-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi all,
ok, next 2.6-WOLK release. Apply ontop of a vanilla 2.6.4 from kernel.org.
Many thanks to Rik van Ballegooijen and all the other people sending fixes!
... Have fun and a nice weekend :)
BIG NOTE:
- ---------
We now have 5 addons in place, named: grsecurity2, Win4Lin, gobohide, SysTrace
and finally the Staircase Scheduler by Con. I won't merge any of these
patches into mainline 2.6-WOLK in the near future. You have to decide what
addon feature you like to have :) ... Another reason is, I won't break users
of 4g/4g, 4k stacks nor uml-skas3, currently a no-go with grsecurity. Apply
an addon/the addons manually. All five patches are located in the patch itself
in the ./ADDON-patches directory after applying the main patch. Apply as
usual.
wolk2.2 was never released (ok, it was available for some hours but I removed
it because it was broken. Therefore, changelog included too)
KNOWN ISSUES:
- -----------------
- - Loop-AES won't work with per-backing dev unpluggin code in WOLK.
But fortunately loop-AES coders updated their code. So get the newest
version before trying loop-AES with WOLK (includes also -mm tree)
Changelog from v2.6.4-wolk2.2 -> v2.6.4-wolk2.3
- -----------------------------------------------
merged with 2.6.5-rc2 (Linus Torvalds)
merged with 2.6.5-rc2-mm3 (Andrew Morton)
- w/o 4k-stacks-always-on.patch (NVIDIA folks)
- w/o bk-acpi stuff from 2.6-bk tree (broken)
- w/ cpu detection for 2.6.5-rc1-mm2 (Manfred Spraul)
- w/ "2.6.5-rc2-mm* machine won't boot" fix (Andrew Morton)
o added: objrmap for file mappings (2.6.5-rc2-aa4) (Andrea Arcangeli)
o added: anon_vma for memory unmapping (2.6.5-rc2-aa4) (Andrea Arcangeli)
o added: prio trees (2.6.5-rc2-aa4) (Andrea Arcangeli)
o added: morse code panic (Tomas Szepe)
o added: VFS: Soft-/Hard Limit of file descriptors ()
o added: Maximum amount of unix sockets via sysctl ()
o added: Renice processes as a user in a special GID (me)
o added: rICMP support (Nail)
o added: improved fdmap support (Matt Miller)
o added: ADDON: systrace v1.5 (Niels Provos)
o added: ADDON: staircase scheduler (Con Colivas)
(WOLK port by Rik van Ballegooijen)
o fixed: fb colour used to clear top and bottom margins (Jakub Bogusz)
o fixed: fb background used for "clear" terminal commands (Jakub Bogusz)
o fixed: ReiserFS v3 buffer refcount problem (Chris Mason)
o fixed: ReiserFS v3 buffer head leak in invalidatepage (Chris Mason)
o fixed: ReiserFS v3 bogus warning message about locked (Chris Mason)
buffers during logging.
