I have uploaded an updated version of the radix-tree pagecache patch
against 2.4.19-pre2-ac2. News in this release:
* fix a deadlock when vmtruncate takes i_shared_lock twice by introducing
a new mapping->page_lock that mutexes mapping->page_tree. (akpm)
* move setting of page->flags back out of move_to/from_swap_cache. (akpm)
* put back lost page state settings in shmem_unuse_inode. (akpm)
* get rid of remove_page_from_inode_queue - there was only one caller. (me)
* replace add_page_to_inode_queue with ___add_to_page_cache. (me)
Please give it some serious beating while I try to get 2.5 working and
port the patch over 8)
Location:
ftp://ftp.kernel.org/pub/linux/kernel/people/hch/patches/v2.4/2.4.19-pre2-ac2/linux-2.4.19-radixtree.patch.gz
ftp://ftp.kernel.org/pub/linux/kernel/people/hch/patches/v2.4/2.4.19-pre2-ac2/linux-2.4.19-radixtree.patch.bz2
On March 3, 2002 03:03 pm, Christoph Hellwig wrote:
> I have uploaded an updated version of the radix-tree pagecache patch
> against 2.4.19-pre2-ac2. News in this release:
>
> * fix a deadlock when vmtruncate takes i_shared_lock twice by introducing
> a new mapping->page_lock that mutexes mapping->page_tree. (akpm)
> * move setting of page->flags back out of move_to/from_swap_cache. (akpm)
> * put back lost page state settings in shmem_unuse_inode. (akpm)
> * get rid of remove_page_from_inode_queue - there was only one caller. (me)
> * replace add_page_to_inode_queue with ___add_to_page_cache. (me)
>
> Please give it some serious beating while I try to get 2.5 working and
> port the patch over 8)
Got this after a couple of hours with pre2-ac2+preempth+radixtree.
Reverted to pre2-ac2.
Hope this helps,
Ed Tomlinson
--------------------------------------------------------
ksymoops 2.4.3 on i586 2.4.19-pre2-ac2. Options used
-V (default)
-k 20020303231146.ksyms (specified)
-l 20020303231146.modules (specified)
-o /lib/modules/2.4.19-pre2-ac2-prert (specified)
-m /boot/System.map-2.4.19-pre2-ac2-prert (specified)
Warning: loading /lib/modules/2.4.19-pre2-ac2-prert/kernel/net/ipv4/netfilter/ipchains.o will taint the kernel: non-GPL license - BSD without advertisement clause
ac97_codec: AC97 Audio codec, id: 0x4352:0x5903 (Cirrus Logic CS4297)
kernel BUG at page_alloc.c:239!
invalid operand: 0000
CPU: 0
EIP: 0010:[<c012d247>] Tainted: P
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010282
eax: 00000020 ebx: c026a73c ecx: ffffffdd edx: d40e0000
esi: c13ca53c edi: 00000001 ebp: d40e0000 esp: d40e1cb0
ds: 0018 es: 0018 ss: 0018
Process kdeinit (pid: 786, stackpage=d40e1000)
Stack: c022d19c 000000ef c026a73c c026a8f8 000001ff 00000000 c01117e4 d40e0000
000150d8 00000292 c026a77c 00000000 c026a73c c012d435 000000f0 c15b2b14
00000000 00007b9c 00007b9c 00000001 c026a8f4 000000f0 c012d2e6 00000003
