00:24.58 | *** mode/#elinux [+v file[laptop]] by ChanServ |
01:04.07 | *** join/#elinux noclouds (n=mhfan@60.166.47.24) |
02:03.40 | *** join/#elinux pigeon (n=pigeon@adsl-60-232.swiftdsl.com.au) |
02:14.18 | *** join/#elinux Soopaman (n=soopaman@69.156.149.27) |
02:51.58 | *** part/#elinux TorbaX (n=TorbaX@host-84-220-144-168.cust-adsl.tiscali.it) |
02:57.47 | *** join/#elinux we0 (n=weo@p5499976E.dip0.t-ipconnect.de) |
03:08.51 | *** join/#elinux Soopaman (n=soopaman@69.156.149.27) |
03:11.29 | Russ | cosmicpenguin is in dresden at amd? |
03:11.51 | file[laptop] | ^_^ |
03:35.46 | Genesis | bonne nuit |
03:46.51 | *** join/#elinux hufnus (n=slonsiki@DSL135-071.labridge.com) |
05:18.34 | T0mW | Russ: help! |
05:18.43 | file[laptop] | oh no, not help |
05:19.01 | T0mW | Russ: how do I decode what "stmia" and "stmdb" mean? |
05:19.49 | cdm | T0mW, get the ARM ARM |
05:20.04 | T0mW | cdm: got it, still cannot figure it out. |
05:20.33 | cdm | stmia - store multiple increment after |
05:20.35 | T0mW | "stm[<cond>]<addressing mode> |
05:20.42 | T0mW | thanks |
05:20.54 | T0mW | couldn't think of the mnemonic |
05:20.57 | cdm | stmdb - store multiple decement before |
05:21.07 | cdm | decrement |
05:21.10 | T0mW | WTF is that in the book, I looked under A3-5 |
05:22.08 | cdm | http://www.arm.com/pdfs/QRC0001H_rvct_v2.1_arm.pdf |
05:22.21 | cdm | page 5, Addressing Mode 4 table |
05:23.02 | T0mW | A3-21 mentions the modes for "load and store multiples", I'll have to dig deeper I guess. |
05:24.21 | T0mW | aha! |
05:24.28 | T0mW | page A5-48 |
05:24.37 | cdm | did you check that PDF? |
05:24.47 | cdm | the quick ref guide is a handy document to have around |
05:24.48 | T0mW | no, I've the paper copy at hand right now |
05:25.06 | T0mW | cdm: quick ref in the PDF? |
05:25.12 | cdm | the pdf I liked |
05:25.14 | cdm | linked |
05:25.17 | cdm | is a quick ref guide |
05:25.29 | cdm | from ARM |
05:26.48 | T0mW | cdm: printing it out now, thank you. |
05:27.06 | T0mW | I'm hand assembling my first "real" arm asm program.. :-/ |
05:27.39 | T0mW | I know what I want to do, trouble is saying it ARM |
05:27.44 | cdm | :) |
05:27.47 | cdm | I love ARM assembly |
05:27.50 | cdm | it's really nice |
05:28.10 | T0mW | arm != 80c188 != 8051 != m68k != <various> |
05:28.22 | T0mW | another language to learn <sigh> |
05:29.54 | T0mW | this program is almost done. Now, it has to detect that it was launched in low addr Flash, erase upper flash area and move itself up into that upper flash region. hence the assembly lang. |
05:30.41 | T0mW | hmmm, maybe I should take another look at the -fPIC flag and write this stupid thing in C. |
05:31.11 | T0mW | cdm: naw! it is time to learn arm asm! |
06:02.06 | *** join/#elinux Spetznaz (n=spetznaz@tac.mukachevo.net) |
06:26.27 | *** part/#elinux cdm (n=cdm@adsl-69-236-127-131.dsl.pltn13.pacbell.net) |
06:28.51 | *** join/#elinux cdm (n=cdm@adsl-69-236-127-131.dsl.pltn13.pacbell.net) |
06:50.59 | Russ | if you learn arm assembly, you immediately can pick up so many others |
06:51.24 | Russ | T0mW: did you get that 6 page quick reference yet? |
06:51.29 | T0mW | yes |
06:51.44 | T0mW | Russ: I already have many many other assemblers I know. |
06:51.44 | Russ | do you actually want to decode a hex instruction? |
06:52.09 | T0mW | naw, just couldn't understand organization of the reference material |
06:52.39 | T0mW | I've done a ton of 8051, 80c188, 6502, ..., ... |
06:52.57 | Russ | mips is really similar |
06:53.13 | T0mW | Russ: I just have to get my tail in gear and just do the ARM stuff. |
06:53.15 | T0mW | heh |
06:54.15 | T0mW | The aritmatic functions, branches are similar, but the cond<ition> code and shifter fields are new to me. |
06:54.33 | T0mW | like try this with 8051: mov r0, r0, lsr #12 |
06:54.40 | T0mW | no workee |
06:54.59 | Russ | well, given you don't have a 32 bit register |
06:55.03 | Russ | it'll take a while |
06:55.27 | T0mW | heh |
06:56.27 | MonMotha | T0mW: yeah, the whole predication thing was new to me |
06:56.36 | T0mW | I'm writing this flash mover in asm rather than c. have to take about 40K of code and move from 0x0000:0000 to 0x0006:8000 |
06:56.39 | MonMotha | and shifters |
06:57.20 | Russ | into ram? |
06:57.54 | T0mW | no, flash --> flash. got to erase before, prog into upper flash, erase lower |
06:57.59 | Russ | bummer |
06:58.28 | T0mW | not really, the bootloader has the commands to mangle the flash in it. you just make function calls. |
06:58.42 | Russ | I did some asm that wrote out one 32 bit word to a specific location, and some that parsed the cfi table to find the end of flash |
06:58.48 | Russ | but thats about it on the asm<->flash |
06:59.26 | T0mW | yeah, I've done some of that, but I wrote it in C, compiled it, then copied the asm code. heh |
06:59.49 | T0mW | this is from scratch so I can "punish" myself into knowing what I'm doing. |
