Following is the latest version of my status report web page. It can be
found at:
http://members.cox.net/tmolina/kernprobs/status.html
I've seen a lot of positive feedback for Martin's proposal to create a
bugzilla for kernel bug reports so this is likely to be my last formal
posting on this subject. I intend to enter these as the "seed" bug
reports for his effort, so any comment on this is welcome.
2.5 Kernel Problem Reports as of 22 Oct
Status Discussion Problem Title
open 04 Oct 2002 AIC7XXX boot failure
1. http://marc.theaimsgroup.com/?l=linux-kernel&m=103356254615324&w=2
--------------------------------------------------------------------------
open 05 Oct 2002 oops in lock_get_status
2. http://marc.theaimsgroup.com/?l=linux-kernel&m=103244657605155&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 problems loading/unloading ide-scsi
modules
3. http://marc.theaimsgroup.com/?l=linux-kernel&m=103446296810822&w=2
--------------------------------------------------------------------------
open 08 Oct 2002 IDE problems on prePCI
4. http://marc.theaimsgroup.com/?l=linux-kernel&m=103277899317468&w=2
--------------------------------------------------------------------------
open 09 Oct 2002 USB Mass Storage problems
5. http://marc.theaimsgroup.com/?l=linux-kernel&m=103404393623200&w=2
--------------------------------------------------------------------------
open 18 Oct 2002 init_irq() function doing unsafe
things inside ide_lock
6. http://marc.theaimsgroup.com/?l=linux-kernel&m=103316967724891&w=2
--------------------------------------------------------------------------
open 04 Oct 2002 register_console() called in illegal
context
7. http://marc.theaimsgroup.com/?l=linux-kernel&m=103282695403237&w=2
--------------------------------------------------------------------------
open 09 Oct 2002 eata2x_detect() calls port_detect()
under driver_lock
8. http://marc.theaimsgroup.com/?l=linux-kernel&m=103281310122580&w=2
--------------------------------------------------------------------------
open 04 Oct 2002 sym_eh_handler does down(&ep->sem)
and might sleep
9. http://marc.theaimsgroup.com/?l=linux-kernel&m=103372067026942&w=2
--------------------------------------------------------------------------
open 07 Oct 2002 illegal sleeping function called
from acpi_os_wait_semaphore()
10. http://marc.theaimsgroup.com/?l=linux-kernel&m=103404677824885&w=2
--------------------------------------------------------------------------
open 09 Oct 2002 migration_thread atomicity error
11. http://marc.theaimsgroup.com/?l=linux-kernel&m=103408159014496&w=2
--------------------------------------------------------------------------
open 08 Oct 2002 snd_via8233 atomicity error
12. http://marc.theaimsgroup.com/?l=linux-kernel&m=103410375210315&w=2
--------------------------------------------------------------------------
open 19 Oct 2002 atomicity error in
sound/pci/via82xx.c
13. http://marc.theaimsgroup.com/?l=linux-kernel&m=103459664021147&w=2
--------------------------------------------------------------------------
open 11 Oct 2002 scheduling while atomic in
autofs4_root_lookup
14. http://marc.theaimsgroup.com/?l=linux-kernel&m=103426998326969&w=2
--------------------------------------------------------------------------
open 14 Oct 2002 atomicity error in
drivers/net/ppp_async.c
15. http://marc.theaimsgroup.com/?l=linux-kernel&m=103456920802806&w=2
--------------------------------------------------------------------------
open 14 Oct 2002 atomicity error in bond_enslave
16. http://marc.theaimsgroup.com/?l=linux-kernel&m=103462775624793&w=2
--------------------------------------------------------------------------
open 17 Oct 2002 swsusp atomicity error
17. http://marc.theaimsgroup.com/?l=linux-kernel&m=103489821623783&w=2
--------------------------------------------------------------------------
possible fix available 19 Oct 2002 atomicity error in snd_pcm/emufx.c
18. http://marc.theaimsgroup.com/?l=linux-kernel&m=103502805324053&w=2
--------------------------------------------------------------------------
open 03 Oct 2002 ACPI Mutex failure
19. http://marc.theaimsgroup.com/?l=linux-kernel&m=103369523011536&w=2
--------------------------------------------------------------------------
open 16 Oct 2002 initrd breakage
20. http://marc.theaimsgroup.com/?l=linux-kernel&m=103364305822611&w=2
--------------------------------------------------------------------------
open 05 Oct 2002 2.5.x and 8250 UART problems
21. http://marc.theaimsgroup.com/?l=linux-kernel&m=103383019409525&w=2
--------------------------------------------------------------------------
open 19 Oct 2002 mouse/keyboard freeze in X
22. http://marc.theaimsgroup.com/?l=linux-kernel&m=103441624616220&w=2
--------------------------------------------------------------------------
open 07 Oct 2002 bug related to virtual consoles
23. http://marc.theaimsgroup.com/?l=linux-kernel&m=103403138113853&w=2
--------------------------------------------------------------------------
open 07 Oct 2002 oops in kmem_cache_create
24. http://marc.theaimsgroup.com/?l=linux-kernel&m=103403423716317&w=2
--------------------------------------------------------------------------
open 07 Oct 2002 USB Hub failure
25. http://marc.theaimsgroup.com/?l=linux-kernel&m=103402696809279&w=2
--------------------------------------------------------------------------
open 08 Oct 2002 boot problem on 440GX
26. http://marc.theaimsgroup.com/?l=linux-kernel&m=103399796506960&w=2
--------------------------------------------------------------------------
open 08 Oct 2002 oops in run_timer_tasklet
27. http://marc.theaimsgroup.com/?l=linux-kernel&m=103393743102152&w=2
--------------------------------------------------------------------------
open 08 Oct 2002 oops while running kjournald
28. http://marc.theaimsgroup.com/?l=linux-kernel&m=103408191314814&w=2
--------------------------------------------------------------------------
open 09 Oct 2002 64GB highmem BUG()
29. http://marc.theaimsgroup.com/?l=linux-kernel&m=103399745406334&w=2
--------------------------------------------------------------------------
open 09 Oct 2002 Attempt to release TCP socket errors
30. http://marc.theaimsgroup.com/?l=linux-kernel&m=103409524231641&w=2
--------------------------------------------------------------------------
closed 09 Oct 2002 raid 0/1 problems in 2.5
31. http://marc.theaimsgroup.com/?l=linux-kernel&m=103414903003887&w=2
--------------------------------------------------------------------------
open 18 Oct 2002 raid5 hangs system
32. http://marc.theaimsgroup.com/?l=linux-kernel&m=103495428502729&w=2
--------------------------------------------------------------------------
open 06 Oct 2002 analog joystick oops
33. http://marc.theaimsgroup.com/?l=linux-kernel&m=103393598801189&w=2
--------------------------------------------------------------------------
open 07 Oct 2002 DRI not working
34. http://marc.theaimsgroup.com/?l=linux-kernel&m=103403348315804&w=2
--------------------------------------------------------------------------
open 10 Oct 2002 keyboard generates bogus key results
35. http://marc.theaimsgroup.com/?l=linux-kernel&m=103423327423623&w=2
--------------------------------------------------------------------------
closed 19 Oct 2002 no mouse wheel
36. http://marc.theaimsgroup.com/?l=linux-kernel&m=103351918416613&w=2
--------------------------------------------------------------------------
open 10 Oct 2002 PCMCIA trouble
37. http://marc.theaimsgroup.com/?l=linux-kernel&m=103420230730597&w=2
--------------------------------------------------------------------------
open 13 Oct 2002 apm hangs instead of suspending
38. http://marc.theaimsgroup.com/?l=linux-kernel&m=103454656623320&w=2
--------------------------------------------------------------------------
open 11 Oct 2002 tcp packets lost
39. http://marc.theaimsgroup.com/?l=linux-kernel&m=103429736523667&w=2
--------------------------------------------------------------------------
open 11 Oct 2002 shutdown problems in
driverfs_remove_file
40. http://marc.theaimsgroup.com/?l=linux-kernel&m=103443278524877&w=2
--------------------------------------------------------------------------
open 11 Oct 2002 broke ARM zImage/Image
41. http://marc.theaimsgroup.com/?l=linux-kernel&m=103442271819464&w=2
--------------------------------------------------------------------------
open 20 Oct 2002 loadlin boot failure
42. http://marc.theaimsgroup.com/?l=linux-kernel&m=103444415832048&w=2
--------------------------------------------------------------------------
open 13 Oct 2002 dual pointing device problem on
laptop
43. http://marc.theaimsgroup.com/?l=linux-kernel&m=103454188820088&w=2
--------------------------------------------------------------------------
open 14 Oct 2002 fbcon oops
44. http://marc.theaimsgroup.com/?l=linux-kernel&m=103458863514865&w=2
--------------------------------------------------------------------------
open 14 Oct 2002 ACPI/Suspend with an Acer Travelmate
350
45. http://marc.theaimsgroup.com/?l=linux-kernel&m=103463029127750&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 nfsd oops in auth_domain function
46. http://marc.theaimsgroup.com/?l=linux-kernel&m=103462971527133&w=2
--------------------------------------------------------------------------
open 15 Oct 2002 BUG in put_device during rmmod
47. http://marc.theaimsgroup.com/?l=linux-kernel&m=103470283114965&w=2
--------------------------------------------------------------------------
open 15 Oct 2002 BUG in kmem_cache_alloc_one_tail on
2.5.42
48. http://marc.theaimsgroup.com/?l=linux-kernel&m=103472220913410&w=2
--------------------------------------------------------------------------
open 15 Oct 2002 oops stopping serial
49. http://marc.theaimsgroup.com/?l=linux-kernel&m=103470900729987&w=2
--------------------------------------------------------------------------
open 15 Oct 2002 kernel hangs executing rpcinfo
50. http://marc.theaimsgroup.com/?l=linux-kernel&m=103462345019675&w=2
--------------------------------------------------------------------------
open 17 Oct 2002 reboot kills Dell Latitude keyboard
51. http://marc.theaimsgroup.com/?l=linux-kernel&m=103484425027884&w=2
--------------------------------------------------------------------------
open 19 Oct 2002 power down fails after 2.5.41
52. http://marc.theaimsgroup.com/?l=linux-kernel&m=103479527518536&w=2
--------------------------------------------------------------------------
open 16 Oct 2002 ACPI/Sb16 IRQ conflict
53. http://marc.theaimsgroup.com/?l=linux-kernel&m=103480163226174&w=2
--------------------------------------------------------------------------
open 17 Oct 2002 oops booting via ide controller
54. http://marc.theaimsgroup.com/?l=linux-kernel&m=103480082625264&w=2
--------------------------------------------------------------------------
open 17 Oct 2002 IDE not powered down on shutdown
55. http://marc.theaimsgroup.com/?l=linux-kernel&m=103476420012508&w=2
--------------------------------------------------------------------------
closed 20 Oct 2002 scsi/raid-related smp boot crash
56. http://marc.theaimsgroup.com/?l=linux-kernel&m=103485010600696&w=2
--------------------------------------------------------------------------
open 17 Oct 2002 nfs-related oops
57. http://marc.theaimsgroup.com/?l=linux-kernel&m=103477312121275&w=2
--------------------------------------------------------------------------
open 17 Oct 2002 neofb oops on shutdown
58. http://marc.theaimsgroup.com/?l=linux-kernel&m=103485950708944&w=2
--------------------------------------------------------------------------
open 17 Oct 2002 oops inserting xircom_cb network
card
59. http://marc.theaimsgroup.com/?l=linux-kernel&m=103474343128893&w=2
--------------------------------------------------------------------------
open 20 Oct 2002 usb-related boot hang
60. http://marc.theaimsgroup.com/?l=linux-kernel&m=103463093028435&w=2
--------------------------------------------------------------------------
open 18 Oct 2002 io-apic bug and spinlock deadlock
61. http://marc.theaimsgroup.com/?l=linux-kernel&m=103482589715521&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 buslogic scsi broke
62. http://marc.theaimsgroup.com/?l=linux-kernel&m=103496938421117&w=2
--------------------------------------------------------------------------
open 18 Oct 2002 color problem with atyfb
63. http://marc.theaimsgroup.com/?l=linux-kernel&m=103424151129857&w=2
--------------------------------------------------------------------------
open 18 Oct 2002 ipv4 /proc/net/route bug
64. http://marc.theaimsgroup.com/?l=linux-kernel&m=103497845730726&w=2
--------------------------------------------------------------------------
open 18 Oct 2002 crash with shared page table
65. http://marc.theaimsgroup.com/?l=linux-kernel&m=103499186007896&w=2
--------------------------------------------------------------------------
open 18 Oct 2002 qlogic 2x00 driver broke
66. http://marc.theaimsgroup.com/?l=linux-kernel&m=103470985631070&w=2
--------------------------------------------------------------------------
open 19 Oct 2002 tcq causes filesystem corruption
67. http://marc.theaimsgroup.com/?l=linux-kernel&m=103498823305987&w=2
--------------------------------------------------------------------------
open 19 Oct 2002 ncr adaptor doesn't see devices
68. http://marc.theaimsgroup.com/?l=linux-kernel&m=103506893016255&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 ide-cd module broke
69. http://marc.theaimsgroup.com/?l=linux-kernel&m=103508472223894&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 unable to eject zip disk
70. http://marc.theaimsgroup.com/?l=linux-kernel&m=103523397807029&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 isdn badly broken
71. http://marc.theaimsgroup.com/?l=linux-kernel&m=103513416515540&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 ide hangs on boot
72. http://marc.theaimsgroup.com/?l=linux-kernel&m=103515327029718&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 scsi hang on shutdown
73. http://marc.theaimsgroup.com/?l=linux-kernel&m=103504174230947&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 oops in ieee1394
74. http://marc.theaimsgroup.com/?l=linux-kernel&m=103519819428268&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 oops on boot in parport_pc module
75. http://marc.theaimsgroup.com/?l=linux-kernel&m=103524170815346&w=2
--------------------------------------------------------------------------
open 21 Oct 2002 ZONE_NORMAL exhaustion (dcache slab)
76. http://marc.theaimsgroup.com/?l=linux-kernel&m=103523368106684&w=2
--------------------------------------------------------------------------
open 22 Oct 2002 2.5.44 fs corruption
77. http://marc.theaimsgroup.com/?l=linux-kernel&m=103532467828806&w=2
--------------------------------------------------------------------------
open 22 Oct 2002 CS4236B stopping working as of
2.5.44
78. http://marc.theaimsgroup.com/?l=linux-kernel&m=103532492529636&w=2
--------------------------------------------------------------------------
open 22 Oct 2002 2.5.44-mm1 numa-q panic on boot
79. http://marc.theaimsgroup.com/?l=linux-kernel&m=103533122402278&w=2
--------------------------------------------------------------------------
open 22 Oct 2002 poisoned oops unmounting ramfs
80. http://marc.theaimsgroup.com/?l=linux-kernel&m=103530750609277&w=2
--------------------------------------------------------------------------
> From: Thomas Molina [mailto:[email protected]]
> open 07 Oct 2002 illegal sleeping
> function called
> from acpi_os_wait_semaphore()
> 10.
> http://marc.theaimsgroup.com/?l=linux-kernel&m=103404677824885&w=2
Fixed.
(the other 3 ACPI bugs are still open)
Regards -- Andy
> --------------------------------------------------------------------------
> open 21 Oct 2002 oops in ieee1394
> 74. http://marc.theaimsgroup.com/?l=linux-kernel&m=103519819428268&w=2
>
> --------------------------------------------------------------------------
Oh this is a real ass biter.
I have a workqueue that I setup once. Same data, same function, it never
changes. Every so often I call schedule_work() for the task.
Is queue_task() not reentrant? IOW, can I not schedule work that was
already scheduled similar to how tasklets worked?
Also, after the task has been run, does the workqueue struct's list
member not get cleared?
I'm a bit confused by this, as I expected behavior similar to before.
Ben
--
Debian - http://www.debian.org/
Linux 1394 - http://www.linux1394.org/
Subversion - http://subversion.tigris.org/
Deqo - http://www.deqo.com/
On Tue, Oct 22, 2002 at 10:28:52PM -0400, Ben Collins wrote:
> > --------------------------------------------------------------------------
> > open 21 Oct 2002 oops in ieee1394
> > 74. http://marc.theaimsgroup.com/?l=linux-kernel&m=103519819428268&w=2
> >
> > --------------------------------------------------------------------------
>
> Oh this is a real ass biter.
>
> I have a workqueue that I setup once. Same data, same function, it never
> changes. Every so often I call schedule_work() for the task.
Wait. I was looking at the wrong queue. This was a queue that I had list
linked myself, and then went through one at a time and attempted to
schedule it.
Already gone in my current code, so this will be closed the next time I
sync to Linus (which should be by the weekend).
--
Debian - http://www.debian.org/
Linux 1394 - http://www.linux1394.org/
Subversion - http://subversion.tigris.org/
Deqo - http://www.deqo.com/
On Tue, Oct 22, 2002 at 09:07:13PM -0500, Thomas Molina wrote:
> open 11 Oct 2002 broke ARM zImage/Image
> 41. http://marc.theaimsgroup.com/?l=linux-kernel&m=103442271819464&w=2
Fixed in 2.5.44, with thanks to Sam and Kai.
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
> Following is the latest version of my status report web page. It can be
> found at:
>
> http://members.cox.net/tmolina/kernprobs/status.html
>
> I've seen a lot of positive feedback for Martin's proposal to create a
> bugzilla for kernel bug reports so this is likely to be my last formal
> posting on this subject. I intend to enter these as the "seed" bug
> reports for his effort, so any comment on this is welcome.
> --------------------------------------------------------------------------
> open 17 Oct 2002 IDE not powered down on shutdown
> 55. http://marc.theaimsgroup.com/?l=linux-kernel&m=103476420012508&w=2
>
> --------------------------------------------------------------------------
This doesn't happen on every shutdown, but randomly on about 30% of
shutdowns. Only observed with 2.5.43. Has somebody changed the order
of the flush and spindown commands to the IDE devices?
I will try 2.5.44 on the machine later today, and report
success/failiure.
John.
On Tue, Oct 22, 2002 at 09:07:13PM -0500, Thomas Molina wrote:
> Following is the latest version of my status report web page. It can be
> found at:
>
> http://members.cox.net/tmolina/kernprobs/status.html
>
> I've seen a lot of positive feedback for Martin's proposal to create a
> bugzilla for kernel bug reports so this is likely to be my last formal
> posting on this subject. I intend to enter these as the "seed" bug
> reports for his effort, so any comment on this is welcome.
> --------------------------------------------------------------------------
> open 06 Oct 2002 analog joystick oops
> 33. http://marc.theaimsgroup.com/?l=linux-kernel&m=103393598801189&w=2
Closed, fixed by properly calling init_input_dev() in analog.c
> --------------------------------------------------------------------------
> open 10 Oct 2002 keyboard generates bogus key results
> 35. http://marc.theaimsgroup.com/?l=linux-kernel&m=103423327423623&w=2
Closed, PS/2 Active Multiplexing support was fixed.
--
Vojtech Pavlik
SuSE Labs
> 2.5 Kernel Problem Reports as of 22 Oct
> Status Discussion Problem Title
> --------------------------------------------------------------------------
> open 05 Oct 2002 2.5.x and 8250 UART problems
> 21. http://marc.theaimsgroup.com/?l=linux-kernel&m=103383019409525&w=2
>
> --------------------------------------------------------------------------
Tried 2.5.44 with preemption disabled, (it has been enabled for all
2.5.x up to now), and the problem is still there. Infact, it seems
worse - ZModem transfers are going down to 512 block size, because of
the errors :-(
> --------------------------------------------------------------------------
> open 17 Oct 2002 IDE not powered down on shutdown
> 55. http://marc.theaimsgroup.com/?l=linux-kernel&m=103476420012508&w=2
>
> --------------------------------------------------------------------------
Still happens with 2.5.44
John.
> 2.5 Kernel Problem Reports as of 22 Oct
> Status Discussion Problem Title
>
> --------------------------------------------------------------------------
> open 17 Oct 2002 IDE not powered down on shutdown
> 55. http://marc.theaimsgroup.com/?l=linux-kernel&m=103476420012508&w=2
>
> --------------------------------------------------------------------------
>
> --------------------------------------------------------------------------
> open 22 Oct 2002 2.5.44 fs corruption
> 77. http://marc.theaimsgroup.com/?l=linux-kernel&m=103532467828806&w=2
>
> --------------------------------------------------------------------------
Any possibility that the above two problems are related - I.E. disks
are not being flushed properly on shutdown?
John.
On Wed, 2002-10-23 at 13:26, [email protected] wrote:
> > 2.5 Kernel Problem Reports as of 22 Oct
> > Status Discussion Problem Title
> >
> > --------------------------------------------------------------------------
> > open 17 Oct 2002 IDE not powered down on shutdown
> > 55. http://marc.theaimsgroup.com/?l=linux-kernel&m=103476420012508&w=2
> >
> > --------------------------------------------------------------------------
> >
> > --------------------------------------------------------------------------
> > open 22 Oct 2002 2.5.44 fs corruption
> > 77. http://marc.theaimsgroup.com/?l=linux-kernel&m=103532467828806&w=2
> >
> > --------------------------------------------------------------------------
>
> Any possibility that the above two problems are related - I.E. disks
> are not being flushed properly on shutdown?
Possibly. I would be suprised however
On Wed, 23 Oct 2002 04:10:08 +0200, in linux.kernel you wrote:
>Following is the latest version of my status report web page. It can be
>found at:
>
>http://members.cox.net/tmolina/kernprobs/status.html
>
>I've seen a lot of positive feedback for Martin's proposal to create a
>bugzilla for kernel bug reports so this is likely to be my last formal
>posting on this subject. I intend to enter these as the "seed" bug
>reports for his effort, so any comment on this is welcome.
>
> 2.5 Kernel Problem Reports as of 22 Oct
> Status Discussion Problem Title
>
> open 19 Oct 2002 atomicity error in
> sound/pci/via82xx.c
> 13. http://marc.theaimsgroup.com/?l=linux-kernel&m=103459664021147&w=2
>
This problem is still present in 2.5.44-ac1.
Greetz Mu
> > > 2.5 Kernel Problem Reports as of 22 Oct
> > > Status Discussion Problem Title
> > >
> > > --------------------------------------------------------------------------
> > > open 17 Oct 2002 IDE not powered down on shutdown
> > > 55. http://marc.theaimsgroup.com/?l=linux-kernel&m=103476420012508&w=2
> > >
> > > --------------------------------------------------------------------------
> > >
> > > --------------------------------------------------------------------------
> > > open 22 Oct 2002 2.5.44 fs corruption
> > > 77. http://marc.theaimsgroup.com/?l=linux-kernel&m=103532467828806&w=2
> > >
> > > --------------------------------------------------------------------------
> >
> > Any possibility that the above two problems are related - I.E. disks
> > are not being flushed properly on shutdown?
>
> Possibly. I would be suprised however
Alan - have there been any changes to the flush/spindown code between
2.5.42 and 2.5.44? I remember a discussion about a month ago where
you said that it's necessary to do both, but that the order could be
wrong. I am seriously begining to suspect that something is
definitely wrong, because I can actually hear the disk spindown for a
fraction of a second, then spin up again, (at least with 2.5.43, so
far not with 2.5.44).
John.
On Wed, Oct 23, 2002 at 01:26:38PM +0100, [email protected] wrote:
> > 2.5 Kernel Problem Reports as of 22 Oct
> > Status Discussion Problem Title
> >
> > --------------------------------------------------------------------------
> > open 17 Oct 2002 IDE not powered down on shutdown
> > 55. http://marc.theaimsgroup.com/?l=linux-kernel&m=103476420012508&w=2
> >
> > --------------------------------------------------------------------------
> >
> > --------------------------------------------------------------------------
> > open 22 Oct 2002 2.5.44 fs corruption
> > 77. http://marc.theaimsgroup.com/?l=linux-kernel&m=103532467828806&w=2
> >
> > --------------------------------------------------------------------------
>
> Any possibility that the above two problems are related - I.E. disks
> are not being flushed properly on shutdown?
Nope. Mine didn't get to shutdown. Basically the whole system went
nuts - I couldn't run commands other than shell builtins, eth0
transmit timeouts, IDE complaining about attempts to access beyond the
end of device. I couldn't do a proper shutdown - hard to sudo
shutdown -r now when you can execute anything.
The first thing to show in the log was:
Oct 22 06:27:40 hagbard kernel: is_tree_node: node level 0 does not match to the expected one 1
Oct 22 06:27:40 hagbard kernel: vs-5150: search_by_key: invalid format found in block 2162976. Fsck?
Oct 22 06:27:40 hagbard kernel: is_tree_node: node level 0 does not match to the expected one 1
Oct 22 06:27:40 hagbard kernel: vs-5150: search_by_key: invalid format found in block 2162976. Fsck?
Oct 22 06:27:40 hagbard kernel: is_tree_node: node level 3748 does not match to the expected one 1
Oct 22 06:27:40 hagbard kernel: vs-5150: search_by_key: invalid format found in block 80570. Fsck?
Oct 22 06:27:40 hagbard kernel: is_tree_node: node level 4831 does not match to the expected one 1
Oct 22 06:27:40 hagbard kernel: vs-5150: search_by_key: invalid format found in block 80626. Fsck?
Oct 22 06:27:40 hagbard kernel: is_tree_node: node level 3760 does not match to the expected one 1
Oct 22 06:27:40 hagbard kernel: vs-5150: search_by_key: invalid format found in block 80655. Fsck?
