RE: [PATCH 3/3] Drivers: hv: hv_balloon: Don't post pressure status from interrupt context

[KY Srinivasan]

> -----Original Message-----
> From: Sitsofe Wheeler [mailto:sitsofe@xxxxxxxxx]
> Sent: Wednesday, December 10, 2014 12:50 PM
> To: KY Srinivasan
> Cc: gregkh@xxxxxxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx;
> devel@xxxxxxxxxxxxxxxxxxxxxx; olaf@xxxxxxxxx; apw@xxxxxxxxxxxxx;
> jasowang@xxxxxxxxxx
> Subject: Re: [PATCH 3/3] Drivers: hv: hv_balloon: Don't post pressure status
> from interrupt context
> 
> On Mon, Dec 08, 2014 at 06:04:35AM +0000, KY Srinivasan wrote:
> >
> > Greg has not committed these patches yet. One of the patches changes
> the balloon floor.
> > This means that the guest will not be ballooned down below the floor.
> > Is this what you are seeing? In our testing we did not see anything
> > unusual other than the floor being elevated (as per the design).
> 
> I applied the following:
> 
> drivers-scsi-storvsc-Fix-a-bug-in-handling-ring-buffer-failures-that-may-
> result-in-I-O-freeze.patch
> V2-1-3-Drivers-hv-hv_balloon-Make-adjustments-in-computing-the-
> floor.patch
> V2-2-3-Drivers-hv-hv_balloon-Fix-a-locking-bug-in-the-balloon-driver.patch
> V2-3-3-Drivers-hv-hv_balloon-Don-t-post-pressure-status-from-interrupt-
> context.patch
> 
> Initially things looked OK but now I'm starting to see the following which is
> rather worrying:
> 
> Dec 10 20:37:11 a kernel: BUG: unable to handle kernel NULL pointer
> dereference at           (null)
> Dec 10 20:37:11 a kernel: IP: [<ffffffff811c30a0>] commit_charge+0x20/0x90
> Dec 10 20:37:11 a kernel: PGD e44cb067 PUD e4495067 PMD 0 Dec 10 20:37:11
> a kernel: Oops: 0000 [#1] SMP DEBUG_PAGEALLOC Dec 10 20:37:11 a kernel:
> CPU: 5 PID: 1490 Comm: ruby Not tainted 3.18.0.x86_64-01967-g86c6a2f-dirty
> #163 Dec 10 20:37:11 a kernel: Hardware name: Microsoft Corporation Virtual
> Machine/Virtual Machine, BIOS 090006  05/23/2012 Dec 10 20:37:11 a kernel:
> task: ffff8800e9bce040 ti: ffff880003890000 task.ti: ffff880003890000 Dec 10
> 20:37:11 a kernel: RIP: 0010:[<ffffffff811c30a0>]  [<ffffffff811c30a0>]
> commit_charge+0x20/0x90 Dec 10 20:37:11 a kernel: RSP:
> 0018:ffff880003893a88  EFLAGS: 00010246 Dec 10 20:37:11 a kernel: RAX:
> 0000000000000000 RBX: ffffea00048d0380 RCX: 0000000000000006 Dec 10
> 20:37:11 a kernel: RDX: 0000000000000480 RSI: ffff880108829bd8 RDI:
> 000000000012340e Dec 10 20:37:11 a kernel: RBP: ffff880003893ac8 R08:
> 0000000000000000 R09: 0000000000000000 Dec 10 20:37:11 a kernel: R10:
> 0000000000000001 R11: 0000000000000000 R12: 0000000000000000 Dec 10
> 20:37:11 a kernel: R13: ffff880108829bd8 R14: ffff880017669c58 R15:
> 0000000000000000 Dec 10 20:37:11 a kernel: FS:  00007f4dc62fa740(0000)
> GS:ffff88010d4a0000(0000) knlGS:0000000000000000 Dec 10 20:37:11 a kernel:
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033 Dec 10 20:37:11 a kernel:
> CR2: 0000000000000000 CR3: 00000000f1459000 CR4: 00000000000406e0 Dec
> 10 20:37:11 a kernel: Stack:
> Dec 10 20:37:11 a kernel:  ffff8800e9bce040 ffffffff816f3950
> 0000000000000000 ffff880017669c58 Dec 10 20:37:11 a kernel:
> ffff880003893ac8 ffffea00048d0380 ffff880108829bd8 0000000000000000 Dec
> 10 20:37:11 a kernel:  ffff880003893af8 ffffffff811c6b36 ffff880003893af8
> ffffea00048d0380 Dec 10 20:37:11 a kernel: Call Trace:
> Dec 10 20:37:11 a kernel:  [<ffffffff816f3950>] ?
