History log of /opensolaris-onvv-gate/usr/src/uts/sfmmu/vm/hat_sfmmu.c
Revision Date Author Comments
# 12908:80a39220b451 26-Jul-2010 Pavel Tatashin <Pavel.Tatashin@Sun.COM>

6948098 freemem_lock contention causes poor concurrent DISM MC_LOCK/MC_UNLOCK performance
6946742 sfmmu_mlspl_enter lock contention on mml_table
6956454 ani_free_pool lock contention with multi processed ISM


# 12230:12df01648ca5 23-Apr-2010 Peter Rival <Frank.Rival@oracle.com>

6778289 vm locks need to scale with the size of system (strands/memory size)


# 11713:03615b084875 19-Feb-2010 Pavel Tatashin <Pavel.Tatashin@Sun.COM>

6892591 per-MMU context id domains for sun4v


# 11178:d671c0911ce4 23-Nov-2009 Dave Plauger <Dave.Plauger@Sun.COM>

6898505 kmem_verify oddities around Fast Crash Dump integration


# 10843:eb166ea6e74e 22-Oct-2009 Dave Plauger <Dave.Plauger@Sun.COM>

6828976 Fast Crash Dump
6878030 live crash dump is much slower than reboot dump
6626023 Crash dump size is excessive on large memory machines


# 10701:dabf35f459b8 30-Sep-2009 Feri Chua <Feri.Chua@Sun.COM>

6778854 System panics in unix:sfmmu_ismtlbcache_demap


# 10271:7c80b70bb8de 06-Aug-2009 Jason Beloro <Jason.Beloro@Sun.COM>

6858457 Remove Solaris support for UltraSPARC-AT10 processor


# 9281:67a3435ea50e 03-Apr-2009 Prakash Sangappa <Prakash.Sangappa@Sun.COM>

6811473 VPM interfaces should provide multiple pagelength mappings per request


# 8819:4df632c9b303 13-Feb-2009 Jason Beloro <Jason.Beloro@Sun.COM>

FWARC 2008/592 MMU_EXT API group name and number change
6729256 kernel routines such as bcopy, bzero need to be optimized for UltraSPARC-AT10.
6757032 Use TLB search order register API
6771318 Some packets get dropped from the performance counter ringbuffer
6781121 CPC: Rock PCBE does not correctly handle overflow profiling. 64-bit extension is incorrect.
6791429 SUNW,UltraSPARC-AT10 should not be using ASI_NQUAD_LD since it is redefined as ASI_STICK_CMPR


# 8574:34b33d2c8168 19-Jan-2009 Jason Beloro <Jason.Beloro@Sun.COM>

6755072 Frequency of calls to demap all could be reduced
6757028 Improve CRP error packet handling
6791608 SW support for humidity sensors
6793427 Optimize update to mmu primary context id 1


# 8187:07a2f708e698 21-Nov-2008 Amritpal Sandhu <Paul.Sandhu@Sun.COM>

6638112 eliminate HMELOCK_ENTER() to avoid xcall timeout panic and improve GET_TTE() performance
6620705 sfmmu_hblk_hash_add/rm should not access hme buckets via va
6638127 TSB_UPDATE_TL()/TSB_UPDATE_TL_PN() shouldn't spin forever in TSB_LOCK_ENTRY()


# 7869:10227e8ccea5 17-Oct-2008 Sean McEnroe <Sean.McEnroe@Sun.COM>

6729139 Shared context causes SEGVs and panics on OPL


# 7718:555eee33ff6e 26-Sep-2008 Jason Beloro <Jason.Beloro@Sun.COM>

PSARC/2007/003 Solaris Support for the Rock Processor
6368478 Solaris needs to support the Rock processor
6440653 stores to sun4v error queue head registers are missing a membar #Sync
6639717 FLUSH instruction used more than necessary in sun4v


# 7393:8a8d73648b44 22-Aug-2008 Donghai Qiao <Donghai.Qiao@Sun.COM>

6672470 Looping within uiomove()/xcopyout()/copyout_move()
6707987 kpm has some VAC handling issues
6388567 VAC flushing is broken on US III-IV+ for large pages


