Patch-ID# 111335-23 Keywords: dr copy-rename cfgadm dr drmach sunw,sun-fire-15000 hw4 Synopsis: SunOS 5.8: Sun-Fire-15000 dr, gptwo_pci & sc_gptwocfg drivers patch Date: Aug/30/2004 Install Requirements: Install in Single User Mode Reboot immediately after patch is installed Solaris Release: 8 SunOS Release: 5.8 Unbundled Product: Unbundled Release: Xref: Topic: SunOS 5.8: Sun-Fire-15000 dr, gptwo_pci & sc_gptwocfg drivers patch *********************************************************** NOTE: This patch may contain one or more OEM-specific platform ports. See the appropriate OEM_NOTES file within the patch for information specific to these platforms. DO NOT INSTALL this patch on an OEM system if a corresponding OEM_NOTES file is not present (or is present, but instructs not to install the patch), unless the OEM vendor directs otherwise. *********************************************************** NOTE: Refer to Special Install Instructions section for IMPORTANT specific information on this patch. Relevant Architectures: sparc.sun4u BugId's fixed with this patch: 4346494 4379810 4385107 4397382 4399138 4399165 4399184 4400692 4404492 4404532 4406374 4406573 4406903 4407043 4407520 4409977 4410012 4412070 4414010 4414500 4414580 4416528 4416546 4416565 4416645 4417163 4418267 4418315 4418721 4419158 4419171 4421238 4422498 4422946 4423982 4434338 4446470 4451073 4451120 4451354 4451879 4451940 4453074 4453614 4455850 4456609 4458013 4458210 4459820 4460715 4461330 4461811 4464336 4464593 4465918 4466463 4467405 4467793 4468133 4468450 4469104 4469251 4470271 4470641 4475481 4476158 4477894 4478178 4478393 4481262 4483884 4485627 4487620 4489166 4489890 4491299 4495650 4495712 4496385 4500100 4503287 4506806 4508895 4509391 4509968 4510008 4512458 4513778 4514400 4519712 4520227 4521645 4522151 4522581 4523099 4523109 4523979 4525053 4525286 4526043 4527408 4532551 4609938 4613988 4614002 4616340 4620790 4622938 4630898 4642161 4646811 4652995 4658527 4661605 4664749 4669462 4670106 4673645 4675057 4676706 4680826 4682022 4682824 4692462 4692542 4696700 4700564 4707287 4712228 4728572 4746505 4749709 4756790 4759384 4769147 4806726 4810029 4828704 4831702 4841716 4848215 4856452 4858710 4860955 4893666 4901725 4921936 4927078 4929059 4964679 5018945 Changes incorporated in this version: 4831702 4921936 4929059 5018945 Patches accumulated and obsoleted by this patch: 111395-02 116206-01 Patches which conflict with this patch: Patches required with this patch: 108528-28 or greater 108528-29 or greater 110826-09 or greater 110836-05 or greater 110837-04 or greater 117000-05 or greater 117350-03 or greater Obsoleted by: Files included with this patch: /platform/SUNW,Sun-Fire-15000/kernel/drv/sparcv9/dr /platform/SUNW,Sun-Fire-15000/kernel/misc/sparcv9/drmach /platform/SUNW,Sun-Fire-15000/kernel/misc/sparcv9/fcgp2 /platform/SUNW,Sun-Fire-15000/kernel/misc/sparcv9/gptwo_pci /platform/SUNW,Sun-Fire-15000/kernel/misc/sparcv9/gptwo_wci /platform/SUNW,Sun-Fire-15000/kernel/misc/sparcv9/sc_gptwocfg Problem Description: 4831702 dr copy-rename causes dstop if target board reprogrammed with +ve slice offset 4929059 panic in kphysm_add_memory_dynamic routine during cfgadm memory configure 4921936 unconfigure dynamic APs without golden IOSRAM shouldn't have failed tunnel swith 5018945 I/O cage CPU selection should avoid known Jaguar XIR problem (from 111335-22) 4410012 cfgadm reports source memory as unconfigured during copy-rename 4860955 page removal causes a good page to be removed after DR operation 4893666 page_retire does not update retired page list in some cases (from 111335-21) 4806726 sequence of copy-rename/reboots events caused OS to hang during quiesce stage 4964679 Rated proc speed used instead of actual with DR operation in Starcat (from 111335-20) 4810029 Solaris needs to support UltraSPARC-IV 4453074 To configure cpu 0 or 1 its partner needs to be unconfigured 4464593 debug assertion panic while detaching CPUs 4483884 CPU "powered off" after DR, can't be powered back on via psradm 4658527 Need to add error injection model for testing sbdp driver 4673645 Incorrect use of REMAP_CHEETAH_MASK for cheetah+ processors 4841716 PSARC/2003/196 psrinfo CMP support 4901725 cfgadm / prtdiag mismatch 4828704 the dispatcher should be made CMP aware 4856452 Dispatcher causes performance degradation & variability on US-IV systems 4927078 Eliminate dependency between drmach and platmod (from 111335-19) 4712228 AXQ requests Slot0 Dstop; Home board issues Unmapped on Remote Writeback 4858710 drmach_slice_table_lock not released in error flow (from 111335-18) 4848215 schpc and sc_gptwocfg drivers need to support RSV_EMPTY_CASSETTE (from 111335-17) 4416645 Error returned for parallel cfgadm commands (from 111335-16) 4495650 Segmentation Fault in eFCode interpreter when DR connecting IO board 4523099 mailbox timeouts need to be revisited 4613988 race condition between dr status and connect causes assert failure 4616340 DN_BUSY_CHANGING domain panic at IO DR 4630898 assertion failed: slot1_paused[axq_exp] == NULL 4646811 memory leak in drmach_board_find_devices_cb 4661605 failed assert in plat_assign_lgrphand_to_mem_node 4669462 slot1 dr needs to be enabled by default 4670106 hsPCI+ needs to support DR 4675057 cfgadm shows incorrect occupant state, speed/ecache for unrecovered IO test cpus 4676706 unconfigure IO starts failing while running SUNWsgdr test suite. 4680826 cfgadm -c configure operation fails on wPCI board 4682022 IO Configure failed with Invalid state transition error. 4682824 IKP does not support DR connect of wPCI boards 4692462 DR support for mixed LPA mode domains 4700564 wPCI slot with empty cassette usually shows incorrect state after DR operation 4707287 Panic: DN_BUSY_CHANGING at scsi_hba 4728572 hsPCI+ slot 1 (5V) is missing under cfgadm after DR operation. 4746505 Slot 1 receptacle state is NOT marked empty after hot plug out of IO board 4749709 cpus unrecovered with test IO4: Mailbox framework failure: outgoing 4756790 Boards marked unusable on mailbox reinitialization failures during disconnect 4759384 panic: in drmach_board_status() while unconfiguring boards 4769147 CDC indicates an owner outside the domain Dstop during DR operations (from 111335-15) 4692542 panic: in fcode:fc_run_priv() with FC_RUN_PRIV: Priv violation (from 111335-14) 4696700 DR copy-rename deadlocks modloading klmmod (from 111335-13) 4525286 unable to move SB using DR 3.0 4664749 E15K copy-rename blocks with cpus paused (from 111335-12) 4622938 race condition between drmach_status and drmach_unassign 4652995 pcisch streaming cache flush TO panic during copy-rename suspend-resume (from 111335-11) 4508895 IO board DR leads to CPQ_TO Dstop 4510008 Enable SF15K DR by default 4513778 GDCD/LDCD updates not made on copy-rename events 4520227 use sms allocated iocage and remove platmod iocage allocation 4521645 sdrfunc_04.pl panics: dr_with_proc: alignment error: deadbeefdeadbeef 4522151 cfgadm status not killable during DR disconnect 4522581 incorrect dip handed to man_dr_attach() 4523109 panic in module "axq" due to a NULL pointer dereference 4523979 domain panic:ptl1 trap reason 0x9 while configuring CPU board of MCPU 4526043 need to identify source of slot1 transactions during copy-rename 4527408 misleading errors reported during DR suspend error flow 4532551 incorrect bit mask for schizo safari unmapped error 4609938 drmach mailbox versioning code needs cleanup 4614002 incorrect decoding of Schizo internal interrupts 4620790 system hangs on parallel dr on i_ndi_block_device_tree_change() (from 111335-10) 4642161 struct cpu_node needs to maintain same size and offsets (from 111335-09) 4525053 moveboard hang - cannot cancel (from 111335-08) 4509968 REDSTATE on copy-rename if slot1 has MCPU boards and slot0 memory is unconfigure 4512458 Panic disconnecting cpu/mem board when MAXCAT present 4514400 memory unconfigure on E15K does not