o fixed: when CONFIG_IP_NMAP_FREAK was not set, the kernel (anonymous)
did not reply to any icmp echo requests :(
o updated: O(1) Batch scheduler for 2.6 (Rik van Ballegooijen)
o updated: ReiserFS v3 extended attributes (Chris Mason)
o updated: ReiserFS v3 POSIX Access Control Lists (Chris Mason)
o updated: ReiserFS v3 Security Labels (Chris Mason)
o updated: ADDON: grsecurity2 (Brad Spengler)
o updated: netdev-random: Drivers v2.6.4 (Michal Purzynski)
o changed: max loop devices to 32 (from 8) (me)
o changed: max msg queue identifiers to 512 (from 16) (me)
o changed: max semaphore identifiers to 1024 (from 128) (me)
o removed: stupid/nonsense warning in Supermount (Rik van Ballegooijen)
Changelog from v2.6.4-wolk2.1 -> v2.6.4-wolk2.2
- -----------------------------------------------
merged with 2.6.5-rc1-mm2 (Andrew Morton)
- w/o 4k-stacks-always-on.patch
- w/ "anti 2.6.5-rc1-mm2-is-dog-slow-patch" ;) (Jens Axboe)
o added: O(1) Batch scheduler for 2.6 (Con Kolivas)
(WOLK port by Rik van Ballegooijen)
o added: O(1) ISO scheduler for 2.6 (Con Kolivas)
o fixed: classic OOM killer (the default) compile error (Robert F?hricht)
o fixed: compile warning in rc-update.h with gcc 3.4 (Robert F?hricht)
o fixed: compile problems with gcc 2.95.x (me)
o fixed: bootsplash breakage (Rik van Ballegooijen)
o fixed: shfs compilation error (Rik van Ballegooijen)
o fixed: PRAMfs compilation error (Rik van Ballegooijen)
o fixed: SCSI oops() in -mm tree (Kai Makisara)
o fixed: compile problems on ppc64 (Paul Mackerras)
o removed: some SATA fixes (did not really fix anything)
Todo
- ----
o /proc restrictions w/o the need for grsecurity2 addon
o madwifi
o Linux Trustees
o rsbac as an addon
o menu cleanups
o DRBD once it's ported to 2.6
o Bind Bount Extensions 0.05
o vservers for 2.6 once Herbert comes up with a patch
o probably changing O(1) Scheduler addons to a config option
o _____ <add more things if you want>
md5sums:
- --------
02d50db9872e44f686dcc21361d019d3 linux-2.6.4-wolk2.3.patch.bz2
6fe224fd75e3c15b776b87a810dec2d4 linux-2.6.4-wolk2.3.patch.gz
- --
Kind regards
Marc-Christian Petersen
http://sourceforge.net/projects/wolk
PGP/GnuPG Key: 1024D/569DE2E3DB441A16
Fingerprint: 3469 0CF8 CA7E 0042 7824 080A 569D E2E3 DB44 1A16
Key available at http://pgp.mit.edu. Encrypted e-mail preferred
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: !! No Risk - No Fun !! - Try to crack this ;-)
iD8DBQFAZIqPVp3i49tEGhYRAiTxAJ9FcVnDVEOjsf7NcEgamoTy3PzcbgCdGYPW
s22a478UkeLo4ki/g9Uvaug=
=hsR2
-----END PGP SIGNATURE-----
Hi Marc,
I took the patch for a spin and found a few glitches.
A patch is attached to fix them.
1. The kernel oops'ed in vt_ioctl due to driver_data
being NULL in the tty. I copied the relevant fix from
2.6.5-rc2-mm2.
2. I did not configure the in-kernel statd daemon for
NFS. That caused a BUG() in mon.c:nsm_monitor()
because the handle was never initialized.
3. A typo fix in ipmi/af_ipmi.c.
Thanks,
Shantanu
__________________________________
Do you Yahoo!?
Yahoo! Finance Tax Center - File online. File on time.
http://taxes.yahoo.com/filing.html
Just a thought...
If it is supposed to be _working_ overload linux kernel, so
maybe Marc should drop the -mm patches as they contain testing
code, which is more prone to bugs than vanilla kernels.
Besides that, WOLK is great with all the stuff you might need in
there, thanks.
Regards,
Maciej
Hi,
When running modem_run on 2.6.4-WOLK2.3 it locks in D state on one of USB
ioctls. It works at least on 2.6.2-rc2. I have no idea what causes this
bug so I sent it to all lists.
Please help if you can.
Grzegorz Kulewski
On Sat, 27 Mar 2004, Grzegorz Kulewski wrote:
> Hi,
>
> When running modem_run on 2.6.4-WOLK2.3 it locks in D state on one of USB
> ioctls. It works at least on 2.6.2-rc2. I have no idea what causes this
> bug so I sent it to all lists.
>
> Please help if you can.
>
> Grzegorz Kulewski
Try applying this patch:
http://marc.theaimsgroup.com/?l=linux-usb-devel&m=108016447231291&q=raw
Alan Stern
On Sunday 28 March 2004 00:51, Alan Stern wrote:
Hi Grzegorz,
> > When running modem_run on 2.6.4-WOLK2.3 it locks in D state on one of USB
> > ioctls. It works at least on 2.6.2-rc2. I have no idea what causes this
> > bug so I sent it to all lists.
> > Please help if you can.
