#milkymist IRC log for Friday, 2011-08-12

kristianpaulwpwrak: (ram corruption) well, not to alarm, but i remenber first ethe core that uses DMA had some issues with memory in RTEMS, dunno if just was something tricky in software, but lekernel just decided to re-implement as a FIFO using ram from fgpa02:48
kristianpauland yes there are other DMA capable cores, but mostly for video/audio part, so i wonder if ram corruption can be noticed easilly..02:49
kristianpaulanyway, i always tought that, just droping my comments02:50
kristianpaulalso fyi the only flash corruption (as the m1 neverd loaded neither bitstream or bios) was in the time i dint knew it urjtag had a command for reconfigure the fpga. so i can avoid that manual unplug/plug DC power..02:52
wpwrakokay, so you just stopped power cycling ?02:57
kristianpaulsure time ago03:04
kristianpaulat least not so often as i reflash several times at weekend..03:04
wpwrakmaybe, if you were to start power cycling again, you could also reproduce NOR corruption03:07
kristianpaul10 time right? :)03:08
kristianpaulxiangfu: you have script for read nor, just to be sync i can try that powecycle as you did03:09
kristianpauland confirm same issue in rc203:09
kristianpauli guess i need flash flickernoise and render :)03:09
xiangfukristianpaul, this one https://github.com/milkymist/scripts/blob/master/read_flash_m1.sh03:10
wolfspraulkristianpaul: wow, thanks for helping!03:12
kristianpaulwolfspraul: haha, whatervr i can do remotelly :)03:13
kristianpaulxiangfu: wich flicernoise bins, bitstream?03:13
xiangfukristianpaul, wait one second03:13
wpwrakmaybe testing will be faster with the CRC check running on the M1. e.g., like adam's test sw does03:13
xiangfukristianpaul, forward one email to you. how to reflash the m1. and where the files.03:18
xiangfumaybe those steps already wrote down in wiki by aw_ .03:18
kristianpaulxiangfu: jsut in case, you pluged ethernet cable as well?03:27
kristianpaulxiangfu: how are you diffing?03:41
kristianpaulhexdump to txt then diff?03:41
xiangfuyes. hexdump03:41
xiangfuthen diff03:41
kristianpaulxiangfu: for reply you test03:46
kristianpaulpoweron wait to render wait N seconds and powercycle?03:46
kristianpaulokay seven times all OK :)03:47
kristianpauli just waited 3 seconds before powercycle tought..03:47
xiangfumaybe a little longer 10 second03:48
xiangfuseconds :)03:48
Action: kristianpaul goes for a stopwatch03:49
kristianpaul10 times powercycle wait 10 seconds to cut power and try again all OK03:59
kristianpaulaparently :)03:59
kristianpaulnow lets read the standtby bitstream just to wonder ;)04:00
wpwrakcan you run a CRC check on all the partitions ?04:00
wpwraki.e., do whatever magic adam's test does04:00
kristianpauldont know, how i do it?04:01
kristianpaulxiangfu: ? :)04:01
wolfspraulkristianpaul: need to load the test software over the serial console04:01
kristianpauljust that?04:02
kristianpaulsame md5sums :)04:04
kristianpaulok where is that test software binary..04:04
Action: kristianpaul compiles04:04
aw_kristianpaul, use this test image: http://milkymist.org/updates/current/for-rc3/boot.4e53273.bin04:11
kristianpaulaw_: thnks !04:11
kristianpaulstarted getting mad about makefile not happy with is current location..04:12
aw_time to rework 13 pcs usb transceivers after lunch. i updated wiki results.04:12
kristianpaul  i: Mking : standby.fpg : bios-rescue.bin(CRC)CRC passed (got e4cng : splash-rescue.rawsed (got e8ff824f)04:18
kristianpaul  Checking : flickernoise.fbi(rescue)(CRC)CRC )04:18
kristianpaul  Checking : bios.bin(CRC)CRC passed (4)04:18
kristianpaul  Checking : splash.rawCRC passed (got a2m PASSED  ]mages04:18
kristianpaulImages CRC: Checking : standby.fpg8905) Checking : soc-rescue.fpgChecking : bios-rescue.bin(CRC)ed (got aa12a56a) Checking : soc.fpg3a31e737) Checking : bios.bin(CRC)CRC passed (got 86e23684) Checking : splash.rawlow, or hit ENTER to run all tests:04:18
kristianpaulok, sorry messy now04:18
wolfspraulkristianpaul: the key is to try to get your board into the flash/d2 dimly lit/cannot reconfigure problems04:19
wolfspraulthen it gets interesting :-)04:20
kristianpaul(dimly) oh now, i hope not ;)04:20
kristianpaulbtw what is the adam latop OS and libusb version?04:21
kristianpaulis it Fedora 15? :-D04:21
wolfsprauldon't know04:21
kristianpaulworth to ask, as ubuntu/fedora like follow upstream a lot... well libusb..04:22
kristianpauljust in case04:22
xiangfuadam use ubuntu04:23
kristianpaul7.04? :)04:24
wpwrakwolfspraul: (can't reconfigure condition) if the corruption strikes at a random place, then you'd have a 50x higher probability to find it if you check the entire NOR after each cycle, instead of waiting for that little 640 kB partition with the bitstream to get hit ;)05:16
