Skip to contentSkip to conversation

a bug on 325d0eab4f31

DISCUSSION Alex Shi 2 participants 4 messages Sep 22, 2020 – Sep 24, 2020 Memory management

lore
Share & tools
lore

Alex Shi

author
#0 lore raw

I just found a bug on recent upstream kernel on my x86 box.
325d0eab4f31 Merge branch 'akpm' (patches from Andrew)
The last good kernel I known is 5.9-rc2.

Thanks
Alex

Linux aliy8 5.9.0-rc5-00156-g325d0eab4f31 #88 SMP PREEMPT Mon Sep 21 08:48:19 CST 2020 x86_64 x86_64 x86_64 GNU/Linux

1841.488609] BUG: Bad page state in process userfaultfd pfn:1a7d7

[ 1841.494930] page:00000000ec5d2f1e refcount:0 mapcount:0 mapping:0000000000000000 index:0x7fd6181a6 pfn:0x1a7d7
[ 1841.505080] flags: 0xfffffc0080004(uptodate|swapbacked)
[ 1841.510427] raw: 000fffffc0080004 dead000000000100 dead000000000122 0000000000000000
[ 1841.518318] raw: 00000007fd6181a6 0000000000000000 00000000ffffffff 0000000000000000
[ 1841.526207] page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
[ 1841.532775] Modules linked in: loop xt_addrtype br_netfilter xt_CHECKSUM iptable_mangle xt_MASQUERADE iptable_nat nf_nat xt_conntrack nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 libcrc32c ipt_REJECT nf_reject_ipv4 tun bridge stp llc overlay ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter dm_mirror dm_region_hash dm_log dm_mod ipmi_ssif intel_rapl_msr intel_rapl_common x86_pkg_temp_thermal coretemp kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel glue_helper crypto_simd cryptd pcspkr i2c_i801 joydev i2c_smbus mei_me mei ipmi_si ipmi_devintf ipmi_msghandler ip_tables ext4 mbcache jbd2 crc32c_intel drm_vram_helper drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm_ttm_helper ttm drm ahci libahci libata virtio_net net_failover failover
[ 1841.603813] CPU: 0 PID: 26221 Comm: userfaultfd Kdump: loaded Tainted: G S  B   W         5.9.0-rc5-00156-g325d0eab4f31 #88
[ 1841.615080] Hardware name: Alibaba X-Dragon CN 01/20G4B, BIOS 1ALSP017 04/09/2019
[ 1841.622701] Call Trace:
[ 1841.625280]  dump_stack+0x8d/0xc0
[ 1841.628726]  bad_page.cold.130+0x63/0x93
[ 1841.632776]  free_pcp_prepare+0x20b/0x250
[ 1841.636908]  free_unref_page+0x18/0x90
[ 1841.640780]  wp_page_copy+0x2cd/0x5d0
[ 1841.644569]  __handle_mm_fault+0x791/0x7a0
[ 1841.648790]  ? handle_mm_fault+0x4b/0x3f0
[ 1841.652925]  handle_mm_fault+0x16f/0x3f0
[ 1841.656968]  exc_page_fault+0x3cb/0x6a0
[ 1841.660920]  ? asm_exc_page_fault+0x8/0x30
[ 1841.665131]  asm_exc_page_fault+0x1e/0x30
[ 1841.669259] RIP: 0033:0x7fd625da6d5d
[ 1841.672957] Code: 00 00 89 77 10 89 f2 49 8d 70 16 4c 89 c7 81 e2 80 00 00 00 e9 04 6c 00 00 0f 1f 40 00 81 e6 80 00 00 00 bf 01 00 00 00 31 c0 <f0> 41 0f b1 38 0f 85 24 01 00 00 64 8b 04 25 d0 02 00 00 41 89 40
[ 1841.691889] RSP: 002b:00007fd6147b2e58 EFLAGS: 00010246
[ 1841.697239] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 00000000000027da
[ 1841.704497] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000001
[ 1841.711759] RBP: 00007fd6147b2f10 R08: 00007fd6181a6000 R09: 00007fd6147b2eb0
[ 1841.719028] R10: 000000000000000f R11: 0000000000000206 R12: 0000000000000000
[ 1841.726286] R13: 0000000001001000 R14: 0000000000000000 R15: 00007fd6147b3700
author
#1 lore raw

Re: a bug on 5.9-rc6

Quoted 4 lines · Alex Shi

I just found a bug on recent upstream kernel on my x86 box.
325d0eab4f31 Merge branch 'akpm' (patches from Andrew)
The last good kernel I known is 5.9-rc2.

This bug still exist on 5.9-rc6.

I found it in doing stress kernel build on swapping, while run kernel selftest's vmtest like:

	for ((i=0;i<8000;i++)); do ./run_vmtests ; sleep 1; done

It take time.
I will try to do bisect when things down in hands.

Thanks
Alex

Quoted 36 lines

Thanks
Alex

Linux aliy8 5.9.0-rc5-00156-g325d0eab4f31 #88 SMP PREEMPT Mon Sep 21 08:48:19 CST 2020 x86_64 x86_64 x86_64 GNU/Linux

1841.488609] BUG: Bad page state in process userfaultfd pfn:1a7d7

[ 1841.494930] page:00000000ec5d2f1e refcount:0 mapcount:0 mapping:0000000000000000 index:0x7fd6181a6 pfn:0x1a7d7
[ 1841.505080] flags: 0xfffffc0080004(uptodate|swapbacked)
[ 1841.510427] raw: 000fffffc0080004 dead000000000100 dead000000000122 0000000000000000
[ 1841.518318] raw: 00000007fd6181a6 0000000000000000 00000000ffffffff 0000000000000000
[ 1841.526207] page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
[ 1841.532775] Modules linked in: loop xt_addrtype br_netfilter xt_CHECKSUM iptable_mangle xt_MASQUERADE iptable_nat nf_nat xt_conntrack nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 libcrc32c ipt_REJECT nf_reject_ipv4 tun bridge stp llc overlay ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter dm_mirror dm_region_hash dm_log dm_mod ipmi_ssif intel_rapl_msr intel_rapl_common x86_pkg_temp_thermal coretemp kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel glue_helper crypto_simd cryptd pcspkr i2c_i801 joydev i2c_smbus mei_me mei ipmi_si ipmi_devintf ipmi_msghandler ip_tables ext4 mbcache jbd2 crc32c_intel drm_vram_helper drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm_ttm_helper ttm drm ahci libahci libata virtio_net net_failover failover
[ 1841.603813] CPU: 0 PID: 26221 Comm: userfaultfd Kdump: loaded Tainted: G S  B   W         5.9.0-rc5-00156-g325d0eab4f31 #88
[ 1841.615080] Hardware name: Alibaba X-Dragon CN 01/20G4B, BIOS 1ALSP017 04/09/2019
[ 1841.622701] Call Trace:
[ 1841.625280]  dump_stack+0x8d/0xc0
[ 1841.628726]  bad_page.cold.130+0x63/0x93
[ 1841.632776]  free_pcp_prepare+0x20b/0x250
[ 1841.636908]  free_unref_page+0x18/0x90
[ 1841.640780]  wp_page_copy+0x2cd/0x5d0
[ 1841.644569]  __handle_mm_fault+0x791/0x7a0
[ 1841.648790]  ? handle_mm_fault+0x4b/0x3f0
[ 1841.652925]  handle_mm_fault+0x16f/0x3f0
[ 1841.656968]  exc_page_fault+0x3cb/0x6a0
[ 1841.660920]  ? asm_exc_page_fault+0x8/0x30
[ 1841.665131]  asm_exc_page_fault+0x1e/0x30
[ 1841.669259] RIP: 0033:0x7fd625da6d5d
[ 1841.672957] Code: 00 00 89 77 10 89 f2 49 8d 70 16 4c 89 c7 81 e2 80 00 00 00 e9 04 6c 00 00 0f 1f 40 00 81 e6 80 00 00 00 bf 01 00 00 00 31 c0 <f0> 41 0f b1 38 0f 85 24 01 00 00 64 8b 04 25 d0 02 00 00 41 89 40
[ 1841.691889] RSP: 002b:00007fd6147b2e58 EFLAGS: 00010246
[ 1841.697239] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 00000000000027da
[ 1841.704497] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000001
[ 1841.711759] RBP: 00007fd6147b2f10 R08: 00007fd6181a6000 R09: 00007fd6147b2eb0
[ 1841.719028] R10: 000000000000000f R11: 0000000000000206 R12: 0000000000000000
[ 1841.726286] R13: 0000000001001000 R14: 0000000000000000 R15: 00007fd6147b3700
#2 lore raw

Re: a bug on 5.9-rc6

Looks like the same problem as
https://lore.kernel.org/linux-mm/[email protected]/

Quoted 57 lines · Alex Shi
Quoted 4 lines · Alex Shi

I just found a bug on recent upstream kernel on my x86 box.
325d0eab4f31 Merge branch 'akpm' (patches from Andrew)
The last good kernel I known is 5.9-rc2.

This bug still exist on 5.9-rc6.

I found it in doing stress kernel build on swapping, while run kernel selftest's vmtest like:

        for ((i=0;i<8000;i++)); do ./run_vmtests ; sleep 1; done

It take time.
I will try to do bisect when things down in hands.

Thanks
Alex

Quoted 36 lines

Thanks
Alex

Linux aliy8 5.9.0-rc5-00156-g325d0eab4f31 #88 SMP PREEMPT Mon Sep 21 08:48:19 CST 2020 x86_64 x86_64 x86_64 GNU/Linux

1841.488609] BUG: Bad page state in process userfaultfd pfn:1a7d7

[ 1841.494930] page:00000000ec5d2f1e refcount:0 mapcount:0 mapping:0000000000000000 index:0x7fd6181a6 pfn:0x1a7d7
[ 1841.505080] flags: 0xfffffc0080004(uptodate|swapbacked)
[ 1841.510427] raw: 000fffffc0080004 dead000000000100 dead000000000122 0000000000000000
[ 1841.518318] raw: 00000007fd6181a6 0000000000000000 00000000ffffffff 0000000000000000
[ 1841.526207] page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
[ 1841.532775] Modules linked in: loop xt_addrtype br_netfilter xt_CHECKSUM iptable_mangle xt_MASQUERADE iptable_nat nf_nat xt_conntrack nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 libcrc32c ipt_REJECT nf_reject_ipv4 tun bridge stp llc overlay ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter dm_mirror dm_region_hash dm_log dm_mod ipmi_ssif intel_rapl_msr intel_rapl_common x86_pkg_temp_thermal coretemp kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel glue_helper crypto_simd cryptd pcspkr i2c_i801 joydev i2c_smbus mei_me mei ipmi_si ipmi_devintf ipmi_msghandler ip_tables ext4 mbcache jbd2 crc32c_intel drm_vram_helper drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm_ttm_helper ttm drm ahci libahci libata virtio_net net_failover failover
[ 1841.603813] CPU: 0 PID: 26221 Comm: userfaultfd Kdump: loaded Tainted: G S  B   W         5.9.0-rc5-00156-g325d0eab4f31 #88
[ 1841.615080] Hardware name: Alibaba X-Dragon CN 01/20G4B, BIOS 1ALSP017 04/09/2019
[ 1841.622701] Call Trace:
[ 1841.625280]  dump_stack+0x8d/0xc0
[ 1841.628726]  bad_page.cold.130+0x63/0x93
[ 1841.632776]  free_pcp_prepare+0x20b/0x250
[ 1841.636908]  free_unref_page+0x18/0x90
[ 1841.640780]  wp_page_copy+0x2cd/0x5d0
[ 1841.644569]  __handle_mm_fault+0x791/0x7a0
[ 1841.648790]  ? handle_mm_fault+0x4b/0x3f0
[ 1841.652925]  handle_mm_fault+0x16f/0x3f0
[ 1841.656968]  exc_page_fault+0x3cb/0x6a0
[ 1841.660920]  ? asm_exc_page_fault+0x8/0x30
[ 1841.665131]  asm_exc_page_fault+0x1e/0x30
[ 1841.669259] RIP: 0033:0x7fd625da6d5d
[ 1841.672957] Code: 00 00 89 77 10 89 f2 49 8d 70 16 4c 89 c7 81 e2 80 00 00 00 e9 04 6c 00 00 0f 1f 40 00 81 e6 80 00 00 00 bf 01 00 00 00 31 c0 <f0> 41 0f b1 38 0f 85 24 01 00 00 64 8b 04 25 d0 02 00 00 41 89 40
[ 1841.691889] RSP: 002b:00007fd6147b2e58 EFLAGS: 00010246
[ 1841.697239] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 00000000000027da
[ 1841.704497] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000001
[ 1841.711759] RBP: 00007fd6147b2f10 R08: 00007fd6181a6000 R09: 00007fd6147b2eb0
[ 1841.719028] R10: 000000000000000f R11: 0000000000000206 R12: 0000000000000000
[ 1841.726286] R13: 0000000001001000 R14: 0000000000000000 R15: 00007fd6147b3700
author
#3 lore raw

Re: a bug on 5.9-rc6

Quoted 3 lines · Yang Shi

Looks like the same problem as
https://lore.kernel.org/linux-mm/[email protected]/

Right, At least they share the same path:

[ 1841.628726]  bad_page.cold.130+0x63/0x93
[ 1841.632776]  free_pcp_prepare+0x20b/0x250
[ 1841.636908]  free_unref_page+0x18/0x90
CC Qian Cai.

