xfs
[Top] [All Lists]

Re: [PATCH] xfstests: v2 xfs directory unbalance assert test

To: "Michael L. Semon" <mlsemon35@xxxxxxxxx>
Subject: Re: [PATCH] xfstests: v2 xfs directory unbalance assert test
From: Mark Tinguely <tinguely@xxxxxxx>
Date: Wed, 18 Sep 2013 12:43:20 -0500
Cc: "xfs@xxxxxxxxxxx" <xfs@xxxxxxxxxxx>
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <CAJzLF9kacDY3gu7moxfFDRMTCxjQOzp_574Ojagk-kvNySedfg@xxxxxxxxxxxxxx>
References: <20130917192538.230164044@xxxxxxx> <CAJzLF9kacDY3gu7moxfFDRMTCxjQOzp_574Ojagk-kvNySedfg@xxxxxxxxxxxxxx>
User-agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:9.0) Gecko/20120122 Thunderbird/9.0
On 09/18/13 11:36, Michael L. Semon wrote:
Looks good.  I tried it on an 11-GB partition with an unpatched git kernel
3.12.0 + xfs-oss/master, coupled with the latest git xfsprogs.  It took a
reasonable amount of time, relatively speaking.  It put forth the error I was
expecting, and a new 32-bit core is available upon your request but not
uploaded yet.

As for the assert itself, according to kgdb, these are the first, last, and
BBTOB(bp->b_length) numbers for this run:

1544 4591 4096

Good work!

Michael

Thanks for the feedback.

last > BBTOB(bp->b_length) was also true for the longer version of the test. The only difference is that it happens earlier in the remove.

I take a closer look at the code.

--Mark.

<Prev in Thread] Current Thread [Next in Thread>