This is the mail archive of the mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: PATCH for Re: 3.0.1 performance

On Fri, Aug 17, 2001 at 09:16:05PM +0200, Gerald Pfeifer wrote:
> retrieving revision 1.44
> diff -u -3 -p -r1.44 invoke.texi
> --- invoke.texi	2001/08/03 01:19:20	1.44
> +++ invoke.texi	2001/08/17 19:13:47
> @@ -3310,7 +3310,8 @@ allows the control of this limit for fun
>  inline (ie marked with the inline keyword or defined within the class
>  definition in c++).  @var{n} is the size of functions that can be inlined in
>  number of pseudo instructions (not counting parameter handling).  The default
> -value of @var{n} is 10000.  Increasing this value can result in more inlined code at
> +value of @var{n} is 600.
> +Increasing this value can result in more inlined code at
>  the cost of compilation time and memory consumption.  Decreasing usually makes
>  the compilation faster and less code will be inlined (which presumably
>  means slower programs).  This option is particularly useful for programs that


I'd add some sentence telling that this vvalue used to be higher in previous
versions of the compiler and that a possible slowdown of code execution 
might be caused by this change and might be helped by trying higher values.

I still hope somebody will increase this value ...

Kurt Garloff                   <>         [Eindhoven, NL]
Physics: Plasma simulations  <K.Garloff@Phys.TUE.NL>  [TU Eindhoven, NL]
Linux: SCSI, Security          <>    [SuSE Nuernberg, DE]
 (See mail header or public key servers for PGP2 and GPG public keys.)

PGP signature

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]