NetBSD Problem Report #9403

Received: (qmail 16624 invoked from network); 12 Feb 2000 20:40:45 -0000
Message-Id: <200002122040.OAA19781@fb.sa.enteract.com>
Date: Sat, 12 Feb 2000 14:40:31 -0600 (CST)
From: Frederick Bruckman <fb@enteract.com>
Reply-To: fb@enteract.com
To: gnats-bugs@gnats.netbsd.org
Subject: Build of debugging kernel fails at "strip -d".
X-Send-Pr-Version: 3.95

>Number:         9403
>Category:       port-m68k
>Synopsis:       Build of debugging kernel fails at "strip -d".
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    port-m68k-maintainer
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sat Feb 12 12:42:00 +0000 2000
>Closed-Date:    Sun Feb 13 12:30:09 +0000 2000
>Last-Modified:  Sun Feb 13 12:31:00 +0000 2000
>Originator:     Frederick Bruckman
>Release:        NetBSD-1.4.2_ALPHA-2000-02-11
>Organization:

>Environment:

System: NetBSD corwin.home 1.4.2_ALPHA
NetBSD 1.4.2_ALPHA (CORWIN) #11:
Mon Feb 7 12:16:59 CST 2000
fredb@corwin.home:/usr/src/sys/arch/mac68k/compile/CORWIN mac68k


>Description:

	Attempting to build a kernel with debug symbols fails, as follows.

ld -n -Ttext 0 -e start -X -o netbsd ${SYSTEM_OBJ} vers.o
text	data	bss	dec	hex	filename
1146804	68044  	223852 	1438700	15f3ec 	netbsd
cp netbsd netbsd.gdb
strip -d netbsd

strip: Can not allocate 1146804 bytes after allocating 31515800 bytes
*** Error code 1

Stop.
>How-To-Repeat:

	Simply configure and make a kernel with "makeoptions DEBUG=-g",
	as suggested in the GENERIC kernel config.
>Fix:

	Invoking ld with "-S" rather than "-X" works better. Doing so via
	"make netbsd DEBUG=" almost works, but it has the side-effect of
	incrementing the version number. So the work around is:
	1) "make build"
	2) "rm netbsd"; "rm st[1-9]*" # clean up after "strip".
	3) "vi version" # decrement version number
	4) "make netbsd DEBUG="

	Whatever's wrong with strip, it would make sense, perhaps, to edit 
	/sys/arch/${arch}/conf/Makefile.${arch} to make the two kernels in
	two link steps (but without incrementing the version number).


>Release-Note:
>Audit-Trail:

From: Dave Huang <khym@bga.com>
To: Frederick Bruckman <fb@enteract.com>
Cc: gnats-bugs@gnats.netbsd.org
Subject: Re: port-m68k/9403: Build of debugging kernel fails at "strip -d".
Date: Sat, 12 Feb 2000 15:18:41 -0600 (CST)

 On Sat, 12 Feb 2000, Frederick Bruckman wrote:
 > 	Whatever's wrong with strip, it would make sense, perhaps, to edit 
 > 	/sys/arch/${arch}/conf/Makefile.${arch} to make the two kernels in
 > 	two link steps (but without incrementing the version number).

 In -current, it does a strip -g -o netbsd netbsd.gdb, which is much more
 efficient than the two-step "cp netbsd.gdb netbsd; strip -d netbsd". I
 don't know why 1.4.2_ALPHA's broken though; it worked fine in 1.4.1...


From: Frederick Bruckman <fb@enteract.com>
To: Dave Huang <khym@bga.com>
Cc: gnats-bugs@gnats.netbsd.org
Subject: Re: port-m68k/9403: Build of debugging kernel fails at "strip -d".
Date: Sun, 13 Feb 2000 06:24:45 -0600 (CST)

 On Sat, 12 Feb 2000, Dave Huang wrote:

 > On Sat, 12 Feb 2000, Frederick Bruckman wrote:
 > > 	Whatever's wrong with strip, it would make sense, perhaps, to edit 
 > > 	/sys/arch/${arch}/conf/Makefile.${arch} to make the two kernels in
 > > 	two link steps (but without incrementing the version number).
 > 
 > In -current, it does a strip -g -o netbsd netbsd.gdb, which is much more
 > efficient than the two-step "cp netbsd.gdb netbsd; strip -d netbsd". I
 > don't know why 1.4.2_ALPHA's broken though; it worked fine in 1.4.1...

 Ack! All I had to do was increase the data seg limit in my shell. I'll
 close the PR. Thanks for the response.


State-Changed-From-To: open->closed 
State-Changed-By: fredb 
State-Changed-When: Sun Feb 13 06:30:09 CST 2000 
State-Changed-Why:  
filed in error 

From: Todd Whitesel <toddpw@best.com>
To: fb@enteract.com
Cc: gnats-bugs@gnats.netbsd.org
Subject: Re: port-m68k/9403: Build of debugging kernel fails at "strip -d".
Date: Fri, 18 Feb 2000 22:19:46 -0800 (PST)

 > strip: Can not allocate 1146804 bytes after allocating 31515800 bytes
 ...
 > 	Invoking ld with "-S" rather than "-X" works better. Doing so via

 No offense, but this smells like a "limit datasize" problem, which BTW I
 recently ran into on next68k -- the sources I checked out had debugging
 enabled already, but I didn't realize that until after strip gave an error.

 Todd Whitesel
 toddpw @ best.com
>Unformatted:

Home	PR Database Search