| 00:01.25 | MonMotha | prpplague: I've found that multiple small vias are often more effective than one larger one |
| 00:01.41 | MonMotha | just have to be careful not to slice up your power/ground planes too badly |
| 00:07.55 | g1powermac_apl | wonders what would be best first to route, power planes or data lines |
| 00:08.05 | g1powermac_apl | its a simple two layer board |
| 00:08.46 | MonMotha | in general, I route signals first then flood the power planes then clean up any parts of the power I don't like by push/shoving |
| 00:09.20 | MonMotha | depending on topology, though, you may have better luck doing power busses first to establish layer direction then routing signals |
| 00:15.51 | prpplague | MonMotha yea, the problem is the EE wont acknowledge to the problem |
| 00:16.18 | prpplague | RobotGuy its more like saying "aint" in the US |
| 00:16.33 | RobotGuy | Oh, OK. |
| 00:18.45 | g1powermac_apl | is going to have humdinger of a time routing this board |
| 00:22.37 | RobotGuy | Which board is that? |
| 00:23.42 | CosmicPenguin | heh /me read hamburger of a time |
| 00:25.09 | MonMotha | g1powermac_apl: PCB routing is something you really just have to keep playing with |
| 00:25.14 | MonMotha | every time I do a board, I learn new tricks |
| 00:25.34 | MonMotha | I rarely use autorouters, though I do get good use out of my CAD software's "interactive" routing |
| 00:50.50 | *** join/#edev linac (n=lin@121.41.36.212) |
| 00:56.37 | g1powermac_apl | sigh, that was fun |
| 00:56.52 | g1powermac_apl | needed to correct a mistake on my schematic and I lost my entire routing |
| 00:57.16 | g1powermac_apl | now figured out the proper way of doing that so he doesn't lose his routing |
| 00:58.08 | T0mW | MonMotha: where do you put your power planes in a multilayer pcb? |
| 00:58.12 | *** join/#edev ce_geek (n=ce_geek@c-24-61-92-197.hsd1.nh.comcast.net) |
| 01:03.41 | GPSFan | T0mW: usually in the middle with additional ones as required once you go over 6-8 layers |
| 01:04.05 | T0mW | GPSFan: any reason for putting them in the middle and not an outer layer? |
| 01:04.52 | T0mW | GPSFan: I would think that the flood of foil that the power planes tend to make would create something of a faraday shield. |
| 01:05.01 | GPSFan | well there are several schools of thought, the best IMHO is tha by putting them in the middle you get the highest inter-layer capacitance; |
| 01:05.12 | T0mW | true |
| 01:05.32 | T0mW | less noise filtering needed (caps) |
| 01:06.07 | T0mW | I would rather put the noisy stuff in the inner layers and keep the FCC happy |
| 01:06.31 | GPSFan | you will be using distributed bypassing in any case, the plane cap just helps. in RF designs there are often ground planes near the HF chips and stripline signals' |
| 01:07.06 | T0mW | RF is a special case |
| 01:07.28 | GPSFan | the other school of thought is put the power planes on the outside and to hell with inter-layer cap, and use the shielding of the plane for FCC |
| 01:08.15 | GPSFan | YMMV ;>) |
| 01:09.29 | T0mW | LOL |
| 01:10.32 | T0mW | not everybody sells to consumer level (B) some people have type A to consider. |
| 01:11.18 | GPSFan | T0mW: I'm glad prpplague got his eabi stuff for arm working, I was concerned that there were issues that hadn't been discovered yet. |
| 01:11.51 | GPSFan | and some have to meet MILSTD-461 and varients. |
| 01:12.14 | GPSFan | or even worse Tempe$t |
| 01:15.23 | T0mW | GPSFan: I would think that in that case, external screening of the board area is needed. |
| 01:15.40 | GPSFan | some really high layer count designs do both, center VCC & GND and multiple shielding layers. but then if you are doing 12+ layers you have lots of $$ and requirements that yoou probably can't talk about. |
| 01:16.45 | T0mW | GPSFan: well, Dave was using an s3c2440 for the eabi. I've not yet gotten that working on the s3c2410. But, then, I have been doing a lot of backtracking in my bitbake setup. |
| 01:17.26 | T0mW | GPSFan: got nailed by using a symlink to my bitbake projects dir: /home/tom/OEstuff --> /speedy/OpenEmbeddedProjects/ |
| 01:17.52 | GPSFan | T0mW: my hat is off to both you & dave for going into the maze of "twisty little passages" that is OE/poky. |
| 01:18.19 | T0mW | GPSFan: all kinds of weird problems disappeared after I mounted that volume (raid-0) at the point where I kept the symlink. |
| 01:19.09 | T0mW | GPSFan: it is getting simpler with OE, just stay the heck away from "org.openembedded.dev" and use the stable branch of "org.openembedded.stable" |
| 01:19.28 | T0mW | "dev" is a nightmare of changes and broken things |
| 01:19.44 | GPSFan | time to go watch a movie, too much wine with dinner......... cul... |
| 01:19.56 | T0mW | GPSFan: nite |
| 01:20.18 | T0mW | watches GPSFan fall asleep in front of TV |
| 01:25.15 | RobotGuy | prpplague: Got a question about getting children from another country whose mother is here in the USA. |
| 01:25.54 | T0mW | have the father fly to NYC, rent a motel room... |
| 01:45.31 | *** join/#edev sunrunner20 (n=jamesdoe@pool-71-170-134-104.dllstx.fios.verizon.net) |
| 01:49.26 | MonMotha | T0mW: I tend to put ground on layer 2 (from top) and power on the 2nd layer from the bottom |
| 01:49.30 | MonMotha | unless I have reasons not to |
| 01:50.27 | MonMotha | I'm told that attempting to put the planes on the outside and using them as a shield isn't very effective |
| 01:50.30 | MonMotha | I've never tried it |
| 01:50.54 | MonMotha | supposedly keeping the ground plane near the top (for a PCB with most high-speed components on the top) helps by keeping loop area as low as possible wrt IOs |
| 01:51.21 | MonMotha | some people suggest putting ground right in the middle of the stack-up if you don't have a dedicated power plane (sometimes happens on 4 layer designs) for similar purposes |
| 01:51.51 | MonMotha | I have heard of designs using lots of buried vias and grounded full floods on the top and bottom and plated sides |
| 01:51.56 | MonMotha | that can supposedly work as a "cage" |
| 01:52.24 | MonMotha | but apparently if you have lots of through-hole vias and open edges like a normal PCB, attempting to use outer planes as shields doesn't help |
| 01:53.47 | MonMotha | most people seem to just suggest dedicated planes as possible on 8 or fewer layer layouts with plenty of high-quality bypass caps as near the power pins as possible |
| 01:53.57 | MonMotha | and "bulk" low ESR tantalums as often as you can fit them in |
