|Re: Procedure calls; compilers and register windows firstname.lastname@example.org (1992-02-04)|
|Re: Procedure calls; compilers and register windows email@example.com.Virginia.EDU (1992-02-06)|
|Posted-Date:||Thu, 6 Feb 92 14:43:02 EST|
|Date:||Thu, 6 Feb 92 14:43:02 EST|
>[From comp.arch -John]
>> On 3 Feb 92 23:46:05 GMT, firstname.lastname@example.org (Daniel Walter Krueger) said:
>> I am hoping to find some information on both static and dynamic
>> mechanisms of reducing the overhead involved in procedure calls.
>> [Details deleted.]
If you would consider reducing the overhead involved in procedure calls by
eliminating those calls through subprogram inlining, you may be interested
in a paper by Jack Davidson and me that is slated to appear in the
February 1992 issue of IEEE Transactions on Software Engineering. The
paper is titled "Subprogram Inlining: A Study of its Effects on Program
Execution Time". In addition to reducing program execution time, inlining
was found in our study to yield modest improvements in caching and paging
behavior. Also, we did not find a performance penalty associated with the
increased size of our inlined test programs on a set of demand-paged
virtual memory machines.
Anne M. Holler (email@example.com)
Return to the
Search the comp.compilers archives again.