[distcc] How execution time (not compilation time) of a process can be reduced

Martin Pool mbp at sourcefrog.net
Wed Jul 23 05:34:38 GMT 2003


On 23 Jul 2003, murad ali <murad_77 at hotmail.com> wrote:

> Not only this , almost 60/70 lines are scrolled, and all objects
> files are created , it means gmake work properly with distcc. All
> objects files were created , but when I issue command for execution
> as % ./executable then program takes so much time as without
> distcc. Kindly tell me that >how this execution time can be reduced.

> All things work Properly , But all processors of systems are not
> utilized during EXECUTION of programs.

Oh, I see: you want distcc to make your program run faster, rather
than compile faster?  

To do that, distcc would need to mechanically transform an arbitrary C
program into a distributed parallel program.  It is not quite
impossible (people have written parallelizing compilers) but it's very
very hard, especially for a shared-nothing high-latency setup like a
cluster of workstations.  Consider even just the possible aliasing
effects that are present in a C program, and the fact that every
shared-memory update needs to be sent across the network.

If you want your program to be distributed at run time across several
machines, you need to design in parallelism and distribution,
possibly using a framework such as MPI or OpenMOSIX.

Did something in the distcc documentation make you think it could do
this?  Can you tell me what it is so that it can be made more clear?

-- 
Martin 



More information about the distcc mailing list