Aw: Re: Prelinking on ARM with Debug Link

["Torsten Polle" <Torsten dot Polle at gmx dot de>]

Hi Mark,

> Gesendet: Mittwoch, 13. April 2016 um 11:25 Uhr
> Von: "Mark Wielaard" <mjw@redhat.com>
> An: "Torsten Polle" <Torsten.Polle@gmx.de>
> Cc: systemtap@sourceware.org
> Betreff: Re: Prelinking on ARM with Debug Link
> Hi Torsten,
> 
> On Tue, 2016-04-12 at 22:26 +0200, Torsten Polle wrote:
> > > Am 11.04.2016 um 23:01 schrieb Mark Wielaard <mjw@redhat.com>:
> > >
> > > On Mon, Apr 11, 2016 at 08:47:00PM +0200, Torsten Polle wrote:
> > >> Iâve checked your patch. As a result the backtrace calculations
> > >> break in my environment.
> > >
> > > I assume this is an in-kernel backtrace, does it involve kernel
> > modules?
> > > Or is it a user backtrace, executable only? shared libraries?
> > > Could you show a probe script and example backtrace?
> >
> > in principle, Iâm talking about the example in [1]. In short, Iâm
> > talking about getting a proper backtrace in the prelinked library
> > libc-2.18.so. Iâm saying âin principleâ because the example does not
> > set a real probe, but only makes SystemTap include the necessary
> > unwinding information for libc-2.18.so.
> >
> > The real (actually already stripped down) script can be found in [2].
> >
> > In the script taptrek_run_izA4.stp in line 240, Iâm using the
> > following statement
> > uaddr = __ustack_raw(depth);
> > to get backtrace information. Please donât get distracted by the usage
> > of the function __ustack_raw(). The function print_ubacktrace() would
> > provide a similar result.
> 
> I am not sure why the example needs to be so complicated.
> The usage of mixed user/kernel space probes/backtraces makes it really
> hard to understand what is going on. Can't you show what you get with a
> simple probe inside libc.so with a print_ubacktrace?

This is what I had in mind. ;-) I use the following probe. Please find the result below.

probe process("/lib/libc-2.18.so").function("_int_malloc").call
{
    print_ubacktrace();
}

> > > Please do include some verbose output. If you can provide the output
> > > of stap -DDEBUG_UNWIND=1 that might be helpful.

