry about now :)21:57
kristianpaulyeap21:57
wpwrak_there are several clock rate ranges. the host gets to pick what it likes.21:57
kristianpaulok i'll wire it later, well on fpga i already wired SiGE and ANT cable :)21:59
kristianpaulgn821:59
wpwrak_sweet dreams ! :)21:59
kristianpauli'll dream with clocks for sure ;)21:59
--- Tue Dec 7 201000:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!