Oct 22 06:27:40 hagbard kernel: vs-13070: reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1040516 1315261 0x0 SD]
Oct 22 06:27:40 hagbard kernel: is_tree_node: node level 269 does not match to the expected one 1
after 3 seconds of that:
Oct 22 06:27:43 hagbard kernel: attempt to access beyond end of device
Oct 22 06:27:43 hagbard kernel: ide0(3,2): rw=0, want=928907664, limit=6152895
Oct 22 06:27:43 hagbard kernel: attempt to access beyond end of device
Oct 22 06:27:43 hagbard kernel: ide0(3,2): rw=0, want=1208618096, limit=6152895
which went on for a few iterations, then:
Oct 22 06:28:33 hagbard kernel: NETDEV WATCHDOG: eth0: transmit timed out
Oct 22 06:28:33 hagbard kernel: eth0: Transmit timed out: status 0150 0c00 at 3896/3957 command 00001622.
Oct 22 06:29:23 hagbard kernel: NETDEV WATCHDOG: eth0: transmit timed out
Oct 22 06:29:23 hagbard kernel: eth0: Transmit timed out: status 0150 0c00 at 3958/4018 command 00020000.
Interestingly enough, reiserfs didn't complain on reboot. it was only
the root fs (on ext2/3) that complained - and had lots of problems.
Erik
On Wed, 2002-10-23 at 15:14, [email protected] wrote:
> definitely wrong, because I can actually hear the disk spindown for a
> fraction of a second, then spin up again, (at least with 2.5.43, so
> far not with 2.5.44).
Someone broke the power management code. When they fix it then I expect
the IDE powerdown stuff will behave better again. If you had scsi then
the scsi stuff may have been what broke it all.
> > definitely wrong, because I can actually hear the disk spindown for a
> > fraction of a second, then spin up again, (at least with 2.5.43, so
> > far not with 2.5.44).
>
> Someone broke the power management code. When they fix it then I expect
> the IDE powerdown stuff will behave better again. If you had scsi then
> the scsi stuff may have been what broke it all.
Ah, right, OK. On the subject of SCSI, recent 2.5.x kernels have
caused the SCSI bus activity LED on my Adaptec 2940AU to stay on after
powerdown as well, is this related?
John.
On Wed, 23 Oct 2002 [email protected] wrote:
> > > > 2.5 Kernel Problem Reports as of 22 Oct
> > > > Status Discussion Problem Title
> > > >
> > > > --------------------------------------------------------------------------
> > > > open 17 Oct 2002 IDE not powered down on shutdown
> > > > 55. http://marc.theaimsgroup.com/?l=linux-kernel&m=103476420012508&w=2
> > > >
> > > > --------------------------------------------------------------------------
> > > >
> > > > --------------------------------------------------------------------------
> > > > open 22 Oct 2002 2.5.44 fs corruption
> > > > 77. http://marc.theaimsgroup.com/?l=linux-kernel&m=103532467828806&w=2
> > > >
> > > > --------------------------------------------------------------------------
> > >
> > > Any possibility that the above two problems are related - I.E. disks
> > > are not being flushed properly on shutdown?
> >
> > Possibly. I would be suprised however
>
> Alan - have there been any changes to the flush/spindown code between
> 2.5.42 and 2.5.44? I remember a discussion about a month ago where
> you said that it's necessary to do both, but that the order could be
> wrong. I am seriously begining to suspect that something is
> definitely wrong, because I can actually hear the disk spindown for a
> fraction of a second, then spin up again, (at least with 2.5.43, so
> far not with 2.5.44).
It's my fault. At least the the problem about the disk not spinning down.
The driver core code was changed in 2.5.44 to call ->shutdown() instead of
->remove() during system power transitions, and none of the drivers got
converted over before 2.5.44 went out. I'm really sorry about this, and I
sincerely hope that it hasn't bitten any too bad..
Concerning the actual shutdown, I'm simply calling the ide driver's
->standby() method. At least in the case of ide disks, there is a call in
the driver's ->cleanup() method to flush the cache. Should this be
moved to ->standby()? Or, should we call ->flushcache() for all drives
from ->shutdown()?
Initial patch appended.
