Derrick Brashear [Fri, 1 Dec 2006 20:45:28 +0000]
DEVEL15-openafs-1-5-12-
20061201
1.5.12
Derrick Brashear [Fri, 1 Dec 2006 18:50:51 +0000]
DEVEL15-give-s390x-syscall-table-the-big-hammer-
20061201
Neale Ferguson contributed the assembler.
The false illusion of security some people want to play under is usually good for wasting a few hours; it was here. making the sys_call_table read nly for real of course fixes it, let's just hope keyrings come along before then on such platforms.
(cherry picked from commit
b81763506fdf29f3df66d811e2d299cc4c9acec0)
Derrick Brashear [Fri, 1 Dec 2006 18:35:33 +0000]
DEVEL15-redhat-packaging-updates-
20061201
updates from derek
updates for s390x
updates to handle suse
(cherry picked from commit
9ba1d32b4ba8aac1d7e2c7f7489739e0e4ca5475)
Derrick Brashear [Fri, 1 Dec 2006 13:00:44 +0000]
DEVEL15-redhat-initscripts-update-for-suse-
20061201
suse lacks /etc/init.d/functions
(cherry picked from commit
af6af0bbc4d874a42f0640eaafd762ebfc1bc872)
Jeffrey Altman [Thu, 30 Nov 2006 06:50:58 +0000]
DEVEL15-windows-notes-
20061130
updates for 1.5.12
Jeffrey Altman [Thu, 30 Nov 2006 06:46:28 +0000]
DEVEL15-STABLE15-windows-version-
20061130
1.5.1200 = 1.5.12
Jeffrey Altman [Thu, 30 Nov 2006 04:40:07 +0000]
DEVEL15-windows-release-notes-
20061129
1.5.12 Release Notes
Derrick Brashear [Wed, 29 Nov 2006 21:25:09 +0000]
DEVEL15-cprivate-ticketlen-is-32-not-16-
20061129
everyone else thinks ticketLen is 32 bits. contrary to what arby's says, different is not necessarily better.
(cherry picked from commit
45425b2f222d5c0a8dd56695587b6cdeba581b6a)
Jeffrey Altman [Wed, 29 Nov 2006 06:23:44 +0000]
DEVEL15-windows-dirty-buffer-optimization-
20061128
The old dirty buffer synchronization algorithm had a buf_IncrSyncer
thread walking the all buffer list periodically searching for dirty
buffers to write to the file server. This had several negative
results. The alogirithm ate up ever increasing amounts of CPU time
even when AFS is idle as the size of the cache increases. Also,
buffers were written to the file server in an order based upon the
original buffer allocation which has nothing to do with the order
in which the buffers became dirty.
The new algorithm maintains a dirty buffer list. Items are added
when the buffer is originally marked dirty. A buffer is only
removed from the list by the buf_IncrSyncer when the buffer is no
longer dirty. If the list is empty the thread goes back to thread
immediately without additional processing requirements.
(cherry picked from commit
d253bde4574e34ee08cf326ec4c481b57c230476)
Jeffrey Altman [Tue, 28 Nov 2006 09:29:28 +0000]
DEVEL15-volser-cleanup-
20061128
cleanup the return code handling from
DELTA volser-dump-validate-input-
20060417
Jeffrey Altman [Tue, 28 Nov 2006 09:17:14 +0000]
DEVEL15-windows-smb-dont-crash-vista-
20061128
Unlike previous versions of the OS, Vista performs a shutdown on
the Microsoft Loopback adapter just like it would on a real adapter.
This causes the smb_Listener threads to trigger a panic during a
suspend/hibernate power event.
The fix is to unbind from the network adapters in response to a
suspend/hibernate power event and then rebind to the adapters
when a resume power event is received. Note that the resume events
are not reliably delivered so it is possible the afs service will
not be accessible. However, this is the best we can do.
