Package Details: rocksdb 5.14.2-1

Git Clone URL: https://aur.archlinux.org/rocksdb.git (read-only)
Package Base: rocksdb
Description: Embedded key-value store for fast storage
Upstream URL: http://rocksdb.org
Licenses: Apache, leveldb
Submitter: anatolik
Maintainer: timlyo
Last Packager: timlyo
Votes: 19
Popularity: 0.290260
First Submitted: 2013-12-24 14:34
Last Updated: 2018-07-15 09:33

Latest Comments

1 2 3 Next › Last »

farseerfc commented on 2018-08-03 01:48

Lastest build needs the fix in upstream commit: https://github.com/facebook/rocksdb/commit/e5ae1bb46564689e56a38f3509daffa4aca3b29c A working PKGBUILD is here: https://fars.ee/025s/sh

FallenSnow commented on 2018-07-15 15:57

Latest build fails with:

util/status.cc: In static member function ‘static const char* rocksdb::Status::CopyState(const char*)’:
util/status.cc:28:15: error: ‘char* strncpy(char*, const char*, size_t)’ output truncated before terminating nul copying as many bytes from a string as its length [-Werror=stringop-truncation]
   std::strncpy(result, state, cch - 1);
   ~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
util/status.cc:19:18: note: length computed here
       std::strlen(state) + 1; // +1 for the null terminator
       ~~~~~~~~~~~^~~~~~~
cc1plus: all warnings being treated as errors
make: *** [Makefile:650: shared-objects/util/status.o] Error 1
make: *** Waiting for unfinished jobs....

Using $ gcc --version gcc (GCC) 8.1.1 20180531

lantw44 commented on 2018-05-29 13:32

rocksdb automatically adds -march=native to compiler flags. Can we add PORTABLE=1 in PKGBUILD to make it build portable binaries instead?

Jonhoo commented on 2018-05-18 03:18

@timlyo hmm, the current package does not build for me with all other packages up-to-date:

CC shared-objects/memtable/skiplistrep.o In file included from memtable/skiplistrep.cc:6: ./memtable/inlineskiplist.h: In instantiation of ‘void rocksdb::InlineSkipList<comparator>::Node::StashHeight(int) [with Comparator = const rocksdb::MemTableRep::KeyComparator&]’: ./memtable/inlineskiplist.h:612:3: required from ‘rocksdb::InlineSkipList<comparator>::Node rocksdb::InlineSkipList<comparator>::AllocateNode(size_t, int) [with Comparator = const rocksdb::MemTableRep::KeyComparator&; size_t = long unsigned int]’ ./memtable/inlineskiplist.h:574:13: required from ‘rocksdb::InlineSkipList<comparator>::InlineSkipList(Comparator, rocksdb::Allocator</comparator></comparator>, int32_t, int32_t) [with Comparator = const rocksdb::MemTableRep::KeyComparator&; int32_t =</comparator></comparator>

I believe this is one of the errors I linked above that is caused by gcc8.1. See specifically https://github.com/facebook/rocksdb/issues/2705#issuecomment-389580914.

timlyo commented on 2018-05-17 18:45

@Jonhoo I wasn't aware of that, thanks for raising it.

Even so I'm not comfortable with making those changes across the board to benefit a few when it has potential, even if small, to negatively impact the majority.

Jonhoo commented on 2018-05-17 18:19

The facebook developers are good about fixing warnings (that's why they enable -Werror in the first place). The concern in these cases is that the warnings aren't there due to changes to the code, but because of new warnings introduced by new versions of gcc. See for example https://github.com/facebook/rocksdb/issues/3486, https://github.com/facebook/rocksdb/issues/3716, and https://github.com/facebook/rocksdb/issues/2705.

timlyo commented on 2018-05-17 16:35

@Jonhoo I'm of the opinion that maintainers should take the minimal number of risks possible. Turning off warnings may break things, if people want to do that for themselves then the option is there, but I'd rather be a version or two behind than risk introducing vulnerabilities to all users.

The latest version seems to compile ok for me though, I'm going to push it out.

Jonhoo commented on 2018-05-16 16:28

Instead of holding the newer version back, wouldn't it be better to call

$ make DISABLE_WARNING_AS_ERROR=1

Which would leave these warnings as warnings, and let compilation succeed.

timlyo commented on 2018-03-16 15:40

@RcrdBrt That's great news, thanks for the update.

RcrdBrt commented on 2018-03-16 08:35

v5.11 definitely contains the "fall through" build fixes. Also, it compiles without problems on armv7h (raspberry). Please add it to the supported architectures