linux-kernel - DRM range manager selftest run time

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAMuHMdU9JXh34tGrpuUgQsMfGRNsZtEFvJ=9KtWT1MWosumW+w@mail.gmail.com>
Date:   Thu, 9 Feb 2017 10:58:38 +0100
From:   Geert Uytterhoeven <geert@...ux-m68k.org>
To:     Chris Wilson <chris@...is-wilson.co.uk>,
        David Airlie <airlied@...ux.ie>
Cc:     DRI Development <dri-devel@...ts.freedesktop.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: DRM range manager selftest run time

Hi Chris, Dave,

Recently, I started seeing ca. 2 minute delays during boot-up on Renesas
R-Car Gen2 boards.

It turned out the DRM range manager selftests started taking quite some time.
With debug output, I get:

[    2.310472] drm_mm: Testing DRM range manger (struct drm_mm), with
random_seed=0x637f8ba8 max_iterations=8192 max_prime=128
[    2.321625] drm_mm: drm: Running sanitycheck
[    2.325905] drm_mm: igt_sanitycheck - ok!
[    2.329915] drm_mm: drm: Running init
[    2.333590] drm_mm: drm: Running debug
[    2.337341] drm_mm: drm: Running reserve
[    2.862878] drm_mm: drm: Running insert
[   49.205174] drm_mm: drm: Running replace
[   95.570840] drm_mm: drm: Running insert_range
[   98.971590] drm_mm: drm: Running align
[   98.975614] drm_mm: drm: Running align32
[   98.979601] drm_mm: drm: Running align64
[   98.983652] drm_mm: drm: Running evict
[  100.321925] drm_mm: drm: Running evict_range
[  101.226411] drm_mm: drm: Running bottomup
[  101.316911] drm_mm: drm: Running topdown
[  101.404385] drm_mm: drm: Running color
[  103.889639] drm_mm: drm: Running color_evict
[  107.318608] drm_mm: drm: Running color_evict_range
[  110.485926] drm_mm: run_selftests: Completed

Perhaps some tests can be simplified, to reduce run time?

Alternatviely, the pr_debug() calls in drivers/gpu/drm/selftests/drm_selftest.c
could be turned into pr_info() calls (cfr. the output above), to give a cue?

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@...ux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds