Supporting Multiple Page Sizes in the Solaris
Contents
1. Ss The mapping is shared such that changes made in the observed address space are committed to the mapped file and are visible from all other processes sharing the mapping R Swap space is not reserved for this mapping Mappings created with MAP_NORESERVE and System V ISM shared memory mappings do not reserve swap space Mapping Name Mapped File A descriptive name for each mapping The following major types of names are displayed for mappings o A mapped file For mappings between a process and a file the pmap command attempts to resolve the file name for each mapping If the file name cannot be resolved pmap displays the major and minor number of the device containing the file and the file system inode number of the file o Anonymous memory Memory not relating to any named object or file within the file system is reported as anon continued on next page 10 Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page The pmap command displays common names for cer tain known anonymous memory mappings such as heap The process heap stack The process stack If the common name for the mapping is unknown pmap displays anon as the mapping name o System V Shared Memory Mappings created using System V shared memory system calls are reported with the names shown below shmid n The mapping is a System V share2. Ce produit ou document est prot g par un copyright et distribu avec des licences qui en restreignent l utilisation la copie la distribution et la d compilation Aucune partie de ce produit ou document ne peut tre reproduite sous aucune forme par quelque moyen que ce soit sans l autorisation pr alable et crite de Sun et de ses bailleurs de licence s il y en a Le logiciel d tenu par des tiers et qui comprend la technologie relative aux polices de caract res est prot g par un copyright et licenci par des fournisseurs de Sun Des parties de ce produit pourront tre d riv es des syst mes Berkeley BSD licenci s par l Universit de Californie UNIX est une marque enregistree aux Etats Unis et dans d autres pays et licenci e exclusivement par X Open Company Ltd Sun Sun Microsystems le logo Sun Sun BluePrints Sun Cluster Solstice DiskSuite StarOffice et Solaris sont des marques de fabrique ou des marques d pos es ou marques de service de Sun Microsystems Inc aux Etats Unis et dans d autres pays Toutes les marques SPARC sont utilis es sous licence et sont des marques de fabrique ou des marques d pos es de SPARC International Inc aux Etats Unis et dans d autres pays Les produits portant les marques SPARC sont bas s sur une architecture d velopp e par Sun Microsystems Inc L interface d utilisation graphique OPEN LOOK et Sun a t d velopp e par Sun Microsystems Inc pour ses utilisateurs e
3. A program compiled with this option will not link on Solaris 7 and 8 If you specify xpagesize default the Solaris environ ment sets the page size xpagesize without an argument is the equivalent to xpagesize default Compiling with this option has the same effect as set ting the LD_PRELOAD environment variable to mpss so 1 with the equivalent options or running the Solaris 9 command ppgsz 1 with the equivalent options before running the program See the Solaris 9 man pages for details This option is a macro for xpagesize_heap and xpagesize _stack These two options accept the same arguments as xpagesize 8K 64K 512K 4M 32M 256M 2G 16G default You can set them both with the same value by specifying xpagesize n or you can specify them individually with different values xpagesize_heap n SPARC Set the page size in memory for the heap n can be 8K 64K 512K 4M 32M 256M 2G 16G or default You must specify a valid page size for the Solaris operating environment on the target platform as returned by getpagesize 3C If you do not specify a valid page size the request is silently ignored at run time You can use pmap l or meminfo 2 to determine page size at the target platform continued on next page cc 1 continued from preceding page If you specify xpagesize_heap default the Solaris environment sets the page size xpagesize_heap without an argument is the equivalent to xpage
4. Due to resource constraints the setting of the preferred page size does not necessarily guarantee that the target process es will get the preferred page size Use pmap 1 to view the actual heap and stack page sizes of the target process es see pmap s option Large pages are required to be mapped at addresses that are multiples of the size of the large page Given that the heap is typically not large page aligned the starting portions of the heap below the first large page aligned address are mapped with the system memory page size See getpagesize 3C To provide a heap that will be mapped with a large page size an application can be built using a link editor ld 1 mapfile containing the bss segment declaration directive Refer to the section Mapfile Option in the Linker and Libraries Guide for more details of this direc tive and the template mapfile provided in usr 1ib 1d map bssalign Users are cautioned that an align ment specification may be machine specific and may lose its benefit on different hardware platforms A more flexible means of requesting the most optimal underlying page size may evolve in future releases mpss so 1 1 a preloadable shared object can also be used to set the preferred stack and or heap page sizes Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 trapstat 1M NAME trapstat report trap statistics SYNOPSIS usr platform pl
5. in the pid list after ppgsz completes will inherit the preferred heap and stack page sizes The preferred page sizes are set back to the default system page size on exec 2 see getpagesize 3C o option option The options are heap size This option specifies the preferred page size for the heap of the target process es heap is defined to be the bss uninitialized data and the brk area that immediately follows the bss see brk 2 The preferred heap page size is set for the existing heap and for any additional heap memory allocated in the future See NOTES stack size This option specifies the preferred page size for the stack of the target process es The preferred stack page size is set for the exist ing stack and newly allocated parts of the stack as it expands continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page At least one of the above options must be specified size must be a supported page size see pagesize 1 or 0 in which case the system will select an appropriate page size see memcntl 2 size defaults to bytes and can be specified in octal 0 decimal or hexadecimal 0x The numeric value can be qualified with K M G or T to specify Kilo bytes Megabytes Gigabytes or Terabytes respec tively 4194304 0x400000 4096K 0x1000K and 4M are different ways to specify 4 Megabytes EXAMPLE
