Jim Rees [Thu, 10 Oct 2002 23:45:14 +0000]
openbsd-
20021010
More OpenBSD merge.
Jim Rees [Thu, 10 Oct 2002 22:22:45 +0000]
netsend-proto-and-obsd-includes-
20021010
Put in a prototype for osi_NetSend(), and try to fix all versions
of rx_knet.c. Sorry if I broke yours.
Fix includes for OpenBSD.
Nickolai Zeldovich [Wed, 9 Oct 2002 23:36:54 +0000]
A better way to do cross-cell-only fakestat, which should
also avoid problems with Linux's attribute validation.
Jim Rees [Wed, 9 Oct 2002 19:25:54 +0000]
openbsd-continues-
20021009
More OpenBSD merge.
Please let me know if this breaks other architectures.
Jim Rees [Wed, 9 Oct 2002 16:57:18 +0000]
no-stddef-in-kernel-
20021009
Don't include stddef.h in the kernel.
Untested except on OpenBSD, but I warned the mailing list.
Let me know if this breaks your favorite sysname.
Derrick Brashear [Wed, 9 Oct 2002 16:54:48 +0000]
fileserver-cbs-raise-limit-by-killing-goofy-u-short-indexing-
20021008
i probably missed something here.
====================
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.
====================
i did miss something. also, change FileEntry padding and axe useless comment
====================
try, try again
====================
once more with feeling
====================
make typing of the indexes into timeout[] match
====================
make struct CallBack a multiple of 32 bit so the weird math works again
Hartmut Reuter [Wed, 9 Oct 2002 02:02:47 +0000]
aix51-support-
20021008
aix 5.1 support, not yet ready for prime-time
Derrick Brashear [Wed, 9 Oct 2002 01:54:48 +0000]
osi-wakeup-return-zero-on-doing-a-wakeup-
20021008
based on patch submitted by reuter@rzg.mpg.de, but i actually ported it
to all platforms.
Jim Rees [Tue, 8 Oct 2002 18:48:42 +0000]
curcred-
20021008
Add osi_curcred().
Re-format a bit for readability.
Hartmut Reuter [Tue, 8 Oct 2002 05:15:48 +0000]
aix51-support-
20020926
aix 5.1 support, not yet ready for prime-time
Nickolai Zeldovich [Mon, 7 Oct 2002 10:29:45 +0000]
Make cmdebug -addrs work on platforms other than Solaris.
Nickolai Zeldovich [Mon, 7 Oct 2002 00:00:14 +0000]
Change the meaning of the -fakestat switch to only enable fakestat
for cross-cell mountpoints. The -fakestat-all switch can now be
used to enable fakestat for all mountpoints (old behavior).
Nickolai Zeldovich [Sun, 6 Oct 2002 23:58:32 +0000]
Add JUAFS to the list of things that make clean should remove.
Jim Rees [Thu, 3 Oct 2002 22:44:46 +0000]
openbsd31-cm-
20021003
OpenBSD 3.1 changes for the cache manager
eliminate some unused variables
Derrick Brashear [Thu, 3 Oct 2002 22:24:08 +0000]
fileserver-may-not-have-valid-diskDataHandle-if-volume-was-offlined-
20021003
also avoid potential null deref in ih_open but with the ReadVolumeHeader patch
we should never get there.
thanks to Kris Van Hees for discovering this
Nickolai Zeldovich [Thu, 3 Oct 2002 01:42:43 +0000]
Clean up the #ifdefs around getuerror calls: change KERNEL_HAVE_SETUERROR
to KERNEL_HAVE_UERROR, and use that for both setuerror and getuerror.
Jim Rees [Wed, 2 Oct 2002 22:48:49 +0000]
unused-vars-
20021002
remove (some) unused variables
fix a few compiler warnings
this is part of OpenBSD merge
Jim Rees [Wed, 2 Oct 2002 22:45:19 +0000]
openbsd-mbuf-before-in-
20021002
include mbuf.h before in.h
let me know if this breaks anything
Jim Rees [Wed, 2 Oct 2002 16:58:18 +0000]
openbsd31-cm-
20021002
OpenBSD 3.1 changes for the cache manager
eliminate some unused variables
Nickolai Zeldovich [Wed, 2 Oct 2002 01:48:51 +0000]
Return EINVAL when the user tries to create a FIFO under afs,
instead of silently creating a regular file.
Nathan Neulinger [Tue, 1 Oct 2002 19:09:40 +0000]
linux-dcache-flushing-experimental-
20020819
fix missed ref
Mattias Amnefelt [Tue, 1 Oct 2002 10:51:09 +0000]
dux51-support-
20021001
this was missed. oops.
Love Hörnquist-Åstrand [Tue, 1 Oct 2002 10:50:26 +0000]
configure-check-for-timegm-
20021001
check for timegm. lha needs it for rxkad but it's "safe"
Derrick Brashear [Tue, 1 Oct 2002 04:32:22 +0000]
solaris-dont-truncate-inode-number-
20020930
if ino_t is a larger type than afs_int32, we lose bits, then treat it as
if nothing happened...
Nathan Neulinger [Mon, 30 Sep 2002 20:36:44 +0000]
hpux-rx-cleanup-
20020930
make the hpux client work
Nathan Neulinger [Mon, 30 Sep 2002 20:29:10 +0000]
lock-header-cleanup-
20020930
make sure we include what we mean to
Derrick Brashear [Mon, 30 Sep 2002 20:25:29 +0000]
completion-deal-with-redhat-patch-
20020930
just include the header directly if we caN
Nathan Neulinger [Mon, 30 Sep 2002 20:23:36 +0000]
hpux-bin-is-a-symlink-so-install-elsewhere-
20020930
update install
Nathan Neulinger [Mon, 30 Sep 2002 20:21:40 +0000]
handle-cow-differently-if-volume-wont-be-salvaged-
20020930
as edited by shadow@dementia.org
Derrick Brashear [Mon, 30 Sep 2002 19:38:11 +0000]
wakeup-only-returns-int-on-aix-
20020930
reported by several people.
Derrick Brashear [Fri, 27 Sep 2002 01:51:47 +0000]
hpux-updated-rx-kmutex-from-ibm-
20020926
as provided by Srikanth Vishwanathan
Nathan Neulinger [Fri, 27 Sep 2002 00:42:32 +0000]
hpux-kmutex-cleanup-
20020926
clean up missed and won't-compile stuff
Derrick Brashear [Thu, 26 Sep 2002 20:48:30 +0000]
hpux-rx-kmutex-h-
20020926
basically verbatim import from ibm afs, per laura stentz:
Derrick,
I have gotten approval to allow you to publish this header in OpenAFS.
Enjoy!
Laura
Mattias Amnefelt [Thu, 26 Sep 2002 08:13:06 +0000]
tru64-51-gets-a-sysname-
20020926
the binaries from 5.0 don't work, so...
Nathan Neulinger [Thu, 26 Sep 2002 08:09:14 +0000]
xdr-build-ours-entirely-
20020926
If we're doing our own xdr, we should be doing ALL of of own xdr, not piecemeal. This impacted on HP w/ duplicate symbols if
-lnsl, and missing symbols if no -lnsl.
Nathan Neulinger [Thu, 26 Sep 2002 08:07:07 +0000]
shlib-linker-handle-objdir-builds-on-solaris-
20020926
at least he notices when i screw this up
Hartmut Reuter [Thu, 26 Sep 2002 08:01:05 +0000]
aix51-support-
20020926
aix 5.1 support, not yet ready for prime-time
David Howells [Thu, 26 Sep 2002 07:39:46 +0000]
linux-init-rwlocks-
20020926
init linux rwlocks we missed before
Nick Ingolia [Thu, 26 Sep 2002 07:33:16 +0000]
linux-pag-deal-with-no-groups-
20020926
If ngroups in the parent is 0, but the first two elements of the groups
array represent a valid PAG (perhaps because they weren't zeroed), then
set_pag_in_parent won't add 2 to ngroups and the parent process won't
join the PAG.
Garry Zacheiss [Thu, 26 Sep 2002 04:47:32 +0000]
Recognize MacOS X 10.2.1 as ppc_darwin_60. It's compatible.
Kris Van Hees [Thu, 26 Sep 2002 03:57:34 +0000]
salvager-dont-mask-uniq-comparison-in-vnodes-only-in-inodes-
20020920
the masking operation is for inodes when using "inode spares" fields which
overflow, not for vnode operations.
====================
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.
====================
remove an extraneous right paren
Garry Zacheiss [Thu, 26 Sep 2002 00:38:28 +0000]
Update NEWS file with user visible changes and other noteworthy things
about OpenAFS 1.2.7.
Jim Rees [Fri, 20 Sep 2002 17:32:11 +0000]
update-openbsd-31-
20020920
update OpenBSD to 3.1
Derrick Brashear [Sat, 14 Sep 2002 00:52:57 +0000]
nfs-trans-comment-unused-variables-
20020913
make solaris xlator module load
Derrick Brashear [Fri, 13 Sep 2002 03:48:45 +0000]
rxkad-client-offer-better-bad-token-warnings-
20020912
if it's too large to be a valid token, offer a useful error.
Derrick Brashear [Wed, 11 Sep 2002 17:15:31 +0000]
linux-use-kernel-threads-and-completions-if-possible-
20020911
configure glue is all my fault
====================
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.
====================
with minor rewriting by shadow@dementia.org
====================
rest of the configure glue
====================
fix ifdef to encapsulate correct code
====================
fix ifdef
====================
and deal with change to afsd ops
====================
sys_exit not exported everywhere
Michael Niksch [Wed, 11 Sep 2002 09:16:06 +0000]
aix-namei-fileserver-fix-
20020911
make namei fileserver work on aix
Derrick Brashear [Wed, 11 Sep 2002 08:19:57 +0000]
ptserver-acl-dumpentry-to-disallow-access-based-on-record-
20020911
tighten up acl enforcement
Nathan Neulinger [Tue, 10 Sep 2002 22:41:18 +0000]
parisc-linux-kernel-headers-
20020909
link in stuff from asm
Derrick Brashear [Tue, 10 Sep 2002 05:00:52 +0000]
evalmountpoint-save-stack-
20020909
use "small" memory pool instead of 128 bytes off the stack
Derrick Brashear [Tue, 10 Sep 2002 02:12:54 +0000]
shlib-updates-
20020909
based on reports from nneul@umr.edu
====================
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.
====================
fix dest for real
====================
based on reports from nneul@umr.edu
====================
based on reports from nneul@umr.edu
====================
fix link lines for real
====================
install everything in toplibdir
Nathan Neulinger [Tue, 10 Sep 2002 00:43:24 +0000]
parisc-linux-lwp-update-
20020909
gets all the test cases working
Derrick Brashear [Mon, 9 Sep 2002 18:00:40 +0000]
uss-fix-typo-
20020909
reported anonymously
Derrick Brashear [Fri, 6 Sep 2002 03:44:03 +0000]
revert-kaserver-set-realm-name-in-tgts-
20020708
based on feedback from Pavel Semerad
fair warning; without this if you live in a world with multiple kdc vendors in
the same realm serving the same data you will be sad. so don't.
Derrick Brashear [Wed, 28 Aug 2002 07:05:43 +0000]
shared-libafsrpc-and-libafsauthent-
20020828
shared library versions of libafsrpc and libafsauthent
Derrick Brashear [Wed, 28 Aug 2002 05:30:06 +0000]
linux-tasklist-lock-export-detection-
20020829
if you don't have a versioned kernel, so sorry.
====================
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.
====================
add macro i forgot to commit
Derrick Brashear [Tue, 27 Aug 2002 23:18:50 +0000]
pin-function-definition-and-ifdef-cleanup-
20020827
define if kernel uses pin/unpin to pin memory; cleanup ifdef usage
Derrick Brashear [Tue, 27 Aug 2002 22:42:25 +0000]
cachemgr-setuerror-ifdef-cleanup-
20020827
feature-ifdef for setuerror, and clean up where it's called
Derrick Brashear [Tue, 27 Aug 2002 21:46:16 +0000]
namei-close-directory-behind-us-always-
20020827
more or less verbatim as suggested by kolya@mit.edu
avoid leaking directory fds if we get an error
Nathan Neulinger [Mon, 26 Aug 2002 22:04:48 +0000]
prototypes-empty-files-to-diff-against-
20020826
literally empty
Derrick Brashear [Mon, 26 Aug 2002 22:01:00 +0000]
packaging-rationale-
20020826
why we bothered to put it here
Sam Hartman [Mon, 26 Aug 2002 21:52:02 +0000]
debian-packagings-
20020826
packagings for debian.
Derek Atkins [Mon, 26 Aug 2002 21:46:51 +0000]
redhat-packagings-
20020826
afsmodname actually from jhutz@cmu.edu but i don't feel like committing twice
Chaskiel M Grundman [Mon, 26 Aug 2002 21:24:11 +0000]
packaging-add-macos-bits-
20020826
the macos x packaging
Nathan Neulinger [Mon, 26 Aug 2002 16:48:50 +0000]
linux-kdump-cleanup-build-process-
20020826
so we don't remake this on each make
Nathan Neulinger [Mon, 26 Aug 2002 16:47:26 +0000]
fs-prototyping-
20020826
prototype the fs command source
Nathan Neulinger [Mon, 26 Aug 2002 16:46:32 +0000]
viced-prototypes-
20020826
prototyping for viced/callback.c
Nathan Neulinger [Mon, 26 Aug 2002 16:45:28 +0000]
vos-move-more-diagnostics-
20020826
in verbose mode, print better diagnostics
Derrick Brashear [Mon, 26 Aug 2002 16:33:03 +0000]
rpc-authent-shlib-support-mostly-
20020826
necessary bits to build shlibs, except since we can't easily decide
when to deal, we don't, yet
Derrick Brashear [Mon, 26 Aug 2002 15:09:35 +0000]
osconf-cleanup-
20020826
don't set shlib stuff in makefile fragments. do include sun4_413 in
osconf
Derrick Brashear [Mon, 26 Aug 2002 15:01:19 +0000]
libafsrpc-use-our-xdr-for-consistency-
20020826
suggested by Srikanth Vishwanathan
Derrick Brashear [Fri, 23 Aug 2002 05:52:56 +0000]
prototypes-fixes-
20020821
add externs removed and missed
====================
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.
====================
ok, another try. don't include protosw.h here and let sysincludes.h pull it in
====================
one more missed extern
====================
keep trying...
====================
fix weird redefintion problem with kmem_alloc
====================
can't know this sort of thing (and it's wrong at least sometimes)
Garry Zacheiss [Fri, 23 Aug 2002 05:38:53 +0000]
dux-juafs-make-it-compile-
20020823
fix this so make will bother
Derrick Brashear [Fri, 23 Aug 2002 05:29:38 +0000]
dux-call-flushpages-with-cred-
20020823
we weren't passing in creds before. do so.
Nathan Neulinger [Fri, 23 Aug 2002 04:36:29 +0000]
hpux-dont-build-login-
20020822
don't build login on hpux
Nickolai Zeldovich [Fri, 23 Aug 2002 04:31:38 +0000]
This is mostly a rewrite of src/afs/afs_cell.c, and associated changes
to other things. Features of the new cell code:
* Persistent name-to-cell-number mapping across reboots, to ensure the
cell numbers in the dcache are meaningful. Stored in the new cache
file, /usr/vice/cache/CellInfo.
* Allow cmdebug to query the cell number to cell name mapping.
* Allow clients to have an empty CellServDB as long as AFSDB is enabled
and ThisCell has AFSDB records.
* The home cell retains the setuid bit even if looked up via AFSDB.
* Dynroot gets its own cell, rather than piggybacking on cell 1.
* Cell 1 is no longer special; cell numbers are now semi-opaque.
* Convert cell traversal code to using GetCellByIndex rather than
directly poking at CellLRU / afs_xcell.
* Separate cells from aliases, which shouldn't have been the same
in the first place.
* Cleaner code (IMHO).
====================
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.
====================
make the arguments afs_CellNumValid takes match the prototype (since it
really is a afs_int32 it looks like)
====================
update irix to deal
====================
update irix to deal
====================
Fix some more references to the now-nonexistant cell->index
Derrick Brashear [Thu, 22 Aug 2002 23:37:11 +0000]
prototypes-fixes-
20020821
hopefully last of the lwp fun
====================
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.
====================
make it build on the bsds
====================
nope. this time for sure (more lwp createprocess casting)
====================
clean up osi_UFSOpen prototyping
====================
aix gets bitter at this
====================
fix missed argument
====================
again, aix doesn't like this. so, we don't do it
====================
yup, aix still hates this
====================
dux compiler doesn't like this (i think this one is actually my fault)
====================
already prototyped in rxkad_prototypes.h
====================
causes multiple inclusion of protosw.h which isn't safe on irix.
====================
for some reason you just can't declare the same thing the same way twice
====================
make dux compiler happy. oops.
Derrick Brashear [Thu, 22 Aug 2002 22:59:12 +0000]
dux-force-lex-compatible-lex-
20020822
hopefully to correct problem observed by garry zacheiss
Nickolai Zeldovich [Thu, 22 Aug 2002 20:43:30 +0000]
Drop GLOCK over the entire loop iterating over the iovec, since
it's not needed for any operations performed in the loop (rather
than dropping it just for the memcpy inside the loop).
