Kernel oops in mm/rmap.c in 2.6.11rc4-bk9
Subsystems (3)
Hi,
Fresher trace of my oops captured over a serial tty link (kernel now 2.6.11rc4-bk11)
Kernel 2.6.11colin on an i686 / ttyS0 verbier.straightrunning.com login: ------------[ cut here ]------------ kernel BUG at mm/rmap.c:482!
invalid operand: 0000 [#1]
Modules linked in: parport_pc lp parport ipt_LOG ipt_REJECT ipt_state ipt_limite CPU: 0 EIP: 0060:[<c013cc28>] Not tainted VLI EFLAGS: 00010296 (2.6.11colin)
EIP is at page_remove_rmap+0x38/0x50
eax: fffff000 ebx: 00064000 ecx: c039878c edx: c123ca00
esi: d2389f30 edi: 00098000 ebp: d10e0e4c esp: d10e0e4c
ds: 007b es: 007b ss: 0068
Process cc1 (pid: 10543, threadinfo=d10e0000 task=ced11a80)
Stack: d10e0e80 c013661b c12335c0 cef90904 00000001 00000000 11e50067
c123ca00
b7b68000 c039878c b7f68000 d1f10b7c b7c00000 d10e0ea8 c01367a1
00098000
00000000 c039878c b8241000 c039878c b7b68000 d1f10b7c b7c00000
d10e0ed0
Call Trace:
[<c0102736>] show_stack+0xa6/0xb0
[<c01028a9>] show_registers+0x149/0x1c0
[<c0102a8b>] die+0xbb/0x150
[<c0102b9e>] do_trap+0x7e/0xc0
[<c0102e15>] do_invalid_op+0xa5/0xb0
[<c01023af>] error_code+0x2b/0x30
[<c013661b>] zap_pte_range+0x13b/0x270
[<c01367a1>] zap_pmd_range+0x51/0x70
[<c01367fd>] zap_pud_range+0x3d/0x70
[<c0136897>] unmap_page_range+0x67/0x80
[<c01369a4>] unmap_vmas+0xf4/0x1e0
[<c013afe8>] exit_mmap+0x68/0x130
[<c01100d1>] mmput+0x21/0x70
[<c011343e>] exit_mm+0xbe/0xe0
[<c0113d9d>] do_exit+0x9d/0x2c0
[<c0114032>] do_group_exit+0x32/0x70
[<c011407f>] sys_exit_group+0xf/0x20
[<c0102145>] sysenter_past_esp+0x52/0x75
Code: 42 08 ff 0f 98 c0 84 c0 74 15 8b 42 08 40 78 1b ba ff ff ff ff b8 1000 0
Entering kdb (current=0xced11a80, pid 10543) Oops: invalid operand due to oops @ 0xc013cc28
eax = 0xfffff000 ebx = 0x00064000 ecx = 0xc039878c edx = 0xc123ca00 esi = 0xd2389f30 edi = 0x00098000 esp = 0xd10e0e4c eip = 0xc013cc28 ebp = 0xd10e0e4c xss = 0xc0210068 xcs = 0x00000060 eflags = 0x00010296 xds = 0xc122007b xes = 0x0000007b origeax = 0xffffffff ®s = 0xd10e0e18 kdb>
Thanks
Colin Harrison
Quoted 64 lines
-----Original Message-----
From: Colin Harrison [mailto:[email protected]]
Sent: 21 February 2005 18:45
To: '[email protected]'
Subject: Kernel oops inmm/rmap.cin 2.6.11rc4-bk9Hi
I've been getting a hang on my firewall machine for some weeks when pushed.
Kernel freezes when heavily loaded either compiling or under lots of web traffic.
Machine is a Pentium III 450MHz i440BX Intel Mobo with 384MHz memory running a firewall (iptables-1.3.0-200050220) via two natsemi network cards.
(old machine but should be good as firewall only...been solid for years!) Kernel 2.6.11rc4-bk9 + supermount patch.I have another machine which I am going to connect via serial cable to help debug.
(this has also been known to freeze on 2.6.11...a common denominator is that I use Netgear natsemi net cards in both...probably not significant?)I've today built in KDB and get the following trace:-
In this case while compiling a kernel (make bzImage...running cc1)
Typed by hand:-
kernel BUG at
mm/rmap.c:482!
invalid operand: 0000 [#1]Modules linked in: parport_pc lp parport ipt_LOG ipt_REJECT ipt_state ipt_limit iptable_mangle iptable_filter ip_nat_ftp iptable_nat ip_conntrack_ftp floppy natsemi supermount intel_agp agpgart cs4243 ad1848 uart401 sound soundcore CPU: 0 EIP: 0060:[<c013cc28>] Not tainted VLI EFLAGS: 00010296 (2.6.11colin)EIP is at
page_remove_rmap+0x38/0x50
eax:fffff000ebx:0002f000ecx:c038d78cedx: c123sd20
esi:d0e7e44cedi: 00090000 ebp:d2a6ae4cesp:d2a6ae4c
ds: 007b es: 007b ss: 0068
Process cc1 (pid: 10835, threadinfo=d2a6a000 task=cf57aaa0)Stack: d2a6ae80etc......
Entered kdb (current=0xcf57aaa0, pid 10835) Oops: invalid operand due to oops @
0xc013cc28etc...More info can be supplied if required and patches compiled in to trace etc.
Note that I have run a memory checker (memtest-86 v3.2 boot CD) for ~ 3hours with no errors.
I can usually repeat the crash while compiling a kernel..which after a reboot recovers and allows 'make bzImage' 'make modules' to finish!Thanks
Colin Harrison
Quoted 6 lines · Colin Harrison
Fresher trace of my oops captured over a serial tty link (kernel now 2.6.11rc4-bk11)
kernel BUG at
mm/rmap.c:482!
EIP is atpage_remove_rmap+0x38/0x50
Process cc1 (pid: 10543, threadinfo=d10e0000 task=ced11a80)
Welcome to the exclusive club of those who've seen this - check the archives for other sightings.
Quoted 5 lines
Quoted 5 lines
-----Original Message-----
From: Colin Harrison [mailto:[email protected]]
Sent: 21 February 2005 18:45
To: '[email protected]'
Subject: Kernel oops inmm/rmap.cin 2.6.11rc4-bk9
Sorry, that message never reached the list - marc hasn't got it.
[ I've snipped out a lot of info you were right to supply,
thank you, though it'll probably turn out to be irrelevant. ]
Quoted 2 lines
Quoted 2 lines
More info can be supplied if required and patches compiled in to trace etc.
Thank you: please apply the patch below, and mail me back any interesting messages you see - should be captured by dmesg, or better /var/log/messages which will show the times too, to help group them. Not just the "Bad rmap:" errors this patch adds, there might be relevant "Bad page state" errors or "swap_free:" errors. The system should stay up. Perhaps you'll be swamped with messages (don't spam the list with them if so, just me).
Quoted 2 lines
Quoted 2 lines
Note that I have run a memory checker (memtest-86 v3.2 boot CD) for ~ 3hours with no errors.
Right thing to try first, though may not have been long enough. Since you've already tried that, let's try the patch below next.
Quoted 3 lines
Quoted 3 lines
I can usually repeat the crash while compiling a kernel..which after a reboot recovers and allows 'make bzImage' 'make modules' to finish!
If you can reproduce this fairly easily, that would be _wonderful_! It's dragged on for months, but nobody sees it often enough to get an understanding of it. It's possible that most sightings are in fact due to bad memory, and the rmap.c BUG a good (but tiresome) checker itself - I do hope your info will give us more of a clue.
Thanks,
Hugh
p.s. Ammar, I've cc'ed you for interest as the most recent member of the club; but this patch below is unlikely to apply cleanly to your Fedora kernel - if you're used to moving patches from one release to another, you might want to try it, but probably not.
include/linux/rmap.h +1 -1
@@ -72,7 +72,7 @@ void __anon_vma_link(struct vm_area_stru*/ void page_add_anon_rmap (struct page *,struct vm_area_struct *,unsigned long );void page_add_file_rmap (struct page *); -void page_remove_rmap (struct page *); +void page_remove_rmap (struct page *,struct vm_area_struct *,unsigned long );/** * page_dup_rmap - duplicate pte mapping to a page
mm/fremap.c +1 -1
@@ -37,7 +37,7 @@ static inline void zap_pte(struct mm_strif (!PageReserved (page)) {if (pte_dirty (pte))set_page_dirty (page); -page_remove_rmap (page); +page_remove_rmap (page, vma, addr);page_cache_release (page); mm->rss--; }
mm/memory.c +8 -5
@@ -452,6 +452,7 @@ next_pgd: }static void zap_pte_range (struct mmu_gather *tlb, +struct vm_area_struct *vma,pmd_t *pmd,unsigned long address,unsigned long size,struct zap_details *details) {
@@ -517,7 +518,7 @@ static void zap_pte_range(struct mmu_gatelse if (pte_young (pte))mark_page_accessed (page); tlb->freed++; -page_remove_rmap (page); +page_remove_rmap (page, vma, address+offset);tlb_remove_page (tlb, page);continue ; }
@@ -535,6 +536,7 @@ static void zap_pte_range(struct mmu_gat }static void zap_pmd_range (struct mmu_gather *tlb, +struct vm_area_struct *vma,pud_t *pud,unsigned long address,unsigned long size,struct zap_details *details) {
@@ -553,13 +555,14 @@ static void zap_pmd_range(struct mmu_gatif (end > ((address +PUD_SIZE ) &PUD_MASK )) end = ((address +PUD_SIZE ) &PUD_MASK );do { -zap_pte_range (tlb, pmd, address, end - address, details); +zap_pte_range (tlb, vma, pmd, address, end - address, details); address = (address +PMD_SIZE ) &PMD_MASK ; pmd++; }while (address && (address < end)); }static void zap_pud_range (struct mmu_gather *tlb, +struct vm_area_struct *vma,pgd_t * pgd,unsigned long address,unsigned long end,struct zap_details *details) {
@@ -574,7 +577,7 @@ static void zap_pud_range(struct mmu_gat } pud =pud_offset (pgd, address);do { -zap_pmd_range (tlb, pud, address, end - address, details); +zap_pmd_range (tlb, vma, pud, address, end - address, details); address = (address +PUD_SIZE ) &PUD_MASK ; pud++; }while (address && (address < end));
@@ -595,7 +598,7 @@ static void unmap_page_range(struct mmu_ next = (address +PGDIR_SIZE ) &PGDIR_MASK ;if (next <= address || next > end) next = end; -zap_pud_range (tlb, pgd, address, next, details); +zap_pud_range (tlb, vma, pgd, address, next, details); address = next; pgd++; }
@@ -1343,7 +1346,7 @@ static int do_wp_page(struct mm_struct *acct_update_integrals ();update_mem_hiwater (); }else -page_remove_rmap (old_page); +page_remove_rmap (old_page, vma, address);break_cow (vma, new_page, address, page_table);lru_cache_add_active (new_page);page_add_anon_rmap (new_page, vma, address);
mm/rmap.c +40 -5
74 lines not on this page · raw message