| 07:37:37 | Sammy | russ_: help |
| 07:39:12 | Sammy | how can I currect use ldconfig make ld.so.cache into my designate dir ? |
| 07:39:34 | Sammy | ldconfig /jffs2-file ? |
| 07:39:51 | Sammy | or need some special command ? |
| 07:40:45 | Sammy | or ldconfig -C <directory> ? |
| 09:12:04 | Sammy | ha ha :) |
| 10:21:07 | prpplague | morning all |
| 10:25:27 | erikm | hello prpplague |
| 10:34:14 | Sammy | hi <prpplague | erikm > |
| 10:47:52 | prpplague | needs more coffee |
| 10:48:04 | prpplague | all nighter working on raid cabinet |
| 10:48:21 | erikm | prpplague: :( |
| 10:48:46 | prpplague | tell me about it, the only good thing is that i'm in barbados |
| 10:49:02 | erikm | heh, so what's the weather like over there? |
| 10:49:17 | prpplague | bout 85 with a little rain |
| 10:49:33 | erikm | so quite a high humidity |
| 10:49:58 | prpplague | i'm going to the hotel in about 3 hours and drink a beer and pass out on the beach |
| 10:49:58 | prpplague | ya pretty humid |
| 10:50:17 | prpplague | what about the UK today? |
| 10:50:32 | erikm | is in .nl |
| 10:50:43 | erikm | but that's pretty close |
| 10:51:18 | prpplague | hmm, why did i think you were in the uk, maybe because of alpeha |
| 10:52:07 | erikm | I guess it's about 5C, foggy (500m visibility) |
| 10:52:21 | prpplague | sounds like fun |
| 10:52:58 | erikm | better than a couple of days ago, where we head fog with -2C, so the fog froze immediately on cars :( |
| 10:53:06 | erikm | s/head/had/ |
| 10:53:58 | prpplague | ewww, sounds dangerious |
| 10:54:21 | erikm | *nod* |
| 10:54:37 | erikm | luckily it was on sunday morning, so not too many people on the road |
| 10:55:10 | prpplague | ppl from texas are really dangerious on the road when it gets icy |
| 10:56:13 | erikm | yeah, I've seen that. two weeks ago when I was in NY, they had a couple of winter storms in north texas. pretty scare pictures on the weather channel |
| 10:56:48 | prpplague | i'm based out of east texas |
| 11:00:13 | erikm | SIGLUNCH |
| 11:28:53 | prpplague | sammy: knee how |
| 11:31:30 | Sammy | prpplague: knee how, :-D |
| 11:31:57 | prpplague | wo han how |
| 11:32:10 | Sammy | 's iPAQ have brother .... hehe |
| 11:32:23 | Sammy | I by another one |
| 11:32:26 | Sammy | er buy |
| 11:32:55 | Sammy | only cost $289 us dollars |
| 11:33:02 | prpplague | thats a good price |
| 11:33:24 | prpplague | why not a girlfriend so they can multiply |
| 11:33:54 | Sammy | he he I know that |
| 11:34:21 | Sammy | that's why I want to buy another Tuxscreen ... |
| 11:34:33 | Sammy | I have two ipaq now... |
| 11:35:16 | Sammy | but ,the new one I will let WXX CX live until 2001.12.31 and then ... |
| 11:35:41 | Sammy | I'll change it with some QPE and QT App on it , |
| 11:36:01 | prpplague | sounds fun |
| 11:37:46 | Sammy | and this one is waiting erikm's blob stuff stable then his job is useing for blob test platform ... ^_^ |
| 11:45:16 | prpplague | erikm: find something good for lunch? |
| 11:46:15 | Sammy | suggest chinese food ... |
| 11:47:01 | prpplague | Sammy: we have lots of chinese food places in texas, but no india places |
| 11:47:50 | erikm | prpplague: just regular lunch, moderate amount of bread, some fruit, and milk. |
| 11:48:18 | erikm | closest chinese restaurant is 2km away |
| 11:48:38 | erikm | has been to an india place yesterday evening |
| 11:49:06 | erikm | we have lots of chinese restaurants (but where don't you have a chinese restaurant) |
