| 00:15.35 | *** join/#openezx TimRiker (n=timr@pdpc/supporter/bronze/TimRiker) |
| 00:18.38 | *** join/#openezx TimRiker (n=timr@pdpc/supporter/bronze/TimRiker) |
| 01:18.48 | *** join/#openezx _kenws (n=ken@dslb-088-064-153-139.pools.arcor-ip.net) |
| 02:17.00 | *** join/#openezx LoneStar (n=xinzhen@211.144.200.92) |
| 03:16.41 | nh_home | How goes the JTAG stuff guys? |
| 06:23.40 | *** join/#openezx proberts (n=proberts@67-20-86-248.vnnyca.adelphia.net) |
| 06:26.58 | proberts | Is there source for mux_cli.o module from Motorola? |
| 08:22.15 | *** join/#openezx blkhawk (i=blkhawk@pD9E07D7D.dip.t-dialin.net) |
| 09:07.32 | *** join/#openezx webmind (i=webmind@194.145.200.126) [NETSPLIT VICTIM] |
| 09:07.32 | *** join/#openezx jamesmck (n=james@131.111.64.221) [NETSPLIT VICTIM] |
| 09:07.32 | *** join/#openezx SuN (i=unices@86.82.173.228) [NETSPLIT VICTIM] |
| 10:03.37 | *** join/#openezx ao2 (n=u@2001:1418:117:0:0:0:0:1) |
| 10:19.59 | *** join/#openezx leoncamel (n=leoncame@219.142.137.2) |
| 10:39.48 | *** join/#openezx |dion| (n=dionua@xen.inhex.net) |
| 10:46.32 | stefan_schmidt | *yawn* |
| 10:46.35 | stefan_schmidt | morning all |
| 10:46.51 | stefan_schmidt | nh_home: We are still trying |
| 10:47.06 | stefan_schmidt | proberts: Yes it is offered from moto. |
| 10:47.55 | ao2 | hi stefan_schmidt |
| 10:48.06 | stefan_schmidt | hi ao2 |
| 10:49.25 | stefan_schmidt | proberts: http://svn.openezx.org/branches/motorola/mux |
| 11:27.17 | *** join/#openezx |dion| (n=dionua@xen.inhex.net) |
| 11:39.44 | *** join/#openezx jim_rayner (n=jim_rayn@beyondvoice.plus.com) |
| 12:01.23 | *** join/#openezx alphaone (n=alphaone@c120.apm.etc.tu-bs.de) |
| 12:25.18 | stefan_schmidt | hi alphaone |
| 12:32.09 | alphaone | hi |
| 13:07.51 | Shoragan | vmaster, are you here? |
| 13:09.28 | vmaster_ | Shoragan: hey |
| 13:09.41 | Shoragan | hi |
| 13:10.02 | Shoragan | i've tried openocd with the E680, but it doesn't seem to work :/ |
| 13:10.33 | Shoragan | it loops after the reset, i think installing the debughander fails |
| 13:10.50 | vmaster | have you been able to identify both resets? |
| 13:11.24 | Shoragan | both resets? |
| 13:11.34 | vmaster | nTRST and nSRST |
| 13:11.42 | vmaster | you neet both to be able to debug an XScale core |
| 13:11.54 | Shoragan | i've found nSRST |
| 13:12.00 | Shoragan | and that is workin ok |
| 13:12.11 | Shoragan | how can i verify that both are working? |
| 13:12.26 | vmaster | comment out the target line in the log |
| 13:12.37 | vmaster | s/log/config/ |
| 13:13.49 | vmaster | then start the openocd |
| 13:13.53 | vmaster | telnet to port 4444 |
| 13:14.17 | Shoragan | ok |
| 13:14.22 | Shoragan | done |
| 13:14.46 | vmaster | jtag_reset <ntrst> <nsrst> allows you to assert (1) or deassert (0) them |
| 13:14.54 | vmaster | try jtag_reset 1 1 |
| 13:14.58 | vmaster | that should keep the board in reset |
| 13:15.10 | vmaster | what jtag interface do you use? |
| 13:16.06 | stefan_schmidt | JTAGKey-tiny |
| 13:16.26 | stefan_schmidt | Shoragan and alphaone are my local hacker team. :) |
| 13:16.33 | vmaster | ah, hehehe, nice |
| 13:16.55 | vmaster | can you measure nTRST and nSRST? they should both be ~0V |
| 13:17.33 | alphaone | But doesn't the device stay in reset even if you just set nSRST? |
| 13:17.44 | Shoragan | n[TS]RST on the JTAGKey-tiny? |
| 13:18.40 | vmaster | alphaone: you need nTRST to initialize some of the debug functionality - it /might/ work without, but i'm not sure about that. intel says you need both |
| 13:18.54 | vmaster | Shoragan: guess it's easier to measure at the board |
| 13:22.47 | stefan_schmidt | vmaster: we measure right now... |
| 13:24.39 | Shoragan | ok, both reset lines are correct |
| 13:25.06 | Shoragan | SRST resets the system, but i'm not sure how to verify that TRST is correct |
| 13:27.24 | alphaone | If we pull down TRST by itself there is no reaction whatsoever |
| 13:30.54 | Shoragan | vmaster: the scan command always gives the same result: |
| 13:30.57 | Shoragan | > scan |
| 13:30.57 | Shoragan | 0: idcode: 0x00000000 ir length 7, ir capture 0x1, ir mask 0x7f, current instruction 0x7f |
| 13:31.03 | nh_home | has hacking weekend begun? :) |
| 13:31.16 | Shoragan | yes :) |
| 13:31.54 | vmaster | Shoragan: scan_chain just prints the layout of the scanchain |
| 13:32.00 | vmaster | and the idcode field isn't used currently |
| 13:32.04 | Shoragan | ah |
| 13:32.14 | vmaster | i wanted to implement it, but it's not much use anyway |
| 13:32.25 | vmaster | you could modify a line in jtag.c |
| 13:32.59 | vmaster | comment out line 354 |
| 13:32.59 | vmaster | <PROTECTED> |
| 13:33.04 | vmaster | d'oh |
| 13:33.31 | vmaster | <PROTECTED> |
| 13:33.53 | vmaster | then try: jtag_reset 1 1 |
| 13:33.55 | vmaster | endstate rti |
| 13:34.26 | Shoragan | in revision 121, xscale branch? |
| 13:35.29 | vmaster | yeah |
| 13:35.41 | vmaster | ah |
| 13:35.42 | vmaster | sorry |
| 13:35.57 | vmaster | i looked at trunk |
| 13:36.12 | vmaster | line 374 |
| 13:36.26 | vmaster | in jtag_add_ir_scan() |
| 13:39.29 | Shoragan | ok, it gives no output |
| 13:39.54 | vmaster | irscan 0 0x7e |
| 13:40.20 | Shoragan | value captured during scan didn't pass the requested check: captured: 0x7f check_value: 0x01 check_mask: 0x7f |
| 13:40.30 | vmaster | ok, cool, that means it works |
| 13:41.11 | Shoragan | ah? |
| 13:43.40 | Shoragan | ok, i tried it with toggeling trst only. with trst high irscan fails |
| 13:44.20 | Shoragan | and after switching it to low again, it works :) |
| 13:44.28 | vmaster | ok, cool |
| 13:44.40 | vmaster | could you try again with the target line enabled |
| 13:44.45 | vmaster | and send me a log file |
| 13:44.52 | vmaster | to Dominic.Rath <at> gxm.de |
| 13:44.56 | vmaster | gmx |
| 13:45.34 | Shoragan | is it possible that the voltages for the data lines are wrong/too low? |
| 13:46.32 | vmaster | ah, did you connect Vtref? |
| 13:46.43 | alphaone | Yes, we did |
| 13:47.02 | vmaster | the XScale is 3.3V IO, right? |
| 13:47.30 | vmaster | mhh, anyway, the voltage is right, otherwise the irscan 0 0x7e would have failed with TRST deasserted |
| 13:47.34 | alphaone | I measured ~2.7V on the pads |
| 13:48.29 | vmaster | the JTAGkey-tiny is speced as 2.8V to 5V |
| 13:48.49 | vmaster | which xscale is on that phone? PXA27x? |
| 13:48.54 | Shoragan | 271 |
| 13:50.12 | Shoragan | what should i try with the target line enabled? |
| 13:52.10 | vmaster | start the openocd |
| 13:52.17 | vmaster | with -l <logfile> |
| 13:52.19 | vmaster | and -d |
| 13:53.24 | Shoragan | with daemon_startup attach or reset? |
| 13:53.32 | vmaster | doesn't matter at the moment |
| 13:53.35 | vmaster | i think |
| 13:53.41 | vmaster | reset to be safe |
| 13:54.42 | vmaster | ah, yeah, you have to be specify reset, otherwise it should tell you that you have to reset the target to debug an xscale |
| 13:55.59 | vmaster | mail me the log, and i'll have a look at it in about 10 minutes |
| 13:56.01 | Shoragan | ok |
| 13:56.06 | vmaster | gotta get myself something for lunch |
| 13:56.36 | Shoragan | i always need to run openocd twice before i tries anything :/ |
| 13:58.07 | Shoragan | k, mail is sent |
| 14:10.38 | vmaster | which FT2232 are you using? |
| 14:10.41 | vmaster | libftdi? |
| 14:12.36 | Shoragan | libftdi |
| 14:12.56 | Shoragan | i didn't get the other one to work... |
| 14:17.18 | vmaster | hmm, the communication isn't reliable |
| 14:17.39 | vmaster | all "Warning: jtag.c:1074 jtag_read_buffer(): value captured during scan didn't pass the requested check:" indicate a problem with the jtag communication |
| 14:17.53 | alphaone | Maybe the voltage? |
| 14:18.13 | *** join/#openezx florian (n=fuchs@87.193.4.167) |
| 14:18.18 | stefan_schmidt | hi florian |
| 14:18.25 | vmaster | hmm, 2.7V is at the tiny's limit |
| 14:18.45 | stefan_schmidt | florian: Do you have time to tell me about your 2.4 status? |
| 14:18.48 | vmaster | try increasing the jtag_speed divisor |
| 14:18.55 | vmaster | currently it's set to 2, right? |
| 14:19.08 | vmaster | how long are your wires? |
| 14:19.30 | stefan_schmidt | florian: I would like to know, if you were able to build a 2.4, flash it and it don't turn off. |
| 14:19.42 | Shoragan | speed is currently 2 |
| 14:19.43 | alphaone | vmaster: about 20cm |
| 14:20.10 | vmaster | try jtag_speed 10 |
| 14:20.38 | vmaster | 20cm is a bit long, but it should work - jtag is a bit picky about the cables used |
| 14:20.51 | vmaster | and only 2.7v isn't going to improve the situation |
| 14:21.32 | alphaone | :-) |
| 14:22.26 | florian | hi all |
| 14:22.41 | nh_home | wow, lots of updates on the wiki recently |
| 14:22.47 | nh_home | florian: heya. |
| 14:23.58 | alphaone | vmaster: We'll try with jtag_speed 10 and if that doesn't work with shorter wires |
| 14:24.21 | Shoragan | how log should it take with speed 10 before openocd becomes ready? |
| 14:25.08 | vmaster | only a couple of seconds |
| 14:25.27 | Shoragan | it is still running until the phones switches itself off |
| 14:26.03 | vmaster | i have to run with jtag_speed 21, because the funny engineers that designed my PXA250 board decided to put a low-pass filter on TCK, limiting me to ~250kHz |
| 14:26.03 | Shoragan | and then it gives some errors and aborts |
| 14:26.20 | Shoragan | hmm |
| 14:26.23 | vmaster | are there still Warnings? |
| 14:26.36 | florian | stefan_schmidt: right, boot_usb turned off the asian one but not the european one. |
| 14:26.59 | stefan_schmidt | florian: hmm, I mean flashed |
| 14:27.04 | florian | stefan_schmidt: flashing worked on both but for some reason my userland didn't work at all |
| 14:27.44 | stefan_schmidt | florian: Phone don't turn off after 1,5 minutes? |
| 14:27.59 | stefan_schmidt | florian: Which kernel release do you use for it? |
| 14:28.06 | nh_home | heh, someone made a swapfile mpkg for the E2. |
| 14:28.34 | florian | stefan_schmidt: hmm... it turned off after some time iirc, but i could try to verify this evening |
| 14:28.39 | stefan_schmidt | nh_home: still now E2 here. :) |
