[linux-mm-cc] [RFC] LZO de/compression support - take 3

[Nitin Gupta]

On 5/25/07, Bret Towe <magnade at gmail.com> wrote:
> On 5/24/07, Nitin Gupta <nitingupta910 at gmail.com> wrote:
> > On 5/23/07, Bret Towe <magnade at gmail.com> wrote:
> > > On 5/23/07, Nitin Gupta <nitingupta910 at gmail.com> wrote:
> >
> > > > For now, tested on x86 only.
> > >
> > > If you have a program to test this I can run it on an amd64 and a g4 ppc
> > >
> >
> > Attached is the kernel module (compress-test) to test this LZO code.
> > Just compile this module against 2.6.22-rc2 with this LZO patch. Then
> > testing can be done as:
> > 1- Mount DebugFS somewhere e.g:
> > mkdir /debug; mount -t debugfs debugfs /debug
> > 2- Load the module and do:
> > cat /path/to/some_file > /debug/compress_test/compress
> > (/var/log/messages should show that compression was successful)
> > 3- Then decompress this file as:
> > cat /debug/compress_test/decompress > /tmp/t
> > (/var/log/messages should show that decompression was successful)
> > 4- For extra verification do:
> > diff /tmp/t /path/to/some_file   -- O/P must be empty
> >
> >
>
> the test worked fine on amd64
> from dmesg:
> LZO compress successful: orig_size=17448, comp_size=8183
> LZO decompress successful: decomp_size=17448
>
> and input and output files I gave it:
> sha1sum test-input output
> 2221c586e3eb869af7f4333d4f56b441b9aa8414  test-input
> 2221c586e3eb869af7f4333d4f56b441b9aa8414  output
>

Good to know it worked correctly on 64-bit system too. I will also
add exporting benchmarking figures soon.

> (will be giving the ppc box the same file to test btw when I get to it)
>
> and I don't know if it matters much but I tried feeding it a 260k file
> and it didn't like it
>
> cat /usr/bin/yelp > /debug/compress_test/compress
> cat: write error: No space left on device
>

Ah! I forgot to mention that max file size to feed is 256K (this was
just for simplicity of compress-test module implementation).

Thanks for your testing.

- Nitin