| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
Tested on: 64-bit Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
This is part I of the fix for object refresh: no refresh action needed if the file is the writer.
Tested on jam, platypus, os1010test, emu, moohan, ostrich.
|
|
|
|
|
| |
Tested on: 64-bit Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
| |
|
|
|
|
|
|
|
|
| |
HDFFV-9273: cmakehdf5 --script failed in linux platypus
Solution: remove --script option from cmakehdf5.
Tested: platypus
|
|
|
|
|
|
|
| |
the debug/verbose function.
Tested on: 64-bit Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on: 64-bit Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
|
|
|
| |
B-tree package.
This data structure cannot be used with SWMR so its flush dependency code
has been removed.
Tested on: 64-bit Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
|
| |
Also purged obsolete /*ARGSUSED*/ from the library.
Tested on:
Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
|
| |
Minor code cleanups during review/review/merge of branch
Tested on:
MacOSX/64 10.11.2 (amazon) w/serial & parallel
(h5committest not required on this branch)
|
|
|
|
|
|
| |
Tested on:
Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
|
|
| |
Local heaps can't be used in SWMR due to a lack of checksums. The
flush dependency code has been removed as in the v1 B-trees.
Tested on:
Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
|
| |
Minor changes.
Tested on:
Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
|
|
|
| |
Revisions to normalize revise chunks vs. trunk, plus push dataset metadata
changes into the cache immediately (instead of holding them within the dataset
data structures).
Tested on:
MacOSX/64 10.11.2 (amazon) w/serial & parallel
(h5committest not required on this branch)
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1 w/ CMake 3.3.2
serial only
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
SWMR-95 -- specifically:
1) Code in H5C__flush_single_entry() taking down all remaining
flush dependencies before an entry is destroyed.
Replaced this with code asserting that all flush dependencies
have been taken down by the cache client as required by the
protocol for managing flush dependencies. This exposed:
2) Both the object header proxy and the object header continuation
chunk cache clients were failing to track their flush dependency
parents, and thus were relying on the above code in
H5C__flush_single_entry() to take down their flush dependencies
instead of handling this issue in their notify callbacks.
Fixing item 2 for the object header continuation chunks was straight
forward, as each object header continuation chunk can have at most
one flush dependency parent.
However, there is no upper bound on the number of flush dependency parents
for an object header proxy (one must be an object header -- any additional
flush dependency parents must be object header continuation chunks).
This said, in most cases object header proxies have only one flush
dependency parent.
To handle this, I adapted Neil's code for tracking multiple flush
dependency parents in the metadata cache for use in the object header
proxy, with the difference that I set the initial array size to 1
(Neil uses 8). To change this, modify H5O_FD_PAR_LIST_BASE in H5Opkg.h
Potential issues with this fix:
1) While the fix passes the existing regression tests, I don't believe
these test the code for tracking multiple flush dependency parents
in the object header proxy adequately.
From discussions with Quincey, I gather that the obvious way to
do this is to add and delete large numbers of properties to/from the
object header, forcing increases and decreases in the number of
object header continuation chunks.
However, as I am unfamiliar with the layout of the SWMR test code,
it seemed inefficient for me to address this issue -- hence I gather
that it will be handled elsewhere.
2) My fix tracks both pointers to flush dependency parents and their
addresses. As the addresses are used purely for sanity checking,
once we are reasonably confident of my fix, the code can be
modified to either to remove the address tracking and the associated
sanity checks completely, or to maintain the addresses in debug
builds only.
Note that the address tracking code presumes that object headers
and object header continuation chunks cannot change addresses. If
this ever changes, the sanity checking code will complain.
Tested (debug only) serial and parallel on Mercury, serial on Jam.
|
|
|
|
|
|
|
| |
Minor bin/cmakehdf5 change
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1 w/ CMake 3.3.2
serial only
|
|
|
|
|
|
|
| |
check-vfd.
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
| |
Tested on: 64-bit Windows 10 Pro w/ VS 2015 Pro & CMake 3.3.1
|
|
|
|
|
|
|
| |
truncation check only if the file
is marked for SWMR writing mode and the file has version 3 superblock for SWMR support.
Tested on jam, platypus, osx1010test, moohan, ostrich, emu, kite.
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
| |
indexing. This passes with no further code changes.
Tested on: Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
| |
data structure.
Tested on: Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on: Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
|
| |
Based on the original error report. Confirmed to fail in r28476
(pre-fix) and pass in the current version.
Tested on: Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
| |
Tested on: 64-bit Windows 10 Pro w/ VS2015 Pro & CMake 3.3.1
|
|
|
|
| |
Tested on: 64-bit Windows 10 Pro w/ VS2015 Pro & CMake 3.3.1
|
|
|
|
|
|
|
|
| |
- Re-disabled the cork test, which is still failing on Windows.
Tested on:
64-bit Windows 10 w/ VS 2015 Pro & CMake 3.3.1
Ubuntu 15.10 (Linux 4.2.0 x86_64) w/ gcc 5.2.1 & CMake 3.3.2
|
|
|
|
|
|
|
|
|
| |
Correct error where dataset metadata wasn't getting tagged correctly when
brought into the cache through the get object info calls.
Tested on:
MacOSX/64 10.11.1 (amazon) w/serial & parallel
(h5committest not required on this branch)
|
|
|
|
| |
should work on windows.
|
|
|
|
| |
only.(Although it could be built shared - but this could make things more complicated).
|
|
|
|
|
|
|
|
| |
Clean up changes to extensible array flush dependencies.
Tested on:
MacOSX/64 10.11.1 (amazon) w/serial & parallel
(h5committest not required on this branch)
|
|
|
|
|
|
| |
condition in the build phase by separating the static and shared build directories.
Tested on: Windows 10 Pro w/ VS 2015 Pro and CMake 3.3.1.
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1 CMake 3.3.2
serial only
|
|
|
|
|
|
| |
a no-op until a Win-32 based Windows VFD is available.
Tested on: Windows 10 Pro w/ VS 2015 Pro and CMake 3.3.1.
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1 & CMake 3.3.2
serial only
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
H5Oget_info_by_name() resulted in an assertion failure
in the metadata cache.
Investigation of this bug uncovered a number of issues,
only some of which are addressed in this checkin --
specifically incorrect takedown of the flush dependency
between the oject header proxy and the extensible array
header, and attempts to setup this flush dependency
repeatedly.
tested (serial/debug only) on mercury and jam
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1 & CMake 3.3.2
serial only
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
|
| |
Warning & code style cleanups, along with normalizations against trunk.
Tested on:
MacOSX/64 10.11.1 (amazon) w/serial & parallel
(h5committest not required on this branch)
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
|
|
| |
was negative from H5Pset_append_flush() and the swmr.c test.
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|
|
|
|
|
| |
Tested on Ubuntu 15.10 (Linux 4.2.0 x86_64) gcc 5.2.1
serial only
|