<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.5730.11" name=GENERATOR></HEAD>
<BODY>
<DIV><SPAN class=967300522-01122006><FONT face=Arial color=#0000ff size=2>The
warning message here is a little strange.</FONT></SPAN></DIV>
<DIV><SPAN class=967300522-01122006></SPAN><FONT face=Arial color=#0000ff
size=2><SPAN class=967300522-01122006>The symbol matches as being in the
'bss', but appears to be</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=967300522-01122006>being
located outside the bss segment.</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN
class=967300522-01122006></SPAN></FONT> </DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=967300522-01122006>Once I
think I've identified a segment, I try to skip over the </SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=967300522-01122006>rest
of it. There's an assumption (and I probably should have
</SPAN></FONT></DIV>
<DIV><FONT><SPAN class=967300522-01122006></SPAN></FONT><FONT face=Tahoma><SPAN
class=967300522-01122006><FONT face=Arial color=#0000ff size=2>written an
assertion) that symbols we scan over starting at
mod_symtable</FONT></SPAN></FONT></DIV>
<DIV><FONT face=Tahoma><SPAN class=967300522-01122006><FONT face=Arial
color=#0000ff size=2>are monotonically increasing. Is that not
true? </FONT></SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN
class=967300522-01122006></SPAN></FONT> </DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=967300522-01122006>The
clipping on the longs is annoying. I tend to rely on gcc
catching</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=967300522-01122006>those
for me. </SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN
class=967300522-01122006></SPAN></FONT> </DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=967300522-01122006>Thanks
for the testing!</SPAN></FONT></DIV>
<DIV><FONT face=Tahoma><FONT size=2><SPAN class=967300522-01122006><FONT
face=Arial color=#0000ff></FONT></SPAN></FONT></FONT> </DIV>
<DIV><FONT face=Tahoma><FONT size=2>-----Original Message-----<BR><B>From:</B>
crash-utility-bounces@redhat.com [mailto:crash-utility-bounces@redhat.com]<B>On
Behalf Of </B>Dave Anderson<BR><B>Sent:</B> Friday, December 01, 2006 11:46
AM<BR><B>To:</B> Discussion list for crash utility usage, maintenance and
development<BR><B>Subject:</B> Re: [Crash-utility] modules and data / bss
initialization<BR><BR></DIV></FONT></FONT>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">Castor Fu wrote:
<BLOCKQUOTE TYPE="CITE"> <SPAN class=923310002-01122006><FONT
face=Arial><FONT color=#0000ff><FONT size=-1>I don't think this made it out
earlier...</FONT></FONT></FONT></SPAN><SPAN
class=923310002-01122006></SPAN><SPAN class=923310002-01122006><FONT
face=Arial><FONT color=#0000ff><FONT size=-1>Here's a fix. I've also
added something so 'MODULES_IN_CWD' will work on
2.6</FONT></FONT></FONT></SPAN><SPAN class=923310002-01122006><FONT
face=Arial><FONT color=#0000ff><FONT size=-1>since modules will end with
.ko</FONT></FONT></FONT></SPAN><SPAN class=923310002-01122006></SPAN><SPAN
class=923310002-01122006><FONT face=Arial><FONT color=#0000ff><FONT
size=-1>I hope this looks good to
others....</FONT></FONT></FONT></SPAN></BLOCKQUOTE>
<P><BR><TT>Hi Castor,</TT><TT></TT>
<P><TT>Upon quick testing with RHEL4 and RHEL5 x86_64 kernels,</TT>
<BR><TT>this patch certainly looks promising...</TT><TT></TT>
<P><TT>Although I don't particularly care to see these
messages:</TT><TT></TT>
<P><TT>ffffffff8810ae80
serio_raw 41157
/lib/modules/2.6.18-1.2747.el5/kernel/drivers/input/serio/serio_raw.ko</TT>
<BR><TT>ffffffff8811b580
uhci_hcd 59353
/lib/modules/2.6.18-1.2747.el5/kernel/drivers/usb/host/uhci-hcd.ko</TT>
<BR><TT>ffffffff88130b00
shpchp
73069
/lib/modules/2.6.18-1.2747.el5/kernel/drivers/pci/hotplug/shpchp.ko</TT>
<BR><B><TT>unexpected sym __key.10825 8814a180 sec .bss offset e180 mod_base
8813c000</TT></B> <BR><B><TT>XXX sym __key.10825 @ 8814a180 bfd val 0
section .bss</TT></B> <BR><B><TT>unexpected sym __key.10826 8814a180 sec .bss
offset e180 mod_base 8813c000</TT></B> <BR><B><TT>XXX sym __key.10826 @
8814a180 bfd val 0 section .bss</TT></B> <BR><TT>ffffffff88141f80
i2c_core 57793
/lib/modules/2.6.18-1.2747.el5/kernel/drivers/i2c/i2c-core.ko</TT><TT></TT>
<P><TT>I would think they could be CRASHDEBUG(1)'d, couldn't they?</TT>
<BR><TT>Plus, those error messages will clip 64-bit values as shown</TT>
<BR><TT>above.</TT><TT></TT>
<P><TT>I'm also presuming that the new add-symbol-file operation will</TT>
<BR><TT>harmlessly take a "0" mod_data_start, mod_rodata_start or</TT>
<BR><TT>mod_bss_start address argument; seemingly it does, since several</TT>
<BR><TT>of my test modules have 0 as one or more of those start
addresses.</TT><TT></TT>
<P><TT>Anyway, I also would be interested in the experiences of others</TT>
<BR><TT>on the list who are using different architectures and kernel</TT>
<BR><TT>versions.</TT><TT></TT>
<P><TT>Thanks,</TT> <BR><TT> Dave</TT> <BR><TT></TT> <BR>
<BR> <BR> <BR> <BR> </P></BLOCKQUOTE></BODY></HTML>