00:00.11 | T0mW | memvars |
00:00.39 | jacques | it was challenging - when we started the toolchains hadn't even been tested in BE (mostly) |
00:00.48 | jacques | and definitely not armeb softfloat |
00:01.14 | T0mW | IIRC, there were type-coercion issues with big endian which little endian avoids |
00:01.16 | jacques | there were problems with glibc too |
00:01.43 | jacques | now we are in the process of getting a new arch (armeb) added to debian (officially) |
00:02.14 | T0mW | uint foo; uchar bar; if (foo != (uint)bar) .... |
00:02.44 | T0mW | something like that anway |
00:02.51 | T0mW | s/anway/anyway/ |
00:03.06 | T0mW | jacques: but, you have to deal with it, right, heh |
00:03.37 | T0mW | jacques: so, what are you doing that is usefull with your unit? |
00:04.49 | jacques | useful? I don't do useful :-) |
00:05.23 | jacques | but seriously, I was using mine to compile things that don't cross-compile properly - like perl and emacs |
00:06.11 | jacques | since I've been moving around last few months I haven't done much with the units themselves tho I have been playing a lot with emulators lately |
00:06.52 | T0mW | oh, where are you earning your rent money lately? |
00:07.27 | jacques | I haven't earned money in quite a while - was living on savings in oregon and now am in the process of moving around and trying to find a job |
00:09.06 | jacques | BIAB |
00:11.05 | T0mW | sigh |
00:11.22 | T0mW | ld still insists on linking with the wrong libraries |
00:20.13 | thraxisp | jacques: what toolchain are you using? |
00:20.45 | T0mW | my own |
00:20.48 | T0mW | gcc |
00:21.31 | T0mW | looks like when I build gcc, the multilib option means both thumb + thumb_iw, but not plain arm32 library for libgcc |
00:21.48 | T0mW | newlib is building all three |
00:29.15 | *** join/#elinux hufnus (n=slonsiki@DSL135-071.labridge.com) |
00:32.22 | jacques | back |
00:33.04 | jacques | thraxisp, crosstool and oe generated |
00:34.39 | jacques | since we have multiple dists (one is backwards compatible with linksys binaries) we use a gcc 3.3/ glibc 2.2.5 and a gcc 3.4 / glibc 2.3.x |
00:34.58 | jacques | we also have uclibc based dists |
00:39.32 | thraxisp | thanks. I'm trying to update tools for a similar platform. |
00:41.55 | thraxisp | I was trying to build them with buildroot. The toolchain we have is missing some libraries. |
00:42.22 | jacques | for uclibc based, buildroot should be good |
00:43.03 | thraxisp | I've been working on one for a few weeks. The big-endian and no FPU seems to be a problem for buildroot. |
00:44.09 | jacques | hmm, I had built armeb uclibc softfloat toolchains with it in late 2004 |
00:44.18 | jacques | well like sept 2004 |
00:44.53 | jacques | well I'm 99% sure they were softfloat |
00:46.03 | thraxisp | recent versions of buildroot seem to be problematic. |
00:46.18 | jacques | ah I wonder what changed |
00:47.59 | thraxisp | Oct 15 fails and I left Oct 1 building at work. |
00:49.18 | jacques | what type of errors are you seeing? |
00:51.45 | thraxisp | a few specific errors. The most recent ones are mixed host and target libraries when building libstdc++. |
00:52.04 | thraxisp | There is also a problem building a shared uclibc. |
00:52.11 | jacques | ugh, I've seen stuf flike that - it required patches I think |
00:52.17 | thraxisp | both were reported in the tracker. |
00:54.28 | thraxisp | I have a patch for the latter but it doesn't make sense. There are functions missing from libgcc when uclibc is compiled. |
01:01.48 | *** join/#elinux noclouds (n=mhfan@60.166.49.146) |
01:40.08 | *** join/#elinux noclouds (n=mhfan@60.166.49.146) |
02:06.20 | *** join/#elinux megapixel_ (n=megapixe@p54B3F51A.dip.t-dialin.net) |
02:21.27 | *** join/#elinux T0mW (n=Tom@24.238.70.125.res-cmts.sth.ptd.net) |
02:37.06 | *** join/#elinux hufnus (n=slonsiki@DSL135-071.LABridge.com) |
02:42.46 | *** join/#elinux Soopaman (n=soopaman@CPE00112f7ebeec-CM000a739b3c5c.cpe.net.cable.rogers.com) |
02:44.58 | *** join/#elinux orifice- (n=orifice0@host-222-126-79-184.dhcp.infocom.ph) |
02:47.12 | *** join/#elinux noclouds (n=mhfan@60.166.49.146) |
02:57.53 | *** join/#elinux weo1 (n=weo@p5499ED5B.dip.t-dialin.net) |
04:31.02 | *** join/#elinux megapixel__ (n=megapixe@p54B3F51A.dip.t-dialin.net) |
04:50.14 | *** join/#elinux Soopaman (n=soopaman@69.156.148.136) |
04:53.29 | *** join/#elinux Soopaman- (n=soopaman@CPE00112f7ebeec-CM000a739b3c5c.cpe.net.cable.rogers.com) |
04:57.33 | *** join/#elinux jipi (n=jipi@cm191.delta245.maxonline.com.sg) |
06:33.22 | *** join/#elinux fog_proxy (n=user@mail.sernet.com.cn) |
06:33.52 | fog_proxy | hi all.I am searching for a CLI emaxple code,any ideas? |
06:38.58 | Russ | #include <stdio.h> |
06:39.06 | Russ | int main(void) |
06:39.08 | Russ | { |
06:39.16 | Russ | <PROTECTED> |
06:39.20 | Russ | <PROTECTED> |
06:39.21 | Russ | } |
06:39.30 | Russ | anything else? |
06:39.48 | Russ | oh, oops, have a typo, extra " |
06:41.53 | fog_proxy | Russ: ... |
06:42.08 | Russ | in the printf line |
06:42.15 | fog_proxy | Russ: I want a example which support telnet and complex cli functions |
06:42.16 | Russ | there's an extra quote before the \n |
06:42.50 | Russ | ~# telnet remotehost |
06:42.57 | Russ | username: someuser |
06:43.03 | Russ | password: ******** |
06:43.05 | fog_proxy | Russ: faint |
06:43.23 | Russ | someuser@remotehost ~# ./helloworld |
06:43.28 | Russ | hello world |
06:43.32 | Russ | someuser@remotehost ~# |
06:43.37 | Russ | k, we have telnet support |
06:43.39 | Russ | what next? |
06:57.56 | *** join/#elinux Soopaman- (n=soopaman@CPE00112f7ebeec-CM000a739b3c5c.cpe.net.cable.rogers.com) |
07:14.18 | Russ | fog_proxy: you want getopt support, what? |
07:16.05 | *** join/#elinux neofob|dsl (n=neofob@pool-71-254-64-152.ronkva.east.verizon.net) |
07:35.05 | *** part/#elinux fog_proxy (n=user@mail.sernet.com.cn) |
07:44.40 | *** join/#elinux noclouds (n=mhfan@60.166.49.146) |
10:00.35 | *** join/#elinux CosmicPenguin (n=nobody@aus-ext-proxy02.amd.com) |
10:00.35 | *** mode/#elinux [+v CosmicPenguin] by ChanServ |
10:18.12 | *** join/#elinux zinosat (n=Davide@adsl-ull-238-138.44-151.net24.it) |
10:53.58 | *** join/#elinux ade|desk (n=adavey@213-232-83-137.dsl.prodigynet.co.uk) |
10:58.17 | *** join/#elinux Soopaman- (n=soopaman@CPE00112f7ebeec-CM000a739b3c5c.cpe.net.cable.rogers.com) |
12:35.10 | *** join/#elinux chewy (n=dieu@r2351064.cidc.net) |
12:37.46 | *** join/#elinux prpplague (n=billybob@72.22.144.105) |
12:37.46 | *** mode/#elinux [+o prpplague] by ChanServ |
12:41.37 | prpplague | ho ho ho |
13:26.22 | *** join/#elinux Soopaman- (n=soopaman@CPE00112f7ebeec-CM000a739b3c5c.cpe.net.cable.rogers.com) |
13:48.07 | *** join/#elinux zinosat (n=Davide@adsl-ull-238-138.44-151.net24.it) |
14:10.00 | *** join/#elinux GPSFan (n=Ken@65.121.49.208) |
14:41.22 | *** join/#elinux Soopaman (n=soopaman@66.59.138.125) |
14:43.15 | prpplague | anyone have access to a hp laser printer with a duplexer? |
14:46.09 | *** join/#elinux supc (n=kennysu@sw169-242-104.adsl.seed.net.tw) |
15:42.27 | *** join/#elinux thraxisp (n=thraxisp@ottgate.precidia.com) |
16:24.56 | prpplague | interesting groklaw seems to be down |
16:40.34 | *** join/#elinux Genesis (n=genesis@dev.aimao.org) |
16:41.53 | Genesis | bonsoir |
17:37.46 | *** join/#elinux joe_bleau (n=joe_blea@adsl-69-152-195-74.dsl.fyvlar.swbell.net) |
17:52.14 | *** join/#elinux jipi (n=jipi@cm191.delta245.maxonline.com.sg) |
18:12.25 | prpplague | anyone going to the CELF conference? |
18:46.46 | *** join/#elinux jamie (n=jamie@pdpc/supporter/active/jamie) |
18:50.41 | jamie | Is anyone using uclinux-2.4.26 (or thereabouts) with a hard disk and ext3? |
18:51.36 | jamie | I'm seeing data loss and other problems, and wonder if it's just my system or a known uclinux issue. |
18:53.10 | prpplague | jamie: harddisk specific or ata interface such as a cf card to ide adapter? |
18:53.43 | jamie | Ordinary hard disk - 40G Travelstar. |
18:53.59 | jamie | This is on a proprietary ARM board. |
18:54.38 | prpplague | jamie: sorry no, all i have is cf used as hd's |
18:54.45 | jamie | With potentially dodgy IDE drivers or even dodgier electronics. However, it might be uclinux memory management or scheduling. |
18:54.55 | jamie | Have you used ext3 on them? |
18:55.09 | jamie | I'm wondering if ext3's journalling is not working properly with uclinux 2.4.x. |