6. parentheses below Virtual Address Address The first column of output represents the starting virtual address of each mapping Virtual addresses are displayed in ascending order Virtual Mapping Size Kbytes The virtual size in kilobytes of each mapping Resident Physical Memory RSS The amount of physical memory in kilobytes that is resident for each mapping including that which is shared with other address spaces Anonymous Memory Anon The number of pages counted by using the system page size of anonymous memory associated with the speci fied mapping Anonymous memory shared with other address spaces is not included unless the a option is specified Anonymous memory is reported for the process heap stack for copy on write pages with mappings mapped with MAP_PRIVATE see mmap 2 continued on next page pmap 1 9 continued from preceding page Locked Locked The number of pages locked within the mapping Typical examples are memory locked with mlock and System V shared memory created with SHM SHARE _MMU Permissions Flags Mode The virtual memory permissions are shown for each map ping Valid permissions are Le The mapping may be read by the process w The mapping may be written by the process User Commands pmap 1 xX Instructions that reside within the mapping may be executed by the process Flags showing additional information for each mapping may be displayed
7. 16 The amount of incremental memory used by Mode r x rwx rwx re rwx r x rwx r x r x rwx r x 9e r x rwx rw each Mapped File sh sh heap en_US 1S08859 1 s0 2 en_US 1S08859 1 s0 2 libc so 1 liberso 1 libc_psr so l libgen so 1 libgen so 1 11bd1 50 1 anon ld so 1 1d so l stack additional instance of a process can be estimated by using the resident and anonymous memory counts of each mapping In the above example size of 1032Kbytes However memory used by the shell is shared with other shell Another physical memory with the other shell where identical the bourne shell has a resident memory a large amount of the physical instances of instance of the shell will share possible and allocate anonymous memory for any non shared portion In the above example each additional bourne shell uses approximately 56Kbytes of additional physical memory A more complex example shows the output format for a process containing different mappings are as follows 0001000 0002000 0002200 0300000 0400000 0500000 Executable text Executable data Program heap A mapped file A mapped file A mapped file continued on next page mapping types In this example the mapped from maps program mapped from maps program mapped MAP_SHARED mapped MAP_PRIVATE mapped MAP_PRIVATE MAP_N
8. MC_LOCKAS or MC_UNLOCKAS is specified the arg argument is not valid for the func tion specified mha_pagesize or mha_cmd is invalid or MC_HAT_ADVISE is specified and not all pages in the specified region have the same access permissions within the given size boundaries ENOMEM When the selection criteria match some or all of the addresses in the range addr addr len are invalid for the address space of a process or specify one or more pages which are not mapped EPERM The process s effective user ID is not superuser and MC_LOCK MC_LOCKAS MC_UNLOCK or MC _UNLOCKAS was specified ATTRIBUTES See attributes 5 for descriptions of the following attri butes ATTRIBUTE TYPE ATTRIBUTE VALUE MT Level MT Safe SEE ALSO ppgsz 1 fork 2 mmap 2 mprotect 2 getpagesizes 3C mctl 3UCB mlock 3C mlockall 3C msync 3C plock 3C sysconf 3C attributes 5 42 Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 About the Author Richard has over 15 years of UNIX experience including application design kernel development and performance analysis Richard specializes in operating system tools and architecture Ordering Sun Documents The SunDocs program provides more than 250 manuals from Sun Microsystems Inc If you live in the United States Canada Europe or Japan you can purchase documentation sets or individual manuals th
9. Permissions within each mha_pagesize aligned portion of the region must be consistent When a size of 0 is specified the system selects an appropriate size based on the size and alignment of the memory region type of processor and other con siderations MHA_MAPSIZE_STACK sets the preferred hardware address translation mapping size of the process main thread stack segment The addr and len arguments must be NULL and 0 respectively MHA_MAPSIZE_BSSBRK sets the preferred hardware address translation mapping size of the process heap The addr and len arguments must be NULL and 0 respectively See the NOTES section of the ppgsz l manual page for additional information on process heap alignment The attr argument must be 0 for all MC_HAT_ADVISE operations continued on next page 40 Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page The mask argument must be 0 it is reserved for future use Locks established with the lock operations are not inherited by a child process after fork 2 The memcntl function fails if it attempts to lock more memory than a system specific limit Due to the potential impact on system resources all opera tions except MC_SYNC are restricted to processes with superuser effective user ID USAGE The memcntl function subsumes the operations of plock 3C and mctl 3UCB MC_HAT_ADVISE is intended to improve performa
10. on page 17 m trapstat 1M on page 21 m mpss so 1 1 on page 32 a mementl 2 on page 36 cc 1 NAME SYNOPSIS ec cc C compiler continued on next page Aname tokens B static dynamic C c Dname tokens d yIn dalign E errfmt no error erroff t t errshort i errtags a errwarn t t fast fd flags fnonstd fns yes no fprecision p fround r fsimplel n fsingle fstore ftrap t t G g H hname I dir i KPIC Kpic keeptmp Ldir lname mc misalign misalign2 mr string mt native nofstore O ofilename P p Qlyin qp RATE Pedi sd S s Uname V v Wc arg w X claltis x386 x486 xa xalias_level a xarch a xautopar xbuiltin a xCC xc99 0 xcache c xcg89 xcg92 xchar o xchar_byte_order o xcheck n xchip c xcode v xcrossfile n xcsi xdebugformat stabs dwarf xdepend yes no xdryrun xe xexplicitpar xF xhelp f xhwcprof enable disable xild off xildon xinline v v xipo a xjobs n xldscope v xlibmieee xlibmil xlic_lib sunperf xlicinfo xlinkopt level xloopinfo xM xM1 xMerge xmaxopt v xmemalign ab xnativeconnect a a xnolib xnolibmil xOn xo
11. 0 1 121 0 0 1 6 3 K 54 0 0 0 0 0 102682 12 6 0220 4 a 4 4 u 3064 0 3 0 0 0 10832 1 120 0 0 1 6 4k 31 0 0 0 0 0 107977 13 236 Oa De 134 T EE CELL 14816 0 3 CTO 585937 14 1009 0 0 114 5 continued from preceding page Example 4 Using trapstat with TLB Statistics and Page Size Information By specifying the T option trapstat shows TLB misses bro ken down by page size In this example CPU 0 is spending 7 9 percent of its time handling user mode TLB misses on 8K pages and another 2 3 percent of its time handling user mode TLB misses on 64K pages example trapstat T c 0 cpu m size itlb miss Stim itsb miss tim dtlb miss tim dtsb miss tim tim 4 Ou 8k 1300 0 1 15 0 0 104897 7 9 90 0 0 8 0 Ou 64k Qi 0 0 Qi 10 50 29935 2 53 To 05 10 253 0 u 512k 0 0 0 0 0 0 3569 0 2 2 050 0 2 Ou 4m 0 0 0 0 0 0 233 gt 0 0 2 0 0 0 0 gt t 7 t t 0 k 8k 15 0 30 0 0 0 1733 6 5 110 0 0 6 5 O k 64k 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 k 512k 0 0 0 0 0 0 0 0 0 206 0 1 0 1 O k 4m 0 0 0 0 0 0 0 0 0 0 0 0 0 0 RE EEL 1313 0 1 15 0 0 210367 17 1 AUD OS 12S Example 5 Using trapstat with Entry Filtering By specifying the e option trapstat displays statistics for only specific trap types Using this op
