diff options
author | Nicolai Hähnle <nicolai.haehnle@amd.com> | 2016-10-04 16:06:31 +0200 |
---|---|---|
committer | Nicolai Hähnle <nicolai.haehnle@amd.com> | 2016-10-05 15:39:39 +0200 |
commit | 8915f0c0de84fa593ca6c31518c1292f94b3bb7b (patch) | |
tree | 2c719a8879546dc843ed07aa14375eff7291ce6c /doxygen | |
parent | d51c1f9d51ef0e80873a9a32c48606cdce25a209 (diff) |
util: use GCC atomic intrinsics with explicit memory model
This is motivated by the fact that p_atomic_read and p_atomic_set may
somewhat surprisingly not do the right thing in the old version: while
stores and loads are de facto atomic at least on x86, the compiler may
apply re-ordering and speculation quite liberally. Basically, the old
version uses the "relaxed" memory ordering.
The new ordering always uses acquire/release ordering. This is the
strongest possible memory ordering that doesn't require additional
fence instructions on x86. (And the only stronger ordering is
"sequentially consistent", which is usually more than you need anyway.)
I would feel more comfortable if p_atomic_set/read in the old
implementation were at least using volatile loads and stores, but I
don't see a way to get there without typeof (which we cannot use here
since the code is compiled with -std=c99).
Eventually, we should really just move to something that is based on
the atomics in C11 / C++11.
Acked-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Diffstat (limited to 'doxygen')
0 files changed, 0 insertions, 0 deletions