Chas Williams [Sun, 30 Jun 2002 07:39:12 +0000]
linux-fix-readdir-hinting-
20020630
an earlier patch erroneously removed this
Derrick Brashear [Tue, 25 Jun 2002 16:25:43 +0000]
linux-gcpags-support-
20020625
might even work
Derrick Brashear [Tue, 25 Jun 2002 16:02:11 +0000]
linux-kernel-detect-gfp-mask-in-struct-address-space-
20020624
based on report that new redhat kernels need this
Chas Williams [Mon, 24 Jun 2002 23:18:18 +0000]
linux-sleep-wakeup-cleanup-
20020624
"cleanup osi_sleep.c to remove some additional races. this patches
removes AFS_GLOBAL_SUNLOCK and just uses that case to handle both
smp and non smp cases. osi_TimedSleep always sleeps atleast 1 hz.
(it also always sleeps interruptibly -- you might get some extra
loops but this is better than the load average running away)
sleep_on's are replaced with the more appropriate add_waitq/schedule
sequence. getevent is called after addevent since there is no
danger of GLOCK dropping. getevent/addevent isnt re-merged since
osi_Wakeup doesnt create entries in the event table."
Kris Van Hees [Mon, 24 Jun 2002 20:09:59 +0000]
vol-allocate-partition-names-dynamically-to-avoid-problems-with-vx-stuff-
20020624
otherwise long device paths can screw you
Derrick Brashear [Mon, 24 Jun 2002 18:34:02 +0000]
linux-dont-install-osi-vfs-h-when-no-kernel-module-
20020611
do --disbale-kernel-mdoule build works
====================
This delta was composed from multiple commits as part of the CVS->Git migration.
The checkin message with each commit was inconsistent.
The following are the additional commit messages.
====================
the rest of the patch
Marcus Watts [Mon, 24 Jun 2002 17:30:34 +0000]
ubik-pause-collapsing-
20020624
I'm told that after we introduce ptserver nested groups we can expect to
see periodic pauses in ubik operations, and this fixes the problem. if it
happens, we can start with the UBIK_PAUSE code and go from there
Chas Williams [Mon, 24 Jun 2002 16:40:41 +0000]
linux-more-osi-sleep-wakeup-fun-
20020624
"the attached patch 'fixes', in my opinion, in osi_alloc. objects <=
AFS_SMALLOCSIZ are alloced with kmalloc, other vmalloc. if kmalloc
fails it doesnt fail over to vmalloc. TASK_RUNNING is set after
schedule. it doesnt drop the GLOCK. i moved the allocator init
later so that it doesnt need to toggle the semaphore just before
going to linux_alloc."
Derrick Brashear [Mon, 24 Jun 2002 16:23:40 +0000]
ufsread-save-the-oops-for-where-we-expect-it-
20020624
based on comment from matt@slackers.net
Pavel Semerad [Mon, 24 Jun 2002 16:03:34 +0000]
dcache-locking-cleanup-
20020624
don't oops if getDCache failed
Richard Basch [Sat, 22 Jun 2002 05:32:45 +0000]
Update_TargetVnodeStatus() checks and strips setuid bits for non admin
clients; it should do the same thing for setgid files.
Derrick Brashear [Sat, 22 Jun 2002 05:26:52 +0000]
Add afsUUID_to_string() and afsUUID_from_string() utility functions, from
arla.
Garry Zacheiss [Sat, 22 Jun 2002 05:24:49 +0000]
Use:
#define CONFIG_SMP 1
instead of:
#define CONFIG_SMP
to fix recent Redhat kernel which used #if CONFIG_SMP instead of
#ifdef CONFIG_SMP in some kernel headers.
Derrick Brashear [Sat, 15 Jun 2002 18:11:18 +0000]
linux-newvcache-make-sure-we-dont-try-to-iterate-null-pointer-
20020615
based on report from wilder@us.ibm.com
not clear this is sufficient
Derrick Brashear [Sat, 15 Jun 2002 17:52:00 +0000]
afsd-dont-print-garbage-on-bad-cacheinfo-file-
20020615
based on report from matt@slackers.net
don't bother printing anything we parsed, since we're guaranteed at least
one item is trash anyhow
Mattias Amnefelt [Fri, 14 Jun 2002 18:45:33 +0000]
dux-sysent-update-
20020614
cast this to deal with changes in sysent in later tru64
Derrick Brashear [Fri, 14 Jun 2002 18:22:37 +0000]
junk-xdr-in-non-common-makefiles-
20020614
based on report from chas@cmf.nrl.navy.mil
Derrick Brashear [Thu, 13 Jun 2002 20:33:55 +0000]
vos-listaddrs-cleanup-and-allow-printing-of-ip-or-uuid-or-both-
20020613
lha@stacken.kth.se complained that we sometimes printed garbage, so i took
the opportunity to both fix it and update ours to support the options
the arla vos listaddrs does
Derrick Brashear [Wed, 12 Jun 2002 23:07:08 +0000]
vol-fssync-ack-before-breaking-callbacks-
20020612
given that we don't bother to do anything to return value, we can ack
the fssync op earlier and not tie up the volserver unnecessarily
found due to comments from brent.johnson@jpl.nasa.gov
Derrick Brashear [Wed, 12 Jun 2002 22:41:36 +0000]
pioctl-always-null-terminate-after-copy-
20020612
based on report from lha@e.kth.se
note that this works because PIGGYSIZE is always less than AFS_LRALLOCSIZ
Marc Dionne [Wed, 12 Jun 2002 19:22:24 +0000]
windows-update-dirpath-header-
20020612
add missing definitions
Richard Basch [Wed, 12 Jun 2002 18:06:45 +0000]
Don't give directory owners any implicit rights. Volume owners still
have their implicit rights.
Derrick Brashear [Wed, 12 Jun 2002 07:02:02 +0000]
arla-pioctls-now-bogus-
20020612
as reported by lha@stacken.kth.se
Joakim Fallsjö [Wed, 12 Jun 2002 01:25:33 +0000]
solaris-x86-
20020611
solaris 7 and 8 for x86
Quanah Gibson-Mount [Wed, 12 Jun 2002 01:03:17 +0000]
solaris-9-support-
20020611
update hrestime interface
Derrick Brashear [Wed, 12 Jun 2002 00:33:40 +0000]
windows-callback-interface-updates-
20020611
make callback interface like unix
Manuel Pereira [Wed, 12 Jun 2002 00:29:18 +0000]
libjuafs-target-
20020611
create a libjuafs library for the java api to use
Derrick Brashear [Thu, 13 Jun 2002 20:35:28 +0000]
revert-dont-spawn-kernel-thread-for-rxk-listener-
20020426
in some cases this causes shutdowns to hang forever. notably shutdown -i0 -g6 -y works while reboot does not
Derrick Brashear [Sat, 8 Jun 2002 12:01:59 +0000]
getcellservdb-cleanup-
20020608
based on comments and suggestion from Srikanth Vishwanathan <vsrikanth@in.ibm.com>
avoid deadlock. send back variable length list. allocate what we need to, and
be care to clean up after ourselves if there's an error. whether this
should be variable length is arguable but it's pretty clear no one was using
this before.
Derrick Brashear [Sat, 8 Jun 2002 08:44:09 +0000]
use-private-xdr-in-kernel-to-avoid-conflicts-over-memory-ownership-
20020608
allocating memory from a pool and then having it be freed out from under you is unwise
Garry Zacheiss [Sat, 8 Jun 2002 05:54:50 +0000]
Update NEWS file with important changes in 1.2.5 release.
Garry Zacheiss [Sat, 8 Jun 2002 02:29:50 +0000]
Recognize MacOS X 10.1.5 as ppc_darwin_14. It's compatible.