Call Trace: [<c01117e4>] [<c012d435>] [<c012d2e6>] [<c0124b23>] [<c0137c91>]
[<c0137f20>] [<c0135f37>] [<c0136124>] [<c0171f32>] [<c016df25>] [<c0136124>]
[<c015cfe1>] [<c015d568>] [<c015d68f>] [<c01467f7>] [<c013d940>] [<c013e2e9>]
[<c013d5fe>] [<c013e57a>] [<c013ea21>] [<c0132ba0>] [<c0106da3>]
Code: 0f 0b 83 c4 08 8d 74 26 00 8b 46 14 a8 80 74 19 68 ef 00 00
>>EIP; c012d246 <rmqueue+256/2e0> <=====
Trace; c01117e4 <schedule+230/268>
Trace; c012d434 <__alloc_pages+70/2bc>
Trace; c012d2e6 <_alloc_pages+16/18>
Trace; c0124b22 <find_or_create_page+2a/d8>
Trace; c0137c90 <grow_dev_page+20/ac>
Trace; c0137f20 <grow_buffers+f4/13c>
Trace; c0135f36 <getblk+2a/40>
Trace; c0136124 <bread+18/bc>
Trace; c0171f32 <reiserfs_bread+16/24>
Trace; c016df24 <search_by_key+5c/d40>
Trace; c0136124 <bread+18/bc>
Trace; c015cfe0 <search_by_entry_key+1c/1c0>
Trace; c015d568 <reiserfs_find_entry+7c/134>
Trace; c015d68e <reiserfs_lookup+6e/e0>
Trace; c01467f6 <d_alloc+1a/194>
Trace; c013d940 <real_lookup+70/118>
Trace; c013e2e8 <link_path_walk+79c/a14>
Trace; c013d5fe <getname+5e/9c>
Trace; c013e57a <path_walk+1a/1c>
Trace; c013ea20 <__user_walk+34/50>
Trace; c0132ba0 <sys_access+94/128>
Trace; c0106da2 <system_call+32/40>
Code; c012d246 <rmqueue+256/2e0>
00000000 <_EIP>:
Code; c012d246 <rmqueue+256/2e0> <=====
0: 0f 0b ud2a <=====
Code; c012d248 <rmqueue+258/2e0>
2: 83 c4 08 add $0x8,%esp
Code; c012d24a <rmqueue+25a/2e0>
5: 8d 74 26 00 lea 0x0(%esi,1),%esi
Code; c012d24e <rmqueue+25e/2e0>
9: 8b 46 14 mov 0x14(%esi),%eax
Code; c012d252 <rmqueue+262/2e0>
c: a8 80 test $0x80,%al
Code; c012d254 <rmqueue+264/2e0>
e: 74 19 je 29 <_EIP+0x29> c012d26e <rmqueue+27e/2e0>
Code; c012d256 <rmqueue+266/2e0>
10: 68 ef 00 00 00 push $0xef
<1>Unable to handle kernel paging request at virtual address e5746608
c0129ece
*pde = 17423067
Oops: 0000
CPU: 0
EIP: 0010:[<c0129ece>] Tainted: P
EFLAGS: 00010082
eax: c906e0c0 ebx: f0415880 ecx: c158e2f0 edx: 00000007
esi: c158e2f0 edi: 00000246 ebp: 00000020 esp: d3a29de8
ds: 0018 es: 0018 ss: 0018
Process kdeinit (pid: 789, stackpage=d3a29000)
Stack: d495fb74 00000001 ffffffff 00000001 c0222fc5 c158e2f0 00000020 ffffffff
c1362970 d495fb74 00000000 c022304b d495fb74 00000001 d3a28000 c1362970
d495fb74 00000001 c02230f7 d495fb74 00000001 d3a29e40 c026a73c c01245d2
Call Trace: [<c0222fc5>] [<c022304b>] [<c02230f7>] [<c01245d2>] [<c01246dc>]
[<c012474a>] [<c0125c7a>] [<c0121fad>] [<c0122263>] [<c01108bf>] [<c01107a8>]
[<c013b3fe>] [<c0106ef4>]
Code: 8b 44 81 18 89 41 14 83 f8 ff 75 18 8b 41 04 8b 11 89 42 04
>>EIP; c0129ece <kmem_cache_alloc+92/d4> <=====
Trace; c0222fc4 <radix_tree_extend+54/9c>
Trace; c022304a <radix_tree_reserve+3e/d8>
Trace; c02230f6 <radix_tree_insert+12/2c>
Trace; c01245d2 <add_to_page_cache+32/ac>
