Re: [ANNOUNCE] "Fast Kernel Headers" Tree -v2

[Ingo Molnar <mingo@kernel.org>]

* Arnd Bergmann <arnd@arndb.de> wrote:

> > If we include comments & line-markers then the bloat goes up by another
> > ~2x:
> >
> >  kepler:~/mingo.tip.git> ./st include/linux/sched.h
> >   #include <linux/sched.h>                | LOC:  2,186 | headers:  118
> >  kepler:~/mingo.tip.git> ./st include/linux/sched.h
> >   #include <linux/sched.h>                | LOC:  4,092 | headers:    0
> 
> The metric I've been focusing on is bytes of the preprocessed header, 
> which is more sensitive to function definitions that get generated from 
> macros, and I multiply this by the number of inclusions (from scanning 
> the .file.o.cmd files). It probably helps to have a couple of metrics and 
> look at all of them occasionally to not miss something important.

Actual inclusions don't just depend on .file.o.cmd files though, that won't 
catch indirect inclusions, right?

> In the meantime, I have made some progress on reducing the headers
> for arm64, on top of your tree from Jan 8, but I have not looked at
> later changes from your side, and I need to work on this a bit more
> to ensure this doesn't break other architectures.

Sure & great!

> For an arm64 allmodconfig build, my additional improvements on top
> of yours are significant but not as good as I had hoped for, this
> can still improve I hope:
> 
> 5.16-rc8-vanilla  32640 seconds user, 3286 seconds sys
> 5.16-rc8-mingo  22990 seconds user, 2304 seconds sys
> 5.16-rc8-arnd   19007 seconds user, 1853 seconds sys

~71% build throughput speedup for allmodconfig is very impressive to me. :-)

> As my tree builds any randconfig cleanly, [...]

Yeah, same here - having a few thousand randconfig build tests is normal 
for each version:

  /* This file is auto generated, version 3288 */
  #define UTS_MACHINE "x86_64"
  #define UTS_VERSION "#3288 Fri Jan 14 18:20:14 CET 2022"

My testing is mostly concentrated on x86 - but I often test ARM64 
randconfig as well.

> I keep looking at different configs and find that this has a big impact, 
> some options end up eliminating most of the benefits until I add further 
> changes to clean up certain files. This happened with kasan, kprobes, and 
> lse-atomics for instance. After eliminating all circular includes, I was 
> also able to revisit my old script to visualize the inclusions, see[1] 
> for the current arm64 defconfig output. This version uses my arbitrary 
> metric as font-size, and uses labels for the number of inclusions.

This is really nice!

I was concentrating on optimizing a generic distro config - which doesn't 
include the tons of extreme instrumentation measures that allmodconfig 
includes but production distro kernels rarely do.

allmodconfig definitely needs more work, but 71% is a pretty good starting 
point ...

Feel free to send in patches, I can help with the testing too.

Thanks,

	Ingo