Chas Williams [Sat, 8 Jun 2002 00:55:01 +0000]
des-fix-linux-64bit-problems-more-usefully-
20020607
with a minor fix from shadow@dementia.org
long isn't 32bit on ia64 or alpha
Nickolai Zeldovich [Wed, 5 Jun 2002 16:56:09 +0000]
Delay calling afs_ProcessFS() and upgrading the shared vcache lock
until after the dcache locks have been released, to obey lock ordering
and avoid deadlock.
Garry Zacheiss [Wed, 29 May 2002 21:59:51 +0000]
uuencode binary files used by the test suite (KeyFile and sample kaserver
database). Keeping binary files in CVS is annoying, and this way "cvs rdiff"
will work again.
Derrick Brashear [Fri, 24 May 2002 23:05:47 +0000]
viced-malloc-dont-make-assumptions-about-the-compiler-
20020524
based on report from lha@stacken.kth.se
"identP = (struct Identity *)malloc(1);
This can't be right, there should not be an assumption how the
c-compiler will lay out code. I find it acceptable to do something like this
/* sizeof valid + offset of valid. */
i.. = malloc(sizeof(identP->valid) + (int)&(((struct Identity *)0)->valid));
But then, why not just allocate a whole identP and be happy. A afsUUID
is not that bloted (16 bytes) so I can't see why way this is
chosen. 20 bytes will probably be allocated on 32 bit platforms for
the whole struct Identity.
"
Derrick Brashear [Thu, 23 May 2002 06:21:13 +0000]
afs-dont-allow-shutdown-before-umount-
20020523
fix afs to enforce that /afs must be unmounted, and afsd to report it
Nickolai Zeldovich [Tue, 21 May 2002 07:37:53 +0000]
Zero out some fields to avoid random crashes in "cmdebug host -cache".
Nickolai Zeldovich [Tue, 21 May 2002 07:35:57 +0000]
Report correct cache size when -memcache is used.
Nathan Neulinger [Tue, 21 May 2002 06:35:45 +0000]
aix-fix-objdir-build-
20020521
make objdir builds work on aix
Chaskiel M Grundman [Tue, 21 May 2002 06:30:27 +0000]
pts-sysviewer-should-be-able-to-listowned-
20020521
fix so the "any reader" can really read anythinmg
Derrick Brashear [Thu, 16 May 2002 14:54:42 +0000]
NT-dirpath-copy-in-useful-file-
20020516
basically until NT port has been properly set up to deal with dirpath, make
it work as before
Derrick Brashear [Thu, 16 May 2002 14:47:28 +0000]
afs-attrs-cast-date-to-int-to-avoid-negative-dates-
20020516
cheap hack to avoid negative date problem which occasionally plagues solaris
Nathan Neulinger [Thu, 16 May 2002 13:54:56 +0000]
pioctl-avoid-inadvertantly-setting-new-pags-due-to-gcc-
20020516
some versions of gcc compile the if used to decide whether to allocate a pag
badly. make life easier.
Derek Atkins [Wed, 15 May 2002 21:04:15 +0000]
osi_linux_unmask() needs to explicitly flush_signals() before calling
recalc_sigpending(), otherwise recent kernels don't appear to deliver
new instances of any signals already received.
Revert back to using SIGKILL at shutdown time, but use SIGKILL instead of
explicitly using signal 9, because abstraction is your friend.
Derek Atkins [Tue, 14 May 2002 23:56:40 +0000]
Seems to fix a hang at shutdown time seen on Redhat 7.3 kernels; doesn't
appear to affect older kernels.
Derek Atkins [Tue, 14 May 2002 23:48:35 +0000]
Allows us to build modules for the bigmem kernels redhat started shipping.
Derrick Brashear [Tue, 14 May 2002 21:21:09 +0000]
dir-hold-afs-bufferlock-across-increment-of-buffer-lockers-to-prevent-newslot-from-walking-on-stuff-
20020514
liberal debugging code and hammering produced a failure where newslot was
actually walking on another call to newslot because the bufferlock was
being dropped before newslot incremented lockers on the buffer it
was allocating, allowing someone else to come along and also think they
could allocate it.
by holding the bufferlock just a little longer we avoid this
Garry Zacheiss [Mon, 13 May 2002 02:20:40 +0000]
Update NEWS file with user visible changes in 1.2.4.
Nickolai Zeldovich [Thu, 9 May 2002 21:37:29 +0000]
Avoid kernel panic when dynroot and memcache are used simultaneously.
Derrick Brashear [Thu, 9 May 2002 16:56:46 +0000]
viced-cleanup-dirhandle-references-after-ourselves-
20020509
doesn't fix any problem that i can find, but results in ihandles not
having thousands of bogus references (and may fix problems if refcnt were to wrap)
Garry Zacheiss [Tue, 7 May 2002 21:01:27 +0000]
Use vsu_ExtractName() to get the RW volume name before doing the vid lookup
for addsite and remsite operations, since "vos addsite foo.bar.readonly" is
such a common thinko.
Nickolai Zeldovich [Mon, 6 May 2002 08:13:10 +0000]
Undo need-prototype-for-osi-dnlc-lookup-
20020503, as it created
double inclusion of afs_osidnlc.h.
An equivalent patch was already present in the mainline as part of
linux-dcache-inconsistency-fix-
20020115, but its pullup to the stable
branch, STABLE12-linux-dcache-inconsistency-fix-
20020120, was oddly
missing this particular part of the diff.
Chas Williams [Fri, 3 May 2002 20:03:34 +0000]
Include afs_osidnlc.h for osi_dnlc_lookup() prototype. Apparently
fixes a panic on ia64.
Chas Williams [Thu, 2 May 2002 16:11:14 +0000]
linux-ia64-build-pam-module-
20020502
keep small globals from being gprel
Chas Williams [Thu, 2 May 2002 15:38:16 +0000]
linux-ia64-no-asm-syscall-stubs-
20020502
rewrite syscall stubs not in machine code
Kevin Coffman [Thu, 2 May 2002 07:35:33 +0000]
linux-revalidate-root-always-good-
20020502
avoid ESTALE in 2.4.19-pre7
Chaskiel M Grundman [Tue, 30 Apr 2002 01:24:11 +0000]
README-NT-provide-file-locations-
20020429
give people a clue where to get things from
Srikanth Vishwanathan [Tue, 30 Apr 2002 00:40:22 +0000]
vol-ihandle-cleanup-
20020429
this does fix some potential problems, even if none of them are the
CopyOnWrite problem. basically, ih_reallyclose() could reinsert a now-unref'd
fd handle into the wrong list when cleaning up, and this cleans up the code
considerably.
Chaskiel M Grundman [Mon, 29 Apr 2002 16:31:54 +0000]
darwin-use-namei-fileserver-
20020429
build the right fileserver on darwin
Chaskiel M Grundman [Mon, 29 Apr 2002 16:32:24 +0000]
cvsignore-updates-
20020429
add missed fc_test
Nickolai Zeldovich [Fri, 26 Apr 2002 22:33:55 +0000]
If the inode is not allocated, don't bother updating timestamps.
Pointed out by Frank.Batschulat@Sun.COM.
Nickolai Zeldovich [Fri, 26 Apr 2002 22:27:29 +0000]
Don't spawn a separate kernel thread for rx kernel listener under
Solaris, but instead just use the afsd fork for the listener. Use
of newproc() is deprecated, according to Frank.Batschulat@Sun.COM.
Chaskiel M Grundman [Thu, 25 Apr 2002 22:44:53 +0000]
darwin-map-more-versions-
20020425
5.3 and 5.4 are "Same" enough
Chaskiel M Grundman [Thu, 25 Apr 2002 22:18:23 +0000]
rxkad-stats-initialize-to-placate-darwin-compiler-
20020425
because it deals badly with common symbols
Chas Williams [Tue, 23 Apr 2002 14:45:51 +0000]
linux-ia64-cleanup-casting-
20020423
so we don't need the SYSCALL2POINTER change
Chaskiel M Grundman [Tue, 23 Apr 2002 04:26:36 +0000]
sparc64-linux-and-setgroups32-cleanup-
20020422
update setgroups32 for 2.4 and make sparc64 linux2.4 work again
Chas Williams [Tue, 23 Apr 2002 04:03:29 +0000]
linux-ia64-updates-
20020422
updates to make it work with later 2.4
SYSCALL2POINTER change may be unnecessary (sparc64 way should work)
but without any way to test...