(cherry picked from commit
1ed0460435ec05150526c951d1ff32dd3b5a9c39)
Jeffrey Altman [Tue, 28 Nov 2006 09:08:04 +0000]
DEVEL15-windows-conn-analyze-
20061128
force an rx connection reset if VICECONNBAD or VICETOKENDEAD are
received
(cherry picked from commit
cb1997aad8a62b31f1e890dfe093a2d4845e0086)
Jeffrey Altman [Tue, 28 Nov 2006 09:01:14 +0000]
DEVEL15-windows-afslogon-dont-modify-environment-
20061128
don't allow environment variables we set to be inherited by child
processes
(cherry picked from commit
97733f8dcfd08968da5173ceb5976143728b5eab)
Russ Allbery [Tue, 21 Nov 2006 00:56:56 +0000]
DEVEL15-thiscell-whitespace-
20061120
Be more liberal when parsing ThisCell. Accept and ignore leading and
trailing whitespace and anything after the first whitespace character on
the first line. Return an error for a read error or for an empty cell
name.
(cherry picked from commit
769f4e49b60b15b27eb2898a5b28c2d99fc0238c)
Russ Allbery [Mon, 20 Nov 2006 23:49:18 +0000]
DEVEL15-relative-path-canonicalization-
20061120
Path canonicalization for commands such as bos getlog was only applied for
absolute paths; relative paths were still constructed relative to the
canonical directory. Modify the path canonicalization routines to also
canonicalize the base directory for relative paths.
(cherry picked from commit
778e2d5f18c1d0c93a073c137bb36e56f64747ef)
Jeffrey Altman [Mon, 20 Nov 2006 18:54:31 +0000]
DEVEL15-windows-netidmgr-plugin-vista-vs2005-
20061120
Define _USE_32BIT_TIME_T since that is what tokens contain.
Use cflagsdll instead of cflagsmt to avoid multiply defined entry points
(cherry picked from commit
848d2f1caa8d196831a1fe599694fdce0018136d)
Jeffrey Altman [Mon, 20 Nov 2006 18:04:53 +0000]
DEVEL15-windows-remove-unused-libs-
20061120
rpcndr.lib and largeint.lib are no longer used.
rpcndr.lib no longer exists in the Vista SDK.
(cherry picked from commit
4a3482f9e6b5fdac731da45cc4f4179fd5db7046)
Jeffrey Altman [Sun, 12 Nov 2006 00:22:59 +0000]
DEVEL15-windows-notes-
20061111
more notes for 1.5.11
Jeffrey Altman [Sat, 11 Nov 2006 07:13:48 +0000]
DEVEL15-windows-smb-rename-offline-folders-
20061111
FIXES 45692
when offline folders are enabled, the old file name sent in a rename
operation is sent in all uppercase even when the file name is not.
this patch attempts a case insensitive match after the case sensitive
match fails.
(cherry picked from commit
7fff96e0cf113a34d3001282807b6bd04375f5b1)
Derrick Brashear [Fri, 10 Nov 2006 04:36:34 +0000]
DEVEL15-revert-inactivevcache-hold-lock-
20061109
fails verification. come back to it
(cherry picked from commit
1d6da37c594e3f379473aebc2a7345dc41bfff15)
Derrick Brashear [Fri, 10 Nov 2006 04:11:33 +0000]
DEVEL15-macos-preupgrade-script-
20061109
update the preupgrade script
(cherry picked from commit
6766285fc24e8ebc087e57085097106e1bb61425)
Chas Williams [Fri, 10 Nov 2006 04:05:31 +0000]
DEVEL15-kill-stat-blksize-
20061109
FIXES 42663
don't provide a bogus hint
(cherry picked from commit
2110d5e843aef69b0753370cd5ad705b5d5afb8e)
Derrick Brashear [Fri, 10 Nov 2006 00:22:27 +0000]
DEVEL15-openafs-1-5-11-
20061109
make 1.5.11
Derrick Brashear [Fri, 10 Nov 2006 00:20:19 +0000]
DEVEL15-leopard-updates-
20061109
make things work with current seeds
(cherry picked from commit
24487b5ba24ef5a5d6ee269e880dbcdfd2b7cc6d)
Chas Williams [Fri, 10 Nov 2006 00:11:01 +0000]
DEVEL15-linux-blkbits-
20061109
FIXES 42671
set blkbits if we have it
(cherry picked from commit
789bf060d28465edc996b07b7f4bc21fc051515d)
Russ Allbery [Fri, 10 Nov 2006 00:01:26 +0000]
DEVEL15-aklog-doc-improvements-
20061105
Document (at least partially) AFS's mapping of Kerberos v5 principal names
to Kerberos v4 format in the aklog man page. Also document that -setpag
may not always work.
