[LU-13846] move iov iter forward by ourself

• Type: Bug
• Resolution: Fixed
• Priority: Major
• Fix Version/s: Lustre 2.14.0
• Severity: 3
• Assignee: Wang Shilong (Inactive)
• Reporter: Wang Shilong (Inactive)

Description

Newer kernel will reward iov iter back to original
position for direct IO, see following codes:

    
            iov_iter_revert(from, write_len -iov_iter_count(from));--------->here
    out:
            return written;
    }
    EXPORT_SYMBOL(generic_file_direct_write);

This break assumptions from Lustre and caused problem
when Lustre need split one IO to several io loop, considering
4M block IO for 1 MiB stripe file, it will submit first 1MiB IO
4 times and caused data corruptions finally.

Since generic kernel varies from different kernel versions,
we'd better fix this problem by move iov iter forward by
Lustre itself.

Issue Links

• is related to: LU-13835 reuse same cl_dio_aio for one IO

Comments

[James A Simmons]

Patch has landed. If more work is needed please reopen

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/39565/
Subject: LU-13846 llite: move iov iter forward by ourself
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 689714eb511d39e225da64d8bfae42a4b2554626

[Wang Shilong (Inactive)]

A bit deep look when we will have this bug:

The problem existed since kernel v3.16 from following upstream linux commit:

commit 26978b8b4d83c46f4310b253db70fa9e65149e7c
Author: Al Viro <[e-mail]>
Date:   Mon Mar 10 14:08:45 2014 -0400
    give ->direct_IO() a copy of iov_iter
    
    the thing is, we want to advance what's given to ->direct_IO() as we
    are forming the request; however, the callers care about the amount
    of data actually transferred, not the amount we tried to transfer.
    It's more convenient to allow ->direct_IO() instances do use
    iov_iter_advance() on the copy of iov_iter, leaving the actual
    advancing of the original to caller.
    
    Signed-off-by: Al Viro <[e-mail]>

[Wang Shilong (Inactive)]

Patch is here:
https://review.whamcloud.com/#/c/39565/2