Nickolai Zeldovich [Mon, 22 Apr 2002 18:16:16 +0000]
The heldp argument to h_Lookup should be an int* rather than int,
as pointed out by amu@alum.mit.edu.
Chaskiel M Grundman [Mon, 22 Apr 2002 17:05:10 +0000]
darwin-vnops-make-unlink-fix-fake-references-
20020422
so unlink doesn't panic us
Jeremy Stribling [Mon, 22 Apr 2002 16:42:21 +0000]
libadmin-pts-allow-listing-all-users-and-groups-
20020422
so the java api can take advantage of this
Derrick Brashear [Wed, 17 Apr 2002 23:29:04 +0000]
ubik-allow-netinfo-netrestrict-to-mask-address-pointed-to-by-hostname-
20020417
if you really want to mask the IP address from gethostbyname(gethostname())
there's not really any reason to stand in your way
Nickolai Zeldovich [Mon, 15 Apr 2002 22:36:00 +0000]
Make h_Lookup_r return the host held, and ensure that it's not
deleted at that point; otherwise, the host may be deleted by
another thread between a call to h_Lookup_r and the time when
we actually get around to h_Hold'ing it.
Nickolai Zeldovich [Thu, 11 Apr 2002 08:18:11 +0000]
Pass &treq to afsrename() when preserving a currently-open file.
Nickolai Zeldovich [Tue, 9 Apr 2002 20:21:59 +0000]
Fix more places where afs_EvalFakeStat arguments were out of order
Pavel Semerad [Tue, 9 Apr 2002 19:08:10 +0000]
fakestat-fix-eval-args-ordering-
20020409
fixed bad args ordering in call to afs_EvalFakeStat
Derrick Brashear [Fri, 5 Apr 2002 19:13:32 +0000]
linux-newvcache-more-simplification-
20020405
based on suggestion from vsrikanth@in.ibm.com
eliminate extra for loop, flatten function to drop dentries, fix
locking error i introduced
Derrick Brashear [Thu, 4 Apr 2002 20:03:13 +0000]
solaris-nfstrans-make-it-compile-
20020404
it compiles. it's not pretty, but it does compile
Nickolai Zeldovich [Wed, 3 Apr 2002 07:20:07 +0000]
Clear the RX_CONN_ATTACHWAIT flag when an asymmetric client's calls
time out; otherwise, the client could be locked out until the conn
is recycled.
Nickolai Zeldovich [Tue, 2 Apr 2002 18:35:03 +0000]
Don't make tvc a register, since we want to dereference it later on.
Pointed out by nneul@umr.edu.
Nickolai Zeldovich [Tue, 2 Apr 2002 06:09:48 +0000]
Initial fakestat support (fake mountpoint directory attributes).
Useful for quickly stat'ing everything under /afs, for instance.
Reasonably tested on Solaris, and appears to work on Linux too.
Enable with 'afsd -fakestat'.
Nickolai Zeldovich [Mon, 1 Apr 2002 20:18:50 +0000]
Fix build on Darwin (missing close paren).
David Bailey [Mon, 1 Apr 2002 17:25:37 +0000]
winnt-servermgr-use-existing-creds-option-
20020401
This adds /useexisting to use current AFS creds, bypassing the initial authentication
dialog box. It's useful if you're not running kaserver (i.e. if you've moved
over to kerberos 5) and can't authenticate again by the normal AFS means.
Nickolai Zeldovich [Sat, 30 Mar 2002 18:02:40 +0000]
Add osi_Assert()'s around pthread_{cond,mutex}_* calls to make sure
we aren't getting errors anywhere.
Update the documentation/comments about Rx lock ordering.
Fix possible deadlock in asymmetric client detection code.