(cherry picked from commit
46cfe728987b8d2f45a7dc718234f83fccbda9eb)
Russ Allbery [Thu, 9 Nov 2006 23:58:43 +0000]
DEVEL15-afsd-shutdown-doc-improvement-
20061105
Change the CAUTIONS about afsd -shutdown to be less dire and more accurate,
as this now mostly works on Linux.
(cherry picked from commit
f82012552166a51edf6931b05abd0380c429d6b6)
Marc Dionne [Thu, 9 Nov 2006 23:43:29 +0000]
DEVEL15-linux-statfs-dentry-
20061109
FIXES 42788
statfs changed in recent linux kernels. cope.
(cherry picked from commit
c12c958febfd6f7667345baec164b617f44f8bb9)
Chas Williams [Thu, 9 Nov 2006 23:18:45 +0000]
DEVEL15-linux-config-h-died-
20061109
FIXES 42662
config.h vanished. deal.
(cherry picked from commit
ee8ae61a4d9610fea16473d87b471cd4fc128804)
Chas Williams [Thu, 9 Nov 2006 23:03:15 +0000]
DEVEL15-inactivevcache-hold-lock-
20061109
FIXES 42798
hold a lock while we are in InactiveVCache
(cherry picked from commit
6975fcb202f2a599814ce9f26cda1999163050f8)
Chas Williams [Thu, 9 Nov 2006 22:56:11 +0000]
DEVEL15-linux26-disable-backing-readahead-
20061109
FIXES 42797
disable vm readahead, we don't want it
(cherry picked from commit
dd508cf7b4553b13872335ec1eb84401bc61435f)
Jeffrey Altman [Thu, 9 Nov 2006 21:52:21 +0000]
DEVEL15-windows-notes-
20061109
notes for 1.5.11
Jeffrey Altman [Thu, 9 Nov 2006 21:35:43 +0000]
DEVEL15-windows-release-notes-
20061109
updates for 1.5.11
Jeffrey Altman [Thu, 9 Nov 2006 21:29:21 +0000]
DEVEL15-windows-version-
20061109
1.5.11 == 1.5.1100
Jeffrey Altman [Thu, 9 Nov 2006 21:26:51 +0000]
DEVEL15-windows-buf-flush-
20061109
FIXES 44330
The return value of buf_CleanAsync is not an error value but a flag
indicating whether or not the buffer was in fact dirty. Do not treat
it as an error. The error state is stored in the buffer itself.
(cherry picked from commit
398ceb0775a0663829ce4c44a8ec08d9e63bd1ca)
Jeffrey Hutzelman [Thu, 9 Nov 2006 02:16:23 +0000]
DEVEL15-bozo-avoid-garbage-in-allprocs-
20061108
if we error, don't put garbage in allProcs
(cherry picked from commit
ea4e9f1841dccebf37e9287334442a1ac2f928b3)
Jeffrey Hutzelman [Wed, 8 Nov 2006 18:57:45 +0000]
DEVEL15-amd64-better-range-checking-
20061108
FIXES 44198
check entire mapped space
(cherry picked from commit
9ea216fc9837eb6922ce365aa3eedafdb8564d94)
Jeffrey Altman [Sat, 4 Nov 2006 23:55:22 +0000]
DEVEL15-rxkad-server-bad-ticket-part-two-
20061103
FIXES 43862
Ensure that tkt_DecodeTicket and rxkad_CheckResponse return the right
RXKAD errors for ticket expiration or invalidity. Avoid calling
tkt_CheckTimes twice in rxkad_CheckResponse
(cherry picked from commit
bf5bfc06638b028335ecdc19ce35315195fc6ffe)
Jeffrey Altman [Sat, 4 Nov 2006 15:39:01 +0000]
DEVEL15-rxkad-server-bad-ticket-
20061103
FIXES 43862
The file server has been returning RXKADNOAUTH when the time between
the client and server is out of sync. tkt_CheckTimes returns -1 for
recently expired tickets and -2 for tickets that have been expired for
a while. In the -2 case we must return RXKADBADTICKET not RXKADNOAUTH.
