**** BEGIN LOGGING AT Tue Apr 27 02:59:57 2010 Apr 27 03:02:58 hi! it's better to use binutils 2.20.1 vs 2.19.1? Apr 27 06:16:11 gmorning Apr 27 08:39:09 acoul * r21180 /trunk/include/kernel-version.mk: add 2.6.33.3 & 2.6.32.12 kernel checksums Apr 27 08:40:59 acoul * r21181 /trunk/target/linux/adm5120/Makefile: adm5120: use kernel 2.6.32.12 Apr 27 08:46:24 acoul * r21182 /trunk/target/linux/adm5120/ (40 files in 4 dirs): adm5120: add 2.6.33 & 2.6.34 kernel support Apr 27 10:09:38 rhk * r21183 /packages/net/hiawatha/Makefile: [packages] hiawatha: update to 7.2 Apr 27 11:42:50 acoul * r21184 /trunk/package/grub/Makefile: package/grub: fix cross compile options Apr 27 11:50:41 nico * r21185 /trunk/package/busybox/patches/000-upstream-sed.patch: package/busybox: refresh patches Apr 27 11:59:20 nico * r21186 /trunk/package/busybox/ (4 files in 2 dirs): package/busybox: add 2 upstream fixes (ash & cpio), bumb release number Apr 27 13:07:53 jow * r21187 /trunk/package/cyassl/patches/120-makefile_destdir.patch: [package] cyassl: remove stray "\r" in hunk header of 120-makefile_destdir.patch which incorrectly triggers CR stripping in "patch" resulting in non-fitting hunks (#6954) Apr 27 13:08:16 acoul * r21188 /trunk/target/linux/rb532/ (config-2.6.32 config-2.6.33 config-2.6.34): rb532: refresh config options Apr 27 13:08:31 jow * r21189 /branches/backfire/package/cyassl/patches/120-makefile_destdir.patch: [backfire] merge r21187 Apr 27 13:11:29 acoul * r21190 /branches/backfire/target/linux/rb532/config-2.6.32: [backfire] merge r21188 Apr 27 13:46:24 rhk * r21191 /trunk/target/linux/brcm63xx/ (2 files in 2 dirs): [brcm63xx] add extended platform devices patch, thx Miguel & Florian Apr 27 13:46:56 ooo what's that? Apr 27 13:48:27 rhk * r21192 /trunk/target/linux/brcm63xx/ (2 files in 2 dirs): [brcm63xx] add NXP 74HC164 chip driver support, thx Miguel & Florian Apr 27 13:49:32 rhk * r21193 /trunk/target/linux/brcm63xx/ (3 files in 3 dirs): [brcm63xx] add neufbox 4 board support, thx Miguel & Florian Apr 27 14:08:40 ralph * r21194 /trunk/package/ifxmips-dsl-api/ (9 files in 3 dirs): [ifxmips-dsl-api] fix .32 kernel, add support for .33 kernel Apr 27 14:13:30 hi ppl! Apr 27 14:15:00 What is the status of openwrt for the RB411? Apr 27 14:15:27 rock stable Apr 27 14:15:39 bring image! Apr 27 14:22:58 my wireless is not on SPI Apr 27 14:23:11 gotta wonder how it is connected, maybe USB? Apr 27 14:23:26 <[florian]> ali1234: I doubt so, I do not think bcm6338 even has usb Apr 27 14:23:34 <[florian]> ali1234: could be PCI, but I am still not sure about that Apr 27 14:23:58 it looks like it only has 5 wires going to the wifi module Apr 27 14:24:13 <[florian]> your PCB is probably 4 or 6 layers Apr 27 14:24:21 <[florian]> so the others might be in under layers Apr 27 14:24:44 <[florian]> which is a must if doing pci, since the tracks are quite sensitive to external perturbations Apr 27 14:25:15 hmm i have an idea Apr 27 14:25:39 i can dump the registers on jtag while running original firmware and see if the PCI registers are there Apr 27 14:25:56 <[florian]> yes, you can do that, indeed Apr 27 14:27:46 well SPI is working, the switch is device 1 Apr 27 14:27:55 there doesn't seem to be anything else connected on it Apr 27 14:30:12 acoul * r21195 /trunk/ (24 files in 24 dirs): sync to kernel 2.6.32.12 Apr 27 14:43:42 acoul * r21196 /trunk/target/linux/x86/ (patches/ patches-2.6.32/): x86: prepare for kernels 2.6.33 & 2.6.34 Apr 27 15:30:37 this is what the original firmware says when booting: http://pastebin.com/wbMVzW2U Apr 27 15:30:56 well, part of it, and after removing the cross talk from two things writing to the serial port at the same time Apr 27 15:35:51 ralph * r21197 /trunk/package/ifx-tapi/ (6 files in 3 dirs): [ifx-tapi] add .33 support Apr 27 15:37:17 ralph * r21198 /trunk/package/ifx-vmmc/ (5 files in 3 dirs): [ifx-vmmc] add .33 support Apr 27 15:45:07 rhk * r21199 /trunk/target/linux/brcm63xx/ (2 files in 2 dirs): [brcm63xx] extend bcm63xx_wdt driver for debugging purposes, thx Miguel Apr 27 15:45:46 ralph * r21200 /trunk/target/linux/ifxmips/ (16 files in 3 dirs): [ifxmips] add ar9 target Apr 27 15:46:47 ralph * r21201 /trunk/target/linux/ifxmips/files-2.6.30/arch/mips/include/asm/mach-ifxmips/ (ifxmips.h ifxmips_irq.h): [ifxmips] add external irq Apr 27 15:47:31 http://www.dd-wrt.com/phpBB2/viewtopic.php?p=153493&sid=bb3b133191629c977df4ff68652ad180 this stuff seems to indicate that the wireless is on SDIO Apr 27 15:47:52 the wifi firmware used is the same, and it's also 6338 Apr 27 15:47:52 ralph * r21202 /trunk/target/linux/ifxmips/files-2.6.32/include/ (. asm-mips/ asm-mips/bitsperlong.h): [ifxmips] fix atm compilation Apr 27 15:48:52 ralph * r21203 /trunk/target/linux/ifxmips/danube/ (config-2.6.30 config-2.6.32 config-2.6.33 target.mk): [ifxmips] update danube target Apr 27 15:50:58 weird though, spi does not seem connected to the wifi Apr 27 15:51:11 i wonder if this chip actually has some sdio interface on it Apr 27 15:53:05 ralph * r21204 /trunk/target/linux/ifxmips/Makefile: [ifxmips] update Makefile Apr 27 15:56:03 broadcom code defines a separate interrupt for SDIO, i think it could Apr 27 15:58:26 blogic * r21205 /trunk/package/mountd/Makefile: fixes file perms of mountd uci file and init.d script, thx yann.lopez Apr 27 17:30:15 nico * r21206 /branches/backfire/target/linux/rdc/modules.mk: [backfire] merge r21129 Apr 27 18:50:58 nbd * r21207 /branches/backfire/target/linux/ar71xx/files/arch/mips/ar71xx/irq.c: [backfire] ar71xx: backport pci/dma fix from r21143 Apr 27 18:55:27 KanjiMonster: what bcm63xx do you have? Apr 27 19:25:04 looks like the SDIO host driver is somehow tied in to the DSL stuff we don't have: http://pastebin.com/wXciqa1s Apr 27 19:28:42 nico * r21208 /branches/backfire/ (24 files in 24 dirs): [backfire] merge r20834, r20840, r20854 & r20855 Apr 27 19:32:41 <[florian]> ali1234: I do not think so, they are completely separate Apr 27 19:32:41 ali1234: you sure the sdio init doesn't happen after dsl init Apr 27 19:32:52 maybe but check this out Apr 27 19:34:18 "ATM and SDIOH share the same reset bit" Apr 27 19:34:30 it then proceeds to toggle the SAR reset bit Apr 27 19:34:45 that's from broadcom: kernel/linux/arch/mips/brcm-boards/bcm963xx/setup.c Apr 27 19:35:02 so they are not "completely" separate :) Apr 27 19:35:20 <[florian]> :) Apr 27 19:36:29 <[florian]> ali1234: you should be able to make bcm4318 work with the spi driver + adding the stub to support an sdio irq Apr 27 19:36:38 nope, it isn't connected on SPI Apr 27 19:36:50 <[florian]> ali1234: oh there is a dedicated sdhc controller then? Apr 27 19:36:57 yes Apr 27 19:37:06 <[florian]> good luck then Apr 27 19:37:07 well that's the way it looks Apr 27 19:37:23 why else would it haveits own irq? Apr 27 19:37:39 <[florian]> because the irq line cannot be signaled using spi Apr 27 19:38:54 the 5 lines that look like they go into the wifi are not electrically connected to the spi bus anyway Apr 27 19:39:05 well, they don't look like they go in, they definitely do :) Apr 27 19:41:30 there's a lot of stuff in 0x2000 space that could be SDIO, more so if this thing doesn't have USB Apr 27 19:43:12 <[florian]> http://android.git.kernel.org/?p=platform/system/wlan/broadcom.git;a=blob;f=src/bcmsdio/sys/bcmsdh.c;h=afbf5e0d6885423bbf5fe77ebe1ea35e76c36e30;hb=f42f81d9f222732884ecf14adfeaee6a58e8a534 Apr 27 19:43:39 was looking at that earlier, it runs on top of a sdio host Apr 27 19:43:58 <[florian]> it seems to have a spi binding actually Apr 27 19:44:05 <[florian]> http://android.git.kernel.org/?p=platform/system/wlan/broadcom.git;a=blob;f=src/bcmsdio/sys/bcmsdspi.c;h=d22fbb2de788bda1c79fde658a7785d458ed8db9;hb=f42f81d9f222732884ecf14adfeaee6a58e8a534 Apr 27 19:45:47 well... if i could get it to respond on spi... Apr 27 19:45:55 but only device 1 responds, and that's the switch Apr 27 19:46:13 maybe it has pass-thru of some kind Apr 27 19:47:16 <[florian]> or maybe you are holding it in reset Apr 27 19:47:26 how? Apr 27 19:47:54 actually i do see something interesting Apr 27 19:48:08 on the original firmware the spi cmd reg contains 0x213 Apr 27 19:48:38 1 = device, 3 = command, 2 = ? Apr 27 19:49:02 <[florian]> no idea Apr 27 19:49:02 oh, byte ount Apr 27 19:49:08 not interesting Apr 27 19:51:40 i'm not sure spi would be fast enough anyway, with max clock of 25mhz and a 63 byte fifo... Apr 27 19:52:05 <[florian]> sdio is not made for fast rates anyway ;~ Apr 27 19:52:54 according to wikipedia it can go up to 80mbit Apr 27 19:54:05 hmmmmm Apr 27 19:55:45 well, maybe it *is* on spi... apparently sdio standard is 25mhz clock too Apr 27 19:55:57 [florian]: looks like the ar525w image for backfire is kinda broken Apr 27 19:56:11 [florian]: the kernel seems to be only 17kb Apr 27 19:56:42 ali1234: I don't know, but i think its a 6348 (its hidden benath the daughter bord) Apr 27 19:56:48 <[florian]> rtz2: ok, will check that Apr 27 19:57:01 <[florian]> KanjiMonster: if it has a bcm6341 with it, yes it's a bcm6348 Apr 27 19:58:14 hmm ok, if it is not a separate controller, why does it need a separate reset bit? Apr 27 20:04:46 <[florian]> it says it reset bit is shared apparently Apr 27 20:05:07 [florian]: it may be a problem with the "strip unused kernel symbols" option, it seems to be too aggresive on rdc Apr 27 20:05:21 <[florian]> rtz2: ok Apr 27 20:05:24 [florian]: but this is only a guess Apr 27 20:05:42 [florian]: yes shared with the SAR... Apr 27 20:05:50 <[florian]> rtz2: or simply, there is neither a kernel nor the rootfs :) Apr 27 20:06:26 [florian]: no, both are there Apr 27 20:06:57 <[florian]> ali1234: no idea why it is shared Apr 27 20:07:16 [florian]: and the rootfs looks good as far as the size goes Apr 27 20:08:30 [florian]: and the intial x86 startup code also seems to be there Apr 27 20:08:44 but the actual kernel code is kinda missing Apr 27 20:08:53 <[florian]> rtz2: I guess your assumption is right **** BEGIN LOGGING AT Tue Apr 27 20:12:00 2010 Apr 27 20:16:30 nbd * r21209 /trunk/target/linux/ar71xx/profiles/ (00-default.mk 01-madwifi.mk 01-minimal.mk 02-madwifi.mk): ar71xx: add a new default profile containing all drivers for all boards Apr 27 20:55:52 could somebody explain to me the conecpt of the mtd unlock command? Apr 27 20:56:20 the ioctl only produces an EOPNOTSUPP error for me Apr 27 21:01:09 some flash chips can lock blocks to prevent them from being erased or written to Apr 27 21:01:18 mtd unlock calls the ioctl to unlock blocks Apr 27 21:01:42 if the driver returns -EOPNOTSUPP, we should just ignore that Apr 27 21:02:09 nbd: ahh, ok Apr 27 21:02:10 thanks Apr 27 21:57:35 nbd * r21210 /trunk/tools/Makefile: mkimage fails to build on some systems if sed is not installed, add a dependency to fix parallel build issues Apr 27 21:59:19 nbd * r21211 /branches/backfire/target/linux/ar71xx/profiles/ (00-default.mk 01-madwifi.mk 01-minimal.mk 02-madwifi.mk): [backfire] ar71xx: backport profile change from r21209 Apr 27 21:59:19 nbd * r21212 /branches/backfire/tools/Makefile: [backfire] backport mkimage dependency from r21210 Apr 27 22:16:42 jow * r21213 /trunk/target/imagebuilder/ (Makefile files/Makefile): Apr 27 22:16:42 [imagebuilder] Apr 27 22:16:42 - correctly copy .config when scripts/env is used in the buildroot Apr 27 22:16:42 - prevent user provides PACKAGES from overriding per-profile defaults Apr 27 22:22:13 jow * r21214 /branches/backfire/target/imagebuilder/ (Makefile files/Makefile): [backfire] merge r21213 Apr 27 23:35:40 80443148: 1462000d bne $3,$2,0x80443180 Apr 27 23:35:45 gotcha suckers Apr 27 23:35:55 that's the wifi chip ID Apr 27 23:37:18 wait Apr 27 23:37:29 it is, but it's used as an address :( Apr 28 00:26:48 hah, i booted that beetel 6338W image Apr 28 00:27:03 it did not detect wireless orrectly Apr 28 00:27:21 it did detect the ethernet switch though Apr 28 00:27:41 gah its passworded Apr 28 00:30:58 this is the most ridiculously crippled shell ever Apr 28 00:41:50 that's weird, even though the wifi does not work, the wifi led lit. never seen that before, it isn't on a normal gpio Apr 28 01:39:29 nico * r21215 /packages/libs/libdlna/Makefile: [packages] libdlna: fix build failure introduced by [21056], restore fix from [12678] (closes: #7200, #7206) Apr 28 01:51:08 nico * r21216 /packages/multimedia/ushare/Makefile: [packages] ushare: apply the same fix as libdlna in [21215], use PKG_INSTALL **** ENDING LOGGING AT Wed Apr 28 02:59:56 2010