wpwrakoh wait. it's 1.5 MB. well, so 20x :)05:16
wolfspraulchecking the entire nor is too slow now, via jtag05:19
wolfspraul4.5 hours I think05:19
wolfspraula lot of power cycles in those 4.5 hours05:19
wpwrakthe production test sw runs locally, no ?05:19
wolfspraulor maybe run the test software again?05:19
wpwrakyup. or maybe FN can do it, too05:20
wolfspraulthat's a good idea! between each power cycle, add a run of the test sw, loaded via serial05:20
wpwrakis there something like a shell ?05:20
wolfspraulyes I think so - rtems shell05:21
wolfsprauladd crc checks there, and/or in the fn gui05:21
wpwrakthen you could probably run a NOR checker from there05:21
wolfspraulanother good idea05:21
wolfspraula stream of good ideas05:21
wolfspraulnow only Xiangfu needs to type faster05:21
wpwrakthey're kinda easy ;-)05:21
wolfspraulyes definitely, also adding it in the fn gui is a good idea05:22
wolfspraulcould come in handy once the units are in the field05:22
wpwrakthat means, i should get them patented. the more obvious the idea, the more likely someone else will have it too :)05:22
wolfsprauldid you see the anti-Apple boot patent?05:23
wpwrakyeah. boot speed limits strictly enforced by trolls ;-)05:24
wolfspraulif it's 'fast', it's patented05:25
wolfspraul"a method for fast booting a computer system05:25
wolfspraulRead more: http://news.cnet.com/8301-13506_3-20089381-17/apple-sued-over-speedy-mac-os-x-startup/#ixzz1Un09SEPa05:25
wolfsprauloops sorry05:25
wolfspraul"a method for fast booting a computer system05:25
wolfspraulRead more: http://news.cnet.com/8301-13506_3-20089381-17/apple-sued-over-speedy-mac-os-x-startup/#ixzz1Un09SEPa05:25
wpwraki saw it a few days ago on heise05:26
wolfspraulok I take notes - crc checks for all partitions in rtems shell and fn gui05:26
wolfspraulwill do05:26
wolfspraulflickernoise could check on every boot, but then what? beep?05:27
wpwrakif the CRC doesn't cover the whole partition, you will have some means to know the partition size. then you could just check that the rest is 0xffff05:27
wolfspraulI think it should not fall back to the GUI and display an error message, that could do more harm than good05:27
wpwrakmaybe the BIOS could ?05:27
wolfspraulno beeper05:27
wpwrakboot into recovery05:28
wpwrakunless that one is broken too. if yes, beeo noisily for a few times before continuing anyway :)05:28
wolfspraulyes definitely continue05:28
wolfspraulotherwise we may cause more harm than good too easily05:28
wolfspraulanyway, let's add checks slowly and where they make sense. good idea! rtems shell first, and fn gui05:29
wpwrak(continue) yeah. not every corruption affects operation.05:29
wolfsprauland in the meantime, adam could do the power cycle tests with running the test image in between05:29
wpwrakah, and mwalle mentioned that gdb has a faster NOR download than jtag. so that may be an option too05:29
wolfspraulbut only after the full round is completed and we are zooming in on this, not now otherwise we may confuse ourselves05:29
wolfspraulyes, all has been added to xiangfu's todo list :-)05:30
wpwrakmaybe that's the quickest approach. then analyze the data on the PC. would also allow just erasing all the other partitions and never running RTEMS.05:30
wpwrak(and this narrowing the set of things the M1 does between cycles. if the problem suddenly disappears, when we can suspect it was one of the skipped things)05:31
wolfspraulah you mean you want to try to reproduce it with power cycles alone?05:32
wolfspraulso just power cycle 100 times (2 seconds in between), then crc check05:33
wpwrakmaybe check every ten05:34
wpwrakif you examine the whole NOR, the corruption should show up quickly. and once it happens, you want to analyze the single event. if you get multiple corruption events between checks, things get blurry again05:35
wolfspraulok but then setting a baseline of 100% erase is not bad05:38
wpwrakyup, then it would help05:38
wolfsprauland checking the entire area outside of the written partitions for anything non-105:38
wpwraki think you still need the "BIOS"05:38
wolfspraulk gotta run, things are moving05:40
wolfspraulAdam will need at least another day or two for more testing and fixing. and in parallel we improve the test tools, erasing, crc checks, etc.05:40
GitHub20[scripts] xiangfu pushed 2 new commits to master: http://bit.ly/rkFUwp08:46
GitHub20[scripts/master] build: match latest development - Xiangfu Liu08:46
GitHub20[scripts/master] move all shell script file to one folder - Xiangfu Liu08:46
methril_workcongratulations for the qemu lm32 support!! :)11:10
methril_worknow it`s oficial11:10
