Status of hfs support as of 2.4.0-test1-ac21

I played around with hfs support in this developer kernel and
penetrated my hfs partion with some paralell cp-s and rm-s
(I copied "the linux kernel" documentation around :-)
After every step I rebootet into MacOs and ran DiskFirstAid (DFA)
but nothing was found. cp and rm complained about problems with the
.finderinfo and .resource directories, but I expected it this way.
I used my iBook for testing.

Is there anything else i _must_ check for a fs to be considered
stable ?

In case anyone reports a bug to me in this area I'll try to fix
it.

Meanwhile I'll have a look at the hfsutils
( http://www.mars.org/home/rob/proj/hfs/ ) and will start
moving them to HFS+.

I found however some bugs not related to hfs and may reproduce
them in detail in case someone asks:
  - the console driver messed up some pixels while editing
  - hdparm did not always work correctly
    (I use pmud and it calls hdparm depending on the battery load)
    Sometimes I got stuck with the message "hda interrupt lost"
    or similar.


Greetings,

|       |                   Klaus Halfmann    khalfmann@libra.de
|       |
|    i  | --.  r---  --- |  Libra Software    Tel +49 (0)621 41997-0
|    |  b-- |  |    a--- |  Erzbergerstr. 17  Fax +49 (0)621 41997-30
L--- |  --- |  |    .----.  68165 Mannheim

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Hi Klaus,

On  23 Jun, this message from Halfmann, Klaus echoed through cyberspace:

Is there anything else i _must_ check for a fs to be considered
stable ?

Did you try accessing CDRoms ith their 2K block size? While late
2.3.99/2.4 kernels have always been able to access hfs disks in a more
or less stable way, CDRoms have not been working for a very long time...

The LinuxPPC 2K CDs are an example of such hfs CDs.

Thanks

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

On Fri, Jun 23, 2000, Halfmann, Klaus [off-list ref] wrote:

I found however some bugs not related to hfs and may reproduce
them in detail in case someone asks:
 - the console driver messed up some pixels while editing
 - hdparm did not always work correctly
   (I use pmud and it calls hdparm depending on the battery load)
   Sometimes I got stuck with the message "hda interrupt lost"
   or similar.

That's interesting. Could you give me more details about what you do
excatly with hdparm that is causing the lost interrupt ? Can you try to
use it manually (without pmud) and reproduce the problem ?

Ben.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

At 06:35 -0500 6/23/00, Michel Lanners wrote:

Did you try accessing CDRoms ith their 2K block size? While late
2.3.99/2.4 kernels have always been able to access hfs disks in a more
or less stable way, CDRoms have not been working for a very long time...

Really?  All the CDs I burn in Toast must have 512 byte sectors.  I've
always been able to access them from both my PPC and X86 Linux boxes.

Or maybe it's because I threw out the Philips and got a Pioneer.  I dunno.
_______________________________________________________________________________
Mike DeSimone        Applied Research Laboratories: The University of Texas
Research Engineer    Mail to: PO Box 8029       Shipments to: 10000 Burnet Rd.
Sonar Development             Austin, TX 78713                Austin, TX 78758
Office (512) 835-3061-------FAX (512) 835-3259--------desimone@arlut.utexas.edu

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

  - hdparm did not always work correctly
    (I use pmud and it calls hdparm depending on the battery load)
    Sometimes I got stuck with the message "hda interrupt lost"
    or similar.

What sort of machine is that on? The lost interrupt should only happen
(copiously) on Pismo when the pwrctl script forces a disk shutdown on
battery low. Having HD requests time out often enough to force a disk
reset is the only way to power up the disk again there.

	Michael


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Halfmann, Klaus wrote:

Meanwhile I'll have a look at the hfsutils
( http://www.mars.org/home/rob/proj/hfs/ ) and will start
moving them to HFS+.

I don't know if you want to look at my code, but I've been working on a
module for the kernel to do HFS+. I'd be willing to help you out with the
updates to hfsutils. I'll be putting up new HFS+ code shortly, perhaps
this weekend if I get a chance.

	Brad Boyer
	flar@pants.nu


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/