Whamcloud - gitweb
fs/lustre-release.git
19 years agob=5628
phil [Sat, 12 Feb 2005 22:59:54 +0000 (22:59 +0000)]
b=5628
r=jacob
- add a per-mdc/osc /proc file that lets you do a Lustre ping
- the old way, used by lfs check, would block when the MDS is down, because it
  opened the mountpoint
- also exported it via liblustreapi as llapi_ping(obd_type, obd_name)

19 years agob=5656
phil [Sat, 12 Feb 2005 21:32:17 +0000 (21:32 +0000)]
b=5656
Clean up a lot of calls to LDLM_DEBUG that weren't protected by an ns_lock.
This is probably not a comprehensive fix, but addresses all of the violations
that I saw.

19 years agoAdd an l_check_ns_lock to go with l_check_no_ns_lock.
phil [Sat, 12 Feb 2005 21:03:22 +0000 (21:03 +0000)]
Add an l_check_ns_lock to go with l_check_no_ns_lock.

19 years agoremove the last vestiges of test_ldlm from {l,obd}ctl
phil [Sat, 12 Feb 2005 21:00:20 +0000 (21:00 +0000)]
remove the last vestiges of test_ldlm from {l,obd}ctl

19 years agoUpdated config files from 2.4.21-37EL, which enable a handful of new modules
phil [Sat, 12 Feb 2005 20:53:59 +0000 (20:53 +0000)]
Updated config files from 2.4.21-37EL, which enable a handful of new modules
(at least on x86_64)

19 years agobump the version to 1.4.0.10 for 1.4.1 rc1
phil [Sat, 12 Feb 2005 19:23:03 +0000 (19:23 +0000)]
bump the version to 1.4.0.10 for 1.4.1 rc1

19 years agoA customer mentioned that we didn't have a get_hostaddr method for the LDAP
phil [Sat, 12 Feb 2005 06:25:35 +0000 (06:25 +0000)]
A customer mentioned that we didn't have a get_hostaddr method for the LDAP
backend.  LDAP is on its way out of favour, so I'm hardly shocked.

19 years agoland b1_4_mmap on b1_4 (20050211_1829)
nic [Fri, 11 Feb 2005 23:42:08 +0000 (23:42 +0000)]
land b1_4_mmap on b1_4 (20050211_1829)

19 years agoLand b1_4_smallfix onto b1_4 (20050210_1537)
adilger [Thu, 10 Feb 2005 23:43:50 +0000 (23:43 +0000)]
Land b1_4_smallfix onto b1_4 (20050210_1537)
- fix incorrect LASSERT in mds_getattr_name (5635)
- fix configure check for grab_cache_page_nowait_gfp
- fix lsattr -v for directories so we can get inode generation from userspace
- fix mtime return from glimpse AST reply again (4829)

19 years ago- BUGS should just point at bugzilla
phil [Thu, 10 Feb 2005 23:33:09 +0000 (23:33 +0000)]
- BUGS should just point at bugzilla
- nodist doesn't need to reference obdext2 anymore
- rename COPYING to LICENSE and make it somewhat clearer
- remove the unused mgmt/ module; we know how to get it back.

19 years agoremove the lustre book-building bits from Makefile.am, in preparation for
phil [Thu, 10 Feb 2005 23:31:40 +0000 (23:31 +0000)]
remove the lustre book-building bits from Makefile.am, in preparation for
removal from the tree.
r=jacob

19 years agob=5277
phil [Thu, 10 Feb 2005 05:55:19 +0000 (05:55 +0000)]
b=5277
r=adilger
Andreas's patch to avoid starting duplicate acceptors on servers with multiple
IP hostaddrs.

19 years agob=5544
phil [Wed, 9 Feb 2005 19:50:22 +0000 (19:50 +0000)]
b=5544
A checkin was made on 17 Jan to move lov_get_info from llite to obdclass.
But here it is, still in llite, which is causing depmod to cry foul about
recursive dependencies with the lov.

19 years agob=5492
phil [Sat, 5 Feb 2005 06:34:04 +0000 (06:34 +0000)]
b=5492
Back out the KMS locking change; I saw a problem with it yesterday, and the
new truncate buffalo failures confirmed my suspicion.

19 years agoAt some point, somebody added a 5th upcall argument, but didn't print it
phil [Fri, 4 Feb 2005 19:55:46 +0000 (19:55 +0000)]
At some point, somebody added a 5th upcall argument, but didn't print it
with the rest of them.

19 years agob=5654
phil [Fri, 4 Feb 2005 19:30:22 +0000 (19:30 +0000)]
b=5654
The lock ordering of i_sem and i_alloc_sem changed between 2.4 and 2.6,
which causes deadlocks if multiple threads try to truncate a single file.

I don't really like this solution, but it's the best I can come up with
right now.

19 years ago * Landed portals:b_port_step as follows...
eeb [Fri, 4 Feb 2005 18:49:21 +0000 (18:49 +0000)]
* Landed portals:b_port_step as follows...

  - removed CFS_DECL_SPIN*
    just use 'spinlock_t' and initialise with spin_lock_init()

  - removed CFS_DECL_MUTEX*
    just use 'struct semaphore' and initialise with init_mutex()

  - removed CFS_DECL_RWSEM*
    just use 'struct rw_semaphore' and initialise with init_rwsem()

  - renamed cfs_sleep_chan -> cfs_waitq
            cfs_sleep_link -> cfs_waitlink

          - fixed race in linux version of arch-independent socknal
       (the ENOMEM/EAGAIN decision).

  - Didn't fix problems in Darwin version of arch-independent socknal
      (resetting socket callbacks, eager ack hack, ENOMEM/EAGAIN decision)

  - removed libcfs types from non-socknal header files (only some types
    in the header files had been changed; the .c files hadn't been
    updated at all).

  - Updated lustre b1_4 to match

19 years agoignore _netdev as well
jacob [Thu, 3 Feb 2005 19:49:20 +0000 (19:49 +0000)]
ignore _netdev as well

19 years agothis is upstream now
jacob [Thu, 3 Feb 2005 17:09:23 +0000 (17:09 +0000)]
this is upstream now

19 years agofix automake warning
jacob [Thu, 3 Feb 2005 16:58:13 +0000 (16:58 +0000)]
fix automake warning

19 years agob=5383
jacob [Wed, 2 Feb 2005 23:29:07 +0000 (23:29 +0000)]
b=5383
r=phil

allow readahead to fail when low on memory.

19 years agoland b1_4_smallfix on b1_4(20050202_1817)
nic [Wed, 2 Feb 2005 23:13:21 +0000 (23:13 +0000)]
land b1_4_smallfix on b1_4(20050202_1817)

- readahead could read extra page into cache that wasn't ejected (5388)
- prevent races in class_attach/setup/cleanup/detach (5260)
- don't dereference de->d_inode after l_dput of de (5458)
- use "int" for stripe value returned from lock_to_stripe (5544)
- mballoc allocation and error-checking fixes in 2.6 (5504)
- block device patches to fix I/O request sizes in 2.6 (5482)
- protect KMS changes with its own lock, not i_sem (5492, 5624, 3453)
- look up hostnames for IB nals (5602)
- fix mount usage message, return errors per mount(8) (5168)
- remove ancient LOV_MAGIC_V0 EA support (5047)
- add "disk I/Os in flight" and "I/O req time" stats in obdfilter

19 years agoupdate to SP1 kernel: 2.6.5-7.141
jacob [Wed, 2 Feb 2005 04:17:16 +0000 (04:17 +0000)]
update to SP1 kernel: 2.6.5-7.141

19 years ago* ranal debugging and lconf support
eeb [Tue, 1 Feb 2005 11:58:00 +0000 (11:58 +0000)]
*   ranal debugging and lconf support

19 years ago* 5602 fix improves checks that NID is set correctly and causes incorrect
eeb [Mon, 31 Jan 2005 12:30:33 +0000 (12:30 +0000)]
*  5602 fix improves checks that NID is set correctly and causes incorrect
   configurations to fail rather than trip a server assertion.

*  5606 fix ensures both peers agree on maximum "immediate" get.

*  ranal updates

19 years agoFix for bug 5511
nkj [Fri, 28 Jan 2005 20:22:15 +0000 (20:22 +0000)]
Fix for bug 5511

19 years agocorrected filename
nkj [Fri, 28 Jan 2005 15:30:09 +0000 (15:30 +0000)]
corrected filename

19 years agob=5594
nic [Thu, 27 Jan 2005 00:38:22 +0000 (00:38 +0000)]
b=5594

remove hardcoded path to llmount that breaks rpm based testing

19 years agolanding patch for bug 5272 - making copy of list instead of using reference
nkj [Wed, 26 Jan 2005 12:44:13 +0000 (12:44 +0000)]
landing patch for bug 5272 - making copy of list instead of using reference

19 years agob=5570
nic [Wed, 19 Jan 2005 22:39:00 +0000 (22:39 +0000)]
b=5570

shorten name of suse drop to something we can use with lbuild and buffalo testing

19 years ago* Made openib not use the subnet manager to discover connection parameters
eeb [Tue, 11 Jan 2005 03:37:38 +0000 (03:37 +0000)]
*   Made openib not use the subnet manager to discover connection parameters

    + Openib listens on a TCP/IP port for service queries and responds with the
      service id, port GID and pkey.

    + Openib peer table entries have become (NID, IP, port) tuples, where IP
      and port connect to the peer's service query server.

    + /proc interfaces for port, listener timeout and backlog (restarts kernel
      listener on update).

    + lmc/lconf support for new openib peers, including only running acceptor
      if net == 'tcp'

    + Changed connection daemon to conduct both sides of the service query as
      well as IB connection establishment.  Spawning several since tcp/ip reads
      can block for a timeout.

    + Added a reaper thread to do connection cleanup and timeout checks (the
      single connection daemon used to do that too).

*   Removed some unused lconf default constants

*   Better openib automagic compilation check.

*   Consistent usage (flipping + optional checksum) of all openib messages
    (service query, connection requests and "normal" message flow).

*   Exhaustive openib "are-you-still-the-same-person" checks

*   Fixed bug which stopped idle persistent peers from getting removed from the
    peer table.

*   Fixed some ranal bugs which became obvious when similar problems were
    debugged in openibnal

19 years agob=5445
phil [Mon, 10 Jan 2005 21:51:53 +0000 (21:51 +0000)]
b=5445
Fix the x86-64 large-stack RHEL3 patch.  The old patch would allocate 1 page
in alloc_task_struct, then free THREAD_ORDER pages in free_task_struct.

19 years ago* Included changes from *old* b1_4/b_cray
eeb [Fri, 7 Jan 2005 12:41:05 +0000 (12:41 +0000)]
* Included changes from *old* b1_4/b_cray

    *  OBD_FL_DEBUG_CHECK for consistent obdecho end-to-end checking

    *  struct echo_obd cleanup

    *  obdecho support for "difficult" reply testing (test_setattr)

    *  replaced cowardly eap_from_cookie() usage with EAP_FROM_COOKIE()
       which asserts it is being called correctly

    *  bracketed echo_client_iocontrol() guts with unlock/lock_kernel()
       for increased concurrency during echo ioctl execution

    *  better thread control and stats in obd.c

    *  n-threads-over-m-objects test_brw support (required for
       obdfilter-survey)

* openibnal changes

    *  re-enabled openib on 2.5/2.6 builds
       configure --with-openib=no can be used to disable openib explicitly if
       required.

    *  fixed some 64 bit casts

    *  just use kernel's page_to_phys()

* fixed CERROR format problem in libcfs/module.c

19 years ago- keep version in line with latest tarball available on the FTP site;
buffalo-pull [Wed, 29 Dec 2004 23:12:57 +0000 (23:12 +0000)]
- keep version in line with latest tarball available on the FTP site;
- keep ia64 configs up-to-date with same tarball

19 years agoUpdate series to U4 kernel (2.4.21-27.EL)
jacob [Tue, 21 Dec 2004 21:39:32 +0000 (21:39 +0000)]
Update series to U4 kernel (2.4.21-27.EL)

19 years agoLand b1_4_smallfix onto b1_4 (20041221_0116)
adilger [Tue, 21 Dec 2004 09:43:56 +0000 (09:43 +0000)]
Land b1_4_smallfix onto b1_4 (20041221_0116)
- fix 2.6 build problems
- fix mount usage message, return errors per mount(8)
- change grep [] to grep "[]" in tests so they work in more UMLs

19 years agoFix non-module build
jacob [Mon, 20 Dec 2004 23:55:00 +0000 (23:55 +0000)]
Fix non-module build

19 years agoRemove obd_ptlbd.h, and fix "make dist"
phil [Mon, 20 Dec 2004 23:08:38 +0000 (23:08 +0000)]
Remove obd_ptlbd.h, and fix "make dist"

19 years agoadd bigstack support to rhel 2.4.21
jacob [Mon, 20 Dec 2004 22:54:27 +0000 (22:54 +0000)]
add bigstack support to rhel 2.4.21

19 years agoRemove one missed smfs/snapfs bit from lustre-core.m4
phil [Mon, 20 Dec 2004 22:52:15 +0000 (22:52 +0000)]
Remove one missed smfs/snapfs bit from lustre-core.m4

19 years agoremoving unused b1_4 code modules: cobd, ptlbd, smfs, snapfs
phil [Mon, 20 Dec 2004 22:49:36 +0000 (22:49 +0000)]
removing unused b1_4 code modules: cobd, ptlbd, smfs, snapfs

I'll send a separate mail to eng, but BE CAREFUL WHEN MERGING.

Merging into b1_2 or b_cray is fine, but be careful when you merge
into HEAD!  If you want my help, please ask.

19 years agoRemove chaos series. chaos76 era kernels are compatible with our rhel patches
jacob [Mon, 20 Dec 2004 22:40:52 +0000 (22:40 +0000)]
Remove chaos series.  chaos76 era kernels are compatible with our rhel patches

19 years agoI remember very clearly making this change, but I don't know where it
phil [Mon, 20 Dec 2004 21:01:39 +0000 (21:01 +0000)]
I remember very clearly making this change, but I don't know where it
got saved.  Probably in the wrong source tree.

19 years agoCopyright cleanup phase 1:
phil [Mon, 20 Dec 2004 20:30:52 +0000 (20:30 +0000)]
Copyright cleanup phase 1:
- add some HP copyright statements to the code that has survived our
  rewrites
- remove copyright and license boilerplate from non-copyrightable headers
- remove the unused ldlm_test
- remove the kernel copyright from llite/namei.c

19 years agoMinor b1_4 changelog revision; a few items were in the 1.4.0 list, when
phil [Mon, 20 Dec 2004 18:22:14 +0000 (18:22 +0000)]
Minor b1_4 changelog revision; a few items were in the 1.4.0 list, when
they did not, in fact, make it.

19 years agofix in-tree path to portals
jacob [Mon, 20 Dec 2004 17:56:34 +0000 (17:56 +0000)]
fix in-tree path to portals

19 years agoRemove files that moved to lustre-build
jacob [Mon, 20 Dec 2004 00:51:35 +0000 (00:51 +0000)]
Remove files that moved to lustre-build

19 years agofixes for make rpms
jacob [Mon, 20 Dec 2004 00:19:29 +0000 (00:19 +0000)]
fixes for make rpms

19 years agoAdd missed files.
jacob [Sun, 19 Dec 2004 23:28:18 +0000 (23:28 +0000)]
Add missed files.

19 years agoSplit portals into its own CVS module (b1_4)
jacob [Sun, 19 Dec 2004 23:20:10 +0000 (23:20 +0000)]
Split portals into its own CVS module (b1_4)

19 years agoThis commit was manufactured by cvs2svn to create branch 'b1_4'.
cvs2svn [Sun, 19 Dec 2004 23:11:17 +0000 (23:11 +0000)]
This commit was manufactured by cvs2svn to create branch 'b1_4'.

19 years agoMove portals into its own CVS module for HEAD.
jacob [Sun, 19 Dec 2004 23:11:15 +0000 (23:11 +0000)]
Move portals into its own CVS module for HEAD.

19 years agoUse the newer libsysio tags.
jacob [Sun, 19 Dec 2004 22:58:14 +0000 (22:58 +0000)]
Use the newer libsysio tags.

19 years agoThis commit was generated by cvs2svn to compensate for changes in r45926,
jacob [Sun, 19 Dec 2004 22:56:35 +0000 (22:56 +0000)]
This commit was generated by cvs2svn to compensate for changes in r45926,
which included commits to RCS files with non-trunk default branches.

19 years agoimport libsysio for b_newsysio
jacob [Sun, 19 Dec 2004 22:56:35 +0000 (22:56 +0000)]
import libsysio for b_newsysio

19 years agoThis commit was generated by cvs2svn to compensate for changes in r45924,
jacob [Sun, 19 Dec 2004 22:54:41 +0000 (22:54 +0000)]
This commit was generated by cvs2svn to compensate for changes in r45924,
which included commits to RCS files with non-trunk default branches.

19 years agoimport older libsysio snapshot.
jacob [Sun, 19 Dec 2004 22:54:41 +0000 (22:54 +0000)]
import older libsysio snapshot.

19 years ago* Updated ranal prior to the Great Schism
eeb [Fri, 17 Dec 2004 18:25:27 +0000 (18:25 +0000)]
*    Updated ranal prior to the Great Schism

19 years ago* Updated ranal prior to the Great Schism
eeb [Fri, 17 Dec 2004 18:22:35 +0000 (18:22 +0000)]
*   Updated ranal prior to the Great Schism

19 years agonew upstream bluesmoke patch
jacob [Fri, 17 Dec 2004 17:49:01 +0000 (17:49 +0000)]
new upstream bluesmoke patch

19 years ago* Added ranal
eeb [Fri, 17 Dec 2004 15:46:50 +0000 (15:46 +0000)]
*   Added ranal

19 years ago* more ranal lconf fixes
eeb [Fri, 17 Dec 2004 15:37:49 +0000 (15:37 +0000)]
*    more ranal lconf fixes

19 years ago* fixed ranal lconf support
eeb [Fri, 17 Dec 2004 14:57:22 +0000 (14:57 +0000)]
*   fixed ranal lconf support

19 years agoThis commit was manufactured by cvs2svn to create branch 'b1_4'.
cvs2svn [Fri, 17 Dec 2004 14:35:57 +0000 (14:35 +0000)]
This commit was manufactured by cvs2svn to create branch 'b1_4'.

19 years ago* Added ranal subdir
eeb [Fri, 17 Dec 2004 14:35:56 +0000 (14:35 +0000)]
*    Added ranal subdir

19 years ago* Added ranal
eeb [Fri, 17 Dec 2004 14:29:44 +0000 (14:29 +0000)]
*   Added ranal

*   Removed S_???NAL in favour of a single S_NAL

*   NB liblustre needs attention

19 years ago* fixed subsystem_names in lconf
eeb [Fri, 17 Dec 2004 12:32:51 +0000 (12:32 +0000)]
*   fixed subsystem_names in lconf

19 years ago- support UPCALLs and failover
yusy [Fri, 17 Dec 2004 07:35:15 +0000 (07:35 +0000)]
- support UPCALLs and failover

19 years agoLand b1_4_smallfix onto b1_4 (20041216_1438)
adilger [Thu, 16 Dec 2004 22:52:19 +0000 (22:52 +0000)]
Land b1_4_smallfix onto b1_4 (20041216_1438)
- add per-fs limit of Lustre pages in page cache, avoid OOM (4699)
- no transno return for symlink open, don't save no-trasno open (3440)
- don't try to complete elan receive that already failed (4012)
- free RPC server reply state on error (5406)
- fix ppc64/x86_64 spec to use %{_libdir} instead of /usr/lib (5389)
- unlock page after async write error during send (3677)
- return original error code on reconstructed replies (3761)
- no transno return for symlink open, don't save no-trasno open (3440)
- bump the watchdog timeouts -- we can't handle 30sec yet
- extra debugging for orphan dentry/inode bug (5259)
- clean up thread from ptlrpc_start_thread() on error (5160)
- minor changes from b_cray to reduce size of diff
- fix mount usage message, return errors per mount(8)

19 years agoMove these macros to lustre-portals.m4 and lustre-core.m4 where they really belong
jacob [Thu, 16 Dec 2004 01:22:27 +0000 (01:22 +0000)]
Move these macros to lustre-portals.m4 and lustre-core.m4 where they really belong

19 years ago * fix cvs up for HEAD branches
jacob [Wed, 15 Dec 2004 22:35:21 +0000 (22:35 +0000)]
 * fix cvs up for HEAD branches
 * get rid of LC_{DARWIN,LINUX}_SUPPORTED in favor of more general
   LC_TARGET_SUPPORTED
 * fix some typos and spacings

19 years agoAbstract modules other than lustre-build from the lustrecvs into its
jacob [Tue, 14 Dec 2004 23:21:10 +0000 (23:21 +0000)]
Abstract modules other than lustre-build from the lustrecvs into its
own buildcvs scriptlet.

19 years agothese are in share now
jacob [Tue, 14 Dec 2004 02:59:03 +0000 (02:59 +0000)]
these are in share now

19 years agohandle !=
jacob [Tue, 14 Dec 2004 01:21:31 +0000 (01:21 +0000)]
handle  !=

19 years ago - add missing files to EXTRA_DIST
jacob [Mon, 13 Dec 2004 23:38:58 +0000 (23:38 +0000)]
 - add missing files to EXTRA_DIST
 - change fshooks tests to work on HEAD (and better in general)
 - fix init script tests

19 years agouse HEAD portals and fix HEAD checkouts
jacob [Mon, 13 Dec 2004 23:27:43 +0000 (23:27 +0000)]
use HEAD portals and fix HEAD checkouts

19 years agoFix lconf --dump handling:
adilger [Sat, 11 Dec 2004 09:55:24 +0000 (09:55 +0000)]
Fix lconf --dump handling:
- don't generate a temp filename for a raw dump
- don't unlink non-regular files (e.g. /dev/null)
- don't report an error if kernel doesn't dump anything (fix mismerge?)

19 years agoLimit setting kernel debug buffer size to 512MB to avoid lowmem starvation.
adilger [Fri, 10 Dec 2004 09:44:46 +0000 (09:44 +0000)]
Limit setting kernel debug buffer size to 512MB to avoid lowmem starvation.

19 years agob_port_step files
jacob [Thu, 9 Dec 2004 23:27:44 +0000 (23:27 +0000)]
b_port_step files

19 years agob_port_step files
jacob [Thu, 9 Dec 2004 23:08:26 +0000 (23:08 +0000)]
b_port_step files

19 years agoadd autoconf dirs to build
jacob [Thu, 9 Dec 2004 22:14:36 +0000 (22:14 +0000)]
add autoconf dirs to build

19 years agorefine skeletal configure.ac
jacob [Thu, 9 Dec 2004 21:39:28 +0000 (21:39 +0000)]
refine skeletal configure.ac

19 years agofinish last commit
jacob [Thu, 9 Dec 2004 19:35:46 +0000 (19:35 +0000)]
finish last commit

19 years agoSplit up configure thusly:
jacob [Thu, 9 Dec 2004 19:34:54 +0000 (19:34 +0000)]
Split up configure thusly:

 - build/autoconf/configure.ac: main skel
 - build/autoconf/lustre-build.m4: non-arch macros
 - build/autoconf/lustre-build-linux.m4: linux-specific macros
 - build/autoconf/lustre-build-darwin.m4: darwin-specific macros
 - portals/autoconf/lustre-portals.m4: portals checks
 - lustre/autoconf/lustre-core.m4: lustre checks
 - lustre/autoconf/lustre-version.ac: the canonical lustre version

19 years ago* removed extra line
eeb [Thu, 9 Dec 2004 01:15:06 +0000 (01:15 +0000)]
*   removed extra line

19 years ago* removed a diff that crept in somehow
eeb [Thu, 9 Dec 2004 01:03:34 +0000 (01:03 +0000)]
*     removed a diff that crept in somehow

19 years ago* applied patch for missing <linux/init.h> b5396
eeb [Thu, 9 Dec 2004 00:43:09 +0000 (00:43 +0000)]
*    applied patch for missing <linux/init.h> b5396

19 years agofile writemany.c was initially added on branch b_B1_4_BUG4834.
nathan [Tue, 7 Dec 2004 22:09:15 +0000 (22:09 +0000)]
file writemany.c was initially added on branch b_B1_4_BUG4834.

19 years ago* Moved the CDEBUG message in the exit path of log_commit_thread. It was
eeb [Tue, 7 Dec 2004 14:11:39 +0000 (14:11 +0000)]
*   Moved the CDEBUG message in the exit path of log_commit_thread.  It was
    possible for portals_debug_msg() to crash dereferencing the format
    string if the CDEBUG happened after the thread had signalled that it
    was about to exit.  I guess the data segment got unloaded before the
    log_commit_thread had exited!

19 years ago* fixed compiler warning
eeb [Mon, 6 Dec 2004 22:07:15 +0000 (22:07 +0000)]
*     fixed compiler warning

19 years ago* Merged changes from HEAD and b_cray in preparation for
eeb [Mon, 6 Dec 2004 20:55:57 +0000 (20:55 +0000)]
*   Merged changes from HEAD and b_cray in preparation for
    portals separation

19 years agobranch scripts
jacob [Fri, 3 Dec 2004 18:40:00 +0000 (18:40 +0000)]
branch scripts

19 years ago- fixes in fsfilt interface about 2.6/2.4 versions of kernel. Fixes in ext3 implement...
yury [Fri, 3 Dec 2004 09:07:53 +0000 (09:07 +0000)]
- fixes in fsfilt interface about 2.6/2.4 versions of kernel. Fixes in ext3 implementation.
- removed not needed lookup_by_path() from lustre_smfs.h
- small cleanups about names of variables to make them coherent with another parts.

19 years agoFix lbuild building and 2.6
jacob [Thu, 2 Dec 2004 23:55:22 +0000 (23:55 +0000)]
Fix lbuild building and 2.6

19 years agoblah blah blah
jacob [Thu, 2 Dec 2004 16:34:59 +0000 (16:34 +0000)]
blah blah blah

19 years ago* fixed 4012: NULL deref in ep_complete_rpc
eeb [Wed, 1 Dec 2004 11:39:44 +0000 (11:39 +0000)]
*   fixed 4012: NULL deref in ep_complete_rpc
    This could crash a qswnal client engaged in bulk READS with a crashing server

19 years ago* fixed 4012: NULL deref in ep_complete_rpc
eeb [Wed, 1 Dec 2004 11:35:34 +0000 (11:35 +0000)]
*   fixed 4012: NULL deref in ep_complete_rpc
    This could crash a qswnal client engaged in bulk READS with a crashing server

19 years ago* backing out 4012 fix due to incomplete patch application
eeb [Tue, 30 Nov 2004 17:54:23 +0000 (17:54 +0000)]
*   backing out 4012 fix due to incomplete patch application

19 years ago* fixed 4012 (NULL pointer deref in ep_complete_rpc)
eeb [Tue, 30 Nov 2004 17:47:03 +0000 (17:47 +0000)]
*   fixed 4012 (NULL pointer deref in ep_complete_rpc)

19 years agoBranch: b1_4
adilger [Mon, 29 Nov 2004 18:20:03 +0000 (18:20 +0000)]
Branch: b1_4
Don't try to preallocate too many xmit buffers, qsnet doesn't like that.
Tested to work on TDEV.
b=5291
r=eeb