Fix undefined NULL pointer arithmetic

Clang 3.3 with -O2 optimisations is especially picky about arithmetic
on NULL pointers. This undefined behavior is turned into optimized out
NULL checks by clang 3.3. Fix the undefined behavior by checking against
the pointer directly, without going back and forth around NULL with
pointer arithmetic.

Reported-by: Zifei Tong <soariez@gmail.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Blacklist ARM gcc 4.8.0, 4.8.1, 4.8.2

It produces clobbered frame accesses, which can lead to stack corruption
when racing with signal handlers nested on stack.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

rculfhash: document max_nr_buckets = 0

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Version 0.8.1

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

tls-compat: fix comment typo

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Keep ABI compatible with already compiled LGPL applications

Applications with _LGPL_SOURCE defined that were compiled against bogus
tls-compat.h header *and* which are using multiple urcu flavors
concurrently need to be told that they need to be recompiled against a
fixed tls-compat.h header. Detect these usage, and abort() with a
message error on stderr.

This is needed for stable-0.7 and stable-0.8 branches of userspace RCU
only. The upcoming 0.9 will bump the soname version, and therefore does
not have to care about this aspect of ABI compatibility.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: tls-compat multi-lib conflict

When configured with the TLS pthread key fallback either:

- explicitly with ./configure --disable-compiler-tls,
- or if compiler TLS is not usable,

(this can be confirmed by looking at the configure output:
Thread Local Storage (TLS): pthread_getspecific().)

There is an issue when using multiple flavors of RCU within the same
program. Unit tests concerned:

tests/unit/test_urcu_multiflavor
tests/unit/test_urcu_multiflavor_dynlink

Vladimir Nikulichev noticed crashes when using this setup. The problem
can be pinpointed to a missing macro expansion in urcu/tls-compat.h:

looking at the output of

nm tests/unit/.libs/test_urcu_multiflavor :

                 U __tls_access_rcu_reader

this seems to be the issue. We're missing macro expansion in
tls-compat.h. With this commit, it becomes:

                 U __tls_access_rcu_reader_bp
                 U __tls_access_rcu_reader_mb
                 U __tls_access_rcu_reader_memb
                 U __tls_access_rcu_reader_sig

Please note that this affects an unusual configuration of userspace RCU
(with TLS pthread key fallback), needed for some BSD that don't support
compiler TLS. Strictly speaking, this requires bumping the URCU library
soname version major number, because it breaks the ABI presented to
applications on those unusual configurations.

A following commit will handle the ABI migration: for stable releases
(stable-0.7 and stable-0.8 branches), the ABI is kept compatible, and
bogus usage are detected. For the upcoming stable-0.9, the soname will
simply be bumped.

Reported-by: Vladimir Nikulichev <nvs@tbricks.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Use cross compiler for doc examples

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

gcc warning fixes: -Wsign-compare and -Wextra

When compiling code using the rcu_xchg_pointer() family of functions,
with the following define:

 #define URCU_INLINE_SMALL_FUNCTIONS

prior to including urcu headers, when compiling with gcc with
-Wsign-compare and -Wextra, gcc warns about:

urcu-xchg.c: In function ‘reload’:
urcu-xchg.c:19:1: warning: ordered comparison of pointer with integer zero [-Wextra]
urcu-xchg.c:19:1: warning: signed and unsigned type in conditional expression [-Wsign-compare]

For the "ordered comparison of pointer with integer zero" warning, fix
this by comparing (type) -1 against (type) 0 instead of just 0, so if
"type" is a pointer type, this pointer type will be applied to the right
operand too, thus fixing the warning.

For the "signed and unsigned type in conditional expression" warning, we
need caa_cast_long_keep_sign() to always evaluate to the same type
signedness. In order to do so, when we need to sign-extend the value,
cast it to unsigned long after first casting it to long.

Reported-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: urcu-qsbr: reversed logic on RCU_DEBUG

* Dmitri Shubin <sbn@tbricks.com> wrote:
> Shouldn't the condition in line 94 actually be
>
>      94  #if (!defined(BUILD_QSBR_LIB) && !defined(RCU_DEBUG))
>
> So when RCU_DEBUG is _not_ defined we get static inlines for
> rcu_read_{,un}lock() ?

Indeed!

Reported-by: Dmitri Shubin <sbn@tbricks.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: urcu-bp segfault in glibc pthread_kill()

This fixes an issue that appears after this recent urcu-bp fix is
applied:

  Fix: urcu-bp: Bulletproof RCU arena resize bug

Prior to this fix, on Linux at least, the behavior was to allocate
(and leak) one memory map region per reader thread. It worked, except
for the unfortunate leak. The fact that it worked, even though not the
way we had intended it to, is is why testing did not raise any red flag.
That state of affairs has prevailed for a long time, but it was
side-tracking some issues. After fixing the underlying bug that was
causing the memory map leak, another issue appears.

