Re: Kernel 4.17.4 lockup

From: Dave Hansen
Date: Mon Jul 09 2018 - 20:45:27 EST

... cc'ing a few folks who I know have been looking at this code
lately. The full oops is below if any of you want to take a look.

OK, well, annotating the disassembly a bit:

> (gdb) disass free_pages_and_swap_cache
> Dump of assembler code for function free_pages_and_swap_cache:
> 0xffffffff8124c0d0 <+0>: callq 0xffffffff81a017a0 <__fentry__>
> 0xffffffff8124c0d5 <+5>: push %r14
> 0xffffffff8124c0d7 <+7>: push %r13
> 0xffffffff8124c0d9 <+9>: push %r12
> 0xffffffff8124c0db <+11>: mov %rdi,%r12 // %r12 = pages
> 0xffffffff8124c0de <+14>: push %rbp
> 0xffffffff8124c0df <+15>: mov %esi,%ebp // %ebp = nr
> 0xffffffff8124c0e1 <+17>: push %rbx
> 0xffffffff8124c0e2 <+18>: callq 0xffffffff81205a10 <lru_add_drain>
> 0xffffffff8124c0e7 <+23>: test %ebp,%ebp // test nr==0
> 0xffffffff8124c0e9 <+25>: jle 0xffffffff8124c156 <free_pages_and_swap_cache+134>
> 0xffffffff8124c0eb <+27>: lea -0x1(%rbp),%eax
> 0xffffffff8124c0ee <+30>: mov %r12,%rbx // %rbx = pages
> 0xffffffff8124c0f1 <+33>: lea 0x8(%r12,%rax,8),%r14 // load &pages[nr] into %r14?
> 0xffffffff8124c0f6 <+38>: mov (%rbx),%r13 // %r13 = pages[i]
> 0xffffffff8124c0f9 <+41>: mov 0x20(%r13),%rdx //<<<<<<<<<<<<<<<<<<<< GPF here.
%r13 is 64-byte aligned, so looks like a halfway reasonable 'struct page *'.

%R14 looks OK (0xffff93d4abb5f000) because it points to the end of a
dynamically-allocated (not on-stack) mmu_gather_batch page. %RBX is
pointing 50 pages up from the start of the previous page. That makes it
the 48th page in pages[] after a pointer and two integers in the
beginning of the structure. That 48 is important because it's way
larger than the on-stack size of 8.

It's hard to make much sense of %R13 (pages[48] / 0xfffbf0809e304bc0)
because the vmemmap addresses get randomized. But, I _think_ that's too
high of an address for a 4-level paging vmemmap[] entry. Does anybody
else know offhand?

I'd really want to see this reproduced without KASLR to make the oops
easier to read. It would also be handy to try your workload with all
the pedantic debugging: KASAN, slab debugging, DEBUG_PAGE_ALLOC, etc...
and see if it still triggers.

Some relevant functions and structures below for reference.

void free_pages_and_swap_cache(struct page **pages, int nr)
{
for (i = 0; i < nr; i++)
free_swap_cache(pages[i]);
}


static void tlb_flush_mmu_free(struct mmu_gather *tlb)
{
for (batch = &tlb->local; batch && batch->nr;
batch = batch->next) {
free_pages_and_swap_cache(batch->pages, batch->nr);
}

zap_pte_range()
{
if (force_flush)
tlb_flush_mmu_free(tlb);
}

... all the way up to the on-stack-allocated mmu_gather:

void zap_page_range(struct vm_area_struct *vma, unsigned long start,
unsigned long size)
{
struct mmu_gather tlb;


#define MMU_GATHER_BUNDLE 8

struct mmu_gather {
...
struct mmu_gather_batch local;
struct page *__pages[MMU_GATHER_BUNDLE];
}

struct mmu_gather_batch {
struct mmu_gather_batch *next;
unsigned int nr;
unsigned int max;
struct page *pages[0];
};

#define MAX_GATHER_BATCH \
((PAGE_SIZE - sizeof(struct mmu_gather_batch)) / sizeof(void *))



> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: general protection
> fault: 0000 [#1] SMP PTI
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: Modules linked in:
> rpcsec_gss_krb5 nfsv4 dns_resolver nfs fscache devlink ebtable_filter
> ebtables ip6table_filter ip6_tables intel_rapl x86_pkg_temp_thermal
> intel_powerclamp coretemp snd_hda_codec_hdmi snd_hda_codec_realtek
> kvm_intel snd_hda_codec_generic snd_hda_intel kvm snd_hda_codec
> snd_hda_core snd_hwdep irqbypass crct10dif_pclmul crc32_pclmul snd_seq
> mei_wdt ghash_clmulni_intel snd_seq_device intel_cstate ppdev
> intel_uncore iTCO_wdt gpio_ich iTCO_vendor_support snd_pcm
> intel_rapl_perf snd_timer snd mei_me parport_pc joydev i2c_i801 mei
> soundcore shpchp lpc_ich parport nfsd auth_rpcgss nfs_acl lockd grace
> sunrpc i915 i2c_algo_bit drm_kms_helper r8169 drm crc32c_intel mii
> video
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: CPU: 7 PID: 7093 Comm:
> cc1 Not tainted 4.17.4-200.0.fc28.x86_64 #1
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: Hardware name: Gigabyte
> Technology Co., Ltd. H87M-D3H/H87M-D3H, BIOS F11 08/18/2015
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RIP: 0010:free_pages_and_swap_cache+0x29/0xb0
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RSP: 0018:ffffb2cd83ffbd58 EFLAGS: 00010202
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RAX: 0017fffe00040068 RBX: ffff93d4abb5ec80 RCX: 0000000000000000
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RDX: 0017fffe00040068 RSI: 00000000000001fe RDI: ffff93d51e3dd2a0
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RBP: 00000000000001fe R08: fffff0809df82d20 R09: ffff93d51e5d5000
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: R10: ffff93d51e5d5e20 R11: ffff93d51e5d5d00 R12: ffff93d4abb5e010
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: R13: fffbf0809e304bc0 R14: ffff93d4abb5f000 R15: ffff93d4cbcee8f0
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: FS: 0000000000000000(0000) GS:ffff93d51e3c0000(0000) knlGS:0000000000000000
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: CR2: 00007ffb255e753c CR3: 00000005e820a002 CR4: 00000000001606e0
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: Call Trace:
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: tlb_flush_mmu_free+0x31/0x50
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: arch_tlb_finish_mmu+0x42/0x70
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: tlb_finish_mmu+0x1f/0x30
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: exit_mmap+0xca/0x190
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: mmput+0x5f/0x130
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: do_exit+0x280/0xae0
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: ? __do_page_fault+0x263/0x4e0
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: do_group_exit+0x3a/0xa0
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: __x64_sys_exit_group+0x14/0x20
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: do_syscall_64+0x65/0x160
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: entry_SYSCALL_64_after_hwframe+0x44/0xa9
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RIP: 0033:0x7ffb2542b3c6
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RSP: 002b:00007ffd9e7e33b8 EFLAGS: 00000246 ORIG_RAX: 00000000000000e7
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RAX: ffffffffffffffda RBX: 00007ffb2551c740 RCX: 00007ffb2542b3c6
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RDX: 0000000000000000 RSI: 000000000000003c RDI: 0000000000000000
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RBP: 0000000000000000 R08: 00000000000000e7 R09: fffffffffffffe70
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: R10: 00007ffd9e7e3250 R11: 0000000000000246 R12: 00007ffb2551c740
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: R13: 0000000000000037 R14: 00007ffb25525708 R15: 0000000000000000
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: Code: 40 00 0f 1f 44 00 00 41 56 41 55 41 54 49 89 fc 55 89 f5 53 e8 29 99 fb ff 85 ed 7e 6b 8d 45 ff 4c 89 e3 4d 8d 74 c4 08 4c 8b 2b <49> 8b 55 20 48 8d 42 ff 83 e2 01 49 0f 44 c5 48 8b 48 20 48 8d
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: RIP: free_pages_and_swap_cache+0x29/0xb0 RSP: ffffb2cd83ffbd58
> Jul 05 14:33:32 gnu-hsw-1.sc.intel.com kernel: ---[ end trace 5960277fd8a3c0b5 ]---