12. 2 512 z 8K rw R dev 0 2 ino 4628487 05080000 512 512 ee 3 rw R dev 0 2 ino 4628487 06000000 1024 1024 1024 3 8K rw anon 07000000 512 512 512 8K rw R anon 08000000 8192 8192 8192 rwxs dism shmid 0x5 09000000 4096 4096 7 8K rwxs dism shmid 0x4 04000000 4096 3 z LWXS dism shmid 0x2 0B000000 8192 8192 8192 4M rwxsR ism shmid 0x3 FF280000 136 136 SK r z libes FF2A2000 120 120 Sugeges Jibe se T FF2C0000 128 128 SK r k 11156 60 1 FF2E0000 200 200 PR Libo sa l FF312000 48 48 8K r x libc so l FF31E000 48 40 rege Lib so FF33A000 32 32 32 8k rwx libc so l FF390000 8 8 8K r x libc psr so 1 FF3A0000 8 8 ok r x e libdl s6 1 FF3B0000 8 8 8 8K rwx anon FF3C0000 152 152 BR r x ldvso l FF3F6000 8 8 8 8K rwx ld so l FFBFA000 24 24 24 8K rwx stack total Kb 50464 42264 18888 16384 continued on next page pmap 1 15 continued from preceding page Example 4 Displaying swap reservations The S option can be used to describe the swap reservations for a process The amount of swap space reserved is displayed for each mapping within the process Swap reserva tions are reported as zero for shared mappings since they are accounted for only once system wide exampleS pmap S 15492 15492 maps Address Kbytes Swap Mode Mapped File 00010000 8 FoR maps 00020000 8 8 rwx maps 00022000 203
13. 44 20344 rwx heap 03000000 1024 rw s dev 0 2 ino 4628487 04000000 1024 1024 rw dev 0 2 ino 4628487 05000000 1024 512 rw R dev 0 2 ino 4628487 06000000 1024 1024 rw anon 07000000 512 512 rw R anon 08000000 8192 rwxs dism shmid 0x5 09000000 8192 rwxs dism shmid 0x4 0A000000 8192 rwxs dism shmid 0x2 0B000000 8192 rwxsR ism shmid 0x3 FF280000 680 rex Tibe so FF33A000 32 92 fwe Libe so 1 FF390000 8 tex Libs psr so l FF3A0000 8 Sgen Vi bd eo L FF3B0000 8 8 rwx anon FF3C0000 152 Res Idees FF3F6000 8 8 rwx ld so 1 FFBFAOOO 24 24 rwx stack total Kb 50464 23496 The swap reservation information can be used to estimate the amount of virtual swap used by each additional process Each process consumes virtual swap from a global virtual swap pool Global swap reservations are reported by the avail field of the swap 1M command EXIT STATUS The following exit values are returned 0 Successful operation non zero An error has occurred FILES proc process files usr proc lib proc tools supporting files continued on next page 16 Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page ATTRIBUTES See attributes 5 for descriptions of the following attributes ATTRIBUTE TYPE ATTRIBUTE VALUE Availability SUNWesu 32 bit SUNWesxu 64 bit Int
14. 8336 38057 6c dtlb prot 1046 549 417 545 370 84 spill user 32 66551 29480 301588 26522 213032 88 spill user 64 0 318652 111239 299829 221716 8c spill user 32 cln 856 347 331 416 293 90 spill user 64 cln 0 55 21 59 39 98 spill kern 64 66464 31803 24758 34004 22277 a4 spill asuser 32 1423 569 560 698 483 a8 spill asuser 64 0 74 32 98 46 ac spill asuser 32 cln 4875 2250 1728 2384 1584 b0 spill asuser 64 cln 0 2 0 1 0 c4 fill user 32 64193 28418 287516 27055 202093 c8 fill user 64 0 305016 106692 288542 210654 cc fill user 32 cln 6733 3520 15185 2396 12035 d0 fill user 64 cln 0 13226 3506 12933 11032 d8 fill kern 64 66220 31680 24674 33892 22196 108 syscall 32 2446 967 817 1196 755 Example 3 Using trapstat with TLB Statistics The t option displays in depth TLB statistics including the amount of time spent performing TLB miss processing The following example shows that the machine is spending 14 1 percent of its time just handling D TLB misses example trapstat t cpu m itlb miss tim itsb miss tim dtlb miss tim dtsb miss Stim tim ae ee pee ooo Ss SS E eS E ee ee Ou 2571 0 3 Ou 00 10802 O i O0 16 0k 0 0 0 0 0 0 106420 13 184 0 1 113 6 oo lu 3069 0 3 0 0 0 10983 1 100 0 0 1 436 Ink 27 0 0 0 0 0 106974 12 19 0 0 112 7 ne Te pon M dr ve pee 2u 3033 04 8 0 0 0 11045 1 1O5 gt 0 30 1 6 2k 43 0 0 0 0 0 107842 12 108 0 0 112 8 oo 3 0 2924 0 3 0 0 0 1038
15. ORESERVE pmap 1 13 14 continued from preceding page 0600000 0700000 0800000 0900000 0A00000 0B00000 examples 15492 Address 00010000 00020000 00022000 03000000 04000000 05000000 06000000 07000000 08000000 09000000 0A000000 0B000000 FF280000 FF33A000 FF390000 FF3A0000 FF3B0000 FF3C0000 FF3F6000 FFBFA000 Anonymous memory created by mapping Anonymous memory created by mapping wit AD wit AD wit AD wit h MAP_NORESERVE ISM shared memory mapping created h 8MB locked via mlock 2 ISM shared memory mapping created h 4MB of its pages touched ISM shared memory mapping created h none of its pages touched dev zero dev zero with SHM_PAGEABLE with SHM_PAGEABLE with SHM_PAGEABLE A ISM shared memory mapping created with SHM_SHARE_MMU pmap xs 15492 maps Kbytes RSS Anon Locked Mode 8 8 a 8 8 8 LWX 20344 16248 16248 rwx 1024 1024 rw s 1024 1024 512 Ew 1024 1024 512 rw R 1024 1024 1024 rw 512 512 512 rw R 8192 8192 Pad 8192 rwxs 8192 4096 rwxs 8192 8192 S 8192 rwxsR 8192 8192 8192 rwxsR 680 672 So ee 32 32 32 LWX 8 8 Lx 8 8 EX 8 8 8 LWX 152 152 7 SSRs 8 8 8 LWX 24 24 24 LWX 50464 42264 18888 16384 continued on next page Mapped File maps maps heap dev 0 2 ino 4628487 dev 0 2 ino 4628487 dev 0 2 i
16. S Example 1 Setting the preferred heap and stack page size The following example sets the preferred heap page size to 4M and the preferred stack page size to 512K for all ora owned processes running commands that begin with ora nA nN example ppgsz o heap 4M stack 512K p pgrep u ora ora EXIT STATUS If cmd is specified and successfully invoked see exec 2 the exit status of ppgsz will be the exit status of cmd Otherwise ppgsz will exit with one of the following values 0 Successfully set preferred page size s for processes in the pid list 125 An error occurred in ppgsz Errors include invalid argument invalid page size s specified and failure to set preferred page size s for one or more processes in the pid list or cmd 126 cmd was found but could not be invoked 127 cmd could not be found FILES proc Process files usr lib 1d map bssalign A template link editor mapfile for aligning bss see NOTES continued on next page ppgsz 1 19 20 continued from preceding page ATTRIBUTES See attributes 5 for descriptions of the following attributes ATTRIBUTE TYPE ATTRIBUTE VALUE Availability SUNWesu 32 bit SUNWesxu 64 bit Interface Stability Evolving SEE ALSO ld 1 mpss so 1 1 pagesize 1 pgrep 1 pmap 1 proc l brk 2 exec 2 fork 2 memcntl 2 sbrk 2 getpagesize 3C proc 4 attributes 5 Linker and Libraries Guide NOTES
17. Supporting Multiple Page Sizes in the Solaris Operating System Appendix Richard McDougall PAE Sun BluePrints OnLine March 2004 amp Sun microsystems http www sun com blueprints Sun Microsystems Inc 4150 Network Circle Santa Clara CA 95045 U S A 650 960 1300 Part No 817 6242 10 Revision 1 0 3 10 04 Edition March 2004 Copyright 2004 Sun Microsystems Inc 4150 Network Circle Santa Clara California 95045 U S A All rights reserved Sun Microsystems Inc has intellectual property rights relating to technology described in this document In particular and without limitation these intellectual property rights may include one or more patents or pending patent applications in the U S or other countries This product or document is protected by copyright and distributed under licenses restricting its use copying distribution and decompilation No part of this product or document may be reproduced in any form by any means without prior written authorization of Sun and its licensors if any Third party software including font technology is copyrighted and licensed from Sun suppliers Parts of the product may be derived from Berkeley BSD systems licensed from the University of California UNIX is a registered trademark in the United States and other countries exclusively licensed through X Open Company Ltd Sun Sun Microsystems the Sun logo Sun BluePrints Sun Cluster Solstice DiskSuite StarOffice a
18. To accommodate such workloads the operating system supports multiple page sizes larger page sizes increase the effec tive TLB reach and thereby reduce the number of TLB misses To provide insight into the relationship between page size and TLB miss rate trapstat optionally provides in depth TLB miss information broken down by page size using the T option The information provided by the T option is a superset of that provided by the t option only one of t and T can be specified OPTIONS The following options are supported a Displays the number of traps as accumulating monoton ically increasing values instead of per second or per interval rates c cpulist Enables trapstat only on the CPUs specified by cpu list cpulist can be a single processor ID for example 4 a range of processor IDs for example 4 6 ora comma separated list of processor IDs or processor ID ranges for example 4 5 6 or 4 6 8 C processor_set_id Enables trapstat only on the CPUs in the processor set specified by processor_set_id continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page trapstat modifies its output to always reflect the CPUs in the specified processor set If a CPU is added to the set trapstat modifies its output to include the added CPU if a is removed from the set trapstat modifies its output to exclu
19. address space on which to operate If an operation shall not be constrained by the selection criteria attr must have the value 0 The operation to be performed is identified by the argument cmd The symbolic names for the operations are defined in lt sys mman h gt as follows MC_LOCK Lock in memory all pages in the range with attributes attr A given page may be locked multiple times through different mappings however within a given mapping page locks do not nest Multiple lock opera tions on the same address in the same process will all be removed with a single unlock operation A page locked in one process and mapped in another or visi ble through a different mapping in the locking pro cess is locked in memory as long as the locking pro cess does neither an implicit nor explicit unlock operation If a locked mapping is removed or a page is deleted through file removal or truncation an unlock operation is implicitly performed If a writ able MAP_PRIVATE page in the address range is changed the lock will be transferred to the private page The arg argument is not used but must be 0 to ensure compatibility with potential future enhancements MC_LOCKAS Lock in memory all pages mapped by the address space with attributes attr The addr and len arguments are not used but must be NULL and 0 respectively to ensure compatibility with potential future enhance ments The arg argument is a bit pattern built from the flags T
20. atform name sbin trapstat t T e entry C processor_set_id c cpulist P a r rate interval count command args usr platform platform name sbin trapstat 1 DESCRIPTION The trapstat utility gathers and displays run time trap statistics on UltraSPARC based systems The default output is a table of trap types and CPU IDs with each row of the table denoting a trap type and each column of the table denoting a CPU If standard output is a terminal the table contains as many columns of data as can fit within the ter minal width if standard output is not a terminal the table contains at most six columns of data By default data is gathered and and displayed for all CPUs if the data cannot fit in a single table it is printed across multiple tables The set of CPUs for which data is gathered and displayed can be optionally specified with the c or C option Unless the r option or the a option is specified the value displayed in each entry of the table corresponds to the number of traps per second If the r option is speci fied the value corresponds to the number of traps over the interval implied by the specified sampling rate if the a option is specified the value corresponds to the accumu lated number of traps since the invocation of trapstat By default trapstat displays data once per second and runs indefinitely both of these behaviors can be optionally con trolled with the interval and c
21. ation or exe cutable The corresponding preferred page size s are set for newly created processes see getexecname 3C that match the first exec name found in the file exec name can be a full pathname a base name or a pattern string See File Name Generation in sh 1 for a discussion of pattern matching If heap size and or stack size are not specified the corresponding preferred page size s will not be set MPSSCFGFILE takes precedence over MPSSHEAP and MPSSSTACK continued on next page mpss so 1 1 33 34 continued from preceding page MPSSERRFILE pathname By default error messages are logged via syslog 3C using level LOG_ERR and facility LOG_USER If MPSSERRFILE contains a valid pathname such as dev stderr error messages will be logged there instead EXAMPLES Example 1 Configuring preferred page sizes using MPSSCFGFILE The following Bourne shell commands see sh 1 configure the preferred page sizes to a select set of applications with exec names that begin with foo using the MPSSCFGFILE environment variable The MPSS configuration file mpsscfg is assumed to have been previously created via a text editor like vi l The cat 1 command is only dumping out the con tents example LD_PRELOAD SLD_PRELOAD mpss so 1 example MPSSCFGFILE mpsscfg exampleS export LD_PRELOAD MPSSCFGFILE exampleS cat SMPSSCFGFILE foo 512K 64K Once the application has been started pmap see proc l can be u
22. c itlb miss dtlb miss dtlb prot spill user 32 spill user 64 spill user 32 cln spill user 64 cln spill kern 64 spill asuser 32 spill asuser 64 spill asuser 32 cln spill asuser 64 cln fill user 32 fill user 64 fill user 32 cln fill user 64 cln fill kern 64 syscall 32 self xcall The c option can be trapstat is enabled through 15 example trapstat c 1 12 15 cleanwin level 4 level 9 level 13 continued on next page used to 15 0 2813 2636 90528 819 175768 0 681 0 52158 TIIS 0 3816 0 170744 0 8550 0 51968 1839 0 11 0 2833 1925 70639 675 39933 241348 513 42 40914 856 64 2942 0 38444 230381 3790 10726 40760 1495 0 Using trapset with CPU Filtering limit the 22 0 2738 3133 107786 988 2811 96907 153 16 62305 1251 16 4515 0 2876 92941 3612 4495 62053 2144 0 CPUs on 11 0 2714 3029 103425 954 2742 118298 730 20 60141 1208 24 4381 0 2784 111694 3553 5845 59922 2083 0 which This example limits CPU 1 and CPUs 12 PE ee ee ee ee ee D ATN ENEA EE RR TE 2 cpul4 cpuls 3518 2261 1 1 100 100 19 14 trapstat 1M 27 28 continued from preceding page continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 60 int vec 2559 2699 2752 2688 21 92 64 itlb miss 3296 1548 1174 1698 1087 68 dtlb miss 114788 54313 43040 5