| 14:28.58 | nh_home | stefan: Are you going to buy one? |
| 14:29.08 | florian | stefan_schmidt: i believe that might be caused by lacking input processing. |
| 14:29.14 | stefan_schmidt | nh_home: student... |
| 14:29.20 | Shoragan | vmaster, some warnings in the beginning, and then only when the phone has switched of |
| 14:29.21 | alphaone | florian: Do you have the config of that kernel handy? |
| 14:29.21 | nh_home | stefan: ah, ok |
| 14:29.34 | vmaster | Shoragan: could you send me another logfile? |
| 14:29.49 | nh_home | stefan: I graduated about 2 1/2 years ago. |
| 14:30.05 | florian | alphaone: i think so yes... the machine should be on... i hope |
| 14:30.05 | Shoragan | vmaster, i tried it with lower speed |
| 14:30.29 | nh_home | stefan: I worked for a military contractor for 1 1/2 years, but I hated it so I came back to the University and now work for the Supercomputer Institute. :) |
| 14:30.54 | alphaone | florian: Would be great if you could send me the config |
| 14:32.09 | stefan_schmidt | florian: atm only the original zImage of the e680 booted into userland and we have a working phone. Get this to work with a selfcompiled 2.4 would help us to find the problem. |
| 14:32.52 | Shoragan | vmaster, mail sent :) |
| 14:33.21 | florian | stefan_schmidt: indeed |
| 14:44.22 | Shoragan | vmaster, any ideas? |
| 14:44.40 | florian | http://linuxtogo.org/~florian/files/a780-config |
| 14:44.48 | florian | iirc that is the one i used |
| 14:44.52 | florian | bbiab |
| 14:48.04 | alphaone | thanks, I'll play around with that one |
| 14:53.24 | stefan_schmidt | florian: which kernel release you used this config with? |
| 14:53.46 | stefan_schmidt | florian: zImage fails for me |
| 14:54.36 | vmaster | Shoragan: there are still communication problems |
| 14:54.42 | vmaster | you could try with shorter wires |
| 14:55.05 | vmaster | if that doesn't work, i have a JTAGkey (not tiny) prototype Amontec sent me, that goes down to 1.8V |
| 14:55.58 | vmaster | you don't happen to have a scope to verify the signals? |
| 14:56.11 | stefan_schmidt | vmaster: We have one here |
| 14:56.21 | stefan_schmidt | old analog |
| 14:56.46 | stefan_schmidt | vmaster: alphaone shorts the wires now |
| 14:57.14 | vmaster | stefan_schmidt: ok, you could check TCK and TDO |
| 14:57.18 | stefan_schmidt | vmaster: We can try with the scope afterwards. But you need to tell us what signals we should look for. |
| 14:57.51 | vmaster | TCK is driven by the JTAGkey-Tiny with VTref, while TDO is driven by the target |
| 14:58.15 | stefan_schmidt | ok |
| 14:58.47 | Shoragan | the address for the debughandler and cache is ok? |
| 14:59.21 | vmaster | the problem is definitely the failing communication |
| 14:59.28 | vmaster | the addresses should be fine |
| 14:59.44 | stefan_schmidt | ok |
| 15:05.48 | Shoragan | the cables are about 5 cm now, but it dosn't seem to make a difference :/ |
| 15:05.54 | Shoragan | well try the scope now |
| 15:22.22 | *** join/#openezx LoneStar (n=xinzhen@218.81.34.26) |
| 15:26.09 | Shoragan | vmaster, TDO on the JTAGkey should be connected to TDI or TDO on the phone? |
| 15:27.35 | vmaster | to TDO |
| 15:27.40 | vmaster | but you've got that right |
| 15:27.56 | vmaster | otherwise the openocd would fail in jtag_validate_chain |
| 15:28.30 | vmaster | it might really be because of the voltage |