offline all mc nodes 4519712 DR mismanages cached AXQ dip (from 111335-07) 4470271 E15K DR incorrectly reports cfgadm condition of connected cpus 4476158 Safari bus error while configuring an IO board 4478178 panic: cpu 418 failed to start during moveboard command 4481262 Add I$ and P$ flush for all procs, add D$ flush for master proc 4485627 NULL pointer dereference panic after DR mailbox corruption 4487620 E15K DR must verify receipt of mailbox messages 4489166 panic: unowned mutex 4491299 panic: assertion failed while connecting IO board 4495712 panic: Timeout Event on CPUxxx Privileged Data Access 4500100 need enabling mechanism for slot 1 DR on E15K 4503287 CPU unconfigure operation can cause dstop 4506806 Dstop while disconnecting System board 4509391 dsmd dump; system crash during aloha dr test (from 111335-06) 4496385 "panic: ptl1 trap reason 0x2" while resuming drivers (from 111335-05) 4475481 cannot hotplug adapter board into empty pci slot (from 111335-04) 4489890 need enabling mechanism for SUNW,Sun-Fire-15000 DR (from 111335-03) 4346494 unable to connect new io boards 4379810 invalid cfgadm -o platform options not rejected 4416565 "cfgadm -av" does not report proper "Busy" and "Condition" status for sb 4422498 sgenv: minimize amount of work done in interrupt context 4434338 sghsc module fails to attach for devices on Wildcat (SSM) systems 4451354 Unable to force unconfigure permanent memory due to RT process 4451879 SUNW,Sun-Fire DR events should use sysevent defined hints 4453614 sghsc module causes kernel panic when iowib is present 4458013 Error case : panic when unconfigure permanent memory board and there is sync err 4458210 ERROR case : cannot disconnect the unsuccessfully DR configured board 4459820 panic unconfiguring I/O board 4461330 CPU SRAM shutdown area not found after firmware change 4466463 CPUs go offline if DR unconfigure is rejected 4467793 BAD TRAP panic when swapping 2-way and 4-way boards then dr connect 4468133 openeepr: WARNING: nodeid not found 4468450 DR: sghsc: unable to stat slot 4470641 sbd can pass garbage in the handle to sbdp causing panic 4478393 memory condition is set to failed if a CPU fails 4477894 DR: panic running SUNWsgdr suite 4399184 copy-rename does not support target with smaller memory size 4414580 drmach_io_pre_release and drmach_io_post_attach should detail errors 4451940 copy-rename leaves source attachment point marked "Busy" after it completes 4455850 cfgadm can hang with offlined cpus 4464336 AXQ h/w restrictions require iocage re-alignment 4465918 Domain hangs when copy-rename master cpu is on the source board 4467405 DR walks device tree without locks 4469104 CDC must be flushed during disconnect, regardless of board type 4469251 SUNW,Sun-Fire-15000 DR miscalculates MCPU cpuids 4446470 Changes to wci device tree per FWARC/2000/552 4451073 The "bank-status" property incorrectly formed by IKP (from 111335-02) This patch revision was generated to accumulate and obsolete the feature changes introduced in feature point patch 111395-02. (from 111335-01) 4385107 "Cannot stop user thread" error labels permanent memory as unconfigured 4399138 cfgadm incorrectly reports permanent pages on board being configured in 4399165 Incorrect output from cfgadm when configuring a board already configured. 4400692 get "Unexpected internal condition: drmach.c 2433" unconfiguring CPU/Mem board 4404492 Panic from copy-rename using deleteboard 4404532 DR should not mark board as disconnected if SUNW,Sun-Fire-15000 fails over during operation. 