07:00.35 | T0mW | sure is easier doing this stuff in 32bit instead of 8 / 16 though |
07:01.12 | T0mW | nice, big, fat registers to use! |
07:02.14 | T0mW | Russ: what still boggles my mind is the seperate sets of registers: FIQ, IRQ, Superv, etc. |
07:03.09 | MonMotha | T0mW: you've never done that on 8051? |
07:04.22 | T0mW | MonMotha: lots and lots of '51 asm |
07:04.54 | T0mW | MonMotha: I'll just have to think of this as an 8051 with 32bit registers and 8 potential DPTRs |
07:05.30 | MonMotha | heh |
07:05.37 | T0mW | heh, having only two methods of accessing external RAM (DPTR, P2:Rx) is a PITA |
07:05.55 | T0mW | especially when doing block moves |
07:05.56 | MonMotha | I'm just curious that you've never done separate registers for normal and irq on 8051 |
07:06.09 | MonMotha | yes, I did like the EZ-USB's double DPTR |
07:06.17 | T0mW | no, the "fast" IRAM was too valuable as storage |
07:06.24 | MonMotha | ah |
07:06.54 | T0mW | yeah, the app got very big very fast and I had to lose the register banking |
07:07.13 | T0mW | I always seem to run out of IRAM |
07:07.28 | MonMotha | I take it your system had a variable cycle count? |
07:07.39 | MonMotha | the original 8051 was 12 cycles/instruction no matter what |
07:08.41 | T0mW | figure it this way: if you bank the registers you are reserving 8 bytes, but if you push a couple of regs onto the stack you can usually write most interrupt handlers in less than 8 registers |
07:09.58 | MonMotha | indeeed |
07:10.05 | MonMotha | at the expense of the time requierd to stack everything |
07:10.18 | T0mW | MonMotha: nope, still a standard 12. We never could use the dallas stuff as most equipment needed lower power, not faster speed. then, if you went with dallas, you were locked into a single vendor. |
07:11.07 | T0mW | Although, OKI was disqualified as a supplier as their 8051's had a mistake in the die |
07:11.24 | MonMotha | the interrupt handler I wrote with banked registers needed to be fairly fast since I was juggling two tasks at once |
07:11.41 | MonMotha | what did OKI do? |
07:12.53 | T0mW | well, you know how "movx acc, @Rx" works, right? P2 forms the high byte of the 16bit address to external RAM |
07:13.33 | MonMotha | actually, I wasn't aware of that particular instruction |
07:14.27 | MonMotha | ok, yes |
07:14.31 | MonMotha | that's in my documentation though |
07:14.58 | MonMotha | I usually have all my IOs tied up :) |
07:15.07 | Russ | when I wrote code for the 8051 that handled 4 tasks, I didn't bother with interrupts |
07:15.07 | MonMotha | since most of my uC apps are interfacing |
07:15.11 | Russ | I just did time slices |
07:15.17 | T0mW | well, it somewhat obscure, but the "movx" uses R0..R1 (Rx) to emit the lower 8 bits, the Port 2 pins (P2) becomes the upper address. P2 normally is used by the external ROM access to ghost the address counter onto the bus. |
07:15.55 | Russ | each function excecuted for the amount of processor cycles for the time slice |
07:15.58 | MonMotha | Russ: I'm tempted to do that here. I went with interrupts because I was using the serial port. I'm thinking that may not be the best way to do things here |
07:16.04 | Russ | and then jumped to the next function for the next time slice |
07:16.09 | T0mW | MonMotha: well, the OKI 8051s did not drive the value of the Port 2 latches out the upper address, they wrote a 0 instead. |
07:16.27 | Russ | mine was a interceptor for ps/2 signals |
07:17.20 | MonMotha | Russ: I'm interfacing MIDI to an SPI bus |
07:17.31 | MonMotha | SPI with a bunch of extra handshaking |
07:18.24 | MonMotha | T0mW: that would effectively render that instruction useless, no? |
07:18.45 | T0mW | yup |
07:19.21 | T0mW | renders the 8051 pretty useless as well, then you have to swap the h^ll out of DPTR when you want to move something |
07:19.33 | MonMotha | T0mW: indeed |
07:19.36 | T0mW | or copy, or fill |
07:20.04 | MonMotha | you can address dptrl directly on a normal 8051, right? |
07:20.15 | T0mW | ah, erase code works, now to do the programming loop |
07:20.39 | T0mW | you mean dpl and dph, not dptrl |
07:20.52 | MonMotha | ok, the cypress stuff calls them differently |
07:21.01 | MonMotha | and the last thing I did major work with was the cypress |
07:21.53 | T0mW | well, the compiler / assembler probably calls them differently: dptrl, dptrh instead of dpl, dph correct? |
07:22.16 | T0mW | all 8051 assemblers / compilers are so different! |
07:22.36 | T0mW | all the ones I've run into, that is |
07:22.47 | MonMotha | yeah |
07:23.00 | MonMotha | I tend to use as31, which I've gathered is far from standard |
07:23.06 | MonMotha | and for the cypress part, I just .equ'd them |
07:23.07 | T0mW | different syntax for the asm opcodes, equates, defb vs db vs ?? |
07:23.17 | MonMotha | yeah, as31 is .db |
07:24.04 | T0mW | IIRC, the devtools that I use on the 8051 originally cost something like $2500 ? |
