**** BEGIN LOGGING AT Thu Mar 09 10:59:57 2006 Mar 09 14:06:06 <[g2-lap]> vmaster ping Mar 09 14:06:32 [g2-lap]: did the jtag gpio stuff work out? Mar 09 14:06:56 <[g2-lap]> lennert I haven't wrapped that up yet :( Mar 09 14:07:01 okay, np Mar 09 14:07:03 <[g2-lap]> everything looks good Mar 09 14:07:03 no hurry Mar 09 14:07:15 <[g2-lap]> well I'd like to do two things Mar 09 14:07:32 <[g2-lap]> 1) get the changes upstream to vmaster Mar 09 14:07:40 0) test the changes? :-) Mar 09 14:07:44 <[g2-lap]> 2) have it really working :) Mar 09 14:08:20 <[g2-lap]> I am testing the changes Mar 09 14:08:23 [g2-lap]: pong Mar 09 14:08:29 <[g2-lap]> hey vmaster Mar 09 14:09:09 <[g2-lap]> it's a little like groundhog day Mar 09 14:09:30 * [g2-lap] wakes up and I'm exactly where I was yesterday morning :) Mar 09 14:09:38 hehehe Mar 09 14:09:56 <[g2-lap]> openOCD 99% there and need to fix and get upstream Mar 09 14:10:09 <[g2-lap]> make it available for koen and dyoung-zzzz Mar 09 14:11:14 <[g2-lap]> vmaster thoughts on upstream to you ? Mar 09 14:12:05 send me a patch ('svn add ep93xx.c'), and i'll push it into the repository Mar 09 14:12:56 would be nice if it worked, of course Mar 09 14:13:17 but that's not strictly a requirement - i don't think anyone is going to try this himself within the next few days Mar 09 14:13:43 <[g2-lap]> vmaster ok Mar 09 14:14:04 <[g2-lap]> and have you added the _ep93xx.cnf files ? Mar 09 14:14:26 hmm? Mar 09 14:14:46 <[g2-lap]> for the ep93xx port Mar 09 14:14:58 ah, you mean an example configuration? sure, i can add that, too Mar 09 14:15:01 <[g2-lap]> yeah Mar 09 14:15:24 <[g2-lap]> and the Makefile changes do you want a diff for that too ? Mar 09 14:15:35 yes, please Mar 09 14:15:37 <[g2-lap]> I think I"m just missing the build= issue Mar 09 14:15:50 i can fix it myself, if something is missing Mar 09 14:16:19 <[g2-lap]> I think that should only take 5-10 minutes to run down Mar 09 14:17:18 <[g2-lap]> vmaster is r47 still the latest ? Mar 09 14:18:19 yeah Mar 09 14:20:11 <[g2-lap]> Ok how's this I'll make a fresh pull and work out all the diffs for upstream Mar 09 14:21:20 <[g2-lap]> then we should be able to build in OE and dyoung can have images to test and tweak if I don't have it all working by the time he wakes up Mar 09 14:21:37 well, if you want to do that, wait half an hour - i'm about to push r48 online soon, just a few more tests Mar 09 14:21:52 then you could already send diffs against that Mar 09 14:22:44 <[g2-lap]> Ok that's fine Mar 09 14:23:00 <[g2-lap]> my diff's from 47 or 48 should be really small Mar 09 14:23:07 if any Mar 09 14:23:11 <[g2-lap]> nod Mar 09 14:23:13 <[g2-lap]> :) Mar 09 16:22:55 <[g2-lap]> koen ping Mar 09 16:34:58 pong Mar 09 16:35:29 <[g2-lap]> koen how is your configure/autotools/auto* foo ? Mar 09 16:36:19 <[g2-lap]> I've got a patchset against r47 for openocd now (it builds but needs to be debugged a little more) Mar 09 16:36:52 <[g2-lap]> I"m just missing one piece where the build for ep93xx is nailed up in config file Mar 09 16:37:15 <[g2-lap]> I'd like to switch that but my auto* foo is weak :( Mar 09 16:37:26 mine is non-existant Mar 09 16:37:39 <[g2-lap]> Ok well thx then Mar 09 16:37:51 <[g2-lap]> I was just checking Mar 09 16:38:26 <[g2-lap]> I'll wait for vmaster to roll 48 and give him a patchset for that and focus on the debugging Mar 09 16:38:56 * [g2-lap] heads off for lunch Mar 09 17:06:04 [g2-lap]: r48 is available. sorry it took a bit longer than i hoped. Mar 09 17:34:48 <[g2-lap]> vmaster np Mar 09 17:35:11 <[g2-lap]> it's been taking longer than I hoped to make stuff work / release too :) Mar 09 17:59:26 always does Mar 09 19:44:39 <[g2-lap]> vmaster ping Mar 09 19:45:35 <[g2-lap]> vmaster fyi: http://ep93xx.giantshoulderinc.com:8001/ep93xx/browser/openocd/openocd.diff Mar 09 19:45:53 <[g2-lap]> there are 3 issues with that patchset on R48 Mar 09 19:46:29 <[g2-lap]> 1 and 2) build_ep93xx and bitbang are nailed up in .in file Mar 09 19:46:39 <[g2-lap]> 3) it doesn't fully work :) Mar 09 19:46:46 <[g2-lap]> but it damn close Mar 09 19:46:54 <[g2-lap]> s/it/it is/ Mar 09 19:46:55 [g2-lap] meant: but it is damn close Mar 09 19:47:49 <[g2-lap]> I guess 4 and 5) the timing needs to be better for the looping and the it's for the PORT C on the 9312 Mar 09 19:47:59 <[g2-lap]> but really #3 is the issue Mar 09 19:48:33 <[g2-lap]> koen|away could you add that patch to OE for openocd ? Mar 09 19:49:06 <[g2-lap]> openocd for machine == ep93xx Mar 09 19:49:21 <[g2-lap]> kas11 HEY !! Mar 09 19:49:58 ok...so what is the hot setup for usb to serial?....I need a serial port on this laptop so I can see what is locking up the kernel on another box Mar 09 19:50:38 hiya [g2]...that question was for the general population but I bet you know the answer ;) Mar 09 19:50:42 <[g2-lap]> a simple usb-serial adapter Mar 09 19:51:29 I was thinking maybe a trip to Radio Shack for a cell phone thingee was in order... Mar 09 19:51:46 <[g2-lap]> dunno if they still have them Mar 09 19:52:03 <[g2-lap]> it depends on what you are trying to optimize Mar 09 19:52:10 then I suppose I need a null modem cable or do some surgery on the gizmo Mar 09 19:52:31 <[g2-lap]> do you have serial on the device ? Mar 09 19:52:39 <[g2-lap]> like a DB-9 Mar 09 19:53:01 yeah...the box is a std PC...gonna use minicom on the laptop Mar 09 19:53:40 yeah DB-9M Mar 09 19:53:49 kas11: there are 3 or 4 good chips out there Mar 09 19:53:59 <[g2-lap]> yeah a simple usb-serial adapter and having a null modem around Mar 09 19:54:01 kas11: the ft232 series is pretty popular Mar 09 19:54:11 kas11: as well as the prolific chips Mar 09 19:54:31 <[g2-lap]> ka6sox I've got a couple and that's how the laptop talks to the Lofts and other boxes Mar 09 19:54:46 <[g2-lap]> I've even run minicom from the Loft and NSLU2 to another Mar 09 19:55:08 k...I remember both those names...thx prpplague :) Mar 09 19:55:39 I don't need a gizmo thats gonna compound my problems Mar 09 19:58:15 I get along with minicom ok...now kermit is an entirely different story... Mar 09 19:59:35 it sure would be nice if I hadn't tossed that televideo 970 a couple of moves ago Mar 09 20:00:05 hi kas11 :) Mar 09 20:00:19 hiya lennert :) Mar 09 20:01:55 hi kas11 Mar 09 20:03:03 ;) @ ka6sox Mar 09 20:14:04 brilliant...compusa carries a null modem cable...DB25 to DB25...and they say rocket science is dead in this country Mar 09 20:15:36 when i get time i will do a diagram for a bit wigler level transator forthe ts72xx dio port for those takingthat route Mar 09 20:16:06 AchiestDragon: great Mar 09 20:16:25 will want to know the pins your using Mar 09 20:17:12 for now, i think it was 0-TDO 1-TDI 2-TCK 3-TMS 4-TRST 5-SRST Mar 09 20:17:18 but that can change at any moment :) Mar 09 20:17:23 k Mar 09 20:18:06 will do one that uses std dip components and can be built on strip board Mar 09 20:18:45 <[g2-lap]> vmaster dyoung-zzzz koen|away lennert fyi: http://ep93xx.giantshoulderinc.com:8001/ep93xx/wiki/OpenOCD Mar 09 20:18:52 <[g2-lap]> there's a link off the main page Mar 09 20:34:53 googling for SBC9312 gives me a bunch of very oriental looking links...anyone got a cluebat handy? Mar 09 20:35:36 I was feeling very out-of-the-loop...and I am naturally nosey :) Mar 09 20:36:23 www.glomationinc.com Mar 09 20:36:28 <[g2-lap]> kas11 can you grace us with your presence in #ep93xx ? Mar 09 20:36:41 http://www.glomationinc.com/products_9312-sx.html in particular Mar 09 20:37:56 <[g2-lap]> kas11 our wiki is at http://ep93xx.giantshoulderinc.com:8001/ep93xx Mar 09 20:59:34 [g2-lap]: commited as revision 49, with a proper configure.in Mar 09 21:01:04 <[g2-lap]> vmaster ok cool Mar 09 21:01:12 <[g2-lap]> lennert may have found the issue Mar 09 21:01:19 <[g2-lap]> my nanosleep was bogus Mar 09 21:01:31 <[g2-lap]> I'm building now Mar 09 21:01:36 <[g2-lap]> but thanks Mar 09 21:01:52 okay Mar 09 21:02:11 is nanosleep on arm better than on x86? Mar 09 21:02:50 <[g2-lap]> dunno, but it's what lennert suggested. it's local to the ep93xx which should be arm Mar 09 21:05:22 vmaster: no, it has HZ resolution too Mar 09 21:05:29 vmaster: although there are high-res timer patches Mar 09 21:05:39 vmaster: the nanosleep was to test whether we were over-driving the bus Mar 09 21:05:42 ah, ok Mar 09 21:12:38 <[g2-lap]> that's buiding on the glomation now Mar 09 21:23:18 <[g2-lap]> vmaster is there something to pass config to include bitbang ? Mar 09 21:23:26 <[g2-lap]> and ep93xx ? Mar 09 21:24:49 bitbang? Mar 09 21:24:53 --enable-ep93xx Mar 09 21:24:53 is that a util? Mar 09 21:25:24 if either parport or ep93xx is enabled, then bitbang is included, too Mar 09 21:25:41 <[g2-lap]> vmaster heh Mar 09 21:25:55 <[g2-lap]> I just tried --enable-ep93xx --enable-bitban Mar 09 21:25:58 <[g2-lap]> with a g Mar 09 21:26:50 i've disabled all by default, and require everyone to select the interfaces they want - people hat a lot of problems finding the disable switches... Mar 09 21:27:00 *had Mar 09 21:27:19 <[g2-lap]> right and it there an enable-all ? :) Mar 09 21:27:24 no Mar 09 21:27:31 but that doesn't make a lot of sense anymore Mar 09 21:27:46 the parallel port ones won't work on arm anyway Mar 09 21:27:49 with interfaces specific to one hardware platform Mar 09 21:27:53 <[g2-lap]> oh as it's kinda target specific now Mar 09 21:27:53 right Mar 09 21:27:56 unless you have some isa bus on your board Mar 09 21:28:12 <[g2-lap]> like the pc/104 Mar 09 21:28:19 <[g2-lap]> there are none of those around :) Mar 09 21:28:26 naah :) Mar 09 21:28:45 <[g2-lap]> lennert I guess you beat me to the punch on that one :) Mar 09 21:31:59 <[g2-lap]> vmaster r49 didn't build Mar 09 21:32:24 <[g2-lap]> so I modified the wiki svn pull -r48 for the patch for right now Mar 09 21:32:28 <[g2-lap]> -r 48 Mar 09 21:32:47 <[g2-lap]> I think the .am file for ep93xx is missing Mar 09 21:32:55 <[g2-lap]> .am mods Mar 09 21:34:57 <[g2-lap]> it looks like the ep93xx.c file is missing Mar 09 21:35:56 oh, yeah Mar 09 21:36:00 fixed in r50 Mar 09 21:36:33 <[g2-lap]> heh I just saw 50 come across the line Mar 09 21:36:45 <[g2-lap]> that still has the bad nanosleeps Mar 09 21:42:12 <[g2-lap]> vmaster koen r50 builds Mar 09 21:42:30 <[g2-lap]> I'll update the wiki and change the extra patch Mar 09 21:42:51 <[g2-lap]> the