Trace; c01246dc <page_cache_read+90/d8>
Trace; c012474a <read_cluster_nonblocking+26/40>
Trace; c0125c7a <filemap_nopage+12a/220>
Trace; c0121fac <do_no_page+70/250>
Trace; c0122262 <handle_mm_fault+d6/178>
Trace; c01108be <do_page_fault+116/42c>
Trace; c01107a8 <do_page_fault+0/42c>
Trace; c013b3fe <sys_fstat64+5a/6c>
Trace; c0106ef4 <error_code+34/40>
Code; c0129ece <kmem_cache_alloc+92/d4>
00000000 <_EIP>:
Code; c0129ece <kmem_cache_alloc+92/d4> <=====
0: 8b 44 81 18 mov 0x18(%ecx,%eax,4),%eax <=====
Code; c0129ed2 <kmem_cache_alloc+96/d4>
4: 89 41 14 mov %eax,0x14(%ecx)
Code; c0129ed4 <kmem_cache_alloc+98/d4>
7: 83 f8 ff cmp $0xffffffff,%eax
Code; c0129ed8 <kmem_cache_alloc+9c/d4>
a: 75 18 jne 24 <_EIP+0x24> c0129ef2 <kmem_cache_alloc+b6/d4>
Code; c0129eda <kmem_cache_alloc+9e/d4>
c: 8b 41 04 mov 0x4(%ecx),%eax
Code; c0129edc <kmem_cache_alloc+a0/d4>
f: 8b 11 mov (%ecx),%edx
Code; c0129ede <kmem_cache_alloc+a2/d4>
11: 89 42 04 mov %eax,0x4(%edx)
kernel BUG at page_alloc.c:239!
invalid operand: 0000
CPU: 0
EIP: 0010:[<c012d247>] Tainted: P
EFLAGS: 00013282
eax: 00000020 ebx: c026a73c ecx: ffffffe0 edx: d8ce4000
esi: c13c541c edi: 00000001 ebp: d8ce4000 esp: d8ce5e44
ds: 0018 es: 0018 ss: 0018
Process XFree86 (pid: 736, stackpage=d8ce5000)
Stack: c022d19c 000000ef c026a73c c026a918 000001ff 00000000 00003246 d8ce4000
00014f00 00003296 c026a77c 00000000 c026a73c c012d435 000001d2 c1002ccc
43b81000 d7420e04 d39d8d80 00000001 c026a914 000001d2 c012d2e6 d8ce4000
Call Trace: [<c012d435>] [<c012d2e6>] [<c0121e6c>] [<c0121f72>] [<c0122263>]
[<c01108bf>] [<c01107a8>] [<c011b9b6>] [<c0118ace>] [<c0118a08>] [<c01187e1>]
[<c010826e>] [<c0106ef4>]
Code: 0f 0b 83 c4 08 8d 74 26 00 8b 46 14 a8 80 74 19 68 ef 00 00
>>EIP; c012d246 <rmqueue+256/2e0> <=====
Trace; c012d434 <__alloc_pages+70/2bc>
Trace; c012d2e6 <_alloc_pages+16/18>
Trace; c0121e6c <do_anonymous_page+58/128>
Trace; c0121f72 <do_no_page+36/250>
Trace; c0122262 <handle_mm_fault+d6/178>
Trace; c01108be <do_page_fault+116/42c>
Trace; c01107a8 <do_page_fault+0/42c>
Trace; c011b9b6 <timer_bh+2e/2a4>
Trace; c0118ace <bh_action+26/70>
Trace; c0118a08 <tasklet_hi_action+5c/80>
Trace; c01187e0 <do_softirq+60/c8>
Trace; c010826e <do_IRQ+d6/e4>
Trace; c0106ef4 <error_code+34/40>
Code; c012d246 <rmqueue+256/2e0>
00000000 <_EIP>:
Code; c012d246 <rmqueue+256/2e0> <=====
0: 0f 0b ud2a <=====
Code; c012d248 <rmqueue+258/2e0>
2: 83 c4 08 add $0x8,%esp
Code; c012d24a <rmqueue+25a/2e0>
5: 8d 74 26 00 lea 0x0(%esi,1),%esi
Code; c012d24e <rmqueue+25e/2e0>
9: 8b 46 14 mov 0x14(%esi),%eax
Code; c012d252 <rmqueue+262/2e0>
c: a8 80 test $0x80,%al
Code; c012d254 <rmqueue+264/2e0>
e: 74 19 je 29 <_EIP+0x29> c012d26e <rmqueue+27e/2e0>
Code; c012d256 <rmqueue+266/2e0>
10: 68 ef 00 00 00 push $0xef
kernel BUG at inode.c:515!
