**** BEGIN LOGGING AT Thu Jan 14 02:59:57 2010 Jan 14 09:53:05 asac, i milestoned bug #457878 (since it still exists with the recent BSP drop) Jan 14 09:53:07 Launchpad bug 457878 in linux-fsl-imx51 "imx51 on board ethernet plug/unplug events not detected" [Medium,Confirmed] https://launchpad.net/bugs/457878 Jan 14 09:53:10 cooloney, ^^^ Jan 14 09:56:09 ogra: cool, will take a look when I am back to Shanghai, heh Jan 14 10:43:46 ogra: you are sure your modules are loaded? Jan 14 10:44:06 aka are you sure your kernel/initramfs you used for your uboot is identical with the one you have no your filesystem Jan 14 10:44:15 i had the same behaviour yesterday, but it was all a version mismatch Jan 14 10:44:28 dmart: thx for the email, could you please file a NEON bug in launchpad and assign it to me? Jan 14 10:44:43 dmart: then we can tracker the discussion Jan 14 10:45:43 ogra: asked for dmesg/syslog in bug Jan 14 10:50:54 asac, ?? Jan 14 10:51:08 ogra in your ethernet bug Jan 14 10:51:09 are you referring to the bug above ? Jan 14 10:51:15 or rather the bug you hi jacked Jan 14 10:51:19 yes Jan 14 10:51:19 sure the modules are loaded Jan 14 10:51:28 its not related to uboot Jan 14 10:51:40 yesterday you said there is no interface Jan 14 10:51:44 now its just LOWER_UP? Jan 14 10:51:51 thats on a redboot booted system and is a long standing missing feature in the FEC driver Jan 14 10:52:03 yeah Jan 14 10:52:08 ok Jan 14 10:52:13 it was missing in all kernel versions we ever had Jan 14 10:52:18 so that "no net at all" feature is fixed for you? Jan 14 10:52:23 no Jan 14 10:52:38 where is that bug? Jan 14 10:52:47 thats what i'm looking at atm, i just wanted to make sure the above bug isnt forgotten Jan 14 10:52:53 heh ok. Jan 14 10:53:06 i havent filed it yet, just looking at uboot code to make sure we build the right driver Jan 14 10:53:36 do you see your interface at all? Jan 14 10:53:39 i saw it Jan 14 10:53:46 though i only have a read-only fs Jan 14 10:53:59 yes, you can see it, but you cant bring it up Jan 14 10:54:29 usually you see the kernel setting the MAC in demsg, that part is missing, i see it initializing the driver but the line with MAC is missing Jan 14 10:54:46 wireless is working flawlessly here :) Jan 14 10:54:48 and ifconfig says it cant assign an address Jan 14 10:54:54 so i think its not a blocker to stop the uboot move in debian-cd Jan 14 10:55:02 (where it refers to the MAC) Jan 14 10:55:06 yeah Jan 14 10:55:19 the only interface the bnoard has doesnt work .. Jan 14 10:55:27 i would consider that as a blocker :) Jan 14 10:55:37 a release blocker yes, Jan 14 10:55:40 but not a roll out blocker Jan 14 10:55:49 well, at least highest prio Jan 14 10:56:01 sure, but we should do the uboot move before that Jan 14 10:56:14 well, uboot needs an upload anyway Jan 14 10:56:20 for the changed defaults Jan 14 10:56:42 we also need to consider that an installed system might not have /boot on the second partition Jan 14 10:57:36 i'm not sure how to solve that yet but have some ideas (if uboot doesnt stop script execution we can just try to load from part1 and then from part2, but i'm not sure that works, needs testing) Jan 14 10:57:52 ogra: atm we have a boot floppy and an install floppy. we can assume that its always on partition 2 for the boot floppy thing Jan 14 10:57:58 its not a real solution Jan 14 10:58:05 but we need to work on scripts to work on that Jan 14 10:58:11 how are you doing that on redboot? Jan 14 10:58:15 we cant assume that once we install uboot to mtd flash Jan 14 10:58:24 redboot never goes into flash Jan 14 10:58:29 once we install it there Jan 14 10:58:40 as soon as you boot from flash the first partition will be gone Jan 14 10:58:46 at that point we must be able to interate Jan 14 10:58:52 thats what davidm wants as final install setup Jan 14 10:59:00 sure Jan 14 10:59:07 but we are not there yet ;) Jan 14 10:59:14 so users just flip a DIP switch in the end Jan 14 10:59:29 no, but i'd like to know in advance if it works :) Jan 14 10:59:34 right. thats understood. just doesnt need to happen on the first iteration Jan 14 10:59:47 it has to be made working. Jan 14 11:00:09 it works already :) Jan 14 11:00:20 so you have an iterating boot.scr? Jan 14 11:00:24 but i dont feel good with all the hardcoding Jan 14 11:00:28 no Jan 14 11:00:32 it works as is Jan 14 11:00:33 right. so dont bother about that Jan 14 11:00:45 with the hardcoded part2 loading Jan 14 11:00:45 what i meant with has to be made working is that we can do that Jan 14 11:00:49 yes. Jan 14 11:00:57 11:59 < asac> right. thats understood. just doesnt need to happen on the first iteration Jan 14 11:01:01 11:59 < asac> it has to be made working. Jan 14 11:01:02 -> thats iterating boot.scr Jan 14 11:01:07 ok Jan 14 11:01:14 i am getting cigarettes and coffee Jan 14 11:01:21 and then will stop bothering you ; Jan 14 11:01:22 ) Jan 14 11:01:46 if i can put a bootscript into the defaults that tries part1 and moves on to part2 (without .scr since we need to load that as well) i prefer to do that directly Jan 14 11:02:12 currently the only thing i want the scr for is the kernel cmdline Jan 14 11:02:13 only boot.scr can test for something afaik Jan 14 11:02:18 so you cannot test if something succeeded Jan 14 11:02:32 boot.scr just replaces the bootcmd in the uboot defaults afaik Jan 14 11:02:41 it gets compiled afaik Jan 14 11:02:52 the whole for and if stuff is not available at command prompt Jan 14 11:02:53 it just gets turned into a file uboot can load Jan 14 11:03:08 but the content is what you see in printenv at the uboot prompt Jan 14 11:03:25 so everything we do in .scr files should be possible directly too Jan 14 11:04:17 afaik you can do "if ... then ... etc" in the hardcoded bootcmd Jan 14 11:05:15 i dont see any docs about that Jan 14 11:05:18 if you find that let me know Jan 14 11:05:54 uboot has really sucky documentation Jan 14 11:06:55 http://www.denx.de/wiki/view/DULG/CommandLineParsing Jan 14 11:07:31 bootm $addr1 || bootm $addr2 || tftpboot $loadaddr $loadfile && bootm Jan 14 11:07:33 ogra: ^ Jan 14 11:07:38 so you can test that etc. Jan 14 11:08:15 well, thats already what i look for Jan 14 11:08:48 so we need hush shell support builtin Jan 14 11:10:18 aha, needs the SHELL_HUSH option set at build time Jan 14 11:10:26 * ogra looks in the code Jan 14 11:13:22 CONFIG_SYS_HUSH_PARSER :) there it is Jan 14 11:16:34 yes Jan 14 11:18:04 http://paste.ubuntu.com/356511/ has the full config in case someone wants to look :) Jan 14 11:18:41 * ogra wonders if #define CONFIG_FEC0_PHY_ADDR 0x1F is our prob with the NIC Jan 14 11:35:12 * ogra builds Jan 14 11:35:17 asac, how about ext2 support while i'm at it, do you want that ? Jan 14 11:38:49 make[2]: Entering directory `/uboot-imx-2009.08/fs/fdos' Jan 14 11:38:49 ar crv /uboot-imx-2009.08/build/build-imx51/fs/fdos/libfdos.a Jan 14 11:38:49 make[2]: Leaving directory `/uboot-imx-2009.08/fs/fdos' Jan 14 11:38:49 hmm Jan 14 11:39:04 i wonder if fdos is actually the better fat filesystem :) Jan 14 11:40:08 mxc_fec.c: In function 'mxc_fec_initialize': Jan 14 11:40:09 mxc_fec.c:781: warning: passing argument 1 of 'mxc_fec_set_mac' from incompatible pointer type Jan 14 11:40:09 mxc_fec.c:725: note: expected 'struct fec_info_s *' but argument is of type 'struct fec_info_s (*)[1]' Jan 14 11:40:10 AHA ! Jan 14 11:40:26 i guess thats the NIC issue Jan 14 11:54:03 ogra: you can just include that i guess Jan 14 11:54:12 i think we should Jan 14 11:54:21 do it Jan 14 11:54:27 having /boot ext2 formatted will gain us a lot more flexibility Jan 14 11:54:34 i.e. symlinks Jan 14 11:55:20 so we can have uimage being linked to uimage-$version Jan 14 11:55:23 yes, go for it Jan 14 11:55:31 yup, will do Jan 14 11:55:37 testing hush support first Jan 14 11:55:40 its just a config flag after all Jan 14 11:55:42 right Jan 14 11:55:47 hush + ext2 + fix fec Jan 14 11:55:53 if you want me to lok at that code i can check Jan 14 11:56:49 plars: so when did dove regress exactly? do we know that? Jan 14 11:57:28 asac, fec ? yeah, that would help, so i can concentrate on the script Jan 14 11:57:46 ok ... let me use jocote ;) Jan 14 11:57:57 ogra: that happens on current archive uboot too, right? Jan 14 11:58:02 gah, i hate that sshd doesnt log me out if i shut down the babbage Jan 14 11:58:12 asac, yea, thats what i'm building Jan 14 12:08:08 BBG U-Boot > setenv myload 'if mmcinfo; then echo MMC found;else echo MMC not found;fi' Jan 14 12:08:08 BBG U-Boot > print myload Jan 14 12:08:08 myload=if mmcinfo; then echo MMC found;else echo MMC not found;fi Jan 14 12:08:08 BBG U-Boot > run myload Jan 14 12:08:08 Device: FSL_ESDHC Jan 14 12:08:08 Manufacturer ID: 1 Jan 14 12:08:10 OEM: 5041 Jan 14 12:08:12 Name: SM06G Jan 14 12:08:16 Tran Speed: 25000000 Jan 14 12:08:18 Rd Block Len: 512 Jan 14 12:08:20 SD version 2.0 Jan 14 12:08:22 High Capacity: Yes Jan 14 12:08:24 Capacity: 1945882623 Jan 14 12:08:26 Bus Width: 4-bit Jan 14 12:08:28 MMC found Jan 14 12:08:30 BBG U-Boot > Jan 14 12:08:32 SWEET ! Jan 14 12:08:48 ogra: can you say something like: Jan 14 12:08:51 so now we can iterate over partitions, look for /casper etc Jan 14 12:08:54 if fatinfo mmc 0:2; then Jan 14 12:09:01 sure Jan 14 12:09:07 if fatload mmc 0:2 /boot/...; then Jan 14 12:09:11 well, fatls rather Jan 14 12:09:16 ogra: where is that documented Jan 14 12:09:16 yeah, exactly Jan 14 12:09:21 i didnt see that on that patge Jan 14 12:09:39 14.2.16.3. Hush shell scripts Jan 14 12:10:28 it supports similar syntax to shell Jan 14 12:10:28 ok Jan 14 12:10:34 still thats not a reference Jan 14 12:10:36 its just examples Jan 14 12:10:39 bad doc Jan 14 12:10:45 well, as long as it works i dont care :) Jan 14 12:10:45 is hushshell some wider standard Jan 14 12:10:48 * asac checks google Jan 14 12:10:53 no, its uboot specific Jan 14 12:11:05 seems not Jan 14 12:11:07 yeah Jan 14 12:11:15 so ok. as long as it just works ;) Jan 14 12:11:17 i'm happy with if, for, while and until :) Jan 14 12:12:00 we can write a clever load script now that loads the .scr from predefined locations ... in the scr we can then put all the load commands Jan 14 12:12:09 that gives us most flexibility Jan 14 12:13:07 it expecially solves the prob that we have to set the UUID on cmdline after install Jan 14 12:14:49 BBG U-Boot > if fatinfo mmc 0:0; then echo fat found on partition0; else echo no fat on partition0;fi Jan 14 12:14:49 ** Partition 0 not valid on device 0 ** Jan 14 12:14:49 ** Unable to use mmc 0:0 for fatinfo ** Jan 14 12:14:49 no fat on partition0 Jan 14 12:14:50 heh Jan 14 12:15:35 BBG U-Boot > if fatinfo mmc 0:2; then echo fat found on partition2; else echo no fat on partition2;fi Jan 14 12:15:35 Interface: MMC Jan 14 12:15:35 Device 0: Vendor: Man 015041 Snr 41a4343b Rev: 9.4 Prod: SM06G Jan 14 12:15:35 Type: Removable Hard Disk Jan 14 12:15:35 Capacity: 5743.0 MB = 5.6 GB (11761664 x 512) Jan 14 12:15:36 Partition 2: Filesystem: FAT16 " " Jan 14 12:15:38 fat found on partition2 Jan 14 12:15:40 cute Jan 14 12:16:50 * ogra goes to make some coffee and will then work out a little script :) Jan 14 12:29:24 hmm first i should enable ext2 i guess Jan 14 12:30:01 * asac grabs more coffee while ogra makes good progress Jan 14 12:30:28 * asac takes the cheerleader role for now Jan 14 12:30:34 heh Jan 14 12:39:42 ok, uploaded Jan 14 12:46:54 ogra: wait ;) Jan 14 12:46:59 to late :) Jan 14 12:47:03 ogra: jocote:~asac/fix_fec_warnings.patch Jan 14 12:47:05 try that Jan 14 12:47:14 let me know if that works, there is a bunch of odd code surrouding that Jan 14 12:47:16 will do Jan 14 12:47:25 so it could well be that it doesnt and i have to fix a lot of code Jan 14 12:47:33 but the warning is gone that way Jan 14 12:47:44 but fec_initialize is a mystery code wise ;) Jan 14 12:47:57 int mxc_fec_initialize(bd_t *bis) Jan 14 12:47:57 { Jan 14 12:48:03 for (i = 0; i < sizeof(fec_info) / sizeof(fec_info[0]); i++) { Jan 14 12:48:08 feels they are on crack ;) Jan 14 12:48:14 heh Jan 14 12:48:18 i is used like fec_info[i] Jan 14 12:48:33 but lets see ... if it just starts working let me know ... i can do the upload Jan 14 12:50:05 i'm doing a testbuild anyway so i can quickly shove that in Jan 14 12:50:31 you were too eager ;) Jan 14 12:50:40 but well, you probably need some uploads just to excersize Jan 14 12:50:45 nah, i just didnt want the changes to get lost :) Jan 14 12:51:10 we have anout free version numbers :) Jan 14 12:51:13 *enough Jan 14 12:51:31 yeah. but if you upload quickly, it will end up with a upload failure Jan 14 12:51:38 and the ftbfs will be stuck forever on ubuntuwire Jan 14 12:51:55 e.g. .deb doesnt get accepted because newer source is avail Jan 14 12:52:43 nah, ubuntuwire is clever enough Jan 14 12:53:31 * ogra fires off dpkg-buildpackage ... Jan 14 12:53:36 20min or so Jan 14 12:59:32 i wonder if we should have other filesystems too Jan 14 12:59:56 ubifs, cramfs and jffs2 might be some that developers might use Jan 14 13:00:24 (ext2 passes btw) Jan 14 13:01:27 asac, mxc_fec.c passes the build with no errors now Jan 14 13:01:30 I'd like ubifs support. Jan 14 13:01:50 But I'm not sure I'd want a ubifs /boot, just because it'd be a pain to set up initially. Jan 14 13:01:59 right Jan 14 13:02:05 its only about /boot atm Jan 14 13:02:58 i wonder what would happen if i just enabled USB support :) Jan 14 13:02:58 something to play with on a weekend though Jan 14 13:03:09 For /boot, don't bother about ubifs or jffs2. Jan 14 13:03:16 yeah Jan 14 13:03:26 I'm less familiar with cramfs: dunno if that'd be useful. Jan 14 13:03:34 ext3 or 4 would be nice, but i doubt anyone ever implemented that Jan 14 13:03:38 I *do* want to be able to put / on ubifs, but that's just a kernel thing. Jan 14 13:03:46 yeah Jan 14 13:03:47 ogra: let me know if it *works* Jan 14 13:03:53 i know that the warning is gone ;) Jan 14 13:03:57 heh Jan 14 13:04:00 I'm not a fan of journaled boot. I use ext2 even on my amd64 systems. Jan 14 13:04:02 will do, still building Jan 14 13:04:06 just not sure if it works at all because there is strange code Jan 14 13:04:14 persia, to avoid fsck Jan 14 13:04:18 s$boot$/boot$ Jan 14 13:04:29 fsck of 50MB is fast :p Jan 14 13:04:39 still slows down the boot Jan 14 13:04:47 especially on SD Jan 14 13:05:03 I guess. I can't remember the last time I needed to fsck my /boot Jan 14 13:05:03 ogra: enable all fs that are available Jan 14 13:05:12 unless size explodes of course Jan 14 13:05:17 asac, thats overkill Jan 14 13:05:18 (but I tend to reboot only for kernel updates) Jan 14 13:05:19 not sure how big uboot can be for flash usually Jan 14 13:05:27 ogra: its not overkill if size is ok Jan 14 13:05:44 otherwise stick with what we have now Jan 14 13:05:47 as persia said, its unlikely that anyone puts /boot on ubifs ot jffs2 Jan 14 13:06:06 asac: The rationale to *not* enable ubifs or jffs2 for /boot is that these filesystems need separate preparation and writing to flash, which gets ugly and complicated. Jan 14 13:06:15 if we enable stuff it should add useful features Jan 14 13:06:20 USB would rock Jan 14 13:06:27 /boot on jffs2 basically means no kernel updates (in any sane way) Jan 14 13:06:38 /boot on ubifs could be made to work, but it'd be a very ugly hack. Jan 14 13:06:58 iots like squashfs, isnt it ? Jan 14 13:07:07 you would have to recompress it each time Jan 14 13:07:17 jffs2 is like squashfs Jan 14 13:07:24 ubifs is just different. Jan 14 13:07:26 (for a user perspective i mean) Jan 14 13:07:31 dont care. if we have reasons against those - and uboot bin gets bigger it can be ignored Jan 14 13:07:37 There's a procedure to create an initial filesystem, but writes thereafter are persistent. Jan 14 13:07:46 i want usb-storage ;) Jan 14 13:07:50 thats more important Jan 14 13:07:52 uboot.bin gets surely bigger with each option i enable Jan 14 13:07:52 Yes. Jan 14 13:07:59 and musb :) Jan 14 13:08:09 usb-storage is essential Jan 14 13:08:20 but according to FSL it doesnt work Jan 14 13:08:28 Depends on who you ask. Jan 14 13:08:40 i'll tinker with it in a spare minute some day Jan 14 13:08:42 they dont have the code for uboot there yet Jan 14 13:08:52 only thing i found was support for some fsl powerpc stuff Jan 14 13:08:58 well, should be only the driver init Jan 14 13:09:01 I once confused an FSL engineer by demonstrating a redboot provided by FSL that booted from ext2 when told that FSL redboot didn't support ext2. Jan 14 13:09:08 USB generally is in uboot Jan 14 13:09:22 i know Jan 14 13:09:33 see yourself. support isnt there imo Jan 14 13:09:38 yes Jan 14 13:09:54 but there is a kernel driver Jan 14 13:10:06 most uboot drivers are just based on kernel drivers Jan 14 13:10:26 ogra: Which means you can port it on saturday ? :) Jan 14 13:10:35 asac: no, that's why I was wondering if tobin might know Jan 14 13:10:37 nah Jan 14 13:10:41 i'm no kernel hacker Jan 14 13:10:47 but i can take a look :) Jan 14 13:11:39 plars: i remember that we had working dove images Jan 14 13:11:45 then at some day you said it broke Jan 14 13:11:59 plars, well, according to isotracker tobin didnt test dove at all Jan 14 13:14:26 did we have dove for A1 ? Jan 14 13:14:30 iirc we didnt Jan 14 13:22:49 asac, no go Jan 14 13:23:00 still comes up with a MAC of 00:00:00:.... Jan 14 13:23:55 SIOCSIFFLAGS Cannot assign requested address ... is what ifconfig gives me on ifconfig eth0 up Jan 14 13:24:21 i guess we need FSL help here Jan 14 13:25:32 * ogra hugs his USB NIC Jan 14 13:26:00 Hi guys... is there an lp bug in the pixman/NEON issues we were seeing? This needs tracking, because we may revert to CONFIG_NEON=n by default... which will make the problem appear to be solved even though it's not. Jan 14 13:26:17 s/lp bug in/lp bug on/ Jan 14 13:28:24 dmart, Bug #385553 ?? Jan 14 13:28:25 Launchpad bug 385553 in pixman "Integrate NEON optimisations for armel" [Wishlist,Fix released] https://launchpad.net/bugs/385553 Jan 14 13:29:04 we could recycle that one i assume Jan 14 13:29:05 ogra: yeah. i assume the code is completely busted Jan 14 13:29:13 it looked really bad hackish Jan 14 13:29:36 What code? Jan 14 13:29:39 asac, the fun thing is that it defaults to tftpboot Jan 14 13:29:42 ogra: the mcx_initialize function Jan 14 13:29:49 dmart, different issue :) Jan 14 13:29:51 dmart: the uboot fec driver has odd stuff Jan 14 13:29:58 Oh, right :) Jan 14 13:30:07 seems they moved to an array approach, but only did so at half of the places Jan 14 13:30:08 like: Jan 14 13:30:10 dmart, seems FEC doesnt get initialized at all when using uboot Jan 14 13:30:18 13:48 < asac> for (i = 0; i < sizeof(fec_info) / sizeof(fec_info[0]); i++) { Jan 14 13:30:24 and then they use fec_info[i] Jan 14 13:30:39 is that going to work? Jan 14 13:30:43 dmart, leaving us without NIC since the kernel cant figure out a MAC for it Jan 14 13:31:02 e.i would have expected it to be more like sizeof(struct fec_info_s) Jan 14 13:31:11 doesn't look right Jan 14 13:31:21 it looks completely wrong ;) Jan 14 13:31:25 dmart, you boot with an older uboot, right, do you have NIC support on your babbage ? Jan 14 13:31:35 esepecially since the fec_info array is a constant array Jan 14 13:31:37 (we might just be able to roll back to an older patch) Jan 14 13:31:37 that doesnt grow Jan 14 13:31:44 it should just be i < 1 ;)( Jan 14 13:31:53 let me check Jan 14 13:31:54 Does the NIC work if a MAC is just arbitrarily assigned by userspace later? Jan 14 13:32:01 It was on a nettop, and I don't remember whether the network was OK... I didn't check that. Jan 14 13:32:02 persia, nope Jan 14 13:32:13 Ugh. Jan 14 13:32:16 SIOCSIFFLAGS Cannot assign requested address Jan 14 13:32:21 no matter what i do Jan 14 13:32:29 NIC not promiscuous? Jan 14 13:32:30 the HW isnt completely up Jan 14 13:32:43 we had such issues before with redboot Jan 14 13:32:52 in jaunty iirc Jan 14 13:34:05 ogra: get the new patch from same location Jan 14 13:34:21 let me first try something Jan 14 13:34:34 * ogra issues setenv ethaddr 00:04:9F:00:E8:C2 Jan 14 13:34:46 ogra: the _initialize code doesnt initialize anything without that patch Jan 14 13:34:53 sizeof(fec_info) / sizeof(fec_info[0]) is probably 0 Jan 14 13:35:01 so no device gets initialized Jan 14 13:35:05 yeah Jan 14 13:35:22 but uboot reports FEC0 as primary interface Jan 14 13:35:38 yes Jan 14 13:35:50 so it knows its there Jan 14 13:35:51 ogra: it knows about it, but doesnt initialize it Jan 14 13:36:05 because the loop isnt entered Jan 14 13:36:37 yes, thats why i try to check if initialzing it manually changes a thing Jan 14 13:36:42 no Jan 14 13:36:48 you cant initialize it manually Jan 14 13:36:55 because the code that initializes it does nothing ;) Jan 14 13:37:06 unless you can run C code at the prompt Jan 14 13:37:34 at least i would try that patch first ;) Jan 14 13:38:43 e.g. the drivers vtable isnt filled if it really doesnt go in there Jan 14 13:39:36 same patch name ? Jan 14 13:39:52 yes Jan 14 13:39:58 hmm. if it knows the name it seems to have happened. but try anyway Jan 14 13:40:53 ah cool, alternates got added to iso tracker today Jan 14 13:41:05 yeah Jan 14 13:41:38 so i think that is probably 1 ;) ... too bad Jan 14 13:41:53 do you see any other errors ogra ? Jan 14 13:42:03 just applying the change ... Jan 14 13:42:23 one sec, cdbs-edit-patch isnt the fastest on the babbage :) Jan 14 13:43:12 ogra: copy the patch in there Jan 14 13:43:16 ogra: you should use quilt Jan 14 13:43:25 asac, the package is 100% cdbs Jan 14 13:43:33 and i wont touch quilt if i dont have to Jan 14 13:43:53 its the biggest pain in the ass i've encountered after git Jan 14 13:43:55 just replace the patch then :) Jan 14 13:44:04 oh wait, then it fails to clean Jan 14 13:44:12 hmm. quilt doesnt have that problem ;) Jan 14 13:45:22 ogra: if you are still at it, grab a new version of the patch that enables debugging Jan 14 13:45:29 CDBS simple-patchsys doesn't have that problem either. Jan 14 13:45:33 ARGH! Jan 14 13:46:01 fact is that simple-patchsys is for the weak ;) Jan 14 13:46:04 did you notice there is fec_mxc.c and mxc_fec.c in the drivers dir Jan 14 13:46:09 insanity !!! Jan 14 13:46:24 really. the one i touched at least gets compiled ;) Jan 14 13:46:26 * asac checks Jan 14 13:46:39 Well, someone decided that simple-patchsys was the easiest to teach, so lots of interesting packages end up using it. Jan 14 13:47:12 yes, but for professional development - aka if you really fix bugs and touch code - its inferior ;) Jan 14 13:47:28 throwing existing patches together is simplest yes Jan 14 13:47:44 but rebasing patches that dont apply anymore is a mess ;) Jan 14 13:47:49 building Jan 14 13:47:55 Depends on one's viewpoint. I also like quilt, which makes me a poor candidate as a simple-patchsys apologist :) Jan 14 13:47:56 another 20min ... Jan 14 13:48:26 often there are viewpoints/opinions ... but in this case there are only facts :-P Jan 14 13:48:48 ogra: make O=debian/build-*/ Jan 14 13:48:50 then debuild -nc Jan 14 13:48:59 dpkg-buildpackage -b Jan 14 13:49:25 i want a proper package to work with Jan 14 13:49:27 yeah. thats why it takes 20 instead of 2 minutes Jan 14 13:49:39 debuild -nc gives you a good package if you just change one patch Jan 14 13:51:22 ok Jan 14 13:51:29 me hopes the other fec source isnt used Jan 14 13:52:44 me too :) Jan 14 13:53:10 but given that the warnings go away after your first fix it doesnt look like Jan 14 13:53:43 * ogra takes a coffebreak Jan 14 13:53:59 ogra: i would abort the break. the enabling of debug fails to build Jan 14 13:54:03 what a bad code :( Jan 14 14:23:17 asac, to late to abort it :) Jan 14 14:28:29 asac, no change Jan 14 14:33:05 ogra: so you didnt pick the DEBUG change? Jan 14 14:33:07 ok Jan 14 14:33:10 well. then... Jan 14 14:33:26 no, i just picked the fec change Jan 14 14:33:42 i updated that a minute after you picked it ... thats why i am not sure Jan 14 14:39:32 oh, bte, we'Re at 152k for .bin Jan 14 14:39:55 vs 136k before switching on ext2 and hush Jan 14 14:40:02 *btw Jan 14 14:43:54 All, for the pixman issue I raised a new bug #507503 (since the current understanding is that it's not really a pixman bug) Jan 14 14:43:55 Launchpad bug 507503 in linux-fsl-imx51 "VFP/NEON state is not preserved around signal handlers, causing state corruption between user processes" [Undecided,New] https://launchpad.net/bugs/507503 Jan 14 15:08:30 sihg, is there any way to disable DCC offers in xchat ... Jan 14 15:08:42 that spam is driving me insane Jan 14 15:33:49 NCommander, do you happen to know if there is any way of command substitution in hush shell ? Jan 14 15:34:50 ogra, define command substitution Jan 14 15:35:14 files=$(fatls mmc 0:2) Jan 14 15:36:08 ogra, yeah, you set the command you want in a variable and then run it Jan 14 15:36:14 i.e. Jan 14 15:36:23 setenv testcmd fatls mmc 0:2 Jan 14 15:36:25 run testcmd Jan 14 15:36:29 an that gives me the flie list in the files variable ? Jan 14 15:36:39 ogra, oh, no Jan 14 15:36:42 i want a variable i can work with Jan 14 15:36:46 ogra, can't do that unfortantely Jan 14 15:36:52 right, thats what i thought Jan 14 15:36:54 * NCommander notes hush is very limited Jan 14 15:37:02 yes, found that :) Jan 14 15:37:09 How about: Jan 14 15:37:11 setenv tmp mmcload ${file} Jan 14 15:37:16 run tmp Jan 14 15:37:37 (hang on, are you still talking about UBoot?) Jan 14 15:37:41 yes :) Jan 14 15:38:01 Something like the above might work then Jan 14 15:38:06 what i want is some way to detect if uimage exists in a partition/dir Jan 14 15:38:22 but i have no grep and fatls is pretty noisy Jan 14 15:38:23 ogra, what I did for dove is just try and unconditionally load the boot.scr file Jan 14 15:38:37 ogra, if that fails, then assume the uimage is MIA, and move on Jan 14 15:38:44 "fatls mmc 0:2 uimage" unfortunately doesnt work Jan 14 15:38:45 That might work Jan 14 15:38:48 I dunno if fatls/ext2ls return an error code Jan 14 15:38:52 ogra, it should work ... Jan 14 15:38:56 That does work on Dove I think Jan 14 15:39:04 doesnt here Jan 14 15:39:13 *grummmmmmble* Jan 14 15:39:16 BBG U-Boot > fatls mmc 0:2 uimage Jan 14 15:39:16 No Fat FS detected Jan 14 15:39:24 BBG U-Boot > fatls mmc 0:2 Jan 14 15:39:24 3085760 uimage Jan 14 15:39:24 3282905 uinitrd Jan 14 15:39:35 ? Jan 14 15:39:48 it somehow interprets the last parameter Jan 14 15:39:52 Hmmm, not sure I ever tried that Jan 14 15:40:03 ogra, try fatls mmc 0:2 /uimage Jan 14 15:40:05 What does help fatload say? Jan 14 15:40:19 it says the last param is a directory :P Jan 14 15:40:32 BBG U-Boot > fatls mmc 0:2 / Jan 14 15:40:32 3085760 uimage Jan 14 15:40:32 3282905 uinitrd Jan 14 15:40:36 so yes, that works Jan 14 15:40:47 but i still dont know if uimage exists Jan 14 15:40:58 since i cant parse the return value Jan 14 15:41:24 What 's the end goal heer? Jan 14 15:41:31 s/heer/here/ Jan 14 15:41:42 i want to loop over all fat and ext2 partitions and find if there is a boot.scr Jan 14 15:42:03 if boot.scr and uimage exist i want to load boot.scr and execute it Jan 14 15:42:06 hmm arm SIMD fast paths are disabled in the pixman in lucid Jan 14 15:42:11 ogra, look at how I did it on Dove, and rewrite :-) Jan 14 15:42:19 ogra, I only check for boot.scr and not uImage Jan 14 15:42:20 boot.scr then should have all env variables needed Jan 14 15:42:25 Is there a way only to iterate over partitions that exist, or do we have to be dump and poll everything (slow?) Jan 14 15:42:38 argh s/dump/dumb/ Jan 14 15:42:41 dmart, the later, but it isn't too bad, only 3-4 seconds on Dove boot time Jan 14 15:42:43 for i in 1 2 3 4 5;do if fatls mmc 0:$i;then fatload mmc 0:$i ${loadaddr} uImage; fatload mmc 0:$i ${loadaddr_initrd} uinitrd;fi;done Jan 14 15:42:48 dmart, ^^^ that works fine Jan 14 15:43:15 and isnt to slow Jan 14 15:44:08 ogra, http://paste.ubuntu.com/356625/ Jan 14 15:44:14 my point is i want to detect if i'm using a live image or not as the very first thing ... because then i know for sure where what file is Jan 14 15:44:46 Can we do something like for ...; do if