> > Grzegorz Kulewski
> Try applying this patch:
> http://marc.theaimsgroup.com/?l=linux-usb-devel&m=108016447231291&q=raw
Did this help Grzegorz?
ciao, Marc
On Saturday 27 March 2004 06:21, Shantanu Goel wrote:
Hi Shantanu,
> 1. The kernel oops'ed in vt_ioctl due to driver_data
> being NULL in the tty. I copied the relevant fix from
> 2.6.5-rc2-mm2.
> 2. I did not configure the in-kernel statd daemon for
> NFS. That caused a BUG() in mon.c:nsm_monitor()
> because the handle was never initialized.
> 3. A typo fix in ipmi/af_ipmi.c.
thank you :) ... Applied.
ciao, Marc
On Wed, 31 Mar 2004, Marc-Christian Petersen wrote:
> On Sunday 28 March 2004 00:51, Alan Stern wrote:
>
> Hi Grzegorz,
>
> > > When running modem_run on 2.6.4-WOLK2.3 it locks in D state on one of USB
> > > ioctls. It works at least on 2.6.2-rc2. I have no idea what causes this
> > > bug so I sent it to all lists.
> > > Please help if you can.
> > > Grzegorz Kulewski
>
> > Try applying this patch:
> > http://marc.theaimsgroup.com/?l=linux-usb-devel&m=108016447231291&q=raw
>
> Did this help Grzegorz?
I will probably compile new kernel before friday. I had some urgent tasks
to solve in normal differetial equations on my university :).
But this seems the right solution/workaround for the problem and probably
should work (of course I do not know if there are other bugs preventing
modem_run from continuing :)).
I will post the anwser as soon as possible.
thanks once more
Grzegorz
On Sat, Mar 27, 2004 at 06:51:36PM -0500, Alan Stern wrote:
> On Sat, 27 Mar 2004, Grzegorz Kulewski wrote:
>
> > Hi,
> >
> > When running modem_run on 2.6.4-WOLK2.3 it locks in D state on one of USB
> > ioctls. It works at least on 2.6.2-rc2. I have no idea what causes this
> > bug so I sent it to all lists.
> >
> > Please help if you can.
> >
> > Grzegorz Kulewski
>
> Try applying this patch:
>
> http://marc.theaimsgroup.com/?l=linux-usb-devel&m=108016447231291&q=raw
>
> Alan Stern
>
This patch seems to be included in 2.6.5-rc2 bk-curent but it don't
solve the same problem with xsane (or scanimage) which hangs in
ioctl(n, USBDEVFS_SETCONFIGURATION) when accessing my Epson USB Sc
anner.Not the first time I run it but each time after that.
The process remains in D state for many hours before returning for no
evident reason.Was working fine in 2.6.3.
On Wed, 31 Mar 2004, Leopold Gouverneur wrote:
> On Sat, Mar 27, 2004 at 06:51:36PM -0500, Alan Stern wrote:
> >
> > Try applying this patch:
> >
> > http://marc.theaimsgroup.com/?l=linux-usb-devel&m=108016447231291&q=raw
> >
> > Alan Stern
> >
> This patch seems to be included in 2.6.5-rc2 bk-curent but it don't
> solve the same problem with xsane (or scanimage) which hangs in
> ioctl(n, USBDEVFS_SETCONFIGURATION) when accessing my Epson USB Sc
> anner.Not the first time I run it but each time after that.
> The process remains in D state for many hours before returning for no
> evident reason.Was working fine in 2.6.3.
The short answer is that programs (like xsane or scanimage) generally
shouldn't be using the SETCONFIGURATION call. The program should be
fixed.
The longer answer is that several patches have been submitted in the last
week that touch on this and other related problems. They may appear in
2.6.5-rc4 or later.
Alan Stern
On Wed, 31 Mar 2004, Marc-Christian Petersen wrote:
> On Sunday 28 March 2004 00:51, Alan Stern wrote:
>
> Hi Grzegorz,
>
> > > When running modem_run on 2.6.4-WOLK2.3 it locks in D state on one of USB
> > > ioctls. It works at least on 2.6.2-rc2. I have no idea what causes this
> > > bug so I sent it to all lists.
