irclog2html for #edev on 20051117

00:12.28*** join/#edev joe_bleau (n=joe_blea@adsl-69-151-117-13.dsl.fyvlar.swbell.net)
00:20.51*** join/#edev sorphin (n=dans@70.249.196.81) [NETSPLIT VICTIM]
00:20.51*** join/#edev fishhead (i=why@unaffiliated/fishhead)
00:20.51*** join/#edev GPSFan (n=Ken@65.121.49.208)
00:20.51*** join/#edev sunrunner20 (n=jamesdoe@carrell27.august.net) [NETSPLIT VICTIM]
00:20.51*** join/#edev catastr0p (n=catastr0@pcp04232047pcs.plyntv01.mi.comcast.net)
00:20.51*** mode/#edev [+o sorphin] by irc.freenode.net
00:23.29fishheadugh I spend $25 today on a used royal davinci and it's daPIECEofSHIT
01:40.47*** join/#edev fishhead (i=why@c-67-165-88-141.hsd1.pa.comcast.net) [NETSPLIT VICTIM]
01:41.21*** join/#edev sorphin (n=dans@70.249.196.81) [NETSPLIT VICTIM]
01:41.21*** join/#edev fishhead (i=why@unaffiliated/fishhead)
01:41.21*** join/#edev catastr0p (n=catastr0@pcp04232047pcs.plyntv01.mi.comcast.net) [NETSPLIT VICTIM]
01:41.21*** mode/#edev [+o sorphin] by irc.freenode.net
01:42.21*** join/#edev FISHHEAD- (i=why@c-67-165-88-141.hsd1.pa.comcast.net)
07:52.46*** join/#edev ChanServ (ChanServ@services.)
07:52.46*** mode/#edev [+o ChanServ] by irc.freenode.net
11:56.44*** join/#edev purl_ (i=ibot@rikers.org)
11:56.44*** topic/#edev is discussion for small embedded device development and hacking || http://www.elinux.org/wiki/ || http://www.hackaday.com || http://www.linux-hacker.net
12:11.28*** join/#edev prpplague (n=billybob@216.110.99.43)
13:08.45*** join/#edev sunrunner20 (n=jamesdoe@carrell27.august.net)
13:35.03*** join/#edev GPSFan (n=Ken@65.121.49.208)
14:22.23*** join/#edev ShortWave (n=shortwav@67.151.73.146)
14:22.23*** mode/#edev [+o ShortWave] by ChanServ
15:13.51GPSFanprpplague: here is my .mmc.o.flags file, I thank that provides all the flags you are needing.
15:14.09GPSFanifeq (-D__KERNEL__ -I/mnt/hdb4/usr/local/linux-2.4.21.32/linux-2.4.21/include -Wall -Wstrict-prototypes -Wno-trigraphs -Os -fno-strict-aliasing -fno-common -Uarm -fno-common -pipe -mapcs-32 -D__LINUX_ARM_ARCH__=4 -march=armv4 -mtune=arm7tdmi -msoft-float -Uarm -DMODULE -nostdinc -iwithprefix include -DEXPORT_SYMTAB,$(strip $(subst $(comma),:,$(CFLAGS) $(EXTRA_CFLAGS_nostdinc) $(CFLAGS_mmc.o) -DEXPORT_SYMTAB)))
15:14.09GPSFanFILES_FLAGS_UP_TO_DATE += mmc.o
15:14.09GPSFanendif
15:15.37prpplagueGPSFan: interesting
15:20.23GPSFansorphin: morning..
15:30.37sorphinGPSFan: howdy mr ken, how's things in the land of Paeonia(sp?)?
15:33.32prpplagueGPSFan: no luck with the compile, i'm gonna go back and try to find out why it doesn't work with the standard register_blkdev
15:34.29prpplagueGPSFan: the only thing i can think of is that devfs is taking care of something that needs to be added to the code
15:40.11sorphinprpplague: hey, you're an embedded hillbilly
15:41.08sorphinhtf do you get a CF to boot dos? we've tried a buncha diff ways, so we can do a bios upgrade on these nokias (no floppy, but have CF as primary master, then a hdd as sec master).. but the cf always fails when it comes to dos
15:42.27prpplaguesorphin: hmm, the only thing that i can think of that could be an issue is if you don't have the primary partiton on the cf card set as active
15:42.36prpplaguesorphin: haven't used DOS in years
15:42.52prpplagueGPSFan: found the problem
15:43.03prpplagueGPSFan: the minor numbering is off by one
15:43.20prpplagueGPSFan: the first partition on the sd/mmc card is minor number 2
15:43.29sorphinprpplague: all the things we've found actually write the dos image to /dev/sda directly... when freedos goes to boot, all i see is Starting.... and FAT ROOT or ROOT FAT, forget which
15:43.32sorphinand i don't use dos either
15:43.37sorphinexcept to flash the bios
15:43.38sorphinheh
15:49.11GPSFanprpplague:  so I have mmca as 121,0 and mmca1 as 121,1
15:52.15*** join/#edev prpplague (n=billybob@216.110.99.43)
15:59.20prpplaguesorphin: the other thing that i find is that some systems don't see the h/c/s the same which causes boot problems
16:00.48prpplagueGPSFan: looks like there is something off
16:01.16prpplagueGPSFan: minor 2 is getting the base info from minor 1 but trying to mount minor2 specs
16:04.22GPSFanprpplague: what is your mmc formatted for?
16:06.23prpplagueGPSFan: 1 partition, with fat16
16:07.25GPSFanah, mine is 1 partition ext2.
16:07.46prpplagueGPSFan: i also tried ext2
16:08.06GPSFanwhat version of the kernel?
16:08.15prpplague2.4.26
16:08.24GPSFan2.4.21-rmk1
16:08.36GPSFan+zipit & km patches
16:08.55prpplagueGPSFan: yea i dl'd your patch and had a look
16:10.13*** join/#edev jbevren (n=jbevren@dumt-02-164.dsl.netins.net)
16:11.00GPSFanI found that include/linux/devfs_fs_kernel.h maps devfs_register_blkdev to register_blkdev, but I havent found how it's included in mmc.c
16:15.02prpplaguei wonder if simply including that .h file would do the trick?
16:19.23GPSFanprpplague: maybe, I just re-compiled with devfs_register_blkdev replaced with register_blkdev and the mmc.o's are identical
16:20.31prpplagueinteresting
16:21.16GPSFanso somehow in my configuration the devfs_fs_kernel.h file is being included in my mmc.c during the compile.
16:22.35prpplagueGPSFan: which version of gcc are you using?
16:25.24GPSFanlooks like 3.3.3
16:32.15prpplagueGPSFan: hmm, still like to know why just plain register_blkdev doesn't work properly
16:35.51GPSFanyeah, that puzzles me too. It indicates (along with the partiton issue) that there is something fundamentaly different about how we are building the code.
16:36.22GPSFannot that I understand it  :>)
16:39.35prpplagueor i
16:39.58prpplagueGPSFan: but it does seem to be mix-matching the minor numbers and information
16:41.58GPSFanI still need to play around with vfat, fro some reason it wouldn't insmod, bunch of undefined externals.
16:42.32GPSFani probably don't have the kernel config right.
16:45.53prpplagueGPSFan: np
16:48.54prpplagueGPSFan: the only difference i can find in your code vs. mine is that you have the gendisk defined and i have it as an external
17:12.46GPSFanprpplague: time for a beer.. that usually leads to less confusion. ;>)
17:51.03prpplagueGPSFan: interesting, looks like something is shifted as the start sector using standard register_blkdev is being read as the nr_sectors
18:02.50prpplagueGPSFan: yea it looks like the arangement of some of the structs are different when devfs isn't used
18:21.12prpplagueproblem found
18:21.43prpplagueGPSFan: starting after 2.4.21 the genhd.h file has ifdef's for the hd_struct
18:21.54prpplagueGPSFan: which depend on devfs
18:31.11prpplagueGPSFan: or atleast part of it is working
18:31.25prpplagueGPSFan: i'm still getting some bogus values when using df
18:31.35prpplagueGPSFan: its reporting values for the entire card
18:44.31jbevrenhmm
18:44.36jbevreninteresting
18:44.37jbevrenhehe
19:50.11GPSFanprpplague: very interesting, lots of things changed around 2.4.21, some say for the better, I'm of mixed feelings. Somethings definately got better, others....
19:56.27GPSFanprpplague: gee, genhd.h from 2.4.21 (unpatched) is identical to one from 2.4.31 ???
20:25.42prpplagueGPSFan: hmm 2.4.20 is different from 2.4.21 from what i see
20:25.59prpplagueGPSFan: but thats not the only issue
20:26.12prpplagueGPSFan: this code doesn't handle the partitions very well at all
20:27.57GPSFanprpplague: yep, I'm using 2.4.21, so the genhd.h changes must really be after 2.4.20.
20:28.44GPSFanprpplague: I've only teied 1 partition, I'll give a couple a try.
20:33.17prpplagueGPSFan: are you actually partitioning the card? or just formating the entire card?
20:40.01GPSFanfor the 1 partition I use fdisk, then mkfs.ext2
20:46.18GPSFanprpplague:  2 partitions seems to work I fdisk'd them on another machine, did mkfs.ext2 then copied some files to them.
20:48.46GPSFanI mounted bot on the zipit to seperate places, and df sees the correct size for the second one, but the full disk for the first one.
20:50.36GPSFaninterestingly the other machine sees the same thing! other machine is ubuntu 5.10 ppc 2.6.12-9
20:53.22prpplagueinteresting indeed
20:53.43prpplagueGPSFan: thats similiar to the problem i'm running into
20:56.06prpplagueGPSFan: what does cat /proc/partitions show
21:00.00GPSFanmajor 121, minor 0 mmca 29888, minor 1 mmca1 12655, minor 2 17198.
21:01.03prpplagueGPSFan: just out of curiosity if you try mounting /dev/mmca what do you get?
21:02.05GPSFanfailed, invalid argument
21:05.02GPSFaninteresting, the ubuntu system automounts 3 partitions, the first as the whole device vfat, the next 2 as ext2 of the right size. the Zipit doesnt support vfat in the kernel, probably why the error message.
21:05.24prpplagueahh
21:05.53prpplagueGPSFan: its soooo stange that your build will go without the devfs
21:06.27GPSFanyeah, I know.. hmmm it's snowing ;>)
21:08.34prpplaguehehe
21:12.24GPSFanthe size reporting problem in df may not have anything to do with your/my build. I'm seeing the same /proc/partions results on both machines, and the same df results as well (except for the vfat on the Zipit)
21:13.47GPSFandf on the zipit is busybox 1.1-pre1 the df on the ubuntu box in df from gnu coreutils
21:15.02prpplaguehmm
21:21.38GPSFanI'm going to try re-partitioning using cfdisk instead of fdisk.
21:21.51prpplaguei still understand why your's is compiling with the devfs options,
21:26.58GPSFanI made anf formatted 3 10MB partitions (32M sd card) ubuntu still sees the whole thing as a seperate vfat partition, and now sees 3 properly sized ext2 partitions. I started with a zero partition table instead of reading the original and erasing all the partitions.
21:28.55prpplagueGPSFan: just out of curiosity do you have different files on each partition?
21:29.27prpplagueGPSFan: i.e. to make sure you are getting the right parition mounted?
21:31.18GPSFanyes I have a file logx on each partition where x is the partition number
21:32.46prpplagueahh cool
21:33.29GPSFannow on to the zipit...
21:33.55prpplagueGPSFan: oh i thought you were testing on the zipit
21:37.04GPSFannot yet, just did, df now reports the correct sizes for each partition, and each of the files is in its right partition.
21:37.15GPSFanon the zipit.
21:37.21prpplagueGPSFan: interesting
21:38.09GPSFan<PROTECTED>
21:38.09prpplagueGPSFan: what node numbers are you using?
21:39.09GPSFanmmca 121, 0 - mmca1 121, 1 - mmca2 121, 2 - mmca3 121, 3
21:39.51prpplaguehmm same here
21:40.08prpplagueGPSFan: yours was built during your kernel build correct?
21:40.22GPSFanyep, make modules.
21:42.00prpplagueguess i'm just gonna have to go with devfs for the momment
21:42.38GPSFanbummer, is this for the littlechips board?
21:42.51prpplaguenaw, another board
21:42.54prpplaguelh79520 based
21:43.03GPSFanah.
21:43.07prpplaguejust reallly really confused as to why this isn't workin
21:43.22prpplaguethis should be really simple and straight forward
21:44.49GPSFanI think my bad size in df was not starting with a clean partition table if fdisk. still don't know why ubuntu sees a fullsize vfat partition. oh well...
21:45.21GPSFanmaybe it's a bug in its automounting code.
21:45.46GPSFanor a feature, yeah, that's it a feature...
21:46.09GPSFanI'll keep looking at the devfs issue.
21:46.13prpplaguethanks
21:46.19prpplaguei'm out here for today
21:46.19prpplaguelater

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.