| 15:28.58 | stefan_schmidt | vmaster: Shoragan uploads some images of the scope for you. moment |
| 15:29.25 | stefan_schmidt | vmaster: But good to know that we don't wired up TDI/TDO wrong. |
| 15:30.33 | alphaone | vmaster: Which are the relevant signals? |
| 15:31.45 | alphaone | I'd try to boost the voltage with a transistor for all the signals from the device to the JTAG |
| 15:32.38 | vmaster | alphaone: TCK and TDO should do |
| 15:33.46 | alphaone | I thought TCK was driven by the JTAG? |
| 15:34.21 | vmaster | yeah, but the JTAGKey could have problems when VTRef is only 2.7V |
| 15:34.29 | alphaone | okay |
| 15:34.30 | vmaster | ah, btw., what is VTref? |
| 15:34.34 | vmaster | is it 2.7V, too? |
| 15:34.49 | alphaone | yeah |
| 15:34.59 | stefan_schmidt | florian: I'm not able to compile a 2.4 with your config with all three different kernel releses from moto. Would be nice if you can test it and report later. |
| 15:37.54 | vmaster | sorry I didn't think about voltages when i recommended the -tiny |
| 15:38.20 | alphaone | I just measured 2.77V with a cheap multimeter |
| 15:38.41 | Shoragan | http://www.sicherheitsschwankung.de/~jluebbe/ezx/IMG_8226.JPG |
| 15:39.20 | stefan_schmidt | hmm, we still have the wiggler here... |
| 15:40.27 | vmaster | self-made? |
| 15:40.32 | stefan_schmidt | bought |
| 15:40.37 | alphaone | The link is to a picture of TDI |
| 15:40.44 | stefan_schmidt | http://www.sander-electronic.de/es0016.html |
| 15:40.44 | alphaone | self-bought :-) |
| 15:41.04 | alphaone | TDO, sorry |
| 15:41.41 | vmaster | looks quite clean to me |
| 15:42.01 | vmaster | that's a olimex ARM-JTAG, it's limited to 3.0 to 3.6V |
| 15:42.24 | stefan_schmidt | aaargh |
| 15:42.30 | alphaone | heh |
| 15:42.55 | alphaone | Maybe it's the other way round? .-) |
| 15:43.01 | vmaster | hehehe |
| 15:43.24 | stefan_schmidt | funny, funny. ;) |
| 15:43.38 | vmaster | I can send you the JTAGkey prototype on monday - I still have a final JTAGkey, a Signalyzer and a ARM-USB-OCD here, so I wont miss it |
| 15:43.57 | vmaster | ah, and a JTAGkey-Tiny |
| 15:44.11 | stefan_schmidt | heh |
| 15:44.36 | stefan_schmidt | hmm, perhaps the normal JTAGKey would be worth a try. |
| 15:44.49 | vmaster | you could rule out voltage problems with it |
| 15:44.57 | stefan_schmidt | yeah |
| 15:45.00 | alphaone | Is TDI supposed to be a flatline? |
| 15:45.15 | vmaster | not really |
| 15:45.40 | vmaster | it's not active all the time, but it definitely should toggle sometmes |
| 15:45.51 | alphaone | Hmm, the other signals are: |
| 15:46.15 | alphaone | TCK toggles up and down in regular intervals |
| 15:47.30 | alphaone | TMS is up only when TCK is up but during the time where TCK is up TMS toggles state in regular intervals, maybe 10-20 times a TCK period |
| 15:48.36 | vmaster | huh? |
| 15:48.48 | vmaster | TMS toggles several times during one TCK cycle? |
| 15:48.52 | alphaone | yeah |
| 15:48.55 | alphaone | mom |
| 15:49.17 | vmaster | TMS should change at the falling edge of TCK |
| 15:49.21 | vmaster | and only then |
| 15:49.25 | vmaster | it's sampled at the rising edge |
| 15:49.37 | alphaone | hmm |
| 15:49.56 | alphaone | We'll take more pictures |
| 16:09.00 | alphaone | Okay, TMS seems to behave okay |
| 16:11.37 | alphaone | How often should TDI toggle? |
| 16:26.07 | alphaone | AFK: shopping |
| 16:26.40 | stefan_schmidt | + fresh air. :) |
| 17:03.38 | *** join/#openezx neolynx (n=u@167-123-204-62-pool.dsl.fcom.ch) |
| 18:16.52 | *** join/#openezx vmaster_ (i=vmaster@p549B7CAB.dip.t-dialin.net) |
| 18:36.32 | *** join/#openezx |dion| (n=dionua@xen.inhex.net) |
| 18:38.25 | *** join/#openezx TimRiker (n=timr@pdpc/supporter/bronze/TimRiker) |
| 18:46.28 | *** join/#openezx ao2 (n=u@2001:1418:117:0:0:0:0:1) |
| 18:50.58 | *** join/#openezx GvzEvxre (n=timr@pdpc/supporter/bronze/TimRiker) |
| 18:53.12 | *** join/#openezx florian (n=fuchs@87.193.4.167) |
| 18:55.34 | *** part/#openezx neolynx (n=u@167-123-204-62-pool.dsl.fcom.ch) |
| 18:56.11 | florian | re |
| 18:57.27 | stefan_schmidt | wb florian |
| 18:57.51 | stefan_schmidt | you read my statement about the .config? |
| 18:59.57 | florian | yes, try again... that was the wrong one :} |
| 19:00.48 | stefan_schmidt | ok, but some clarification before. :) |
| 19:01.03 | florian | i'm not sure what i changed, but if necessary i could make a patch |
| 19:01.06 | stefan_schmidt | You build, flashed and used this kernel with the moto uderspace? |
| 19:01.22 | stefan_schmidt | s/uderspace/userspace/ |
| 19:02.48 | florian | stefan_schmidt: i must admit that never tried with the moto userspace stuff :} |
| 19:03.01 | stefan_schmidt | hmm, ok |
| 19:03.21 | stefan_schmidt | florian: But you flashed and it don't turn off after 90 seconds? |
| 19:03.35 | florian | is the driver for the doc included? |
| 19:04.30 | stefan_schmidt | which doc? |
| 19:12.02 | *** join/#openezx jaebird (n=jae@53-89.netblk-69-41-89.coolaccess.net) |
| 19:12.35 | florian | isn't the internal flash a doc module? |
| 19:19.47 | stefan_schmidt | ah you mean disc-on-chip and not documentataion. |
| 19:20.10 | stefan_schmidt | florian: It's an m-systems doc |
| 19:24.15 | florian | :-) |
| 19:25.58 | *** join/#openezx z3ro (n=z3ro@60-234-138-172.bitstream.orcon.net.nz) |
| 19:34.57 | stefan_schmidt | ~botsnack |
| 19:34.57 | purl | stefan_schmidt: :) |
| 19:54.51 | *** join/#openezx jim_rayner (n=jim_rayn@beyondvoice.plus.com) |
| 19:57.56 | stefan_schmidt | wb jim_rayner |
| 19:58.32 | jim_rayner | stefan_schmidt: how goes it? |
| 19:59.53 | stefan_schmidt | getting better all the time. |
| 20:00.09 | stefan_schmidt | We were able to complete the passthrough mode. :)) |
| 20:00.21 | jim_rayner | :) |
| 20:00.54 | stefan_schmidt | In this mode the phone (BP) appears as exactly the PID:VID the mux_cli like to use. :) |
| 20:01.11 | stefan_schmidt | This means we can work on mux_cli on normal pc hardware. |
| 20:01.57 | stefan_schmidt | Unfortunately Harald's mux_cli patch have some depends on arm. We don't started yet to port it over to x86 |
| 20:58.32 | ao2 | stefan_schmidt, are you keeping a log of your actions beside this channel? :) |
| 20:59.50 | stefan_schmidt | ao2: : A mail with all we found and got working will follow to the ml |
| 21:00.03 | stefan_schmidt | ao2: atm we focus on hacking |
| 21:00.35 | stefan_schmidt | ao2: I hope to write the mail today night or tommorow evening. |
| 21:10.19 | *** join/#openezx kenws (n=ken@dslb-088-064-153-139.pools.arcor-ip.net) |
| 22:01.25 | *** join/#openezx |dion| (n=dionua@xen.inhex.net) |
| 22:45.45 | *** join/#openezx nh_home (n=nh@aurora.msi.umn.edu) |
| 23:16.50 | proberts | .away |