# 7240:c4957ab6a78e 31-Jul-2008 rh87107

6698108 Sun Studio 12 finds many lint errors in ON source.


# 6695:12d7dd4459fd 22-May-2008 aguzovsk

6423097 segvn_pagelock() may perform very poorly
6526804 DR delete_memory_thread, AIO, and segvn deadlock
6557794 segspt_dismpagelock() and segspt_shmadvise(MADV_FREE) may deadlock
6557813 seg_ppurge_seg() shouldn't flush all unrelated ISM/DISM segments
6557891 softlocks/pagelocks of anon pages should not decrement availrmem for memory swapped pages
6559612 multiple softlocks on a DISM segment should decrement availrmem just once
6562291 page_mem_avail() is stuck due to availrmem overaccounting and lack of seg_preap() calls
6596555 locked anonymous pages should not have assigned disk swap slots
6639424 hat_sfmmu.c:hat_pagesync() doesn't handle well HAT_SYNC_STOPON_REF and HAT_SYNC_STOPON_MOD flags
6639425 optimize checkpage() optimizations
6662927 page_llock contention during I/O


# 6665:1515bc919a93 20-May-2008 paulsan

6701890 panic: hat_unlock_region: addr not found panic


# 6127:3731a676ec5e 29-Feb-2008 sm142603

6642758 Support Shared Context for SPARC64-VII


# 6044:56828186ab76 20-Feb-2008 paulsan

6656203 Stress shm24 test - panic[cpu2]/thread=300daf0e020: ptl1_panic reason 0x15 at TL 1


# 5110:6c913b66a74a 21-Sep-2007 paulsan

6606499 Fix cstyle error in hat_sfmmu.c


# 5075:199eb2ec2c2c 17-Sep-2007 paulsan

6597746 Don't create srd_cache, region_cache and scd_cache on systems that don't use shared context
6597751 SFMMU_HASH_UNLOCK() call is missing in hat_unlock_region()
6600745 shared context code needs to be cleaned up
6603269 assertion failed: sfhme->hme_tte.ll != 0, file: ../../sfmmu/vm/hat_sfmmu.c, line: 11613


# 4528:9ad45715d2ab 24-Jun-2007 paulsan

PSARC 2006/266 Shared Context Support
PSARC 2006/267 Shared Region HME Block support
6388600 do_virtual_coloring checks should be expunged from sfmmu
6449192 Integrate support for MMU Shared Contexts
6449195 Integrate support for Shared Region HME Blocks


# 4421:8ddf0d909338 08-Jun-2007 mb158278

6491350 Auto mpss and kpr combo can make a system crawl in case of fragmented memory.


# 4324:696de332c6cf 25-May-2007 qiao

6538758 Sporadic I/O error will cause data corruption in Veritas CFS with VMODSORT mode on.


# 4266:c151d8b35988 17-May-2007 dp78419

6547155 mptest timing out because of bad page coloring after fork
6550904 kcage_glist_alloc() may not call kmem_zalloc()


# 4236:de290ec1e532 14-May-2007 dm120769

backout 6547155/6550904: sparc build failed


# 4230:9ab9040697c4 14-May-2007 dp78419

6547155 mptest timing out because of bad page coloring after fork
6550904 kcage_glist_alloc() may not call kmem_zalloc()


# 3896:1d5fb739bf40 26-Mar-2007 vb70745

6509040 hang in sfmmu_hblk_steal()


# 3764:74844940a161 05-Mar-2007 dp78419

6510878 Kernel Boot time data structures can be mapped with Large pages to improve performance


# 3687:96d6b6ad3918 20-Feb-2007 jb145095

6244343 sun4v mmu code should support 16-bit context and correctly mask mmu fault addresses


# 3351:853fb8df244a 02-Jan-2007 aguzovsk

6498368 Reduce va_to_pa overheads
6500938 tte_remap panics with inconsistent sf_hments


# 3290:256464cbb73c 20-Dec-2006 johansen

4894692 caching data in heap inflates crash dump
6499454 time to increase size of kmem default allocation caches
6499459 vm should stop checking kvp directly


# 3258:077f19d9741d 15-Dec-2006 trevtom

6351793 sfmmu_mlist_enter recursive mutex_enter.
6467201 Duplicate hrmstat structures cause long hat_statlock hold times.


# 3000:f09e516c394d 27-Oct-2006 stans

6393251 data corruption caused by VMODSORT problem (missing changes)


# 2991:4b13d6c49c6b 26-Oct-2006 susans

6254029 memcntl() MC_HAT_ADVISE with page size 0 may cause segment page sizes to be demoted
6325885 map_pgszstk() uses p->p_brkpageszc rather than p->p_stkpageszc
6371967 assign large pages to anon segment created using mmap /dev/zero
6483208 unify and cleanup OOB (out of the box) large pagesize selection code
6483216 use intermediate pagesizes to map the beginning of bss/heap and stack when it may help performance
6483226 bss size is not properly taken into account by LP OOB policy at exec() time
6483230 grow_internal() doesn't properly align stack bottom for large pages
6483231 memcntl.c: ASSERT(IS_P2ALIGNED(p->p_brkbase + p->p_brksize, pgsz));
6483233 provide a mechanism to enable the use of 32M text pages on OPL by default
6485171 memcntl() shouldn't silently fail when stack space is unavailable with requested pagesize


# 2874:ef6ddc19af4e 06-Oct-2006 vb70745

6457396 missing 2nd sfmmu_info causes failure of hypervisor call 0x21 (MMU_TSB_CTXNON0) and panic


# 2659:de6ef283fa99 31-Aug-2006 susans

6458128 SAP performs 5% better with 4M pages than with 32M pages


# 2414:4349245e130c 21-Jul-2006 aguzovsk

4614772 MPSS to be extended to shared memory
6413095 madvise() freeing up whole pages even when passed sizes smaller than a page
6416714 anon_map_getpages panic during stress test
6428347 multilple pagelocks/softlocks on the same large page should decrement availrmem just once


# 2368:4f70fed83e4f 13-Jul-2006 jimand

6402533 LP OOB policies must not allow the use of more than one large pagesize on USIII+-USIV+ systems


# 2359:7716bbee8e29 12-Jul-2006 ae112802

6447334 sfmmu_kpm_vatopfn incorrectly returns -1


# 2296:4644f3dc5596 28-Jun-2006 ae112802

6261101 VAC alias handling code should be compiled out for sun4v


# 2251:99824a0e7def 21-Jun-2006 elowe

6331963 Fix for 6272550 could be improved
6426868 race between hat_add_callback() and hat_page_relocate() causes system panic
6429406 pci nexus driver must use a unique 'key' in hat_register_callback()


# 2241:592fbc504a44 20-Jun-2006 huah

6373298 Integrate support for MMU context ID domains


# 1772:78cca3d2cc4b 08-Apr-2006 jl139090

PSARC 2004/750 Solaris for OPL
6379529 Solaris for OPL Project


# 1517:4c6d6b1ea65a 01-Mar-2006 jb145095

6349262 sun4v ppcopy and pagezero shouldn't use ppmapin/ppmapout


# 1338:c2e71173ece3 27-Jan-2006 elowe

6344261 segkmem should not return a shadow list with holes
6354842 ecc_page_zero() should try to lock pages with SE_RETIRED
6372151 page_retire can mangle PR_MESSAGEs of PAE pages
6372720 setting kpm_enable to 0 may lead to system panic at dr operations
6376283 page retire would panic if page_reclaim() were to fail
6376967 TSufs induces panic in hat_kpm_fault on amd64


# 824:e020aac5419c 02-Nov-2005 elowe

6342112 E20K with 118822-19 panics when b_addr is a segkmap address


# 819:3f228c52a5e6 01-Nov-2005 aguzovsk

6340768 Erie 2 GB system panics while executing memory testing


# 735:be9b33572cf3 21-Oct-2005 elowe

6333712 SunFire 25K panics with IB card installed with HSI/P or ATM installed


# 670:fead92dad0b1 10-Oct-2005 elowe

6277224 7.6% Cadence regression in snv_12
6289141 reading from /dev/kmem may crash the system
6313259 panic: hat_register_callback: out of callback IDs


# 0:68f95e015346 14-Jun-2005 stevel@tonic-gate

OpenSolaris Launch


# 12908:80a39220b451 26-Jul-2010 Pavel Tatashin <Pavel.Tatashin@Sun.COM>

6948098 freemem_lock contention causes poor concurrent DISM MC_LOCK/MC_UNLOCK performance
6946742 sfmmu_mlspl_enter lock contention on mml_table
6956454 ani_free_pool lock contention with multi processed ISM


# 12230:12df01648ca5 23-Apr-2010 Peter Rival <Frank.Rival@oracle.com>

6778289 vm locks need to scale with the size of system (strands/memory size)


# 11713:03615b084875 19-Feb-2010 Pavel Tatashin <Pavel.Tatashin@Sun.COM>

6892591 per-MMU context id domains for sun4v


# 11178:d671c0911ce4 23-Nov-2009 Dave Plauger <Dave.Plauger@Sun.COM>

6898505 kmem_verify oddities around Fast Crash Dump integration


# 10843:eb166ea6e74e 22-Oct-2009 Dave Plauger <Dave.Plauger@Sun.COM>

6828976 Fast Crash Dump
6878030 live crash dump is much slower than reboot dump
6626023 Crash dump size is excessive on large memory machines


# 10701:dabf35f459b8 30-Sep-2009 Feri Chua <Feri.Chua@Sun.COM>

6778854 System panics in unix:sfmmu_ismtlbcache_demap


# 10271:7c80b70bb8de 06-Aug-2009 Jason Beloro <Jason.Beloro@Sun.COM>

6858457 Remove Solaris support for UltraSPARC-AT10 processor


# 9281:67a3435ea50e 03-Apr-2009 Prakash Sangappa <Prakash.Sangappa@Sun.COM>

6811473 VPM interfaces should provide multiple pagelength mappings per request


# 8819:4df632c9b303 13-Feb-2009 Jason Beloro <Jason.Beloro@Sun.COM>

FWARC 2008/592 MMU_EXT API group name and number change
6729256 kernel routines such as bcopy, bzero need to be optimized for UltraSPARC-AT10.
6757032 Use TLB search order register API
6771318 Some packets get dropped from the performance counter ringbuffer
6781121 CPC: Rock PCBE does not correctly handle overflow profiling. 64-bit extension is incorrect.
6791429 SUNW,UltraSPARC-AT10 should not be using ASI_NQUAD_LD since it is redefined as ASI_STICK_CMPR


# 8574:34b33d2c8168 19-Jan-2009 Jason Beloro <Jason.Beloro@Sun.COM>

6755072 Frequency of calls to demap all could be reduced
6757028 Improve CRP error packet handling
6791608 SW support for humidity sensors
6793427 Optimize update to mmu primary context id 1


# 8187:07a2f708e698 21-Nov-2008 Amritpal Sandhu <Paul.Sandhu@Sun.COM>

6638112 eliminate HMELOCK_ENTER() to avoid xcall timeout panic and improve GET_TTE() performance
6620705 sfmmu_hblk_hash_add/rm should not access hme buckets via va
6638127 TSB_UPDATE_TL()/TSB_UPDATE_TL_PN() shouldn't spin forever in TSB_LOCK_ENTRY()


# 7869:10227e8ccea5 17-Oct-2008 Sean McEnroe <Sean.McEnroe@Sun.COM>

6729139 Shared context causes SEGVs and panics on OPL


# 7718:555eee33ff6e 26-Sep-2008 Jason Beloro <Jason.Beloro@Sun.COM>

PSARC/2007/003 Solaris Support for the Rock Processor
6368478 Solaris needs to support the Rock processor
6440653 stores to sun4v error queue head registers are missing a membar #Sync
6639717 FLUSH instruction used more than necessary in sun4v


# 7393:8a8d73648b44 22-Aug-2008 Donghai Qiao <Donghai.Qiao@Sun.COM>

6672470 Looping within uiomove()/xcopyout()/copyout_move()
6707987 kpm has some VAC handling issues
6388567 VAC flushing is broken on US III-IV+ for large pages


# 7240:c4957ab6a78e 31-Jul-2008 rh87107

6698108 Sun Studio 12 finds many lint errors in ON source.


# 6695:12d7dd4459fd 22-May-2008 aguzovsk

6423097 segvn_pagelock() may perform very poorly
6526804 DR delete_memory_thread, AIO, and segvn deadlock
6557794 segspt_dismpagelock() and segspt_shmadvise(MADV_FREE) may deadlock
6557813 seg_ppurge_seg() shouldn't flush all unrelated ISM/DISM segments
6557891 softlocks/pagelocks of anon pages should not decrement availrmem for memory swapped pages
6559612 multiple softlocks on a DISM segment should decrement availrmem just once
6562291 page_mem_avail() is stuck due to availrmem overaccounting and lack of seg_preap() calls
6596555 locked anonymous pages should not have assigned disk swap slots
6639424 hat_sfmmu.c:hat_pagesync() doesn't handle well HAT_SYNC_STOPON_REF and HAT_SYNC_STOPON_MOD flags
6639425 optimize checkpage() optimizations
6662927 page_llock contention during I/O


# 6665:1515bc919a93 20-May-2008 paulsan

6701890 panic: hat_unlock_region: addr not found panic


# 6127:3731a676ec5e 29-Feb-2008 sm142603

6642758 Support Shared Context for SPARC64-VII


# 6044:56828186ab76 20-Feb-2008 paulsan

6656203 Stress shm24 test - panic[cpu2]/thread=300daf0e020: ptl1_panic reason 0x15 at TL 1


# 5110:6c913b66a74a 21-Sep-2007 paulsan

6606499 Fix cstyle error in hat_sfmmu.c


# 5075:199eb2ec2c2c 17-Sep-2007 paulsan

6597746 Don't create srd_cache, region_cache and scd_cache on systems that don't use shared context
6597751 SFMMU_HASH_UNLOCK() call is missing in hat_unlock_region()
6600745 shared context code needs to be cleaned up
6603269 assertion failed: sfhme->hme_tte.ll != 0, file: ../../sfmmu/vm/hat_sfmmu.c, line: 11613


# 4528:9ad45715d2ab 24-Jun-2007 paulsan

PSARC 2006/266 Shared Context Support
PSARC 2006/267 Shared Region HME Block support
6388600 do_virtual_coloring checks should be expunged from sfmmu
6449192 Integrate support for MMU Shared Contexts
6449195 Integrate support for Shared Region HME Blocks


# 4421:8ddf0d909338 08-Jun-2007 mb158278

6491350 Auto mpss and kpr combo can make a system crawl in case of fragmented memory.


# 4324:696de332c6cf 25-May-2007 qiao

6538758 Sporadic I/O error will cause data corruption in Veritas CFS with VMODSORT mode on.


# 4266:c151d8b35988 17-May-2007 dp78419

6547155 mptest timing out because of bad page coloring after fork
6550904 kcage_glist_alloc() may not call kmem_zalloc()


# 4236:de290ec1e532 14-May-2007 dm120769

backout 6547155/6550904: sparc build failed


# 4230:9ab9040697c4 14-May-2007 dp78419

6547155 mptest timing out because of bad page coloring after fork
6550904 kcage_glist_alloc() may not call kmem_zalloc()


# 3896:1d5fb739bf40 26-Mar-2007 vb70745

6509040 hang in sfmmu_hblk_steal()


# 3764:74844940a161 05-Mar-2007 dp78419

6510878 Kernel Boot time data structures can be mapped with Large pages to improve performance


# 3687:96d6b6ad3918 20-Feb-2007 jb145095

6244343 sun4v mmu code should support 16-bit context and correctly mask mmu fault addresses


# 3351:853fb8df244a 02-Jan-2007 aguzovsk

6498368 Reduce va_to_pa overheads
6500938 tte_remap panics with inconsistent sf_hments


# 3290:256464cbb73c 20-Dec-2006 johansen

4894692 caching data in heap inflates crash dump
6499454 time to increase size of kmem default allocation caches
6499459 vm should stop checking kvp directly


# 3258:077f19d9741d 15-Dec-2006 trevtom

6351793 sfmmu_mlist_enter recursive mutex_enter.
6467201 Duplicate hrmstat structures cause long hat_statlock hold times.


