On Tue, Jan 31, 2012 at 09:48:42AM +0100, Eric Dumazet wrote:
> Le mardi 31 janvier 2012 à 15:59 +0800, Shaohua Li a écrit :
>> Herbert Poetzl reported a performance regression since 2.6.39. The test
>> is a simple dd read, but with big block size. The reason is:
>> T1: ra (A, A+128k), (A+128k, A+256k)
>> T2: lock_page for page A, submit the 256k
>> T3: hit page A+128K, ra (A+256k, A+384). the range isn't submitted
>> because of plug and there isn't any lock_page till we hit page A+256k
>> because all pages from A to A+256k is in memory
>> T4: hit page A+256k, ra (A+384, A+ 512). Because of plug, the range isn't
>> submitted again.
>> T5: lock_page A+256k, so (A+256k, A+512k) will be submitted. The task is
>> waitting for (A+256k, A+512k) finish.
>> There is no request to disk in T3 and T4, so readahead pipeline breaks.
>> We really don't need block plug for generic_file_aio_read() for buffered
>> I/O. The readahead already has plug and has fine grained control when I/O
>> should be submitted. Deleting plug for buffered I/O fixes the regression.
>> One side effect is plug makes the request size 256k, the size is 128k
>> without it. This is because default ra size is 128k and not a reason we
>> need plug here.
>> Signed-off-by: Shaohua Li <shaohua.li@xxxxxxxxx>
>> Tested-by: Herbert Poetzl <herbert@xxxxxxxxxxxx>
>> Tested-by: Eric Dumazet <eric.dumazet@xxxxxxxxx>
> Hmm, this is not exactly the patch I tested from Wu Fengguang
> I'll test this one before adding my "Tested-by: ..."
second that
>> Signed-off-by: Wu Fengguang <fengguang.wu@xxxxxxxxx>
>> diff --git a/mm/filemap.c b/mm/filemap.c
>> index 97f49ed..b662757 100644
>> --- a/mm/filemap.c
>> +++ b/mm/filemap.c
>> @@ -1400,15 +1400,12 @@ generic_file_aio_read(struct kiocb *iocb, const struct iovec *iov,
>> unsigned long seg = 0;
>> size_t count;
>> loff_t *ppos = &iocb->ki_pos;
>> - struct blk_plug plug;
>> count = 0;
>> retval = generic_segment_checks(iov, &nr_segs, &count, VERIFY_WRITE);
>> if (retval)
>> return retval;
>> - blk_start_plug(&plug);
>> -
>> /* coalesce the iovecs and go direct-to-BIO for O_DIRECT */
>> if (filp->f_flags & O_DIRECT) {
>> loff_t size;
>> @@ -1424,8 +1421,12 @@ generic_file_aio_read(struct kiocb *iocb, const struct iovec *iov,
>> retval = filemap_write_and_wait_range(mapping, pos,
>> pos + iov_length(iov, nr_segs) - 1);
>> if (!retval) {
>> + struct blk_plug plug;
>> +
>> + blk_start_plug(&plug);
> This part was not on the tested patch yesterday.
yep, will test this one tonight ...
>> retval = mapping->a_ops->direct_IO(READ, iocb,
>> iov, pos, nr_segs);
>> + blk_finish_plug(&plug);
>> }
>> if (retval > 0) {
>> *ppos = pos + retval;
>> @@ -1481,7 +1482,6 @@ generic_file_aio_read(struct kiocb *iocb, const struct iovec *iov,
>> break;
>> }
>> out:
>> - blk_finish_plug(&plug);
>> return retval;
>> }
>> EXPORT_SYMBOL(generic_file_aio_read);
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>