Re: Add Morton,Peter and David for discussion//Re: [PATCH -next] uprobes: fix two zero old_folio bugs in __replace_page()

[David Hildenbrand <david@redhat.com>]

On 18.02.25 03:47, Tong Tiangen wrote:
> 
> 
> 在 2025/2/17 20:38, Tong Tiangen 写道:
>> We triggered the following error logs in syzkaller test:
>>
>>     BUG: Bad page state in process syz.7.38  pfn:1eff3
>>     page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x1eff3
>>     flags: 0x3fffff00004004(referenced|reserved|node=0|zone=1|lastcpupid=0x1fffff)
>>     raw: 003fffff00004004 ffffe6c6c07bfcc8 ffffe6c6c07bfcc8 0000000000000000
>>     raw: 0000000000000000 0000000000000000 00000000fffffffe 0000000000000000
>>     page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
>>     Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1.1 04/01/2014
>>     Call Trace:
>>      <TASK>
>>      dump_stack_lvl+0x32/0x50
>>      bad_page+0x69/0xf0
>>      free_unref_page_prepare+0x401/0x500
>>      free_unref_page+0x6d/0x1b0
>>      uprobe_write_opcode+0x460/0x8e0
>>      install_breakpoint.part.0+0x51/0x80
>>      register_for_each_vma+0x1d9/0x2b0
>>      __uprobe_register+0x245/0x300
>>      bpf_uprobe_multi_link_attach+0x29b/0x4f0
>>      link_create+0x1e2/0x280
>>      __sys_bpf+0x75f/0xac0
>>      __x64_sys_bpf+0x1a/0x30
>>      do_syscall_64+0x56/0x100
>>      entry_SYSCALL_64_after_hwframe+0x78/0xe2
>>
>>      BUG: Bad rss-counter state mm:00000000452453e0 type:MM_FILEPAGES val:-1
>>
>> The following syzkaller test case can be used to reproduce:
>>
>>     r2 = creat(&(0x7f0000000000)='./file0\x00', 0x8)
>>     write$nbd(r2, &(0x7f0000000580)=ANY=[], 0x10)
>>     r4 = openat(0xffffffffffffff9c, &(0x7f0000000040)='./file0\x00', 0x42, 0x0)
>>     mmap$IORING_OFF_SQ_RING(&(0x7f0000ffd000/0x3000)=nil, 0x3000, 0x0, 0x12, r4, 0x0)
>>     r5 = userfaultfd(0x80801)
>>     ioctl$UFFDIO_API(r5, 0xc018aa3f, &(0x7f0000000040)={0xaa, 0x20})
>>     r6 = userfaultfd(0x80801)
>>     ioctl$UFFDIO_API(r6, 0xc018aa3f, &(0x7f0000000140))
>>     ioctl$UFFDIO_REGISTER(r6, 0xc020aa00, &(0x7f0000000100)={{&(0x7f0000ffc000/0x4000)=nil, 0x4000}, 0x2})
>>     ioctl$UFFDIO_ZEROPAGE(r5, 0xc020aa04, &(0x7f0000000000)={{&(0x7f0000ffd000/0x1000)=nil, 0x1000}})
>>     r7 = bpf$PROG_LOAD(0x5, &(0x7f0000000140)={0x2, 0x3, &(0x7f0000000200)=ANY=[@ANYBLOB="1800000000120000000000000000000095"], &(0x7f0000000000)='GPL\x00', 0x7, 0x0, 0x0, 0x0, 0x0, '\x00', 0x0, @fallback=0x30, 0xffffffffffffffff, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x10, 0x0, @void, @value}, 0x94)
>>     bpf$BPF_LINK_CREATE_XDP(0x1c, &(0x7f0000000040)={r7, 0x0, 0x30, 0x1e, @val=@uprobe_multi={&(0x7f0000000080)='./file0\x00', &(0x7f0000000100)=[0x2], 0x0, 0x0, 0x1}}, 0x40)
>>
>> The cause is that zero pfn is set to the pte without increasing the rss
>> count in mfill_atomic_pte_zeropage() and the refcount of zero folio does
>> not increase accordingly. Then, the operation on the same pfn is performed
>> in uprobe_write_opcode()->__replace_page() to unconditional decrease the
>> rss count and old_folio's refcount.
>>
>> Therefore, two bugs are introduced:
>> 1. The rss count is incorrect, when process exit, the check_mm() report
>>      error "Bad rss-count".
>> 2. The reserved folio (zero folio) is freed when folio->refcount is zero,
>>      then free_pages_prepare->free_page_is_bad() report error "Bad page state".
>>
>> To fix it, add zero folio check before rss counter and refcount decrease.
>>
>> Fixes: 7396fa818d62 ("uprobes/core: Make background page replacement logic account for rss_stat counters")
>> Fixes: 2b1444983508 ("uprobes, mm, x86: Add the ability to install and remove uprobes breakpoints")
>> Signed-off-by: Tong Tiangen <tongtiangen@huawei.com>
>> ---
>>    kernel/events/uprobes.c | 6 ++++--
>>    1 file changed, 4 insertions(+), 2 deletions(-)
>>
>> diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
>> index 46ddf3a2334d..ff5694acfa68 100644
>> --- a/kernel/events/uprobes.c
>> +++ b/kernel/events/uprobes.c
>> @@ -213,7 +213,8 @@ static int __replace_page(struct vm_area_struct *vma, unsigned long addr,
>>    		dec_mm_counter(mm, MM_ANONPAGES);
>>    
>>    	if (!folio_test_anon(old_folio)) {
>> -		dec_mm_counter(mm, mm_counter_file(old_folio));
>> +		if (!is_zero_folio(old_folio))
>> +			dec_mm_counter(mm, mm_counter_file(old_folio));
>>    		inc_mm_counter(mm, MM_ANONPAGES);
>>    	}
>>    
>> @@ -227,7 +228,8 @@ static int __replace_page(struct vm_area_struct *vma, unsigned long addr,
>>    	if (!folio_mapped(old_folio))
>>    		folio_free_swap(old_folio);
>>    	page_vma_mapped_walk_done(&pvmw);
>> -	folio_put(old_folio);
>> +	if (!is_zero_folio(old_folio))
>> +		folio_put(old_folio);
 >>    >>    	err = 0;
>>     unlock:
> 

The whole "manually replace pages" logic is fragile. I tried to rewrite 
it a while back:

https://lkml.kernel.org/r/20240604122548.359952-1-david@redhat.com

But didn't get to follow-up yet.

I'm not sure if the page_vma_mapped_walk() really does what we would 
expect here.

The folio_remove_rmap_pte(old_folio, old_page, vma); is certainly wrong 
as well for ero folios.


I don't think there is a sane use case right now where we would hit the 
shared zeropage.

So for the time being, I think we should just reject them immediately 
after get_user_page_vma_remote().

At some point I'll follow up with my rewrite that will clean this 
nastiness here up a bit.

-- 
Cheers,

David / dhildenb