**** BEGIN LOGGING AT Sat Feb 09 02:59:58 2013 Feb 09 06:37:03 wookey / cjwatson: arm64-cross-toolchain-base is up to date again, enjoy. Feb 09 08:45:24 * ogra_ grumbles ... seems one sick cat isnt enough ... Feb 09 09:06:53 robclark , hrw oh ... okay Feb 09 09:07:05 then I'm back to square one basically Feb 09 09:07:16 can't find compatible hardware that I can play with Feb 09 09:07:29 or the ones that are compatible are ETOOEXPESIVE Feb 09 09:08:08 *EXPENSIVE Feb 09 15:38:02 hello, I have a question about the /debian/* files in a source directory Feb 09 15:38:39 is it possible to manually install them if I compiled the program and installed it manually or do i have to build a deb package? Feb 09 17:41:38 Hello, I have a few newbie questions. 1. Please tell me is there any differance between Ubuntu x86 and ARM root filesystem? Feb 09 17:42:17 Differance in source Feb 10 01:46:07 I have an arm cpu question: http://www.ti.com/lit/ug/spruf98x/spruf98x.pdf page 2614: OVF_Rate = (0xFFFF FFFF - GPTn.TLDR + 1) * (timer-functional clock period) * PS table16-11 the example for TLDR=0xFFFF FFF0 and PS=1 is equal to 524 μs. however, (0xFFFF FFFF - 0xFFFF FFF0 + 1) * (1/32768) * 1 = 488 μs Feb 10 01:46:17 can anyone explain why I calculate the value different than the TRM? Feb 10 02:21:16 infinity: toolchain seems to basically work, but for some reason gpm fails to build. Feb 10 02:21:37 hmm same failure for native build. Maybe it just doesn't like 2.17 Feb 10 02:23:17 Does it build for you? Feb 10 02:24:06 I'm getting http://people.linaro.org/~wookey/buildd/raring-arm64/gpm_1.20.4-6-raring-arm64-20130210-0142.log with a lot of warnings, then missing major() Feb 10 02:25:40 aha. OE people have fixed it: http://patches.openembedded.org/patch/42449/ Feb 10 02:32:17 ok, that does the trick Feb 10 02:34:35 I suspect that's the "wrong" fix, but it works for now. Feb 10 02:42:18 is there a list of 2.17 changes that may break things like this? Feb 10 02:44:22 This is the only one I've run into so far. And yeah, actually, that may be the right fix for gpm. Feb 10 02:44:40 Looks like they were relying on fcntl.h misatkenly including sys/types.h, which it shouldn't. Feb 10 02:44:58 Of course, you can also get sys/types.h for free with enabling some system extensions. Feb 10 02:45:15 Since stdlib.h has this: Feb 10 02:45:16 #if defined __USE_SVID || defined __USE_XOPEN_EXTENDED || defined __USE_BSD Feb 10 02:45:16 # include /* we need int32_t... */ Feb 10 02:55:43 wookey: Anyhow, uploading the fix to Ubuntu now, and submitting to Debian. Feb 10 02:57:25 Ah OK. I'll not file this patch then Feb 10 02:57:31 cheers Feb 10 02:57:55 Guess it's about time for another rebuild test in Ubuntu. Feb 10 02:58:19 I know of one other case of "I included fcntl.h, but I actually meant sys/types", but I don't know how rampant it is. Feb 10 02:58:39 do you recall which package? Feb 10 02:58:52 alsa-lib Feb 10 02:59:31 (Which ends up manifesting in a cascaded fashion to many things that build-dep on libasound-dev, since the alsa headers have the broken assumption) Feb 10 02:59:57 I should probably fix that right now too. **** ENDING LOGGING AT Sun Feb 10 02:59:58 2013