Tue Aug 25 2022 Łukasz Plewa <lukasz.plewa@intel.com>
* Version 1.12.1
This release allows usings the movdir64b instruction by setting an
environment variable, and fixes a number of smaller bugs.
Detailed list of bug fixes:
- obj: always perform cleanup of logs with leftover state
- pmem2: actually skip msync with noflush on non-pmem
- pmem2: make sure async functions use proper vdm flags
- common: fix default cflags in makefiles
- obj: ensure zones are reclaimed prior to free
- pmem2: enable movdir64b compilation but disable it in the runtime
Tue May 24 2022 Łukasz Plewa <lukasz.plewa@intel.com>
* Version 1.12.0
This release:
- Deprecates librpmem library and rpmemd tool.
Using the librpmem library API will result in warnings
and is no longer recommended. Those interested in a remote
persistent memory support should use the new rpma library.
- Introduces a new set of APIs in libpmem2 to perform asynchronous data
movement operations. To use this feature, software needs to include an
optional miniasync(7) dependency.
- Adds new API to machine safe read/write operations in the pmem2 library.
- Introduces support for movdir64b instruction for memory operation
functions in the libpmem2 library.
- Adds experimental support for RISC-V.
Other changes and notable bug fixes:
- common: supress false positve of '-Wunused-parameter'
- examples: remove unnecessary persists in pminvaders
- obj: fix invalid type when setting cache size ctl (pmem/issues#5291)
- pmem: fix eADR memmove and memset (pmem/issues#5364)
- pmem2: add addr alignment prediction to vm reservation
- pmreorder: update docs with required last arg - file_name
- pmreorder: add more debug/info logs
Fri Sep 24 2020 Łukasz Plewa <lukasz.plewa@intel.com>
* Version 1.11.1
This release fixes a missing sfence in non-temporal
version of memcpy function
(https://github.com/pmem/pmdk/issues/5292),
and fixes a number of smaller bugs.
Detailed list of bug fixes:
- doc: remove exprimental moniker from libpmem2(7)
- common: fix missing sfence in non-temporal memcpy
- common: fix a mismatch between prototype and body
- common: the yearly spellchecker run
- common: force no LTO for rpm build
- common: fix mismatched function args
- obj: rename vars clashing with those of a containing block
- pmem2: don't force smaller alignment for fsdax mappings
- pool: don't trample upon users of localtime()
- rpmem: Fix RPMEM_RAW_BUFF_SIZE and LANE_ALIGN_SIZE for powerpc64le
- rpmem: drop a redundant check
Mon Sep 24 2021 Łukasz Plewa <lukasz.plewa@intel.com>
* Version 1.10.1
This release fixes a missing sfence in non-temporal
version of memcpy function
(https://github.com/pmem/pmdk/issues/5292),
and fixes a number of smaller bugs.
Detailed list of bug fixes:
- common: add disable lto to cflags
- doc: remove exprimental moniker from libpmem2(7)
- common: fix missing sfence in non-temporal memcpy
- common: fix a mismatch between prototype and body
- common: the yearly spellchecker run
- common: force no LTO for rpm build
- common: fix mismatched function args
- obj: rename vars clashing with those of a containing block
- pmem2: don't force smaller alignment for fsdax mappings
- pool: don't trample upon users of localtime()
- rpmem: Fix RPMEM_RAW_BUFF_SIZE and LANE_ALIGN_SIZE for powerpc64le
- rpmem: drop a redundant check
Fri Sep 24 2021 Łukasz Plewa <lukasz.plewa@intel.com>
* Version 1.9.3
This release fixes a missing sfence in non-temporal
version of memcpy function
(https://github.com/pmem/pmdk/issues/5292),
and fixes a number of smaller bugs.
Detailed list of bug fixes:
- common: add disable lto to cflags
- common: fix missing sfence in non-temporal memcpy
- common: fix mismatched function args
- common: fix a mismatch between prototype and body
- common: the yearly spellchecker run
- common: force no LTO for rpm build
- rpmem: Fix RPMEM_RAW_BUFF_SIZE and LANE_ALIGN_SIZE for powerpc64le
Fri Jul 2 2021 Piotr Balcer <piotr.balcer@intel.com>
* Version 1.11.0
This release:
- Adds new APIs for libpmem2, most notably there are new functions
to shrink and extend an existing reservation and a new iterator API
for mappings contained within an existing reservation. There's also
a new function to retrieve a numa node for a source.
- Makes the pmemobj_open() and pmemobj_close() functions from libpmemobj
thread-safe. It's now easier to correctly manage persistent memory
pools in a parallel environment.
- Introduces a new API in libpmemobj to globally change the method of
assigning arenas to threads. The default is to rely on a OS per-thread
key to store arena information. This release introduces an option
to avoid the use of thread-local keys by simply using one global
arena for all threads in a pool.
Other changes and notable bug fixes:
- pmem2: don't force smaller alignment for fsdax mappings
- rpmem: various fixes for powerpc64le
- doc: fix documentation of pmem_is_pmem()
- common: fix various minor problems found by static analysis
- pmem2: arm64: fix possible data loss on ARMv8.2+ (improper flushing)
This release introduces no changes to the on-media layout and is
fully compatible with the previous version of PMDK.
Wed Oct 28 2020 Piotr Balcer <piotr.balcer@intel.com>
* Version 1.10
This release introduces a new stable PMDK library, libpmem2, which
is the next major release of libpmem.
This library has an entirely new, but familiar, API that addresses many
shortcomings of the previous version, while retaining all of its
functionality. To learn more, see https://pmem.io/pmdk/libpmem2/ or
libpmem2(7).
The old library, libpmem, is still going to be maintained for the
foreseeable future, but we'd like to encourage any new applications to
leverage libpmem2.
Wed Oct 28 2020 Piotr Balcer <piotr.balcer@intel.com>
* Version 1.9.2
This release reverts an incorrect change in SDS handling
"pool: disable SDS check if not supported", and introduces
a proper fix for the issues that patch attempted to correct.
Wed Sep 16 2020 Piotr Balcer <piotr.balcer@intel.com>
* Version 1.9.1
Detailed list of bug fixes:
- common: fix LIBFABRIC flags
- common: Add runtime SDS check and disable
- pool: disable SDS check if not supported
- obj: fix failure atomicity bug in huge allocs
- obj: add missing drain after ulog processing
Fri Jul 3 2020 Piotr Balcer <piotr.balcer@intel.com>
* Version 1.9
This release:
- Switches the default instruction set for memcpy, memmove and memset
implementations on x86 to AVX512, and introduces numerous performance
improvements for those operations on AVX and SSE2 fallback paths.
- Optimizes transactional allocations in libpmemobj by avoiding one
extraneous cache miss and reducing the amount of work required to perform
a reservation.
- Introduces a new API in libpmemobj, pmemobj_tx_set_failure_behavior,
that enables the application to control the behavior of aborting
transactions.
- Improves performance of pool creation on Windows by avoiding expensive
physical page allocation during file allocation.
- Stabilizes support for ppc64.
Other changes:
- pmem: mem[cpy|set] optimization when eADR is available
- obj: detect msync failures in non-pmem variants of mem[cpy|move|set]
Notable bug fixes:
- core: fix Last_errormsg_get when NO_LIBPTHREAD is defined
- pmem: read Unsafe Shutdown Count from region instead of interleave set
- common: fix deep_flushes failing on platforms that don't need them
- pmem: fix data cache flush on ppc64
- obj: fix run allocated recalculation
Fri Jan 31 2020 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.8
This release:
- Introduces new API in libpmemobj for user-assisted defragmentation
(see pmemobj_defrag man page for details).
- Introduces experimental support for PPC64LE architecture.
- Introduces new flag in libpmemobj for opting-out of transaction aborts
on failure (POBJ_TX_NO_ABORT), along with new variants of existing APIs
that didn't accept flags (pmemobj_tx_xfree, pmemobj_tx_xpublish,
pmemobj_tx_xlock, pmemobj_tx_xlog_append_buffer, pmemobj_tx_xstrdup,
pmemobj_tx_xwcsdup).
- Moves out libvmem and libvmmalloc to the new repository
(https://github.com/pmem/vmem).
Other changes:
- obj: introduce new statistics useful for defrag
- obj: introduce transient/persistent statistics enabling
- obj: introduce pmemobj_tx_(set/get)_user_data funcs
- obj: introduce pmemobj_(set/get)_user_data funcs
- obj: disable workaround for offsetof() since VS 15.5 in public header
- common: drop support for libndctl < 63 on Linux
- pool: rename -N --no-exec to -d --dry-run
Notable bug fixes:
- obj: fix zone size calculations
- obj: fix potential NULL-dereference in ulog_store
- obj: fix unintended recursive locking during palloc
- obj: fix lock release order in palloc publish
- obj: fix transient redo log of more than 64 entries
- obj: fix capacity ulog calculation
- obj: fix check of unaligned capacity size
- rpmem: add a missing case for GPSPM + FLUSH_STRICT
- pmem: fix pmemcheck support on platforms with eADR
- pool: fix possible memory leak
- rpmem: fix possible memory leak in rpmemd_config_read
- rpmem: fix possible memory leak in rpmemd_log_init
- rpmem: fix possible use-after-free
Mon Sep 30 2019 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.7
This release:
- Introduces new APIs in libpmemobj for managing space used by transactions.
(see pmemobj_tx_log_append_buffer man page for details)
- Introduces new APIs in librpmem, splitting rpmem_persist into rpmem_flush
and rpmem_drain, allowing applications to use the flush + drain model
already known from libpmem. (libpmemobj does not use this feature yet)
- Optimizes large libpmemobj transactions by significantly reducing
the amount of memory modified at the commit phase.
- Optimizes tracking of libpmemobj reservations.
- Adds new flags for libpmemobj's pmemobj_tx_xadd_range[_direct] API:
POBJ_XADD_NO_SNAPSHOT and POBJ_XADD_ASSUME_INITIALIZED, allowing
applications to optimize how memory is tracked by the library.
To support some of the above changes the libpmemobj on-media layout had
to be changed, which means that old pools have to be converted using
pmdk-convert >= 1.7.
Other changes:
- obj: fix merging of ranges when NOFLUSH flag is used (pmem/issues#1100)
- rpmem: fix closing of ssh connection (pmem/issues#995, pmem/issues#1060)
- obj: abort transaction on pmemobj_tx_publish failure
Internal changes:
- test: fault injection tests for pmemblk, pmemlog, and pmemobj
- test: improved Python testing framework
- test: support real pmem in bad blocks tests
- common: allow not building examples and benchmarks
Tue Aug 27 2019 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.6.1
This release fixes possible pool corruptions on Windows
(see https://github.com/pmem/pmdk/pull/3728 for details),
improves compatibility with newer Linux kernels with
respect to Device DAX detection, fixes pmemobj space management
for large pools, improves compatibility with newer toolchains,
incorporates build fixes for FreeBSD and fixes a number of
smaller bugs.
Detailed list of bug fixes:
- common: (win) fix possible pool file coruption (pmem/issues#972,
pmem/issues#715, pmem/issues#603)
- common: implement correct / robust device_dax_alignment
(pmem/issues#1071)
- obj: fix recycler not locating unused chunks
- doc: update pmemobj_tx_lock documentation wrt behavior on fail
- common: fix persistent domain detection (pmem/issues#1093)
- common: vecq: fix a pointer-to-struct aliasing violation
(crash on arm64)
- common: fix minor issues related to ndctl linking
- obj: drop recursion from pmemobj_next
- common: fix bug in badblock file error handling
- obj: fix handling of malloc failures
- common: fix handling of malloc failures (ctl)
- jemalloc: fix build with gcc 9
- obj: don't overwrite errno when palloc_heap_check_remote fails
- doc: fix pmreorder emit log macro
- rpmem: change order of rpmem init (pmem/issues#1043)
- common: Fix build failure due to unused macro PAGE_SIZE
- common: support older versions of pkg-config
- tools: link with release variant of pmemcommon
- common: add PMDK prefix to local copy of queue.h (pmem/issues#990)
- rpmem: switch to using an open coded basename (FreeBSD)
- common: posix_fallocate: guard against integer underflow in
check (FreeBSD)
- test: support Valgrind 3.15
- test: skip if fi_info is missing
- test: (win) fix sparsefile error handling
- test: fix libpmempool_feature tests that match logs
- test: remove vmem_delete test (pmem/issues#1074)
- test: adjust matchfiles in vmem_valgrind_region test (pmem/issues#1087)
- test: remove old log files for windows (pmem/issues#1013)
- test: remove invalid expect_normal_exit (pmem/issues#1092)
- test: suppress ld leak (pmem/issues#1098)
- test: Expose necessary symbols in libvmmalloc_dummy_funcs
(FreeBSD)
- test: fix tests failing because `tput` fails (FreeBSD)
- test: avoid obj_critnib_mt taking very long on many-core machines
- test: deal with libndctl's path without build system
- test: overwrite old log in pmempool_create/TEST14.PS1
- test: fix match files in tests which use dax devices
- test: fix match file in rpmem_addr_ext test
- test: fix pmempool_check test
Wed Aug 28 2019 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.5.2
This release fixes possible pool corruptions on Windows
(see https://github.com/pmem/pmdk/pull/3728 for details),
improves compatibility with newer Linux kernels with
respect to Device DAX detection, fixes pmemobj space management
for large pools, improves compatibility with newer toolchains
and fixes a number of smaller bugs.
Detailed list of bug fixes:
- common: (win) fix possible pool file coruption (pmem/issues#972,
pmem/issues#715, pmem/issues#603)
- common: implement correct / robust device_dax_alignment
(pmem/issues#1071)
- obj: fix crash after large undo log recovery
- obj: fix recycler not locating unused chunks
- doc: update pmemobj_tx_lock documentation wrt behavior on fail
- common: fix build of rpm packages on suse (pmem/issues#1023)
- common: fix persistent domain detection (pmem/issues#1093)
- common: vecq: fix a pointer-to-struct aliasing violation
(crash on arm64)
- rpmem: lock file prior to unlink (pmem/issues#833)
- common: fix for pool_set error handling (pmem/issues#1036)
- pmreorder: fix handling of store drain flush drain pattern
- obj: fix possible memory leak in tx_add_lock
- pool: free bad_block vector
- common: fix bug in badblock file error handling
- obj: fix handling of malloc failures
- common: fix handling of malloc failures (ctl)
- jemalloc: fix build with gcc 9
- obj: don't overwrite errno when palloc_heap_check_remote fails
- doc: fix typos in pmreorder configuration
- doc: fix pmreorder emit log macro
- tools: link with release variant of pmemcommon
- test: support Valgrind 3.15
- test: skip if fi_info is missing
- test: split test obj_tx_lock into two test cases (pmem/issues#1027)
- test: (win) fix sparsefile error handling
- test: fix libpmempool_feature tests that match logs
- test: remove vmem_delete test (pmem/issues#1074)
- test: adjust matchfiles in vmem_valgrind_region test (pmem/issues#1087)
- test: remove old log files for windows (pmem/issues#1013)
- test: remove invalid expect_normal_exit (pmem/issues#1092)
- test: suppress ld leak (pmem/issues#1098)
- test: fix failing pmemdetect on Windows
- test: fix match files in tests which use dax devices
- test: fix pmempool_check test
Fri Aug 30 2019 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.4.3
This release fixes possible pool corruptions on Windows
(see https://github.com/pmem/pmdk/pull/3728 for details)
and improves compatibility with newer Linux kernels with
respect to Device DAX detection.
Bug fixes:
- common: (win) fix possible pool file coruption (pmem/issues#972,
pmem/issues#715, pmem/issues#603)
- common: implement correct / robust device_dax_alignment
(pmem/issues#1071)
- common: fix device dax detection
- obj: fix pmemobj_check for pools with some sizes (pmem/issues#975)
- obj: fix type numbers for pmemobj_list_insert_new
- obj: fix pmemobj_tx_lock error handling
- obj: fix possible memory leak in tx_add_lock
- common: fix ctl_load_config during libpmemobj initialization
(pmem/issues#917)
- common: win: fix getopt returning "option is ambiguous"
- common: fix persistent domain detection (pmem/issues#1093)
- pool: do not copy same regions in update_uuids
- test: split test obj_tx_lock into two test cases
- test: remove checking errno in obj_tx_add_range_direct
- test: remove invalid expect_normal_exit
- test: fix int overflow in pmem_deep_persist test
- test: fix pmempool_check test
- test: (win) fix a few issues related to long paths
Tue Aug 27 2019 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.3.3
Bug fixes:
- pmem: fix clflush bit position
- common: implement correct / robust device_dax_alignment
- common: fix device dax detection
- common: fix library dependencies (pmem/issues#767)
- common: use rpm-config CFLAGS/LDFLAGS when building packages
(pmem/issues#768)
- test: fix vmmalloc_malloc_hooks (pmem/issues#773)
- test: fix compilation with clang-5.0 (pmem/issues#783)
- pool: fix set convert of v3 -> v4
- common: generate pkg-config files on make install
(pmem/issues#610)
- common: fix dependencies for Debian's dev packages
- test: add missing include in unittest.h
- common: (win) fix timed locks
- common: provide src version in GitHub tarballs
- common: fix free function in tls
Tue Aug 27 2019 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.2.4
Bug fixes:
- common: fix device dax detection (compatibility with newer kernels)
Tue Mar 26 2019 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.6
This release:
- Enables unsafe shutdown and bad block detection on Linux
on systems with libndctl >= 63. It is expected that systems with
libndctl >= 63 has necessary kernel support (Linux >= 4.20).
However, due to bugs in libndctl = 63 and Linux = 4.20, it is
recommended to use libndctl >= 64.1 and Linux >= 5.0.4.
On systems with libndctl < 63, PMDK uses old superuser-only
interfaces. Support for old or new interfaces is chosen at BUILD
time.
- Introduces arena control interface in pmemobj, allowing applications
to tweak performance and scalability of heap operations. See
pmemobj_ctl_get man page ("heap" namespace) for details.
- Introduces copy_on_write mode, which allows testing applications
using pmemobj with pmreorder. See pmemobj_ctl_get man page
("copy_on_write" namespace) for details.
Other changes:
- allocate file space when creating a pool on existing file
(pmem/issues#167)
- initial support for testing using fault injection
- initial Python test framework
- improve performance of pmemobj_pool_by_ptr
Bug fixes:
- common: work around tmpfs bug during pool creation
(pmem/issues#1018)
- pool: race-free pmempool create --max-size
- obj: don't modify remote pools in pmemobj_check
Tue Feb 19 2019 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.5.1
This release fixes minor bugs and improves compatibility with newer
tool chains.
Notable bug fixes:
- common: make detection of device-dax instances more robust
- obj: fix pmemobj_check for pools with some sizes
- obj: don't use anon struct in an union (public header)
- obj: fix pmemobj_tx_lock error handling
- obj: don't use braces in an expression with clang (public header)
- obj: suppress pmemcheck warnings for statistics
- pmreorder: fix markers nontype issue
Fri Oct 26 2018 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.5
This release has had two major focus areas - performance and RAS
(Reliability, Availability and Serviceability). Beyond that, it introduces
new APIs, new tools and many other improvements. As a side effect
of performance optimizations, the libpmemobj on-media layout had to be
changed, which means that old pools have to be converted using pmdk-convert.
libpmemcto experiment has been finished and removed from the tree.
For more details, please see https://pmem.io/2018/10/22/release-1-5.html.