| 01:54.58 | MonMotha | also, apparently cheaping out on ceramic bypass caps and using Y5V instead of X7R can have amazingly detrimental effects. I just don't cheap out, so I've not experienced that. |
| 01:56.19 | ds2 | MonMotha: you solved your EMI problem? |
| 01:59.46 | MonMotha | ds2: still coming up with solutions |
| 01:59.52 | MonMotha | I've got some copper foil tape on order |
| 02:00.00 | MonMotha | according to some usenet posts, the micro itself is known to radiate |
| 02:00.10 | MonMotha | so I was going to slap some copper on top and bottom and ground it |
| 02:00.27 | MonMotha | I need 5dB, which is a notable amount, but shields can be pretty effective |
| 02:01.18 | ds2 | heh |
| 02:01.33 | ds2 | if you want it in a hurry, they sell copper foil tape at the garden center (for slug control) |
| 02:01.55 | ds2 | I take it changing the clock to something else didn't do enough? |
| 02:32.34 | MonMotha | ds2: well, I don't have the gear to measure any of this |
| 02:32.42 | MonMotha | so I can't do really quick stuff like that and check |
| 02:35.35 | T0mW | MonMotha: I'm that way: lots of X7R bypass caps as close as possible to the devices, with a liberal sprinkle of 10uF tants |
| 02:36.22 | MonMotha | T0mW: yup |
| 02:36.32 | MonMotha | seems to be pretty effective |
| 02:36.36 | MonMotha | just this one problem across two boards |
| 02:37.07 | T0mW | MonMotha: maybe you will have to put hte micro in a shield-can |
| 02:38.41 | MonMotha | too expensive, so it's a last resort |
| 04:01.30 | *** join/#edev DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 04:29.20 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 05:02.22 | *** join/#edev redblue (i=star@ppp019.108-253-207.mtl.mt.videotron.ca) |
| 05:07.07 | *** join/#edev kajaka (i=dirc@chil-nrp1-cs-1345.vdsl.bright.net) |
| 05:16.00 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 05:20.07 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 05:24.07 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 06:07.01 | *** join/#edev drath_ (i=vmaster@p5B07C56C.dip.t-dialin.net) |
| 06:49.04 | *** join/#edev g1powermac_ (n=g1powerm@unaffiliated/g1powermac) [NETSPLIT VICTIM] |
| 06:49.04 | *** join/#edev T0mW (n=Tom@24.115.218.135.res-cmts.sth.ptd.net) [NETSPLIT VICTIM] |
| 06:49.04 | *** join/#edev chouimat|away (n=dieu@kde/developer/chouinard) [NETSPLIT VICTIM] |
| 06:49.04 | *** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) [NETSPLIT VICTIM] |
| 07:02.02 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 07:23.40 | *** join/#edev [AD]Turbo (n=turbo@host24-241-dynamic.22-79-r.retail.telecomitalia.it) |
| 07:24.36 | [AD]Turbo | hi all |
| 07:40.31 | *** join/#edev Appleman1234 (n=Appleman@131.181.99.244) |
| 07:56.25 | *** join/#edev lyakh (n=lyakh@p57BD21C8.dip0.t-ipconnect.de) |
| 08:08.59 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 08:17.31 | *** join/#edev roxfan (n=dunno@162.214-200-80.adsl-dyn.isp.belgacom.be) |
| 09:27.07 | *** join/#edev linac (n=lin@222.76.91.139) |
| 09:46.11 | *** join/#edev Appleman1234 (n=Appleman@131.181.99.244) |
| 09:48.43 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 10:16.23 | *** join/#edev ali_as (n=ali_as@ambix.plus.com) |
| 10:30.04 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 10:45.24 | *** join/#edev czr_ (n=czr@ebi0.eke.fi) |
| 10:45.40 | czr_ | (OT) does anyone know whether there's a channel for openocd somewhere? |
| 10:52.35 | Crofton | czr_, no, but during US daytime there are some guys here that know about it |
| 10:52.46 | czr_ | Crofton, thanks |
| 10:55.53 | Crofton | they should be around in a couple of hours |
| 10:58.22 | czr_ | I'll continue cursing meanwhile :-), thanks for the hint. |
| 10:58.33 | *** join/#edev methril (n=Methril@213.27.233.98) |
| 10:59.45 | *** join/#edev grummel (n=christia@cl-13.dus-01.de.sixxs.net) |
| 11:02.22 | *** join/#edev ebbe (n=ebbe@users173.kollegienet.dk) |
| 11:47.22 | *** join/#edev g1powermac_apl (i=g1powerm@CPE-76-177-156-202.natcky.res.rr.com) |
| 11:53.40 | *** join/#edev redblue (i=star@ppp228.108-253-207.mtl.mt.videotron.ca) |
| 11:55.50 | *** join/#edev sjhill (n=sjhill@real.realitydiluted.com) |
| 11:55.50 | *** mode/#edev [+o sjhill] by ChanServ |
| 12:15.45 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 12:15.45 | *** mode/#edev [+o GPSFan] by ChanServ |
| 12:16.31 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 12:39.42 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 12:47.51 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 13:03.29 | *** join/#edev chouimat|work (n=dieu@209.217.106.98) |
| 13:05.23 | *** join/#edev linac (n=lin@121.41.36.212) |
| 13:19.37 | *** join/#edev sunrunner20 (n=jamesdoe@pool-71-170-134-104.dllstx.fios.verizon.net) |
| 13:19.59 | T0mW | prpplague: The problem with the SPI MMC errors cleared up with "mem=32M". |
| 13:20.20 | prpplague | T0mW: ?? |
| 13:20.24 | T0mW | prpplague: btw, I have maxspeed = 20000000 |
| 13:20.35 | prpplague | T0mW: oh you were detecting at 64? |
| 13:20.49 | T0mW | I was specifying @ 32M |
| 13:20.55 | T0mW | in the kernel command line |
| 13:21.06 | prpplague | T0mW: you were telling it 32 or telling 64? |
| 13:21.20 | T0mW | I was telling it 64M |
| 13:21.23 | prpplague | ahh |
| 13:21.47 | prpplague | T0mW: i'm going thru a bunch of pcb realestate before the microsd, i suspect i was getting noise |
| 13:22.03 | T0mW | AH, yes |
| 13:22.04 | prpplague | T0mW: 10mhz seems like reasonable performanc for my needs |
| 13:22.19 | T0mW | no no! you MUST go FASTER! |
| 13:22.24 | T0mW | heh |
| 13:22.33 | prpplague | T0mW: the one i'm concerned about is the usb host |
| 13:23.10 | T0mW | yeah, I would be too. You recall the prob I had with the host not detecting the keyboard? |
| 13:23.26 | prpplague | T0mW: yea, you find the problem? |
| 13:23.46 | czr_ | hmm. any suggestions on how to flash a nand chip using jtag of an mcu/soc? |
| 13:23.46 | T0mW | yeah, traces on the PCB too close to each other |
| 13:24.01 | czr_ | (the nand is connected to a flash controlled embedded on the soc) |
| 13:24.08 | czr_ | controller even. |
| 13:24.51 | prpplague | czr_: openocd works find for most nand chipset |
| 13:24.57 | prpplague | s/find/fine |
| 13:25.08 | czr_ | prpplague, I'm having hard time finding any examples on this, any hints where to look for that? |
| 13:25.18 | czr_ | (having hard time with openocd to start with too ;-) |
| 13:25.22 | T0mW | prpplague: I cut the traces out, then use 30ga wwrap wire to jumper to the connector, that greatly improved detection. |
| 13:25.43 | prpplague | czr_: the current openocd trunk has full nand support |
| 13:25.54 | prpplague | czr_: there should be a couple examples |
| 13:26.22 | prpplague | czr_: what SoC are you using? |
| 13:26.45 | czr_ | prpplague, pxa270 |
| 13:27.20 | prpplague | looks |
| 13:28.48 | czr_ | prpplague, any hints of known good revisions of openocd? using the latest but havin't not much luck in getting it to co-operate. |
| 13:29.15 | prpplague | czr_: just a quick look doesn't show any support for the pxa nand controller |
| 13:29.34 | czr_ | it's a programmable generic flash controller. would require setting it up for nand first. |
| 13:29.43 | T0mW | czr_: what are you using to drive the jtag lines? USB wiggler ? parallel port banger? |
| 13:29.55 | czr_ | although I'm desperate enough to start banging at the pins that are connected to the nand soon :-) |
| 13:30.06 | czr_ | T0mW, amontec jtagkey (tiny and normal) at the moment |
| 13:30.19 | T0mW | ok |
| 13:30.31 | chouimat|work | morning |
| 13:30.37 | prpplague | czr_: yea, look at the s3c24xx nand controller setup in openocd |
| 13:30.55 | prpplague | czr_: have you even gotten openocd to talk to the pxa270 yet? |
| 13:31.35 | T0mW | czr_: I tried the USB wiggler from sparkfun which should be the same. It seemed to work ok. That is my "spare" and hasn't been used in a while. |
| 13:31.59 | prpplague | T0mW: you don't use your flyswatter? |
| 13:32.10 | prpplague | T0mW: iirc you have a bdi right? |
| 13:32.33 | czr_ | prpplague, not really no. I'm getting "Warning: jtag.c:1211 jtag_check_value(): value captured during scan didn't pass the requested check: captured: 0x00 check_value: 0x02 check_mask: 0x07" and similar messages, which I'm assuming aren't good |
| 13:32.47 | mbuf | on laptops, SIM cards can be fitted with I2C bus interface? |
| 13:32.56 | czr_ | those lead to errors while writing DCSR -> openocd not really useful. |
| 13:33.04 | T0mW | prpplague: yeah, I have the bdi2000 |
| 13:33.16 | prpplague | czr_: yea, that usually means something is wrong with the communication |
| 13:33.23 | prpplague | czr_: i'd get that working first |
| 13:33.38 | prpplague | mbuf: iirc yes, you can talk to some SIM cards via i2c |
| 13:33.43 | mbuf | prpplague, thanks |
| 13:33.54 | czr_ | prpplague, I'm in the dark here really. where would I even start? I've tried both of the amontecs -> same messages. I've tried another PXA270 (different board) -> same messages. |
| 13:34.32 | T0mW | czr_: try a different computer? |
| 13:35.12 | methril | <czr_> i had some problems connecting between Amontec and an ATMEL ARM9 board |
| 13:35.18 | T0mW | czr_: It might be something stupid like EHCI grabbing the USB device instead of the OHCI host |
| 13:35.39 | czr_ | hmm. I'll do that next, thanks for the hint. |
| 13:35.40 | methril | with openocd, y had to change the JTAGKey name sometimes |
| 13:35.48 | methril | it worked with other boards? |
| 13:35.49 | prpplague | czr_: might start by joining the #openjtag channel , they can spend more time helping you there |
| 13:36.01 | czr_ | prpplague, thanks |
| 13:36.07 | prpplague | np |
| 13:36.18 | T0mW | I have that problem with a flash media reader on USB, the high-speed USB host tries to grab it and it is a USB-1.0 device. |
| 13:37.55 | T0mW | czr_: other than getting the config file setup for my target, my only objection to the USB wiggler was its speed. I did program my target flash, run breakpoints, etc. without problems. |
| 13:38.17 | T0mW | that was an LPC2214 target though, fairly simple SOC |
| 13:39.40 | prpplague | T0mW: generally the error czr_ is seeing is between the jtag device and the soc |
| 13:40.08 | T0mW | speed problem then on the wiggling |
| 13:40.28 | prpplague | yea |
| 13:40.47 | T0mW | prpplague: did xfbdev require inet6 ? |
| 13:41.48 | prpplague | T0mW: when built via poky, yea |
| 13:42.01 | prpplague | T0mW: yea here i found the problem with my segfaults? |
| 13:42.13 | T0mW | prpplague: looks like the same when built with OE |
| 13:42.22 | T0mW | prpplague: something about the backlight driver? |
| 13:43.00 | prpplague | T0mW: yea it was setting one of the registers causing the tft format to change from 565 to 555 |
| 13:43.32 | T0mW | heh |
| 13:45.53 | g1powermac_PB | hmm |
| 13:45.55 | T0mW | prpplague: well, here is the deal: got rid of the symlink in the path to my OE project, used angstrom-2007.1 as the distro, and, the stable branch of OE (org.openembedded.stable). Now gpe builds fine. |
| 13:46.04 | g1powermac_PB | my printer doesn't print to exact scale |
| 13:46.11 | g1powermac_PB | not good when printing pcbs |
| 13:46.32 | prpplague | T0mW: interesting |
| 13:46.41 | T0mW | g1powermac_: that might be a funtion of your CAD package rather than the printer? |
| 13:47.06 | prpplague | looks like landley is after another busybox gpl violator |
| 13:47.10 | T0mW | prpplague: I have had a dearth of errors with OE and it looks like the symlink was the biggest culprit |
| 13:47.23 | prpplague | T0mW: i'm sticking with poky |
| 13:47.25 | T0mW | s/dearth/wealth/ |
| 13:47.33 | prpplague | T0mW: it atleast has some sanity |
| 13:47.50 | g1powermac_PB | T0mW: can't be, I exported to pdf (printed from the pdf reader but isn't correct) and to gerber files (the gerber viewer shows the correct sizes, but prints wrong) |
| 13:48.26 | T0mW | prpplague: I went up to the house yesterday to turn off the main breaker so the electric company could connect the power. I met the current owners. |
| 13:48.28 | g1powermac_PB | both prints have the same sizing issue, even though its printed from two different programs |
| 13:48.40 | prpplague | T0mW: oh? |
| 13:48.55 | g1powermac_PB | and the gerber viewer is showing the correct sizes on the screen |
| 13:49.05 | T0mW | prpplague: I was hoping that I would meet them. They told me stories about the place and showed me around the property for the history of things. |
| 13:49.15 | prpplague | T0mW: thats cool |
| 13:49.51 | T0mW | prpplague: we got along fine. I told them I sympathized with them losing the house, that I had had that happen once to me. |
| 13:49.53 | prpplague | T0mW: hehe, when you get all settled, i'll have to take a road trip and stop by with some brew |
| 13:50.18 | T0mW | prpplague: sure, I plan to keep a guest room |
| 13:50.18 | prpplague | T0mW: ahh it was foreclosed? |
| 13:50.36 | T0mW | prpplague: in the process of settlement to head off foreclosure. |
| 13:50.40 | prpplague | ahh |
| 13:51.27 | T0mW | prpplague: dave, I can now invite people into my home and entertain them without their having to stare at my workshop in the living room! |