Fallenouyep :)11:23
kristianpaul9.25 hrs render now crash :)14:28
Fallenousometimes spelling makes a difference :)14:31
xiangfukristianpaul, have you connect the jtag/serial to your PC?14:33
xiangfukristianpaul, oh. NOT14:33
kristianpaullol , yes all fine14:35
kristianpaulbut jtag is connected ;)14:35
xiangfukristianpaul, just finish crc command in flickernoise. but my system have problem open terminal.14:36
kristianpaulbusy port?14:37
lekerneli've run mine for a full week without crashing14:40
lekernelintermittent bugs that only hit once in dozen days with serious consequences are the most pesky ... :(14:41
xiangfufull week that is long time14:41
xiangfulekernel, can I put this 'crc' command to branch 'stable_1.0' then cherry-pick it to 'master' later.14:44
lekernelwhat is that crc command?14:44
xiangfulekernel, just 'crc address length' then output the result14:44
xiangfuvery simple.14:44
xiangfubut Werner think we can setup a automatic test : like power-on, crc test all images. power-off, then do it again and again. until meet the NOR flash bug :)14:45
lekerneljust commit it to master14:46
lekernelor use Werner's test :)14:46
lekernelfaster and does not clutter the flickernoise code base with spurious features14:46
lekernelactually, that crc command might belong in RTEMS instead14:47
xiangfuoh. yes. let me push to flickernoise first :)14:48
lekernelnah, add it to RTEMS14:49
GitHub122[flickernoise] xiangfu pushed 1 new commit to master: http://bit.ly/o6LrU014:49
GitHub122[flickernoise/master] new command: crc address length - Xiangfu Liu14:49
GitHub198[flickernoise] xiangfu pushed 1 new commit to stable_1.0: http://bit.ly/pBuahO14:49
GitHub198[flickernoise/stable_1.0] new command: crc address length - Xiangfu Liu14:49
lekernelwe are following cvs head closely now btw14:50
lekernelthere's just this annoying bug that causes the whole system to lag after rendering has been run14:51
xiangfuoh. reverted.14:51
lekernelwhat I propose is:14:53
lekernel1. implement it in RTEMS14:53
lekernel2. add it to https://github.com/milkymist/rtems-old so it can be used in flickernoise stable-1.0 branch14:53
lekernel3. add it to https://github.com/milkymist/rtems so it can be used in flickernoise master branch14:53
xiangfuok. got it, reboot my system first. I only have one working terminal now :(14:53
lekernel(use the branch mmstaging for the rtems repository)14:54
lekernel4. submit a patch into the RTEMS PR system with the CRC command so it gets merged14:54
xiangfuthen I will work on the lock command in urjtag.14:56
xiangfuI am thinking add one command like "flashlock ADD BLOCK_COUNT"14:56
lekernelit's not supported already?14:56
lekernelsurprising ...14:56
xiangfunot support in urjtag.14:57
xiangfuhave to reboot, read IRC log later14:57
wpwrak(werner's test) hmm, i really have to build that little USB relay switch ...16:36
wpwrakat openmoko, i once made some parallel port to transistor switch, but that was for lower currents. i mention this not because it was a great technical achievement, but because, for some unfathomable reason, nobody else had that idea. pressing buttons gets kinda boring after the 1000th time or so ...16:38
kristianpaul(transistor switch) lovelly:)16:41
mwallewpwrak: mh is a software write protection persistent?20:04
mwalle(nor flashes)20:04
wpwrakyes, it survives power cycling20:05
wpwrakat least the one that's documented. the chip seems to have additional, undocumented protection modes20:05
mwallemh ok20:06
mwallelekernel: do i need ise 13.2 to synthesize the newest mm? or is 13.1 sufficient enough?20:28
lekernelgit head needs 13.2 because it enables the block ram initialization workaround in bitgen20:29
lekernelif you disable this option, 13.1 should work (and the block ram initialization bug doesn't seem to affect the fpgas we have anyway)20:30
mwallecould you point me to the commit/file i have to look at?20:30
mwallefound it :)20:31
lekernelit's just the -g INIT_9K:Yes20:31
mwallewhats 40000 ? lol20:31
mwalles6 errata20:32
mwalleFür das Evaluationsboard LM32 gibt es eine grundlegende Unterstützung, das SoC Milkymist wird vollständig, inklusive Video-Rendering, unterstützt.20:42
mwallelol golem.de20:42
lekernelis it just me, or do you need a user account to _read_ messages from the ubuntu forum?21:32
wpwrakwhere ?21:33
lekernelmaybe the post was just deleted, according to google it dates from Nov 200721:34
mwalleother posts are working for me without registration21:35
wpwraklekernel: btw, what's your intuition about the NOR corruption ? same situation as in rc2 ? better ? something different ?23:16
--- Sat Aug 13 201100:00

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