MIME-Version: 1.0
From: Denys Vlasenko <vda.linux@googlemail.com>
Date: Sun, 12 Jul 2015 16:05:18 +0200
Message-ID: <CAK1hOcMFp9YcTKFen_ysXpv0dxst0v_Q=1V=7C8+GXLQhXKdkQ@mail.gmail.com>
Subject: Automated search for inline bloat
To: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Andrew Morton <akpm@linux-foundation.org>
Content-Type: multipart/mixed; boundary=001a11474bac81ebb0051aae1919
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 12797
Lines: 237

--001a11474bac81ebb0051aae1919
Content-Type: text/plain; charset=UTF-8

The scripts are in attached tarball.

Is there interest in putting them in, say, scripts/inline_bloat/
in the kernel tree?

============================================
        Inline hunting.

There are outrageously big inlines in kernel. Finding them by hand
is inefficient, if you want to handle the worst of them first.

Let's automate it.

Inline hunting in header files is done by replacing "inline" with "noinline",
rebuilding vmlinux, and seeing how much smaller it becomes.

(The accounting is more complex than simple comparison of "size vmlinux"
before and after, since deinlined function body is repeated in each
object file it is called from).


    HOWTO.

* Find a machine with lots of CPUs.
* Start with an empty work directory.
* Copy these scripts into it.
* git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
* cd linux && make oldconfig && make -j99 && cp vmlinux vmlinux.original
* Edit conf.py - this is crucial to optimally parallelize the next step,
  which consumes most of the time (usually a day or more).
* Run ./1start. Now parallel build jobs are running in linux.N/ dirs.
  N = conf.py::dir_count.
* Run ./2merge_and_sort_results:
  ...
  Total inlines to measure:22640, processed:15883
  Total inlines to measure:22640, processed:15889 (after 60 seconds)
  ...
  (This script generates the final result, inlines.log.measured.sorted,
  every minute. When all inlines are processed, it will stop)

Result files are:

inlines_err.log      - error messages from inline finding script
linux.N/inlines_err.log - each parallel job's stderr
inlines.log          - all inlines
inlines.log.sorted   - sorted by line count
inlines.log.filtered - selected for measurement (via conf.py::min_lines)
inlines.log.measured - measured code size change
inlines.log.measured.sorted - sorted, format is:

drivers/gpu/drm/radeon/radeon.h:2696:radeon_ring_write:7:46196:73
filename:lineno^^^^^^^^^^^^^^^^^^^^^|inline_name^^^^^^|^|saved|size_of_deinlined_fn
                                                         !lines_of_source_code
IOW:
Deinlining radeon_ring_write() shrinks kernel by 46196 bytes.
Deinlined function body is 73 bytes of code.


    Which config to use?

Well, we want to look at *all* code, so allyesconfig is a natural choice.

However, some options are clearly "heavy debugging" stuff.
IOW: many developers run their work machines with lock debugging and such,
but only few would *constantly* use something which slows kernel down by
a factor of 3. We don't care if inlined function is "too big" only when
this sort of config option is in effect.

So, CONFIG_KASAN is off.

CONFIG_STAGING is also off: we probably don't bother covering
semi-broken drivers.

CONFIG_CC_OPTIMIZE_FOR_SIZE=y (iow: -Os build) is useful because it eliminates
code padding and thus reduces random jitter in code sizes. However, gcc has
a problem where it spuriously deinlines functions marked "inline", and it's
much more pronounced wit -Os:

    https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66122

Therefore, CONFIG_OPTIMIZE_INLINING needs to be disabled with -Os.
It's a good idea anyway: when it's off, build system turns "inline"
into "__always_inline", making deinlining code size change much more robust.
And we do want to find bad inlines - for that, we should prevent gcc
from hiding them!

Nevertheless. CONFIG_OPTIMIZE_INLINING=y build is also useful:
after it,

        nm --size-sort vmlinux \
    | grep -iF ' t ' \
    | uniq -c | grep -v '^ *1 ' | sort -rn

nicely shows duplicate same-sized functions, most of them are bogus
deinlines by gcc.
If you plan to create a patch which forces their inlining (or you work
on fixing gcc!),
this list is useful.

CONFIG_CMDLINE_BOOL should be disabled, otherwise test boots in e.g.
qemu may fail.

[What else to disable?]
[CONFIG_FRAME_POINTER?]


    What do we miss?

Macros. (They are also "inlines" of sorts).
Some macros even define inline functions (!). Example:
    #define DECLARE_EVENT_CLASS(call, proto, args, tstruct, assign, print)  \
    static inline notrace int ftrace_get_offsets_##call(                    \
            struct ftrace_data_offsets_##call *__data_offsets, proto)       \
    {                                                                       \
            int __data_size = 0;                                            \
            int __maybe_unused __item_length;                               \
            struct ftrace_raw_##call __maybe_unused *entry;                 \
                                                                            \
            tstruct;                                                        \
                                                                            \
            return __data_size;                                             \
    }

Bad coding practice of this type:
Fourteen separate *.c files with
#include "echoaudio.c"   // contains static functions, they get duplicated...

--001a11474bac81ebb0051aae1919
Content-Type: application/x-gzip; name="inline_hunting_v1.tar.gz"
Content-Disposition: attachment; filename="inline_hunting_v1.tar.gz"
Content-Transfer-Encoding: base64
X-Attachment-Id: f_ic0jwnka0
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--001a11474bac81ebb0051aae1919--
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/