The garbage collection scheme reclaiming the thread tracking structures
in urcu-bp fails in stress tests to due a bug in glibc (tested against
glibc 2.13 and 2.17). Under this workload, on a 2-core/hyperthreaded i7:

./test_urcu_bp 40 4 10

we can easily trigger a segmentation fault in the pthread_kill() code.

Program terminated with signal 11, Segmentation fault.

Backtrace:

 #0  __pthread_kill (threadid=140723681437440, signo=0) at ../nptl/sysdeps/unix/sysv/linux/pthread_kill.c:42
 42 ../nptl/sysdeps/unix/sysv/linux/pthread_kill.c: No such file or directory.
 (gdb) bt full
 #0  __pthread_kill (threadid=140723681437440, signo=0) at ../nptl/sysdeps/unix/sysv/linux/pthread_kill.c:42
         __x = <optimized out>
         pd = 0x7ffcc90b2700
         tid = <optimized out>
         val = <optimized out>
 #1  0x0000000000403009 in rcu_gc_registry () at ../../urcu-bp.c:437
         tid = 140723681437440
         ret = 0
         chunk = 0x7ffcca0b8000
         rcu_reader_reg = 0x7ffcca0b8120
         __PRETTY_FUNCTION__ = "rcu_gc_registry"
 #2  0x0000000000402b9c in synchronize_rcu_bp () at ../../urcu-bp.c:230
         cur_snap_readers = {next = 0x7ffcb4888cc0, prev = 0x7ffcb4888cc0}
         qsreaders = {next = 0x7ffcb4888cd0, prev = 0x7ffcb4888cd0}
         newmask = {__val = {18446744067267100671, 18446744073709551615 <repeats 15 times>}}
         oldmask = {__val = {0, 140723337334144, 0, 0, 0, 140723690351643, 0, 140723127058464, 4, 0, 140723698253920, 140723693868864, 4096, 140723690370432,
             140723698253920, 140723059951840}}
         ret = 0
         __PRETTY_FUNCTION__ = "synchronize_rcu_bp"
 #3  0x0000000000401803 in thr_writer (_count=0x76b2f0) at test_urcu_bp.c:223
         count = 0x76b2f0
         new = 0x7ffca80008c0
         old = 0x7ffca40008c0
 #4  0x00007ffcc9c83f8e in start_thread (arg=0x7ffcb4889700) at pthread_create.c:311
         __res = <optimized out>
         pd = 0x7ffcb4889700
         now = <optimized out>
         unwind_buf = {cancel_jmp_buf = {{jmp_buf = {140723337336576, 6546223316613858487, 0, 140723698253920, 140723693868864, 4096, -6547756131873848137,
                 -6547872135220034377}, mask_was_saved = 0}}, priv = {pad = {0x0, 0x0, 0x0, 0x0}, data = {prev = 0x0, cleanup = 0x0, canceltype = 0}}}
         not_first_call = 0
         pagesize_m1 = <optimized out>
         sp = <optimized out>
         freesize = <optimized out>
         __PRETTY_FUNCTION__ = "start_thread"
 #5  0x00007ffcc99ade1d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:113

It appears that the memory backing the thread information can be
relinquished by NPTL concurrently with execution of pthread_kill()
targeting an already joined thread and cause this segfault. We were
using pthread_kill(tid, 0) to discover if the target thread was alive or
not, as documented in pthread_kill(3):

       If  sig  is 0, then no signal is sent, but error checking is still per‐
       formed; this can be used to check for the existence of a thread ID.

but it appears that the glibc implementation is racy.

Instead of using the racy pthread_kill implementation, implement cleanup
using a pthread_key destroy notifier for a dummy key. This notifier is
called for each thread exit and destroy.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix urcu-bp: don't move registry

It is not correct to move the registry address range, since there are
external references from reader threads. This will trigger on workloads
with many threads.

Typically, on Linux, mremap can expand the existing range, which is OK.
However, if there is not enough space around the existing range, it may
try to map it at a different address, which is incorrect.

It is more likely that this bug will be observed on operating systems
where urcu uses the mmap/munmap fallback instead of mremap.

Moreover, prior to commit:

  "Fix: urcu-bp: Bulletproof RCU arena resize bug"

this issue was hidden by the fact that each thread ended up with their
own memory mapping (leaked), on Linux at least.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: compat futex duplicated lock and completion

compat_futex.c has one instance included in each urcu shared object, as
well as within some of the test applications. However, it is expected
that an entire program interact with the same lock and completion
variables. Therefore, define them as globally visible, but weak, so the
entire program agree on which object should be used.

Reported-by: Vladimir Nikulichev <nvs@tbricks.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: i386 compat code duplicated mutex instances

compat_arch_x86.c is linked into many .so and even into test programs.
The basic problem with this is that it contains a statically defined
mutex, which will fail to protect concurrent use of this compat code by
different shared objects.

Fix this by defining both the mutex (now called __urcu_x86_compat_mutex)
and __rcu_cas_avail as weak symbols. Therefore, the first symbol that
gets loaded in a program will by used by everyone.

Reported-by: Vladimir Nikulichev <nvs@tbricks.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: urcu-bp: Bulletproof RCU arena resize bug

> From: "Milosz Tanski" <milosz@adfin.com>

> While trying to use the BP flavor of RCU I ran into random crashes. I
> tracked it down to issues with resizing of the BP RCU memory pool.
>
> The problem is in the urcu-bp.c file in the resize_arena() function.
> On successful allocation / remapping the len member of the
> registry_arena struct is never set anywhere function. On the second
> resize of the arena the code in resize_arena() still thinks the
> previous size is equal to the original mapping size. I've fixed this
> issue locally by just adding the following code at the bottom of
> resize_arena().

Good catch !!

However, I think your fix misses one case: if we happen to re-use the
same region, we want to update the length too.

Reported-by: Milosz Tanski <milosz@adfin.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: test_mutex.c uninitialized mutex

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Version 0.8.0

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: hash table growth (for small tables) should be limited

Buckets with many entries encountered in a hash table could cause it to
grow to a large size, beyond the scope for which this mechanism is
expected to play a role when node accounting is available. Indeed, when
the hash table grows to larger size, split-counter node accounting is
expected to deal with resize/shrink rather than relying on an heuristic
based on the largest bucket size.

This is fixing an issue where we see hash tables sometimes reaching 65k
entries index (65536*8 = 524288 bytes) for a workload limited to adding
1000 entries and then removing all of them, done in a loop (random
keys).

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: doc/examples cross-build

We need to pass automake's CPPFLAGS and LDFLAGS to examples. Add CFLAGS
too for completeness.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Introduce URCU_INLINE_SMALL_FUNCTIONS

Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Add missing tests/common/Makefile.am

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

README: document make check/regtest/bench

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

tests: split in check, regtest and bench targets

Allow make check to run without requiring build machines to use hours of
CPU time.

make check: short unit tests
make regtest: long regression tests
make bench: long benchmarks

Fixes #611

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Cleanup: doc/examples makefile

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: doc/examples VPATH build

Fixes #568

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: Move the LIBS after the OBJECTS in the Makefile

Fixes #569.

Signed-off-by: Alexandre Montplaisir <alexmonthy@voxpopuli.im>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Document build work-around on MacOS X

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix tests: use of uninitialized variables

Were working fine by luck, since they were allocated at the start of
newly spawned test programs. Identified by Coverity.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

test_urcu_hash*: initialize TLS seeds

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_lfht_for_each_entry_duplicate

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_lfht_lookup

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_lfht_destroy

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_lfht_add_replace

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_lfht_add_unique

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_lfht_add/cds_lfht_del

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: add rculfqueue example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: add synchronize_rcu()

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: add bp flavor

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: add dist toplevel makefile

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: add membarrier flavor

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: document call_rcu()

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: update qsbr example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

urcu signal: remove assertion on exit

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: signal flavor

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: add mb flavor

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: update qsbr

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: introduce urcu-flavors examples directory

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: enhance rcu-flavor-qsbr example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: rename qsbr-minimal to rcu-flavor-qsbr

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: automake stop on error

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: hlist

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

hlist/rcuhlist update

introduce:

- cds_hlist_for_each()
- cds_hlist_for_each_safe()
- CDS_HLIST_HEAD()
- CDS_HLIST_HEAD_INIT()
- cds_hlist_for_each_entry_2() (takes 3 argument, like the Linux kernel
  API),
- cds_hlist_for_each_entry_safe_2() (takes 4 arguments, like the Linux
  kernel API),
- cds_hlist_for_each_rcu()
- cds_hlist_for_each_entry_rcu_2() (takes 3 arguments, like the Linux
  kernel API).

Left cds_hlist_for_each_entry(), cds_hlist_for_each_entry_safe() and
cds_hlist_for_each_entry_rcu() as-is (different from the ones found in
the Linux kernel) because those APIs were already exposed by Userspace
RCU.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: fix typo in list example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

rcuhlist: make pointer stores atomic

Use rcu_assign_pointer() to store into head->next in
cds_hlist_add_head_rcu(). This includes the write barrier needed before
publishing the new node.

Use CMM_STORE_SHARED() to store into elem->prev->next in
cds_hlist_del_rcu().

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

hlist, rcuhlist: cleanup coding style

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: lfstack

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: update cds_wfs_pop_all_blocking

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_wfs_pop_all_blocking

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_wfs_pop

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: add missing Makefile

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_wfs_push

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/wfcqueue: cds_wfcq_splice

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: add cds_wfcq_dequeue

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: wfcq needs to link against urcu-common

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: update queue comment

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: fix make clean

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

gitignore: add qsbr-minimal

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_wfcq_enqueue

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: Move LIBS to each makefile

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_list_for_each_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_list_for_each_entry_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_list_replace_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_list_add_tail_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_list_del_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

doc/examples: cds_list_add_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

rculist: ensure atomic updates of next pointers

in cds_list_add_rcu, use rcu_assign_pointer to update head->next
atomically and provide the memory barrier before publishing head->next.
Notice that we don't need the wmb() prior to store to prev, because RCU
traversals only go forward, and thus only use "next".

in cds_list_del_rcu, use CMM_STORE_SHARED() to store to elem->prev->next
atomically.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

rculist: implement cds_list_add_tail_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

rculist.h and list.h style cleanup

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

example makefile: add missing cd ..

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Update gitignore

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: examples Makefile on FreeBSD

Support BSD-style make.

Fixes #565

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

hash table test: don't redefine CACHE_LINE_SIZE

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

tests: use thread-id.h wrapper

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Implement thread-id.h wrapper

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

tests: add missing unsigned long casts to pthread_self()

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: don't build examples in static builds

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Add QSBR minimal example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

compiler.h: implement CAA_ARRAY_SIZE()

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

document rcu barrier

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

rcu barrier: handle OOM die urcu_die

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Implement rcu_barrier()

Awaits for all in-flight call_rcu handlers to complete execution before
returning.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

rculfhash: document destroy context limitations

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Add MIPS to README

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Update README

Detail i386/amd64 as tested for FreeBSD.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Update README testing info about FreeBSD

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

test: fix api.h missing if brackets

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

tests: fix incorrect counter

- not initialized to 0 in 2 tests,
- within inner loop in 3rd test, thus always 0.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: membarrier fallback symbol conflict

* Lai Jiangshan (laijs@cn.fujitsu.com) wrote:
> Hi, Mathieu,
>
> There is a big compatible problem in URCU which should be fix in next round.
>
> LB: liburcu built on the system which has sys_membarrier().
> LU: liburcu built on the system which does NOT have sys_membarrier().
>
> LBM: liburcu-mb ....
> LUM: liburcu-mb ...
>
> AB: application(-lliburcu) built on the system which has sys_membarrier().
> AU: application(-lliburcu) built on the system which does NOT have
> sys_membarrier().
>
> ABM application(-lliburcu-mb) ...
> AUM application(-lliburcu-mb) ...
>
> AB/AU + LB/LU: 4 combinations
> ABM/AUM + LBM/LUM: 4 combinations
>
> I remember some of the 8 combinations can't works due to symbols are
> miss match.  only LU+AB and LB+AU ?
>
> could you check it?
>
> How to fix it: In LU and AU, keep all the symbol name/ABI as LA and
> AB, but only the behaviors falls back to URCU_MB.

Define membarrier() as -ENOSYS when SYS_membarrier is not found in the
system headers. Check dynamically for membarrier availability to ensure
ABI compatibility between applications and librairies.

Reported-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

Fix: Use a filled signal mask to disable all signals

Changelog from David Pelton's original patch:

While using lttng-ust with an application that was calling fork()
with pending signals, I found that all signals were getting unmasked
shortly before the underlying call to fork().  After some
investigation, I found that the rcu_bp_before_fork() function was
unmasking all signals.  Based on the comments for this function, it
should be masking all signals.  Inspection of the rest of the code
in urcu-bp.c revealed the same pattern in two other functions.

This patch changes the code to use a filled signal mask to disable
all signals.  The change to rcu_bp_before_fork() addressed the
problem I was seeing while using lttng-ust.  The changes to the
other two functions appear to fix other instances of the same
problem.

Updates by Mathieu Desnoyers:

- Use SIG_BLOCK instead of SIG_SETMASK when setting a filled mask. This
  has the same behavior in this case (since we're blocking all signals),
  but is semantically neater: if we ever some signals from that mask,
  we'd like to to a union with the signal mask already blocked by the
  application.
- Also fix incorrect signal masking in compat_arch_x86.c.

Reported-by: David Pelton <dpelton@ciena.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

urcu-bp: introduce struct urcu_gp

Make urcu-bp similar to urcu-qsbr and other urcu flavors.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>