# 3000:f09e516c394d 27-Oct-2006 stans

6393251 data corruption caused by VMODSORT problem (missing changes)


# 2991:4b13d6c49c6b 26-Oct-2006 susans

6254029 memcntl() MC_HAT_ADVISE with page size 0 may cause segment page sizes to be demoted
6325885 map_pgszstk() uses p->p_brkpageszc rather than p->p_stkpageszc
6371967 assign large pages to anon segment created using mmap /dev/zero
6483208 unify and cleanup OOB (out of the box) large pagesize selection code
6483216 use intermediate pagesizes to map the beginning of bss/heap and stack when it may help performance
6483226 bss size is not properly taken into account by LP OOB policy at exec() time
6483230 grow_internal() doesn't properly align stack bottom for large pages
6483231 memcntl.c: ASSERT(IS_P2ALIGNED(p->p_brkbase + p->p_brksize, pgsz));
6483233 provide a mechanism to enable the use of 32M text pages on OPL by default
6485171 memcntl() shouldn't silently fail when stack space is unavailable with requested pagesize


# 2874:ef6ddc19af4e 06-Oct-2006 vb70745

6457396 missing 2nd sfmmu_info causes failure of hypervisor call 0x21 (MMU_TSB_CTXNON0) and panic


# 2659:de6ef283fa99 31-Aug-2006 susans

6458128 SAP performs 5% better with 4M pages than with 32M pages


# 2414:4349245e130c 21-Jul-2006 aguzovsk

4614772 MPSS to be extended to shared memory
6413095 madvise() freeing up whole pages even when passed sizes smaller than a page
6416714 anon_map_getpages panic during stress test
6428347 multilple pagelocks/softlocks on the same large page should decrement availrmem just once


# 2368:4f70fed83e4f 13-Jul-2006 jimand

6402533 LP OOB policies must not allow the use of more than one large pagesize on USIII+-USIV+ systems


# 2359:7716bbee8e29 12-Jul-2006 ae112802

6447334 sfmmu_kpm_vatopfn incorrectly returns -1


# 2296:4644f3dc5596 28-Jun-2006 ae112802

6261101 VAC alias handling code should be compiled out for sun4v


# 2251:99824a0e7def 21-Jun-2006 elowe

6331963 Fix for 6272550 could be improved
6426868 race between hat_add_callback() and hat_page_relocate() causes system panic
6429406 pci nexus driver must use a unique 'key' in hat_register_callback()


# 2241:592fbc504a44 20-Jun-2006 huah

6373298 Integrate support for MMU context ID domains


# 1772:78cca3d2cc4b 08-Apr-2006 jl139090

PSARC 2004/750 Solaris for OPL
6379529 Solaris for OPL Project


# 1517:4c6d6b1ea65a 01-Mar-2006 jb145095

6349262 sun4v ppcopy and pagezero shouldn't use ppmapin/ppmapout


# 1338:c2e71173ece3 27-Jan-2006 elowe

6344261 segkmem should not return a shadow list with holes
6354842 ecc_page_zero() should try to lock pages with SE_RETIRED
6372151 page_retire can mangle PR_MESSAGEs of PAE pages
6372720 setting kpm_enable to 0 may lead to system panic at dr operations
6376283 page retire would panic if page_reclaim() were to fail
6376967 TSufs induces panic in hat_kpm_fault on amd64


# 824:e020aac5419c 02-Nov-2005 elowe

6342112 E20K with 118822-19 panics when b_addr is a segkmap address


# 819:3f228c52a5e6 01-Nov-2005 aguzovsk

6340768 Erie 2 GB system panics while executing memory testing


# 735:be9b33572cf3 21-Oct-2005 elowe

6333712 SunFire 25K panics with IB card installed with HSI/P or ATM installed


# 670:fead92dad0b1 10-Oct-2005 elowe

6277224 7.6% Cadence regression in snv_12
6289141 reading from /dev/kmem may crash the system
6313259 panic: hat_register_callback: out of callback IDs


# 0:68f95e015346 14-Jun-2005 stevel@tonic-gate

OpenSolaris Launch