> _raw_spin_unlock_irq+0x30/0x50 Dec 10 20:37:11 a kernel:
> [<ffffffff811c6b36>] mem_cgroup_commit_charge+0x76/0x140
> Dec 10 20:37:11 a kernel:  [<ffffffff8115d8d5>]
> __add_to_page_cache_locked+0x1e5/0x2d0
> Dec 10 20:37:11 a kernel:  [<ffffffff8115dfb8>]
> add_to_page_cache_lru+0x28/0x80 Dec 10 20:37:11 a kernel:
> [<ffffffff8115f347>] pagecache_get_page+0x197/0x220 Dec 10 20:37:11 a
> kernel:  [<ffffffff81160cc3>] grab_cache_page_write_begin+0x33/0x50
> Dec 10 20:37:11 a kernel:  [<ffffffff81254fd7>]
> ext4_da_write_begin+0x157/0x340 Dec 10 20:37:11 a kernel:
> [<ffffffff81160da1>] generic_perform_write+0xc1/0x1d0 Dec 10 20:37:11 a
> kernel:  [<ffffffff81161138>] __generic_file_write_iter+0x288/0x340
> Dec 10 20:37:11 a kernel:  [<ffffffff8124a693>]
> ext4_file_write_iter+0x2f3/0x3b0 Dec 10 20:37:11 a kernel:
> [<ffffffff811cde47>] ? vfs_write+0xa7/0x1d0 Dec 10 20:37:11 a kernel:
> [<ffffffff811cdc31>] new_sync_write+0x81/0xb0 Dec 10 20:37:11 a kernel:
> [<ffffffff811cde6b>] vfs_write+0xcb/0x1d0 Dec 10 20:37:11 a kernel:
> [<ffffffff811ce069>] SyS_write+0x49/0xb0 Dec 10 20:37:11 a kernel:
> [<ffffffff816f45a9>] system_call_fastpath+0x12/0x17 Dec 10 20:37:11 a
> kernel: Code: 5d c3 66 0f 1f 84 00 00 00 00 00 66 66 66 66 90 55 48 89 e5 41 55 49
> 89 f5 41 54 41 89 d4 53 48 89 fb 48 83 ec 28 e8 90 3e 00 00 <f6> 00 01 74 1b 48 c7
> c6 e0 f1 9e 81 48 89 df e8 cc 4f fc ff 0f Dec 10 20:37:11 a kernel: RIP
> [<ffffffff811c30a0>] commit_charge+0x20/0x90 Dec 10 20:37:11 a kernel:  RSP
> <ffff880003893a88> Dec 10 20:37:11 a kernel: CR2: 0000000000000000 Dec 10
> 20:37:11 a kernel: BUG: unable to handle kernel Dec 10 20:37:11 a kernel: ---[
> end trace 0ae405bbdfb1f416 ]--- Dec 10 20:37:11 a kernel: NULL pointer
> dereference
> Dec 10 20:37:11 a kernel:  at           (null)
> Dec 10 20:37:11 a kernel: IP: [<ffffffff811c30a0>] commit_charge+0x20/0x90
> Dec 10 20:37:11 a kernel: PGD f17d4067 PUD f1567067 PMD 0 Dec 10 20:37:12
> a kernel: Oops: 0000 [#2] SMP DEBUG_PAGEALLOC
> Dec 10 20:37:12 a kernel: CPU: 2 PID: 25465 Comm: ruby Tainted: G      D
> 3.18.0.x86_64-01967-g86c6a2f-dirty #163
> Dec 10 20:37:12 a kernel: Hardware name: Microsoft Corporation Virtual
> Machine/Virtual Machine, BIOS 090006  05/23/2012 Dec 10 20:37:12 a kernel:
> task: ffff880011a16040 ti: ffff880098754000 task.ti: ffff880098754000 Dec 10
