diff mbox series

[1/4] erofs: fix setting up pcluster for temporary pages

Message ID 20201022145724.27284-1-hsiangkao@aol.com
State New
Headers show
Series [1/4] erofs: fix setting up pcluster for temporary pages | expand

Commit Message

Gao Xiang Oct. 22, 2020, 2:57 p.m. UTC
From: Gao Xiang <hsiangkao@redhat.com>

pcluster should be only set up for all managed pages instead of
temporary pages. Since it currently uses page->mapping to identify,
the impact is minor for now.

Fixes: 5ddcee1f3a1c ("erofs: get rid of __stagingpage_alloc helper")
Cc: <stable@vger.kernel.org> # 5.5+
Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
---
 fs/erofs/zdata.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

Comments

Vladimir Zapolskiy Oct. 30, 2020, 12:20 p.m. UTC | #1
Hello Gao Xiang,

On 10/22/20 5:57 PM, Gao Xiang via Linux-erofs wrote:
> From: Gao Xiang <hsiangkao@redhat.com>

> 

> pcluster should be only set up for all managed pages instead of

> temporary pages. Since it currently uses page->mapping to identify,

> the impact is minor for now.

> 

> Fixes: 5ddcee1f3a1c ("erofs: get rid of __stagingpage_alloc helper")

> Cc: <stable@vger.kernel.org> # 5.5+

> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>


I was looking exactly at this problem recently, my change is one-to-one
to your fix, thus I can provide a tag:

Tested-by: Vladimir Zapolskiy <vladimir@tuxera.com>



The fixed problem is minor, but the kernel log becomes polluted, if
a page allocation debug option is enabled:

     % md5sum ~/erofs/testfile
     BUG: Bad page state in process kworker/u9:0  pfn:687de
     page:0000000057b8bcb4 refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x687de
     flags: 0x4000000000002000(private)
     raw: 4000000000002000 dead000000000100 dead000000000122 0000000000000000
     raw: 0000000000000000 ffff888066758690 00000000ffffffff 0000000000000000
     page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
     Modules linked in:
     CPU: 1 PID: 602 Comm: kworker/u9:0 Not tainted 5.9.1 #2
     Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1 04/01/2014
     Workqueue: erofs_unzipd z_erofs_decompressqueue_work
     Call Trace:
      dump_stack+0x84/0xba
      bad_page.cold+0xac/0xb1
      check_free_page_bad+0xb0/0xc0
      free_pcp_prepare+0x2c8/0x2d0
      free_unref_page+0x18/0xf0
      put_pages_list+0x11a/0x120
      z_erofs_decompressqueue_work+0xc9/0x110
      ? z_erofs_decompress_pcluster.isra.0+0xf10/0xf10
      ? read_word_at_a_time+0x12/0x20
      ? strscpy+0xc7/0x1a0
      process_one_work+0x30c/0x730
      worker_thread+0x91/0x640
      ? __kasan_check_read+0x11/0x20
      ? rescuer_thread+0x8a0/0x8a0
      kthread+0x1dd/0x200
      ? kthread_unpark+0xa0/0xa0
      ret_from_fork+0x1f/0x30
     Disabling lock debugging due to kernel taint

--
Best wishes,
Vladimir
Gao Xiang Oct. 30, 2020, 12:47 p.m. UTC | #2
Hi Vladimir,

On Fri, Oct 30, 2020 at 02:20:31PM +0200, Vladimir Zapolskiy wrote:
> Hello Gao Xiang,
> 
> On 10/22/20 5:57 PM, Gao Xiang via Linux-erofs wrote:
> > From: Gao Xiang <hsiangkao@redhat.com>
> > 
> > pcluster should be only set up for all managed pages instead of
> > temporary pages. Since it currently uses page->mapping to identify,
> > the impact is minor for now.
> > 
> > Fixes: 5ddcee1f3a1c ("erofs: get rid of __stagingpage_alloc helper")
> > Cc: <stable@vger.kernel.org> # 5.5+
> > Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
> 
> I was looking exactly at this problem recently, my change is one-to-one
> to your fix, thus I can provide a tag:
> 
> Tested-by: Vladimir Zapolskiy <vladimir@tuxera.com>

Many thanks for confirming this!
I found this when I was killing magical stagingpage page->mapping,
it's somewhat late :-)

> 
> 
> The fixed problem is minor, but the kernel log becomes polluted, if
> a page allocation debug option is enabled:
> 
>     % md5sum ~/erofs/testfile
>     BUG: Bad page state in process kworker/u9:0  pfn:687de
>     page:0000000057b8bcb4 refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x687de
>     flags: 0x4000000000002000(private)
>     raw: 4000000000002000 dead000000000100 dead000000000122 0000000000000000
>     raw: 0000000000000000 ffff888066758690 00000000ffffffff 0000000000000000
>     page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
>     Modules linked in:
>     CPU: 1 PID: 602 Comm: kworker/u9:0 Not tainted 5.9.1 #2
>     Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1 04/01/2014
>     Workqueue: erofs_unzipd z_erofs_decompressqueue_work
>     Call Trace:
>      dump_stack+0x84/0xba
>      bad_page.cold+0xac/0xb1
>      check_free_page_bad+0xb0/0xc0
>      free_pcp_prepare+0x2c8/0x2d0
>      free_unref_page+0x18/0xf0
>      put_pages_list+0x11a/0x120
>      z_erofs_decompressqueue_work+0xc9/0x110
>      ? z_erofs_decompress_pcluster.isra.0+0xf10/0xf10
>      ? read_word_at_a_time+0x12/0x20
>      ? strscpy+0xc7/0x1a0
>      process_one_work+0x30c/0x730
>      worker_thread+0x91/0x640
>      ? __kasan_check_read+0x11/0x20
>      ? rescuer_thread+0x8a0/0x8a0
>      kthread+0x1dd/0x200
>      ? kthread_unpark+0xa0/0xa0
>      ret_from_fork+0x1f/0x30
>     Disabling lock debugging due to kernel taint

Yeah, I can make a pull-request to Linus if you need this to be in master
now, or I can post it for v5.11-rc1 since 5.4 LTS isn't effected (and it
would be only a print problem with debugging option.)

Thanks,
Gao Xiang

> 
> --
> Best wishes,
> Vladimir
>
Vladimir Zapolskiy Oct. 30, 2020, 1:32 p.m. UTC | #3
Hi Gao Xiang,

On 10/30/20 2:47 PM, Gao Xiang wrote:
> Hi Vladimir,
> 
> On Fri, Oct 30, 2020 at 02:20:31PM +0200, Vladimir Zapolskiy wrote:
>> Hello Gao Xiang,
>>
>> On 10/22/20 5:57 PM, Gao Xiang via Linux-erofs wrote:
>>> From: Gao Xiang <hsiangkao@redhat.com>
>>>
>>> pcluster should be only set up for all managed pages instead of
>>> temporary pages. Since it currently uses page->mapping to identify,
>>> the impact is minor for now.
>>>
>>> Fixes: 5ddcee1f3a1c ("erofs: get rid of __stagingpage_alloc helper")
>>> Cc: <stable@vger.kernel.org> # 5.5+
>>> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
>>
>> I was looking exactly at this problem recently, my change is one-to-one
>> to your fix, thus I can provide a tag:
>>
>> Tested-by: Vladimir Zapolskiy <vladimir@tuxera.com>
> 
> Many thanks for confirming this!
> I found this when I was killing magical stagingpage page->mapping,
> it's somewhat late :-)
> 

sure, for me it was an exciting immersion into the filesystem code :)

>>
>>
>> The fixed problem is minor, but the kernel log becomes polluted, if
>> a page allocation debug option is enabled:
>>
>>      % md5sum ~/erofs/testfile
>>      BUG: Bad page state in process kworker/u9:0  pfn:687de
>>      page:0000000057b8bcb4 refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x687de
>>      flags: 0x4000000000002000(private)
>>      raw: 4000000000002000 dead000000000100 dead000000000122 0000000000000000
>>      raw: 0000000000000000 ffff888066758690 00000000ffffffff 0000000000000000
>>      page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
>>      Modules linked in:
>>      CPU: 1 PID: 602 Comm: kworker/u9:0 Not tainted 5.9.1 #2
>>      Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1 04/01/2014
>>      Workqueue: erofs_unzipd z_erofs_decompressqueue_work
>>      Call Trace:
>>       dump_stack+0x84/0xba
>>       bad_page.cold+0xac/0xb1
>>       check_free_page_bad+0xb0/0xc0
>>       free_pcp_prepare+0x2c8/0x2d0
>>       free_unref_page+0x18/0xf0
>>       put_pages_list+0x11a/0x120
>>       z_erofs_decompressqueue_work+0xc9/0x110
>>       ? z_erofs_decompress_pcluster.isra.0+0xf10/0xf10
>>       ? read_word_at_a_time+0x12/0x20
>>       ? strscpy+0xc7/0x1a0
>>       process_one_work+0x30c/0x730
>>       worker_thread+0x91/0x640
>>       ? __kasan_check_read+0x11/0x20
>>       ? rescuer_thread+0x8a0/0x8a0
>>       kthread+0x1dd/0x200
>>       ? kthread_unpark+0xa0/0xa0
>>       ret_from_fork+0x1f/0x30
>>      Disabling lock debugging due to kernel taint
> 
> Yeah, I can make a pull-request to Linus if you need this to be in master
> now, or I can post it for v5.11-rc1 since 5.4 LTS isn't effected (and it
> would be only a print problem with debugging option.)
> 

As for myself I don't utterly need this fix on the master branch ASAP, however
it might be reasonable to get it included right into the next v5.10 release,
because I believe it'll be an LTS. Eventually it's up to you to make a decision,
from my side I won't urge you, the fixed issue is obviously a non-critical one.

Thank you for the original fix and taking my opinion into consideration :)

--
Best wishes,
Vladimir
Gao Xiang Oct. 30, 2020, 2:10 p.m. UTC | #4
On Fri, Oct 30, 2020 at 03:32:55PM +0200, Vladimir Zapolskiy wrote:
> Hi Gao Xiang,
> 
> On 10/30/20 2:47 PM, Gao Xiang wrote:
> > Hi Vladimir,
> > 
> > On Fri, Oct 30, 2020 at 02:20:31PM +0200, Vladimir Zapolskiy wrote:
> > > Hello Gao Xiang,
> > > 
> > > On 10/22/20 5:57 PM, Gao Xiang via Linux-erofs wrote:
> > > > From: Gao Xiang <hsiangkao@redhat.com>
> > > > 
> > > > pcluster should be only set up for all managed pages instead of
> > > > temporary pages. Since it currently uses page->mapping to identify,
> > > > the impact is minor for now.
> > > > 
> > > > Fixes: 5ddcee1f3a1c ("erofs: get rid of __stagingpage_alloc helper")
> > > > Cc: <stable@vger.kernel.org> # 5.5+
> > > > Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
> > > 
> > > I was looking exactly at this problem recently, my change is one-to-one
> > > to your fix, thus I can provide a tag:
> > > 
> > > Tested-by: Vladimir Zapolskiy <vladimir@tuxera.com>
> > 
> > Many thanks for confirming this!
> > I found this when I was killing magical stagingpage page->mapping,
> > it's somewhat late :-)
> > 
> 
> sure, for me it was an exciting immersion into the filesystem code :)

Thanks for your effort on this!

You could also post related kernel message in advance and
I will definitly look into that as well. :)

> 
> > > 
> > > 
> > > The fixed problem is minor, but the kernel log becomes polluted, if
> > > a page allocation debug option is enabled:
> > > 
> > >      % md5sum ~/erofs/testfile
> > >      BUG: Bad page state in process kworker/u9:0  pfn:687de
> > >      page:0000000057b8bcb4 refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x687de
> > >      flags: 0x4000000000002000(private)
> > >      raw: 4000000000002000 dead000000000100 dead000000000122 0000000000000000
> > >      raw: 0000000000000000 ffff888066758690 00000000ffffffff 0000000000000000
> > >      page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
> > >      Modules linked in:
> > >      CPU: 1 PID: 602 Comm: kworker/u9:0 Not tainted 5.9.1 #2
> > >      Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1 04/01/2014
> > >      Workqueue: erofs_unzipd z_erofs_decompressqueue_work
> > >      Call Trace:
> > >       dump_stack+0x84/0xba
> > >       bad_page.cold+0xac/0xb1
> > >       check_free_page_bad+0xb0/0xc0
> > >       free_pcp_prepare+0x2c8/0x2d0
> > >       free_unref_page+0x18/0xf0
> > >       put_pages_list+0x11a/0x120
> > >       z_erofs_decompressqueue_work+0xc9/0x110
> > >       ? z_erofs_decompress_pcluster.isra.0+0xf10/0xf10
> > >       ? read_word_at_a_time+0x12/0x20
> > >       ? strscpy+0xc7/0x1a0
> > >       process_one_work+0x30c/0x730
> > >       worker_thread+0x91/0x640
> > >       ? __kasan_check_read+0x11/0x20
> > >       ? rescuer_thread+0x8a0/0x8a0
> > >       kthread+0x1dd/0x200
> > >       ? kthread_unpark+0xa0/0xa0
> > >       ret_from_fork+0x1f/0x30
> > >      Disabling lock debugging due to kernel taint
> > 
> > Yeah, I can make a pull-request to Linus if you need this to be in master
> > now, or I can post it for v5.11-rc1 since 5.4 LTS isn't effected (and it
> > would be only a print problem with debugging option.)
> > 
> 
> As for myself I don't utterly need this fix on the master branch ASAP, however
> it might be reasonable to get it included right into the next v5.10 release,
> because I believe it'll be an LTS. Eventually it's up to you to make a decision,
> from my side I won't urge you, the fixed issue is obviously a non-critical one.
> 
> Thank you for the original fix and taking my opinion into consideration :)

Yeah, v5.10 is a LTS version, and you are right, I will try to make a
pull-request after I get Chao's RVB.

Thanks,
Gao Xiang

> 
> --
> Best wishes,
> Vladimir
>
Chao Yu Nov. 4, 2020, 1:05 a.m. UTC | #5
On 2020/10/22 22:57, Gao Xiang wrote:
> From: Gao Xiang <hsiangkao@redhat.com>
> 
> pcluster should be only set up for all managed pages instead of
> temporary pages. Since it currently uses page->mapping to identify,
> the impact is minor for now.
> 
> Fixes: 5ddcee1f3a1c ("erofs: get rid of __stagingpage_alloc helper")
> Cc: <stable@vger.kernel.org> # 5.5+
> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>

Reviewed-by: Chao Yu <yuchao0@huawei.com>

Thanks,
Gao Xiang Nov. 4, 2020, 1:11 a.m. UTC | #6
On Wed, Nov 04, 2020 at 09:05:56AM +0800, Chao Yu wrote:
> On 2020/10/22 22:57, Gao Xiang wrote:
> > From: Gao Xiang <hsiangkao@redhat.com>
> > 
> > pcluster should be only set up for all managed pages instead of
> > temporary pages. Since it currently uses page->mapping to identify,
> > the impact is minor for now.
> > 
> > Fixes: 5ddcee1f3a1c ("erofs: get rid of __stagingpage_alloc helper")
> > Cc: <stable@vger.kernel.org> # 5.5+
> > Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
> 
> Reviewed-by: Chao Yu <yuchao0@huawei.com>

Thanks, I've also added a note to the commit message like this,
"
[ Update: Vladimir reported the kernel log becomes polluted
  because PAGE_FLAGS_CHECK_AT_FREE flag(s) set if the page
  allocation debug option is enabled. ]
"
Will apply all of this to -fixes branch.

Thanks,
Gao Xiang

> 
> Thanks,
>
Chao Yu Nov. 4, 2020, 1:44 a.m. UTC | #7
On 2020/11/4 9:11, Gao Xiang wrote:
> On Wed, Nov 04, 2020 at 09:05:56AM +0800, Chao Yu wrote:
>> On 2020/10/22 22:57, Gao Xiang wrote:
>>> From: Gao Xiang <hsiangkao@redhat.com>
>>>
>>> pcluster should be only set up for all managed pages instead of
>>> temporary pages. Since it currently uses page->mapping to identify,
>>> the impact is minor for now.
>>>
>>> Fixes: 5ddcee1f3a1c ("erofs: get rid of __stagingpage_alloc helper")
>>> Cc: <stable@vger.kernel.org> # 5.5+
>>> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
>>
>> Reviewed-by: Chao Yu <yuchao0@huawei.com>
> 
> Thanks, I've also added a note to the commit message like this,
> "
> [ Update: Vladimir reported the kernel log becomes polluted
>    because PAGE_FLAGS_CHECK_AT_FREE flag(s) set if the page
>    allocation debug option is enabled. ]
> "
> Will apply all of this to -fixes branch.

Thanks for noticing that, looks fine to me.

Thanks,

> 
> Thanks,
> Gao Xiang
> 
>>
>> Thanks,
>>
> 
> .
>
diff mbox series

Patch

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index 50912a5420b4..86fd3bf62af6 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -1078,8 +1078,11 @@  static struct page *pickup_page_for_submission(struct z_erofs_pcluster *pcl,
 		cond_resched();
 		goto repeat;
 	}
-	set_page_private(page, (unsigned long)pcl);
-	SetPagePrivate(page);
+
+	if (tocache) {
+		set_page_private(page, (unsigned long)pcl);
+		SetPagePrivate(page);
+	}
 out:	/* the only exit (for tracing and debugging) */
 	return page;
 }