[uml-devel] link failure, multiple definition of 'in6addr_any'

[uml-devel] link failure, multiple definition of 'in6addr_any'

[Antoine Martin]

  CC      init/version.o
  LD      init/built-in.o
  LD      .tmp_vmlinux1
arch/um/drivers/built-in.o: In function `pcap_nametoaddrinfo':
(.text+0x1143f): warning: Using 'getaddrinfo' in statically linked 
applications requires at runtime the shared libraries from the glibc 
version used for linking
arch/um/drivers/built-in.o: In function `pcap_nametoaddr':
(.text+0x11465): warning: Using 'gethostbyname' in statically linked 
applications requires at runtime the shared libraries from the glibc 
version used for linking
arch/um/drivers/built-in.o: In function `pcap_nametonetaddr':
(.text+0x113e5): warning: Using 'getnetbyname' in statically linked 
applications requires at runtime the shared libraries from the glibc 
version used for linking
arch/um/drivers/built-in.o: In function `pcap_nametoproto':
(.text+0x111f5): warning: Using 'getprotobyname' in statically linked 
applications requires at runtime the shared libraries from the glibc 
version used for linking
arch/um/drivers/built-in.o: In function `pcap_nametoport':
(.text+0x11239): warning: Using 'getservbyname' in statically linked 
applications requires at runtime the shared libraries from the glibc 
version used for linking
/usr/lib/gcc/x86_64-pc-linux-gnu/4.1.2/../../../../lib64/libc.a(in6_addr.o):(.rodata+0x0): 
multiple definition of `in6addr_any'
net/built-in.o:(.rodata+0x5380): first defined here
collect2: ld returned 1 exit status

This is from a known "good" config upgraded to 2.6.22 with oldconfig.
I removed IPv6 support and all is well.

Antoine

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] link failure, multiple definition of 'in6addr_any'

[Jeff Dike]

On Fri, Jul 13, 2007 at 12:05:28PM +0100, Antoine Martin wrote:
> /usr/lib/gcc/x86_64-pc-linux-gnu/4.1.2/../../../../lib64/libc.a(in6_addr.o):(.rodata+0x0): 
> multiple definition of `in6addr_any'
> net/built-in.o:(.rodata+0x5380): first defined here
> collect2: ld returned 1 exit status

I hate to do this, but can you check the patch below?

I really need to find a way to separate the kernel and libc namespaces.

				Jeff

-- 
Work email - jdike at linux dot intel dot com

Index: linux-2.6.21-mm/arch/um/Makefile
===================================================================
--- linux-2.6.21-mm.orig/arch/um/Makefile	2007-06-28 16:11:29.000000000 -0400
+++ linux-2.6.21-mm/arch/um/Makefile	2007-07-13 11:59:21.000000000 -0400
@@ -60,7 +60,8 @@ SYS_DIR		:= $(ARCH_DIR)/include/sysdep-$
 
 CFLAGS += $(CFLAGS-y) -D__arch_um__ -DSUBARCH=\"$(SUBARCH)\"	\
 	$(ARCH_INCLUDE) $(MODE_INCLUDE) -Dvmap=kernel_vmap	\
-	-Din6addr_loopback=kernel_in6addr_loopback
+	-Din6addr_loopback=kernel_in6addr_loopback \
+	-Din6addr_any=kernel_in6addr_any
 
 AFLAGS += $(ARCH_INCLUDE)
 

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] link failure, multiple definition of 'in6addr_any'

[Antoine Martin]

Jeff Dike wrote:
> On Fri, Jul 13, 2007 at 12:05:28PM +0100, Antoine Martin wrote:
>   
>> /usr/lib/gcc/x86_64-pc-linux-gnu/4.1.2/../../../../lib64/libc.a(in6_addr.o):(.rodata+0x0): 
>> multiple definition of `in6addr_any'
>> net/built-in.o:(.rodata+0x5380): first defined here
>> collect2: ld returned 1 exit status
>>     
>
> I hate to do this, but can you check the patch below?
>   
Works for me (tm)

btw, if you're in bug squashing mode, I've got one which is 100% 
reproducible - looks like a resource leak to me: simply run a post 
2.6.20 guest kernel on a 2.6.21.5 32-bit host (IIRC with or without 
skas3) and do an "emerge --sync" (gentoo fs) on the guest. It will hang 
half way through, you can interrupt the task with ^Z, but nothing else 
will run after that.

Antoine
> I really need to find a way to separate the kernel and libc namespaces.
>
> 				Jeff
>
>   


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] link failure, multiple definition of 'in6addr_any'

[Jeff Dike]

On Fri, Jul 13, 2007 at 07:09:15PM +0100, Antoine Martin wrote:
> Works for me (tm)

OK.

> btw, if you're in bug squashing mode, 

I always am (or if not, someone needs to give me a good kick in the rump).

> I've got one which is 100% 
> reproducible - looks like a resource leak to me: simply run a post 
> 2.6.20 guest kernel on a 2.6.21.5 32-bit host (IIRC with or without 
> skas3) and do an "emerge --sync" (gentoo fs) on the guest. It will hang 
> half way through, you can interrupt the task with ^Z, but nothing else 
> will run after that.

OK, I'll take a look.

				Jeff

-- 
Work email - jdike at linux dot intel dot com

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Antoine Martin]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Jeff Dike wrote:
> On Fri, Jul 13, 2007 at 07:09:15PM +0100, Antoine Martin wrote:
>> Works for me (tm)
> 
> OK.
> 
>> btw, if you're in bug squashing mode, 
> 
> I always am (or if not, someone needs to give me a good kick in the rump).
> 
>> I've got one which is 100% 
>> reproducible - looks like a resource leak to me: simply run a post 
>> 2.6.20 guest kernel on a 2.6.21.5 32-bit host (IIRC with or without 
>> skas3) and do an "emerge --sync" (gentoo fs) on the guest. It will hang 
>> half way through, you can interrupt the task with ^Z, but nothing else 
>> will run after that.
> 
> OK, I'll take a look.
If it doesn't crash on emerge --sync (when I reboot it doesn't always
crash), try building something big (like glibc or gcc: emerge glibc gcc)
It does look like an fd leak or something: the connection stays alive,
processes already started can still receive keyboard input, but no way
to login or run any commands.

Antoine
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.5 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGr3V/GK2zHPGK1rsRCnHjAJ9MCfCo2IIE1jroW008jhOZGizn8gCdGT/T
TlwijBMcskIJz0E+jAT01l4=
=X9u4
-----END PGP SIGNATURE-----

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Jeff Dike]

On Tue, Jul 31, 2007 at 06:46:39PM +0100, Antoine Martin wrote:
> If it doesn't crash on emerge --sync (when I reboot it doesn't always
> crash), try building something big (like glibc or gcc: emerge glibc gcc)
> It does look like an fd leak or something: the connection stays alive,
> processes already started can still receive keyboard input, but no way
> to login or run any commands.

OK, so I get this:

	Welcome to viper.nagafix.co.uk

	@ERROR: access denied to gentoo-portage from pool-72-70-38-233.bstnma.fios.verizon.net (72.70.38.233)

What are the public servers that I can sync against, and where do I put them?

				Jeff

-- 
Work email - jdike at linux dot intel dot com

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Jeff Dike]

On Tue, Jul 31, 2007 at 06:46:39PM +0100, Antoine Martin wrote:
> If it doesn't crash on emerge --sync

Nevermind that last, I figured it out.

>  (when I reboot it doesn't always crash), 

Didn't you say this was 100%? :-)  Anyway, it works fine here.

> try building something big (like glibc or gcc: emerge glibc gcc)

That's running as I type, apparently happily.

				Jeff

-- 
Work email - jdike at linux dot intel dot com

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Jeff Dike]

On Tue, Jul 31, 2007 at 06:46:39PM +0100, Antoine Martin wrote:
> If it doesn't crash on emerge --sync (when I reboot it doesn't always
> crash), try building something big (like glibc or gcc: emerge glibc gcc)

Well, I did that, and it finished fine.

Does this happen with a current UML?  Do you have any stack traces or
core dumps for me to look at?

				Jeff

-- 
Work email - jdike at linux dot intel dot com

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Antoine Martin]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Jeff Dike wrote:
> On Tue, Jul 31, 2007 at 06:46:39PM +0100, Antoine Martin wrote:
>> If it doesn't crash on emerge --sync (when I reboot it doesn't always
>> crash), try building something big (like glibc or gcc: emerge glibc gcc)
> 
> Well, I did that, and it finished fine.
Damn, so maybe it is interacting with apache. This guest has got apache
running and it crashes every few days (usually when doing an emerge sync
but not always).


> Does this happen with a current UML?  Do you have any stack traces or
> core dumps for me to look at?
No core dumps but I'll try getting some stack traces using SysRq next
time it crashes.

