- Oct 19, 2007
-
-
Alexey Lyashkov authored
b=13730 i=johann i=green
-
Yang Sheng authored
b=13521 i=scjody i=nathan Update which_patch & target file for SLES10 2.6.16.53-0.8 kernel.
-
Yury Umanets authored
r=adilger,nathan - fixes in LVF calculation in sanity test 124a.
-
- Oct 18, 2007
-
-
Elena Gryaznova authored
i=Adilger skip test_125 if mds acl disabled
-
Nathan Rutman authored
i=adilger add error_ignore method to run a test but ignore failures
-
scjody authored
Revert Qian's recent changes related to bug 13099 since they caused build issues and differ significantly from the patch that was inspected.
-
Johann Lombardi authored
b=13828 i=adilger,shadow Don't use kernel's ATTR_* attributes over the network since the client and MDS may run different kernels. Instead, use MDS_ATTR_* attributes (defined in lustre_idl.h) for sa_valid.
-
qian authored
i=adilger,shadow remove the useless 'nob'.
-
qian authored
i=adilger,shadow fix the directio path: submit entire write in one RPC set. change the test script name.
-
Bobi Jam authored
b=13626 i=johann, adilger use k(un)map_atomic instead of k(un)map for short/tight code paths.
-
Elena Gryaznova authored
lost test_7 fix
-
Bobi Jam authored
b=13626 reverse the patch since been reported broken test on box with 2G ram.
-
alex authored
- revert patch landed by mistake
-
alex authored
i=adilger i=eeb - improved sd iostats: * time requests spend in elevator's queue * time requests are served by disk * seek stats * queue depth stats
-
Bobi Jam authored
b=12744 i=adilger, johann get rid of smp_num_cpus, instead use num_online_cpus() and/or num_possible_cpus().
-
Bobi Jam authored
b=13626 i=johann, adilger use k(un)map_atomic instead of k(un)map for short/tight code paths.
-
qian authored
i=adilger,shadow
-
qian authored
i=adilger,shadow
-
qian authored
i=adilger,shadow fix directIO path.
-
scjody authored
Report a patchless kernel if client patches are missing, since that's what we really care about for testing purposes. b=12460 i=adilger
-
qian authored
i=adilger,shadow fix directIO path.
-
- Oct 17, 2007
-
-
scjody authored
Fix ChangeLog bug number to match reality (the wrong bug number was entered) and improve the entry somewhat.
-
kalpak authored
i=adilger, johann Fix bugs in jbd-stats patches spotted by coverity.
-
Wang Di authored
update 13570 Entry
-
alex authored
i=adilger - patch to rhel4 series: keep inode hashed over clear_inode() call
-
kalpak authored
i=adilger, johann Enhance lfs find by adding filetype and filesize parameters. Also multiple OBDs can now be specified for the --obd option.
-
scjody authored
Add test 14c to ALWAYS_EXCEPT b=12652
-
- Oct 16, 2007
-
-
Nathan Rutman authored
i=johann i=adilger add mds req_history details
-
Nathan Rutman authored
i=shadow i=adilger remove client blocking for ost connection
-
Nathan Rutman authored
i=adilger i=johann extend reconnect timer as new clients reconnect fix recovery_status proc file consistently use xNNN for xids for easier debugging
-
Vitaly Fertman authored
b=13693 i=adilger i=green do not forget to cancel locally locks with invalid import; do not cancel locks in ldlm_lock_list_put() when 0 locks are to be canceled;
-
Johann Lombardi authored
b=13635 i=adilger i=shadow Add opc parameter to osc_checksum_bulk
-
- Oct 15, 2007
-
-
Yang Sheng authored
b=13748 i=shadow i=johann Update RHEL4 kernel to fix the vulnerability described in CVE-2007-4573.
-
- Oct 14, 2007
-
-
Wang Di authored
b=13570 In filter_check_grant, for non_grant cache write, we should check the left space by if (*left > ungranted + bytes), instead of (*left > ungranted), because only we are sure the left space is enough for another "bytes", then the ungrant space should be increase. In client, we should update cl_avail_grant only there is OBD_MD_FLGRANT in the reply. i=adilger,green
-
Jian Yu authored
b=13207 i=nathan i=brian 1) Adapt the lustre_config script to support upgrading 1.4 server targets to 1.6 in parallel. 2) Fix a path issue in lustre_createcsv.
-
- Oct 13, 2007
-
-
Yury Umanets authored
r=adilger,nathan - fixed race in ldlm_namespace_new() when new namespace could be aded to its list before initializing its pool. This led to panic when ldlm pools thread find this namespace before ldlm_namespace_new() finishes its init.
-
Yury Umanets authored
r=adilger,nathan - fixed handling user input in lprocfs_wr_lru_size().
-
Elena Gryaznova authored
i=Adilger add recursive symlink test_17e
-
- Oct 12, 2007