xfs
[Top] [All Lists]

Re: [RFC] btree test harness

To: xfs@xxxxxxxxxxx
Subject: Re: [RFC] btree test harness
From: Christoph Hellwig <hch@xxxxxx>
Date: Mon, 1 Sep 2008 00:24:20 +0200
In-reply-to: <20080813223022.GA15025@xxxxxx>
References: <20080813223022.GA15025@xxxxxx>
Sender: xfs-bounce@xxxxxxxxxxx
User-agent: Mutt/1.3.28i
New version below.  Now we use an ioctl to driver the tests for
a given maximum of levels.  There's a small program for xfstests
that calls it.  No real testcases yet because the setup is still a
little weird:

mkfs.xfs -f \
        -b size=512 \
        -d agcount=1 \
        -l logdev=/dev/hdb,size=65000b,lazy-count=1 \
        /dev/hdd

mount -t xfs -o logdev=/dev/hdb /dev/hdd /mnt/

./src/btree-test /mnt 0 <levels to test>

Next step would be to find the largest span of free blocks using real
allocator functions, which should make this a lot safer, and might also
get rid of the requirement of an external log.


Attachment: xfs-btree-test-harness
Description: Text document

Attachment: fsqa-add-btree-testharness
Description: Text document

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