Thanks,
-pat
===== drivers/ide/ide.c 1.33 vs edited =====
--- 1.33/drivers/ide/ide.c Fri Oct 18 12:44:11 2002
+++ edited/drivers/ide/ide.c Wed Oct 23 09:42:27 2002
@@ -3351,6 +3351,14 @@
return 0;
}
+static void ide_drive_shutdown(struct device * dev)
+{
+ ide_drive_t * drive = container_of(dev,ide_drive_t,gendev);
+ ide_driver_t * drive = drive->driver;
+ if (driver && driver->standby)
+ driver->standby(drive);
+}
+
int ide_register_driver(ide_driver_t *driver)
{
struct list_head list;
@@ -3372,6 +3380,7 @@
driver->gen_driver.name = driver->name;
driver->gen_driver.bus = &ide_bus_type;
driver->gen_driver.remove = ide_drive_remove;
+ driver->gen_driver.shutdown = ide_drive_shutdown;
return driver_register(&driver->gen_driver);
}
On Wed, 2002-10-23 at 18:03, Patrick Mochel wrote:
> Concerning the actual shutdown, I'm simply calling the ide driver's
> ->standby() method. At least in the case of ide disks, there is a call in
> the driver's ->cleanup() method to flush the cache. Should this be
> moved to ->standby()? Or, should we call ->flushcache() for all drives
> from ->shutdown()?
I'll take a look. I need to finish porting over the 2.4 shutdown/eject
locking fixes that also touch this area
On Wed, Oct 23, 2002 at 10:03:51AM -0700, Patrick Mochel wrote:
>
> ===== drivers/ide/ide.c 1.33 vs edited =====
> --- 1.33/drivers/ide/ide.c Fri Oct 18 12:44:11 2002
> +++ edited/drivers/ide/ide.c Wed Oct 23 09:42:27 2002
> @@ -3351,6 +3351,14 @@
> return 0;
> }
>
> +static void ide_drive_shutdown(struct device * dev)
> +{
> + ide_drive_t * drive = container_of(dev,ide_drive_t,gendev);
> + ide_driver_t * drive = drive->driver;
Are you sure you didn't introduce a typo here? (missing 'r')
Maybe you meant this line to be:
+ ide_driver_t * driver = drive->driver;
> + if (driver && driver->standby)
> + driver->standby(drive);
> +}
> +
> int ide_register_driver(ide_driver_t *driver)
> {
> struct list_head list;
> @@ -3372,6 +3380,7 @@
> driver->gen_driver.name = driver->name;
> driver->gen_driver.bus = &ide_bus_type;
> driver->gen_driver.remove = ide_drive_remove;
> + driver->gen_driver.shutdown = ide_drive_shutdown;
> return driver_register(&driver->gen_driver);
> }
Regards,
Toon.
--
/"\ |
\ / ASCII RIBBON CAMPAIGN | "Who is this General Failure, and
X AGAINST HTML MAIL | what is he doing on my harddisk?"
/ \
On Sat, 26 Oct 2002, Toon van der Pas wrote:
> On Wed, Oct 23, 2002 at 10:03:51AM -0700, Patrick Mochel wrote:
> >
> > ===== drivers/ide/ide.c 1.33 vs edited =====
> > --- 1.33/drivers/ide/ide.c Fri Oct 18 12:44:11 2002
> > +++ edited/drivers/ide/ide.c Wed Oct 23 09:42:27 2002
> > @@ -3351,6 +3351,14 @@
> > return 0;
> > }
> >
> > +static void ide_drive_shutdown(struct device * dev)
> > +{
> > + ide_drive_t * drive = container_of(dev,ide_drive_t,gendev);
> > + ide_driver_t * drive = drive->driver;
>
> Are you sure you didn't introduce a typo here? (missing 'r')
> Maybe you meant this line to be:
>
> + ide_driver_t * driver = drive->driver;
Uhm..yes. Sorry about that.
-pat
>> 2.5 Kernel Problem Reports as of 22 Oct
>> Status Discussion Problem Title
>>
>
>
>> open 19 Oct 2002 atomicity error in
>> sound/pci/via82xx.c
>> 13. http://marc.theaimsgroup.com/?l=linux-kernel&m=103459664021147&w=2
>>
>
>
>This problem is still present in 2.5.44-ac1.
>Greetz Mu
>
>
And still present in 2.5.45.
But ....after a Linus-is-back-and-loose merge armageddon,my kernel
still compiles and runs fine. Not bad after that huge merging frenzy.
I guess cleanups like this will be worked at after the feature freeze.
Greetz Mu