(cherry picked from commit
799caa43b5417b40e43f38a690962b1d53bca374)
Jeffrey Altman [Fri, 3 Nov 2006 07:57:15 +0000]
DEVEL15-windows-afs-config-stat-entries-
20061103
raise max stat entries in the control from 10000
to 262144
(cherry picked from commit
8dd192a8dd5b516aaa6f1f74d7c1979b004e358f)
Jeffrey Altman [Mon, 23 Oct 2006 05:08:48 +0000]
DEVEL15-windows-readme-
20061023
(cherry picked from commit
eecdb30c47b4c1a12011d2bf9bf0b6eb07797427)
Jeffrey Altman [Sun, 22 Oct 2006 13:25:38 +0000]
DEVEL15-windows-smb_fid_t-deadlock-
20061022
smb_ReleaseFID cannot be called while a cm_scache_t->mx is held
shuffle the order of the smb_ReleaseFID calls so they are always after
cm_XXXRelease calls for performance.
(cherry picked from commit
c484781531ce29d3d1b5c3753322be4a87dd0841)
Jeffrey Altman [Sun, 22 Oct 2006 03:29:58 +0000]
DEVEL15-windows-delonclose-
20061021
set delonclose after obtaining the smb_fid_t
(cherry picked from commit
77d210cde774995bffdebf3347847afebd6f7cfd)
Tom Keiser [Sun, 22 Oct 2006 03:00:12 +0000]
DEVEL15-parallel-attaching-
20061021
FIXES 43332
Some sites expect single-threaded volume attaches to happen in a specific
order. The parallel attach patch also changed the attach order for sites
not specifying -vattachpar.
(cherry picked from commit
e6b719a4c0a33b0b020c07c20124a40a245dbe9e)
Jeffrey Hutzelman [Sun, 22 Oct 2006 02:53:46 +0000]
DEVEL15-kreltime-
20061021
FIXES 43551
kreltime-guard-against-null-tm-
20060731 was apparently an
attempt to deal with the possibility of localtime() returning NULL by
using localtime_r instead, and zeroing the tm structure first. This has
two problems. First, localtime_r is not guaranteed to leave its output
buffer untouched or in a sane state in the event it is given invalid
input. More importantly, the second half of this delta fixed the build
on Windows (which apparently lacks localtime_r) at the expense of using
the wrong level of indirection on other platforms.
Simple is nice, but correct is important too. There is nothing wrong
with switching to localtime_r to avoid thread safety issues. However,
the correct fix for the anticipated problem was and is to check the
return value of localtime (or localtime_r), and if it is NULL, fill in
the ktime_date with suitable default values.
The attached patch does not implement the correct fix, but does fix the
double-indirection problem, so butc will again perform correctly instead
of crashing.
