Re: [PATCH] dio: falling through to buffered I/O when invalidationof a page fails

[Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>]

At 04:15 07/12/15, Zach Brown wrote:
>
>> If anyone has a testcase - I can take a look at the problem again.
>
>I can try and throw something together..
>
>- z

I did a test by using fsstress.

I modified the dio write() of fsstress to check return value, and input
following command;

# fsstress -d /root/testdir/ -l 0 -n 1000 -p 1000 -z -f creat=1 -f write=1 
-f dwrite=1

I got EIO from dio write.


I encountered some PostgreSQL mailing list archive that mentioned about 
this issue.
http://groups.google.com/group/pgsql.hackers/browse_thread/thread/c7515ccd5ac1269c/dc2ddeac3fe0d8c1?lnk=raot&fwc=1

dio can be used in PostgreSQL to write transaction log(WAL). It is said 
that when dio is used,
EIO is returned under some workload.
 

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/