**** BEGIN LOGGING AT Sun Mar 26 10:59:56 2006
Mar 26 11:13:56 <eFfeM> anyone an idea which modules defines xscale_mc_clear_user_page, this is missing when I want to load video-buf.ko
Mar 26 11:33:04 <likewise> arch/arm/mm/copypage-xscale.c
Mar 26 11:33:56 <likewise> I guess this should be built-in your kernel.
Mar 26 11:38:19 <eFfeM> likewise, apparently the symbol isn't exported any more
Mar 26 11:38:47 <eFfeM> there seems to be a change here between .15 and .16
Mar 26 11:38:52 <eFfeM> no ENTRY line for the symbol
Mar 26 11:39:02 * likewise dumps his 2.6.16 kernel symbols...
Mar 26 11:39:39 <eFfeM> the odd thing is that I cannot find a caller 
Mar 26 11:40:13 <eFfeM> likewise: System.map?
Mar 26 11:41:12 <likewise> 00000000c00298c4 T xscale_mc_clear_user_page
Mar 26 11:41:45 <eFfeM> i have a slightly different addres, but it is there
Mar 26 11:41:50 <eFfeM> in my system.map as well
Mar 26 11:42:07 <eFfeM> the problem is that video-buf.ko references this as an external symbol
Mar 26 11:42:29 <eFfeM> and apparently it is not exported any more (at least that is my guess)
Mar 26 11:42:38 * eFfeM is not a wizard on the kernel
Mar 26 11:42:43 <likewise> That seems most likely.
Mar 26 11:42:50 * likewise is neither.
Mar 26 11:44:22 <eFfeM> there has been a change: http://grmso.net:8090/commit/f5d9b97ee0d6d00a29bf881263510d74291fb862/diff/3079ca621e9e09f4593c20a9a2f24237c355f683/
Mar 26 11:44:35 <eFfeM> and the current file does not have an ENTRY ...
Mar 26 11:45:00 <eFfeM> but someone in video_buf.ko still tries to use it
Mar 26 11:51:46 <likewise> void __attribute__((naked)) xscale_mc_clear_user_page(void *kaddr, unsigned long vaddr)
Mar 26 11:51:54 <likewise> No idea what the naked attribute is about...
Mar 26 11:52:14 <likewise> First, there is beer in the kernel, and now there is naked also? :-)
Mar 26 11:53:18 <NAiL> the naked attribute is new, IIRC
Mar 26 11:53:29 <NAiL> I think it's related to gcc4x's inlining
Mar 26 11:54:37 <eFfeM> shouldn't there have been an ENTRY() for the symbol. It is in other files in the video dir ...
Mar 26 11:54:50 <eFfeM> anyway, just to be safe, I'm rebuilding the kernel
Mar 26 11:55:16 <CIA-10> 03orepvik * 10kernel/2.6.16/defconfig: defconfig: Add CONFIG_VIDEO_DECODER=m
Mar 26 11:55:27 <eFfeM> NAiL: tnx
Mar 26 11:56:44 <CIA-10> 03repvik * r297 10/releases/slugos-3.6-beta/openembedded/packages/linux/ixp4xx-kernel/2.6.16/defconfig: Add CONFIG_VIDEO_DECODER=m
Mar 26 11:57:12 <NAiL> hang on... why is the first commit by "orepvik"?
Mar 26 11:57:18 * NAiL switched username on sf.net
Mar 26 12:03:52 <likewise> insmod ixp400_eth gives me a "Division by zero in kernel." bug. Anyone else seen this?
Mar 26 12:05:21 <NAiL> yup
Mar 26 12:05:32 <NAiL> but I think I was the only one
Mar 26 12:05:48 <NAiL> that was on the DS101, and I'm the only one doing stuff with the ds101
Mar 26 12:06:58 <eFfeM> rebuilt the kernel, but the dreaded U xscale_mc_clear_user_page is still there if I nm the .o file
Mar 26 12:11:09 <likewise> NAiL: OK, I am on a IXDP425 development platform here.
Mar 26 12:11:26 <likewise> You used the BSD_ variant of the IXP s/w ?
Mar 26 12:12:03 <NAiL> no, I used the IPL
Mar 26 12:12:59 <likewise> Did you use defconfig, or did you make changes (for example ,kernel debugging) ?
Mar 26 12:14:02 <likewise> NAiL: After the division bug, can you do anything with the NICs? I cannot, ifconfig fails finding them, and I get kernel bugs
Mar 26 12:14:15 <likewise> (sscheduling while atomic etc)
Mar 26 12:15:04 <NAiL> I seem to remember that it worked for me, but I'm not sure. Currently the thing is bricked, so I can't test :(
Mar 26 12:16:14 <pepijn> NAiL: have you read the stuff i wrote about 'turnup disk' ?
Mar 26 12:16:19 <NAiL> no
Mar 26 12:16:51 <NAiL> I just woke up, and everyone wants my attention :-P
Mar 26 12:16:58 <NAiL> I'll get around to it ;)
Mar 26 12:17:04 <likewise> NAiL: Flash contents bricked? Or you bricked? :-)
Mar 26 12:17:39 <NAiL> I managed to hit "enter" instead of "backspace" when trying a dd, so I overwrote RedBoot
Mar 26 12:18:20 <likewise> I unbrick stuff using openjtag and a dongle. 
Mar 26 12:18:23 <NAiL> And so far, I've been unable to find the JTAG pinout (with some help from the openjtag folks)
Mar 26 12:18:31 <likewise> ah ok....
Mar 26 12:18:47 <NAiL> if you have a good idea on how to find the jtag pinout, please tell ;)
Mar 26 12:19:37 <likewise> So this is not right? http://www.nslu2-linux.org/wiki/DS101/DS101Hardware
Mar 26 12:21:01 <NAiL> apparently not
Mar 26 12:22:33 <NAiL> Atleast, I cannot get it to work
Mar 26 12:24:45 <NAiL> I don't really know much about JTAG, so I'm kinda stuck
Mar 26 12:25:00 <likewise> pin 1 = VCC, pin 6 = GND
Mar 26 12:25:13 <NAiL> yeah, that I know for sure
Mar 26 12:25:13 <likewise> (from what I can see on http://www.nslu2-linux.org/gallery/DS101-PCBs/ds101_front_big?full=1)
Mar 26 12:25:22 <likewise> can you measure additional grounds?
Mar 26 12:25:32 <NAiL> additional grounds?
Mar 26 12:25:48 <likewise> i.e. 0 Ohms between GND and the other pins?
Mar 26 12:27:19 <NAiL> no additional grounds
Mar 26 12:32:21 <NAiL> pins 2,4 &5 have 10k ohm to vcc
Mar 26 12:34:07 <NAiL> pin 7 has 20omh to gnd
Mar 26 12:34:23 <likewise> ok, let me lookup some schematics of the IXDP425. Hold on.
Mar 26 12:35:27 <NAiL> pin 3 has 36ohm to vcc, and pin 7 has 36ohm to vcc
Mar 26 12:35:29 * NAiL is confused
Mar 26 12:35:38 <NAiL> (I'm no EE, mind you)
Mar 26 12:36:17 <NAiL> Would a scope help? 
Mar 26 12:36:24 <NAiL> I think I can get access to a scope tomorrow.
Mar 26 12:37:23 <likewise> NAiL: I am EE, but I am not very familiar with JTAG. At least these readings can help...
Mar 26 12:38:48 * likewise is opening the schematics PDF...
Mar 26 12:38:59 <NAiL> takes forever :-P
Mar 26 12:39:35 <dwery-away> iirc you usually place 10k pullups on in, out and clk..
Mar 26 12:39:49 <likewise> SSH -> SMB -> SSH -> SMB -> Unzip -> Unzip -> ... :-)
Mar 26 12:40:18 <NAiL> that'd be 2,4 and 5 then
Mar 26 12:40:44 <NAiL> I wonder how I managed to read wrong the last time
Mar 26 12:41:03 <dwery-away> regarding 7 and 3, I have no particular ideas :)
Mar 26 12:42:14 <dwery-away> NAiL: do you have a quick link to ixdp425 jtag schematic?
Mar 26 12:42:32 <NAiL> no, but likewise might?
Mar 26 12:42:52 <likewise> no link, but I can screenshot or email this PDF
Mar 26 12:43:08 <likewise> 585KB
Mar 26 12:43:29 <dwery-away> yes please. dwery@nslu2-linux.org
Mar 26 12:43:51 <likewise> TDI, TMS and TCK are pull-up with 10kOhms
Mar 26 12:44:01 <likewise> TDO is pulled down with 10 kOhms
Mar 26 12:46:53 <NAiL> uhm... why am I getting 10kOhm between pins 2,4,5 and either GND or VCC?
Mar 26 12:47:10 <NAiL> uhm... why am I getting 10kOhm between pins 2,4,5 and either GND or VCC?
Mar 26 12:47:28 <likewise> laptop battery ran out suddenly - back again,...
Mar 26 12:48:07 <likewise> Because there are resistors "pulling" the lines to a default state if no chip is driving the logic states.
Mar 26 12:48:43 <NAiL> yeah, I'm just confused by what looks like pins being pulled both up and down
Mar 26 12:51:12 <likewise> yes, this can be caused by the device being idle and having internal pull-down in the IXP chip. I am just guessing.
