USB 3.0 ?

2012-03-29 Thread Paul Goyette
Just curious - is anyone out there working on a USB 3.0 driver for 
NetBSD?



-
| Paul Goyette | PGP Key fingerprint: | E-mail addresses:   |
| Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com|
| Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette at juniper.net |
| Kernel Developer |  | pgoyette at netbsd.org  |
-


Re: extattr namespaces

2012-03-29 Thread Bill Stouder-Studenmund
On Mon, Feb 13, 2012 at 11:42:49AM +0100, Manuel Bouyer wrote:
> On Mon, Feb 13, 2012 at 10:37:25AM +, Emmanuel Dreyfus wrote:
> > On Mon, Feb 13, 2012 at 11:27:29AM +0100, Manuel Bouyer wrote:
> > > But then, if you see USER "foo.bar" in the filesystem, you don't know
> > > if it should be mapped to user.foo.bar or foo.bar.
> > > I don't think this can work.
> > 
> > If the disk is from FreeBSD, you mount with  -o stripxattr. When
> > you read it, USER "foo.bar" becomes user.foo.bar. When you write 
> > user.foo.bar, it is stored as USER "foo.bar". Therefore you can use 
> > the Linux-like API and you will not screw up what is stored on disk.
> 
> And what happens if you write "baz.foo.bar" ? You could just error out in
> this case.

Note: users and applications from OS X are not going to expect to have to 
write "user." to get at xattrs. I am not sure what all to recommend here 
as it looks like FreeBSD and Linux apps are going to want to add it. I 
really think Linux's way of putting the name space in the name is wrong. 
If these spaces are really supposed to have different symantic behavior, 
you should have to do different actions to get at them (i.e. flag 
something in the open).

Take care,

Bill


pgpRS4YVGB1YO.pgp
Description: PGP signature


Re: extattr namespaces

2012-03-29 Thread Bill Stouder-Studenmund
On Mon, Feb 06, 2012 at 05:04:45PM +, David Holland wrote:
> On Mon, Feb 06, 2012 at 09:51:19AM +, Emmanuel Dreyfus wrote:
>  > Here is public disuccsion about extended attributs namespaces, following
>  > a private request from yamt@
> 
> Which of the two models does OS X use?

OS X doesn't directly support namespacing nor does it support a difference 
in access. If a file has an extended attribute, a program can read and 
write it.

Put another way, the whole API deals with "user" xattrs.

ls -l@ /

drwxr-xr-x@ 62 root  wheel 2108 Feb 17 20:48 sbin
com.apple.FinderInfo  32 

That's some of the output on my Lion system. Looking in ~/Downloads, we 
see some other interesting EAs:

-rw-r--r--@  1 wrstuden  staff 8381161 Apr  2  2010 things_1.3.3.dmg
com.apple.diskimages.fsck   20 
com.apple.diskimages.recentcksum80 
com.apple.metadata:kMDItemDownloadedDate53 
com.apple.metadata:kMDItemWhereFroms   166 
com.apple.quarantine74 

So just make the whole API do "user" and you're set.

The OS X file systems that have "system" EAs and which I am familiar with 
do not expose them using the normal EA APIs. HFS uses a special EA to 
store a file's ACL, and you never see it (nor should you).

Xsan supports user, system, and system.inherited EAs. Everything that 
comes from userland is a "user" EA. We use the "system" ones for Reverse 
Path data, so we can get a path for a file descriptor. The 
"system.inherited" ones are used by Quantum for their data migration work.

As above, just have the VFS layer handle "user" attributes. It's simpler 
and I don't think anything else will be reliable.

Take care,

Bill


pgpDM4kClFA35.pgp
Description: PGP signature


Potential fix for PR kern/41417 (WAPBL: hang on tstile)

2012-03-29 Thread Brian Buhrow
Hello.  If you're running NetBSD-5, NetBSD-5.1 or any of the
derivatives of those branches and have been having trouble with the system
hanging with all processes stuck in tstile wait, then you're probably
experiencing pr kern/41417.  Even if you're running without WAPBL
journaling, you still could be suffering from this bug.
If you are, then there are test patches for this bug which I believe
fix this problem.  
These patches implement the UFS rename locking fixes David Holland made for
NetBSD-current with the addition that they've been made to work with
softdep, which is supported under NetBSD-5.
I would be interested in having as many people as
possible test these patches so we can insure  the fix is righ before we
commit to the branch.
If you know how to build a custom kernel, then you can download the
patch set for your version of NetBSD from:
http://gnats.NetBSD.org/41417
(there are patches for 5.0.x and 5.1.x branches respectively.)
If you do not know how to build a custom kernel or how to rebuild
GENERIC with the patches installed, let me know, and I'll post some
prebuilt  kernels.
If you test the patches, I'd be interested to know what your findings
are, good or bad.  Feel free to mail me with your results.

-thanks
-Brian