Antoine
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.5 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGuJgKGK2zHPGK1rsRCibpAJ9NLCwY8o192pgebZ04QqUUzYQ8dQCfZvcZ
jlaytfGYM9N4uy3R6QDkRnc=
=XC5D
-----END PGP SIGNATURE-----

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Antoine Martin]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Jeff Dike wrote:
> On Tue, Jul 31, 2007 at 06:46:39PM +0100, Antoine Martin wrote:
>> If it doesn't crash on emerge --sync (when I reboot it doesn't always
>> crash), try building something big (like glibc or gcc: emerge glibc gcc)
>> It does look like an fd leak or something: the connection stays alive,
>> processes already started can still receive keyboard input, but no way
>> to login or run any commands.
> 
> OK, so I get this:
> 
> 	Welcome to viper.nagafix.co.uk
> 
> 	@ERROR: access denied to gentoo-portage from pool-72-70-38-233.bstnma.fios.verizon.net (72.70.38.233)
> 
> What are the public servers that I can sync against, and where do I put them?
Ooops, my update script seems to have been interrupted in the past and
failed to remove the "SYNC=XXX" line it places in /etc/make.conf
Script now fixed and the images have been updated too.

Antoine

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.5 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGuwXSGK2zHPGK1rsRCpWHAJ986mjSod2gLW11+jI9iYgoQI6gqACeMm8A
B+9MBfr2My/xrpFswM2ZuCE=
=ETyK
-----END PGP SIGNATURE-----

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Antoine Martin]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Antoine Martin wrote:
> Jeff Dike wrote:
>> On Tue, Jul 31, 2007 at 06:46:39PM +0100, Antoine Martin wrote:
>>> If it doesn't crash on emerge --sync (when I reboot it doesn't
>>> always crash), try building something big (like glibc or gcc:
>>> emerge glibc gcc)
>> Well, I did that, and it finished fine.
> Damn, so maybe it is interacting with apache. This guest has got
> apache running and it crashes every few days (usually when doing an
> emerge sync but not always).
> 
> 
>> Does this happen with a current UML?  Do you have any stack traces
>> or core dumps for me to look at?
> No core dumps but I'll try getting some stack traces using SysRq next
>  time it crashes.

SysRq doesn't dump anything anymore (or did it ever? IIRC it did)
This crash is 100% reproducible in my systems. Just emerging stuff
triggers it.
Can you try with one of the guests on my site (say 2.6.20.x?)
When it dies, the existing shells are still there - but are unable to
spawn any new processes.

Antoine
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.5 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGwxSJGK2zHPGK1rsRCvcNAJ4wlrfDnxuQVhvw89YXtk9SZSfewQCaA6Xu
bMix9qFnAfAk5lsmoTkxiJw=
=j4Zv
-----END PGP SIGNATURE-----

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Antoine Martin]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Antoine Martin wrote:
> Antoine Martin wrote:
>> Jeff Dike wrote:
>>> On Tue, Jul 31, 2007 at 06:46:39PM +0100, Antoine Martin wrote:
>>>> If it doesn't crash on emerge --sync (when I reboot it doesn't
>>>> always crash), try building something big (like glibc or gcc:
>>>> emerge glibc gcc)
>>> Well, I did that, and it finished fine.
>> Damn, so maybe it is interacting with apache. This guest has got
>> apache running and it crashes every few days (usually when doing an
>> emerge sync but not always).
> 
> 
>>> Does this happen with a current UML?  Do you have any stack traces
>>> or core dumps for me to look at?
>> No core dumps but I'll try getting some stack traces using SysRq next
>>  time it crashes.
> 
> SysRq doesn't dump anything anymore (or did it ever? IIRC it did)
> This crash is 100% reproducible in my systems. Just emerging stuff
> triggers it.
Even better than emerging stuff, I found that simply doing a:
emerge --sync
Will crash during the sync or when it rebuilds the cache.

Antoine
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.5 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGxGXkGK2zHPGK1rsRCl4+AJ9M+RBcrHoGk21+9kEkZDt302Rd2QCeLHAR
/4QhuzJQPxf208rzd4Cczqo=
=jsDz
-----END PGP SIGNATURE-----

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] 2.6.21.5 crashes [was link failure, multiple definition of 'in6addr_any']

[Jeff Dike]

On Wed, Aug 15, 2007 at 03:58:17PM +0100, Antoine Martin wrote:
> SysRq doesn't dump anything anymore (or did it ever? IIRC it did)

I was just looking at that.  You need CONFIG_MAGIC_SYSRQ enabled,
which isn't by default for some reason.  Then, there is a bug in
returning stuff to the mconsole client, which I have a patch for.
Then, there's a problem with large amounts of data being passed back.

However, if sysrq is enabled, data will end up in the kernel log as
usual, where it can be looked at with gdb even if the UML isn't
healthy enough to run dmesg.

				Jeff

-- 
Work email - jdike at linux dot intel dot com

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel