cc0a015671
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-19 21:30:17 -07:00
61914d7403
Merge commit 'refs/top-bases/linux-kernel-module' into linux-kernel-module
2009-03-19 21:30:15 -07:00
bd1685ee4e
Merge commit 'refs/top-bases/zfs-branch' into zfs-branch
2009-03-19 21:28:51 -07:00
1769e10f17
Merge branch 'feature-branch' into refs/top-bases/zfs-branch
2009-03-19 21:28:50 -07:00
6d883d7d40
Merge branch 'feature-pthreads' into refs/top-bases/feature-branch
2009-03-19 21:28:48 -07:00
fed604c844
Add extra () requested by gcc
2009-03-19 21:18:54 -07:00
0173e5dce7
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-19 21:08:22 -07:00
275f4aae5d
Merge commit 'refs/top-bases/linux-kernel-module' into linux-kernel-module
...
Conflicts:
module/zfs/dmu_tx.c
2009-03-19 21:08:17 -07:00
e7f02d3b41
Merge commit 'refs/top-bases/zfs-branch' into zfs-branch
2009-03-19 20:59:33 -07:00
eb811f7fa9
Merge branch 'feature-branch' into refs/top-bases/zfs-branch
...
Conflicts:
module/zfs/include/sys/dmu_tx.h
module/zfs/include/sys/txg.h
2009-03-19 20:59:12 -07:00
f180959bac
Merge branch 'feature-pthreads' into refs/top-bases/feature-branch
2009-03-19 20:51:18 -07:00
29703a5b4e
Add 'feature-commit-cb' branch for DMU commit callbacks.
2009-03-19 20:30:14 -07:00
86681399b8
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-19 15:30:33 -07:00
e1540e9cbb
Merge commit 'refs/top-bases/linux-kernel-module' into linux-kernel-module
2009-03-19 15:30:32 -07:00
db3436fafe
Merge commit 'refs/top-bases/zfs-branch' into zfs-branch
2009-03-19 15:29:35 -07:00
2518c98d9e
Merge branch 'fix-branch' into refs/top-bases/zfs-branch
2009-03-19 15:29:34 -07:00
c8c4948d00
Merge branch 'fix-dnode-cons' into refs/top-bases/fix-branch
2009-03-19 15:29:33 -07:00
60d25f4b93
The patch resolves the extra call to dnode_cons() in dnode_create().
...
The extra call to the constructor was there to reinitialize the non-
trivial primatives in the dnode (lists, mutexs, condvars, avl tree, etc).
This was safe, although not exactly clean, on Solaris because none of
the primitives allocate memory. In the Linux port this is not true.
To keep stack usage to a minimum several of the primatives dynamically
allocate memory thus initializing them twice results in a memory leak.
This patch resolves this problem for Solaris and Linux by ensuring all
*_inits are called in the constructor, and all *_destroys are called
in the destructor. Additionally we ensure that all dnode objects are
properly deconstructed before being freed to the slab, and when the
objects are allocated from the slab all required data members are
explicity initialized to correct values.
2009-03-19 15:22:48 -07:00
099dd7e175
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-19 15:08:18 -07:00
fd3db92fe7
Merge commit 'refs/top-bases/linux-kernel-module' into linux-kernel-module
2009-03-19 15:08:17 -07:00
90bbbd109b
Merge commit 'refs/top-bases/zfs-branch' into zfs-branch
2009-03-19 15:07:19 -07:00
f07d61a8c1
Merge branch 'fix-branch' into refs/top-bases/zfs-branch
2009-03-19 15:07:18 -07:00
cc032b6337
Merge branch 'fix-list' into refs/top-bases/fix-branch
2009-03-19 15:07:16 -07:00
9bbf8f1761
Resolve conflict with fix-dnode-cons
2009-03-19 15:02:05 -07:00
80e98bb0f5
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-17 16:06:50 -07:00
a9b0088f48
Use include missing zfs_config.h and use ZFS_META_VERSION (not VERSION)
2009-03-17 16:02:09 -07:00
7ca5472963
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-16 16:20:03 -07:00
5edabb999f
Merge commit 'refs/top-bases/linux-kernel-module' into linux-kernel-module
2009-03-16 16:20:02 -07:00
df985d53db
Merge commit 'refs/top-bases/zfs-branch' into zfs-branch
2009-03-16 16:17:57 -07:00
f86b52a39d
Merge branch 'gcc-branch' into refs/top-bases/zfs-branch
2009-03-16 16:17:55 -07:00
3eebe1bd0e
Merge branch 'gcc-uninit' into refs/top-bases/gcc-branch
2009-03-16 16:17:54 -07:00
3eb4ba3ebd
Add 3 more uninit errors detected by gcc 3.4.0 compiler.
2009-03-16 16:06:50 -07:00
f21bd57c48
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-13 10:59:15 -07:00
adf984639d
Merge commit 'refs/top-bases/linux-kernel-module' into linux-kernel-module
2009-03-13 10:59:14 -07:00
486df2d11d
Merge commit 'refs/top-bases/zfs-branch' into zfs-branch
2009-03-13 10:57:46 -07:00
36c46c6031
Merge branch 'gcc-branch' into refs/top-bases/zfs-branch
2009-03-13 10:57:45 -07:00
d2910d2415
Merge branch 'gcc-unused' into refs/top-bases/gcc-branch
2009-03-13 10:57:43 -07:00
c98996ded8
Explicitly cast to prevent 'void *' dereference.
2009-03-13 10:51:11 -07:00
5303323f3a
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-12 10:38:45 -07:00
2ff032fd60
Merge commit 'refs/top-bases/linux-kernel-module' into linux-kernel-module
2009-03-12 10:38:44 -07:00
b9b399bf04
Merge commit 'refs/top-bases/zfs-branch' into zfs-branch
2009-03-12 10:37:14 -07:00
a03187bb3b
Merge branch 'gcc-branch' into refs/top-bases/zfs-branch
2009-03-12 10:37:12 -07:00
f8fe8a76f4
Merge branch 'gcc-uninit' into refs/top-bases/gcc-branch
2009-03-12 10:37:11 -07:00
cabd4a69c4
Resolve remaining uninitialized variables
...
Most of these fixes appear to be harmless and should never occur.
However, there were a few cases in this patch which do concern me,
I doubt we're seeing them but they look possible... mainly in the
user tools.
2009-03-12 10:30:41 -07:00
cebf08458e
Merge commit 'refs/top-bases/linux-have-sysevent' into linux-have-sysevent
2009-03-12 10:01:27 -07:00
165b2611d2
Merge commit 'refs/top-bases/linux-kernel-module' into linux-kernel-module
2009-03-12 10:01:25 -07:00
dd30dba016
Merge commit 'refs/top-bases/zfs-branch' into zfs-branch
2009-03-12 10:00:04 -07:00
193ea156e9
Merge branch 'gcc-branch' into refs/top-bases/zfs-branch
2009-03-12 10:00:03 -07:00
eceaf813ad
Merge branch 'gcc-missing-case' into refs/top-bases/gcc-branch
2009-03-12 10:00:01 -07:00
f2d72af596
Add remaining missing cases (or a default) to all switch statements.
...
The previous code was not wrong, but this prevents gcc from warning
us about missing cases for these known safe switch statements. The
-Wno-missing-cases can now be removed to detect places where we
accidentally forgot a case.
2009-03-12 09:53:08 -07:00
Brian Behlendorf