invalid operand: 0000
CPU: 0
EIP: 0010:[<c0148076>] Tainted: P
EFLAGS: 00210282
eax: 0000001b ebx: d4fa03a0 ecx: ffffffe5 edx: d5484000
esi: dd46f8e8 edi: dd46f800 ebp: d5129ac0 esp: d5485f2c
ds: 0018 es: 0018 ss: 0018
Process artsd (pid: 772, stackpage=d5485000)
Stack: c022f787 00000203 d4fa03a0 c0130256 d4fa03a0 c01301b8 d4fa03a0 c0148b16
d4fa03a0 d5484000 d5129ac0 d4fa03a0 c0146ca6 d4fa03a0 d5484000 00000000
d53253e0 c013fa8a d5129ac0 d5129ac0 d5129ac0 d1d49000 d5485fa4 c013fb65
Call Trace: [<c0130256>] [<c01301b8>] [<c0148b16>] [<c0146ca6>] [<c013fa8a>]
[<c013fb65>] [<c0106da3>]
Code: 0f 0b 83 c4 08 90 8d 74 26 00 f6 83 1c 01 00 00 10 75 17 68
>>EIP; c0148076 <clear_inode+26/dc> <=====
Trace; c0130256 <shmem_delete_inode+9e/a4>
Trace; c01301b8 <shmem_delete_inode+0/a4>
Trace; c0148b16 <iput+f2/220>
Trace; c0146ca6 <d_delete+6a/c8>
Trace; c013fa8a <vfs_unlink+15a/190>
Trace; c013fb64 <sys_unlink+a4/118>
Trace; c0106da2 <system_call+32/40>
Code; c0148076 <clear_inode+26/dc>
00000000 <_EIP>:
Code; c0148076 <clear_inode+26/dc> <=====
0: 0f 0b ud2a <=====
Code; c0148078 <clear_inode+28/dc>
2: 83 c4 08 add $0x8,%esp
Code; c014807a <clear_inode+2a/dc>
5: 90 nop
Code; c014807c <clear_inode+2c/dc>
6: 8d 74 26 00 lea 0x0(%esi,1),%esi
Code; c0148080 <clear_inode+30/dc>
a: f6 83 1c 01 00 00 10 testb $0x10,0x11c(%ebx)
Code; c0148086 <clear_inode+36/dc>
11: 75 17 jne 2a <_EIP+0x2a> c01480a0 <clear_inode+50/dc>
Code; c0148088 <clear_inode+38/dc>
13: 68 00 00 00 00 push $0x0
kernel BUG at swap_state.c:141!
invalid operand: 0000
CPU: 0
EIP: 0010:[<c012db30>] Tainted: P
EFLAGS: 00010286
eax: 00000020 ebx: c13d8ab0 ecx: ffffffe0 edx: cb574000
esi: cb574000 edi: c13d8ab0 ebp: d66128f4 esp: cb575ecc
ds: 0018 es: 0018 ss: 0018
Process iceauth (pid: 4904, stackpage=cb575000)
Stack: c022d43d 0000008d 00003202 c012dd1b c13d8ab0 c13d8ab0 00000000 00000001
cb574000 c13d8ab0 00000000 d6612990 00000000 d66128fc c01306d5 c13d8ab0
00000000 d66128f4 d6612974 d6612840 cb575f6c d6612978 00003202 d66128f4
Call Trace: [<c012dd1b>] [<c01306d5>] [<c01308cc>] [<c01311bc>] [<c01312bc>]
[<c0133bea>] [<c0106da3>]
Code: 0f 0b 83 c4 08 b8 04 00 00 00 0f b3 43 14 53 e8 ac 60 ff ff
>>EIP; c012db30 <__delete_from_swap_cache+38/58> <=====
Trace; c012dd1a <move_from_swap_cache+6a/cc>
Trace; c01306d4 <shmem_getpage_locked+1b0/35c>
Trace; c01308cc <shmem_getpage+4c/94>
Trace; c01311bc <do_shmem_file_read+44/e8>
Trace; c01312bc <shmem_file_read+5c/74>
Trace; c0133bea <sys_read+8e/128>
Trace; c0106da2 <system_call+32/40>
Code; c012db30 <__delete_from_swap_cache+38/58>
00000000 <_EIP>:
Code; c012db30 <__delete_from_swap_cache+38/58> <=====
0: 0f 0b ud2a <=====
Code; c012db32 <__delete_from_swap_cache+3a/58>
2: 83 c4 08 add $0x8,%esp
Code; c012db34 <__delete_from_swap_cache+3c/58>
5: b8 04 00 00 00 mov $0x4,%eax
Code; c012db3a <__delete_from_swap_cache+42/58>
a: 0f b3 43 14 btr %eax,0x14(%ebx)