Derrick Brashear [Thu, 22 Aug 2002 19:45:11 +0000]
prototypes-fixes-
20020821
screw it. if winnt can have generic pointers, so can we
====================
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 cast all the pointers
Derrick Brashear [Thu, 22 Aug 2002 18:53:43 +0000]
vsprocs-checkvolume-bug-
20020822
if we're going to make a string we should make it where we expect it
Derrick Brashear [Thu, 22 Aug 2002 18:02:03 +0000]
sunos-support-cleanup-
20020822
some things want ucontext.h. leave it there
Nickolai Zeldovich [Thu, 22 Aug 2002 18:01:51 +0000]
Add two useful functions for string operations: afs_strdup() and
afs_osi_FreeStr(). Will be used by an upcoming patch.
Nickolai Zeldovich [Thu, 22 Aug 2002 18:00:15 +0000]
Make "fs newcell" work even when there are no pre-existing cells in
the kernel yet. Previously, newcell tried to issue a GetCell pioctl
to figure out what version of the pioctl interface the kernel is
using. Lacking such information (if the pioctl returns an error),
we now assume the newer interface.
Chaskiel M Grundman [Thu, 22 Aug 2002 17:19:38 +0000]
struct-buf-hack-cleanup-
20020822
get rid of bogus struct and just declare it
Derrick Brashear [Thu, 22 Aug 2002 17:14:14 +0000]
solaris-make-cleanup-
20020822
avoid make confusion about libafs target
Chaskiel M Grundman [Thu, 22 Aug 2002 01:15:09 +0000]
darwin-head-build-fixes-
20020821
make the head build on darwin
Chaskiel M Grundman [Thu, 22 Aug 2002 01:13:09 +0000]
prototypes-fixes-
20020821
fix the following problems
- including osi_vfs.h on almost all platforms, even though afsincludes.h
already deals with it
- universally declaring afs_globalVFS as a struct vfs *
- declaring afs_stats_XferSumBytes in a header without a storage class
- using afsincludes.h without sysincludes.h
- make clean removes a source file in rxkad
====================
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.
====================
irix needs this, put it back
====================
cast parm argument to CreateProcess
====================
more lwp createprocess fun
====================
more lwp
Derrick Brashear [Thu, 22 Aug 2002 01:05:23 +0000]
osi-vfs-h-only-on-linux-
20020821
based on report from Chaskiel Grundman
no osi_vfs.h on anything but linux
Chaskiel M Grundman [Wed, 21 Aug 2002 23:07:52 +0000]
darwin-updates-
20020821
should be correct support for 10.2. also updates for 10.1
Derrick Brashear [Wed, 21 Aug 2002 22:12:58 +0000]
ptserver-catch-empty-idlist-
20020831
based on report from joda+openafs@pdc.kth.se
consistent with what the opposite transform does
Hartmut Reuter [Wed, 21 Aug 2002 21:57:35 +0000]
int64-operator-updates-
20020821
Changing the length and offset fields from afs_int64 to afs_uint64 gave
strange results for CompareInt64 and SubtractInt64. Therefore it's
better to cast the values as signed in these macros.
Nathan Neulinger [Wed, 21 Aug 2002 21:34:28 +0000]
eliminate-nested-comments-
20020821
cleanup config fragments
Nathan Neulinger [Wed, 21 Aug 2002 21:32:19 +0000]
viced-provide-way-to-not-retraverse-hostlist-
20020821
based on thomas muellers patch
Nathan Neulinger [Wed, 21 Aug 2002 21:30:00 +0000]
ukernel-updateosi-read-write-for-prototypes-
20020821
cleanup
Nathan Neulinger [Wed, 21 Aug 2002 21:02:02 +0000]
volser-prototypes-
20020831
vsprocs prototyping
Derrick Brashear [Wed, 21 Aug 2002 20:57:45 +0000]
linux-22-kernel-support-for-current-state-
20020821
based on report from alfw@SLAC.Stanford.EDU
Derrick Brashear [Wed, 21 Aug 2002 20:49:04 +0000]
bosserver-dont-create-unused-directories-
20020821
in a build with alternate dirpaths we shouldn't create directories we
don't use
Love Hörnquist-Åstrand [Wed, 21 Aug 2002 20:33:15 +0000]
print-cause-of-bosserver-whining-
20020821
instead of a vague error be slightly more useful
Nathan Neulinger [Wed, 21 Aug 2002 20:26:15 +0000]
volser-vsprocs-call-enumerateentry-correctly-
20020821
prototyping caught this