Re: -O2 vs -O3

From: Horst von Brand (vonbrand@sleipnir.valparaiso.cl)
Date: Sat May 27 2000 - 14:13:32 EDT

  • Next message: Horst von Brand: "Re: [KBUILD] Re: Announcing CML2, a replacement for the kbuild system"

    Christopher Thompson <chris@hypocrite.org> said:
    > Here are some stats just from compiling the kernel. Please note that I did
    > *not* try booting any of these kernels so they might not actually work:
    >
    > quoted time is user time on an unloaded system, compiling 2.3.99-pre8
    >
    > -O0: fails compile
    > -O1: 7 min 22 s --> 807594 bytes

    Shouldn't work. Fuction inlining is only done with -O2 and above (strange
    it linked at all). Most probably modules will fail for want of inb(),
    outb() et al.

    > -O2: 8 min 38 s --> 830397 bytes
    > -O3: 8 min 55 s --> 869151 bytes

    Cost of overinlining and overzealous loop replication...

    > -Os: 8 min 36 s --> 819874 bytes

    No inlining at all, will probably kill several places that expect inlining. 

    --
Horst von Brand                             vonbrand@sleipnir.valparaiso.cl
Casilla 9G, Viņa del Mar, Chile                               +56 32 672616

    -
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

    This archive was generated by hypermail 2b29 : Sat May 27 2000 - 20:28:21 EDT