07:24.21 | T0mW | that was 1990 dollars |
07:24.36 | Russ | heh |
07:24.39 | Russ | now its all free |
07:24.42 | T0mW | big money for a glorified Small C Compiler |
07:24.43 | Russ | sucker |
07:24.44 | MonMotha | no kidding |
07:25.00 | MonMotha | you have sdcc and a whole bunch of assemblers to choose from |
07:25.48 | T0mW | Russ: with this new controller, I will be able to, finally(!!!), put that freakin' 8051 system to rest. And code onward in 32bit! |
07:26.06 | T0mW | I hate working in 8bit |
07:26.15 | T0mW | so tedious |
07:27.34 | *** join/#elinux noclouds (n=mhfan@60.166.41.80) |
07:31.14 | T` | anyone know how i can setup a rule inside make to pass some #define to the compiler while compiling a file? |
07:31.54 | T0mW | -D |
07:31.58 | T` | something like $ make LINUX=1 .. and compiler can hopefully see some #define _FOR_LINUX or something like that for the file |
07:32.10 | T` | T0mW, is that a 1 eye bandit laughing ;-) |
07:32.11 | T0mW | -Dfoo=bar |
07:32.19 | T0mW | #ifdef foo |
07:32.20 | T0mW | or |
07:32.23 | T` | ah |
07:32.35 | T` | but i seen in some cases where it just works the way i said above.. no -D |
07:32.55 | T0mW | yeah |
07:33.08 | T0mW | you can define foo as an environ var in the makefile |
07:33.18 | T0mW | e.g. export foo=bar |
07:33.39 | T0mW | bar is just to give foo a non-null value |
07:33.57 | T` | ah.. so how does it relate to $ make foo_en=1 for ex. |
07:33.58 | T0mW | no |
07:34.07 | T` | ? |
07:34.13 | T0mW | if you export, you cannot test with #ifdef |
07:34.26 | T0mW | #ifdef only tests defines not environs |
07:35.01 | T` | uh right.. but i assume there can be some if statement for make? so i can see if some env variable is defined, and if it is, then i pass -Dfoo=bar to the compiler |
07:35.16 | T0mW | exports are only usefull when passing values between makefiles |
07:35.35 | T0mW | or, to child processes (/bin/sh) |
07:35.50 | T` | yea, i'm just looking into passing defined based on env variables |
07:35.58 | T` | s/defined/defines |
07:36.00 | T0mW | ifeq ($foo, "bar") |
07:36.04 | T0mW | something like that |
07:36.14 | T0mW | ifeq ($(foo), "bar") |
07:36.16 | T` | ah.. lemme try it |
07:36.39 | T0mW | I do a lot of that, conditional compile, conditional make |
07:37.12 | T0mW | ifeq "$TOPLEVEL" "" |
07:37.29 | T0mW | tests to see if environ var $TOPLEVEL is null |
07:37.58 | T0mW | export UPGRADE=-DUPGRADE |
07:38.29 | T0mW | $(CC) $(UPGRADE) blah blah |
07:38.57 | T0mW | #ifndef UPGRADE |
07:39.18 | T0mW | the exported UPGRADE controls a conditional compile |
07:40.36 | T0mW | T`: I use the same source for two builds: one build is main binary, second is a bootloader binary. So, I use conditional (UPGRADE) to modify the source |
07:41.44 | T` | aah cool.. that seemed to do something.. but getting some weird error all of a sudden.. hmm |
07:41.51 | T` | make: *** No rule to make target `vxjpeg.c', needed by `vxjpeg.o'. Stop. |
07:42.16 | T0mW | :P now the fun begins! |
07:42.33 | T` | oh crap! i did a rm -rf vxjpeg.* .. which wiped the source :p |
07:42.51 | T0mW | oh, you have it cvs / svn, right? |
07:43.08 | T` | in make clean that is.. |
07:43.12 | T` | yea, i got svn |
07:45.25 | T` | btw.. i dont use tabs in my vi setting.. and makefile need it annoyingly on each line.. anyway to do that easily in vi/ |
07:46.47 | T0mW | T`: so, at the bottom of your makefile, put "# vi:tabstop=3: |
07:47.33 | T0mW | when you run vi, it will see the "# vi:" phrase and parse the remaining options for itself |
07:48.56 | T` | nop.. hmm.. didn't help |
07:50.20 | T` | you mean without the first \" right? |
07:50.32 | T` | well, it didn't do much either way :( |
07:50.43 | T` | may be vi should be told to parse these things in the file? |
07:51.17 | T0mW | any file, it must be either the first line or last line of the file. |
07:51.42 | T` | # vi:tabstop=3: |
07:51.46 | T` | .thats what i got on the first line |
07:51.59 | T0mW | with vim (not vi), a comment followed by a space "vi:" (e.g. // vi:) flags vi to parse the line |
07:52.11 | T0mW | vi won't do it, only vim |
07:52.27 | T` | yup.. i'm trying it in vim for sure |
07:52.34 | T` | do i just type tab to get the special car? |
07:52.37 | T` | char* |
07:52.56 | T` | i just tried adding a new line to my make sequence, and it compiled about missing operator.. |
07:52.56 | T0mW | ? |
07:53.02 | T` | s/operator/seperator |
07:53.44 | T` | set expandtab |
07:53.44 | T` | set tabstop=4 |
07:53.44 | T` | set softtabstop=4 |
07:53.50 | T` | i got those in my vimrc btw.. |
07:54.01 | T` | i think its that "expandtab" which is messing up |
07:54.15 | T0mW | there is no escapement, just plain text, only put "# vi:expandtab:tabstop=4:softtabstop=4:" would do the same |
07:54.33 | T0mW | # vi:expandtab:tabstop=4:softtabstop=4: |
07:54.36 | T0mW | no quotes |
