talos-hostboot/src/usr/hwas, branch master

Improve resource recovery path to handle memory plugging rules

2018-05-15T22:08:42+00:00

Resource recovery is a RAS feature wherein we do not apply certain gard records if those records would result in a boot fail due to a lack of hardware. Change the logic from applying speculative deconfiguration one by one to applying all predictive gard records and then removing those records if minimal hardware is not available. Also add BLOCK_SPEC_DECONFIG attribute to flag when speculative deconfiguration is not allowed. This flag is set if minimum hardware is not available and is cleared after a hardware change. Change-Id: Ia065de3a44ab29fbf33ad4ce98bc42ea5144463f RTC: 191414 CQ: SW424137 Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/58041 Tested-by: Jenkins Server Tested-by: Jenkins OP Build CI Tested-by: Jenkins OP HW Tested-by: FSP CI Jenkins Reviewed-by: Christian R. Geddes Reviewed-by: Daniel M. Crowell

wrap_test: targeting fixups

2018-05-10T03:07:28+00:00

We currently do not support attribute overrides for enum based attributes. Therefore, change the type of ATTR_MFG_WRAP_TEST_ABUS_LINK_SET_ENABLE to uint8_t and add PROC_EFF_FABRIC_GROUP/CHIP_ID attributes to processMrw. Change-Id: I4991e19ab5ceac8e8cf0fbd826b68d7e0fd9ab1d Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/58527 Reviewed-by: Richard Ward Tested-by: Jenkins Server Tested-by: Jenkins OP Build CI Tested-by: FSP CI Jenkins Reviewed-by: Roland Veloz Tested-by: Jenkins OP HW Reviewed-by: Daniel M. Crowell

wrap test: Use MNFG_FLAGS instead of compile time flag

2018-05-07T15:51:15+00:00

Change-Id: I347075dd2424ee8b96805456b3d7d9962f428b64 Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/58347 Tested-by: Jenkins Server Tested-by: Jenkins OP Build CI Tested-by: Jenkins OP HW Tested-by: FSP CI Jenkins Reviewed-by: William G. Hoffa Reviewed-by: Roland Veloz Reviewed-by: Daniel M. Crowell

Fix incorrect parent deconfig in _deconfigByAssocation

2018-05-03T19:53:15+00:00

-The DMI deconfig path was using TYPE_MC as the parent type when it should have been TYPE_MI Change-Id: Iea695f3edbebe651b90ce718f67c95c15df10704 Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/57786 Tested-by: Jenkins Server Tested-by: FSP CI Jenkins Tested-by: Jenkins OP Build CI Tested-by: Jenkins OP HW Reviewed-by: Richard Ward Reviewed-by: Prachi Gupta Reviewed-by: William G. Hoffa

wrap_test: OBUS deconfiguration

2018-04-27T18:54:07+00:00

Due to fabric limitations, we can only have 2 OBUSes enabled at a time. This commit deconfigures the OBUSes that do not belong the set of links to enable in this boot based on attributes. Change-Id: I15f5e9f3ea186367b9c4d55f4dc2e9b32b251f6d Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/57836 Tested-by: Jenkins Server Tested-by: Jenkins OP Build CI Tested-by: FSP CI Jenkins Tested-by: Jenkins OP HW Reviewed-by: Elizabeth K. Liner Reviewed-by: Roland Veloz Reviewed-by: Martin Gloff Reviewed-by: William G. Hoffa

Add deconfigTargetAtRuntime interface

2018-04-12T20:31:07+00:00

-PRD will call this function in response to a core unit checkstop detected by PHYP. The new function will mark the core as non-functional and call a hardware procdure to update the multicast groups for cores and chiplets. The procedure will also update the OCC core configuration status and the quad core status registers which reflect the number of enable cores/quad in the system at runtime. Change-Id: Ia7fc3f47dc1fee1f84d877d3d56b609d1aee1b17 RTC:188918 Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/55154 Tested-by: Jenkins Server Reviewed-by: Richard Ward Reviewed-by: Christian R. Geddes Tested-by: FSP CI Jenkins Tested-by: Jenkins OP Build CI Tested-by: Jenkins OP HW Reviewed-by: William G. Hoffa

Remove logic to deconfigure PHBs in higher position within a pec

2018-04-12T19:29:18+00:00

Current deconfig logic will deconfigure PHBs with a higher unit position under a given pec, if we were deconfiguring the PHB with a lower unit position. For example, PEC1 has PHB1 and PHB2. If we deconfigured PHB1, then we would by association deconfigure PHB2 as well. However, if we deconfigured PHB2, then we would not deconfigure PHB1. This commit removes the logic where if PHB2 is deconfigured, PHB1 is deconfigured as well. To summarize, the end result is to be able to deconfigure PHB1 or PHB2 with no effect on the other. HWPs will handle the hw limitations we have between PHBS. Change-Id: I95b31f4d051c15d896d18adf0a7f7c1b994f5928 CQ: SW417485 Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/56499 Reviewed-by: Martin Gloff Tested-by: Jenkins Server Tested-by: Jenkins OP Build CI Tested-by: Jenkins OP HW Tested-by: FSP CI Jenkins Reviewed-by: William G. Hoffa Reviewed-by: Roland Veloz Reviewed-by: Daniel M. Crowell

Verify deconfig-by-association assumptions still apply to CDIMM scenario.

2018-04-04T14:47:20+00:00

Corrected the deconfig code so that it will only deconfig a MBA if all of its DIMMs have been deconfig'd. Added a testcase to confirm the behavior. Change-Id: Iaf1111073c7229e3d3b72f17b9b905dd244ceb9a RTC: 176408 Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/56481 Tested-by: Jenkins Server Tested-by: Jenkins OP Build CI Tested-by: Jenkins OP HW Tested-by: FSP CI Jenkins Reviewed-by: Richard J. Knight Reviewed-by: Prachi Gupta Reviewed-by: SWATHI M. BHATTIPROLU Reviewed-by: William G. Hoffa

Allow platHwasErrorAddHWCollout now that FSP supports it

2018-03-29T16:10:52+00:00

Previously we tried to add this callout to some hwas code. This caused FSP-CI to fail because the FSP has not implemented support for this platHwasErrorAddHWCollout. Now that the FSP has added support we can allow this callout to happen Change-Id: Ibc4063eecc6607f92e4c1cfe555348d2a0f12785 CQ: SW410022 Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/56398 Tested-by: Jenkins Server Tested-by: Jenkins OP Build CI Tested-by: Jenkins OP HW Tested-by: FSP CI Jenkins Reviewed-by: Matt Derksen Reviewed-by: Martin Gloff Reviewed-by: Daniel M. Crowell

Change deconfig rules to allow for Zeppelin proc config

2018-03-29T14:46:42+00:00

Need to allow for a switch between fabric group and chip ids in the deconfig logic. Currently P1 deconfigs because of step 2 in _symmetryValidation. Change-Id: I54e95847b6078fe981bf6fa9cb9dcebeafa01f94 CQ:SW419526 Reviewed-on: http://ralgit01.raleigh.ibm.com/gerrit1/56223 Reviewed-by: Martin Gloff Tested-by: Jenkins Server Tested-by: Jenkins OP Build CI Tested-by: Jenkins OP HW Reviewed-by: Christian R. Geddes Tested-by: FSP CI Jenkins Reviewed-by: William G. Hoffa