[Intel-gfx] ✗ Fi.CI.BAT: failure for Introduce DG1

Wed Sep 30 09:14:23 UTC 2020

On 30/09/2020 09:01, Lucas De Marchi wrote:
> +Matthew Auld
> 
> On Wed, Sep 30, 2020 at 07:44:25AM +0000, Patchwork wrote:
>> == Series Details ==
>>
>> Series: Introduce DG1
>> URL   : https://patchwork.freedesktop.org/series/82241/
>> State : failure
>>
>> == Summary ==
>>
>> CI Bug Log - changes from CI_DRM_9077 -> Patchwork_18595
>> ====================================================
>>
>> Summary
>> -------
>>
>>  **FAILURE**
>>
>>  Serious unknown changes coming with Patchwork_18595 absolutely need 
>> to be
>>  verified manually.
>>
>>  If you think the reported changes have nothing to do with the changes
>>  introduced in Patchwork_18595, please notify your bug team to allow them
>>  to document this new failure mode, which will reduce false positives 
>> in CI.
>>
>>  External URL: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_18595/index.html
>>
>> Possible new issues
>> -------------------
>>
>>  Here are the unknown changes that may have been introduced in 
>> Patchwork_18595:
>>
>> ### IGT changes ###
>>
>> #### Possible regressions ####
>>
>>  * igt at i915_selftest@live at gt_engines:
>>    - fi-skl-lmem:        [PASS][1] -> [DMESG-WARN][2] +31 similar issues
>>   [1]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9077/fi-skl-lmem/igt@i915_selftest@live@gt_engines.html 
>>
>>   [2]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_18595/fi-skl-lmem/igt@i915_selftest@live@gt_engines.html 
>>
> 
> Matt, this is failing on fi-skl-lmem, basically because we are adding
> checks for IS_DGFX() expecting only really dgfx devices would be true.
> However fake lmem code overwrites device info setting it to 1.
> 
> We will need more decisions in the driver checking for dgfx besides
> lmem. Should we a) go ahead and remove fake lmem or b) add the
> checks explictly for fake lmem, like IS_DGFX() || HAS_FAKE_LMEM()
> or c) something else?

Ideally we would keep fake-lmem, at least until we have real LMEM with 
dg1 being tested in CI, otherwise it's more likely we break something 
and it goes by unnoticed in upstream. For example, patches like[1] are 
at least smoke tested through the selftests with the LMEM path enabled, 
which is pretty useful.

I would prefer option b) for now, but if it's too painful I guess we 
just delete it.