18:55.25 | jamie | (cf should be no different than hd for this). |
18:55.52 | prpplague | jamie: i've used it in the past however i was worried about buring up the cf flash so i switched to ext2 |
18:56.12 | prpplague | jamie: i didn't seem to have any problems that i noticed |
18:56.16 | jamie | I'm more worried about booting to a corrupt disk image, considering I write logs to it quite often. |
18:56.28 | jamie | IT seems to be fine while it's running. |
18:57.08 | jamie | But if I change a file / create a file / whatever, then do "sync" which should commit it, then power cycle, often the changes are absent. |
18:57.17 | jamie | And sometimes I get I/O errors which indicate fs corruption. |
18:57.28 | jamie | Both of those should never happen, in principle, with ext3. |
18:57.58 | jamie | I also find that if I write to the disk _too fast_, the system crashes. I suspect the queuing of I/O requests or dirty data is a bit buggered in uclinux 2.4.x. |
18:58.22 | cdm | I think the ipod linux boys would have seen that if it was the case. |
18:58.23 | prpplague | hmmm |
18:58.25 | jamie | (Because I suspect that the system is filling with dirty data, then panics as it can't allocate any more). |
18:58.27 | cdm | sounds like a bad ATA driver. |
18:58.49 | jamie | It could be. One out of 13 boards corrupts data regularly, though I haven't seen it with the others. |
18:58.50 | prpplague | i'm with cdm |
18:59.05 | jamie | And it is a proprietary driver whose code does not inspire confidence. |
18:59.11 | prpplague | jamie: noisy trace? |
18:59.31 | prpplague | jamie: maybe the pcb has a minor trace problem |
18:59.31 | cdm | If the driver was chewing up memory the OMM killer would still run before the system died. |
18:59.42 | jamie | prpplague: Seems likely there are quite a few, given the various I2C / IDE / random lockup problems on various instances of this board :) |
19:00.27 | jamie | I'm just trying to pin down the problem, and checking if it's a known problem. |
19:01.00 | jamie | cdm: I was wondering if it was a generic I/O buffer management thing, given that uclinux changes that area of linux quite significantly. I'm not sure how mature the 2.4.26 patches were. |
19:01.10 | jamie | Alas, using 2.6 would be nice but not an option. |
19:01.26 | prpplague | jamie: i feel for ya, i spent the first week of this month tracing down an sdram init issue that turned out to be a badly implemented buffer |
19:01.49 | prpplague | jamie: can you scale down the main clock? |
19:02.09 | prpplague | jamie: i.e. if you are running at 200mhz might try running the system at 100mhz |
19:02.16 | prpplague | jamie: see if you have the same issue |
19:02.38 | jamie | prpplague: rumours are that it's possible. If I just wade through the vast directory of unfinished, dodgy, apparently abandoned proprietary code :/ |
19:02.40 | prpplague | jamie: most of the RF trace issues will not show up until after about 100mhz |
19:02.50 | cdm | jamie: the biggest users of uclinux that I know of are the ipod linux guys. And they have been running a 2.4 uclinux kernel driving ATA disks for sometime now. |
19:02.50 | prpplague | jamie: hehe |
19:02.53 | jamie | This thing ticks quite a bit higher than that. |
19:02.58 | jamie | It's a Sigma Designs video processor. |
19:03.04 | prpplague | ugh |
19:03.05 | prpplague | ugh |
19:03.06 | prpplague | ugh |
19:03.18 | jamie | I, of course, have faith in the h/w team to ensure noiseless operation :) |
19:03.39 | joe_bleau | jamie: where is sigma located? I think I may have known someone that worked there |
19:03.43 | jamie | And then I noticed a blatant ripple on the VGA output which rather dented my confidence in the h/w team ;-) |
19:03.44 | prpplague | jamie: hehe, i blame the hardware first |
19:03.54 | jamie | This board isn't from Sigma. |
19:04.06 | jamie | Just the main chip, and most of the code I have. |
19:04.28 | prpplague | jamie: ahh, well sigma for me is associated with their many resellers who are non-GPL compliant |
19:04.59 | jamie | joe_bleau: Somewhere in the USA. I don't know, they didn't respond to a direct attempt to buy stuff - we ended up getting our bits through a Taiwanese company that's bigger than Sigma :) |
19:05.05 | joe_bleau | The sigma I'm thinking of is in PA. They do broadcast video hardware. |
19:05.33 | jamie | joe_bleau: Probably the same. sigmadesigns.com. Used to be (or have something to do with) RealMagic, who designed the kit. |
19:06.05 | jamie | prpplague: I remember at one time, they were threatened with a suit by the Xine folks for nabbing code. |
19:06.23 | joe_bleau | Nope, this isn't the sigma my friend worked for. |
19:06.24 | prpplague | jamie: you might check with cosmicpenguin over in #elinux, iirc before he left CenturySoftware, he did some sigma work |
19:06.31 | jamie | But I gather they replaced the offending code. I haven't seen any blatant GPL-violations in the code that I've got yet - the kernel drivers come with source and it's LGPL'd. |
19:06.55 | jamie | Though I can easily see downstream vendors not complying - but that's not their problem. |
19:07.06 | prpplague | jamie: their resellers are still listed on the busybox hall of shame, and are frequented on the gplviolations.org mailing list |
19:07.39 | jamie | prpplague: Fair point. I think the company I'm getting stuff from don't really comprehend the GPL, but at least they supplied me with code anyway. |
19:07.51 | prpplague | jamie: anyway, cosmipenquin might be able to help ya out some |
19:07.53 | jamie | I never did sign the NDA, as it didn't make sense considering the licensing of the code. |
19:07.55 | jamie | Cheers. |
19:08.46 | prpplague | jamie: i'd still look at droping the clock freq below 100mhz if you can |
19:08.52 | jamie | So, new variation on question: how do I check what's being sent to the IDE driver? |
19:09.11 | jamie | I.e. how to I tell the difference between I/O requests not being sent to the driver, and the driver not sending them to the disk? |
19:09.42 | jamie | prpplague: That would be good anyway. Though I fear video decoding won't look too impressive. |
19:10.58 | prpplague | jamie: well, you will laugh, but when i have a problem like that, i usually load the code down with printk's and log like crazy to a uart |
19:11.33 | jamie | No laughing, I was thinking of it. Don't much like the idea, as "make" in the painfully convoluted source tree takes an hour. |
19:12.08 | prpplague | jamie: yea |
19:12.11 | jamie | And the UART would severally affect the timing, which might hide the problem. Then again, at least I'll know if it hides the problem :) Good plan. |
20:13.40 | *** join/#elinux vittorio|770 (n=andreas@213.174.253.3) |
20:14.27 | prpplague | vittorio|770: hey, we don't like showboating around here! hehe you and your 770 |
20:14.53 | vittorio|770 | :) |
20:16.21 | prpplague | vittorio|770: what kind of foustian deal did you have to make to get that? |
20:17.10 | vittorio|770 | ive ordered it in same hour it was first orderable |
20:18.02 | vittorio|770 | pretty lucky |
20:21.02 | prpplague | hehe cool |
20:21.10 | prpplague | vittorio|770: sooooo...... how ya like it? |
20:21.21 | vittorio|770 | alot |
20:21.29 | prpplague | cool |
20:21.50 | vittorio|770 | browsing and email works nice via wlan |
20:22.01 | vittorio|770 | and xchat |
20:22.15 | vittorio|770 | screen is amazing |
20:22.21 | prpplague | vittorio|770: how ya like matchbox ? |
20:23.09 | vittorio|770 | usefull |
20:23.23 | prpplague | hehe |
20:24.04 | prpplague | hehe next monday, here mallum is with MB |
20:27.12 | vittorio|770 | he wrote it witin a week? |
20:28.00 | vittorio|770 | weekend - impressive |
20:29.27 | prpplague | vittorio|770: pretty much, but that was the first round, its been on the marker for awhile now and lots of dev has gone into it |
20:55.25 | *** join/#elinux LinuxCart (n=kvirc@212.106.228.166.adsl.jazztel.es) |
21:51.43 | *** join/#elinux |LinuxCa| (n=kvirc@212.106.228.166.adsl.jazztel.es) |
22:04.38 | *** join/#elinux bzzbzz (n=franco@MTL-HSE-ppp190398.qc.sympatico.ca) |
22:05.36 | *** join/#elinux lack_ (n=lack@142-165-83-134.sktn.hsdb.sasknet.sk.ca) |
22:25.23 | *** join/#elinux ramsay (n=lack@142-165-83-134.sktn.hsdb.sasknet.sk.ca) |
23:24.06 | *** join/#elinux file[laptop] (n=jcolp@mctnnbsa24w-142167049176.nb.aliant.net) |