(cherry picked from commit
e720779b7c8472a0d928b0e1429b476f073309aa)
Jeffrey Altman [Sat, 21 Oct 2006 21:55:41 +0000]
DEVEL15-windows-version-
20061021
1.5.10 == 1.5.1000
Jeffrey Altman [Sat, 21 Oct 2006 21:53:11 +0000]
DEVEL15-windows-notes-
20061021
1.5.10
Jeffrey Altman [Sat, 21 Oct 2006 21:49:45 +0000]
DEVEL15-windows-release-notes-
20061021
1.5.10
Jeffrey Altman [Sat, 21 Oct 2006 21:47:48 +0000]
DEVEL15-windows-misc-
20061021
* don't DebugBreak unless a debugger is attached
* replace afsdcell.ini with CellServDB in error message
* do not NetBIOS Reset adapters when checking to see if only the loopback
is in use because doing so breaks the existing LANA bind
* remove a deadlock in smb3.c
* do not retry dirty buffer writes when flushing in response to a suspend
* do not flush in response to a suspend when only the loopback is present
(cherry picked from commit
4f26bcfdb3c6a47d17f825e99f9d5a1bc97eba45)
Jeffrey Altman [Tue, 17 Oct 2006 15:35:29 +0000]
DEVEL15-windows-version-
20061017
1.5.903 == 1.5.9a
Jeffrey Altman [Tue, 17 Oct 2006 01:38:52 +0000]
DEVEL15-windows-version-
20061016
1.5.1000 == 1.5.10
Jeffrey Altman [Tue, 17 Oct 2006 01:32:57 +0000]
DEVEL15-windows-dbg-refcount-
20061016
more code to assist in debugging scache refcount leaks
(cherry picked from commit
740013fcbf2273287414dd399b3020ed09cf07ba)
Jeffrey Altman [Tue, 17 Oct 2006 01:21:36 +0000]
DEVEL15-windows-scache-ref-leak-
20061016
more reference count leaks
(cherry picked from commit
6bc62e6c906174ef76759d77b3c6c9067a424455)
Jeffrey Altman [Mon, 16 Oct 2006 15:17:52 +0000]
DEVEL15-windows-notes-
20061016
more notes for 1.5.9
Jeffrey Altman [Mon, 16 Oct 2006 15:13:09 +0000]
DEVEL15-windows-release-notes-
20061016
add info on netidmgr plug-in
Jeffrey Altman [Mon, 16 Oct 2006 14:55:34 +0000]
DEVEL15-windows-netidmgr-plugin-
20061016
fixup build dependencies for afsp_version.h
====================
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.
====================
fixup copyrights
(cherry picked from commit
5aa0347678c472d21b99af516c64578c4af41ea9)
Jeffrey Altman [Mon, 16 Oct 2006 14:42:13 +0000]
DEVEL15-windows-version-
20061016
1.5.902 == 1.5.9
Derrick Brashear [Fri, 13 Oct 2006 20:43:45 +0000]
DEVEL15-audit-one-more-thing-
20061013
it never ends
(cherry picked from commit
3e10dc3a84037bf86a744c6631854b580d5fab4b)
Derrick Brashear [Fri, 13 Oct 2006 18:39:06 +0000]
DEVEL15-audit-remove-nested-valists-
20061013
last straw
make amd64, ppc happy for sure and for ever
(cherry picked from commit
e3d10f048b3cbaa278f8213d8de6874db80f7658)
Tom Keiser [Fri, 13 Oct 2006 14:09:03 +0000]
DEVEL15-aix-cleanup-
20061013
fixes for aix pre 5.2 (statfs64), remove a c++ style comment (makes xlc sad) and
make clean target clean
(cherry picked from commit
453abf6aa149e27622745b818afb88019e22171a)
Derrick Brashear [Fri, 13 Oct 2006 06:12:51 +0000]
DEVEL15-uae-translate-create-
20061013
really the stubs should probably just translate the codes back. maybe for 1.6
(cherry picked from commit
0e8238e7cb90c022ec126c05987d8258197a413a)
Jeffrey Altman [Fri, 13 Oct 2006 00:28:28 +0000]
DEVEL15-windows-netidmgr-plugin-
20061012
afsp_version.h is auto-generated
====================
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 product version generation
(cherry picked from commit
c8a86da7f76a8e71d883cda39c10251b0910302c)
Jeffrey Altman [Thu, 12 Oct 2006 22:42:20 +0000]
DEVEL15-windows-kfw-sdk-
20060921
kfw 3.1
(cherry picked from commit
faee281941502211c3ea8a62ffea85d934d9ba3a)
Asanka Herath [Thu, 12 Oct 2006 22:19:35 +0000]
DEVEL15-windows-netidmgr-plugin-
20061010
Secure Endpoints Inc. is contributing its AFS plug-in for KFW 3.1's
Network Identity Manager to OpenAFS under an MIT style license.
