Re: ✗ Fi.CI.BAT: failure for drm/i915: Add missing NULL check before calling initial_watermarks (rev2)

From: Imre Deak <imre.deak@intel.com>
To: intel-gfx@lists.freedesktop.org
Subject: Re: ✗ Fi.CI.BAT: failure for drm/i915: Add missing NULL check before calling initial_watermarks (rev2)
Date: Tue, 01 Mar 2016 12:32:53 +0200	[thread overview]
Message-ID: <1456828373.3325.25.camel@intel.com> (raw)
In-Reply-To: <20160301072449.18664.79576@emeril.freedesktop.org>

On Tue, 2016-03-01 at 07:24 +0000, Patchwork wrote:
> == Series Details ==
> 
> Series: drm/i915: Add missing NULL check before calling
> initial_watermarks (rev2)
> URL   : https://patchwork.freedesktop.org/series/3930/
> State : failure
> 
> == Summary ==
> 
> Series 3930v2 drm/i915: Add missing NULL check before calling
> initial_watermarks
> http://patchwork.freedesktop.org/api/1.0/series/3930/revisions/2/mbox
> /
> 
> Test kms_flip:
>         Subgroup basic-flip-vs-dpms:
>                 pass       -> DMESG-WARN (ilk-hp8440p) UNSTABLE
>         Subgroup basic-flip-vs-wf_vblank:
>                 pass       -> FAIL       (hsw-gt2)

On HSW this patch shouldn't matter, but I haven't seen this problem
earlier, so opened a bug for it:
https://bugs.freedesktop.org/show_bug.cgi?id=94347

>                 dmesg-warn -> PASS       (hsw-brixbox)
>                 pass       -> DMESG-WARN (bdw-ultra)

This is a hw-access-while-suspended failure on the watermark optimize
path, not relevant to this change. I'm guessing it was introduced by
the two stage WM support patch. A fix would be to take a wake reference
around the WM programming, or avoid the programming in cases where
everything is inactive. Adding Matt and Maarten.

>         Subgroup basic-plain-flip:
>                 pass       -> DMESG-WARN (hsw-gt2)

Same as above.

> Test kms_force_connector_basic:
>         Subgroup force-connector-state:
>                 pass       -> SKIP       (snb-x220t)
>         Subgroup force-edid:
>                 pass       -> SKIP       (snb-x220t)
>         Subgroup force-load-detect:
>                 fail       -> DMESG-FAIL (ivb-t430s)
> Test kms_pipe_crc_basic:
>         Subgroup suspend-read-crc-pipe-c:
>                 pass       -> DMESG-WARN (bsw-nuc-2)
> Test pm_rpm:
>         Subgroup basic-pci-d3-state:
>                 dmesg-warn -> PASS       (snb-dellxps)
>                 dmesg-warn -> PASS       (byt-nuc)
>         Subgroup basic-rte:
>                 dmesg-warn -> PASS       (snb-x220t)
>                 pass       -> DMESG-WARN (byt-nuc) UNSTABLE

For all the above failures, the given platforms are not relevant to
this change.

> 
> bdw-
> nuci7        total:166  pass:155  dwarn:0   dfail:0   fail:0   skip:1
> 1 
> bdw-
> ultra        total:169  pass:154  dwarn:1   dfail:0   fail:0   skip:1
> 4 
> bsw-nuc-
> 2        total:169  pass:137  dwarn:1   dfail:0   fail:1   skip:30 
> byt-
> nuc          total:169  pass:143  dwarn:1   dfail:0   fail:0   skip:2
> 5 
> hsw-
> brixbox      total:169  pass:154  dwarn:0   dfail:0   fail:0   skip:1
> 5 
> hsw-
> gt2          total:169  pass:156  dwarn:1   dfail:1   fail:1   skip:1
> 0 
> ilk-
> hp8440p      total:169  pass:117  dwarn:1   dfail:0   fail:1   skip:5
> 0 
> ivb-
> t430s        total:169  pass:153  dwarn:0   dfail:1   fail:0   skip:1
> 5 
> snb-
> dellxps      total:169  pass:144  dwarn:1   dfail:0   fail:1   skip:2
> 3 
> snb-
> x220t        total:169  pass:142  dwarn:1   dfail:0   fail:2   skip:2
> 4 
> 
> Results at /archive/results/CI_IGT_test/Patchwork_1497/
> 
> a1c053436ca45c5860f842e294ad29d6055bd43b drm-intel-nightly: 2016y-
> 02m-29d-16h-58m-26s UTC integration manifest
> ad1be9242b45df74f512fb0fda89c608c337560d drm/i915: Add missing NULL
> check before calling initial_watermarks
> 
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx