**** BEGIN LOGGING AT Sun May 12 02:59:59 2013 May 12 19:14:08 hi GNUtoo-x60 May 12 19:14:15 hi May 12 19:14:52 did you eventually get your coreboot+kexecboot working? May 12 19:15:03 not yet May 12 19:15:21 but with coreboot->seabios->ipxe->kexecboot it works May 12 19:15:23 trough the SD May 12 19:15:24 I see May 12 19:15:33 I must now add SATA support May 12 19:15:43 btw, are you on multi-cpu? May 12 19:15:43 in the kernel May 12 19:15:47 yes May 12 19:15:52 hm.. ok May 12 19:16:11 dual-core * 1 May 12 19:16:23 model name : Genuine Intel(R) CPU T2300 @ 1.66GHz May 12 19:16:47 I see that kexec-klibc has a 'weird' patch... May 12 19:16:49 -<-----><------>nr_cpus = sysconf(_SC_NPROCESSORS_CONF); May 12 19:16:49 +<-----><------>/*nr_cpus = sysconf(_SC_NPROCESSORS_CONF);*/ May 12 19:16:49 +<-----><------>nr_cpus = 1; May 12 19:16:49 . May 12 19:17:30 this is gross hack... May 12 19:17:53 it happens to work on single-cpu, I've never tested on multi May 12 19:18:21 well, in fact is crashdump, not kexec May 12 19:38:52 ok May 12 19:39:00 should I test something? May 12 19:39:09 because normal booting works May 12 19:39:17 including with klibc version of kexecboot May 12 21:07:46 bah...http://www.gnu.org/software/libc/manual/html_node/Processor-Resources.html May 12 21:08:19 uclibc added it in 2010 May 12 21:08:27 http://busybox.net/~aldot/uClibc/parser/0002-sysconf-implement-_SC_NPROCESSORS_CONF-_SC_NPROCESSO.patch May 12 21:09:11 I'll split it in a separate patch May 12 21:09:17 for now... May 12 21:45:18 GNUtoo-x60, I've removed lot of patches and recipe still builds May 12 21:45:36 pls help with warnings later ;) May 12 21:48:19 ok May 12 21:53:11 I fear I've discovered a multilib flaws in our klibc :/ May 12 22:04:50 GNUtoo-x60, last two questions: May 12 22:05:07 1) I've tested EXTRA_OECONF += "--without-zlib --without-lzma --without-xen" May 12 22:05:13 seems right? May 12 22:05:57 hmmm May 12 22:06:14 2) basename = strrchr(name, '/') + 1; May 12 22:06:17 ok May 12 22:06:27 I did the basename patch too fast May 12 22:08:07 ok 2) sounds good May 12 22:08:16 ok May 12 22:08:21 about 1st... ??? May 12 22:08:22 C_HELP_STRING([--without-zlib],[disable zlib support]) May 12 22:08:56 I'm a bit confused..we load an lzma compressed kernel so I've not tested May 12 22:09:08 it was disabling zlib only May 12 22:09:20 maybe I misunderstand May 12 22:09:42 http://git.kernel.org/cgit/utils/kernel/kexec/kexec-tools.git/tree/configure.ac May 12 22:24:24 hmmm May 12 22:24:43 but the kernel has its own decompression right? May 12 22:24:57 well .. I see both kexec/lzma.c and kexec/zlib.c are compiled May 12 22:25:03 ok May 12 22:25:06 so it defaults to internal May 12 22:25:12 maybe we should test May 12 22:25:20 see thi: May 12 22:25:20 http://lists.infradead.org/pipermail/kexec/2009-November/003622.html May 12 22:25:42 note that I've previously kexec-booted lzma kernels, prior this patch May 12 22:26:03 lzma e bzImage bzImage.lzma May 12 22:26:03 kexec -l bzImage.lzma May 12 22:26:09 when it was EXTRA_OECONF += "--without-zlib" May 12 22:26:15 ok May 12 22:26:26 maybe try it May 12 22:26:29 remove lzma May 12 22:26:36 and try to boot an internal-lzma kernel May 12 22:27:11 I can try on x86 too if you want May 12 22:28:55 0000d100 T lzma_decompress_file May 12 22:29:05 0000d108 T zlib_decompress_file May 12 22:29:12 so it is emnbedded May 12 22:29:37 ok May 12 22:30:05 ok, I'll send patchset in a few mins May 12 22:31:08 ok thanks a lot May 12 22:31:16 sorry for not dedicating enough time to that May 12 23:25:32 patches sent May 12 23:25:35 gn **** ENDING LOGGING AT Mon May 13 02:59:58 2013