| 11:49:28 | erikm | also lots of indonesian restaurants because indonesia is a former dutch colony |
| 11:51:14 | Sammy | prpplague: you must find it and test it once ... |
| 11:51:38 | Sammy | it's deliciou ... |
| 11:51:51 | prpplague | Sammy: chinese? i eat there alot |
| 11:52:33 | prpplague | Sammy: of course its kinda of american/chinese ( too much meat ) |
| 11:53:17 | prpplague | well the stores have started to open, keep your fingers crossed! |
| 11:54:38 | Sammy | prpplague: ohh, that's cool, |
| 11:55:06 | prpplague | hopes the new raid has no problems today |
| 11:55:42 | prpplague | 7:57am up 1:46, 34 users, load average: 0.00, 0.00, 0.00 |
| 11:55:42 | prpplague | User Tty Login@ Idle JCPU PCPU What |
| 11:58:08 | erikm | prpplague: what kind of raid is it? |
| 11:59:31 | prpplague | 4 cabinets with each cabinet containing 6 15000 rpm 36gig with rudundant adaptec hardware raid controlers |
| 12:00:12 | prpplague | raid-1 for every two drives |
| 12:04:55 | erikm | sounds nice |
| 12:06:10 | erikm | last raid I tuned was a 3 cabinet with each cabinet 4 10000 rpm 18GB drive, raid-1 for each cabinet |
| 12:07:48 | erikm | it was tuned for throughput. we needed at least 30MB/s sustained read and write throughput, I got 90MB/s per cabinet when using up to two cabinets, and 70 MB/s per cabinet when using all three |
| 12:08:28 | erikm | the 70MB/s was because I saturated the PCI 64/33 bus with the three scsi controllers :) |
| 12:14:07 | prpplague | erikm: this is not that good, we need to get a better design |
| 12:14:32 | prpplague | argh, permission problems at point of sale |
| 12:14:56 | seletz | hi folks! |
| 12:15:12 | prpplague | morning |
| 12:15:13 | prpplague | 8:17am up 2:06, 102 users, load average: 0.06, 0.07, 0.02 |
| 12:15:13 | prpplague | User Tty Login@ Idle JCPU PCPU What |
| 12:16:10 | prpplague | erikm: should see 1k+ users today |
| 12:18:31 | erikm | prpplague: it was good enough. the previous raid that MTI thought to sell us only reached 25MB/s when using all three voumes |
| 12:18:36 | erikm | s/voumes/volumes/ |
| 12:19:15 | prpplague | erikm: i might have get some info on your system layout |
| 12:19:26 | seletz | woha, heavy loads out there ... :) |
| 12:23:05 | erikm | prpplague: sgi onyx2, 4x MIPS r12k@300MHz with 8MB L2 cache each, 1GB memory. infinite reality graphics pipe, 3x digital video option (good for 20 or 40MB/s sustained data each), XIO to PCI 64/33 adapter with 3x QLocic QL2200 fibre channel controller each connected to a single raid cabinet |
| 12:23:41 | erikm | prpplague: main optimisation criterion was sustained throughput, not redundancy. |
| 12:35:36 | seletz | erikm: Did you found time to take a qick look at the pcmcia stuff? |
| 12:36:28 | erikm | seletz: not yet. I had a project dinner yesterday evening and fell asleep on the couch when I got home |
| 12:36:49 | seletz | erikm: np. |
| 12:37:21 | erikm | seletz: ... only to be awaken at 2am in the morning because of a terrible pain in my neck :( |
| 12:40:10 | seletz | erikm: ouch. |
| 12:40:46 | seletz | erikm: I dreamt of ld and aCC command line switches for HP-SUX shared library binding :( |
| 12:41:59 | erikm | seletz: that's bad. that's why I always read something before I go to sleep. |
| 12:44:36 | seletz | erikm: well, i usually do that too (ATM Terry Pratchet's "Truth"). |
| 12:45:35 | seletz | erikm: i normally dream only of weird things when something did not work and/or i worked too long on it. Well, guess thats our fate anyway. |
| 12:45:39 | seletz | :D |
| 12:49:04 | seletz | erikm: uh-oh. Just read your post on linux-arm-kernel 'bout this bootldr stuff. I guess this is all F**ed up ... |
| 12:49:32 | erikm | seletz: it is. bootldr wants to pass partition information through ATAG_ nodes |
| 12:49:53 | erikm | seletz: which is broken |
| 12:49:54 | seletz | erikm: ah, _this_ thread. |
| 12:49:59 | erikm | seletz: that one, yes |
| 12:50:13 | seletz | erikm: well, i thought rmk was your opinion? |
| 12:50:48 | erikm | seletz: he is |
| 12:51:22 | erikm | seletz: but I think he doesn't want to do that too much right now |
| 12:51:49 | erikm | reads his lkml backlog -- only 3300 to go |
| 12:54:44 | seletz | poking PCMCIA HW |
| 13:10:40 | prpplague | well the raid is holding up, but have some frame relay problems this morning |
| 13:14:58 | prpplague | 9:16am up 3:06, 182 users, load average: 0.02, 0.10, 0.09 |
| 13:24:20 | skanda | guys: any idea how can we implement -1 in Assembly |
| 13:24:31 | skanda | a negative one |
| 13:25:06 | seletz | well, use twos-complement or ones-complement. |
| 13:25:31 | skanda | but does Sa1110 have a specific instruction to do that |
| 13:25:54 | seletz | To do what? load a -1 in a reg? |
| 13:26:00 | skanda | yes |
| 13:26:18 | seletz | hmm. |
| 13:26:24 | skanda | well!!! |
| 13:26:34 | seletz | fetches SA Ref Book |
| 13:26:48 | seletz | one moment pls |
| 13:27:00 | skanda | ok, take your time |
| 13:30:51 | seletz | well, page A6-17 talks about signed and unsigned loads ... |
| 13:31:14 | skanda | just wait lemme look at it |
| 13:34:40 | seletz | hmm, why dont you just write something like ldr r1, #0xffffffff which is -1 in two's complement IIRC |
| 13:35:18 | seletz | i guess in assembler it is you who must take care if or if not data is signed. |
| 13:35:33 | seletz | Just use the richt arithmetic functions. Lets see ... |
| 13:35:33 | skanda | no I thought, suppose I implement a function to return a -1 in case of an error |
| 13:36:10 | erikm | skanda: mov r0, #0 |
| 13:36:12 | skanda | seletz: which book is this |
| 13:36:16 | erikm | skanda: sub r0, r0, 1 |
| 13:36:26 | erikm | ehm, sub r0, r0, #1 |
| 13:36:35 | skanda | I thought so, but then I thought it could gimme an error |
| 13:36:37 | seletz | erikm: oh well. yes. |
| 13:37:08 | erikm | skanda: and then print the -1 in hex |
| 13:37:13 | seletz | erikm: but #0xffffffff should work too (if it fits into the opcode, that is) |
| 13:37:31 | skanda | that will be 0xFFFFFFFD right |
| 13:37:37 | skanda | the ones complement |
| 13:37:38 | erikm | seletz: IIRC it sould |
| 13:38:03 | skanda | the ones complement of -1 or will it be 0xFFFFFFFF |
| 13:38:04 | seletz | skanda: you generally most probably want two's complement |
| 13:38:54 | seletz | skanda: because you can represent 1 additional number. with one's complement there are +0 and -0, which wastes a number. |
| 13:40:04 | seletz | skanda: have "Architecture Reference Manual" for the ARM? |
| 13:40:48 | skanda | i dont think so, can I download it from the net |
| 13:41:57 | skanda | so implemeting a twos complement of -1 will be 0xFFFFFFFE right |
| 13:42:08 | skanda | ( 0xFFFFFFFD + 1 ) |
| 13:42:27 | erikm | skanda: you can't download, just ask your local ARM representative for a copy (either hardcopy or PDF) |
| 13:43:53 | skanda | oh God, I dont think so I could do that |
| 13:44:20 | skanda | I have no idea where my institute sourced the Assabet from |
| 13:45:06 | seletz | skanda: no. its 0xffff. Look:suppose we have 4 bit then 1 is %0001. Negating that gives %1110, which is one's complement. then add 1, which gives %1111. |
| 13:45:07 | erikm | skanda: doesn't matter. Assabet is from Intel, and ARM doesn't care about that |
| 13:45:47 | erikm | skanda: you can also fill out the form on www.arm.com and ask for the ARM ARM. they'll send you the latest version as PDF file (18MB file, so watch out) |
| 13:45:58 | skanda | but thats one's complement, but what about 2 complement |
| 13:46:16 | skanda | i guess I can download that |
| 13:46:41 | seletz | skanda: no. %1110 is one's complement. by adding one you get two's complement and that gives %1111 |
| 13:47:02 | skanda | seletz: terribly sorry about that, how silly of me |
| 13:48:19 | seletz | skanda: np, i just took some paper to check it by myself, so ... :D |
| 13:52:19 | seletz | :D my collegues just testing M$ IE versus Netscape Java 2 performace (or lack thereof). Test program "surrived" 2 GUI clicks ... |
| 13:52:49 | seletz | very happy _not_ using java |
| 13:54:48 | seletz | erikm: ever heard about linux running on a Psion 5 mx pro ? |
| 13:55:34 | skanda | seletz: ltes say I want to transmit data on serial lines 3, so |
| 13:56:08 | skanda | hows the simplest way I can do that |
| 13:56:59 | seletz | skanda: use blob as an example |
| 13:57:23 | seletz | skanda: in configure.in you can select which serial line to use IIRC |
| 13:58:41 | skanda | sorry, i am a little new to this, whats IIRC |
| 14:00:49 | seletz | "if i remember correctly" :D |
| 14:01:15 | skanda | yeah, you are correct, it is defined in configure.in |
| 14:03:23 | erikm | seletz: I know about linux running on some kind of Psion. wookey is involved in that |
| 14:04:02 | seletz | erikm: just got an used psion, i'd like to try this out :D |
| 14:04:52 | prpplague | seletz: ya there is a project for the psion, however psion isn't exactly a linux friendly company, so co-operation has made the develop slow |
| 14:05:50 | prpplague | seletz: last time i check in on linux on psion about a year ago, psion was still telling the developers "oh ya we'll get those specs to you soon..." |
| 14:06:45 | seletz | prpplague: well, now that they dropped the whole handheld buisness, perhaps they really fork those specs over ... |
| 14:08:24 | prpplague | i didn't know the had quit the hh biz |
| 14:08:37 | prpplague | s/the/they |
| 14:09:00 | prpplague | 10:11am up 4 hrs, 205 users, load average: 0.30, 0.39, 0.33 |
| 14:09:08 | prpplague | not too bad for SCO |
| 14:10:27 | seletz | prpplague: well, they did. They now sell soley to "buisness customers". Some sort of rugged, sub-notebook style devices running guess it, a StrongARM at 206 Mhz. |
| 14:11:40 | prpplague | hmm |
| 14:11:52 | prpplague | probably costs around 5 or 6k |
| 14:12:41 | seletz | prpplague: well, they sell on ebay for about DM 1.400.- that would be $700 or so. Called "Psion NetBook" |
| 14:13:46 | seletz | hmm, there's a sourceforge project 'bot linux on psion ... |
| 14:13:58 | prpplague | hmm, i;'ll have to check it out, i wonder what the status of linux support is |
| 14:16:04 | prpplague | i wish i had more time to work on my project, i need to get a prototype finished soon |
| 14:47:23 | Sammy | change place :) |
| 14:47:38 | | sammy: no idea |
| 14:47:38 | Sammy | ibot: why not ? |
| 14:47:49 | Sammy | later all |
| 14:47:56 | erikm | bye |
| 15:06:32 | erikm | answers jamey |
| 23:22:56 | prpplague | evening all |