4406374 Dstop disconnecting a cpu/memory board after failing to set PCD CASM field 4406903 DR driver does not support operation with AXQ cdc enabled 4407520 "bad mutex" panic during random DR test 4414010 DR mailbox fails under multithreaded load 4397382 domain panic's when unconfiguring I/O board 4409977 cfgadm reports source memory of zero KBytes at start of copy-rename 4407043 DR code walks thread list without protection (E15K) 4412070 Need to reconfigure target memory after failed suspend 4416528 Concurrent DR initialization causes DR failure 4416546 DR mailbox times out excessively 4418267 cfgadm status command causes panic during DR disconnect 4418315 cfgadm can report stopped threads as unstoppable following failed suspend 4419158 Panic in drmach_dip2pathname after beginning to configure fibre-channel nodes 4419171 iosram failures due to changes in post leaf structure prevents moving of golden i 4421238 cfgadm misreports memory sizes following copy-rename 4422946 domain panics starting cpu following cpu/mem board connect 4423982 copy-rename panics with smaller target memory board (from 111395-02) 4406573 SUNW,Sun-Fire-15000 DR code grabs adaptive lock and resets spl during OS quiesce 4417163 SUNW,Sun-Fire-15000 DR code leaves CPU affinity set in error case 4418721 Command pool timeout Dstop while running DR tests 4456609 copy-rename can hang with mixed DIMMs 4460715 Panic following cfgadm -t of i/o boards 4461811 Command pool timeout dstops following DR copy-rename operations 4451120 Add support to Solaris for the UltraSPARC-III+ CPU (from 111395-01) 4414500 Provide Next Gen DR feature to SUNW,Ultra-Enterprise-10000 (from 116206-01) 4828704 the dispatcher should be made CMP aware 4856452 Dispatcher causes performance degradation & variability on US-IV systems 4927078 Eliminate dependency between drmach and platmod Patch Installation Instructions: -------------------------------- For Solaris 2.0-2.6 releases, refer to the Install.info file and/or the README within the patch for instructions on using the generic 'installpatch' and 'backoutpatch' scripts provided with each patch. For Solaris 7-9 releases, refer to the man pages for instructions on using 'patchadd' and 'patchrm' scripts provided with Solaris. Any other special or non-generic installation instructions should be described below as special instructions. The following example installs a patch to a standalone machine: example# patchadd /var/spool/patch/104945-02 The following example removes a patch from a standalone system: example# patchrm 104945-02 For additional examples please see the appropriate man pages. Special Install Instructions: ----------------------------- Reboot the system after patch installation. NOTE 1: To get the complete Next Generation Dynamic Reconfiguration feature for Sun Fire 15K, one needs to install the following packages introduced in the Solaris 8 Update 4 (4/01) Release (if they are not already installed): SUNWdcsr (Domain Configuration Server, (Root)) SUNWdcsu (Domain Configuration Server) and also the following patch: 111332-04 (or newer) Domain Configuration Server patch NOTE 2: To get the UltraSPARC-III+ support, one needs to also install the following patches: 111794-01 (or newer) /usr/lib/libcpc.so.1 patch 109873-09 (or newer) prtdiag and platform libprtdiag_psr.so.1 patch 110849-06 (or newer) PICL support for SUNW,Sun-Fire-880 patch 110835-03 (or newer) gptwo_cpu driver patch 109888-11 (or newer) platform drivers patch 110834-03 (or newer) SUNW,Sun-Fire-15000 libprtdiag_psr.so.1 patch NOTE 3: To get the complete fix of bug 4470271 (E15K DR incorrectly reports cfgadm condition of connected cpus), one needs to also install patch 111790-05 (or newer). NOTE 4: To get the complete fix for Phase II SF15K DR implementation, please also install the following patch: 110836-05 (or newer) gptwocfg patch NOTE 5: To get the complete fix for 4410012, one needs to also install the following patch: 111790-20 (or newer) ngdr and ngdrmach drivers patch for Ultra-Enterprise-10000 patch NOTE 6: To get the complete fix for 4860955 and 4893666, one needs to also install the following patches: 111789-06 (or newer) /platform/SUNW,Ultra-Enterprise-10000/ kernel/drv/dr patch 111790-20 (or newer) ngdr and ngdrmach drivers patch for Ultra-Enterprise-10000 patch 116970-01 (or newer) sbdp patch README -- Last modified date: Monday, August 30, 2004