> 20:37:12 a kernel: init_memory_mapping: [mem 0x128000000-0x12fffffff]
> Dec 10 20:37:12 a kernel:  [mem 0x128000000-0x12fffffff] page 4k Dec 10
> 20:37:12 a kernel:  [ffffea0004800000-ffffea00049fffff] PMD ->
> [ffff8800c7400000-ffff8800c75fffff] on node 0 Dec 10 20:37:12 a kernel: RIP:
> 0010:[<ffffffff811c30a0>]  [<ffffffff811c30a0>] commit_charge+0x20/0x90
> Dec 10 20:37:12 a kernel: RSP: 0000:ffff880098757d18  EFLAGS: 00010246 Dec
> 10 20:37:12 a kernel: RAX: 0000000000000000 RBX: ffffea0004915300 RCX:
> 0000000000000001 Dec 10 20:37:12 a kernel: RDX: 0000000000000480 RSI:
> ffff880108829bd8 RDI: 000000000012454c Dec 10 20:37:12 a kernel: RBP:
> ffff880098757d58 R08: 0000000000000006 R09: 0000000000000000 Dec 10
> 20:37:12 a kernel: R10: ffff880011a16040 R11: 0000000000000000 R12:
> 0000000000000000 Dec 10 20:37:12 a kernel: R13: ffff880108829bd8 R14:
> ffff8800f159a5f0 R15: ffff88006b3bc600 Dec 10 20:37:12 a kernel: FS:
> 00007f0836edf700(0000) GS:ffff88010d440000(0000) knlGS:0000000000000000
> Dec 10 20:37:12 a kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> Dec 10 20:37:12 a kernel: CR2: 0000000000000000 CR3: 00000000b8bfd000 CR4:
> 00000000000406e0 Dec 10 20:37:12 a kernel: Stack:
> Dec 10 20:37:12 a kernel:  00000000811bf285 ffff88000723e118
> ffff880108829bd8 ffff88000723e100 Dec 10 20:37:12 a kernel:
> ffffea0004915300 ffffea0004915300 ffff880108829bd8 ffff88000613a280 Dec
> 10 20:37:12 a kernel:  ffff880098757d88 ffffffff811c6b36 ffffffff8118d6fc
> 00007f08200bea58 Dec 10 20:37:12 a kernel: Call Trace:
> Dec 10 20:37:12 a kernel:  [<ffffffff811c6b36>]
> mem_cgroup_commit_charge+0x76/0x140
> Dec 10 20:37:12 a kernel:  [<ffffffff8118d6fc>] ?
> handle_mm_fault+0x62c/0x12a0 Dec 10 20:37:12 a kernel:
> [<ffffffff8118d742>] handle_mm_fault+0x672/0x12a0 Dec 10 20:37:12 a
> kernel:  [<ffffffff81041a13>] ? __do_page_fault+0x1c3/0x4f0 Dec 10 20:37:12
> a kernel:  [<ffffffff81041ce0>] __do_page_fault+0x490/0x4f0 Dec 10 20:37:12
> a kernel:  [<ffffffff810bf2cd>] ? trace_hardirqs_on+0xd/0x10 Dec 10 20:37:12
> a kernel:  [<ffffffff816f3950>] ? _raw_spin_unlock_irq+0x30/0x50 Dec 10
> 20:37:12 a kernel:  [<ffffffff81097a88>] ? finish_task_switch+0x88/0x100 Dec
> 10 20:37:12 a kernel:  [<ffffffff81097a4a>] ? finish_task_switch+0x4a/0x100
> Dec 10 20:37:12 a kernel:  [<ffffffff816ee380>] ? __schedule+0x6a0/0x830
> Dec 10 20:37:12 a kernel:  [<ffffffff813b24ed>] ?