Code; c012db3e <__delete_from_swap_cache+46/58>
e: 53 push %ebx
Code; c012db3e <__delete_from_swap_cache+46/58>
f: e8 ac 60 ff ff call ffff60c0 <_EIP+0xffff60c0> c0123bf0 <__remove_inode_page+0/4c>
On Sun, Mar 03, 2002 at 11:55:57PM -0500, Ed Tomlinson wrote:
> On March 3, 2002 03:03 pm, Christoph Hellwig wrote:
> > I have uploaded an updated version of the radix-tree pagecache patch
> > against 2.4.19-pre2-ac2. News in this release:
> >
> > * fix a deadlock when vmtruncate takes i_shared_lock twice by introducing
> > a new mapping->page_lock that mutexes mapping->page_tree. (akpm)
> > * move setting of page->flags back out of move_to/from_swap_cache. (akpm)
> > * put back lost page state settings in shmem_unuse_inode. (akpm)
> > * get rid of remove_page_from_inode_queue - there was only one caller. (me)
> > * replace add_page_to_inode_queue with ___add_to_page_cache. (me)
> >
> > Please give it some serious beating while I try to get 2.5 working and
> > port the patch over 8)
>
> Got this after a couple of hours with pre2-ac2+preempth+radixtree.
>
Can you try again without preempt?
Christoph Hellwig wrote:
>
> I have uploaded an updated version of the radix-tree pagecache patch
> against 2.4.19-pre2-ac2. News in this release:
>
> * fix a deadlock when vmtruncate takes i_shared_lock twice by introducing
> a new mapping->page_lock that mutexes mapping->page_tree. (akpm)
> * move setting of page->flags back out of move_to/from_swap_cache. (akpm)
> * put back lost page state settings in shmem_unuse_inode. (akpm)
> * get rid of remove_page_from_inode_queue - there was only one caller. (me)
> * replace add_page_to_inode_queue with ___add_to_page_cache. (me)
>
> Please give it some serious beating while I try to get 2.5 working and
> port the patch over 8)
One of my reasons for absorbing ratcache into my current stuff is
just that - to give it a serious beating. The fact that I found a
hitherto-undiscovered BUG() and a deadlock in the first 30 minutes
just shows what a mean beat I have :)
I haven't yet even looked at lib/rat.c, but based on testing, I
believe radix-tree pagecache is ready for 2.5. It would be good
if the other Christoph could check over the shmem.c changes.
As far as I know, the sole remaining "issue" is that block_flushpage()
is being called under spinlock. Well, there's nothing new here.
The kernel is *already* calling block_flushpage() under spinlock
it at least three places. But it just so happens that there are
never (?) any locked buffers against the page from those call sites.
So I don't see the block_flushpage() thing as a blocker for this
patch - it's just general ickiness which needs sorting out separately.
I looked at block_flushpage() a month or so back. I ended up
concluding that we should just create block_flushpage_atomic() and
make it go BUG() if the page has locked buffers. Then call the atomic
version from under spinlock, and leave it at that.
