It also impacts SNB platform.

commit 25cf0551c7d210c8c085c109891dc97a2cc61e27
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Tue Jan 6 09:59:47 2015 +0000

    igt/gem_ctx_thrash: Tweak resource limits
    
    On some systems (ok, most systems!) we may need to enlarge the allowed
    number of open files in order to create enough fd to fill the aperture.
    
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=87572
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

(In reply to Chris Wilson from comment #2)
> commit 25cf0551c7d210c8c085c109891dc97a2cc61e27
> Author: Chris Wilson <chris@chris-wilson.co.uk>
> Date:   Tue Jan 6 09:59:47 2015 +0000
> 
>     igt/gem_ctx_thrash: Tweak resource limits
>     
>     On some systems (ok, most systems!) we may need to enlarge the allowed
>     number of open files in order to create enough fd to fill the aperture.
>     
>     Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=87572
>     Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

This patch causes the case takes more than 10 minutes.

output:
IGT-Version: 1.9-g25cf055 (x86_64) (Linux: 3.19.0-rc3_drm-intel-nightly_296cb8_20150108+ x86_64)
Creating 49152 contexts (assuming of size 65536)
Creating 98304 surfaces (of size 45056)
Subtest processes: SUCCESS (808.140s)

real    13m28.163s
user    0m13.052s
sys     60m36.736s

dmesg:
[   52.063151] [drm:i915_gem_open]
[   52.063154] [drm:i915_gem_open]
[   52.063158] [drm:i915_gem_open]
[   52.063162] [drm:i915_gem_open]
[   52.063166] [drm:i915_gem_open]
[   52.063174] [drm:i915_gem_open]
[   52.063180] [drm:i915_gem_open]
[   52.063184] [drm:i915_gem_open]
[   52.063189] [drm:i915_gem_open]
[   52.063340] systemd-journald[2637]: /dev/kmsg buffer overrun, some messages lost.
[   52.068335] gem_ctx_thrash (4193): drop_caches: 3
[  860.692952] gem_ctx_thrash: exiting, ret=0
[  860.693458] [drm:intel_crtc_set_config] [CRTC:19] [FB:58] #connectors=1 (x y) (0 0)
[  860.693463] [drm:intel_set_config_compute_mode_changes] computed changes for [CRTC:19], mode_changed=0, fb_changed=0
[  860.693465] [drm:intel_modeset_stage_output_state] [CONNECTOR:29:VGA-1] to [CRTC:19]
[  860.693468] [drm:intel_modeset_affected_pipes] set mode pipe masks: modeset: 1, prepare: 1, disable: 0
[  860.693471] [drm:connected_sink_compute_bpp] [CONNECTOR:29:VGA-1] checking for sink bpp constrains
[  860.693474] [drm:ironlake_check_fdi_lanes] checking fdi config on pipe A, lanes 2
[  860.693476] [drm:intel_modeset_pipe_config] plane bpp: 24, pipe bpp: 24, dithering: 0
[  860.693478] [drm:intel_dump_pipe_config] [CRTC:19][modeset] config for pipe A
[  860.693479] [drm:intel_dump_pipe_config] cpu_transcoder: A
[  860.693481] [drm:intel_dump_pipe_config] pipe bpp: 24, dithering: 0
[  860.693483] [drm:intel_dump_pipe_config] fdi/pch: 1, lanes: 2, gmch_m: 6815744, gmch_n: 8388608, link_m: 283989, link_n: 524288, tu: 64
[  860.693485] [drm:intel_dump_pipe_config] dp: 0, gmch_m: 0, gmch_n: 0, link_m: 0, link_n: 0, tu: 0
[  860.693487] [drm:intel_dump_pipe_config] dp: 0, gmch_m2: 0, gmch_n2: 0, link_m2: 0, link_n2: 0, tu2: 0
[  860.693488] [drm:intel_dump_pipe_config] audio: 0, infoframes: 0
[  860.693489] [drm:intel_dump_pipe_config] requested mode:
[  860.693493] [drm:drm_mode_debug_printmodeline] Modeline 0:"1680x1050" 60 146250 1680 1784 1960 2240 1050 1053 1059 1089 0x48 0x6
[  860.693494] [drm:intel_dump_pipe_config] adjusted mode:
[  860.693496] [drm:drm_mode_debug_printmodeline] Modeline 0:"1680x1050" 60 146250 1680 1784 1960 2240 1050 1053 1059 1089 0x48 0x6
[  860.693499] [drm:intel_dump_crtc_timings] crtc timings: 146250 1680 1784 1960 2240 1050 1053 1059 1089, type: 0x48 flags: 0x6
[  860.693500] [drm:intel_dump_pipe_config] port clock: 270000
[  860.693502] [drm:intel_dump_pipe_config] pipe src size: 1680x1050
[  860.693503] [drm:intel_dump_pipe_config] gmch pfit: control: 0x00000000, ratios: 0x00000000, lvds border: 0x00000000
[  860.693505] [drm:intel_dump_pipe_config] pch pfit: pos: 0x00000000, size: 0x00000000, disabled
[  860.693507] [drm:intel_dump_pipe_config] ips: 0
[  860.693508] [drm:intel_dump_pipe_config] double wide: 0
[  860.693510] [drm:intel_crtc_set_config] [CRTC:23] [NOFB]
[  860.693512] [drm:intel_set_config_compute_mode_changes] computed changes for [CRTC:23], mode_changed=0, fb_changed=0
[  860.693514] [drm:intel_modeset_stage_output_state] [CONNECTOR:29:VGA-1] to [CRTC:19]
[  860.693516] [drm:intel_modeset_affected_pipes] set mode pipe masks: modeset: 0, prepare: 0, disable: 0
[  860.693517] [drm:intel_crtc_set_config] [CRTC:27] [NOFB]
[  860.693519] [drm:intel_set_config_compute_mode_changes] computed changes for [CRTC:27], mode_changed=0, fb_changed=0
[  860.693521] [drm:intel_modeset_stage_output_state] [CONNECTOR:29:VGA-1] to [CRTC:19]
[  860.693523] [drm:intel_modeset_affected_pipes] set mode pipe masks: modeset: 0, prepare: 0, disable: 0

(In reply to lu hua from comment #3)
> (In reply to Chris Wilson from comment #2)
> > commit 25cf0551c7d210c8c085c109891dc97a2cc61e27
> > Author: Chris Wilson <chris@chris-wilson.co.uk>
> > Date:   Tue Jan 6 09:59:47 2015 +0000
> > 
> >     igt/gem_ctx_thrash: Tweak resource limits
> >     
> >     On some systems (ok, most systems!) we may need to enlarge the allowed
> >     number of open files in order to create enough fd to fill the aperture.
> >     
> >     Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=87572
> >     Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> 
> This patch causes the case takes more than 10 minutes.

Closing again. Bug 87575 tracks the runtime issue.

Verified.

Closing old verified+fixed.