Re: [PATCH for-4.19] staging: erofs: fix mis-acted TAIL merging behavior
From: Gao Xiang
Date: Thu Mar 07 2019 - 08:04:42 EST
Hi Greg,
On 2019/3/7 20:25, Greg KH wrote:
> On Thu, Mar 07, 2019 at 01:34:04PM +0800, Gao Xiang wrote:
>> commit a112152f6f3a2a88caa6f414d540bd49e406af60 upstream.
>>
>> EROFS has an optimized path called TAIL merging, which is designed
>> to merge multiple reads and the corresponding decompressions into
>> one if these requests read continuous pages almost at the same time.
>>
>> In general, it behaves as follows:
>> ________________________________________________________________
>> ... | TAIL . HEAD | PAGE | PAGE | TAIL . HEAD | ...
>> _____|_combined page A_|________|________|_combined page B_|____
>> 1 ] -> [ 2 ] -> [ 3
>> If the above three reads are requested in the order 1-2-3, it will
>> generate a large work chain rather than 3 individual work chains
>> to reduce scheduling overhead and boost up sequential read.
>>
>> However, if Read 2 is processed slightly earlier than Read 1,
>> currently it still generates 2 individual work chains (chain 1, 2)
>> but it does in-place decompression for combined page A, moreover,
>> if chain 2 decompresses ahead of chain 1, it will be a race and
>> lead to corrupted decompressed page. This patch fixes it.
>>
>> Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
>> Cc: <stable@xxxxxxxxxxxxxxx> # 4.19+
>> Signed-off-by: Gao Xiang <gaoxiang25@xxxxxxxxxx>
>> Reviewed-by: Chao Yu <yuchao0@xxxxxxxxxx>
>> ---
>> drivers/staging/erofs/unzip_vle.c | 69 +++++++++++++++++++++++++--------------
>> 1 file changed, 44 insertions(+), 25 deletions(-)
>>
>
> thanks for the backport, now applied.
Yes, it seems that Sasha's bot cannot cherry-pick it directly,
therefore I made this patch for 4.19 by hand...
BTW, there is another patch called "staging: erofs: compressed_pages should not be accessed again after freed"
in the same condition... And I also send 4.19 and 4.20 version to stable mailing list,
could you also kindly check them out?
Thanks,
Gao Xiang
>
> greg k-h
>