-
Replying to Christoph Hellwig:
> I have uploaded an updated version of the radix-tree pagecache patch
> against 2.4.19-pre2-ac2. News in this release:
60% the patch is broken. I got 2 oopses. Both looking the same.
ksymoops 2.4.1 on i686 2.4.19-pre2-ac2. Options used
-V (default)
-k /proc/ksyms (default)
-l /proc/modules (default)
-o /lib/modules/2.4.19-pre2-ac2/ (default)
-m /boot/System.map-2.4.19-pre2-ac2 (default)
Warning: You did not tell me where to find symbol information. I will
assume that the log matches the kernel and modules that are running
right now and I'll use the default options above for symbol resolution.
If the current kernel and/or modules do not match the log, you can get
more accurate output by telling me the kernel version and where to find
map, modules, ksyms etc. ksymoops -h explains the options.
Warning (compare_maps): mismatch on symbol md_size , md says c4864900, /lib/modules/2.4.19-pre2-ac2/kernel/drivers/md/md.o says c4864720. Ignoring /lib/modules/2.4.19-pre2-ac2/kernel/drivers/md/md.o entry
Warning (compare_maps): mismatch on symbol mddev_map , md says c4864100, /lib/modules/2.4.19-pre2-ac2/kernel/drivers/md/md.o says c4863f20. Ignoring /lib/modules/2.4.19-pre2-ac2/kernel/drivers/md/md.o entry
Warning (compare_maps): mismatch on symbol usb_devfs_handle , usbcore says c484fa34, /lib/modules/2.4.19-pre2-ac2/kernel/drivers/usb/usbcore.o says c484f4f4. Ignoring /lib/modules/2.4.19-pre2-ac2/kernel/drivers/usb/usbcore.o entry
Mar 4 19:57:24 lanserv2 kernel: invalid operand: 0000
Mar 4 19:57:24 lanserv2 kernel: CPU: 0
Mar 4 19:57:24 lanserv2 kernel: EIP: 0010:[shmem_writepage+157/272] Not tainted
Mar 4 19:57:24 lanserv2 kernel: EIP: 0010:[<c013577d>] Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
Mar 4 19:57:24 lanserv2 kernel: EFLAGS: 00010212
Mar 4 19:57:24 lanserv2 kernel: eax: fffffff4 ebx: c3bea0e0 ecx: 0000056b edx: 00000143
Mar 4 19:57:24 lanserv2 kernel: esi: c10887c8 edi: fffffff4 ebp: 00056a00 esp: c3fc7f54
Mar 4 19:57:24 lanserv2 kernel: ds: 0018 es: 0018 ss: 0018
Mar 4 19:57:24 lanserv2 kernel: Process kswapd (pid: 4, stackpage=c3fc7000)
Mar 4 19:57:24 lanserv2 kernel: Stack: 00050d00 00056a00 c10887c8 c10887e0 c0297ca0 0000002a c01307d5 c10887c8
Mar 4 19:57:24 lanserv2 kernel: 00000154 00000015 c0297e1c 00000154 00000086 00000048 c0130bc8 c0297e1c
Mar 4 19:57:24 lanserv2 kernel: 000001d0 00000006 0000003e 000001d0 00000082 00000000 00000000 c0131472
Mar 4 19:57:24 lanserv2 kernel: Call Trace: [page_launder_zone+933/1584] [page_launder+360/768] [do_try_to_free_pages+18/384] [
Mar 4 19:57:24 lanserv2 kernel: Call Trace: [<c01307d5>] [<c0130bc8>] [<c0131472>] [<c0131771>] [<c0105000>]
Mar 4 19:57:24 lanserv2 kernel: [<c01056a6>] [<c0131670>]
Mar 4 19:57:24 lanserv2 kernel: Code: 0f 0b 53 e8 3b f9 ff ff 8b 0f 58 85 c9 74 02 0f 0b 55 56 e8
>>EIP; c013577d <shmem_writepage+9d/110> <=====
Trace; c01307d5 <page_launder_zone+3a5/630>
Trace; c0130bc8 <page_launder+168/300>
Trace; c0131472 <do_try_to_free_pages+12/180>
Trace; c0131771 <kswapd+101/2d0>
Trace; c0105000 <_stext+0/0>
Trace; c01056a6 <kernel_thread+26/30>
Trace; c0131670 <kswapd+0/2d0>
