Re: kbuild vs ccache

[Sam Ravnborg]

On Wed, Jan 07, 2009 at 08:03:36PM -0800, Roland McGrath wrote:
> 
> make clean; ccache -s > before-`git describe`; make -j9; ccache -s > after-`git describe`; diff -u {before,after}-`git describe`
> 
> v2.6.28:
> 
> -cache hit                         737661
> -cache miss                        979624
> -called for link                    78268
> +cache hit                         738093		= +432
> +cache miss                        980524		= +900
> +called for link                    78289		= +21
>  multiple source files                373
>  compile failed                      5679
>  preprocessor error                  1594
> -not a C/C++ file                   44512
> +not a C/C++ file                   44549		= +37
> 
> 1332	cacheable
> 58	uncacheable
> 
> 
> v2.6.28-7485-g9e42d0c:
> 
> -cache hit                         738096		= +9
> +cache hit                         738105
>  cache miss                        980524		= +0
> -called for link                    78293
> +called for link                    79659		= +1366
>  multiple source files                373
>  compile failed                      5679
>  preprocessor error                  1594
> -not a C/C++ file                   44549		= +1384
> +not a C/C++ file                   45933
> 
> 9	cacheable
> 2750	uncacheable
> 
> 
> What changed?

Before we did a straight .c -> .o 
Now we do .c -> .s -> .o to mark __crc_ symbols static so we can
kill them when stripping modules.

It is already being discussed in another thread.
See "[REGRESSION] Recent change to kernel spikes out ccache/distcc"

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