Mar 26 12:52:21 <dwery-away> NAiL: so you miss a 10k pulldown
Mar 26 12:52:33 <likewise> I would say 2,4 and 5 are TDI, TMS and TCK, you just need to try out all combinations :-)
Mar 26 12:52:48 <likewise> permutations, rather ;-)
Mar 26 12:53:19 <likewise> ah wait, the lines are both pulled-up AND down in the IXDP425 schematics
Mar 26 12:53:21 * likewise rubs his eyes
Mar 26 12:53:40 <likewise> except for TDO
Mar 26 12:54:41 <dwery-away> likewise: you forgot the attachment :)
Mar 26 12:57:01 <likewise> Ah, well I didn't forget it, but my laptop went standby moments before I send out the mail. When it woke up, it hadn't restored its samba connection yet (and yes, I clicked through some warning about something not being there :-)
Mar 26 12:57:17 <likewise> should be with you now
Mar 26 12:57:23 <dwery-away> thank :)
Mar 26 12:57:27 <dwery-away> s
Mar 26 12:59:59 <NAiL> hmm
Mar 26 13:00:05 <NAiL> I really don't get this
Mar 26 13:00:29 <NAiL> I tried measuring between pin6 and various GND points on the board
Mar 26 13:01:28 <dwery-away> some resistors are marked as "non populated" on the schematic
Mar 26 13:02:09 <dwery-away> NAiL: do you have any 1K to VCC?
Mar 26 13:02:41 <NAiL> apparently not
Mar 26 13:03:02 <NAiL> got a couple 40ohm, and three 10k.
Mar 26 13:03:29 <dwery-away> :(
Mar 26 13:03:30 * likewise rubs his eyes again and sees
Mar 26 13:03:44 <likewise> Ah yes, DNP = Do not place...
Mar 26 13:07:01 <likewise> Also, this might be interesting: http://www.intel.com/design/iio/applnots/273538.htm
Mar 26 13:13:43 <likewise> NAiL: You should measure the voltage levels on each pin when the unit has power. We then know what pins default to pull-down and which default to pull up. It could also be the pins are both pulled-up and -down, then you'd see 3.3/2 ~= 1.65 Volts on the lines.
Mar 26 13:14:15 * likewise is running out of further ideas.
Mar 26 13:15:13 * dwery-away too
Mar 26 13:16:03 <NAiL> hmm
Mar 26 13:16:08 <NAiL> this is really strange
Mar 26 13:16:31 <NAiL> I get 0v between vcc and gnd?
Mar 26 13:19:11 <likewise> when the unit is ON? 
Mar 26 13:19:14 * NAiL chooses another ground
Mar 26 13:19:27 <likewise> you are measuring with DC mode on the meter, right? ;-)
Mar 26 13:20:07 <NAiL> pin1 is 3.3, 2&3 is 0.5v, 4 is 3.3v, 7 is 0.5 and 8 is 0.
Mar 26 13:20:33 <NAiL> (6 is 0, but I suspect it's somehow become NC)
Mar 26 13:23:06 <NAiL> does that help? :)
Mar 26 13:23:27 <likewise> The TMS signal is the only signal pulled HIGH in the IXDP425 schematics, so that should match with pin 4.
Mar 26 13:23:56 <likewise> (which contradicts the wiki page about this)
Mar 26 13:24:02 <NAiL> yeah
Mar 26 13:25:51 <likewise> However, I suspect the unit does not have JTAG enabled by default. Intel even makes a remark in their schematics telling that "in production units, JTAG_TRST_N must be pulled low" "when not using JTAG".
Mar 26 13:26:26 <likewise> That in turn, should mean there are some resistors involved that can be removed/replaced in case the units needs debricking.
Mar 26 13:27:08 <likewise> However, tracing these resistors can be a hellofajob.. you can't measure the middle PCB layers, nor can you touch the IXP425 BGA pins...
Mar 26 13:30:27 <dwery-away> I'm not sure they did that.. the other boards we tested on are all jtag enabled
Mar 26 13:31:08 <NAiL> but, 4 appears to be TMS, so 2 and 5 should be TDI & TCK? 
Mar 26 13:31:30 <NAiL> GND seems to be screwy, so I guess I should pick that from somewhere else...
Mar 26 13:32:24 <dwery-away> yes
Mar 26 13:33:19 <NAiL> 1 VCC, 2 TDI/TCK, 3 ?, 4 TMS, 5 TDI/TCK, 6 "GND", 7 ?, 8 ?. The three "?"'s should be nTRST, the other RST and TDO, right?
Mar 26 13:37:23 * NAiL guesses that 1=VCC, 2=TDI, 3=NTRST, 4=TMS, 5=TCK, 6=GND
Mar 26 13:37:27 <dwery-away> possible
Mar 26 13:37:50 <NAiL> That would mean they only swapped TDI/NTRST from the arm-jtag-layout
Mar 26 13:38:16 <NAiL> (in the first 5)
Mar 26 13:38:18 <likewise> 3 could be TDO
Mar 26 13:38:50 <likewise> that makes a bit more sense to me (TDI, TDO, TMS, TCK are the main signals AFAIK)
Mar 26 13:39:48 <NAiL> I guess it's time for me to start trying ;)
Mar 26 13:40:08 <likewise> at least if I were a hardware designer I would group the signals :-) but then again I know grounds and voltage should be interleaved inbetween to prevent cross-over.
Mar 26 13:40:08 <dwery-away> NAiL: maybe you should also test your jtags suite using a known working board
Mar 26 13:40:27 <likewise> what dongle do you use?
Mar 26 13:41:27 <NAiL> wiggler
Mar 26 13:42:04 <likewise> ok
Mar 26 13:44:23 <eFfeM> ~seen marceln
Mar 26 13:44:35 <jbot_> marceln <n=marceln@pion.xs4all.nl> was last seen on IRC in channel #nslu2-linux, 1d 21h 14m 46s ago, saying: 'Hi eFfeM'.
Mar 26 13:50:40 * NAiL tries redoing pin6 so he gets proper GND
Mar 26 13:51:05 <NAiL> btw, should VCC be connected to VREF or VTARGET? Or both?
Mar 26 13:52:57 <likewise> See http://www.olimex.com/dev/pdf/arm-jtag.pdf
Mar 26 13:53:18 <likewise> VREF: Target voltage sense. Used to indicate the target's operating voltage to the debug tool.
Mar 26 13:53:38 <likewise> So, VCC connects with VREF
Mar 26 13:54:16 <likewise> if you want to feed the board through the wiggler, VTARGET can be used for that. (But I should not connect it)
Mar 26 13:56:11 <likewise> No, scratch last sentence: VTARGET is used to feed the debug tool (wiggler) from the target.
Mar 26 13:56:57 <CIA-10> 03ingeba * 10unslung/make/rsync.mk: 
Mar 26 13:56:57 <CIA-10> /opt/etc/rsync.secrets was not set up as a conf file and thus
Mar 26 13:56:57 <CIA-10> overwriting it on upgrade.
Mar 26 14:00:22 <eddyp> tbm: ping
Mar 26 14:04:17 <CIA-10> 03azummo * 10kernel/2.6.16/ (linux-2.6.16-i2c.patch series 50-i2c-bus-ixp4xx-hwmon.patch): Upgraded to latest i2c patchset.
Mar 26 14:05:07 <CIA-10> 03azummo * 10kernel/2.6.16/40-rtc-class.patch: upgraded to latest rtc class patchset
Mar 26 14:05:49 <CIA-10> 03azummo * 10kernel/2.6.16/45-eeprom-notifier.patch: Refreshed to cope with i2c patchset
Mar 26 14:09:39 <likewise> NAiL: OK, I am giving up. I hope you can unbrick the unit sometime.
Mar 26 14:10:04 <NAiL> I'm going to try, as soon as I get GND wired up properly
Mar 26 14:10:57 <likewise> NAiL: I nailed down (excuse the pun) the division by zero problem to somewhere in maclist_read()... still debugging.
Mar 26 14:11:13 <dwery-away> likewise: maclist is going to be removed anyhow
Mar 26 14:11:35 <likewise> ah... how would I have known this? 
Mar 26 14:11:47 <dwery-away> well.. a thread in lkml I guess :)
Mar 26 14:11:57 <dwery-away> they did not like the maclist patch
Mar 26 14:12:05 <dwery-away> we tried hard to push it
Mar 26 14:12:32 <dwery-away> the proposed solution is to use an event notification from the netdev subsystem
Mar 26 14:12:38 <dwery-away> so that when the interface goes up
Mar 26 14:12:43 <tbm> eddyp: yeah?
Mar 26 14:12:44 <dwery-away> some code is triggered to read the mac
Mar 26 14:12:56 <dwery-away> I have not time to code that right now :(
Mar 26 14:14:01 <likewise> what will be the alternative? ifconfig works but not for NFS mounted systems, right?
Mar 26 14:14:39 <dwery-away> the event is generated from the kernel, so there should be no differences between that proposal and the maclist code.
Mar 26 14:15:45 <likewise> I am not familiar with the kernel (learning still) but what is the API for this?
Mar 26 14:16:17 <dwery-away> the api is register_netdevice_notifier()
Mar 26 14:16:22 <dwery-away> the notification function
Mar 26 14:16:40 <dwery-away> will be called at some events: NETDEV_REGISTER and NETDEV_UP
Mar 26 14:16:54 <dwery-away> so we can know when the interface gets registered and when it goes up
Mar 26 14:17:05 <likewise> thanks
Mar 26 14:17:20 <dwery-away> the problem is how to read the eeprom from within the kernel
Mar 26 14:17:24 <likewise> also during link down / up status?
Mar 26 14:17:25 <dwery-away> there's no API.
Mar 26 14:18:08 <dwery-away> yes, every status change is reported.
Mar 26 14:19:24 <likewise> The bug in 2.6.16 is in maclist.c: maclist_read(). The line which says "key % count" generates a division by zero when count is zero.
Mar 26 14:19:49 <dwery-away> ok, I can fix the cvs right now if you want
Mar 26 14:20:06 <likewise> I do not know the proper fix yet though. Tryin to understand what happens there.
Mar 26 14:20:24 <dwery-away> an easy fix is to do not print if count == 0
Mar 26 14:20:26 <dwery-away> :)
Mar 26 14:21:20 <dwery-away> uops, I thought it was a printf
Mar 26 14:23:57 <CIA-10> 03azummo * 10kernel/2.6.16/75-dsmg600.patch: Fixed whitespaces and refreshed.
Mar 26 14:24:26 <likewise> I think "index = count? key % count: 0;"
Mar 26 14:24:38 <likewise> i.e. if count is zero, make index zero as well
Mar 26 14:24:47 <dwery-away> but why count is zero in the first place?
Mar 26 14:25:17 <likewise> I hate uncommented code (I hate 99% percent of the kernel for that).
Mar 26 14:26:17 <dwery-away> yep. the rules is comment only when it is not obvious. the problem is that obvious is something objective.
Mar 26 14:26:27 <dwery-away> s/rules/rule/
Mar 26 14:26:28 <jbot_> dwery-away meant: yep. the rule is comment only when it is not obvious. the problem is that obvious is something objective.
Mar 26 14:27:56 <likewise> yes, I know. I should now investigate what the conditions of the variable count exactly are, purely by seeing the surrounding code.
Mar 26 14:28:21 <likewise> count increases only by calls to maclist_add, AFAICS.
Mar 26 14:28:24 <dwery-away> ..but if you want to give a shot to the netdev notifier proposal.... :)
Mar 26 14:29:04 <eddyp> tbm: did something changed with the di image?
Mar 26 14:29:14 <eddyp> anyhting for me to test?
Mar 26 14:29:26 <tbm> eddyp: no
Mar 26 14:29:32 <tbm> eddyp: atm it's broken anyway
Mar 26 14:29:41 <dwery-away> eddyp: but you can test compiling the cvs if you have some spare time :)
Mar 26 14:29:53 <eddyp> because of the removed LE support?
Mar 26 14:30:14 <tbm> eddyp: no, because a) the daily images don't contain the proprietary ethernet driver
Mar 26 14:30:26 <tbm> and b) I broke something but theres no hurry to fix it because of a) ;-)
Mar 26 14:30:39 <tbm> but the lat official release should work... and does work, just not for you.
Mar 26 14:30:39 <eddyp> :))
Mar 26 14:30:58 <eddyp> bah
Mar 26 14:31:15 <tbm> I don't know if I mentioned it to you but I'm in the process of moving to another country
Mar 26 14:31:20 <tbm> so I'm packing right now...
Mar 26 14:31:28 <tbm> I'll do slug stuff again but a few days more or less won't hurt
Mar 26 14:31:30 <likewise> tbm: what's broken?  And where are you moving?!
Mar 26 14:31:38 <pepijn> tbm: where are you moving to?
Mar 26 14:31:40 <tbm> likewise: madrid
Mar 26 14:31:50 <tbm> well, only for 3 months to work with some researchers there
Mar 26 14:31:56 <pepijn> madrid is not a country :-P
Mar 26 14:32:09 <tbm> likewise: I included a program to make "beep", butit fails becaue I forgot to include some kernel modules... *cough*
Mar 26 14:32:15 <tbm> pepijn: spain
Mar 26 14:32:23 <pepijn> hehe
Mar 26 14:32:28 <pepijn> i know *that*
Mar 26 14:32:28 <pepijn> lol
Mar 26 14:32:41 <tbm> pepijn: well, you might be one of those Americans.
Mar 26 14:32:45 * tbm ducks  ;-)
Mar 26 14:33:02 * pepijn is actually also american, just not one of those
Mar 26 14:33:24 <tbm> I thought you were dutch?!?
Mar 26 14:33:31 <tbm> or is that eddyp?
Mar 26 14:33:37 <tbm> don't confuse me, guys
Mar 26 14:33:45 <pepijn> born american, rasied in the netherlands (dutch parents) living in norway ;)
Mar 26 14:33:52 <likewise> No, I am dutch :-)
Mar 26 14:33:57 <pepijn> s/rasied/raised/
Mar 26 14:33:58 <jbot_> pepijn meant: born american, raised in the netherlands (dutch parents) living in norway ;)
Mar 26 14:34:08 * eddyp is Romanian
Mar 26 14:34:15 <tbm> pepijn: and I'm supposed to keep track of that? ;-)
Mar 26 14:34:39 * pepijn has a hard time keeping track of himself
Mar 26 14:35:00 <likewise> Madrid isn't so bad I would assume. Have good stay there.
Mar 26 14:36:01 <tbm> likewise: thanks
Mar 26 14:36:22 <likewise> dwery: well I am too scared coding a network notification handler right away :-)
Mar 26 14:36:51 <dwery-away> it is easy, I will help :)
Mar 26 14:37:14 <likewise> dwery: count = 0 because loft_eeprom_add() probably fails for me (eeprom does not have a valid MAC there). That triggers the count bug.
Mar 26 14:38:04 <likewise> div0 bug rather
Mar 26 14:38:13 <dwery-away> that function should not be called if you are not running on a loft
Mar 26 14:38:30 <dwery-away> well.. you _are_running a loft, right?
Mar 26 14:41:28 <likewise> IXDP425 :-)
Mar 26 14:41:53 <dwery> so that function should not be called.
Mar 26 14:42:17 <likewise> I am using a BE loft image with mach-type 245 (IXDP425).
Mar 26 14:42:42 <likewise> was too lazy to add IXDP425 to the Makefile
Mar 26 14:42:52 <likewise> should have done that though
Mar 26 14:43:06 <dwery> I still can't see why it is called
Mar 26 14:43:16 <dwery> if the mach type != loft
Mar 26 14:43:22 <likewise> it is not, my bad.
Mar 26 14:43:57 <dwery> anyway, the netdev code should look like this:
Mar 26 14:44:11 <dwery> a) add eeprom notification to read and store mac address when called
Mar 26 14:44:30 <dwery> b) add netdev notification to read the macaddress from the stored location and place it in the netdev itself
Mar 26 14:45:24 * likewise is confused now: arch/arm/mach-ixp4xx/ixdp425-setup.c has loft_ functions all over it...
Mar 26 14:45:52 <dwery> because loft is derived fromn that board. but since loft specific code is growing
Mar 26 14:46:03 <dwery> we will have to move it into its own file.
Mar 26 14:48:09 <likewise> understood
Mar 26 14:48:29 <likewise> I will look at [g2] :-)
Mar 26 14:51:43 <likewise> dwery: fix is to replace key % count with "count? key % count: 0"
Mar 26 14:52:17 <dwery> I think i'll rathe rimplement the netdev proposal :)
Mar 26 14:53:23 <likewise> it's a cheap solution to fix this bug for me now. The div0 bug triggers kernels due to scheduling while atomic, so I cannot test beyond this.
Mar 26 14:53:38 <dwery> of course.
Mar 26 14:53:49 <likewise> Should the interface occur as eth0 or as ixp0 ??
Mar 26 14:54:05 <likewise> I can ifconfig eth0 after this. but udhcpc complains:
Mar 26 14:54:09 <NAiL> *sigh*
Mar 26 14:54:18 * NAiL still can't get the bugger properly JTAG'ed
Mar 26 14:54:20 * likewise says uh uh
Mar 26 14:54:22 <likewise> ifconfig: 0.0.0.0: error fetching interface information: Device not found
Mar 26 14:54:29 <dwery> if you are using the latest patches from the cvs it should be eth0
Mar 26 14:54:37 <dwery> NAiL: can you test JTAG with an nslu2?
Mar 26 14:54:44 <likewise> dwery: I am
Mar 26 14:55:01 <NAiL> I don't have a JTAG-enabled nslu2 here... I'll have to make one then ;)
Mar 26 14:55:22 <dwery> likewise: not only the kernel cvs but also the ixp4xx one, ,right?
Mar 26 14:55:44 <NAiL> uhm
Mar 26 14:55:46 <likewise> uh no.
Mar 26 14:56:04 <NAiL> What is wrong *now*
Mar 26 14:56:12 <dwery> likewise: in the same cvs (nslu2 on sf.net) there's the ixp4xx repo
Mar 26 14:56:12 <NAiL> What is wrong *now*Firefox can't find the server at www.nslu2-linux.org.
Mar 26 14:56:33 <likewise> I am working userland from buildroot, not from oe.
Mar 26 14:56:38 * likewise ducks
Mar 26 14:58:23 <likewise> nslu2? http://cvs.sourceforge.net/viewcvs.py/nslu/
Mar 26 14:58:35 <dwery> likewise: i'm not using oe too
Mar 26 14:58:49 <likewise> dwery: phew
Mar 26 14:59:32 <dwery> yes, the page should be that one
Mar 26 14:59:41 <likewise> dwery: I work against a checked out kernel/
Mar 26 14:59:55 <likewise> there is no ixp4* at that level.
Mar 26 15:00:57 <likewise> Do you mean "/kernel/patches/ixp400-eth/" ?
Mar 26 15:01:35 <dwery> i think so, yes. It's been a while since I last built ixp4xx eth
Mar 26 15:01:40 <dwery> i debug using usb adapters
Mar 26 15:01:53 <NAiL> dwery/likewise: Do you know if I can get any useful info by using a scope on the JTAG port?
Mar 26 15:02:21 <dwery> NAiL: probably not.
Mar 26 15:02:28 <dwery> you should jtag your nslu2 :-D
Mar 26 15:05:21 <likewise> dwery: which one specifically? I am looking for a fast adapter.
Mar 26 15:05:47 <likewise> you tried openocd on your usb debugger?
Mar 26 15:05:57 <NAiL> no
Mar 26 15:06:37 <dwery> likewise: rtl8150 based iirc
Mar 26 15:07:22 <CIA-10> 03repvik * r298 10/releases/slugos-3.6-beta/openembedded/packages/meta/slugos-packages.bb: slugos: Add madplay to feed
Mar 26 15:07:24 <dwery> dunno if it is fast
Mar 26 15:07:45 <dwery> likewise: I meant usb/eth not usb/jtag , sorry
Mar 26 15:09:59 <CIA-10> 03repvik * r299 10/releases/slugos-3.6-beta/openembedded/conf/distro/slugos-packages.conf: slugos: add forgotten cups .bb's
Mar 26 15:10:04 <likewise> dwery: ok, an Ethernet over USB adapter?
Mar 26 15:10:28 <dwery> likewise: yes.
Mar 26 15:10:42 <dwery> it is easier to debug with that one instead of recompiling ixp4xx each time I change the kernel
Mar 26 15:11:08 <likewise> dwery: Ah, so that USB->Eth stuff is on your target?
Mar 26 15:11:31 <likewise> Target>(USB->ETH)>network>dev.host ?
Mar 26 15:11:51 <dwery> now I'm confused :)
Mar 26 15:12:05 <dwery> ah ok
Mar 26 15:12:08 <dwery> yes, it is.
Mar 26 15:12:17 <likewise> ok :-)
Mar 26 15:12:46 <likewise> I am going back to 2.6.15 to see if I can get rid of the "Device not found" problem.
Mar 26 15:15:45 <likewise> dwery: you got a EP93xx as well? :-)
Mar 26 15:15:58 <dwery> yes
Mar 26 15:15:59 * likewise is still deciding on which one to buy
Mar 26 15:16:24 <likewise> http://www.glomationinc.com/products_9312-sx.html
Mar 26 15:16:28 <likewise> or the TS-7250
Mar 26 15:18:12 <dwery> dunno... 
Mar 26 15:42:36 <pepijn> ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,00000000000000000000000000000000000000000000000000000000000000000 0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Mar 26 15:42:36 <pepijn> 000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Mar 26 15:42:37 <pepijn> 00000000
Mar 26 15:42:47 <pepijn> oops
Mar 26 15:42:59 <pepijn> son pressing on keyboard ;)
Mar 26 15:44:48 <mwester> Well, he has that character learned, now 35 more letters and numbers, and he can move on to puncuation!  :D
Mar 26 15:45:25 <pepijn> he has three extra vowels to learn as well (him being norwegian)
Mar 26 16:09:16 <likewise> I read about keyboard-cat-detection software. 
Mar 26 16:09:32 <likewise> (i.e. deny keys entered by your cat walking over the keyboard)
Mar 26 16:09:36 <likewise> :-)
Mar 26 16:11:47 <pepijn> yes, have read that also
Mar 26 16:12:02 <pepijn> but luckily, the cat stays away from the keyboard ;)
Mar 26 16:15:11 <mwester> ~seen ByronT
Mar 26 16:15:14 <jbot_> byront <i=byron-po@nslu2-linux/ByronT> was last seen on IRC in channel #nslu2-linux, 2d 20h 15m 45s ago, saying: 'I just want a loft....'.
Mar 26 16:21:12 <likewise> pepijn: not here, only the first year my keyboard was cat free, no it dumbly even lies down straight on the keyboard.
Mar 26 16:21:18 <likewise> no=now
Mar 26 16:44:46 <pepijn> ByronT: famous last words ?
Mar 26 16:48:26 <dwery> likewise, [g2]: I wrote the first part of the new netdev proposal and I'm reading out macs correctly
Mar 26 16:49:07 <likewise> dwery: you naughty coder
Mar 26 16:49:22 <likewise> dwery: you almost swore you didn't have the time, and now this :-)
Mar 26 16:49:36 <dwery> I delayd it too much
Mar 26 17:04:36 <likewise> is anyone actually running a 2.6.16 with the IXP drivers?
Mar 26 17:11:54 <eFfeM> likewise, i think i am
Mar 26 17:12:22 <likewise> without problems over wires Ethernet?
Mar 26 17:12:25 <eFfeM> i'm running the head (build earlier today) but with a hacked defconfig
Mar 26 17:12:26 <eFfeM> yes
Mar 26 17:12:51 <eFfeM> Linux version 2.6.16 (frans@LinuxPC) (gcc version 3.4.4) #1 PREEMPT Sun Mar 26 10:34:25 CEST 2006
Mar 26 17:13:49 <likewise> ok, then I probably ran into a IXDP425 only bug...
Mar 26 17:13:55 <likewise> thanks
Mar 26 17:14:04 <eFfeM> i'm on a slug
Mar 26 17:14:09 <eFfeM> np
Mar 26 17:17:58 <dwery> eFfeM: can you please cat /proc/interrupts on your slug?
Mar 26 17:19:16 <eFfeM> sure, hope I won't be kicked for flooding:
Mar 26 17:19:17 <eFfeM>            CPU0
Mar 26 17:19:17 <eFfeM>   0:          0   osal
Mar 26 17:19:17 <eFfeM>   1:        312   osal
Mar 26 17:19:17 <eFfeM>   2:          0   osal
Mar 26 17:19:17 <eFfeM>   3:          0   ixp400_eth
Mar 26 17:19:18 <eFfeM>   5:    1616844   IXP4xx Timer Tick
Mar 26 17:19:20 <eFfeM>  11:          0   ixp4xx-beeper
Mar 26 17:19:22 <eFfeM>  15:        412   serial
Mar 26 17:19:24 <eFfeM>  18:   63576604   ixp400_eth
Mar 26 17:19:26 <eFfeM>  22:          0   NSLU2 power button
Mar 26 17:19:28 <eFfeM>  26:      86735   ehci_hcd:usb1
Mar 26 17:19:30 <eFfeM>  27:          0   ohci_hcd:usb3
Mar 26 17:19:34 <eFfeM>  28:          0   ohci_hcd:usb2
Mar 26 17:19:36 <eFfeM>  29:          0   NSLU2 reset button
Mar 26 17:19:38 <eFfeM> Err:          0
Mar 26 17:19:56 <dwery> thanks a lot.
Mar 26 17:37:34 <Snide> What kernel version should I use from the debian installer? I believe my slug will not allow me to SSH in because its missing the module for the eth0
Mar 26 17:37:48 <dwery> [g2]: ping
Mar 26 17:40:49 <Snide> Do I need to run the 2.6.15.1?
Mar 26 17:40:56 <Snide> I don't have a USB ethernet
Mar 26 17:49:38 <Daluv> NAiL: you alive?
Mar 26 17:51:22 <Daluv> NAiL: I have been talking with you lately about xfs/jfs support in openslug.  I tried to register at slugbug and assigned the bugs to you but it doesnt let me register.
Mar 26 21:52:23 <dwery-away> [g2]: ping
Mar 26 22:20:55 <CIA-10> 03bzhou * 10unslung/make/swi-prolog.mk: upstream upgrade to 5.6.9
Mar 26 23:00:30 <marceln> Can somebody help me with the optware "autobuild"?
Mar 26 23:00:47 <marceln> I can't find "nmap" any more after my changes.
Mar 26 23:01:32 <marceln> I checked the logfile "http://logs.nslu2-linux.org/buildlogs/autobuild-nudi-last.txt" but didn't find a error for the optware nmap.
Mar 26 23:03:41 <dwery-away> rwhitby: ping
Mar 27 00:39:19 <eno-away> marceln, seems nmap was lost in the optware/Makefile for a long time
Mar 27 00:52:55 <eno-away> Makefile -r1.745 => -r1.746, we lost quite some packages
Mar 27 04:39:28 <mwester> ~seen ByronT
Mar 27 04:39:33 <jbot_> byront <i=byron-po@nslu2-linux/ByronT> was last seen on IRC in channel #nslu2-linux, 3d 8h 40m 8s ago, saying: 'I just want a loft....'.
Mar 27 06:13:57 <eddyp> openslug: I compiled my own kernel with usb_storage bultin, (deps, too), but when I insert the hdd  the only device created is /dev/sda
Mar 27 09:03:09 <CIA-8> 03rwhitby * 10unslung/Makefile: Demoted libmemcache for ds101g.
Mar 27 09:04:04 <CIA-8> 03rwhitby * 10unslung/Makefile: Demoted gambit-c for mss.
Mar 27 10:47:18 <CIA-8> 03rwhitby * 10kernel/2.6.16/40-rtc-class.patch: Fixed errant patch context
Mar 27 10:47:32 <CIA-8> 03rwhitby * 10kernel/2.6.16/defconfig: Updated for 2.6.16 final
**** ENDING LOGGING AT Mon Mar 27 10:59:56 2006