extra patch should just be the nanosleep changes Mar 09 21:45:03 <[g2-lap]> ok wiki updated Mar 09 21:53:12 i've removed the two --disables - they're not required anymore Mar 09 21:54:24 hi Mar 09 21:54:57 hi Mar 09 21:55:46 hey Mar 09 21:55:49 <- Dominic Mar 09 21:56:03 hehe, hi, hoped to meet you here Mar 09 21:56:06 mthomas_ is the WinARM guy Mar 09 21:56:13 excellent Mar 09 21:57:44 yes, just about this, I have seen that you have added the SAM7-flash support. Have you thought about my suggestion to add some command-line option for a script so openocd can be used used for flash-programming from inside a makefile Mar 09 21:58:28 yeah, i've thought about it, but no time yet Mar 09 21:58:48 actually, it might already be possible... Mar 09 21:59:41 oh, and the SAM7 flash support isn't tested at all - it's a patch I've received and added as-is Mar 09 22:00:43 yes, I have seen the contribution in the berlios system Mar 09 22:04:30 How is the state of the native W32-port without Cygwin? Mar 09 22:05:08 in the patch queue, but i've concentrated on making it generally more stable, fixing a few minor and some major bugs Mar 09 22:05:40 i'll let you know when i've incorportated mingw + giveio Mar 09 22:07:31 yes, fixing bugs is more important, i have to admit that I have not tested openocd very much, just checked that it can connect. I had a problem with switching on software-breakpoints but I have not spend a lot of time with this. Mar 09 22:07:51 <[g2-lap]> vmaster I'll try with just --enable-ep93xx Mar 09 22:08:17 <[g2-lap]> I've got the two _ep93xx config files and the ep93xx.c changes Mar 09 22:15:58 <[g2-lap]> vmaster thx for tweaking the wiki already Mar 09 22:21:07 <[g2-lap]> mthomas_ you don't work for the Vole do you ? Mar 09 22:21:22 <[g2-lap]> M$oft Mar 09 22:21:45 <[g2-lap]> interesting thought about running from the makefile Mar 09 22:22:15 <[g2-lap]> koen well have to teach OE to do something like that :) Mar 09 22:22:27 no - i use ms-windows, thats all. Mar 09 22:23:21 <[g2-lap]> that's what I was thinking :) Mar 09 22:23:52 <[g2-lap]> I'd love to hear about the dark-side sometime Mar 09 22:24:00 you may now lpc21isp, which is a software to flash thru presinstalled bootloaders on LPC2k and ADuC-ARMs. its a very convinient method: make all program Mar 09 22:24:34 i'm not that bad, we are running 4 linux servers Mar 09 22:24:50 <[g2-lap]> mthomas_ I'm linux/gnu guy doing embedded linux distributions Mar 09 22:25:01 if it makes you happy, the workgroupd would not "work" without the "good" linux in the background ;-) Mar 09 22:25:55 <[g2-lap]> for me it's kinda like the smoker catagories the cig. co's had Mar 09 22:26:01 <[g2-lap]> smokers Mar 09 22:26:06 <[g2-lap]> and "pre-smokers" Mar 09 22:26:24 * [g2-lap] just wonders how far the gap is Mar 09 22:26:38 <[g2-lap]> well I've got run now. Mar 09 22:26:47 <[g2-lap]> hope to see you around another time Mar 09 22:27:00 <[g2-lap]> vmaster r 50 with a tiny patchset builds Mar 09 22:27:20 <[g2-lap]> now there's just a very tiny patchset I'll update later Mar 09 22:27:43 <[g2-lap]> and workout how to pull the file without the html formatting Mar 09 22:27:47 <[g2-lap]> cheers all Mar 09 22:28:05 bye Mar 09 22:28:08 dominic: http://www.embeddedrelated.com/usenet/embedded/show/43995-1.php Mar 09 22:33:39 yeah, i talked to Antti at the Embedded-World - I'll reply that SAM7 support is now included Mar 09 22:45:29 ok Mar 09 22:48:47 if you have the mingw-patch somehow integrated just send me an e-mail, maybe I can help you if there are problems with it. Mar 09 22:49:04 byebye Mar 09 23:55:09 http://www.whipy.demon.co.uk/wiggler.pdf Mar 09 23:56:28 opps Mar 09 23:56:37 1 min theres an error on that Mar 09 23:59:24 http://www.whipy.demon.co.uk/wiggler.pdf , fixed this should cope with working with ether a 3.3 or 5v dio and translate for 1.2v to 5v on the jtag side with out the need for worrying about vref load or voltage selection Mar 10 00:01:12 the 74hc05 is avalable in 0.1" dip so could easaly be built on stripboard Mar 10 00:02:40 transistor is a bc108 , bc 548 a 2n222 is sutable also should be good to 24mhz Mar 10 00:08:45 ka6sox: ping Mar 10 00:12:07 hopes i got the transistor number right , its a npn and cannot remeber if thats the bc547 or 548 form memory Mar 10 00:13:12 checks :) yes is ok Mar 10 00:14:07 TRANSISTOR, NPN TO-92; Transistor type:Small Signal General Purpose; Transistor polarity:NPN; Case style:TO-92; Voltage, Vceo:30V; Current, Ic max:0.1A; Power, Ptot:500mW; hfe, min:420; Current, Ic hfe:2mA; Voltage, Vcbo:30V; Current, Ic continuous a max:0.1A; Pins, number of:3; Transistors, number of:1; ft, typ:300MHz Mar 10 00:21:37 the 74hc05 has a 10nS delay so it will switch to a max of 30mhz so should be kool at 15mhz Mar 10 01:10:10 AchiestDragon, pong Mar 10 01:11:45 the wigler sch ? thoughts Mar 10 01:12:59 they typically use a '244 Mar 10 01:14:20 but a 244 wil not level shift Mar 10 01:16:24 I just used 7407's once Mar 10 01:25:27 think most of the 244 based ones are derived from people moding the 244 based pargramers rather than a proper understanding of level shifters Mar 10 01:26:50 most likely Mar 10 02:00:07 the 244 design will cause problems if the jtag levels are suposed to be 2.5v and the 244 is a parallel port version opperating at 5v Mar 10 02:03:23 well just checking my parts store , not got any of them in stock , will sort an order out on the morning so i can get one built up Mar 10 02:08:29 bed for me , nn Mar 10 02:09:43 nn Mar 10 09:38:43 yawn Mar 10 09:39:34 AchiestDragon: nSRST should be open-collector Mar 10 09:39:39 ah, and morning Mar 10 09:40:01 k Mar 10 09:40:16 then remove res5 Mar 10 09:40:45 nTRST probably, too Mar 10 09:40:51 the 74hc05 all the outputs are open collector Mar 10 09:40:54 ah, ok Mar 10 09:42:26 the pulup is ok on nTRST as is in most cases though Mar 10 09:43:36 yeah, i'm not really sure about that one, but ARM says it should be open collector Mar 10 09:46:49 if the inputs all have pullups on the board that your testing then res1 to 5 could be ommited , Mar 10 09:48:07 but there better left there for those boards that dont , it should not affect it if they are left , unless the logic level requied for jtag is diferent to vref that it should not be Mar 10 09:52:08 farnell have a £20min order amount so need to get fill up the order with some other bits also Mar 10 10:06:04 74hc05 £0.22 resitors 10 at £0.012 cap already have but £0.02 transitor £ 0.07 connector £0.95 bit of stipboard £0.50 total £1.88 Mar 10 10:10:07 and need the idc connectors and header for the end that connects to the ts72** dio port Mar 10 10:55:27 should point out that only a 74hc05 or 74ac05 should be used for the dio at 3v3 a 74hct05 or 74ls05 ect will only operate from 5v logic , but the jtag end is fine with any to 1.2v logic **** ENDING LOGGING AT Fri Mar 10 10:59:57 2006