(cherry picked from commit
f72c2fc3e75a13e3736a275df35597ad43f620fa)
Jeffrey Altman [Thu, 12 Oct 2006 06:59:25 +0000]
DEVEL15-windows-scache-ref-leak-
20061012
make sure the scache mutex is held across calls to cm_SyncOpDone
(cherry picked from commit
2b96ba59194831ef298585c4f82d574a0790f3a6)
Derrick Brashear [Thu, 12 Oct 2006 05:24:43 +0000]
DEVEL15-auditlog-sublist-fix-
20061012
FIXES 35559
update so we don't botch the va_list child
(cherry picked from commit
e8a6b9977e637e304017cae8ac6d3fd03501b376)
Derrick Brashear [Thu, 12 Oct 2006 03:58:05 +0000]
DEVEL15-linux-aio-functions-improved-test-
20061011
detect properly if we have nothing
(cherry picked from commit
d71d42d29f923750be4f4f518b8d0e72dba8fa23)
Chaskiel M Grundman [Wed, 11 Oct 2006 14:18:10 +0000]
DEVEL15-version-string-generation-consistency-
20061011
FIXES 42320
try harder to reflect actual version
(cherry picked from commit
eca0d2bdc94cbdbaeccd3d56991019c4ac732064)
Chaskiel M Grundman [Wed, 11 Oct 2006 13:27:18 +0000]
DEVEL15-elif-missing-defined-
20061011
FIXES 42321
add missing defined() to avoid warning on macro
(cherry picked from commit
cbc3b26165f18fbcf9e41b493541ed0a8035bc37)
Peter Nelson [Wed, 11 Oct 2006 12:31:38 +0000]
DEVEL15-rx-memset-param-swap-
20061011
discovered via google codesearch
(cherry picked from commit
e870e08280c27e6164b240be74057744352d7371)
Derrick Brashear [Tue, 10 Oct 2006 23:23:16 +0000]
DEVEL15-irix-compiler-hates-commas-
20061010
damn is irix fussy
(cherry picked from commit
59234e93a7aaf2f62bdd19bea39681e74752d1c5)
Chas Williams [Tue, 10 Oct 2006 23:02:42 +0000]
DEVEL15-linux-dont-lock-around-inactivevcache-
20061010
FIXES 42312
iafs_InactiveVCache() calls afs_InvalidateAllSegments() which says
it should be called with the vnode locked. so the lock should
probably be moved to afs_InactiveVCache() so it can be droppped
before calling afs_remunlink().
(cherry picked from commit
ac52e2f3c0bec9298d020de963036409165f380e)
Derrick Brashear [Tue, 10 Oct 2006 21:22:56 +0000]
DEVEL15-irix-fgets-return-type-fun-
20061010
of course, we were just throwing it away, but....
(cherry picked from commit
8fa4ca5b80bb2e0ac43bd9ae4883ac7f5b3928b8)
Jeffrey Altman [Tue, 10 Oct 2006 16:57:44 +0000]
DEVEL15-windows-scache-ref-leak-
20061010
if cm_FindSCache is called the returned scache object
must be released.
(cherry picked from commit
c60736030bc7f7a678fc44af90d001cd8c08acf5)
Jeffrey Altman [Tue, 10 Oct 2006 16:54:33 +0000]
DEVEL15-windows-afscreds-
20061010
prevent a stack overflow if the afsd_service does not successfully
start after a request to do so is issued.
(cherry picked from commit
3c13e5f0c0423d5e724ab224af22d7d3606ef1b0)
Jeffrey Altman [Tue, 10 Oct 2006 16:49:33 +0000]
DEVEL15-windows-build-clean-
20061010
more AFS_component_version_number.h files to remove
(cherry picked from commit
8b7e9da62578c3d10bf699062963c02aa59d1609)
Tracy Di Marco White [Mon, 9 Oct 2006 04:30:40 +0000]
DEVEL15-nbsd-sysnames-
20061008
FIXES 42019
add missing sysnames
(cherry picked from commit
aa7d443b6d8d45b68115c72ede6da68d863e33c7)
Derrick Brashear [Mon, 9 Oct 2006 04:24:51 +0000]
DEVEL15-h-enum-r-avoid-extra-hold-
20061008
FIXES 40699
avoid orphaning an extra hold
(cherry picked from commit
747a8d8f41a34290c207327541372b4e1fcb0dec)
Jeffrey Altman [Mon, 9 Oct 2006 04:20:01 +0000]
DEVEL15-windows-version-
20061008
1.5.900 == 1.5.9
Jeffrey Hutzelman [Mon, 9 Oct 2006 04:15:07 +0000]
DEVEL15-cellservdb-
20061008
cellservb update from grand.central.org
(cherry picked from commit
5ae8cf85e65e18a3d5593706c6f3f8d9435798e2)
Jeffrey Altman [Mon, 9 Oct 2006 02:52:20 +0000]
DEVEL15-windows-scache-ref-leak-fix-
20061008
pointer not pointer to pointer
(cherry picked from commit
1e9533907d4877f6b6aa70eb749e7a3e1717d861)
Jeffrey Altman [Mon, 9 Oct 2006 02:22:30 +0000]
DEVEL15-windows-relnotes-
20061008
Minor updates for 1.5.9
Jeffrey Altman [Sun, 8 Oct 2006 21:38:24 +0000]
DEVEL15-windows-scache-ref-dbg-
20061008
add debugging code that can be activated with #define DEBUG_REFCOUNT
to assist in debugging scache refcount leaks
(cherry picked from commit
8919e0d5175dae85991a3782f659621c5be4056f)
Jeffrey Altman [Sun, 8 Oct 2006 21:31:36 +0000]
DEVEL15-windows-scache-ref-leak-
20061008
plug two more scache refcount leaks
(cherry picked from commit
7231e2efe650b004168d443a290c4fed8b2db0a5)
Jeffrey Altman [Sat, 7 Oct 2006 23:36:03 +0000]
DEVEL15-windows-osi-prototypes-
20061007
osi_LogCreate
(cherry picked from commit
c97374d6fd494ab790da7ae6c591abe0dac3d194)
Jeffrey Altman [Sat, 7 Oct 2006 23:33:29 +0000]
DEVEL15-windows-misc-cleanup-
20061007
reformat some code; make sure that pointers are copied while
mutexes are held; ...
(cherry picked from commit
b4007635064f08d4d0f019acc9888f0d4005708c)
Jeffrey Altman [Sat, 7 Oct 2006 23:29:12 +0000]
DEVEL15-windows-scache-ref-leak-
20061007
return the cm_scache_t whose refCount was increased
(cherry picked from commit
feb5f27c754325dcc5fd9eb13f305bc1c0172977)
Jeffrey Altman [Fri, 6 Oct 2006 18:00:47 +0000]
DEVEL15-windows-notes-
20061006
updates since 1.5.8
Jeffrey Altman [Fri, 6 Oct 2006 17:32:43 +0000]
DEVEL15-windows-misc-fixes-
20061006
#ifdef DEBUG_REFCOUNT the ref count log entries so they aren't always
compiled in
comment out the remaining location where the write lock on cm_scacheLock
is dropped in order to obtain the scache mutex on the object returned
from cm_GetNewSCache(). Dropping the lock results in more than one
thread being given the same cm_scache_t which is more dangerous than
blowing away the contents of the object without holding the mutex
ensure that cm_BufWrite is always called with a non-NULL scp. Add
an assertion to double check that we do it all the time.