Code; c013577d <shmem_writepage+9d/110>
00000000 <_EIP>:
Code; c013577d <shmem_writepage+9d/110> <=====
0: 0f 0b ud2a <=====
Code; c013577f <shmem_writepage+9f/110>
2: 53 push %ebx
Code; c0135780 <shmem_writepage+a0/110>
3: e8 3b f9 ff ff call fffff943 <_EIP+0xfffff943> c01350c0 <shmem_recalc_inode+0/60>
Code; c0135785 <shmem_writepage+a5/110>
8: 8b 0f mov (%edi),%ecx
Code; c0135787 <shmem_writepage+a7/110>
a: 58 pop %eax
Code; c0135788 <shmem_writepage+a8/110>
b: 85 c9 test %ecx,%ecx
Code; c013578a <shmem_writepage+aa/110>
d: 74 02 je 11 <_EIP+0x11> c013578e <shmem_writepage+ae/110>
Code; c013578c <shmem_writepage+ac/110>
f: 0f 0b ud2a
Code; c013578e <shmem_writepage+ae/110>
11: 55 push %ebp
Code; c013578f <shmem_writepage+af/110>
12: 56 push %esi
Code; c0135790 <shmem_writepage+b0/110>
13: e8 00 00 00 00 call 18 <_EIP+0x18> c0135795 <shmem_writepage+b5/110>
4 warnings issued. Results may not be reliable.
Both happens under heavy swapout.
under -ac2 without this patch same case walks fine.
(why I'm still unsure - I have a bit of reiserfs quota in both kernels. This
oopses may be caused by incorrect merge. If youre interested, I can send
merge results.)
--
Paul P 'Stingray' Komkoff 'Greatest' Jr // (icq)23200764 // (irc)Spacebar
PPKJ1-RIPE // (smtp)[email protected] // (http)stingr.net // (pgp)0xA4B4ECA4
On Mon, 2002-03-04 at 00:13, Mike Fedyk wrote:
> On Sun, Mar 03, 2002 at 11:55:57PM -0500, Ed Tomlinson wrote:
>
> > Got this after a couple of hours with pre2-ac2+preempth+radixtree.
>
> Can you try again without preempt?
I've had success with the patch on 2.4.18+preempt and 2.5.5, so I
suspect preemption is not a problem. I also did not see any
preempt_schedules in his backtrace ...
Ah, someone else just posted another oops on 2.4.19-pre2-ac2 ...
Robert Love
On Mon, Mar 04, 2002 at 03:31:52PM -0500, Robert Love wrote:
> On Mon, 2002-03-04 at 00:13, Mike Fedyk wrote:
>
> > On Sun, Mar 03, 2002 at 11:55:57PM -0500, Ed Tomlinson wrote:
> >
> > > Got this after a couple of hours with pre2-ac2+preempth+radixtree.
> >
> > Can you try again without preempt?
>
> I've had success with the patch on 2.4.18+preempt and 2.5.5, so I
> suspect preemption is not a problem. I also did not see any
> preempt_schedules in his backtrace ...
I can repdoduce it locally here. IT looks like we leak a pgae with
incorrect flags in an error path. Still investigating it.
Christoph
--
Of course it doesn't work. We've performed a software upgrade.
On Mon, 4 Mar 2002 23:25:10 +0300
Paul P Komkoff Jr <[email protected]> wrote:
> Replying to Christoph Hellwig:
> > I have uploaded an updated version of the radix-tree pagecache patch
> > against 2.4.19-pre2-ac2. News in this release:
>
> 60% the patch is broken. I got 2 oopses. Both looking the same.
I have this same problem, same place (shmem.c line 498). BUG
triggered on calling shmem_writepage. Through interesting
coincidence, I also have had it happen twice. The second time
I noticed that this happened with RAM basically gone, swap
usage at about 20%. I was doing a 'make install'.
2.4.19-pre2-ac2 +rmap12g +preempt,lockbreak +radix tree pagecache
+ a few fixes from 2.4.18-ac3
- Matt