* Christopher D. Clausen [2007-01-03 11:31:16 -0600]:
Jeffrey Altman [EMAIL PROTECTED] wrote:
If you are in a position to get a new entry added to the Linux
sys/statfs.h header file, please do so. I would suggest a value of
AFS_SUPER_MAGIC 0x5346414F
To be honest, I really don't think
Jeffrey Altman [EMAIL PROTECTED] wrote:
If you are in a position to get a new entry added to the Linux
sys/statfs.h header file, please do so. I would suggest a value of
AFS_SUPER_MAGIC 0x5346414F
To be honest, I really don't think there would be any harm in just
using this value. OAFS
On Dec 25, 2006, at 6:49 PM, Derrick J Brashear wrote:
On Mon, 25 Dec 2006, Adam Megacz wrote:
Marcus Watts [EMAIL PROTECTED] writes:
#1 - what is the process to get a real linux filesystem type
number
assigned?
Ask the Linux Standards Committee.
(No such entity)
Actually I
Steve Simmons wrote:
*chuckle* So as a reducto ad absurdum argument, I can assume this
means there are no actual linux filesystem type numbers because it's
impossible to get agreement. As Descarte didn't say, I think not.
/*chuckle
Since it clearly is possible and there's a minor shitload
Adam Megacz [EMAIL PROTECTED] writes:
Are you referring to this?
http://www.pdc.kth.se/kth-krb/
The code in lib/kafs/afssys.c appears to catch the SIGSYS signal while
doing some scary syscall()-guessing.
This is the closest thing to an answer so far. But that's a pretty
huge amount of
Marcus Watts [EMAIL PROTECTED] writes:
Yes. So why is AFS reporting 0x if that isn't its magic?
linux reports 0 for AFS because that is what it is programmed to do.
(the implied question here is why was it programmed that way?)
This is on or around line 421 in file
Adam Megacz [EMAIL PROTECTED] writes:
To: openafs-info@openafs.org
From: Adam Megacz [EMAIL PROTECTED]
Subject: [OpenAFS] Re: openafs does not put a [correct] value in
fsinfo.f_type?
Date: Sun, 24 Dec 2006 23:59:01 -0800
Marcus Watts [EMAIL PROTECTED] writes:
Yes. So why is AFS
Marcus Watts [EMAIL PROTECTED] writes:
#1 - what is the process to get a real linux filesystem type number
assigned?
Ask the Linux Standards Committee.
(No such entity)
Actually I think you just make one up and advertise it widely.
#2 - how can I reliably determine if a file is
On Mon, 25 Dec 2006, Marcus Watts wrote:
#2 - how can I reliably determine if a file is stored in AFS?
pioctl
You can use code in src/venus/fs.c as an example, or code from Arla.
#3 - how can I tell which kind of locks are truely supported by the (afs)
filesystem?
today? Or
On Mon, 25 Dec 2006, Adam Megacz wrote:
Marcus Watts [EMAIL PROTECTED] writes:
#1 - what is the process to get a real linux filesystem type number
assigned?
Ask the Linux Standards Committee.
(No such entity)
Actually I think you just make one up and advertise it widely.
And then
In message [EMAIL PROTECTED],Adam Megacz writes:
The question is [on Linux], how can my code reliably determine if a
file is stored in AFS without introducing a build-time dependency on
AFS headers/libraries into my code.
perhaps your original question is fallacious. how do you know 0 isnt
the
On Mon, 25 Dec 2006, chas williams - CONTRACTOR wrote:
In message [EMAIL PROTECTED],Adam Megacz writes:
The question is [on Linux], how can my code reliably determine if a
file is stored in AFS without introducing a build-time dependency on
AFS headers/libraries into my code.
perhaps your
Derrick J Brashear [EMAIL PROTECTED] writes:
This won't work for code which must build without dependencies on AFS,
of which might be built on a non-AFS machine and executed on an
AFS-client machine (ie for rpm/deb packaging).
Untrue. The kafs library has its on syscall prober thing which
Adam:
If you want your code to be file system, AFS client, and version
independent you really need to test the functionality and and
not the implementations. Specifically, you must perform a test
that attempts to lock a byte range, spin off a second process attempt
to lock the same byte range,
On Mon, 25 Dec 2006, Adam Megacz wrote:
Derrick J Brashear [EMAIL PROTECTED] writes:
This won't work for code which must build without dependencies on AFS,
of which might be built on a non-AFS machine and executed on an
AFS-client machine (ie for rpm/deb packaging).
Untrue. The kafs
Jeffrey Altman [EMAIL PROTECTED] writes:
If you want your code to be file system, AFS client, and version
independent you really need to test the functionality and and
not the implementations. Specifically, you must perform a test
that attempts to lock a byte range, spin off a second process
Adam Megacz wrote:
Jeffrey Altman [EMAIL PROTECTED] writes:
If you want your code to be file system, AFS client, and version
independent you really need to test the functionality and and
not the implementations. Specifically, you must perform a test
that attempts to lock a byte range, spin
Jeffrey Altman [EMAIL PROTECTED] writes:
Absolutely. In fact, I'd already implemented that. The only problem
is that it doesn't work on Linux since the Linux AFS client does a
more elaborate job of faking the byte-range locks.
The AFS Linux client does not implement this functionality.
We
Adam Megacz wrote:
Jeffrey Altman [EMAIL PROTECTED] writes:
Absolutely. In fact, I'd already implemented that. The only problem
is that it doesn't work on Linux since the Linux AFS client does a
more elaborate job of faking the byte-range locks.
The AFS Linux client does not implement
Jeffrey Altman [EMAIL PROTECTED] writes:
I don't know how the Linux kernel is determining that it should activate
this functionality.
OpenAFS asks it to -- afs_linux_lock() calls posix_lock_file() in
osi_vnodeops.c.
...
AFS_GLOCK();
code = afs_lockctl(vcp, flock, cmd, credp);
chas williams - CONTRACTOR [EMAIL PROTECTED] writes:
In message [EMAIL PROTECTED],Adam Megacz writes:
afs has no magic.
Neither does NFS, but it reports 0x6969.
statfs man page says:
NFS_SUPER_MAGIC 0x6969
Yes. So why is AFS reporting 0x if that isn't its
Adam Megacz [EMAIL PROTECTED] writes:
To: openafs-info@openafs.org
From: Adam Megacz [EMAIL PROTECTED]
Subject: [OpenAFS] Re: openafs does not put a [correct] value in
fsinfo.f_type?
Sender: [EMAIL PROTECTED]
Errors-To: [EMAIL PROTECTED]
Date: Sun, 24 Dec 2006 18:34:10 -0800
chas
22 matches
Mail list logo