> > > Please help if you can.
> > > Grzegorz Kulewski
>
> > Try applying this patch:
> > http://marc.theaimsgroup.com/?l=linux-usb-devel&m=108016447231291&q=raw
>
> Did this help Grzegorz?
No, it did not. This time I tried 2.6.5-rc3-mm4. Seems the patch was
already applied in this release. modem_run did not lock (disk sleep state)
in the same place, but it locked after writing synchronization succeded
to system log. It seems that there is no kernel-stack-for-process-in-proc
option in mm (Andrew, can you add it?), so there is no callstack.
but still thanks for your help
and looking for next :)
Grzegorz Kulewski
On Fri, 2 Apr 2004, Grzegorz Kulewski wrote:
> On Wed, 31 Mar 2004, Marc-Christian Petersen wrote:
>
> > On Sunday 28 March 2004 00:51, Alan Stern wrote:
> >
> > Hi Grzegorz,
> >
> > > > When running modem_run on 2.6.4-WOLK2.3 it locks in D state on one of USB
> > > > ioctls. It works at least on 2.6.2-rc2. I have no idea what causes this
> > > > bug so I sent it to all lists.
> > > > Please help if you can.
> > > > Grzegorz Kulewski
> >
> > > Try applying this patch:
> > > http://marc.theaimsgroup.com/?l=linux-usb-devel&m=108016447231291&q=raw
> >
> > Did this help Grzegorz?
>
> No, it did not. This time I tried 2.6.5-rc3-mm4. Seems the patch was
> already applied in this release. modem_run did not lock (disk sleep state)
> in the same place, but it locked after writing synchronization succeded
> to system log. It seems that there is no kernel-stack-for-process-in-proc
> option in mm (Andrew, can you add it?), so there is no callstack.
Everything seems to work stable on vanilla 2.6.5 + "few debuging and
statistic patches that do not touch anything very important", so probably
on true vanilla 2.6.5 too. So only -mm and wolk are broken?
What do you think about it?
Grzegorz Kulewski
Hi,
I experienced this oops. I have uhci-hcd and two devices. One is usb
camera (TC111 - probably not supported under linux?) and the
second is speedtouch modem. Everytime I shut down my system (Gentoo) with
2.6.5 and newer I get some oops but system log is down before that and I
have no time to hack start scripts to stop shuting syslog. It occures when
removing some usb modules. So I stopped speedtouch and removed the modules
manually (in stop scripts order I hope). But I have not removed uhci-hcd
module (this module is removed in other part of stop scripts). And...
nothing happened. So I unplugged speedtouch and replugged it back. And I
immendiatelly got atached oops. (I think that I should use ksymoops, but
it is searching for /proc/ksyms that is not present in 2.6 and it does not
like /proc/kallsyms... And it produces nothing but warnings. What options
should I use?)
What can I do to help track the problem down?
thanks in advance
Grzegorz Kulewski
PS. I am subscribbed only to LKML, so CC me, please.
Is anybody going to look at it? It prevents my system from shuting down.
On Tue, 27 Apr 2004, Grzegorz Kulewski wrote:
> Hi,
>
> I experienced this oops. I have uhci-hcd and two devices. One is usb
> camera (TC111 - probably not supported under linux?) and the
> second is speedtouch modem. Everytime I shut down my system (Gentoo) with
> 2.6.5 and newer I get some oops but system log is down before that and I
> have no time to hack start scripts to stop shuting syslog. It occures when
> removing some usb modules. So I stopped speedtouch and removed the modules
> manually (in stop scripts order I hope). But I have not removed uhci-hcd
> module (this module is removed in other part of stop scripts). And...
> nothing happened. So I unplugged speedtouch and replugged it back. And I
> immendiatelly got atached oops. (I think that I should use ksymoops, but
> it is searching for /proc/ksyms that is not present in 2.6 and it does not
> like /proc/kallsyms... And it produces nothing but warnings. What options
> should I use?)
>
> What can I do to help track the problem down?
>
>
> thanks in advance
>
> Grzegorz Kulewski
>
>
> PS. I am subscribbed only to LKML, so CC me, please.
Grzegorz Kulewski wrote:
> Hi,
>
> I experienced this oops. I have uhci-hcd and two devices. One is usb
> camera (TC111 - probably not supported under linux?) and the
> second is speedtouch modem. Everytime I shut down my system (Gentoo) with
> 2.6.5 and newer I get some oops but system log is down before that and I
> have no time to hack start scripts to stop shuting syslog. It occures when
> removing some usb modules. So I stopped speedtouch and removed the modules
> manually (in stop scripts order I hope). But I have not removed uhci-hcd
> module (this module is removed in other part of stop scripts). And...
> nothing happened. So I unplugged speedtouch and replugged it back. And I
> immendiatelly got atached oops. (I think that I should use ksymoops, but
> it is searching for /proc/ksyms that is not present in 2.6 and it does not
> like /proc/kallsyms... And it produces nothing but warnings. What options
> should I use?)
>
> What can I do to help track the problem down?
Does it happen without preempt?
--
-bill davidsen ([email protected])
"The secret to procrastination is to put things off until the
last possible moment - but no longer" -me
On Wed, 28 Apr 2004, Bill Davidsen wrote:
> Grzegorz Kulewski wrote:
> > Hi,
> >
> > I experienced this oops. I have uhci-hcd and two devices. One is usb
> > camera (TC111 - probably not supported under linux?) and the
> > second is speedtouch modem. Everytime I shut down my system (Gentoo) with
> > 2.6.5 and newer I get some oops but system log is down before that and I
> > have no time to hack start scripts to stop shuting syslog. It occures when
> > removing some usb modules. So I stopped speedtouch and removed the modules
> > manually (in stop scripts order I hope). But I have not removed uhci-hcd
> > module (this module is removed in other part of stop scripts). And...
> > nothing happened. So I unplugged speedtouch and replugged it back. And I
> > immendiatelly got atached oops. (I think that I should use ksymoops, but
> > it is searching for /proc/ksyms that is not present in 2.6 and it does not
> > like /proc/kallsyms... And it produces nothing but warnings. What options
> > should I use?)
> >
> > What can I do to help track the problem down?
>
> Does it happen without preempt?
Yes, it does. Here is the log:
Apr 28 16:10:50 polb01 usb 1-1: USB disconnect, address 2
Apr 28 16:10:50 polb01 Unable to handle kernel paging request at virtual
address ddc03f64
Apr 28 16:10:50 polb01 printing eip:
Apr 28 16:10:50 polb01 e0943d73
Apr 28 16:10:50 polb01 *pde = 00077067
Apr 28 16:10:50 polb01 *pte = 1dc03000
Apr 28 16:10:50 polb01 Oops: 0002 [#1]
Apr 28 16:10:50 polb01 DEBUG_PAGEALLOC
Apr 28 16:10:50 polb01 CPU: 0
Apr 28 16:10:50 polb01 EIP: 0060:[<e0943d73>] Not tainted
Apr 28 16:10:50 polb01 EFLAGS: 00010282 (2.6.6-rc2-bk3)
Apr 28 16:10:50 polb01 EIP is at driver_disconnect+0x13/0x40 [usbcore]
Apr 28 16:10:50 polb01 eax: db346ef8 ebx: db346ef8 ecx: db346f08
edx: ddc03f38
Apr 28 16:10:50 polb01 esi: e0954be0 edi: db346f08 ebp: dabdfe7c
esp: dabdfe74
Apr 28 16:10:50 polb01 ds: 007b es: 007b ss: 0068
Apr 28 16:10:50 polb01 Process khubd (pid: 5338, threadinfo=dabde000
task=dee39a10)
Apr 28 16:10:50 polb01 Stack: e0954be0 db346f08 dabdfe98 e09380f6 db346ef8
db346ef8 db346f08 db346f08
Apr 28 16:10:50 polb01 e0954c00 dabdfeb0 c030e0c6 db346f08 db346f30
e09544e0 e095452c dabdfecc
Apr 28 16:10:50 polb01 c030e23d db346f08 00000042 db346f08 c03f3488
dbebfcdc dabdfee8 c030d0be
Apr 28 16:10:50 polb01 Call Trace:
Apr 28 16:10:50 polb01 [<e09380f6>] usb_unbind_interface+0x76/0x80
[usbcore]
Apr 28 16:10:50 polb01 [<c030e0c6>] device_release_driver+0x66/0x70
Apr 28 16:10:50 polb01 [<c030e23d>] bus_remove_device+0x6d/0xb0
Apr 28 16:10:50 polb01 [<c030d0be>] device_del+0x6e/0xb0
Apr 28 16:10:50 polb01 [<c030d114>] device_unregister+0x14/0x20
Apr 28 16:10:50 polb01 [<e09401d1>] usb_disable_device+0x71/0xb0 [usbcore]
Apr 28 16:10:50 polb01 [<e0938d26>] usb_disconnect+0xc6/0x120 [usbcore]
Apr 28 16:10:50 polb01 [<e093b77f>] hub_port_connect_change+0x28f/0x2a0
[usbcore]
Apr 28 16:10:50 polb01 [<e093b141>] hub_port_status+0x41/0xb0 [usbcore]
Apr 28 16:10:50 polb01 [<e093baee>] hub_events+0x35e/0x4d0 [usbcore]
Apr 28 16:10:50 polb01 [<e093bc95>] hub_thread+0x35/0x100 [usbcore]
Apr 28 16:10:50 polb01 [<c0118cd0>] default_wake_function+0x0/0x20
Apr 28 16:10:50 polb01 [<e093bc60>] hub_thread+0x0/0x100 [usbcore]
Apr 28 16:10:50 polb01 [<c01022b1>] kernel_thread_helper+0x5/0x14
Apr 28 16:10:50 polb01
Apr 28 16:10:50 polb01 Code: c7 42 2c 00 00 00 00 8b 40 04 0f b6 40 02 0f
b3 82 88 00 00
In my first post there was some warning that spinlock was not initialized.
Isn't this the source of problem?
Grzegorz Kulewski
Grzegorz Kulewski wrote:
> Is anybody going to look at it? It prevents my system from shuting down.
Looks like one of the problems already fixed in the latest bk-usb.patch
found in the MM kernels ...
- Dave
> On Tue, 27 Apr 2004, Grzegorz Kulewski wrote:
>
>
>>Hi,
>>
>>I experienced this oops. I have uhci-hcd and two devices. One is usb
>>camera (TC111 - probably not supported under linux?) and the
>>second is speedtouch modem. Everytime I shut down my system (Gentoo) with
>>2.6.5 and newer I get some oops but system log is down before that and I
>>have no time to hack start scripts to stop shuting syslog. It occures when
>>removing some usb modules. So I stopped speedtouch and removed the modules
>>manually (in stop scripts order I hope). But I have not removed uhci-hcd
>>module (this module is removed in other part of stop scripts). And...
>>nothing happened. So I unplugged speedtouch and replugged it back. And I
>>immendiatelly got atached oops. (I think that I should use ksymoops, but
>>it is searching for /proc/ksyms that is not present in 2.6 and it does not
>>like /proc/kallsyms... And it produces nothing but warnings. What options
>>should I use?)
>>
>>What can I do to help track the problem down?
>>
>>
>>thanks in advance
>>
>>Grzegorz Kulewski
>>
>>
>>PS. I am subscribbed only to LKML, so CC me, please.
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: Oracle 10g
> Get certified on the hottest thing ever to hit the market... Oracle 10g.
> Take an Oracle 10g class now, and we'll give you the exam FREE.
> http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click
> _______________________________________________
> [email protected]
> To unsubscribe, use the last form field at:
> https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
>
On Wednesday 28 April 2004 13:27, Grzegorz Kulewski wrote:
> Is anybody going to look at it? It prevents my system from shuting down.
Sorry about my earlier email - the patches I was thinking about haven't
found their way into Linus's tree yet. They are in the following bitkeeper
tree: bk://kernel.bkbits.net/gregkh/linux/usb-2.6
All the best,
Duncan.
On Wednesday 28 April 2004 13:27, Grzegorz Kulewski wrote:
> Is anybody going to look at it? It prevents my system from shuting down.
If you are lucky I already fixed it. Did you try the latest 2.6 kernel?
Ciao,
Duncan.