07:55.03 | T` | ok i did # vi:set noexpandtab: that seems to work |
07:55.04 | T` | :) |
07:55.09 | T0mW | aha |
07:55.11 | T0mW | :) |
07:55.47 | T` | very well.. now ifeq based thing looks nice.. heh thanks ! |
07:56.14 | T0mW | T`: it makes life easier when you have to normally use tabstop=3 and someone wrote their code with blank tab fill + tabstops at 8, you can comment the file with: // vi:tabstop=8: |
07:57.51 | T` | yea, i agree.. this is a very nice feature... i ahve seen those lines before in files, but never really used them or given much taught to them |
07:59.38 | T` | another thing is.. if i had like.. A= file1.c file2.c file3.c |
07:59.51 | T` | prog.o: $(A) |
07:59.55 | T` | ... |
08:00.03 | T` | A+=mymodule.c |
08:00.12 | T` | mymodule wont be evaluated into prog.o right? |
08:00.35 | T` | i was trying to keep all my changes to the bottom of the makefile, but that didn't work earlier.. i had to place those += before the prog.o was compiled |
08:00.56 | T0mW | makefiles are magic |
08:01.26 | T0mW | IIRC, the makefile is two pass, pass one resolves references, pass two executes rules |
08:02.21 | T` | is that variable considered a reference |
08:02.25 | T0mW | try: |
08:02.41 | T0mW | A=file1.o file2.o file3.o |
08:02.43 | T` | because i had some weird problems of my module file not being compiled with the rest of the files when i had those lines at the end of the file |
08:02.48 | T0mW | all: $(A) |
08:02.59 | T0mW | %.o: %.c |
08:03.08 | T0mW | <PROTECTED> |
08:03.26 | T0mW | it gets wierd |
08:04.21 | T0mW | the logic flow does not follow what you learn to do a language, it has it's own logic |
08:04.33 | T0mW | s/a/with a/ |
08:04.47 | T0mW | bah |
08:04.53 | T` | ibot, shut up |
08:04.54 | ibot | yes, Master t` |
08:05.04 | T` | ibot, T` damnit.. not t` |
08:05.21 | T` | heh |
08:05.29 | T0mW | mov lr, pc appears not be a ret |
08:05.58 | T` | if lr isn't stored on the stack |
08:06.26 | T` | which arch? |
08:06.31 | T` | arm? |
08:06.40 | T0mW | ah, it is not on the stack |
08:07.53 | T0mW | lr contains the return address, so "b lr" ? |
08:08.22 | T` | hmm.. as long as you have the stack in original condition as caller |
08:08.34 | T` | and also perhaps the registers (you can't clobber them) |
08:08.58 | T` | i dunno, i never care about clobbering if i'm writing my own asm.. i make sure i dont touch registers which are used by caller in the called func |
08:09.16 | T` | but some people like to follow conventions like compilers |
08:09.50 | T0mW | mov pc, lr |
08:09.59 | T` | T0mW, also make sure b can branch to far locations.. PC is usually 64-bit, and some b instructions only have a 32-bit in-word |
08:10.01 | T0mW | that works |
08:10.23 | T0mW | thank God that someone invented JTAG! |
08:10.32 | T` | hmm.. is this ATT or INTEL asm syntax? |
08:10.40 | T0mW | gcc |
08:10.43 | T` | pc = lr or lr = pc from that instructino |
08:10.51 | T` | well, then you just moved pc to lr |
08:10.57 | T` | what wont branch |
08:10.59 | T0mW | reverse |
08:11.01 | T` | s/what/that |
08:11.06 | T0mW | moved lr to pc |
08:11.14 | *** join/#elinux toi (n=pleemans@d5152D12D.access.telenet.be) |
08:11.31 | T0mW | mov [dest], [src] |
08:11.51 | T` | yea, but thats ATT ;) |
08:11.56 | T` | gcc is like mov src, dest |
08:12.10 | T` | err wait.. i think ATT is gcc.. intel is what you described |
08:12.21 | MonMotha | yeah, gcc is the forsaken mess that is ATT |
08:12.35 | T` | MonMotha, wow really? i hate intel's syntax :/ |
08:12.40 | T` | MonMotha, i got used to the gcc syntax |
08:13.24 | T0mW | any difference between "ldr r2, =CMD" and "ldr r2, CMD" ? |
08:14.06 | T` | toi, .. one more thing.. if i had this ifeq ($(DEBUG),1) CFLAGS+= -g... and if i did a make.. it will not use debug flag then.. but if i say make DEBUG=1.. it wont clean and recompile.. it will just not do anything since it compiled earlier.. anyway to force recompile on this flag? |
08:14.23 | T` | T0mW, if you have =, thats definetely Intel syntax i would say |
08:14.48 | T` | = is dereference i believe |
08:15.45 | T` | hmm.. i'm actually not sure.. or was it the literal... |
08:16.22 | T0mW | ah, "ldr r2, =CMD" get the address of CMD, "ldr r2, CMD" gets what CMD contains |
08:17.33 | T` | ah' |
08:17.48 | T` | which architecture is this asm for btw? |
08:20.34 | MonMotha | T`: well, the first arch I learned was 8051, and that pretty much universally uses Intel's style |
08:20.46 | MonMotha | then I learned MIPS, which is similar to intel style x86 (at least, more similar than ATT style) |
08:21.31 | T` | yea, mostly i think this whole syntax thing arises from which one you started out first with |
08:22.35 | T` | but i did PIC assembly first, but its too basic to compared to PC (which i did next... ) |
08:23.00 | T` | MonMotha, any ideas on my make question? T0mW seems to have fallen in love with asm ;) |
08:23.24 | T0mW | T`: more so than with Make |
08:23.58 | T0mW | I've been writing makefiles for many years. It takes time to learn those things |
08:24.19 | T` | heh, tell me about it |
08:24.41 | T` | some things are obvious, but i should really read a proper tutorial.. i never did serious study on it |
08:24.57 | T0mW | one thing I use is something like this: |
08:25.16 | T0mW | all: $(PROGRAM) |
08:25.22 | T0mW | . |
08:25.30 | T0mW | $(PROGRAM): |
08:25.31 | T0mW | . |
08:25.51 | T0mW | <PROTECTED> |
08:26.05 | T0mW | note, dots (.) are blank lines. |
08:26.13 | T0mW | doh |
08:26.19 | T0mW | . |
08:26.23 | T0mW | all: $(PROGRAM) |
08:26.25 | T0mW | . |
08:26.29 | T0mW | $(PROGRAM): |
08:26.33 | T0mW | <PROTECTED> |
08:26.36 | T0mW | . |
08:27.12 | T0mW | T`: sometimes you have to use hidden files as a testpoint: |
08:27.24 | T0mW | `touch .foo` |
08:27.47 | T0mW | if [ -e ".foo" ] ; then \ |
08:27.56 | T0mW | <PROTECTED> |
08:27.58 | T0mW | fi |
08:28.03 | T` | but i dont see how defining/undefining DEBUG env.. is going to recompile $(PROGRAM) in the above make |
08:28.19 | T` | yea, but can i include bash shell code into a make file? |
08:28.32 | T0mW | I don't know what is wrong with your makefile, I was just speaking in general |
08:28.46 | T0mW | I just showed you the bash code |
08:28.58 | T0mW | put it under a rule |
08:29.11 | T0mW | as when you do "$(CC) blah blah" |
08:29.34 | T` | ah |
08:29.44 | T0mW | if [ -e ".foo" ] ; then \ |
08:29.59 | T` | yup i got it.. |
08:30.01 | T0mW | $(CC) blah ; \ |
08:30.02 | T0mW | fi |
08:30.18 | T` | and $(PROGRAM): |
08:30.30 | T` | does that mean it wont depend on anything like $(PROGRAM): program.c ? |
08:30.43 | T0mW | no |
08:30.44 | T` | so code below it will always be equaluated? |
08:30.56 | T` | s/qua/va/ |
08:31.16 | T0mW | dependancies are from the output (desired result) tested against resource |
08:31.24 | T0mW | e.g. program.o: program.c |
08:31.41 | T0mW | so |
08:31.48 | T` | right, but what if you dont specify any.. |
08:31.54 | T0mW | you're dead |
08:32.14 | T0mW | you need a "result" of some kind |
08:32.17 | T` | even if that $(PROGRAM) is in all: |
08:32.32 | T0mW | it doesn't have to be an executable. makefiles can also run YACCs, etc. |
08:32.41 | T0mW | or text manglers |
08:32.43 | T0mW | or ... |
08:33.00 | T0mW | it is a "desired" against "input" |
08:33.07 | T0mW | s/input/resource/ |
08:33.25 | T` | hmm ok |
08:33.58 | T0mW | not "take this resource and build that" it is "if that is older than resource, build" |
08:34.38 | T0mW | output from input(s), not, input(s) to output |
08:35.04 | T` | yea, i sorta figured that from experimenting.. heh |
08:35.14 | T` | anyway... i got what i wanted to work :) that ifeq is perfect |
08:35.15 | T` | thanks |
09:23.45 | *** join/#elinux ade|desk (n=adavey@194.200.143.249) |
09:54.13 | *** join/#elinux masked (n=masked@static-203-87-16-192.vic.chariot.net.au) |
11:07.12 | *** join/#elinux darkschneider (n=gab@213-140-6-96.ip.fastwebnet.it) |
11:57.30 | *** join/#elinux mallum (n=mallum@host86-136-17-6.range86-136.btcentralplus.com) |
11:57.30 | *** mode/#elinux [+v mallum] by ChanServ |
12:09.58 | *** join/#elinux Genesis (n=genesis@dev.aimao.org) |
12:11.31 | Genesis | bonjour |
12:12.06 | *** join/#elinux weo (n=weo@p5499976E.dip0.t-ipconnect.de) |
12:12.34 | *** join/#elinux Sgt-Donan (n=Loutre@feu30-1-82-242-59-35.fbx.proxad.net) |
12:55.36 | *** join/#elinux prpplague (n=billybob@72.22.143.75) |
12:56.08 | *** mode/#elinux [+o prpplague] by ChanServ |
12:58.02 | prpplague | ~seen sjhill |
12:58.07 | ibot | sjhill <n=sjhill@eth13.com-link.com> was last seen on IRC in channel #elinux, 6d 19h 13m 33s ago, saying: 'no big deal'. |
13:03.32 | T0mW | last time he said something about going to a meeting about a new project. |
13:03.45 | prpplague | ahh |
13:03.53 | prpplague | anyone here used libelf ? |
13:03.54 | T0mW | so, he is having fun and, I guess, we won't see him for a while |
13:12.03 | prpplague | T0mW: hehe, i've been getting tons of email about making a commercial juicebox cart |
13:12.28 | T0mW | guess you'd better brush up on Eagle then |
13:13.23 | prpplague | T0mW: hehe, naw, i'm not doing one |
13:13.30 | prpplague | T0mW: got other fish to fry |
13:19.33 | chouimat | morning |
13:19.40 | *** join/#elinux dijenerate (n=dijenera@72.22.159.26) |
13:22.30 | T0mW | prpplague: hmm, if I erase the Flash sector, I guess that means if I try to single step code inside that sector, it gives NOPs... |
13:22.38 | T0mW | Oh! I get it now! |
13:22.40 | T0mW | doh |
13:25.23 | T0mW | drat |
13:29.36 | dijenerate | prpplague: hey how's it going? |
13:29.36 | prpplague | dijenerate: ssdd, its monday, hehe |
13:29.55 | dijenerate | prpplague: lol |
13:31.24 | dijenerate | prpplague: did you have a chance to check on anything for me? |
13:31.41 | prpplague | dijenerate: yea i found a couple fpga pci based cards |
13:31.49 | prpplague | dijenerate: just got the specs this morning |
13:31.55 | prpplague | dijenerate: looking them over now |
13:31.58 | dijenerate | prpplague: ok, cool |
13:32.04 | dijenerate | prpplague: great thanx |
13:32.08 | chouimat | hi prpplague T0mW |
13:46.36 | *** join/#elinux GPSFan (n=Ken@65.121.49.208) |
13:48.20 | prpplague | GPSFan: hey hey |
13:48.41 | GPSFan | prpplague: good morning..how'it? |
13:49.30 | prpplague | GPSFan: hehe, i has 27 emails in my inbox with ppl wanting me to build them juicebox carts. heheh |
13:51.17 | GPSFan | prpplague: wow.. |
14:08.03 | chouimat | ROFL ... according to the elections law it's illegal to eat the paper on which you cast your vote |
14:20.59 | *** join/#elinux eggers (n=eggers@pixpat.austin.ibm.com) |
14:43.01 | MonMotha | ok, actual linux question for once |
14:43.29 | MonMotha | I have an application that I wish to link against multiple libraries. Some of them I want to link statically and others dynamically. What's the easiest way to go abotu this? |
14:47.00 | Crofton | http://news.yahoo.com/s/nm/20060123/tc_nm/manufacturing_mattel_dc |
14:52.01 | *** join/#elinux CosmicPenguin (n=nobody@aus-ext-proxy01.amd.com) |
14:52.01 | *** mode/#elinux [+v CosmicPenguin] by ChanServ |
14:52.59 | CosmicPenguin | fear! |
14:53.43 | prpplague | CosmicPenguin: hehe |
14:55.01 | chouimat | hi CosmicPenguin |
14:57.37 | *** join/#elinux TorbaX (n=TorbaX@host-84-220-144-168.cust-adsl.tiscali.it) |
14:58.22 | CosmicPenguin | any suggestions? |
15:03.17 | prpplague | CosmicPenguin: http://applications.linux.com/article.pl?sid=05/10/07/172259&tid=37 |
15:03.26 | prpplague | CosmicPenguin: see if that has anything to help |
15:05.49 | CosmicPenguin | prpplague: thanks - I'm still getting use to this whole kernel maintainer gig |
15:06.51 | CosmicPenguin | prpplague: no, thats pretty basic |
15:07.11 | CosmicPenguin | prpplague: I've been using mutt for a long time now, but I never figured out how to save off a message in a file |
15:07.19 | CosmicPenguin | and its not pratical for me to go finding it in my Mail driectory |
15:08.56 | prpplague | CosmicPenguin: ahh |
15:20.43 | prpplague | CosmicPenguin: hey, whats the rule-of-thumb with adding packages to OE? i.e. what applications are too end user specific for addition to the main branch? |
15:20.57 | CosmicPenguin | nothing |
15:21.24 | CosmicPenguin | generally you don't want binary packages |
15:21.38 | CosmicPenguin | and if there is something that only has benefit to you personally, there isn't much reason to share it |
15:21.49 | CosmicPenguin | but other then that, if other people can benefit from your code, then its legal |
15:22.19 | prpplague | hmm |
15:22.42 | CosmicPenguin | by binary package, I mean, like the ICA tool or something |
15:22.50 | prpplague | yea |
15:22.59 | CosmicPenguin | though /me has written a .bb for ICA |
15:23.06 | prpplague | CosmicPenguin: just trying to consider the rules for buildroot |
15:23.35 | CosmicPenguin | I think the best rules would be: 1) source based, 2) at least a minimal audience |
15:23.42 | prpplague | CosmicPenguin: got a ton of submissions for packages that are very very specific to the end user that submitted it |
15:24.10 | CosmicPenguin | Can you give me an example? |
15:24.29 | prpplague | CosmicPenguin: yea, i think there is one additional (imho) for buildroot as we want to keep the whole package light |
15:24.40 | CosmicPenguin | prpplague: granted |
15:24.51 | CosmicPenguin | prpplague: but eventually somebody will want to toss in Firefox or something, and I think thats cool |
15:24.53 | prpplague | CosmicPenguin: an application to wipe magnetic material such as drives and tape |
15:25.11 | CosmicPenguin | prpplague: is it on freshmeat? |
15:25.24 | CosmicPenguin | I would say that if the package is on freshmeat, it warrents inclusion |
15:25.37 | CosmicPenguin | otherwise, ask the guy why anybody would care other then himself |
15:25.57 | prpplague | yea its on freshmeat |
15:26.04 | CosmicPenguin | then I say take it |
15:26.25 | prpplague | think i'll leave it up for a week for feedback |
15:26.29 | CosmicPenguin | You know, there's also nothing saying you couldn't do a "buildroot-extra" and stick all the questionable packages in there |
15:28.19 | prpplague | CosmicPenguin: yea, thats what i was considering |
15:29.09 | CosmicPenguin | buildroot-overloaded |
15:29.16 | prpplague | yep |
15:29.26 | prpplague | CosmicPenguin: i've got like 137 open buildroot issues |
15:29.33 | prpplague | CosmicPenguin: trying to get them under control |
15:29.48 | prpplague | CosmicPenguin: i'd like to get them down to less than 30 at any one time |
15:29.55 | CosmicPenguin | buildroot is like that cool kid who doesn't care what anybody else thinks |
15:30.09 | prpplague | hehe |
15:30.16 | CosmicPenguin | Everyone is all like "oh, buildroot, can I use you" and buildroot is like "whatever" |
15:30.24 | prpplague | HA |
15:31.09 | prpplague | CosmicPenguin: yea i've found that most of the uclibc folks are very into the compile stuff, but not much for keeping up with bugs and features |
15:31.41 | prpplague | CosmicPenguin: i'm trying to organize some more general rules as such to kinda give the buildroot some direction |
15:32.26 | prpplague | CosmicPenguin: like right now for uclibc, if you need a toolchains for arm7tdmi, you have to select ARM720T and then select no mmu for uclibc to generate the proper tool chain |
15:32.35 | prpplague | CosmicPenguin: which makes no sense at all |
15:33.16 | *** part/#elinux TorbaX (n=TorbaX@host-84-220-144-168.cust-adsl.tiscali.it) |
16:02.34 | *** join/#elinux jipi (n=jipi@cm44.gamma185.maxonline.com.sg) |
16:40.49 | CosmicPenguin | sigh - the spammers are starting to find my work address |
16:41.08 | CosmicPenguin | all spammers must be sent to the deepest darkest depths of hell |
16:41.48 | chouimat | CosmicPenguin: yup ... I'm tired of PE spams .... and now they are able to bypass spamassassin |
16:42.49 | Crofton | PE? |
16:43.13 | chouimat | Crofton: Penis Enlargement stuff |
16:46.04 | *** join/#elinux TimRiker (n=timr@pdpc/supporter/bronze/TimRiker) |
16:46.05 | *** mode/#elinux [+o TimRiker] by ChanServ |
16:46.24 | Crofton | hhe |
16:46.36 | Crofton | the body bouncer guys are claiming the spam isn;t from them |
16:47.43 | CosmicPenguin | seems that everybody has finally emerged from the post holiday blues |
16:47.47 | CosmicPenguin | all the mailing lists are slammed |
16:51.38 | prpplague | hehe |
16:55.37 | prpplague | 123 cases |
16:56.03 | CosmicPenguin | I'm still deperately looking for programmers, if anybody knows anybody |
16:57.19 | chouimat | CosmicPenguin: for what? |
16:57.49 | CosmicPenguin | I need a junior level person as a lab rat, and a senior or very serior level programmer to be my peer |
16:58.06 | chouimat | k |
16:58.28 | CosmicPenguin | Both gigs are in Colorado though, so individuals need to be on-site |
16:58.42 | Crofton | bummer :) |
16:59.00 | chouimat | CosmicPenguin: I thought so ... |
16:59.14 | CosmicPenguin | I'm just saying, is all... :) |
17:00.08 | chouimat | CosmicPenguin: it's ok ... it's just it's not in my current plan to move to the US |
17:02.39 | CosmicPenguin | prpplague: What I'm saying is that if you happened to move to Colorado and wanted a job, I would kill somebody here to get you one at AMD |
17:02.53 | prpplague | CosmicPenguin: hehe, yea |
17:02.55 | CosmicPenguin | not that you would need a job, or want to move to Colorado, I'm just making conversation |
17:03.03 | prpplague | CosmicPenguin: just not all that interested in getting back into x86 |
17:04.04 | CosmicPenguin | I wonder if Micah knows anybody in the CU comp sci department |
17:18.15 | T0mW | prpplague: hey Dave, CosmicPenguin going to be working for you? |
17:19.20 | *** join/#elinux Crofton (n=balister@hc6521c84.dhcp.vt.edu) |
17:39.18 | *** join/#elinux CosmicPenguin (n=nobody@aus-ext-proxy01.amd.com) |
17:39.18 | *** mode/#elinux [+v CosmicPenguin] by ChanServ |
17:39.41 | CosmicPenguin | seesh |
17:39.46 | CosmicPenguin | thats the second time my box has locked up like that |
17:40.15 | CosmicPenguin | prpplague: do me a solid? |
17:40.19 | CosmicPenguin | prpplague: kick CP|Dresden, will ya? |
17:42.41 | CosmicPenguin | there we go |
17:42.43 | CosmicPenguin | stupid screen |
17:54.02 | *** join/#elinux Sgt-Donan (n=Loutre@mto30-1-82-242-53-8.fbx.proxad.net) |
17:56.47 | prpplague | CosmicPenguin: sorry was on the phone |
17:57.30 | CosmicPenguin | danm - I had this todo list all done, and I lost it |
17:59.14 | prpplague | doh |
18:00.01 | prpplague | CosmicPenguin: i love it when there are these frantic posts from ppl who need help with some simple issue, and you do a traceroute and find out they are working for a big "embedded specialist" |
18:00.06 | prpplague | CosmicPenguin: you ok? |
18:00.27 | CosmicPenguin | prpplague: jet lagged |
18:00.37 | CosmicPenguin | prpplague: yeah |
18:00.44 | CosmicPenguin | prpplague: the nano-x maililng list is full of them |
18:00.47 | CosmicPenguin | so is the lkml |
18:01.03 | CosmicPenguin | "I have big job porting OTG to <blah>. Can you help me find documentation on OTG?" |
18:01.14 | CosmicPenguin | Well, geez - probably shouldn't have accepted the job if you didn't know what it was there, champ |
18:03.10 | prpplague | too cool - http://www.cnn.com/2006/WORLD/europe/01/23/uk.russia.row/index.html |
18:13.44 | CosmicPenguin | argh! |
18:17.24 | *** join/#elinux Sgt-Donan (n=Loutre@mto30-1-82-242-53-8.fbx.proxad.net) |
18:29.56 | T0mW | prpplague: it is done, the last section of code has been writ |
18:33.04 | prpplague | hehe |
18:38.12 | *** join/#elinux T`2 (n=total@pdpc/supporter/student/T) |
19:13.56 | *** join/#elinux Crofton|770 (n=balister@mprg5213.mprg.ee.vt.edu) |
20:28.44 | Genesis | bonne nuit |
20:30.39 | *** join/#elinux T`2 (n=total@pdpc/supporter/student/T) |
20:39.28 | *** join/#elinux thraxisp (n=thraxisp@ottgate.precidia.com) |
20:58.11 | *** join/#elinux T`2 (i=debian-t@pdpc/supporter/student/T) |
21:33.03 | *** join/#elinux CosmicPenguin (n=nobody@aus-ext-proxy01.amd.com) |
21:33.04 | *** mode/#elinux [+v CosmicPenguin] by ChanServ |
21:49.32 | *** join/#elinux Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) |
22:44.27 | *** join/#elinux CosmicPenguin (n=nobody@aus-ext-proxy02.amd.com) |
22:44.27 | *** mode/#elinux [+v CosmicPenguin] by ChanServ |
23:18.22 | *** join/#elinux bryankemp (n=bryan@cpe-70-123-129-170.austin.res.rr.com) |
23:18.29 | *** join/#elinux andersee (n=andersee@codepoet.org) |
23:18.30 | *** mode/#elinux [+o andersee] by ChanServ |
23:18.37 | bryankemp | anyone using buildroot? |
23:20.57 | CosmicPenguin | You've found the right room |
23:21.08 | CosmicPenguin | unfortunately our main buildroot dude is gone for theday |
23:21.19 | bryankemp | Anyone have a good generic init script for loading a build root image? |
23:21.23 | CosmicPenguin | unless andersee is really alive and not just here automatically |
23:22.33 | bryankemp | I am building a small 2.6.9 based kernel (which I have compiling properly) and an rootfs from buildroot, but I am having trouble getting the rootfs mounted and switched to... |
23:23.12 | bryankemp | its for x86. |
23:36.47 | *** join/#elinux zx80user (n=adrian@81.170.29.134) |
23:36.55 | MonMotha | I take it you are using an initrd? |
23:37.10 | zx80user | Is anybody else having uname fail on them or other problems with busybox? |
23:37.48 | CosmicPenguin | zx80user: you need to be wayyy more specific then that |
23:38.40 | bryankemp | MonMotha: I am |
23:38.50 | zx80user | CosmicPenguin, sorry I was referring to my email to the list :) uname -r is failing in the first line of my boot time hotplug script, but there are lots of other problems too - on SH build with pthreads |
23:39.23 | CosmicPenguin | zx80user: not many people in here are on that list |
23:39.52 | zx80user | If I comment out the offending uname -r the script still fails, but in a different way - claiming there is no /sys on the system when there is one in plain sight |
23:40.11 | zx80user | CosmicPenguin, my mistake, apologies |
23:41.45 | CosmicPenguin | zx80user: when you say uname fails, what do you mean? |
23:43.15 | MonMotha | bryankemp: does it successfully bring the initrd up? |
23:44.06 | zx80user | CosmicPenguin instead of running uname -r it spews out the help text - it doesn't like the -r: this seems to be a general problem with recent BB builds - commands not understanding options, suggests some code is up the spout |
23:44.37 | CosmicPenguin | have you tried building a native build to verify that the code itself is broken, and not your system? |
23:45.16 | bryankemp | MonMotha: It does and it starts running the init scrip in the initrd, but I can't get that to work. |
23:48.53 | zx80user | CosmicPenguin -- on a Dreamcast?? Or do you mean on an i686 box? No, but then I was building and running code fine until a few days ago |
23:49.21 | CosmicPenguin | Then that points at a configuration issue on your part |
23:50.10 | MonMotha | bryankemp: well, what does it do, and where does it go wrong? |
23:52.04 | bryankemp | MonMotha, I'll be right back. I am going to open irc on my mac (I need this machine to run the new images) |
23:54.10 | *** join/#elinux thraxisp (n=thraxisp@ottawa-hs-64-26-147-146.d-ip.magma.ca) |
23:54.38 | zx80user | CosmicPenguin I don't think so. The builds failed after I updated by buildroot |
23:55.26 | CosmicPenguin | zx80user: so what changed in the buildroot then? |
23:56.01 | zx80user | Well, presumably the uClibc and Busybox sources were updated |
23:56.14 | CosmicPenguin | presumably? |
23:57.03 | zx80user | CosmicPenguin, ok they were. i know they were. Not least because I flagged up some bugs in the uClibc code |
23:57.16 | *** join/#elinux ircleuser (n=fpyanbur@cpe-70-123-129-170.austin.res.rr.com) |
23:57.44 | ircleuser | MonMotha: ok I am on my mac |
23:57.46 | CosmicPenguin | and what do the commit logs have to say about these updates? |
23:58.02 | CosmicPenguin | And like I said, try the busybox code on your host - if it breaks there, then you have a problem |
23:58.14 | CosmicPenguin | if it doesn't, then its time to check the changelogs |
23:58.30 | bryankemp | ok. I got all that worked out. |
23:59.30 | zx80user | That I don't know because my first reaction was to come on here (and the linux-sh channel) and ask whether anyone else was seeing something similar on SH or any other platform. Anyway, got to go now, it's midnight here, see ya and thanks for help |
23:59.41 | CosmicPenguin | have a good night |