2008-05-12 18:27:11

by Eric Sandeen

Subject: [PATCH e2fsprogs] - fix swap sanity tests in blkid, and blkid tests

Swap is actually native-endian on disk, and with the latest
swapspace sanity checks I added we need to have native swapspace
examples in the blkid tests, so re-mkswap them during testing.

One one other required change, though; mkswap requires at least
10 pages of swap, so the image needs to be increased to 10x64k
if mkswap is to succeed...

Maybe it'd be better to just dd it out on the fly?

Addresses-redhat-bugzilla: 445786

Signed-off-by: Eric Sandeen <[email protected]>
---

Index: e2fsprogs-1.41/lib/blkid/probe.c
===================================================================
--- e2fsprogs-1.41.orig/lib/blkid/probe.c
+++ e2fsprogs-1.41/lib/blkid/probe.c
@@ -845,8 +845,7 @@ static int probe_swap1(struct blkid_prob

/* check for wrong version or zeroed pagecount, for sanity */
if (!memcmp(id->bim_magic, "SWAPSPACE2", id->bim_len) &&
- (blkid_le32(sws->sws_version) != 1 ||
- sws->sws_lastpage == 0))
+ (sws->sws_version != 1 || sws->sws_lastpage == 0))
return 1;

/* arbitrary sanity check.. is there any garbage down there? */
Index: e2fsprogs-1.41/lib/blkid/test_probe.in
===================================================================
--- e2fsprogs-1.41.orig/lib/blkid/test_probe.in
+++ e2fsprogs-1.41/lib/blkid/test_probe.in
@@ -20,6 +20,12 @@ do
continue
fi
bunzip2 < $SRCDIR/tests/$i.img.bz2 > tests/tmp/test.img.$$
+ # swap is native-endian, so regenerate before testing
+ if [ "$i" == "swap0" ]; then
+ mkswap -v0 tests/tmp/test.img.$$ > /dev/null
+ elif [ "$i" == "swap1" ]; then
+ mkswap -v1 -L SWAP-TEST -U 8ff8e77f-8553-485e-8656-58be67a81666 tests/tmp/test.img.$$ >/dev/null
+ fi
./tst_probe tests/tmp/test.img.$$ > tests/$i.out
/bin/rm -f tests/tmp/test.img.$$ tests/$i.ok tests/$i.failed
cmp -s tests/$i.out $SRCDIR/tests/$i.results

2008-05-13 06:32:40

by Andreas Dilger

Subject: Re: [PATCH e2fsprogs] - fix swap sanity tests in blkid, and blkid tests

On May 12, 2008 13:26 -0500, Eric Sandeen wrote:
> One one other required change, though; mkswap requires at least
> 10 pages of swap, so the image needs to be increased to 10x64k
> if mkswap is to succeed...
>
> Maybe it'd be better to just dd it out on the fly?

Probably, yes. Little point in bunzipping bytes that you are just
going to overwrite...

> bunzip2 < $SRCDIR/tests/$i.img.bz2 > tests/tmp/test.img.$$
> + # swap is native-endian, so regenerate before testing
> + if [ "$i" == "swap0" ]; then
> + mkswap -v0 tests/tmp/test.img.$$ > /dev/null
> + elif [ "$i" == "swap1" ]; then
> + mkswap -v1 -L SWAP-TEST -U 8ff8e77f-8553-485e-8656-58be67a81666 tests/tmp/test.img.$$ >/dev/null
> + fi

Cheers, Andreas
--
Andreas Dilger
Sr. Staff Engineer, Lustre Group
Sun Microsystems of Canada, Inc.

2008-05-13 12:47:22

by Eric Sandeen

Subject: Re: [PATCH e2fsprogs] - fix swap sanity tests in blkid, and blkid tests

Andreas Dilger wrote:
> On May 12, 2008 13:26 -0500, Eric Sandeen wrote:
>> One one other required change, though; mkswap requires at least
>> 10 pages of swap, so the image needs to be increased to 10x64k
>> if mkswap is to succeed...
>>
>> Maybe it'd be better to just dd it out on the fly?
>
> Probably, yes. Little point in bunzipping bytes that you are just
> going to overwrite...

It's just that right now it iterates over existing files, so that's how
it finds swap0 and swap1 to test...

*shrug*

Could make them bzipped sparse files... :)

>> bunzip2 < $SRCDIR/tests/$i.img.bz2 > tests/tmp/test.img.$$
>> + # swap is native-endian, so regenerate before testing
>> + if [ "$i" == "swap0" ]; then
>> + mkswap -v0 tests/tmp/test.img.$$ > /dev/null
>> + elif [ "$i" == "swap1" ]; then
>> + mkswap -v1 -L SWAP-TEST -U 8ff8e77f-8553-485e-8656-58be67a81666 tests/tmp/test.img.$$ >/dev/null
>> + fi
>
> Cheers, Andreas
> --
> Andreas Dilger
> Sr. Staff Engineer, Lustre Group
> Sun Microsystems of Canada, Inc.
>