> trace_hardirqs_off_thunk+0x3a/0x3c
> Dec 10 20:37:12 a kernel:  [<ffffffff81041d92>] do_page_fault+0x22/0x30 Dec
> 10 20:37:12 a kernel:  [<ffffffff816f6398>] page_fault+0x28/0x30 Dec 10
> 20:37:12 a kernel: Code: 5d c3 66 0f 1f 84 00 00 00 00 00 66 66 66 66 90 55 48 89
> e5 41 55 49 89 f5 41 54 41 89 d4 53 48 89 fb 48 83 ec 28 e8 90 3e 00 00 <f6> 00
> 01 74 1b 48 c7 c6 e0 f1 9e 81 48 89 df e8 cc 4f fc ff 0f Dec 10 20:37:12 a kernel:
> RIP  [<ffffffff811c30a0>] commit_charge+0x20/0x90 Dec 10 20:37:12 a kernel:
> RSP <ffff880098757d18> Dec 10 20:37:12 a kernel: CR2: 0000000000000000 Dec
> 10 20:37:12 a kernel: ---[ end trace 0ae405bbdfb1f417 ]--- Dec 10 20:37:12 a
> kernel: BUG: sleeping function called from invalid context at
> kernel/locking/rwsem.c:41 Dec 10 20:37:12 a kernel: in_atomic(): 1,
> irqs_disabled(): 1, pid: 25465, name: ruby Dec 10 20:37:12 a kernel: INFO:
> lockdep is turned off.
> Dec 10 20:37:12 a kernel: irq event stamp: 2431342 Dec 10 20:37:12 a kernel:
> hardirqs last  enabled at (2431341): [<ffffffff816f38fd>]
> _raw_spin_unlock_irqrestore+0x4d/0x70
> Dec 10 20:37:12 a kernel: hardirqs last disabled at (2431342):
> [<ffffffff816f37dd>] _raw_spin_lock_irq+0x1d/0x60 Dec 10 20:37:12 a kernel:
> softirqs last  enabled at (2431322): [<ffffffff81078458>]
> __do_softirq+0x298/0x340 Dec 10 20:37:12 a kernel: softirqs last disabled at
> (2431317): [<ffffffff810787c8>] irq_exit+0x58/0xc0
> Dec 10 20:37:12 a kernel: CPU: 2 PID: 25465 Comm: ruby Tainted: G      D
> 3.18.0.x86_64-01967-g86c6a2f-dirty #163
> Dec 10 20:37:12 a kernel: Hardware name: Microsoft Corporation Virtual
> Machine/Virtual Machine, BIOS 090006  05/23/2012 Dec 10 20:37:12 a kernel:
> 0000000000000029 ffff8800987578f8 ffffffff816ea99f 0000000000000000 Dec
> 10 20:37:12 a kernel:  ffff880011a16040 ffff880098757918 ffffffff810a2dc5
> ffff880098757948 Dec 10 20:37:12 a kernel:  ffffffff819d796f ffff880098757948
> ffffffff810a2e46 ffffffff82b828c2 Dec 10 20:37:12 a kernel: Call Trace:
> Dec 10 20:37:12 a kernel:  [<ffffffff816ea99f>] dump_stack+0x4e/0x68 Dec 10
> 20:37:12 a kernel:  [<ffffffff810a2dc5>] ___might_sleep+0x115/0x120 Dec 10
> 20:37:12 a kernel:  [<ffffffff810a2e46>] __might_sleep+0x76/0xa0 Dec 10
> 20:37:12 a kernel:  [<ffffffff816f1f04>] down_read+0x24/0x70 Dec 10 20:37:12
> a kernel:  [<ffffffff81082de4>] exit_signals+0x24/0x140 Dec 10 20:37:12 a
> kernel:  [<ffffffff81076714>] do_exit+0x134/0xa80 Dec 10 20:37:12 a kernel:
> [<ffffffff810cb8cc>] ? kmsg_dump+0xfc/0x110 Dec 10 20:37:12 a kernel:
> [<ffffffff810cb7f5>] ? kmsg_dump+0x25/0x110 Dec 10 20:37:12 a kernel:
> [<ffffffff810064e8>] oops_end+0xa8/0xc0 Dec 10 20:37:12 a kernel:
> [<ffffffff816e53bc>] no_context+0x319/0x362 Dec 10 20:37:12 a kernel:
> [<ffffffff816e55d0>] __bad_area_nosemaphore+0x1cb/0x1ea
> Dec 10 20:37:12 a kernel:  [<ffffffff816e5602>]
> bad_area_nosemaphore+0x13/0x15 Dec 10 20:37:12 a kernel:
> [<ffffffff81041a3e>] __do_page_fault+0x1ee/0x4f0 Dec 10 20:37:12 a kernel:
> [<ffffffff811680d5>] ? __alloc_pages_nodemask+0x225/0xaf0
> Dec 10 20:37:12 a kernel:  [<ffffffff813b24ed>] ?
> trace_hardirqs_off_thunk+0x3a/0x3c
> Dec 10 20:37:12 a kernel:  [<ffffffff81041d92>] do_page_fault+0x22/0x30 Dec
> 10 20:37:12 a kernel:  [<ffffffff816f6398>] page_fault+0x28/0x30 Dec 10
> 20:37:12 a kernel:  [<ffffffff811c30a0>] ? commit_charge+0x20/0x90 Dec 10
> 20:37:12 a kernel:  [<ffffffff811c30a0>] ? commit_charge+0x20/0x90 Dec 10
> 20:37:12 a kernel:  [<ffffffff811c6b36>]
> mem_cgroup_commit_charge+0x76/0x140
> Dec 10 20:37:12 a kernel:  [<ffffffff8118d6fc>] ?
> handle_mm_fault+0x62c/0x12a0 Dec 10 20:37:12 a kernel:
> [<ffffffff8118d742>] handle_mm_fault+0x672/0x12a0 Dec 10 20:37:12 a
> kernel:  [<ffffffff81041a13>] ? __do_page_fault+0x1c3/0x4f0 Dec 10 20:37:12
> a kernel:  [<ffffffff81041ce0>] __do_page_fault+0x490/0x4f0 Dec 10 20:37:12
> a kernel:  [<ffffffff810bf2cd>] ? trace_hardirqs_on+0xd/0x10 Dec 10 20:37:12
> a kernel:  [<ffffffff816f3950>] ? _raw_spin_unlock_irq+0x30/0x50 Dec 10
> 20:37:12 a kernel:  [<ffffffff81097a88>] ? finish_task_switch+0x88/0x100 Dec
> 10 20:37:12 a kernel:  [<ffffffff81097a4a>] ? finish_task_switch+0x4a/0x100
> Dec 10 20:37:12 a kernel:  [<ffffffff816ee380>] ? __schedule+0x6a0/0x830
> Dec 10 20:37:12 a kernel:  [<ffffffff813b24ed>] ?
> trace_hardirqs_off_thunk+0x3a/0x3c
> Dec 10 20:37:12 a kernel:  [<ffffffff81041d92>] do_page_fault+0x22/0x30 Dec
> 10 20:37:12 a kernel:  [<ffffffff816f6398>] page_fault+0x28/0x30 Dec 10
> 20:37:12 a kernel: note: ruby[25465] exited with preempt_count 1 Dec 10
> 20:37:16 a kernel: init_memory_mapping: [mem 0x130000000-0x137ffffff]
> Dec 10 20:37:16 a kernel:  [mem 0x130000000-0x137ffffff] page 4k Dec 10
> 20:37:16 a kernel:  [ffffea0004a00000-ffffea0004bfffff] PMD ->
> [ffff880093200000-ffff8800933fffff] on node 0
> Dec 10 20:37:17 a kernel: BUG: unable to handle kernel NULL pointer
> dereference at           (null)
> Dec 10 20:37:17 a kernel: IP: [<ffffffff811c30a0>] commit_charge+0x20/0x90
> 
> Are these Hyper-V related?

I cannot see how the patches that you have applied can cause this problem. What do you need
to run to trigger this problem. Do you see this issue even without these patches.

Regards,

K. Y 
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/