On Wed, Mar 02, 2005 at 05:06:36PM +0300, Michael Tokarev wrote: > Herbert Poetzl wrote: > [] > >BUG_ON() and friends are still broken (at least on x86) > [] > >Freeing unused kernel memory: 244k freed > >------------[ cut here ]------------ > >kernel BUG at <bad filename>:9377! > > ~~~~~~~~~~~~~~~~~~~ > > Have you tried compiling with CONFIG_DEBUG_INFO=y ? > (Looks like CONFIG_DEBUG_BUGVERBOSE should either > depend on CONFIG_DEBUG_INFO or turn it on).
egrep 'CONFIG_DEBUG_(INFO|BUGVERBOSE)' .config # CONFIG_DEBUG_BUGVERBOSE is not set CONFIG_DEBUG_INFO=y good point, just it seems to be the other way round .. with CONFIG_DEBUG_BUGVERBOSE set I get the correct file and line, but no stack information ... Freeing unused kernel memory: 104k freed ------------[ cut here ]------------ kernel BUG at init/main.c:692! invalid operand: 0000 [#1] PREEMPT CPU: 0 EIP: 0060:[<c0100330>] Not tainted VLI EFLAGS: 00000246 (2.6.11-rc1) eax: 00000000 ebx: c01002b0 ecx: 00000000 edx: 00000000 esi: 00000000 edi: 00000000 ebp: 00000000 esp: c10bdff0 ds: 007b es: 007b ss: 0068 Process swapper (pid: 1, threadinfo=c10bc000 task=c10b3aa0) Stack: c0100635 00000000 00000000 00000000 Call Trace: [<c0100635>] Code: 29 c0 01 00 00 00 6a 00 6a 02 68 97 0a 24 c0 e8 47 07 06 00 83 c4 0c 85 c0 78 60 6a 00 e8 a9 80 07 00 6a 00 e8 a2 80 07 00 58 5a <0f> 0b b4 02 13 0a 24 c0 a1 24 21 29 c0 85 c0 75 32 68 a4 0a 24 maybe there is another option which sneaked in, will investigate ... thanks, Herbert PS: was it ever considered to hard code the explicit BUG_* files and lines via a macro instead of that (probably very fancy and efficient but nevertheless) strange address stuff? I mean, are bug reports with missing/wrong file/line info really what the kernel developers want to see in the future? > /mjt - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
