projects / lttng-modules.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 21c2a61) | patch
Performance: mark lib_ring_buffer_write always inline
authorMathieu Desnoyers <mathieu.desnoyers@efficios.com>
Sun, 25 Sep 2016 14:43:22 +0000 (10:43 -0400)
committerMathieu Desnoyers <mathieu.desnoyers@efficios.com>
Sun, 25 Sep 2016 14:43:22 +0000 (10:43 -0400)
commit8617eb9a6f9dd4e63f8cb649120b3b3ae79df4f0
tree3c137cb32c76cc65997ea82a15d798806b1f88c6tree | snapshot
parent21c2a61f57670cabe298a9d53bdce0802a6c137ecommit | diff
Performance: mark lib_ring_buffer_write always inline

The underlying copy operation is more efficient if the size is a
constant, which only happens if this function is inlined in the caller.
Otherwise, we end up calling memcpy for each field.

Force inlining for performance reasons.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
lib/ringbuffer/backend.h diff | blob | blame | history
LTTng 2.0/0.x modules
This page took 0.025085 seconds and 4 git commands to generate.