**** BEGIN LOGGING AT Fri Nov 19 02:59:58 2010
Nov 19 03:20:00 <builder> build #30 of ar71xx is complete: Failure [failed compile_7]  Build details are at http://tksite.gotdns.org:8010/builders/ar71xx/builds/30
Nov 19 03:50:33 <builder> build #29 of sibyte is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/sibyte/builds/29
Nov 19 04:32:42 <CIA-48> flyn * r24034 /packages/libs/libdmapsharing/Makefile: Update libdmapsharing to 2.1.10
Nov 19 05:14:56 <builder> build #26 of iop32x is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/iop32x/builds/26
Nov 19 05:25:41 <CIA-48> flyn * r24035 /packages/libs/libdmapsharing/Makefile: Update libdmapsharing to 2.1.11
Nov 19 05:30:23 <builder> build #29 of avr32 is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/avr32/builds/29
Nov 19 07:31:02 <CIA-48> juhosg * r24036 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-dir-825-b1.c:
Nov 19 07:31:02 <CIA-48> ar71xx: D-Link DIR-825 B1 eth1 MAC address is garbage
Nov 19 07:31:02 <CIA-48> r22700 contained a copy-and-paste error that set eth0's MAC address
Nov 19 07:31:02 <CIA-48> twice and never set eth1's.
Nov 19 07:31:02 <CIA-48> Signed-off-by: Mark Mentovai <mark@moxienet.com>
Nov 19 08:41:03 <builder> build #26 of ep93xx is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/ep93xx/builds/26
Nov 19 08:41:46 <acoul> hello
Nov 19 08:46:50 <CIA-48> jow * r24037 /trunk/package/openssl/Makefile: [package] openssl: update to 0.9.8p
Nov 19 08:48:29 <CIA-48> jow * r24038 /branches/backfire/package/openssl/Makefile: [backfire] merge r24037
Nov 19 09:07:17 <stintel> Weedy: havent tried mixing AP + STA
Nov 19 09:13:16 <russell_> stintel, Weedy: i have on an accton mr3201a (atheros)
Nov 19 09:13:35 <russell_> wds/sta + ap
Nov 19 09:13:54 <stintel> russell_: works well?
Nov 19 09:14:00 <russell_> reasonably
Nov 19 09:14:11 <russell_> it's been up for ... /me looks
Nov 19 09:14:46 <russell_> a week or so
Nov 19 09:15:01 <russell_> has moved a crapload of data
Nov 19 09:15:04 <stintel> k, thanks for info
Nov 19 09:16:10 <russell_> effectively, it's a repeater with a wds backhaul on the same radio
Nov 19 09:32:37 <russell_> i'm wondering why the x86 net48xx image makes weird sized partitions ... with BUILD_LOG=1 V=99 does the image prep get logged anywhere?  grepping didn't find it for me.
Nov 19 09:33:44 <russell_> when this thing is being constructed: openwrt-x86-generic-combined-squashfs.img
Nov 19 09:34:07 <xMff> isnt the partition size stuff configured in menuconfig?
Nov 19 09:34:30 <russell_> there are rough numbers, 4MB and 48MB respectively
Nov 19 09:34:46 <russell_> but, the partitions that get created are kind of odd
Nov 19 09:34:56 <russell_> hda1 is 9072 sectors
Nov 19 09:35:05 <russell_> then there is a 63 sector gap
Nov 19 09:35:14 <russell_> then hda2 is 98721 sectors
Nov 19 09:35:29 <xMff> check  target/linux/x86/image/Makefile
Nov 19 09:35:53 * russell_ checks
Nov 19 09:50:02 <builder> build #27 of ixp4xx is complete: Failure [failed compile_7]  Build details are at http://tksite.gotdns.org:8010/builders/ixp4xx/builds/27
Nov 19 09:50:38 <russell_> in gen_image_x86.sh, why is KERNELOFFSET a function of the output filename?  or am i reading that wrong?
Nov 19 09:52:14 <russell_> oh.  that set `ptgen ... ` is maybe modifying those values.
Nov 19 10:12:21 <russell_> can someone explain this: return result * ((2 << ((10 * exp) - 1)) ?: 1);
Nov 19 10:13:38 <russell_> if exp == 0, the shift becomes 2 << -1, which (at least in perl) seems to give me zero.
Nov 19 10:14:42 <russell_> hmm, maybe perl is non-C like that way
Nov 19 10:14:44 <xMff> ?: 1 turns it into 1 then
Nov 19 10:17:44 <russell_> yeah, but exp is zero if there is no terminating unit specifier or it's 'k'.  so 1024k with still turn into 1, which seems wrong
Nov 19 10:18:01 <russell_> oh.  no.
Nov 19 10:18:36 <russell_> yeah, nm.  result * 1
Nov 19 10:18:39 <russell_> so that's okay
Nov 19 10:31:18 <russell_> hmm.  so it looks like my gaps are due to partitions getting rounded to cylinder boundaries, or what seems to suppose to be cylinder boundaries. (for the fake cylinders)
Nov 19 10:36:18 <russell_> i am still unclear why that to_kbytes expression isn't just: result * (1 << (10 * expr))
Nov 19 10:36:38 <russell_> exp rather
Nov 19 10:36:57 <russell_> why is all the hairyness justified?
Nov 19 10:39:44 <xMff> idk :)
Nov 19 10:40:10 <xMff> the whole sector/cylinder/whatever stuff was always over my head
Nov 19 10:43:52 <EqUaTe> it's because they always count from 0
Nov 19 10:44:01 <EqUaTe> so 1024 == 1023
Nov 19 10:44:36 <russell_> you don't multiply by 1023 though
Nov 19 10:46:08 <russell_> that's not it
Nov 19 10:48:28 <russell_> the cylinder boundary thing seems silly because the first 63 sectors are reserved for the bootloader and partition table crap, so the first partition isn't on a cylinder boundary anyway!
Nov 19 10:49:10 <russell_> but, i am ignorant of the needs of those things, so maybe it really is sane afterall
Nov 19 10:54:14 <russell_> russell@frieda[/src/openwrt/build_dir/host/firmware-utils/bin] $ ./ptgen -o /tmp/test.img -h 16 -s 63 -p 4m -p 48m
Nov 19 10:54:14 <russell_> 32256
Nov 19 10:54:14 <russell_> 4612608
Nov 19 10:54:14 <russell_> 4677120
Nov 19 10:54:14 <russell_> 50545152
Nov 19 10:57:21 <russell_> aha
Nov 19 11:00:01 <russell_> there is a sectors offset being applied to *all* partitions, not just the first one, which accounts for the 63 sector gap between the first and second partition.
Nov 19 11:01:37 <cshore> russell_: I'm pretty sure you've found the problem then....only the first partition should be not exactly on a cylinder boundry
Nov 19 11:01:49 <russell_> yeah
Nov 19 11:01:58 <russell_> seems like
Nov 19 11:02:30 * russell_ reading http://en.wikipedia.org/wiki/GUID_Partition_Table for general enlightenment
Nov 19 11:05:30 <cshore> ping [florian]
Nov 19 11:10:51 <[florian]> cshore: pong
Nov 19 11:11:48 <cshore> Hi there.  Just trying to add a new brcm63xx router...I've figured that the ethernet is on enet1 and the switch is a bcm5325E
Nov 19 11:12:22 <cshore> but when the system boots it tries to attach robo-switch to eth0 and fails to add to eth1
Nov 19 11:12:42 <cshore> There is no actual eth0 ... it's a DSL modem/router
Nov 19 11:13:00 <cshore> any guidance?
Nov 19 11:13:09 <cshore> the ethernet works
Nov 19 11:13:15 <cshore> on eht1
Nov 19 11:13:20 <cshore> *eth1
Nov 19 11:20:52 <cshore> I tried putt the .has_phy and .use_internal_phy on enet0, but when I do that eth1 is not detected (and eth0 doesn't do anything)
Nov 19 11:21:06 <cshore> *putting
Nov 19 11:21:20 <cshore> *that is I put them on enet1
Nov 19 11:22:19 <cshore> In the stock firmware the switch is activated via SPI Slave 1
Nov 19 11:26:43 <cshore> ethernet is a BCM634A1 (part of SoC 6358)
Nov 19 11:27:04 <cshore> BCM635A1 that is
Nov 19 11:29:22 <KanjiMonster> cshore: my solution was to just define .has_enet1 = 1, but not .has_enet0; though I am currently not quite sure it creates the network config correctly
Nov 19 11:30:19 <KanjiMonster> cshore: see https://lists.openwrt.org/pipermail/openwrt-devel/2010-November/008576.html
Nov 19 11:39:56 <cshore> BTW I have committed a patch for imagetag that stores the router name (as used in the filename) in info section1 followed by space and the OpenWRT revision
Nov 19 11:40:22 <cshore> I intend to use this to modify the patch you submitted
Nov 19 11:40:42 <cshore> so that the Tecom boards can have their different GPIOs
Nov 19 11:43:13 <KanjiMonster> ah, great
Nov 19 11:43:19 <KanjiMonster> yeah, I saw the commit
Nov 19 11:48:33 <cshore> does your board also have the second boot crc error?
Nov 19 11:49:38 <cshore> bah, fails to enable switch
Nov 19 11:51:25 <CIA-48> kaloz * r24039 /trunk/target/linux/avr32/image/u-boot/ (3 files in 2 dirs): [avr32]: upgrade u-boot to v2010.12-rc1
Nov 19 11:52:07 <cshore> it works without vlans, but that's it
Nov 19 11:54:30 <KanjiMonster> cshore: nope, mine is booting fine even on second boot. Btw, what's your device?
Nov 19 11:54:46 <cshore> Telsey
Nov 19 11:55:05 <cshore> CPA-ZNTEsomething-or-other
Nov 19 11:55:16 <cshore> CPVA642 board
Nov 19 12:04:28 <KanjiMonster> cshore: mtd already has a fixtrx (or somehing similar) for updating the image size/crcs on first boot; you can try to use that (probably needs to be adapted to bcm63xx); assuming the error is because of a crc missmatch of the data partion (since the jffs2-marker got replaced)
Nov 19 12:04:46 <cshore> that's right
Nov 19 12:04:51 <cshore> that's the plan
Nov 19 12:20:46 <KanjiMonster> cshore: I haven't tested whether the switch on my device works, but it gets identified (it's also a 5325E, identified as a 5325/5350)
Nov 19 12:22:39 <cshore> KanjiMonster: I suspect the MDIO isn't hooked up for the switch and you have program it via SPI (which functionality OpenWRT is missing)...I have another board like that
Nov 19 12:22:56 <cshore> KanjiMonster: you might have board that is hooked up
Nov 19 12:23:15 <cshore> KanjiMonster: depends on that the manufacturer did
Nov 19 12:23:53 <KanjiMonster> cshore: probably; the 5350/5365-identification works by writing to and reading switch registers, which probably requires the switch to be available through MII
Nov 19 12:28:30 <cshore> KanjiMonster: either that or there is a phy that isn't detected
Nov 19 12:59:57 <builder> build #26 of rdc is complete: Failure [failed compile_7]  Build details are at http://tksite.gotdns.org:8010/builders/rdc/builds/26
Nov 19 13:33:18 <builder> build #30 of ubicom32 is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/30
Nov 19 13:42:26 <builder> build #26 of au1000 is complete: Failure [failed compile_7]  Build details are at http://tksite.gotdns.org:8010/builders/au1000/builds/26
Nov 19 13:48:54 <CIA-48> cshore * r24040 /trunk/target/linux/brcm63xx/ (image/Makefile patches-2.6.35/310-CPVA642_board.patch): [brcm63xx] Initial support for Telsey CPVA642 boards (no vlans, no DSL, GPIO led/buttons not known).
Nov 19 13:50:46 <builder> build #24 of xburst is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/xburst/builds/24
Nov 19 15:22:41 <builder> build #27 of adm5120 is complete: Failure [failed compile_7]  Build details are at http://tksite.gotdns.org:8010/builders/adm5120/builds/27
Nov 19 18:53:29 <builder> build #26 of octeon is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/octeon/builds/26
Nov 19 19:15:44 <russell_> someone just pointed this out to me ... fwiw: http://seclists.org/fulldisclosure/2010/Nov/228
Nov 19 20:05:44 <nunojpg> I understand why we can't create a broadcast domain bridge over wireless, but I've just tried vmware player bridge connections, and it appears to do that
Nov 19 20:06:13 <nunojpg> to the host machine have a wireless connection with IP A, and the virtualized machine have IP B
Nov 19 20:06:25 <nunojpg> both are routable from the network
Nov 19 20:06:29 <nunojpg> how is this possible?
Nov 19 20:25:30 <xMff> arpnat
Nov 19 20:25:47 <xMff> and/or proxy arp
Nov 19 21:40:00 <builder> build #26 of ifxmips is complete: Failure [failed shell]  Build details are at http://tksite.gotdns.org:8010/builders/ifxmips/builds/26
Nov 19 22:03:47 <builder> build #28 of ps3 is complete: Failure [failed compile_3]  Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/28
Nov 19 22:09:54 <CIA-48> jow * r24041 /trunk/package/hostapd/patches/600-terminate_on_setup_failure.patch: [package] hostapd: terminate mainloop if interface setup fails, prevents blocking init or /sbin/wifi if an incorrect channel is configured
Nov 19 22:59:32 <builder> build #26 of mpc52xx is complete: Failure [failed shell_13]  Build details are at http://tksite.gotdns.org:8010/builders/mpc52xx/builds/26
Nov 19 23:23:11 <builder> build #30 of pxcab is complete: Failure [failed compile_3]  Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/30
Nov 19 23:31:55 <CIA-48> jow * r24042 /trunk/package/openssl/patches/200-ocf-20100325.patch: [package] openssl: refresh 200-ocf-20100325.patch
Nov 19 23:38:31 <CIA-48> jow * r24043 /branches/backfire/package/openssl/patches/ (4 files): [backfire] openssl: sync patches with trunk
Nov 19 23:48:30 * russell_ has a linksys e2000 open in front of me, looking for something that looks like serial
Nov 19 23:49:07 <russell_> there is something that looks like jtag, although miniature
Nov 19 23:49:17 <russell_> 2x7
Nov 20 00:00:50 <russell_> aha, i think i found it
Nov 20 00:22:48 <roh> russell--: if unsure, make a picture
Nov 20 00:33:28 <russell_> wow, i'm getting better at this.  got the pinout on the first try
Nov 20 00:38:23 <roh> wait till you have done 50 or more different ones
Nov 20 01:20:52 <EqUaTe> russell_: the serial pinout for the e2000 is on the flat things on the edge..
Nov 20 01:21:00 <EqUaTe> there's documentation on the ddwrt forums
Nov 20 01:21:00 <EqUaTe> :P
Nov 20 01:39:28 <builder> build #33 of brcm47xx is complete: Failure [failed compile_7]  Build details are at http://tksite.gotdns.org:8010/builders/brcm47xx/builds/33
Nov 20 02:33:47 <CIA-48> nbd * r24044 /trunk/package/mac80211/ (5 files in 2 dirs): mac80211: update to latest wireless-testing + more stability work
Nov 20 02:50:49 <builder> build #35 of brcm63xx is complete: Failure [failed compile_7]  Build details are at http://tksite.gotdns.org:8010/builders/brcm63xx/builds/35
**** ENDING LOGGING AT Sat Nov 20 02:59:57 2010