New features:
- common: unsafe shutdown detection (SDS)
- common: detection and repair of uncorrectable memory errors (bad blocks)
- pool: new "feature" subcommand for enabling and disabling detection of
unsafe shutdown and uncorrectable memory errors
- common: auto flush detection on Windows (on Linux since 1.4)
- pmreorder: new tool for verification of persistent memory algorithms
- obj: new on media layout
- pmem/obj: new flexible memcpy|memmove|memset API
- obj: new flushing APIs: pmemobj_xpersist, pmemobj_xflush
(PMEMOBJ_F_RELAXED)
- rpmem: new flag RPMEM_PERSIST_RELAXED for rpmem_persist
- obj: lazily initialized volatile variables (pmemobj_volatile)
(EXPERIMENTAL)
- obj: allocation classes with alignment
- obj: new action APIs: pmemobj_defer_free, POBJ_XRESERVE_NEW,
POBJ_XRESERVE_ALLOC
- blk/log: new "ctl" API
Optimizations:
- obj: major performance improvements for AEP NVDIMMs
- obj: better space utilization for small allocations
- common: call msync only on one page for deep drain
Other changes:
- cto: removed
- obj: remove actions limit
- common: new dependency on libndctl on Linux
- pmempool: "convert" subcommand is now a wrapper around pmdk-convert
(please see https://github.com/pmem/pmdk-convert)
- obj: C++ bindings have been moved to a new repository
(please see https://github.com/pmem/libpmemobj-cpp)
Bug fixes:
- obj: fix type numbers for pmemobj_list_insert_new
- pmem: fix inconsistency in pmem_is_pmem
- common: fix windows mmap destruction
- daxio: fix checking and adjusting length
- common: fix long paths support on Windows
Thu Aug 16 2018 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.4.2
This release fixes the way PMDK reports its version via
pkg-config files.
Bug fixes:
- common: fix reported version
- doc: use single "-" in NAME section (pmem/issues#914)
Fri Jun 29 2018 Marcin Ślusarz <marcin.slusarz@intel.com>
* Version 1.4.1
In 1.4 development cycle, we created new daxio utility
(command line tool for performing I/O on Device-DAX), but due
to some complications we had to disable it just before
the 1.4 release.
In 1.4.1 we finally enable it. Daxio depends on ndctl v60.1.
Bug fixes:
- pmem: fix clflush bit position
- obj: fix invalid OOMs when zones are fully packed
- obj: don't register undo logs twice in memcheck
- pool: fix bash completion script
- pool: fix incorrect errno after transform
- obj: fix clang-7 compilation
- obj: test for msync failures in non-pmem path
- doc: add missing field to alloc class entry point
- common: (win) fix timed locks
- common: provide src version in GitHub tarballs
- common: fix free function in tls
- common: fix double close
- test: allow testing installed libraries
- test: fix Valgrind vs stripped libraries issue
- test: fix dependencies between tests and tools
- test: fix races on make pcheck -jN
- test: use libvmmalloc.so.1
- test: fix incorrect number of required dax devices
- test: add suppression for leak in ld.so
- test: fail if memcheck detects overlapping chunks
- test: simplify time measurements in obj_sync
- benchmark: check lseek() return value
- examples: catch exceptions in map_cli
Thu Mar 29 2018 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.4
This is the first release of PMDK under a new name.
The NVML project has been renamed to PMDK (Persistent Memory
Development Kit). This is only the project/repo name change
and it does not affect the names of the PMDK packages.
See this blog article for more details on the reasons and
impact of the name change:
https://pmem.io/2017/12/11/NVML-is-now-PMDK.html
New features:
- common: support for concatenated Device-DAX devices
with 2M/1G alignment
- common: add support for MAP_SYNC flag
- common: always enable Valgrind instrumentation (pmem/issues#292)
- common: pool set options / headerless pools
- pmem: add support for "deep flush" operation
- rpmem: add rpmem_deep_persist
- doc: split man pages and add per-function aliases (pmem/issues#385)
Optimizations:
- pmem: skip CPU cache flushing when eADR is available
(no Windows support yet)
- pmem: add AVX512F support in pmem_memcpy/memset (pmem/issues#656)
Bug fixes:
- common: fix library dependencies (pmem/issues#767, RHBZ #1539564)
- common: use rpm-config CFLAGS/LDFLAGS when building packages
(pmem/issues#768, RHBZ #1539564)
- common: do not unload librpmem on close (pmem/issues#776)
- common: fix NULL check in os_fopen (pmem/issues#813)
- common: fix missing version in .pc files
- obj: fix cancel of huge allocations (pmem/issues#726)
- obj: fix error handling in pmemobj_open (pmem/issues#750)
- obj: validate pe_offset in pmemobj_list_* APIs (pmem/issues#772)
- obj: fix add_range with size == 0 (pmem/issues#781)
- log: add check for negative iovcnt (pmem/issues#690)
- rpmem: limit maximum number of lanes (pmem/issues#609)
- rpmem: change order of memory registration (pmem/issues#655)
- rpmem: fix removing remote pools (pmem/issues#721)
- pool: fix error handling (pmem/issues#643)
- pool: fix sync with switched parts (pmem/issues#730)
- pool: fix sync with missing replica (pmem/issues#731)
- pool: fix detection of Device DAX size (pmem/issues#805)
- pool: fail pmempool_sync if there are no replicas (pmem/issues#816)
- benchmark: fix calculating standard deviation (pmem/issues#318)
- doc: clarify pmem_is_pmem behavior (pmem/issues#719)
- doc: clarify pmemobj_root behavior (pmem/issues#733)
Experimental features:
- common: port PMDK to FreeBSD
- common: add experimental support for aarch64
- obj: introduce allocation classes
- obj: introduce two-phase heap ops (reserve/publish)
(pmem/issues#380, pmem/issues#415)
- obj: provide basic heap statistics (pmem/issues#676)
- obj: implement run-time pool extending (pmem/issues#382)
- cto: add close-to-open persistence library (pmem/issues#192)
The following features are disabled by default, until
ndctl v60.0 is available:
- daxio: add utility to perform I/O on Device-DAX
- RAS: unsafe shutdown detection/handling
Wed Dec 20 2017 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.3.1
Bug fixes:
- rpmem: fix issues reported by Coverity
- rpmem: fix read error handling
- rpmem: add fip monitor (pmem/issues#597)
- test: add rpmemd termination handling test
- cpp: fix pop.persist function in obj_cpp_ptr
- rpmem: return failure for a failed allocation
- rpmem: fix potential memory leak
- common: fix available rm options msg (pmem/issues#651)
- pool: fix pmempool_get_max_size
- obj: fix potential deadlock during realloc
(pmem/issues#635, pmem/issues#636, pmem/issues#637)
- obj: initialize TLS data
- rpmem: fix cleanup if fork() failed (pmem/issues#634)
- obj: fix bogus OOM after exhausting first zone
Thu Jul 13 2017 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.3
This release introduces some useful features and optimizations
in libpmemobj. Most of them are experimental and controlled by
the new pmemobj_ctl APIs. For details, please check the feature
requests identified by the issue numbers listed next to the items below.
Other important changes are related to performance tuning and
stabilization of librpmem library, which is used by libpmemobj to get
remote access to persistent memory and to provide basic data replication
over RDMA. The librpmem is still considered experimental.
NVML for Windows is feature complete (except for libvmmalloc).
This release includes the support for Unicode, long paths and
the NVML installer.
New features:
- common: add support for concatenated DAX Devices
- common: add Unicode support on Windows
- common: add long path support on Windows
- common: add NVML installer for Windows
- pmem: make pmem_is_pmem() true for Device DAX only
- obj: add pmemobj_wcsdup()/pmemobj_tx_wcsdup() APIs
- obj: export non-inlined pmemobj_direct()
- obj: add PMEMOBJ_NLANES env variable
- cpp: introduce the allocator
- cpp: add wstring version of C++ entry points
- vmem: add vmem_wcsdup() API entry
- pool: add pmempool_rm() function (pmem/issues#307)
- pool: add --force flag for create command (pmem/issues#529)
- benchmark: add a minimal execution time option
- benchmark: add thread affinity option
- benchmark: print 99% and 99.9% percentiles
- doc: separate Linux/Windows version of web-based man pages
Optimizations:
- obj: cache _pobj_cached_pool in pmemobj_direct()
- obj: optimize thread utilization of buckets
- obj: stop grabbing a lock when querying pool ptr
- rpmem: use multiple endpoints
Bug fixes:
- common: fix issues reported by static code analyzers
- pmem: fix mmap() implementation on Windows
- pmem: fix mapping addr/length alignment on Windows
- pmem: fix PMEM_MMAP_HINT implementation on Windows
- pmem: fix pmem_is_pmem() on invalid memory ranges
- pmem: fix wrong is_pmem returned by pmem_map_file()
- pmem: fix mprotect() for private mappings on Windows
- pmem: modify pmem_is_pmem() behavior for len==0
- obj: add failsafe to prevent allocs in constructor
- cpp: fix swap implementation
- cpp: fix sync primitives' constructors
- cpp: fix wrong pointer type in the allocator
- cpp: return persistent_ptr::swap to being public
- pool: treat invalid answer as 'n'
- pool: unify flags value for dry run
- pool: transform for remote replicas
- rpmem: persistency method detection
- benchmark: fix time measurement
Experimental features/optimizations:
- obj: pmemobjctl - statistics and control submodule
(pmem/issues#194, pmem/issues#211)
- obj: zero-overhead allocations - customizable alloc header
(pmem/issues#347)
- obj: flexible run size index (pmem/issues#377)
- obj: dynamic range cache (pmem/issues#378)
- obj: asynchronous post-commit (pmem/issues#381)
- obj: configurable object cache (pmem/issues#515)
- obj: add cache size and threshold tx params
- obj: add CTL var for suppressing expensive checks
- rpmem: add rpmem_set_attr() API entry
- rpmem: switch to libfabric v1.4.2
Thu May 18 2017 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.2.3
Bug fixes:
- test: extend timeout for selected tests
- test: reduce number of operations in obj_tx_mt
- test: define cfree() as free() in vmmalloc_calloc
Other changes:
- common: move Docker images to new repo
Sat Apr 15 2017 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.2.2
Bug fixes:
- pmempool: fix mapping type in pool_params_parse
- test: limit number of arenas in vmem_stats
- test: do not run pool_lock test as root
- common: fix pkg-config files
- common: fix building packages for Debian
Tue Feb 21 2017 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.2.1
This NVML release changes the behavior of pmem_is_pmem() on Linux.
The pmem_is_pmem() function will now return true only if the entire
range is mapped directly from Device DAX (/dev/daxX.Y) without an
intervening file system, and only if the corresponding file mapping
was created with pmem_map_file(). See libpmem(7) for details.
Bug fixes:
- jemalloc: fix test compilation on Fedora 26 (rawhide)
- test: fix cpp test compilation on Fedora 26 (rawhide)
- common: use same queue.h on linux and windows
- common: queue.h clang static analyzer fix
- common: fix path handling in build-dpkg.sh
- test: fix match files in pmempool_transform/TEST8
Fri Dec 30 2016 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.2 - Windows Technical Preview #1
This is the first Technical Preview release of NVML for Windows.
It is based on NVML 1.2 version, but not all the 1.2 features
are ported to Windows. In particular, Device DAX and remote access
to persistent memory (librpmem) are not supported by design.
NOTE: This release has not gone through the full validation cycle,
but only through some basic tests on Travis and AppVeyor.
Thus, it cannot be assumed "Production quality" and should not
be used in production environments.
Besides several minor improvements and bug fixes, all the other
changes since NVML 1.2 release were related to Windows support:
- win: port libvmem (and jemalloc)
- win: benchmarks Windows port
- win: fix mapping files of unaligned length
- win: clean up possible race condition in mmap_init()
- win: enable QueryVirtualMemoryInformation() in pmem_is_pmem()
- test: check open handles at START/DONE
- test: port all the remaining unit tests
(scope, pmem_map, obj_debug, util_poolset, pmempool_*)
- win: add resource files for versioning
Known issues and limitations of Windows version of NVML:
- Unicode support is missing. The UTF/USC-encoded file paths
or pool set files may not be handled correctly.
- The libvmmalloc library is not ported yet.
- The on-media format of pmem pools is not portable at the moment.
The pmem pools created using Windows version of NVM libraries
cannot be open on Linux and vice versa.
- Despite the fact the current version of NVML would work
with any recent version of Windows OS, to take full
advantage of PMEM and NVML features and to benefit from
the PMEM performance, the recommended platforms needs be
equipped with the real NVDIMMs hardware and should support
the native, Microsoft's implementation of DAX-enabled file
system (i.e. Windows Server 2016 or later).
In case of using NVML with older versions of Windows or with
the custom implementation of PMEM/DAX drivers, the performance
might not be satisfactory. Please, contact the provider
of PMEM/DAX drivers for your platform to get the customized
version of NVML in such case.
Thu Dec 15 2016 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.2
This NVML release causes a "flag day" for libpmemobj.
The pmemobj pools built under NVML 1.1 are incompatible with pools built
under NVML 1.2 and later. This is because an issue was discovered with
the alignment of locks (pmem/issues#358) and, although rare, the issue potentially
impacts program correctness, making the fix mandatory.
The major version number of the pmemobj pool layout and the version
of the libpmemobj API is changed to prevent the use of the potentially
incorrect layout.
Other key changes introduced in this release:
- Add Device DAX support, providing that "optimized flush" mechanism
defined in SNIA NVM Programming Model can safely be used, even
if PMEM-aware file system supporting that model is not available,
or if the user does not want to use the file system for some reason.
- Add a package for libpmemobj C++ bindings.
C++ API is no longer considered experimental.
Web-based documentation for C++ API is available on https://pmem.io.
- Add "sync" and "transform" commands to pmempool utility.
The "sync" command allows one to recover missing or corrupted part(s)
of a pool set from a healthy replica, while the "transform" command
is a convenient way for modifying the structure of an existing
pool set, i.e. by adding or removing replicas.
- Add experimental support for remote access to persistent memory and
basic remote data replication over RDMA (librpmem). Experimental
support for remote replicas is also provided by libpmemobj library.
New features:
- common: add Device DAX support (pmem/issues#197)
- obj: add C++ bindings package (libpmemobj++-devel)
- obj: add TOID_OFFSETOF macro
- pmempool: add "sync" and "transform" commands
(pmem/issues#172, pmem/issues#196)
Bug fixes:
- obj: force alignment of pmem lock structures (pmem/issues#358)
- blk: cast translation entry to uint64_t when calculating data offset
- obj: fix Valgrind instrumentation of chunk headers and cancelled
allocations
- obj: set error message when user called pmemobj_tx_abort()
- obj: fix status returned by pmemobj_list_insert() (pmem/issues#226)
- obj: defer allocation of global structures
Optimizations:
- obj: fast path for pmemobj_pool_by_ptr() when inside a transaction
- obj: simplify and optimize allocation class generation
Experimental features:
- rpmem: add support for remote access to persistent memory and basic
remote data replication over RDMA
- libpmempool: add pmempool_sync() and pmempool_transform() (pmem/issues#196)
- obj: introduce pmemobj_oid()
- obj: add pmemobj_tx_xalloc()/pmemobj_tx_xadd_range() APIs and
the corresponding macros
- obj: add transaction stage transition callbacks
Thu Jun 23 2016 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.1
This NVML release introduces a new version of libpmemobj pool layout.
Internal undo log structure has been modified to improve performance
of pmemobj transactions. Memory pools created with older versions
of the libpmemobj library must be converted to the new format using
"pmempool convert" command. See pmempool-convert(1) for details.
A new "libpmempool" library is available, providing support for
off-line pool management and diagnostics. Initially it provides only
"check" and "repair" operations for log and blk memory pools, and for
BTT devices.
Other changes:
- pmem: deprecate PCOMMIT
- blk: match BTT Flog initialization with Linux NVDIMM BTT
- pmem: defer pmem_is_pmem() initialization (pmem/issues#158)
- obj: add TOID_TYPEOF macro
Bug fixes:
- doc: update description of valid file size units (pmem/issues#133)
- pmempool: fix --version short option in man page (pmem/issues#135)
- pmempool: print usage when running rm without arg (pmem/issues#136)
- cpp: clarify polymorphism in persistent_ptr (pmem/issues#150)
- obj: let the before flag be any non-zero value (pmem/issues#151)
- obj: fix compare array pptr to nullptr (pmem/issues#152)
- obj: cpp pool.get_root() fix (pmem/issues#156)
- log/blk: set errno if replica section is specified (pmem/issues#161)
- cpp: change exception message (pmem/issues#163)
- doc: remove duplicated words in man page (pmem/issues#164)
- common: always append EXTRA_CFLAGS after our CFLAGS
Experimental features:
- Implementation of C++ bindings for libpmempobj is complete.
Web-based documentation for C++ API is available on https://pmem.io.
Note that C++ API is still considered experimental. Do not use it
in production environments.
- Porting NVML to Windows is in progress. There are MS Visual Studio
solution/projects available, allowing to compile libpmem, libpmemlog,
libpmemblk and libpmemobj on Windows, but the libraries are not fully
functional and most of the test are not enabled yet.
Thu Apr 07 2016 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 1.0
The API of six libraries (libpmem, libpmemblk, libpmemlog,
libpmemobj, libvmem, libvmmalloc) is complete and stable.
The on-media layout of persistent memory pools will be maintained
from this point, and if changed it will be backward compatible.
Man pages are all complete.
This release has been validated to "Production quality".
For the purpose of new features planned for next releases of NVML
there have been some API modifications made:
- pmem: pmem_map replaced with pmem_map_file
- log/blk: 'off_t' substituted with 'long long'
- obj: type numbers extended to 64-bit
- obj: new entry points and macros added:
pmemobj_tx_errno, pmemobj_tx_lock, pmemobj_mutex_timedlock,
TX_ADD_DIRECT, TX_ADD_FIELD_DIRECT, TX_SET_DIRECT
Other key changes since version 0.4 include:
- common: updated/fixed installation scripts
- common: eliminated dependency on libuuid
- pmem: CPU features/ISA detection using CPUID
- obj: improved error handling
- obj: atomic allocation fails if constructor returns error
- obj: multiple performance optimizations
- obj: object store refactoring
- obj: additional examples and benchmarks
This release also introduces a prototype implementation
of C++ bindings for libpmemobj. Note that C++ API is still
experimental and should not be used in production environments.
Fri Dec 04 2015 Krzysztof Czurylo <krzysztof.czurylo@intel.com>
* Version 0.4
This NVML version primarily focuses on improving code quality
and reliability. In addition to a couple of bug fixes,
the changes include:
- benchmarks for libpmemobj, libpmemblk and libvmem
- additional pmemobj tests and examples
- pool mapping address randomization
- added pmempool "rm" command
- eliminated libpmem dependency on libpthread
- enabled extra warnings
- minor performance improvements
Man pages are all complete.
This release is considered "Beta quality" by the team, having
been thoroughly validated, including significant performance analysis.
The pmempool command does not yet support "check" and "repair"
operations for pmemobj type pools.
Sun Sep 13 2015 Andy Rudoff <andy.rudoff@intel.com>
* Version 0.3
NVML is now feature complete, adding support for:
- pool sets
- pmemobj local replication (active/passive)
- experimental valgrind support
- pmempool support for all pool types
Man pages are all complete.
This release is considered "Alpha quality" by the team, having
gone through significant validation but only some performance
analysis at this point.
Tue Jun 30 2015 Andy Rudoff <andy.rudoff@intel.com>
* Version 0.2
NVML now consists of six libraries:
- libpmem (basic flushing, etc)
- libpmemblk, libpmemlog, libpmemobj (transactions)
- libvmem, libvmmalloc (volatile use of pmem)
The "pmempool" command is available for managing pmem files.
Man pages for all the above are complete.
The only things documented in man pages but not implemented are:
- pmem sets (ability to spread a pool over a set of files)
- replication (coming for libpmemobj)
The pmempool command does not yet support pmemobj type pools.
Thu Sep 11 2014 Andy Rudoff <andy.rudoff@intel.com>
* Version 0.1
Initial development done in 0.1 builds
Generated by dwww version 1.15 on Sun Jun 16 14:18:28 CEST 2024.