| 13:51.28 | prpplague | T0mW: i've rebuilt poky 2 times a day for the last 8 days |
| 13:51.39 | prpplague | (8 working days) |
| 13:52.22 | T0mW | a scratch build of gpe-image (3504 packages) on a Q6600 w/8G RAM takes 1hour 38 minutes. |
| 13:53.10 | chouimat|work | needs more RAM |
| 13:53.11 | T0mW | prpplague: I keep the same sources file between projects so that I don't have to continually download them. |
| 13:53.20 | T0mW | s/file/dir |
| 13:53.59 | T0mW | prpplague: "bitbake -c rebuild package_name" |
| 13:54.29 | T0mW | prpplague: if you add a package: "touch build/conf/local.conf", it will force a rescan of the package build list |
| 13:55.55 | T0mW | prpplague: a bit clumsy, but if you are tweaking a package source, try running the scripts in the temp dir of that package (e.g. run.do_compile, run.do_staging, ...) |
| 14:01.19 | *** join/#edev chouimat|work (n=dieu@209.217.106.98) |
| 14:02.02 | prpplague | T0mW: yea i found the shared source dir option |
| 14:02.26 | prpplague | T0mW: ahh so bitbake checks the local.conf for changes? |
| 14:02.51 | prpplague | T0mW: i've scaled the poky build down to around 2500 tasks |
| 14:02.56 | czr_ | T0mW, I only need to flash the boot loader, it can load the rest of the stuff over ethernet, so it's not a biggie. |
| 14:03.03 | prpplague | T0mW: and with shared source it seems to build pretty good |
| 14:03.05 | czr_ | I just want the damn thing to work, speed is not an issue.. |
| 14:03.26 | prpplague | czr_: this custom hardware? |
| 14:03.54 | czr_ | prpplague, the CPU board is on a SO-DIMM, comes from toradex, the carrier board is ours. |
| 14:03.54 | T0mW | prpplague: correct, similar to a make test for source (local.conf) being newer than the object (package list cache). |
| 14:04.08 | czr_ | prpplague, but toradex doesn't want to support linux. |
| 14:04.20 | T0mW | HAL |
| 14:04.23 | T0mW | HA! |
| 14:04.26 | prpplague | czr_: toradex doesn't have any windows utils to flashing? |
| 14:04.49 | T0mW | czr_: I told a customer that! |
| 14:04.52 | czr_ | prpplague, yes. it doesn't work though. the sw doesn't detect the board at any point |
| 14:05.11 | czr_ | T0mW, told what? :-) |
| 14:05.56 | T0mW | czr_: that toradex was windows only, but the customer insisted that they could get linux sources from them. I told them that it was a 2.4 kernel, IIRC. |
| 14:06.00 | czr_ | toradex used to support linux, but they're quite small company, and non-ce stuff is only 10-15% of their boards, so they dropped linux. |
| 14:06.10 | prpplague | czr_: if it doesn't detect the board, that sounds consistant with the errors from openocd |
| 14:06.17 | czr_ | T0mW, yeah, the 2.2 version of their bsp is "floating around", but it's pretty bad. |
| 14:06.35 | *** part/#edev ebbe (n=ebbe@users173.kollegienet.dk) |
| 14:06.39 | T0mW | little chips was that way, they gave out a 2.6.13 kernel source |
| 14:07.01 | czr_ | prpplague, indeed. what I'm thinking doing next is a parallel port <-> jtag-header adapter. there's a simple setup that toradex says should work (there is a buffer on the eval-board side already) |
| 14:07.02 | prpplague | T0mW: which was a hack from mbuf iirc |
| 14:07.30 | T0mW | yeah, on 2.6.16, which is junk. at least 2.6.17 would be usable. |
| 14:07.54 | prpplague | czr_: i think you are going to find that there is probably something wrong with your carrier board |
| 14:08.01 | czr_ | I have another board with pxa270 here as well. from compulab (.il). they have their own "custom & proprietary" boot manager on NOR, which contains the flashing functions for the NAND. |
| 14:08.18 | czr_ | prpplague, well, using toradex eval board still :-) |
| 14:08.53 | prpplague | czr_: ok, so using the toradex eval board, and their windows util, you were still unable to flash the device? |
| 14:09.47 | czr_ | prpplague, yes. |
| 14:09.58 | czr_ | so I'm thinking there's something odd with the amontecs. |
| 14:10.06 | prpplague | czr_: uh, wait |
| 14:10.14 | czr_ | although, the windows in question is "vista". so. well.. |
| 14:10.27 | prpplague | czr_: are you using the hardware that toradex recommends for usage with their windows flash utility? |
| 14:10.33 | czr_ | prpplague, yes. |
| 14:10.44 | czr_ | although they give amontec as one of the options that they recommend. |
| 14:11.03 | czr_ | doing a parallel port adapter next, from their schema. maybe that will work better. |
| 14:11.12 | prpplague | czr_: then it kinda sounds like you need to be calling toradex for support, because it sounds like they have a problem |
| 14:11.50 | czr_ | prpplague, yeah. will do that after testing the parallel thingy. this being my first project with jtag, it's kind of uh oh bang head against the wall - action so far :-) |
| 14:12.14 | prpplague | yea jtag isn't a four letter word by accident |
| 14:12.20 | czr_ | hehe |
| 14:12.49 | czr_ | also the xscale jtag stuff is pretty scary. not that plain jtag isn't. |
| 14:14.16 | prpplague | wonders when cosmic will show up today |
| 14:14.34 | T0mW | Fatal server error: |
| 14:14.34 | T0mW | no screens found |
| 14:14.36 | T0mW | :( |
| 14:15.15 | czr_ | T0mW, apt-get install screen? |
| 14:15.18 | czr_ | hides & runs |
| 14:15.33 | T0mW | heh |
| 14:19.02 | T0mW | I'm stumped, what do I do about this: "wrong resolution or depth 640x480 at 16 bpp" ?? |
| 14:19.21 | T0mW | I have a 12bpp screen, fbset reports the correct values |
| 14:20.15 | T0mW | there doesn't seem to be a configuration file for xorg-kdrive |
| 14:20.16 | prpplague | T0mW: did you try starting it with the -screen option |
| 14:20.33 | prpplague | T0mW: Xfbdev -screen 640x480@0x12 |
| 14:20.53 | T0mW | Xfbdev -screen 640x480@90 |
| 14:21.29 | T0mW | prpplague: heh, Segmentation fault |
| 14:21.35 | prpplague | why 90? |
| 14:22.01 | T0mW | not sure, It looked like a nice number, sort of roundy |
| 14:22.18 | prpplague | T0mW: just try it with: Xfbdev 640x480@0x12 |
| 14:22.24 | prpplague | T0mW: just try it with: Xfbdev -screen \640x480@0x12 |
| 14:22.25 | T0mW | it seg faults |
| 14:22.27 | prpplague | argh |
| 14:22.30 | prpplague | T0mW: just try it with: Xfbdev -screen 640x480@0x12 |
| 14:22.49 | prpplague | T0mW: sounds like maybe the xfbdev code doesn't supprot 12bpp |
| 14:23.01 | T0mW | sigh |
| 14:23.20 | prpplague | T0mW: might try running at 8bpp for testing, thats much more tested |
| 14:27.51 | g1powermac_apl | T0mW, yup, it definitely isn't from my cad software |
| 14:28.06 | g1powermac_apl | I made a 100mm by 100mm square in gimp and printed it out |
| 14:28.24 | g1powermac_apl | its printed size was like 97mm square |
| 14:29.56 | T0mW | g1powermac_apl: I use an HP 4315 and that prints to scale with EagleCAD. |
| 14:30.12 | T0mW | g1powermac_apl: so, it is possible. |
| 14:30.55 | g1powermac_apl | yea |
| 14:31.05 | g1powermac_apl | just not with this printer and/or printer drivers for linux |
| 14:31.23 | g1powermac_apl | which I've been having issues with in other areas |
| 14:31.48 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 14:31.48 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 14:36.01 | g1powermac_apl | heh, I figured it out |
| 14:36.27 | g1powermac_apl | it was set to 'shrink if necessary' and even though it didn't need to shrink it, it still did a little |
| 14:36.33 | g1powermac_apl | once I set it to crop, it worked |
| 14:37.45 | CosmicPenguin | your board has a phoenix bios? |
| 14:37.46 | CosmicPenguin | shew |
| 14:37.52 | prpplague | yea |
| 14:37.53 | CosmicPenguin | I went into the wrong room too... :) |
| 14:37.57 | prpplague | hehe |
| 14:38.19 | CosmicPenguin | I don't think I've ever encountered a geode with Phoenix on it |
| 14:38.23 | CosmicPenguin | its a GX/LX? |
| 14:38.27 | prpplague | not my project, but i'm keeping an eye on the issue |
| 14:38.38 | prpplague | let me check iirc its GX |
| 14:39.17 | CosmicPenguin | I would say the odds are pretty slim |
| 14:40.03 | prpplague | yea gx |
| 14:40.17 | prpplague | hmm, from what i've been reading , alot of people have purchases the tool |
| 14:40.29 | prpplague | but i can't seem to find where to purchase it |
| 14:50.14 | prpplague | waits to see if his post to buildroot mailing list about the switch from svn to git starts a flame war |
| 14:54.44 | sjhill | prpplague: i just posted my opinion |
| 14:54.48 | sjhill | i vote no |
| 14:54.52 | prpplague | sjhill: same as i |
| 14:55.23 | prpplague | reads sjhill's post |
| 14:55.31 | prpplague | sjhill: did my response show up? |
| 14:55.36 | chouimat|work | hi sjhill |
| 14:55.38 | CosmicPenguin | but git is so much easier to develop against |
| 14:56.12 | prpplague | CosmicPenguin: i tend to disagree |
| 14:56.24 | prpplague | CosmicPenguin: everyone has their own needs |
| 14:56.31 | CosmicPenguin | That is true |
| 14:56.50 | CosmicPenguin | Its just so much easier to develop patches in git |
| 14:56.55 | prpplague | CosmicPenguin: sjhill and i are in the realm of heavy commercial needs, and git just doesn't do it |
| 14:56.58 | CosmicPenguin | and figure out when stuff goes south |
| 14:57.16 | CosmicPenguin | prpplague: don't give me that - I am a thousand times more commercial then you are |
| 14:57.19 | CosmicPenguin | probably not sjhill |
| 14:58.03 | prpplague | CosmicPenguin: ok, to argue the point, doesn't OE still use monotone? |
| 14:58.24 | chouimat|work | not this git is so pretty and everything else suck discussion again :) |
| 14:58.38 | prpplague | chouimat|work: indeed |
| 14:58.45 | CosmicPenguin | There is no accounting for taste in RCSs |
| 14:59.22 | CosmicPenguin | I have cvs, svn, monotone, hg, and git trees |
| 14:59.22 | prpplague | CosmicPenguin: hehe agreed |
| 14:59.25 | CosmicPenguin | That doesn't make one right or the others wrong |
| 14:59.29 | chouimat|work | prpplague: sometime I have the impression that git is great because it was written by linus :) we still use CVS here and we are happy :) |
| 14:59.32 | prpplague | CosmicPenguin: agreed |
| 14:59.47 | CosmicPenguin | All I said was that GIT is much easier to develop patches gainst, and to bisect issues |
| 14:59.58 | CosmicPenguin | and seeing as though all of my work involves patches, that seems like a logical fit |
| 15:00.12 | prpplague | CosmicPenguin: i can understand that |
| 15:00.25 | chouimat|work | CosmicPenguin: it had some interesting features but I find it too complicated for me to use |
| 15:01.04 | CosmicPenguin | cranes are complicated, but I think I would probably want to learn how to use one if I was in the construction business |
| 15:01.34 | prpplague | CosmicPenguin: i just don't see any reason to switch buildroot to git, simply for the fact that other developers are "sneering" at the usage of svn instead of git |
| 15:01.57 | prpplague | CosmicPenguin: indeed |
| 15:02.06 | prpplague | CosmicPenguin: i can use git and do the things i need to do |
| 15:02.13 | CosmicPenguin | I'm not always an advocate of switching from a long standing RCS |
| 15:02.28 | CosmicPenguin | especially if the current one is not completely braindead |
| 15:04.32 | prpplague | CosmicPenguin: if you are going to play ball, i.e. with the kernel, you have to play by their rules |
| 15:05.00 | g1powermac_apl | prpplague, interesting, I'm looking at the datasheet for the bluetooth module |
| 15:05.08 | prpplague | CosmicPenguin: thing is, sjhill and i are large contributors and users of buildroot, we prefer svn, our opinion should count |
| 15:05.27 | g1powermac_apl | prpplague, why is it labeled for BlueRadios yet the module is from Mitsumi? |
| 15:05.47 | prpplague | g1powermac_apl: hehe that module is a three layered thing |
| 15:05.52 | CosmicPenguin | prpplague: for sure |
| 15:06.02 | CosmicPenguin | I thought you were the maintainer though? |
| 15:06.03 | prpplague | g1powermac_apl: the actual radio is mitsumi, but contains no software |
| 15:06.09 | prpplague | CosmicPenguin: i am |
| 15:06.10 | CosmicPenguin | if you don't want to switch, then don't switch |
| 15:06.17 | CosmicPenguin | tell em to take a hike |
| 15:06.18 | prpplague | g1powermac_apl: the software is from blueradios |
| 15:06.22 | g1powermac_apl | ahhh |
| 15:06.32 | prpplague | g1powermac_: and the pcb is from sparkfun |
| 15:07.00 | g1powermac_apl | the first and third parts I figured, just didn't know where this blueradios part fit in :-) |
| 15:07.20 | prpplague | g1powermac_apl: pretty much all bt modules come in two parts |
| 15:07.23 | g1powermac_apl | ok, so the reset on this module is active low, that changes my schematic a bit |
| 15:07.28 | prpplague | g1powermac_apl: the hardware and a software stack |
| 15:07.59 | g1powermac_apl | had the bt and the ds89c450 resets hooked together, which is not a good thing it seems |
| 15:08.08 | prpplague | CosmicPenguin: yea, being polite first |
| 15:08.10 | g1powermac_apl | as the ds89c450's reset is active high |
| 15:10.01 | g1powermac_apl | oh bugger |
| 15:10.07 | g1powermac_apl | the bt module is 3.3v |
| 15:11.49 | g1powermac_apl | ponders if he should even keep the bt module on this board now |
| 15:12.05 | g1powermac_apl | too much of a hassle to level shift it |
| 15:16.48 | prpplague | g1powermac_apl: the ds is +5v? |
| 15:18.46 | g1powermac_apl | yup, the ds89c450 is 5v |
| 15:20.30 | prpplague | g1powermac_apl: look at the schematic on sfe for the smirf stuff, you can do a two wire rs-232 with a transistor and a couple of resistors |
| 15:22.04 | g1powermac_apl | k |
| 15:27.06 | prpplague | CosmicPenguin: my poky builds are running alot better |
| 15:27.16 | prpplague | CosmicPenguin: found alot of cruft to remove |
| 15:27.43 | CosmicPenguin | thats totally not fair |
| 15:27.46 | CosmicPenguin | it was running fine before |
| 15:28.05 | CosmicPenguin | its not like you discovered a bunch of dirt in your vaccum cleaner that was clogging the suction or something |
| 15:28.47 | prpplague | rephrases |
| 15:29.00 | prpplague | CosmicPenguin: my poky builds are running alot better for my needs |
| 15:29.03 | CosmicPenguin | ding |
| 15:29.05 | CosmicPenguin | good |
| 15:29.40 | prpplague | CosmicPenguin: got the total tasks down to around 2000 |
| 15:30.00 | CosmicPenguin | cool |
| 15:31.43 | prpplague | sjhill: kind of sounds like there might be some dev work for buildroot in the future |
| 15:33.28 | prpplague | CosmicPenguin: i still need to cut it down more, but for this demo the 2000 tasks should be ok |
| 15:34.08 | CosmicPenguin | whatever gives you an image you can be happy with |
| 15:34.56 | prpplague | CosmicPenguin: yea, mostly its a time factor |
| 15:35.27 | prpplague | CosmicPenguin: doing a fresh build takes so long, i need to have it down small enough that it can be done in a few hours |
| 15:37.02 | *** mode/#edev [+o prpplague] by ChanServ |
| 15:38.00 | solar | prpplague: by chance do you happen to have a static eabi aware busybox handy? |
| 15:38.30 | prpplague | solar: sorry no, i'm right in the middle of 3 eabi builds now |
| 15:38.42 | prpplague | solar: none static |
| 15:39.43 | prpplague | solar: still hacking on the nail? |
| 15:39.56 | solar | it's ok. I'm just fighting with why my init is not found. |
| 15:40.08 | solar | wanted to confirm something. |
| 15:42.18 | prpplague | solar: okie dokie |
| 15:44.40 | solar | this better work this time. |
| 15:44.47 | prpplague | CosmicPenguin: you set the nice on any of your poky builds? |
| 15:44.53 | prpplague | solar: hehe |
| 15:44.58 | CosmicPenguin | prpplague: nah |
| 15:48.50 | prpplague | CosmicPenguin: you are using poky for just the xo or some other AMD projects? |
| 15:50.40 | T0mW | it looks like my problem is down inside the s3c2410fb kernel driver, weird values in mach_info->display array |
| 15:51.09 | T0mW | width=8216, height=58772, ... |
| 15:52.12 | solar | there should of been no reason this should of just failed. |
| 15:53.00 | prpplague | T0mW: uh thats weird |
| 15:53.02 | T0mW | solar: thanks for spotting that 64M error in the ramsize. That cleared up the SPI MMC problem. |
| 15:53.13 | prpplague | T0mW: are you setting some sort of virtual window? |
| 15:53.31 | T0mW | prpplague: yeah, I will have to compile kernel, again, without -O2 and start tracing. |
| 15:53.39 | solar | T0mW: yeah I think that might of been an error in the default hammer config. |
| 15:53.48 | prpplague | T0mW: can you pastebin your machine file? |
| 15:53.58 | T0mW | ok |
| 15:54.21 | prpplague | solar: yea i think its in the default config, but that option is ignored as it is passed from the bootloader |
| 15:54.35 | solar | T0mW: here is what I've been using with pretty great luck. http://dev.gentoo.org/~solar/embedded/tct/linux-2.6.22.config |
| 15:54.39 | prpplague | solar: i'll make the change in the config so as to relieve any confusing |
| 15:55.14 | T0mW | prpplague: http://rafb.net/p/pN6enM56.html |
| 15:55.21 | prpplague | looks |
| 15:55.32 | solar | prpplague: it's ok. You can just start putting 64M on the boards :) |
| 15:56.13 | T0mW | prpplague: hmm, maybe tct_hammer_fb_info needs to be an array? |
| 15:56.48 | prpplague | T0mW: need to remove the __initdata flags |
| 15:56.55 | T0mW | yeah, just saw that |
| 15:57.23 | T0mW | prpplague: you think tct_hammer_lcd_cfg needs to be an array? |
| 15:58.27 | prpplague | T0mW: that looks good just need to remove the __initdata flags, problem is that after the kernel finishes the boot sequence, the fb data is dropped |
| 15:58.45 | T0mW | yup |
| 15:58.47 | prpplague | T0mW: i'd test it with that change first |
| 16:13.17 | T0mW | prpplague: well, that took care of the screen resolution, now I'm down to a segfault with Xfbdev |
| 16:16.58 | solar | bangs head with a nail. this should work. http://rafb.net/p/BfwEV725.html |
| 16:18.44 | agaffney | solar: kernel commandline? got an initramfs somewhere we can peek at it? |
| 16:19.19 | solar | sure I'll make a complete tarball |
| 16:23.10 | solar | http://dev.gentoo.org/~solar/embedded/Nail-eabi.tar.bz2 ; ./flashme |
| 16:31.54 | prpplague | solar: have you got all of the boot log? |
| 16:33.06 | GPSFan | sjhill: "As a maintainer and contributer, I recommend we end this thread and get back to worry about doing releases of buildroot." ..... well said.. |
| 16:34.12 | sjhill | GPSFan: personally i think doing actual releases is great, i just did not have time to do it |
| 16:34.28 | solar | not anymore I disconnected. But the point where it stops is exactly where it's spawning init. Thats with the static one. The non static would simply give me a vfs panic with unable to spawn init. |
| 16:36.39 | prpplague | solar: you can generally get that type of issue when using oabi and have hardfloat code with no fpu emulation in the kernel |
| 16:36.59 | prpplague | solar: thats just an example |
| 16:37.06 | GPSFan | sjhill: indeed, and tangents like changing from svn are counter productive. |
| 16:37.11 | solar | I think I'll have to try with eabi with obai-compat; use the obabi to boot. nfs mount something and confirm the userland works |
| 16:37.32 | solar | nfs mount then chroot into it |
| 16:38.16 | prpplague | solar: i'm guessing something isn't matched up between your eabi and float operations |
| 16:40.07 | solar | prpplague: in userland? |
| 16:40.50 | prpplague | solar: one sec, downloading your tarball for a look |
| 16:41.03 | solar | prpplague: actually can you also post a kernel.config from your working eabi setup? |
| 16:41.40 | prpplague | solar: right now i'm not testing on the hammer, but i can do so |
| 16:47.08 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 16:47.40 | prpplague | solar: hmm, i don't see anything that jumps out as an issue |
| 16:47.53 | prpplague | solar: readelf shows your busybox to be eabi |
| 16:55.41 | solar | yeah |
| 17:36.44 | prpplague | hmm, can someone check http://www.arm.linux.org.uk |
| 17:37.39 | g1powermac_apl | seems kinda dead here |
| 17:38.50 | prpplague | okie dokie |
| 17:40.44 | g1powermac_apl | prpplague, so I ditched the bt module on this board for now |
| 17:41.06 | g1powermac_apl | figured it would just simplify things, and I'll use the bt module for something else later on |
| 17:41.12 | prpplague | okie dokie |
| 17:41.12 | g1powermac_apl | maybe for the zipit :-) |
| 17:41.36 | g1powermac_apl | this whole board thing is really just practice for myself |
| 17:41.50 | g1powermac_apl | need to prepare for the stuff I want to make for the zipit |
| 17:56.59 | *** join/#edev bbradley (n=bbradley@92-235-133-146.cable.ubr17.azte.blueyonder.co.uk) |
| 18:03.02 | *** join/#edev Redhatter (n=vk4fsjl@2001:388:f000:0:0:0:0:279) |
| 18:15.23 | *** join/#edev redblue (i=star@ppp044.109-253-207.mtl.mt.videotron.ca) |
| 18:22.45 | *** join/#edev KeyserSoze (n=Miles@67.107.206.170.ptr.us.xo.net) |
| 18:27.39 | *** join/#edev roxfan2 (n=dunno@129.0-200-80.adsl-dyn.isp.belgacom.be) |
| 18:28.22 | prpplague | CosmicPenguin: there a good gx based module you recommend? |
| 18:28.43 | prpplague | CosmicPenguin: i.e. if we wanted to do all the i/o portion of the board |
| 18:28.58 | CosmicPenguin | you don't want the southbridge either? |
| 18:30.33 | prpplague | CosmicPenguin: most or less the entire system on a module, minus the i/o connectors |
| 18:30.52 | CosmicPenguin | its a two chip solution |
| 18:30.55 | CosmicPenguin | so I think you lose |
| 18:31.05 | prpplague | ? |
| 18:33.03 | prpplague | CosmicPenguin: http://www.linuxdevices.com/news/NS8804345583.html |
| 18:33.15 | prpplague | CosmicPenguin: etx module i guess is the standard |
| 18:33.45 | CosmicPenguin | oh, thats what you want |
| 18:34.20 | prpplague | yea something like that |
| 18:34.29 | prpplague | where we can just make a board with the i/o connectors |
| 18:36.49 | CosmicPenguin | http://wwwd.amd.com/catalog/salescat.nsf/shop?openform |
| 18:37.15 | ds2 | wow prpplague is going x86?! |
| 18:37.37 | CosmicPenguin | Mabye I'll be able to talk him out of it |
| 18:37.46 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 18:37.48 | prpplague | ds2: no, just getting ready incase i have to save a project |
| 18:38.01 | ds2 | oh heh I see |
| 18:38.29 | ds2 | long time ago I asked for quotes and that sort of stuff and most places want you to buy the dev kit at 10x the cost of a board before they will sell you a board :( |
| 18:38.50 | KeyserSoze | anyone have a good link for JTAG? |
| 18:38.54 | CosmicPenguin | its a brave new world |
| 18:39.02 | RobotGuy | thinks most dev kits are way over priced. |
| 18:39.18 | KeyserSoze | i am going to run a microcontroller off of 3.0V (instead of 3.3V), and am wondering if I need level translation on the JTAG lines |
| 18:39.35 | ds2 | CosmicPenguin: are you saying that is no longer the case? |
| 18:39.53 | KeyserSoze | it'd suck to have to put in a 3.3V regulator and level translator, just for JTAG. |
| 18:40.02 | RobotGuy | KeyserSoze: Is that within the specs of the microcontroller? |
| 18:40.14 | KeyserSoze | yeah, the uC runs off of 2.7 to 3.6V |
| 18:40.14 | CosmicPenguin | ds2: heck no |
| 18:40.18 | prpplague | KeyserSoze: what kind of info about jtag do you need? |
| 18:40.25 | KeyserSoze | it's a SiLabs (formerly Cygnal) 8051 |
| 18:40.26 | CosmicPenguin | ds2: but in their defense, they only build about a thousand of the things at a time |
| 18:40.30 | CosmicPenguin | they are as one off as you can get |
| 18:40.51 | KeyserSoze | their docs don't give a recommended schematic for JTAG. All they have is the schematic for the eval board, which runs at 3.3V |
| 18:40.59 | ds2 | CosmicPenguin: yes but if I am not asking for support, I don't see why I have to buy their dev kit esp. if they are windows focused |
| 18:41.07 | KeyserSoze | and the data sheet for the USB JTAG programmer I got from them is pretty vague |
| 18:41.23 | CosmicPenguin | ds2: I agree you shouldn't have to buy the dev kit |
| 18:41.23 | ds2 | KeyserSoze: I think some of the JTAG things will adapt to the voltage... IIRC, flyswatter is like that |
| 18:41.27 | RobotGuy | CosmicPenguin: Yeah, I know. I don't have to like it though. It just cuts out people like me from having access to a lot of interesting technology. |
| 18:41.30 | CosmicPenguin | though most people do, so why not grub some cash |
| 18:42.04 | KeyserSoze | ds2: yeah, i read about another jtag programmer that expects a pull up to Vdd on one of the JTAG lines, so it can sense it and provide signals of appropriate voltage to the device |
| 18:42.19 | KeyserSoze | but, it's on a line that the eval board's schematic has no pull up. |
| 18:42.33 | ds2 | *shrug* |
| 18:42.37 | KeyserSoze | which brings up the whole other issue, of what size pull up (or down!) to use on each JTAG line |
| 18:42.46 | KeyserSoze | i've seen so much conflicting information :( |
| 18:44.19 | prpplague | KeyserSoze: the flyswatter can do 3.3v of course, but you are going to need to have some software to drive it |
| 18:44.28 | prpplague | KeyserSoze: as openocd current is for arm device |
| 18:44.54 | KeyserSoze | is using a different programmer for his 8051. a SiLabs "USB Debug Adapter" |
| 18:44.56 | prpplague | KeyserSoze: jtag is an IEEE standard, what do you expect? |
| 18:45.19 | KeyserSoze | I'm using a Olimex ARM USB Tiny for my ARM, and I've got similar questions about pull up/downs on that board |
| 18:45.34 | KeyserSoze | ST Micro says that NO pull up/downs are needed. |
| 18:45.45 | KeyserSoze | But the eval board I have has pull ups on all JTAG lines :( |
| 18:45.52 | KeyserSoze | prpplague: yeah, it sucks |
| 18:45.52 | ds2 | I thought the voltages are configured by the software driving the JTAG? |
| 18:46.11 | KeyserSoze | from the PC (or host, or whatever you want to call it)? |
| 18:48.23 | KeyserSoze | i ought to just power the 8051 off 3.3V, then I wouldn't have so many questions. But there's a RF attenuator that needs 3.0V, and the whole board if very space constrained, so it'd be nice to drive that directly from the 8051. |
| 18:49.52 | ds2 | I don't think it will be a with the 3.0V 8051 |
| 18:50.16 | ds2 | besides 3.0V vs 3.3V is less then one silicon diode drop so it should be w/i the max ratings |
| 18:50.46 | prpplague | KeyserSoze: yea, jtag is one the most abused standards |
| 18:51.12 | prpplague | KeyserSoze: best you can hope for is to replicate what someone has already got working as part of a reference design |
| 18:52.00 | KeyserSoze | prpplague: i guess i'll see what SiLabs has to say. I really wish they had an appnote or something |
| 18:52.04 | KeyserSoze | thanks guys |
| 18:52.24 | prpplague | KeyserSoze: if you do find out some good info, please post it on the elinux.org wiki |
| 18:53.57 | prpplague | solar: you get your eabi build booting? |
| 18:54.03 | prpplague | solar: i'm having a similiar problem now |
| 18:56.53 | ds2 | I wonder if the SiLabs toolstick will handle it for you |
| 19:00.13 | prpplague | solar: i just a did a fresh build and mine won't boot either |
| 19:02.46 | solar | prpplague: I had to head into work. But it was a no-go for me. |
| 19:02.59 | solar | I did not see any extra VFP stuff unless I enabled OABI compat. |
| 19:03.10 | solar | and did not have to test that before having to head to work. |
| 19:03.18 | prpplague | solar: yea, normally with eabi it shouldn't show you the vfp stuff |
| 19:03.50 | solar | nor with the plain old oabi stuff. |
| 19:04.23 | solar | only with compat. So thats a little new to me. |
| 19:04.40 | solar | there was some eabi patches for uclibc btw. |
| 19:04.50 | prpplague | i'm running glibc |
| 19:04.56 | solar | not sure what the status of those these days might be |
| 19:05.13 | solar | glibc I'd only prefer if I'm having to deal with X. |
| 19:05.42 | solar | cuz uclibc + locales is a pita. And way to many X related programs falsely assume gettext + iconv + locales support |
| 19:05.55 | ds2 | yuck iconv |
| 19:06.10 | ds2 | more reason to write things with just the athena widgets ;) |
| 19:06.45 | prpplague | solar: yea, i'm having to deal with x |
| 20:05.21 | ds2 | 9 |
| 20:28.37 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 21:17.50 | ds2 | for people that have presented talks, is there a good web page on what is expected in an abstract? |
| 21:19.05 | CosmicPenguin | I dunno - examples are always good |
| 21:19.27 | ds2 | where can I find an example? the only "abstract" I am familiar with are for academic papers |
| 21:24.29 | CosmicPenguin | what would be the difference? |
| 21:24.32 | CosmicPenguin | I would go the academic route |
| 21:24.38 | CosmicPenguin | check the OLS papers, as an eample |
| 21:25.42 | ds2 | donno, I am trying find out if there is |
| 21:26.26 | ds2 | cool, that seems like a good example. Thanks |
| 21:32.01 | *** join/#edev landley (n=landley@cpe-70-116-30-223.austin.res.rr.com) |
| 21:32.01 | *** mode/#edev [+o landley] by ChanServ |
| 21:32.20 | solar | landley is here to say he finally found the hardware> |
| 21:32.26 | solar | yes? |
| 21:32.42 | landley | um, haven't looked since yesterday. |
| 21:32.55 | landley | Currently converting all the _defconfig files in the kernel into _miniconfig files. |
| 21:33.06 | landley | Found three separate bugs. |
| 21:33.21 | landley | All of which pretty much boil down to "kconfig is weird and sort of broken"... |
| 21:33.44 | landley | I now have miniconfig files that need a "# BLAH is not set" line in them. |
| 21:33.49 | landley | If you remove it, the resulting .config is different. |
| 21:34.02 | landley | I've found how to put allnoconfig into an endless loop. |
| 21:34.32 | landley | And I've found cases where running "make oldconfig" right after running "make allnoconfig" changes the .config. |
| 21:35.15 | landley | And I've worked around _all_three_ in my miniconfig conversion script, which still has to work its way through the rest of arm, and through mips. |
| 21:35.37 | landley | (I taught it to check existing miniconfig files to make sure they match the defconfig, so I can use it to automatically upgrade the miniconfigs with each -pre release.) |
| 21:36.00 | landley | It also acts as a handy smoke test to see where previous miniconfig conversions didn't work right due to one of the above three bugs (or a few others I've bumped into)... |
| 21:36.19 | landley | Mostly the "it _needs_ a #comment line in the result to produce the right output" bug. That one I'm still boggling over. |
| 21:44.19 | ds2 | landley: heh is that really a bug? I discussed this with someone else and their arguement is that is a problem with they way the Kconfig is written |
| 21:53.56 | landley | It's more of a design flaw than strictly a _bug_, but it's Something It Should Not Be Doing. |
| 21:54.09 | landley | It means that somewhere, somebody disagrees with what "allnoconfig" means. |
| 21:54.22 | landley | Probably somebody using "selects" when they write kconfig... |
| 21:54.27 | landley | Kconfig data files, I mean. |
| 22:09.40 | *** join/#edev CosmicPe1guin (n=nobody@163.181.251.103) |
| 22:10.01 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 22:19.17 | *** join/#edev linac (n=lin@121.41.32.186) |
| 22:21.49 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 22:30.51 | *** join/#edev ce_geek (n=ce_geek@c-24-61-92-197.hsd1.nh.comcast.net) |
| 22:31.05 | ce_geek | prp; hi |
| 22:31.07 | ce_geek | prpplague; hi |
| 22:31.16 | prpplague | ce_geek: hey hey |
| 22:31.22 | prpplague | ce_geek: whats cookin? |
| 22:31.32 | ce_geek | Not much. It's virtual friday. |
| 22:31.44 | ce_geek | How about yourself? |
| 22:31.52 | ce_geek | You must be about set to leave work |
| 22:32.17 | ce_geek | prpplague: Any word when that kit will go out? |
| 22:33.00 | prpplague | ce_geek: i'll ahve to check on it, i've been totally snowed under for the last couple of weeks |
| 22:33.31 | ds2 | need bigger plows on those pickup trucks ;) |
| 22:34.39 | ce_geek | prpplague: Okay, no problem. |
| 22:35.31 | ce_geek | prpplague: I'm going to do some home temperature stuff with it. I have some serious thermal issues here in my apartment. Once I get mine working I'm planning to build one for my folks house as well, probably going to buy a nail for that. |
| 22:35.41 | prpplague | ce_geek: cool |
| 22:36.22 | ce_geek | prpplague: I imagine the stuff I wrote up still needs help making it to the Wiki. I haven't had a chance to look. I can help with that. |
| 22:36.36 | prpplague | ce_geek: gotta run, closing the office |
| 22:36.40 | prpplague | ce_geek: try to go on later |
| 22:36.45 | ce_geek | prpplague: Okay, take care. |
| 22:49.16 | *** join/#edev cajaka (i=dirc@chil-nrp1-cs-1345.vdsl.bright.net) |
| 23:13.15 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 23:39.34 | *** join/#edev sunrunner20 (n=jamesdoe@pool-71-170-134-104.dllstx.fios.verizon.net) |