Thanks!
Alex

Quoted 59 lines · Alex Shi
Quoted 57 lines · Alex Shi
Quoted 4 lines

I just found a bug on recent upstream kernel on my x86 box.
325d0eab4f31 Merge branch 'akpm' (patches from Andrew)
The last good kernel I known is 5.9-rc2.

This bug still exist on 5.9-rc6.

I found it in doing stress kernel build on swapping, while run kernel selftest's vmtest like:

        for ((i=0;i<8000;i++)); do ./run_vmtests ; sleep 1; done

It take time.
I will try to do bisect when things down in hands.

Thanks
Alex

Quoted 36 lines

Thanks
Alex

Linux aliy8 5.9.0-rc5-00156-g325d0eab4f31 #88 SMP PREEMPT Mon Sep 21 08:48:19 CST 2020 x86_64 x86_64 x86_64 GNU/Linux

1841.488609] BUG: Bad page state in process userfaultfd pfn:1a7d7

[ 1841.494930] page:00000000ec5d2f1e refcount:0 mapcount:0 mapping:0000000000000000 index:0x7fd6181a6 pfn:0x1a7d7
[ 1841.505080] flags: 0xfffffc0080004(uptodate|swapbacked)
[ 1841.510427] raw: 000fffffc0080004 dead000000000100 dead000000000122 0000000000000000
[ 1841.518318] raw: 00000007fd6181a6 0000000000000000 00000000ffffffff 0000000000000000
[ 1841.526207] page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
[ 1841.532775] Modules linked in: loop xt_addrtype br_netfilter xt_CHECKSUM iptable_mangle xt_MASQUERADE iptable_nat nf_nat xt_conntrack nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 libcrc32c ipt_REJECT nf_reject_ipv4 tun bridge stp llc overlay ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter dm_mirror dm_region_hash dm_log dm_mod ipmi_ssif intel_rapl_msr intel_rapl_common x86_pkg_temp_thermal coretemp kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel glue_helper crypto_simd cryptd pcspkr i2c_i801 joydev i2c_smbus mei_me mei ipmi_si ipmi_devintf ipmi_msghandler ip_tables ext4 mbcache jbd2 crc32c_intel drm_vram_helper drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm_ttm_helper ttm drm ahci libahci libata virtio_net net_failover failover
[ 1841.603813] CPU: 0 PID: 26221 Comm: userfaultfd Kdump: loaded Tainted: G S  B   W         5.9.0-rc5-00156-g325d0eab4f31 #88
[ 1841.615080] Hardware name: Alibaba X-Dragon CN 01/20G4B, BIOS 1ALSP017 04/09/2019
[ 1841.622701] Call Trace:
[ 1841.625280]  dump_stack+0x8d/0xc0
[ 1841.628726]  bad_page.cold.130+0x63/0x93
[ 1841.632776]  free_pcp_prepare+0x20b/0x250
[ 1841.636908]  free_unref_page+0x18/0x90
[ 1841.640780]  wp_page_copy+0x2cd/0x5d0
[ 1841.644569]  __handle_mm_fault+0x791/0x7a0
[ 1841.648790]  ? handle_mm_fault+0x4b/0x3f0
[ 1841.652925]  handle_mm_fault+0x16f/0x3f0
[ 1841.656968]  exc_page_fault+0x3cb/0x6a0
[ 1841.660920]  ? asm_exc_page_fault+0x8/0x30
[ 1841.665131]  asm_exc_page_fault+0x1e/0x30
[ 1841.669259] RIP: 0033:0x7fd625da6d5d
[ 1841.672957] Code: 00 00 89 77 10 89 f2 49 8d 70 16 4c 89 c7 81 e2 80 00 00 00 e9 04 6c 00 00 0f 1f 40 00 81 e6 80 00 00 00 bf 01 00 00 00 31 c0 <f0> 41 0f b1 38 0f 85 24 01 00 00 64 8b 04 25 d0 02 00 00 41 89 40
[ 1841.691889] RSP: 002b:00007fd6147b2e58 EFLAGS: 00010246
[ 1841.697239] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 00000000000027da
[ 1841.704497] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000001
[ 1841.711759] RBP: 00007fd6147b2f10 R08: 00007fd6181a6000 R09: 00007fd6147b2eb0
[ 1841.719028] R10: 000000000000000f R11: 0000000000000206 R12: 0000000000000000
[ 1841.726286] R13: 0000000001001000 R14: 0000000000000000 R15: 00007fd6147b3700