[1] https://patchwork.freedesktop.org/patch/392327/?series=82063&rev=2

> 
> thanks
> Lucas De Marchi
> 
>>
>>
>> Known issues
>> ------------
>>
>>  Here are the changes found in Patchwork_18595 that come from known 
>> issues:
>>
>> ### IGT changes ###
>>
>> #### Issues hit ####
>>
>>  * igt at kms_chamelium@common-hpd-after-suspend:
>>    - fi-kbl-7500u:       [PASS][3] -> [DMESG-WARN][4] ([i915#2203])
>>   [3]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9077/fi-kbl-7500u/igt@kms_chamelium@common-hpd-after-suspend.html 
>>
>>   [4]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_18595/fi-kbl-7500u/igt@kms_chamelium@common-hpd-after-suspend.html 
>>
>>
>>
>> #### Possible fixes ####
>>
>>  * igt at i915_module_load@reload:
>>    - fi-byt-j1900:       [DMESG-WARN][5] ([i915#1982]) -> [PASS][6]
>>   [5]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9077/fi-byt-j1900/igt@i915_module_load@reload.html 
>>
>>   [6]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_18595/fi-byt-j1900/igt@i915_module_load@reload.html 
>>
>>
>>  * igt at i915_pm_rpm@basic-pci-d3-state:
>>    - fi-bsw-kefka:       [DMESG-WARN][7] ([i915#1982]) -> [PASS][8]
>>   [7]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9077/fi-bsw-kefka/igt@i915_pm_rpm@basic-pci-d3-state.html 
>>
>>   [8]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_18595/fi-bsw-kefka/igt@i915_pm_rpm@basic-pci-d3-state.html 
>>
>>
>>
>> #### Warnings ####
>>
>>  * igt at kms_force_connector_basic@prune-stale-modes:
>>    - fi-kbl-x1275:       [DMESG-WARN][9] ([i915#62] / [i915#92] / 
>> [i915#95]) -> [DMESG-WARN][10] ([i915#62] / [i915#92]) +8 similar issues
>>   [9]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9077/fi-kbl-x1275/igt@kms_force_connector_basic@prune-stale-modes.html 
>>
>>   [10]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_18595/fi-kbl-x1275/igt@kms_force_connector_basic@prune-stale-modes.html 
>>
>>
>>  * igt at kms_pipe_crc_basic@compare-crc-sanitycheck-pipe-c:
>>    - fi-kbl-x1275:       [DMESG-WARN][11] ([i915#62] / [i915#92]) -> 
>> [DMESG-WARN][12] ([i915#62] / [i915#92] / [i915#95]) +4 similar issues
>>   [11]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9077/fi-kbl-x1275/igt@kms_pipe_crc_basic@compare-crc-sanitycheck-pipe-c.html 
>>
>>   [12]: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_18595/fi-kbl-x1275/igt@kms_pipe_crc_basic@compare-crc-sanitycheck-pipe-c.html 
>>
>>
>>
>>  [i915#1982]: https://gitlab.freedesktop.org/drm/intel/issues/1982
>>  [i915#2203]: https://gitlab.freedesktop.org/drm/intel/issues/2203
>>  [i915#62]: https://gitlab.freedesktop.org/drm/intel/issues/62
>>  [i915#92]: https://gitlab.freedesktop.org/drm/intel/issues/92
>>  [i915#95]: https://gitlab.freedesktop.org/drm/intel/issues/95
>>
>>
>> Participating hosts (45 -> 38)
>> ------------------------------
>>
>>  Missing    (7): fi-ilk-m540 fi-hsw-4200u fi-byt-squawks fi-bsw-cyan 
>> fi-ctg-p8600 fi-byt-clapper fi-bdw-samus
>>
>>
>> Build changes
>> -------------
>>
>>  * Linux: CI_DRM_9077 -> Patchwork_18595
>>
>>  CI-20190529: 20190529
>>  CI_DRM_9077: ae1f3f7de609df105aeceed2655656ffc838d720 @ 
>> git://anongit.freedesktop.org/gfx-ci/linux
>>  IGT_5793: c34792447c9fc4d05dd75873cdb99d9ffe57ea23 @ 
>> git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
>>  Patchwork_18595: b609cfe8808be834400aab6819a1b43b7bc5a5d8 @ 
>> git://anongit.freedesktop.org/gfx-ci/linux
>>
>>
>> == Linux commits ==
>>
>> b609cfe8808b drm/i915/dgfx: define llc and snooping behaviour
>> 648249c6dd8b drm/i915/dg1: Change DMC_DEBUG{1, 2} registers
>> 9034c09bbbdf drm/i915/dg1: DG1 does not support DC6
>> 01007b9c648c drm/i915/dg1: Add initial DG1 workarounds
>> f5c37ea500ef drm/i915/dg1: Load DMC
>> 571405def6c9 drm/i915/dg1: enable PORT C/D aka D/E
>> 3056c7ce001a drm/i915/dg1: map/unmap pll clocks
>> d0727f32e9ec drm/i915/dg1: provide port/phy mapping for vbt
>> dc3f763d7b05 drm/i915/dg1: Update voltage swing tables for DP
>> e764c35d3b9e drm/i915/dg1: Update comp master/slave relationships for 
>> PHYs
>> a96cb47a46d4 drm/i915/dg1: Don't program PHY_MISC for PHY-C and PHY-D
>> b5634537768e drm/i915/dg1: Enable first 2 ports for DG1
>> ed7812a70681 drm/i915/dg1: gmbus pin mapping
>> 1556d9b840f8 drm/i915/dg1: invert HPD pins
>> 4caa7e8d99ff drm/i915/dg1: add hpd interrupt handling
>> fc9e91e510d5 drm/i915/dg1: Enable DPLL for DG1
>> 802098453f92 drm/i915/dg1: Add and setup DPLLs for DG1
>> 46ba3b109500 drm/i915/dg1: Add DPLL macros for DG1
>> 313e894b5c43 drm/i915/dg1: Wait for pcode/uncore handshake at startup
>> af7964472d1d drm/i915/dg1: Increase mmio size to 4MB
>> 44b1fb34d9e6 drm/i915/dg1: Add DG1 power wells
>> 90eb98d91e8e drm/i915/dg1: Define MOCS table for DG1
>> 02a93e492718 drm/i915/dg1: Initialize RAWCLK properly
>> 2ab727864817 drm/i915/dg1: add more PCI ids
>>
>> == Logs ==
>>
>> For more details see: 
>> https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_18595/index.html