Re: kgdb and .ko files

From: Julian Elischer <julian_at_freebsd.org>
Date: Mon, 08 Nov 2010 18:08:43 -0800
relying to self with more info..

On 11/8/10 4:34 PM, Julian Elischer wrote:
> my usual command set for doing .ko debugging issomething like:
> %kgdb /sys/amd64/compile/DEBUG/kernel.symbols
> (kgdb) target remote pu_serial2:7005
> (kgdb) sharedlibrary driver.ko
> (kgdb) directory /home/me/branches/blah/obj/
>
> (kgdb) info sharedlibrary
>> From                To                  Syms Read   Shared Object 
>> Library
> 0xffffffff81222000  0xffffffff8129dac0  Yes         
> home/me/branches/blah/freebsd8-amd64/output/driver.ko
>
> but recently the last line has started returning:
>
>> From                To                  Syms Read   Shared Object 
>> Library
>                                         Yes         
> home/me/branches/blah/freebsd8-amd64/output/driver.ko
>
> Now I can guess that the problem might be something to do with 
> readinf symbols as our
> method of generating the .ko changed a while back but when I look at 
> it I do see:
>
> awk -f /sys/conf/kmod_syms.awk driver.ko.debug  export_syms | xargs 
> -J% objcopy %driver.ko.debug
> objcopy --only-keep-debug driver.ko.debug driver.ko.symbols
> objcopy --strip-debug --add-gnu-debuglink=driver.ko.symbols 
> driver.ko.debug driver.ko
>
> so theoretically the plain driver.ko should result in (k)gdb looking 
> up the symbol file driver.ko.symbols
> which should have all the symbol information needed for debugging..
> or am I misreading this?
>
> doesn't seem to work but it does seem to improve if I link the .ko 
> file to the symbols file..

it seems changing the first objcopy line to give the full path name of 
the symbol file seems to help

>
> this is in 8.1 rather than -current but I can't test -current.
>
> _______________________________________________
> freebsd-current_at_freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to 
> "freebsd-current-unsubscribe_at_freebsd.org"
>
Received on Tue Nov 09 2010 - 01:08:42 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:09 UTC