Because memory controller prefetch memory blocks, execution time of functions
calls sequence depend on order this functions in memory. For example:
4 calls:
call func1
call func2
call func3
call func4
faster in case of direct functions order in memmory:
.p2align 4
func1:
ret
.p2align 4
func2:
ret
.p2align 4
func3:
ret
.p2align 4
func4:
ret
and slow in case inverse order:
.p2align 4
func4:
ret
.p2align 4
func3:
ret
.p2align 4
func2:
ret
.p2align 4
func1:
ret
Unfortunately, inverse order is typical for C/C++.
what do you think about this kind optimization?
--
Summary: Optimization by functios reordering.
Product: gcc
Version: 4.4.0
Status: UNCONFIRMED
Severity: enhancement
Priority: P3
Component: c
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: vvv at ru dot ru
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40093