Derrick Brashear [Thu, 28 Mar 2002 22:00:55 +0000]
afs-dont-mask-rights-granted-by-lookup-
20020328
lookup is sufficient for this, fix the client not to bogusly
enforce extra restrictions
Derrick Brashear [Thu, 28 Mar 2002 20:43:03 +0000]
linux-dcache-create-negative-dentries-22-needs-while-still-avoiding-vlru-cycle-
20020328
based on patch by Srikanth Vishwanathan and modifications based on suggestions
from Ted Anderson applicable to this part of the code.
the original theory we operated under, namely that negative dentries
could be banished entirely, causes file creation to stop working
under linux 2.2 kernels. reverting that change means we have to deal with
these negative dentries, and Srikanth's original suggested patch does so.
Nathan Neulinger [Mon, 25 Mar 2002 18:35:07 +0000]
fstrace-fix-memory-leak-
20020325
avoid leaking catalog reference on failure
Nathan Neulinger [Mon, 25 Mar 2002 18:30:09 +0000]
cvsignore-update-for-autoconf-25-
20020325
autoconf 2.5x turds to ignore
Derrick Brashear [Mon, 25 Mar 2002 18:09:22 +0000]
convert-linux-macros-for-inodes-appropriately-
20020325
even though right now a vnode and an inode are basically but not exactly
the same, part of the usefulness of this change is the ability to convert
to inodes linked from vcaches later. make the macros fit with this view of the
fture
Johan Danielsson [Mon, 25 Mar 2002 17:39:55 +0000]
convert-vcache-casts-to-macros-
20020325
convert vcache casting in cache manager to macros
should ease further changes later
Nickolai Zeldovich [Sun, 24 Mar 2002 23:39:01 +0000]
Clean up some more ViceLog calls to consistently refer to hosts in a
format like 18.239.2.102:7001, rather than memory address or hex IP.
Nickolai Zeldovich [Sun, 24 Mar 2002 19:21:35 +0000]
Verify that name lengths are at most AFSNAMEMAX and symlink
contents are at most AFSPATHMAX (notably, the Linux kernel
doesn't check this for us); bug report by nneul@umr.edu.
Nickolai Zeldovich [Sun, 24 Mar 2002 19:08:45 +0000]
Initialize tvc in afs_remove so that we don't call afs_PutVCache
on stack garbage in case of some failure.
Nathan Neulinger [Fri, 22 Mar 2002 21:56:44 +0000]
rx-glock-order-headers-correctly-so-its-defined-or-not-consistently-
20020322
fix for earlier change to RX_AFS_GLOCK()
Derrick Brashear [Thu, 21 Mar 2002 18:44:19 +0000]
afs-memcachetruncate-takes-memcacheentry-not-inode-
20020321
prototyping caught this; the usage is inconsistent with other cases and
i believe this is the correct fix. it's not clear to me why we weren't seeing
problems from this, though
Nathan Neulinger [Thu, 21 Mar 2002 18:11:50 +0000]
afs-checkservers-call-getuser-with-cell-index-
20020321
this means to be calling afs_GetUser with a cell index, not with the cell
structure itself
Love Hörnquist-Åstrand [Wed, 20 Mar 2002 18:59:56 +0000]
rxkad-regression-test-
20020228
regression test for rxkad
Mattias Amnefelt [Wed, 20 Mar 2002 18:38:28 +0000]
tru64-avoid-leaking-vnodes-when-failure-
20020320
tru64 needs vrele when an error happens, rather than just returning. make
sure we do so
Nathan Neulinger [Tue, 19 Mar 2002 22:03:30 +0000]
Derrick Brashear [Tue, 19 Mar 2002 22:01:30 +0000]
sys-add-aix-syscall-exports-removed-incorrectly-earlier-
20020319
to export syscalls as functions you need the magic exp file
====================
This delta was composed from multiple commits as part of the CVS->Git migration.
The checkin message with each commit was inconsistent.
The following are the additional commit messages.
====================
and also reinstall them for the user cache manager library