(cherry picked from commit
ffe26eb18da581d2df02f9531d97e31e1c6dee65)
Jeffrey Altman [Fri, 6 Oct 2006 17:26:41 +0000]
DEVEL15-windows-log-crlf-
20061006
use CR-LF as the afsd.log EOL so that it can be read with notepad.exe
(cherry picked from commit
ffe5092c3107afa72696f8d164c3ec587c2255ad)
Tracy Di Marco White [Fri, 6 Oct 2006 15:58:47 +0000]
DEVEL15-netbsd-30-
20061006
FIXES 40782
add support for netbsd 3.0, netbsd 4.0(ish)
(cherry picked from commit
04b51cc8df7a3210ea18228743cb156c0e2c7115)
Chaskiel M Grundman [Fri, 6 Oct 2006 15:02:56 +0000]
DEVEL15-osi-probe-avoid-out-of-range-memory-reads-
20061006
FIXES 41858
avoid oops on opteron due to accessing unmapped memory
(cherry picked from commit
e39f4d87896c451afb8c38705de4c12fd3940182)
Chas Williams [Fri, 6 Oct 2006 14:29:14 +0000]
DEVEL15-linux-2619-aio-and-read-write-changes-
20061006
changes to support 2.6.19
(cherry picked from commit
775ea6f1b79b0e0e2b700e3894e01279e26ebbfa)
Jeffrey Altman [Fri, 6 Oct 2006 06:15:22 +0000]
DEVEL15-windows-more-misc-fixes-
20061005
replace all DebugBreak() calls with osi_panic.
remove calls to cm_RecycleSCache from VNOVNODE handler as its not
lock safe to drop and re-obtain locks to adhere to the locking
hierarchy
restore reference to CM_SCACHESYNC_STOREDATA that should
not have been removed.
remove attempts to adhere to locking hierarching from cm_GetNewSCache.
dropping the cm_scacheLock is too dangerous
add a check to ensure that the scache refcount is not zero if there
is a valid smb_fid_t pointing to it.
more debug logging
(cherry picked from commit
b7117f61036c63238eb926c6eb71d42a56766501)
Jeffrey Altman [Thu, 5 Oct 2006 21:30:38 +0000]
DEVEL15-windows-misc-fix-
20061005
still tracking down issues when large numbers of temporary files
are created and deleted in the same directory from multiple
machines.
VNOVNODE must return CM_ERROR_BADFD
ensure that all calls to cm_GetCallback are made from cm_SyncOp.
add CM_SCACHESYNC_FORCECB logic to cm_SyncOp so that it can be
removed from cm_GetAccess
don't mix CM_SCACHESYNC_NEEDCALLBACK with CM_SCACHESYNC_STOREDATA
or CM_SCACHESYNC_FETCHDATA
(cherry picked from commit
4cd9eee0155edba89b77779af9aea4f1aa55cd96)
Jeffrey Altman [Thu, 5 Oct 2006 06:41:28 +0000]
DEVEL15-windows-misc-fix-
20061004
more cleanup from recent patches. comment out the recycling code because
it is not possible to implement it using the current locking hierarchy.
change cm_BufWrite to take a pointer to cm_scache_t instead of a fid
which must be used to look up a new reference to the cm_scache_t.
more often than not we already have the scp and in the one case we
don't we can let the caller look up the scp and then call cm_BufWrite
if it is found. If not, we have saved a function call and a bunch
of lock operations.
add a lot more logging.
improve the scp mutex handling within smb_CloseFID
(cherry picked from commit
38d464e5791dbd96d6da85a0c0123f8b575f4e39)
Jeffrey Altman [Wed, 4 Oct 2006 06:32:14 +0000]
DEVEL15-windows-wrong-lock-
20061003
grab the correct mutex so we don't free one we are not holding
(cherry picked from commit
b9fd1d4c7eb672ec092596af7826e87e2d931c40)
Jeffrey Altman [Wed, 4 Oct 2006 02:36:35 +0000]
DEVEL15-windows-oops-
20061003
didn't mean to delete this release mutex
(cherry picked from commit
93b3be7d1adfcf90bb21009891d9e62b440aa629)
Jeffrey Altman [Wed, 4 Oct 2006 01:24:42 +0000]
DEVEL15-windows-multi-fix-
20061003
misc cleanups
hold scp->mx while recycling
add FILE_NOTIFY_CHANGE_CREATION to the notification mask when adding
or deleting files
(cherry picked from commit
de6a01e51d8ea97e943cb4fa93d6bba2ecae9644)
Jeffrey Altman [Wed, 4 Oct 2006 01:20:09 +0000]
DEVEL15-windows-fetchstatus-
20061003
Don't call cm_GetCallback without checking the current callback status
with cm_HaveCallback first
(cherry picked from commit
665a125335c4844d35fa6510a38083602ef7fdbc)