_stp_stack_unwind_one_user:462: STARTING user unwind
 0x4df2b1b8 : _int_malloc+0x0/0x15f0 [/lib/libc-2.18.so]
_stp_stack_unwind_one_user:478: CONTINUING user unwind to depth 1
unwind:1481: pc=4df2b1b8, 4df2b1b8
unwind:1524: trying debug_frame
_stp_search_unwind_hdr:778: binary search for 4df2b1b8
_stp_search_unwind_hdr:843: fde off=7294
_stp_search_unwind_hdr:853: returning fde=7f4275b8 startLoc=4df2b1b8
unwind_frame:1197: /lib/libc-2.18.so: fde=7f4275b8
unwind_frame:1202: /lib/libc-2.18.so: cie=7f427354
parse_fde_cie:156: map retAddrReg value 14 to reg_info idx 14
unwind_frame:1217: startLoc: 4df2b1b8, endLoc: 4df2c7a8
unwind_frame:1266: cie=7f427354 fde=7f4275b8 startLoc=4df2b1b8 endLoc=4df2c7a8, pc=4df2b1b8
unwind_frame:1286: processCFI for CIE
processCFI:313: targetLoc=0 state->loc=4df2b1b8
processCFI:519: map DW_CFA_def_cfa value 13 to reg_info idx 13
processCFI:521: DW_CFA_def_cfa reg=13
processCFI:530: DW_CFA_def_cfa_offset offs=0
processCFI:649: targetLoc=0 state->loc=4df2b1b8
processCFI:650: result: 1
unwind_frame:1294: processCFI for FDE
processCFI:313: targetLoc=4df2b1b8 state->loc=4df2b1b8
advance_loc:241: state->loc=4df2b1bc
processCFI:617: DW_CFA_advance_loc
processCFI:649: targetLoc=4df2b1b8 state->loc=4df2b1bc
processCFI:650: result: 1
unwind_frame:1313: cfa reg=13, off=0
unwind_frame:1318: cfa=7ed94438 startLoc=7ed94438, endLoc=7ed94438
unwind_frame:1325: cie=7f427354 fde=7f4275b8
unwind_frame:1449: returning 0 (4df2de08)
_stp_stack_unwind_one_user:503: ret=0 PC=4df2de08 SP=7ed94438
 0x4df2de08 : __libc_malloc+0x74/0x204 [/lib/libc-2.18.so]
_stp_stack_unwind_one_user:478: CONTINUING user unwind to depth 2
unwind:1481: pc=4df2de07, 4df2de08
unwind:1524: trying debug_frame
_stp_search_unwind_hdr:778: binary search for 4df2de07
_stp_search_unwind_hdr:843: fde off=73e4
_stp_search_unwind_hdr:853: returning fde=7f427708 startLoc=4df2dd94
unwind_frame:1197: /lib/libc-2.18.so: fde=7f427708
unwind_frame:1202: /lib/libc-2.18.so: cie=7f427354
parse_fde_cie:156: map retAddrReg value 14 to reg_info idx 14
unwind_frame:1217: startLoc: 4df2dd94, endLoc: 4df2df98
unwind_frame:1266: cie=7f427354 fde=7f427708 startLoc=4df2dd94 endLoc=4df2df98, pc=4df2de07
unwind_frame:1286: processCFI for CIE
processCFI:313: targetLoc=0 state->loc=4df2dd94
processCFI:519: map DW_CFA_def_cfa value 13 to reg_info idx 13
processCFI:521: DW_CFA_def_cfa reg=13
processCFI:530: DW_CFA_def_cfa_offset offs=0
processCFI:649: targetLoc=0 state->loc=4df2dd94
processCFI:650: result: 1
unwind_frame:1294: processCFI for FDE
processCFI:313: targetLoc=4df2de07 state->loc=4df2dd94
advance_loc:241: state->loc=4df2dd98
processCFI:617: DW_CFA_advance_loc
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
processCFI:530: DW_CFA_def_cfa_offset offs=18
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
processCFI:627: map DW_CFA_offset value 3 to reg_info idx 3
set_offset_rule:259: reg=3, where=2, svalue=ffffffe8
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
processCFI:627: map DW_CFA_offset value 4 to reg_info idx 4
set_offset_rule:259: reg=4, where=2, svalue=ffffffec
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
processCFI:627: map DW_CFA_offset value 5 to reg_info idx 5
set_offset_rule:259: reg=5, where=2, svalue=fffffff0
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
processCFI:627: map DW_CFA_offset value 6 to reg_info idx 6
set_offset_rule:259: reg=6, where=2, svalue=fffffff4
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
processCFI:627: map DW_CFA_offset value 7 to reg_info idx 7
set_offset_rule:259: reg=7, where=2, svalue=fffffff8
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
processCFI:627: map DW_CFA_offset value 14 to reg_info idx 14
set_offset_rule:259: reg=e, where=2, svalue=fffffffc
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
processCFI:322: DW_CFA_nop
processCFI:649: targetLoc=4df2de07 state->loc=4df2dd98
processCFI:650: result: 1
unwind_frame:1313: cfa reg=13, off=18
unwind_frame:1318: cfa=7ed94450 startLoc=7ed94438, endLoc=7ed94450
unwind_frame:1325: cie=7f427354 fde=7f427708
unwind_frame:1439: set register 3 to 7ed94488
unwind_frame:1439: set register 4 to 47
unwind_frame:1439: set register 5 to 7ed94474
unwind_frame:1439: set register 6 to 2f580
unwind_frame:1439: set register 7 to 8
unwind_frame:1439: set register 14 to 23a30
unwind_frame:1449: returning 0 (23a30)
_stp_stack_unwind_one_user:503: ret=0 PC=23a30 SP=7ed94450
 0x23a30 [/lib/systemd/systemd-udevd+0x1ba30/0x35000]
_stp_stack_unwind_one_user:478: CONTINUING user unwind to depth 3
unwind:1481: pc=23a2f, 23a30
unwind:1520: No module found for pc=23a2f
_stp_stack_unwind_one_user:503: ret=-22 PC=23a30 SP=7ed94450


> > Iâll try my best and come back to you. I would provide the
> > information with my patch applied, because otherwise the result is
> > simply garbage.
> 
> Please do include the exact patch you are using.

Please find the patch attached.

> Thanks, that is interesting.
> But it is also wrong that we use section offsets in user space modules.
> The sections really shouldn't play a role, except for kernel modules.
> User space modules are mapped into memory according to their segments
> (phdrs). Could you show the segments plus section to segment mapping of
> the libc-2.18.so and libc-2.18.so.debug files used with eu-readelf -Sl

Please check [1] for the requested output.

This means sec_load_offset should be 0 as you proposed in your
patch. But then the question is where you want to adjust for the
necessary difference between offsets.

> Thanks,
> 
> Mark

Thanks,
Torsten

[1] https://sourceware.org/ml/systemtap/2015-q4/msg00219.html

Attachment: 0001-PATCH-v2-Fix-Compilation-fails-for-prelinked-librari.patch
Description: Binary data