00:38:39 | Sammy | morning :) |
01:09:58 | prpplague | morning sammy |
03:06:33 | Sammy | Russ : you have the blob 2.0.5 pre1 ? |
03:06:44 | Russ | nope |
03:11:05 | Sammy | 2.0.5 the latest update can work on puppy , but I try to change it let the early update still can work ,but something stuck ... |
03:11:32 | Sammy | I don't know if that the problem with because erikm rewrite the flash code |
03:12:27 | Sammy | or something wrong with the BCR control the serila port ? |
03:12:46 | Sammy | do you know if like this how to debug ? |
03:13:45 | Russ | how about your LED? |
03:13:58 | Sammy | work |
03:14:40 | Sammy | I tell it shine the time's I want , it quite do it ... |
03:15:05 | Sammy | but the terminal still don't show any screen on minicom ... |
03:15:06 | Russ | but no output on the serial port? |
03:15:19 | Russ | are you doing everything you need to enable it? |
03:15:23 | Sammy | yeap |
03:15:58 | Sammy | yeah , I write some code to init the BCR control ... |
03:16:34 | Sammy | and also use some code from erikm in assabet.c to control the BCR ... |
03:18:20 | Russ | does the board have a BCR? |
03:18:34 | Sammy | yes. |
03:20:05 | Sammy | I am sure now , because I have already make the blob work on it , but it's come from the erikm's last update (after rewrite the flash part) |
03:21:39 | Sammy | and some magic thing is if I config with assabet , it also can work .... |
03:22:14 | Sammy | but erikm say that's only sheer luck :) |
03:24:08 | Sammy | after the blob into memory, is that have some way to debug ? |
03:33:38 | prpplague | night sammy/Russ |
03:33:56 | prpplague | has been at work for 13 hours 37 minutes |
03:34:00 | Russ | do you hawe the schematic of the bcr? |
03:37:15 | Sammy | yeap |
03:40:02 | Sammy | and also init the same bit with RS232 and GREEN, RED etc... |
03:41:05 | Sammy | so after see the erikm's assabet.c , I even don't add anything in it to let it work... |
03:41:22 | Russ | what would assabet.c have to do with your board/ |
03:41:23 | Russ | ? |
03:42:14 | Sammy | om..because it's init the flash driver and bcr ... |
03:42:26 | Sammy | and puppy also need do that too... |
03:42:53 | Sammy | so I try copy the bcr function from assabet.c to puppy.c ... |
03:42:57 | Russ | what does your bcr do? |
03:43:36 | Sammy | let the RS232 bit enable , and the Green led on... |
03:43:46 | Russ | do you know what address your bcr is at? |
03:44:08 | Sammy | 0x12000000 |
03:44:25 | Russ | and which bit is for the rs232 tranciever? |
03:44:38 | Russ | and if its active low or active high? |
03:45:02 | Sammy | and RS232 is bit 13 and green led is 14 |
03:46:39 | Sammy | sorry RS232 is bit 12 and both of then are active low |
03:47:27 | Russ | ok, so you are setting both to 0? |
03:47:35 | Russ | hmm |
03:47:53 | Russ | are you sure the led functions aren't accidentally setting bit 12? |
03:48:48 | Sammy | which led function ? |
03:49:24 | Sammy | you mean the __initlist() function ? |
03:50:55 | Russ | ok, your are clearing bit 12 before the serial port gets used, right? |
03:51:26 | Sammy | yes |
03:52:11 | Russ | and you double checked that you are using the right serial portl/ |
03:52:12 | Russ | ? |
03:52:46 | Sammy | absolutely right ... |
03:53:30 | Russ | find the function that outputs to the serial port, and try setting it there, to see if its getting cleared somewhere |
03:54:52 | Sammy | ok :) |
04:11:33 | Russ | er, clearing it there, since its active low |
04:13:11 | Sammy | output to serial port ...must be in serial.c right ? |
04:13:58 | Sammy | so clean it there must be in serialinit() function right ? |
04:18:17 | Russ | I don't have a current copy handy |
04:18:21 | Russ | just dig around |
04:20:14 | Sammy | is let the blob naked now... |
04:20:57 | Russ | ? |
04:23:07 | Sammy | nothing , just use the software (Source Insight) to recheck the blob ^_6 |
04:25:23 | Russ | ? |
05:04:25 | amrutha | skanda: hai!! |
05:06:54 | amrutha | anyone knows if i can access the chat window a day or 2 back?? is there any provision for the same?? |
05:07:23 | Russ | only if something is keeping logs |
05:07:53 | amrutha | how do i find out about that?? |
05:14:51 | Russ | I don't think any of the bots are, why? |
05:17:01 | amrutha | Russ: are you familiar with the sleep/wakeup process ? |
05:27:30 | Russ | the log starts out as #nvidia, just skip down to the #blob part |
05:40:06 | skanda | wheres the log |
05:40:28 | Russ | I have a buffer that I saved |
05:41:03 | skanda | how can I get it |
05:41:50 | amrutha | join #nvidia |
05:42:20 | Russ | ? |
05:48:19 | amrutha | Russ: i got the log file..but how do i access only the blob chat room log file...?? |
05:48:32 | Russ | just search it for #blob, and start reading there |
05:48:39 | amrutha | okie.. |
08:01:17 | amrutha | anyone has erik's mail id?? |
08:17:58 | skanda | i have |
09:19:03 | seletz | hi |
09:20:48 | seletz | cvs help update |
09:20:53 | seletz | wrong window. |
09:31:37 | Sammy | Yo seletz ^.^ |
09:41:08 | erikm | hi |
09:44:00 | Sammy | hellooo erikm |
09:52:15 | Sammy | erikm : I'm lost in blob , help... |
09:53:36 | Sammy | I had already let the puppy on blob ,clean and run . but that's your last update ... |
09:54:49 | Sammy | I try use the early version to change the it and make it run ...but |
09:55:22 | Sammy | still stuck on serial port output :( |
09:56:42 | Sammy | but I use the function init_bcr() with __initlist() , but it's sounds not work ? |
09:57:22 | Sammy | beside this is that anything need to hand on ? |
09:57:49 | Sammy | if like now , how to debug that ? |
09:59:08 | erikm | the BCR is an assabet specific register. the assabet is the only board that has a BCR. your puppoy probably has something different |
10:00:28 | Sammy | no , and just I say , your last update both assabet and puppy config all can run ... |
10:01:02 | Sammy | so I think the puppy is 90% the same with assabet ... |
10:02:16 | Sammy | so I think is that anything other reason or some other place stuck it or something ... |
10:03:51 | Sammy | I don't know, maybe your experience can let me get some hint .... |
10:10:39 | seletz | compiling kernel |
10:10:46 | seletz | Sammy: what's the prob? |
10:11:12 | Sammy | do your board has BCR ? |
10:13:20 | Sammy | I think my question always stuck on __initlist(init_flash_driver) and __initlist(init_bcr) |
10:13:29 | seletz | Sammy: no, but something similar. Understand the following: The BCR "register" is located in an CPLD on the assabet only. Has noting to do with the Processor. Our Board has a CPLD, too. (CPLD is a programmable Hardware chip. One can desing registers, conters and other simple logic with them. We use it for adress-decoding, irq multiplexing and other board-specific tasks) |
10:14:08 | seletz | Sammy: so, when you do not have an assabet, you simply dont have a BCR and thus dont need init_BCR() |
10:14:21 | seletz | Sammy: just remove the line. |
10:14:30 | Sammy | ? |
10:14:37 | seletz | Sammy: in (hopefully) puppy.c |
10:15:04 | seletz | Sammy: what do you not understand? |
10:16:13 | Sammy | but in the erikm's last update when I remove __initlist(BCR) line it will _not_ show anything on serial port ... |
10:16:31 | seletz | one moment pls |
10:16:41 | Sammy | because this part , I copy from assabet.c ... |
10:17:22 | seletz | back |
10:17:26 | Sammy | and after I let the // move off , the screen show the blob RUN ... |
10:18:11 | seletz | Sammy: have you the schematics of puppy available? |
10:18:23 | seletz | Sammy: or at last some docs? |
10:18:30 | Sammy | yeap..so I think this line is necessary .... |
10:19:02 | seletz | Sammy: is there a cpld on your board? |
10:19:13 | seletz | Sammy: look at the schematics |
10:19:13 | Sammy | yes .. |
10:19:36 | seletz | Sammy: where to in adress space is this cpld decoded? |
10:19:52 | seletz | Sammy: look at some CSxxx line to the cpld |
10:20:03 | seletz | Sammy: CS1,CS2, CS3 or CS4 |
10:20:23 | Sammy | check... |
10:20:41 | seletz | Sammy: or does the puppy hw spec mention this? (its a hassle to check this out via schematics :( |
10:21:14 | seletz | Sammy: search something like "control register" |
10:21:30 | seletz | Sammy: or similar |
10:31:53 | Sammy | seletz : tell you a bad things the cpld connect each all of CSX but I don't know which one is that you want ? |
10:31:53 | Sammy | but the manual write that BCR Physcription address : 0x12000000 |
10:31:53 | Sammy | and list all bit what it use to do ... |
10:31:53 | seletz | Sammy: ok, that's what we're looking for. |
10:32:04 | seletz | Sammy: ok, do you have assabet's developer manual available? |
10:32:24 | seletz | Sammy: _not_ the sa11xx dev manual, but the assabet board manual. |
10:32:40 | Sammy | yeap ... |
10:33:28 | seletz | Sammy: ok, now compare 1) the adress you found in puppy with the assabet's BCR address, 2) the bit meanings of the assabet BCR and the Puppys CPLD register. |
10:33:58 | seletz | Sammy: if they look equal, then puppy has a CPLD similar to assabet. |
10:34:24 | seletz | Sammy: if not, then you probably want to check your hw guru and shoot him :) |
10:34:47 | Sammy | you mean this the bit 12= RS232_ON and bit 14= GREEN_LED |
10:35:15 | Sammy | and all active low ... |
10:35:25 | seletz | Sammy: yup, thats our final goal (for getting blob to run properly) |
10:35:57 | seletz | Sammy: Same as assabet BCR ?? |
10:36:28 | seletz | Sammy: cant find my assabet doc ATM, so you have to check it. |
10:36:34 | Sammy | yes all the __SAME__ |
10:36:58 | seletz | Sammy: ok, you just prooved that puppy has a BCR like assabet. |
10:37:10 | Sammy | yes , that's why I say this __initlist(puppy_init_bcr... is really need it ... |
10:37:29 | Sammy | and now I am double check with this thing's |
10:37:44 | seletz | Sammy: just for avoiding further confusion: don't name it BCR, name it something else. |
10:38:30 | seletz | Sammy: when you mention "BCR" then everyone thinks "assabet". |
10:38:43 | Sammy | ? why , but assabet.c use it as well |
10:39:08 | seletz | Sammy: yes, thats the reason. puppy is not assabet. |
10:39:26 | seletz | Sammy: or did your hw gurus copy the assabet _exactly_ |
10:39:48 | seletz | Sammy: thinking over it..... |
10:40:08 | seletz | Sammy: actually, it may be a bad idea to rename it. hmmmm. |
10:40:27 | Sammy | and by the erikm's sheer luck I can also use assabet's config to run puppy |
10:40:35 | seletz | Sammy: you have already added puppy.c and include/blob/arch/puppy.h ? |
10:40:51 | Sammy | all prepare .... |
10:41:15 | seletz | Sammy: add a puppy board definition in configure.in, just to make a difference. |
10:41:26 | Sammy | and config it with puppy ... and work , |
10:42:00 | Sammy | but I don't understand is why it can work , but why I use the early version can't ? |
10:42:31 | seletz | Sammy: then you _could_ name it BCR. But beware! you have to check where in the code BCR is used and make sure that it's used if PUPPY defined as well as when defined ASSABEt, clear? |
10:44:04 | seletz | fiddling with kernel |
10:44:47 | Sammy | this I can sure , beside this when I send the diff to you , if it's not ,you will also find it ..and tell me that's worng ... |
10:45:01 | Sammy | _early_ = erikm not rewrite the flash part |
10:45:30 | Sammy | so I think is that the problem with BCR ? or flash part ? |
10:45:37 | Sammy | ohh...lost ... |
10:46:20 | seletz | Sammy: always remember "Don't Panic" |
10:46:23 | seletz | Sammy: :D |
10:47:44 | Sammy | ok ok , just feel trance :-\ |
10:49:11 | Sammy | take some rest ..go eat dinner ... |
10:49:21 | seletz | good idea .... |
10:50:47 | Sammy | I want change my place to home ...see you later :) |
10:51:00 | seletz | bye ... |
10:51:25 | Sammy | later ^_^ .. |
12:09:19 | seletz | hi erik! |
12:09:44 | erikm | hi seletz |
12:10:20 | seletz | erikm: sorry for the confusion/unecessary work i caused yesterday :( It was a _very_ black day for me ... |
12:10:29 | erikm | hehe |
12:10:35 | erikm | no problems :) |
12:11:07 | seletz | erikm: kernel 2.4.13 is just 2 seconds before working with my board. |
12:11:41 | erikm | cool |
12:12:27 | seletz | erikm: btw, i fixed neposet/assabet handling in blob. Just added a "NEPONSET" board which just is a copy of assabet wit an additional "NEPONSET" define. then in linux.c when building memory tags i just "break" after the first bank for neponset. OK? |
12:13:05 | erikm | that's a hack that should work for now |
12:13:30 | seletz | erikm: is neponset worth more support? |
12:13:48 | seletz | erikm: i mean, does someone actually _use_ it? |
12:13:54 | erikm | I'm not sure. I'll ask rmk about it |
12:14:03 | seletz | ok |
12:37:46 | seletz | ok, my system boots with jffs2. Just cant logi because serial line does not work anymore. Hmmm. |
12:37:55 | seletz | thinking |
12:49:46 | erikm | continues hacking a demo |
12:55:26 | seletz | ok, i'm off now. SIGGIRLFRIEND :) |
12:55:44 | seletz | bye folks! |
12:58:53 | erikm | bye |
14:04:30 | prpplague | morning erikm |
14:04:56 | prpplague | erikm: we missed your enlighted presence yesterday |
14:05:10 | erikm | prpplague: morning |
14:05:15 | erikm | prpplague: I was working at home |
14:05:44 | prpplague | erikm: ya ya we know you were hung-over from doing a pub crawl |
14:06:12 | erikm | prpplague: no, it was just because I wanted to get something done. |
14:06:35 | prpplague | erikm: instead of ppl on irc harrassing you? |
14:07:39 | erikm | prpplague: people disturbin me with silly questions. unfortunately people know that I know a lot about unix, C, etc :( |
14:08:40 | prpplague | erikm: no silly questions from me today, got to spend all day doing the cisco thing |
14:08:56 | erikm | prpplague: I wasn't talking about you |
14:09:37 | erikm | ;) |
14:10:31 | prpplague | erikm: lol, thanks |
15:12:32 | sammy_wms | Oh Oh ... |
15:13:32 | sammy_wms | erikm: maybe you are talking the people just like me :( |
15:13:51 | erikm | sammy_wms: no, I was talking about my co workers |
15:14:10 | sammy_wms | ok , I will control the silly speed :) |
15:14:18 | erikm | sammy_wms: it's easy to ignore IRC. it's much more difficult to ignore people standing next to you |
15:15:14 | sammy_wms | yeah. like a FLY |
15:15:42 | sammy_wms | and you are the _candy_ |
15:16:04 | erikm | yup |
15:17:06 | sammy_wms | btw , can you suggest me , how to go on and get more use C to control the hardware ? |
15:17:28 | sammy_wms | I think before this , I always use C to write some App... |
15:17:48 | erikm | sammy_wms: you mean a low level app like diag, or a linux app? |
15:18:34 | sammy_wms | yes diag blob jflash etc... |
15:19:13 | erikm | jflash is a linux app, blob and diag are really low level apps that run without an OS |
15:19:30 | erikm | phone... |
15:20:58 | sammy_wms | omm..that must be _BLOB_ actaully say is that some of your code , I still don't know how it really control the hardware ... |
15:23:42 | sammy_wms | if the code use on high level I can ease to know what it down , but in some low level code expecially in blob(diag) , just a little hard to image what it really to do |
15:25:27 | sammy_wms | even you are really kind to write some hint beside the code ... |
15:31:20 | erikm | sammy_wms: most of the blob code doesn't really control the hardware. |
15:31:48 | erikm | sammy_wms: serial.c speaks with hardware, and led.c as well. |
15:32:26 | erikm | sammy_wms: and the flash drivers. |
15:32:40 | erikm | sammy_wms: and icache.c. but that's about it |
15:33:48 | erikm | sammy_wms: and some of the architectures need some hardware fiddling as well. like h3600.c and assabet.c |
15:34:13 | sammy_wms | erikm: but I think that code are just different the linux app ... |
15:34:42 | erikm | sammy_wms: it doesn't really differ that much with a kernel driver |
15:35:39 | erikm | sammy_wms: except that kernel drivers don't have a user interface ;) |
15:36:17 | sammy_wms | but it's all differ then all the window App , and network App... |
15:36:34 | erikm | yes, so? |
15:37:06 | sammy_wms | maybe I don't spend a lot of time in this way |
15:37:42 | sammy_wms | so I think mnaybe you can suggest me HOW to go through it ... |
15:38:02 | erikm | well, first of all don't develop an application that runs directly on the hardware |
15:38:45 | erikm | develop it to run under linux and your application will be portable to all other linux/unix systems |
15:40:18 | sammy_wms | ya.. |
15:40:25 | sammy_wms | and ? |
15:42:07 | sammy_wms | is that any book talk about this ? |
15:42:19 | sammy_wms | or some doc ? |
15:43:58 | sammy_wms | erikm: I think maybe if I can use this tool(C) as well as , then about the blob grow up "maybe" have a little help... |
15:45:19 | sammy_wms | because, after the blob grow up on puppy, the next step on puppy will let the blob can get kernel and ramdisk from the Ethernet... |
15:45:28 | erikm | sammy_wms: well, there are a couple of books that influenced my coding style |
15:46:00 | erikm | sammy_wms: I have code from rmk for getting kernel and ramdisk from ethernet. we only need to sort out the licensing issues |
15:46:42 | sammy_wms | what's that books ? |
15:47:01 | erikm | sammy_wms: I think I should put the titles on a web site, cause prpplague also asked about them |
15:47:19 | erikm | "The C programming language, ANSI edition", kernighan and ritchie |
15:47:30 | erikm | "The practice of programming", kernighan and pike |
15:47:49 | sammy_wms | book 1 I have :) |
15:47:56 | erikm | "Programming pearls, second edition", jon bentley |
15:48:12 | sammy_wms | write it down... |
15:48:28 | erikm | and to lesser extent: "The mythical man month, 25th anniversary edition", fred brooks |
15:48:33 | prpplague | erikm: i'll put a list on my site |
15:48:58 | erikm | prpplague: a list of books that I recommend? ;) |
15:49:05 | prpplague | erikm: ya |
15:49:11 | erikm | hehe |
15:49:52 | prpplague | "erik says if you don't read these books your an idiot!" |
15:51:08 | erikm | ehm I only said that those books influenced my programming style and knowledge |
15:51:22 | prpplague | just kidding with ya |
15:51:39 | sammy_wms | ok keep it in mind ... |
15:51:51 | prpplague | ./prplague --mode harass |
15:52:38 | prpplague | i've got so much stress on me today, irc and a good laugh is the only thing keeping me going |
15:55:53 | erikm | hehe |
15:58:26 | sammy_wms | erikm: about the download use ethernet , do you need to use some app like Ethload ? |
15:59:12 | erikm | sammy_wms: no, a bootp and tftp server will do |
16:00:00 | erikm | sammy_wms: and I have client code from rmk that can do arp, bootp, and tftp. (all in less than 200 lines of code) |
16:00:18 | sammy_wms | you mean the code eith in the blob ? |
16:00:52 | erikm | sammy_wms: yes, the client code that can be put into blob |
16:02:38 | sammy_wms | that's great, this will be better then Angel, and I think more guys will use blob more and more :D |
16:03:38 | sammy_wms | _BLOB_ small , speed more , strong , easy port ... |
16:04:48 | prpplague | amen |
16:05:19 | erikm | he |
16:05:28 | sammy_wms | maybe someday it's will be the normal bootloader for arm like the DOS for windows |
16:05:31 | erikm | LOL: http://www-1.ibm.com/servers/eserver/linux/fun/evolution.html?t=grdd%2cl=921%2cp=LinuxEvo |
16:08:26 | sammy_ | :( my x-chat is lock |
16:08:47 | sammy_ | erikm: can you send the address again ? |
16:09:45 | erikm | http://www-1.ibm.com/servers/eserver/linux/fun/evolution.html?t=grdd%2cl=921%2cp=LinuxEvo |
16:10:09 | prpplague | erikm: saw those last night, my son got a kick out them "look tux!" |
16:10:38 | erikm | prpplague: hehe, he learns fast :) |
16:11:12 | prpplague | erikm: he's had a tux doll since he was 1 |
16:12:27 | sammy_ | cute... |
16:14:41 | erikm | prpplague: alan cox also has tux dolls around his screen |
16:18:42 | erikm | prpplague: go to www.bbc.co.uk/newsnight/ , click on "latest program" and seek to about 22:20 |
16:20:13 | prpplague | erikm: no audio on this box |
16:20:20 | erikm | prpplague: :( |
16:20:45 | erikm | anyway, it features alan cox on a crypto/security story |
16:20:52 | prpplague | erikm: this is suppose to be a server, i'm using it as a desktop |
16:21:20 | erikm | hehe |
16:21:26 | prpplague | erikm: thats cool, i met ac at the 5th annual linux expo in raleigh, nc |
16:21:42 | sammy_wms | finally it's unluck....~o~ |
16:21:49 | sammy_wms | I am back |
16:24:46 | sammy_wms | thanx erikm ,tomorrow I'll search that's book , now I need some zzz ... |
16:25:00 | erikm | good zzz |
16:25:05 | sammy_wms | night erikm/prpplague |
16:25:17 | prpplague | night sammy_wms |
16:25:57 | sammy_wms | Z.z.z. |
16:27:02 | prpplague | erikm: ok when are you going to do an linux replacement for cisco ios??? |
17:56:21 | prpplague_lunch | erikm: i'm off to lunch, have a good weekend |
17:56:35 | erikm | ok, bye |
20:17:32 | erikm | BZFlag: FYI: rmk just told me that tim@rikers.org times out |
20:17:57 | BZFlag | erikm: my stinking ISP went south with no warning. |
20:18:11 | BZFlag | rikers.org moved IPs but seems to be lagging. |
20:18:14 | erikm | BZFlag: ah, ok. that explains a lot |
20:18:29 | erikm | BZFlag: send rmk a message so he won't unsubscribe you |
20:18:31 | BZFlag | Tim@Lineo.com will work for now, |
20:18:50 | BZFlag | but I am getting mtd mail on tim@rikers.org, just not lak mail. |
20:19:40 | BZFlag | I suspect he has a cached DNS record. They are set to timeout at 8 hours, so if he still has it he has a DNS cache issue. |
20:19:46 | erikm | just contact rmk, maybe mailman already unsubscribed you |
20:20:05 | erikm | (and contact him from your @lineo.com address) |
20:20:19 | BZFlag | I'm no worried. I'll resubscribe to any of the many lists I'm on once things settle down to the new address. |
20:20:35 | erikm | ok. it was just a FYI :) |
20:20:40 | BZFlag | you will note rdns thinks I'm at wwick.dsl.xmission.com instead or rikers.org |
20:20:44 | BZFlag | thanx! |
20:21:01 | erikm | weekend, cu |
20:21:07 | BZFlag | have fun |