23. d memory mapping The shared memory identifier that the mapping was created with is reported ism shmid n The mapping is an Intimate Shared Memory variant of System V shared memory ISM mappings are created with the SHM_SHARE_MMU flag set in accordance with shmat 2 see shmop 2 dism shmid n The mapping is a pageable variant of ISM Pageable ISM is created with the SHM_PAGEABLE flag set in accordance with shmat 2 see shmop 2 o Other Mappings of other objects including dev ices such as frame buffers No mapping name is shown for other mapped objects Page Size Pgsz The page size in kilobytes that is used for hardware address translation for this mapping See memcnt1 2 for further information continued on next page pmap 1 11 continued from preceding page Swap Space Swap The amount of swap space in kilobytes that is reserved for this mapping That is swap space that is deducted from the total available pool of reservable swap space that is displayed with the command swap s See swap 1M EXAMPLES Example 1 Displaying process mappings By default pmap prints one line for each mapping within the address space of the target process The following example displays the address space of a typical bourne shell exampleS pmap 102905 102905 sh 00010000 192K r x usr bin ksh 00040000 8K rwx usr bin ksh 00042000 40K rwx heap FF180000 664K r x usr lib libc
24. de the removed CPU At most one processor set can be specified e entrylist Enables trapstat only for the trap table entry or entries specified by entrylist A trap table entry can be specified by trap number or by trap name for exam ple the level 10 trap can be specified as 74 Ox4A Ox4a or level 10 entrylist can be a single trap table entry or a comma separated list of trap table entries If the specified trap table entry is not valid trapstat prints a table of all valid trap table entries and values A list of valid trap table entries is also found in The SPARC Architecture Manual Version 9 and the Sun Microelec tronics UltraSPARC I amp II User s Manual If the parsable option P is specified in addition to the e option the format of the data is as follows Field Contents 1 Timestamp nanoseconds since start 2 CPU ID 3 Trap number in hexadecimal 4 Trap name 5 Trap rate per interval Each field is separated with whitespace If the format is modified it will be modified by adding potentially new fields beginning with field 6 exant fields will remain unchanged Lists trap table entries By default a table is displayed containing all valid trap numbers their names and a brief description The trap name is used in both the default output and in the entrylist param eter for the e argument If the parsable option P is specified in addition to the l option the format of the data is as follows Fi
25. eld Contents al Trap number in hexadecimal 2 Trap number in decimal 3 Trap name Remaining Trap description continued on next page trapstat 1M 23 24 continued from preceding page P Generates parsable output When run without other data gathering modifying options that is e t or T trapstat s the parsable output has the following format Field Contents Timestamp nanoseconds since start 2 CPU ID 3 Trap number in hexadecimal 4 Trap name 5 Trap rate per interval Each field is separated with whitespace If the format is modified it will be modified by adding potentially new fields beginning with field 6 extant fields will remain unchanged r rate Explicitly sets the sampling rate to be rate samples per second If this option is specified trapstat s output changes from a traps per second to traps per sampling interval Enables TLB statistics A table is displayed with four principal columns of data itlb miss itsb miss dtlb miss and dtsb miss The columns contain both the rate of the corresponding event and the percentage of CPU time spent processing the event The percentage of CPU time is given only in terms of a single CPU The rows of the table correspond to CPUs with each CPU consuming two rows one row for user mode events denoted with u and one row for kernel mode events denoted with k For each row the percentage of CPU time is totalled and displayed in the righ
26. erface Stability Command Syntax Evolving Output Format s Unstable SEE ALSO ldd 1 mdb 1 proc l ps l swap 1M mmap 2 memcntl 2 shmop 2 dlopen 3DL proc 4 attributes 5 ppgsz 1 NAME ppgsz set preferred stack and or heap page size SYNOPSIS usr bin ppgsz F o option option cmd p pid DESCRIPTION The ppgsz utility sets the preferred stack and or heap page size for the target process es that is the launched cmd or the process es in the pid list ppgsz stops the target process es while changing the page size See memcntl 2 continued on next page ppgsz 1 17 18 continued from preceding page OPTIONS The following options are supported F Force Sets the preferred page size options s for target process es even if controlled by other process es Caution should be exercised when using the F flag See proc 1 p pid Sets the preferred page size option s for the target process es in the process id pid list following the p option The pid list can also consist of names in the proc directory Only the process owner or the super user is permitted to set page size cmd is interpreted if p is not specified ppgsz launches cmd and applies page size option s to the new process The heap and stack preferred page sizes are inherited Child process es created see fork 2 from the launched process or the target process es
27. etting preferred page size SYNOPSIS mpss so l DESCRIPTION The mpss so 1 shared object provides a means by which the preferred stack and or heap page size can be selectively configured for launched processes and their descendants To enable mpss so l the following string needs to be present in the environment see ld so 1 1 along with one or more MPSS Multiple Page Size Support environment variables continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page LD_PRELOAD SLD_PRELOAD mpss so 1 ENVIRONMENT VARIABLES Once preloaded the mpss so 1 shared object reads the fol lowing environment variables to determine any preferred page size requirements and any processes these may be specific to MPSSHEAP size MPSSSTACK size MPSSHEAP and MPSSSTACK specify the preferred page sizes for the heap and stack respectively The speci fied page size s are applied to all created processes size must be a supported page size see pagesize l or 0 in which case the system will select an appropriate page size see memcntl 2 size can be qualified with K M G or T to specify Kilobytes Megabytes Gigabytes or Terabytes respec tively MPSSCFGFILE config file config file is a text file which contains one or more mpss configuration entries of the form exec name heap size stack size exec name specifies the name of an applic
28. fg2 exampleS export MPSSCFGFILE exampleS cat SMPSSCFGFILE ES ses ora 4m 4m usr lib 1d map bssalign A template link editor mapfile for aligning bss see NOTES BUTES See attributes 5 for descriptions of the following attri butes ATTRIBUTE TYPE ATTRIBUTE VALUE Availability SUNWesu 32 bit SUNWesxu 64 bit Interface Stability Evolving inued on next page mpss so 1 1 35 continued from preceding page SEE ALSO cat 1 ld 1 ld so 1 1 pagesize 1 ppgsz l proc 1 sh 1 vi l exec 2 fork 2 memcntl 2 getexecname 3C getpagesize 3C syslog 3C proc 4 attributes 5 NOTES The heap and stack preferred page sizes are inherited A child process has the same preferred page sizes as its parent On exec 2 the preferred page sizes are set back to the default system page size unless a preferred page size has been configured via the mpss shared object ppgsz l a proc tool can also be used to set the preferred stack and or heap page sizes It cannot selectively config ure the page size for descendents based on name matches See also NOTES under ppgsz 1 36 mementl 2 NAME memcntl memory management control SYNOPSIS include lt sys types h gt include lt sys mman h gt int memcntl caddr_t addr size_t len int cmd caddr_t arg int attr int mask DESCRIPTION The memcntl function allows the calling process to apply a variety of c
29. for the system as a whole For example running trapstat with the default options on a system that spends 7 of total time handling traps induces a performance degradation of less than one half of one percent running trapstat with the t or T option on a system spending 5 of total time processing TLB misses induce a performance degradation of no more than PR inued on next page trapstat 1M 31 continued from preceding page When run with the t or T option trapstat accounts for its probe effect when calculating the tim fields This assures that the Stim fields are a reasonably accurate indicator of the time a given workload is spending handling TLB misses regardless of the perturbing presence of trapstat While the Stim fields include the explicit cost of executing the TLB miss handler they do not include the implicit costs of TLB miss traps for example pipeline effects cache pol lution etc These implicit costs become more significant as the trap rate grows if high tim values are reported greater than 50 you can accurately infer that much of the balance of time is being spent on the implicit costs of the TLB miss traps Due to the potential system wide degradation induced only the super user can run trapstat Due to the limitation of the underlying statistics gathering methodology only one instance of trapstat can run at a time 32 mpss so 1 1 NAME mpss so l shared object for s
30. he value of arg determines whether the pages to be locked are those currently mapped by the address space those that will be mapped in the future or both If MCL_FUTURE is specified then all mappings subsequently added to the address space will be locked provided sufficient memory is available continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page MCL_CURRENT Lock current mappings MCL_FUTURE Lock future mappings MC_SYNC Write to their backing storage locations all modified pages in the range with attributes attr Optionally invalidate cache copies The backing storage for a modified MAP_SHARED mapping is the file the page is mapped to the backing storage for a modified MAP_PRIVATE mapping is its swap area The arg argument is a bit pattern built from the flags used to control the behavior of the operation S_ASYNC Perform asynchronous writes S_SYNC Perform synchronous writes S_INVALIDATE Invalidate mappings MS_ASYNC Return immediately once all write operations are scheduled with MS_SYNC the function will not return until all write operations are completed MS_INVALIDATE Invalidate all cached copies of data in memory so that further references to the pages will be obtained by the system from their backing storage locations This operation should be used by applica tions that require a memory object to be
31. in a known state MC_UNLOCK Unlock all pages in the range with attributes attr The arg argument is not used but must be 0 to ensure compatibility with potential future enhancements MC_UNLOCKAS Remove address space memory locks and locks on all pages in the address space with attributes attr The addr len and arg arguments are not used but must be NULL 0 and 0 respectively to ensure compatibility with potential future enhancements continued on next page memenil 2 39 continued from preceding page MC_HAT_ADVISE Advise system how a region of user mapped memory will be accessed The arg argument is interpreted as a struct memcntl_mha The following members are defined in a struct memcntl_mha System Calls memcntl 2 uint_t mha_cmd uint_t mha_flags size_t mha pagesize The accepted values for mha_ cmd are MHA_MAPSIZE VA MHA_MAPSIZE_STACK MHA_MAPSIZE_BSSBRK he mha_flags member is reserved for future use and ust always be set to 0 The mha_pagesize member must e a valid size as obtained from getpagesizes 3C or he constant value 0 to allow the system to choose an ppropriate hardware address translation mapping size wt 3 H MHA_MAPSIZE VA sets the preferred hardware address translation mapping size of the region of memory from addr to addr len Both addr and len must be aligned to an mha_pagesize boundary The entire virtual address region from addr to addr len must not have any holes
32. ing of a pipeline continued on next page cpustat 1M 5 6 continued from preceding page t Prints an additional column of processor cycle counts if available on the current architecture USAGE A closely related utility cputrack 1 can be used to moni tor the behavior of individual applications with little or no interference from other activities on the system The cpustat utility must be run by the super user as there is an intrinsic conflict between the use of the CPU perfor mance counters system wide by cpustat and the use of the CPU performance counters to monitor an individual process for example by cputrack Once any instance of this utility has started no further per process or per LWP use of the counters is allowed until the last instance of the utility terminates he times printed by the command correspond to the wallclock nstead of when the program told the kernel to sample them T time when the hardware counters were actually sampled i The time is derived from the same timebase as gethrtime 3C The processor cycle counts enabled by the t option always apply to both user and system modes regardless of the set tings applied to the performance counter registers On some hardware platforms the counters are implemented using 32 bit registers While the kernel attempts to catch all overflows to synthesize 64 bit counters because of hardware implementation restrictions overflows ma
33. ions for shared mappings F Force Grabs the target process even if another pro cess has control 1 Shows unresolved dynamic linker map names iG Prints the process s reserved addresses s Prints HAT page size information USAGE The pmap utility prints information about the address space of a process Process Mappings usr bin pmap rslF pid core By default pmap displays the mappings in the virtual address order they are mapped into the process The mapping size flags and mapped object name are shown continued on next page 8 Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page Process anon locked mapping details usr bin pmap x aslF pid core The x option displays additional information per map ping The size of each mapping the amount of resident physical memory the amount of anonymous memory and the amount of memory locked is shown with this option This does not include anonymous memory taken by kernel address space due to this process Swap Reservations usr bin pmap S alF pid core The S option displays swap reservation information per mapping DISPLAY FORMATS One line of output is printed for each mapping within the process unless the s option is specified where one line is printed for a contiguous mapping of each hardware trans lation page size The column headings are shown in
34. l 10 1 14031200 0 4a level 10 0 15027200 0 4a level 10 0 16027600 0 4a level 10 0 17025000 0 4a level 10 0 18026000 0 4a level 10 0 19027800 0 4a level 10 0 20025600 0 4a level 10 0 21025200 0 4a level 10 0 22025000 0 4a level 10 0 23035400 0 4a level 10 1 24027400 0 4a level 10 0 25026000 0 4a level 10 0 26027000 0 4a level 10 0 continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page ATTRI BUTES See attributes 5 for descriptions of the following attri butes ATTRIBUTE TYPE ATTRIBUTE VALUE Availability SUNWcsu Interface Stability Human Readable Output Unstable Parsable Output Evolving SEE ALSO NOTES cont lockstat 1M pmap 1 psrset 1M psrinfo 1M pbind 1M ppgsz 1 getpagesizes 3C Sun Microelectronics UltraSPARC I amp II User s Manual January 1997 STP1031 The SPARC Architecture Manual Version 9 1994 Prentice Hall When enabled trapstat induces a varying probe effect depending on the type of information collected While the precise probe effect depends upon the specifics of the hardware the following table can be used as a rough guide Option Approximate probe effect default 3 5 per trap e 3 5 per specified trap Pey ST 40 45 per TLB miss trap hitting in the TSB 25 30 per TLB miss trap missing in the TSB These probe effects are per trap not
35. ln pill kern 64 pill asuser 32 pill asuser 64 Koj CO U U U U U U U NUR HA c4 fill user 32 c8 fill user 64 cc fill user 32 cln d0 fill user 64 cln d8 fill kern 64 108 syscall 32 126 self xcall continued on next page pill asuser 32 cln pill asuser 64 cln Without Options trapstat displays a table of trap six columns can fit in the default this example there are more than are displayed cpud cpul cpu4 cpus cpus cpu9 6446 4837 6368 2153 2623 1321 100 0 0 0 al 0 0 fl 1 1 0 0 0 0 0 0 0 0 0 0 0 0 9 0 100 100 100 100 100 100 100 0 0 0 0 0 6 10 7 16 LS rL 100 0 0 0 1 0 2607 2740 2642 2922 2920 3033 3129 2475 3167 1037 1200 569 121061 86162 109838 37386 45639 20269 997 847 1061 379 406 184 2809 2133 2739 200806 332776 454504 45819 207856 93487 228529 68373 77590 784 561 767 274 353 215 9 37 17 39 2 13 62913 50145 63869 21916 28431 11738 1327 947 1288 460 S72 335 26 48 18 54 0 14 4580 3599 4555 1538 1978 857 26 0 0 2 0 0 2862 2161 2798 191746 318115 435850 45813 197781 89179 217668 63905 74281 3802 2833 3723 10153 16419 19475 329 10105 4873 10603 4235 3649 62519 49943 63611 21824 28328 11693 2285 1634 2278 737 957 383 100 0 0 0 0 0 Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page d8 108 126 Example 2 cleanwin level 1 level 4 level 5 level 7 level 9 level 10 level 13 level 14 int ve
36. nce of applica tions that use large amounts of memory on processors that support multiple hardware address translation mapping sizes however it should be used with care Not all processors support all sizes with equal efficiency Use of larger sizes may also introduce extra overhead that could reduce perfor mance or available memory Using large sizes for one appli cation may reduce available resources for other applications and result in slower system wide performance RETURN VALUES Upon successful completion memcntl returns 0 otherwise it returns 1 and sets errno to indicate an error ERRORS The memcntl function will fail if EAGAIN hen the selection criteria match some or all of the emory identified by the operation could not be locked hen MC_LOCK or MC_LOCKAS was specified some or all appings in the address range addr addr len are ocked for I O when MC_HAT_ADVISE was specified or he system has insufficient resources when C_HAT_ADVISE was specified gactrsesB Ss EBUSY When the selection criteria match some or all of the addresses in the range addr addr len are locked and MC_SYNC with the MS_INVALIDATE option was speci fied continued on next page memenil 2 41 continued from preceding page EINVAL The addr argument specifies invalid selection criteria or is not a multiple of the page size as returned by sysconf 3C the addr and or len argument does not have the value 0 when
37. nd Solaris are trademarks or registered trademarks of Sun Microsystems Inc in the United States and other countries All SPARC trademarks are used under license and are trademarks or registered trademarks of SPARC International Inc in the US and other countries Products bearing SPARC trademarks are based upon an architecture developed by Sun Microsystems Inc The OPEN LOOK and Sun Graphical User Interface was developed by Sun Microsystems Inc for its users and licensees Sun acknowledges the pioneering efforts of Xerox in researching and developing the concept of visual or graphical user interfaces for the computer industry Sun holds a non exclusive license from Xerox to the Xerox Graphical User Interface which license also covers Sun s licensees who implement OPEN LOOK GUIs and otherwise comply with Sun s written license agreements U S Government Rights Commercial use Government users are subject to the Sun Microsystems Inc standard license agreement and applicable provisions of the Far and its supplements DOCUMENTATION IS PROVIDED AS IS AND ALL EXPRESS OR IMPLIED CONDITIONS REPRESENTATIONS AND WARRANTIES INCLUDING ANY IMPLIED WARRANTY OF MERCHANTABILITY FITNESS FOR A PARTICULAR PURPOSE OR NON INFRINGEMENT ARE DISCLAIMED EXCEPT TO THE EXTENT THAT SUCH DISCLAIMERS ARE HELD TO BE LEGALLY INVALID Copyright 2004 Sun Microsystems Inc 4150 Network Circle Santa Clara Californie 95045 Etats Unis Tous droits r serv s
38. no 4628487 anon anon dism shmid 0x5 dism shmid 0x4 ism shmid 0x2 ism shmid 0x3 1166 801 libc so l libc_psr so 1 libdl so 1 anon ld so 1 ld so 1 stack Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page Example 3 Displaying Page Size Information The s option can be used to display the hardware transla tion page sizes for each portion of the address space See memcntl 2 for futher information on Solaris multiple page size support In the example below we can see that the majority of the mappings are using an 8K Byte page size while the heap is using a 4M Byte page size Notice that non contiguous regions of resident pages of the same page size are reported as separate mappings In the example below the libc so library is reported as separate mappings since only some of the libc so text is resident example pmap xs 15492 15492 maps Address Kbytes RSS Anon Locked Pgsz Mode Mapped File 00010000 8 8 8K r x maps 00020000 8 8 8 8K rwx maps 00022000 3960 3960 3960 E 8K rwx heap 00400000 8192 8192 8192 a 4M rwx heap 00C00000 4096 a LWX heap 01000000 4096 4096 4096 4M rwx heap 03000000 1024 1024 S 8K rw s dev 0 2 ino 4628487 04000000 512 512 512 A 8K rw dev 0 2 ino 4628487 04080000 512 512 D rw dev 0 2 ino 4628487 05000000 512 51
39. ontrol operations over the address space identi fied by the mappings established for the address range addr addr len continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page The addr argument must be a multiple of the pagesize as returned by sysconf 3C The scope of the control operations can be further defined with additional selection criteria in the form of attributes according to the bit pattern contained in attr The following attributes specify page mapping selection criteria SHARED Page is mapped shared PRIVATE Page is mapped private The following attributes specify page protection selection criteria The selection criteria are constructed by a bit wise OR operation on the attribute bits and must match exactly PROT_READ Page can be read PROT_WRITE Page can be written PROT_EXEC Page can be executed The following criteria may also be specified PROC_TEXT Process text PROC_DATA Process data System Calls memcnt1 2 The PROC_TEXT attribute specifies all privately mapped seg ments with read and execute permission and the PROC_DATA attribute specifies all privately mapped segments with write permission continued on next page memenitl 2 37 38 continued from preceding page Selection criteria can be used to describe various abstract memory objects within the
40. ount parameters respec tively The interval is specified in seconds the count indicates the number of intervals to be executed before exiting Alternatively command can be specified in which case trapstat executes the provided command and continues to run until the command exits A positive integer is assumed to be an interval if the desired command cannot be dis tinguished from an integer the full path of command must be specified continued on next page trapstat 1M 21 22 continued from preceding page UltraSPARC I II and III handle translation lookaside buffer TLB misses by trapping to the operating system TLB miss traps can be a significant component of overall system per formance for some workloads the t option provides in depth information on these traps When run with this option trapstat displays both the rate of TLB miss traps and the percentage of time spent processing those traps Addition ally TLB misses that hit in the translation storage buffer TSB are differentiated from TLB misses that further miss in the TSB The TSB is a software structure used as a translation entry cache to allow the TLB to be quickly filled it is discussed in detail in the UltraSPARC I amp II User s Manual The TLB and TSB miss information is further broken down into user and kernel mode misses Workloads with working sets that exceed the TLB reach may spend a significant amount of time missing in the TLB
41. penmp i xP xpagesize n xpagesize_heap n xpagesize_stack n xparallel xpch v xpchstop xpentium xpg xprefetch val val xprefetch_level 1 xprofile p xprofile_ircache path xprofile pathmap collect _prefix use prefix xreduc tion xregs r r xrestrict f xs xsafe mem xsb xsbfast xsfpconst xspace xstrconst xtarget t xtemp dir xthreadvar o o0 xtime xtransition xtrigraphs yes no xunroll n xustr ascii_utfl6_ushort no xvector yes no xvis xvpara Yc dir YA dir YI dir YP dir YS dir Z11 2 Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page xpagesize n SPARC Set the preferred page size for the stack and the heap The n value must be one of the following 8K 64K 512K 4M 32M 256M 2G 16G or default You must specify a valid page size for the Solaris operating environment on the target platform as returned by getpagesize 3C If you do not specify a valid pagesize the request is silently ignored at run time The Solaris environment offers no guarantee that the page size request will be honored You can use pmap l or meminfo 2 to determine page size of the target platform The xpagesize option has no effect unless you use it at compile time and at link time Note that this feature is not available on Solaris 7 and 8
42. rough this program Accessing Sun Documentation Online The docs sun com web site enables you to access Sun technical documentation online You can browse the docs sun com archive or search for a specific book title or subject The URL is http docs sun com To reference Sun BluePrints OnLine articles visit the Sun BluePrints OnLine Web site at http www sun com blueprints online html About the Author 43
43. sed to view the actual page sizes configured exampleS foobar amp example pmap s pgrep foobar If the desired page size is not configured shown in the pmap output it may be due to errors in the MPSS configura tion file or environment variables Check the error log by default var adm messages for errors If no errors can be found resource or alignment constraints may be responsible See the NOTES section continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 cont FILES ATTRI cont inued from preceding page Example 2 Configuring preferred page sizes using MPSSHEAP and MPSSSTACK The following Bourne shell commands configure 512K heap and 64K stack preferred page sizes for all applications using the MPSSHEAP and MPSSSTACK environment variables example LD_PRELOAD S LD_PRELOAD mpss so 1 exampleS MPSSHEAP 512K exampleS MPSSSTACK 64K exampleS export LD_PRELOAD MPSSHEAP MPSSSTACK Example 3 Precedence rules continuation from Example 2 The preferred page size configuration in MPSSCFGFILE over rides MPSSHEAP and MPSSTACK Appending the following com mands to those in Example 2 would mean that all applications will be configured with 512K heap and 64K stack preferred page sizes with the exception of those applications the ls command and all applications beginning with ora in the configuration file example MPSSCFGFILE mpssc
44. size_heap default Compiling with this option has the same effect as set ting the LD_PRELOAD environment variable to mpss so l with the equivalent options or running the Solaris 9 command ppgsz 1 with the equivalent options before running the program See the Solaris 9 man pages for details Note that this feature is not available on Solaris 7 and 8 A program compiled with this option will not link on Solaris 7 and 8 xpagesize_stack n SPARC Set the page size in memory for the stack n can be 8K 64K 512K 4M 32M 256M 2G 16G default You must specify a valid page size for the Solaris operating environment on the target platform as returned by getpagesize 3C If you do not specify a valid page size the request is silently ignored at run time You can use pmap l or meminfo 2 to determine page size at the target platform If you specify xpagesize_stack default the Solaris environment sets the page size xpagesize_stack without an argument is the equivalent to xpagesize_stack default Compiling with this option has the same effect as set ting the LD_PRELOAD environment variable to mpss so 1 with the equivalent options or running the Solaris 9 command ppgsz 1 with the equivalent options before running the program See the Solaris 9 man pages for details Note that this feature is not available on Solaris 7 and 8 A program compiled with this option will not link on Solaris 7 and 8 4 Supporting M
45. so 1 FF236000 24K rwx usr lib libc so 1 FF23C000 8K rwx usr lib libc so 1 FF250000 8K rwx anon FF260000 16K r x usr lib en_US 1IS08859 1 s0 2 FF272000 16K rwx usr lib en_US 1IS08859 1 s0 2 FF280000 560K r x usr lib libnsl so 1 FF31C000 32K rwx usr lib libnsl so 1 FF324000 32K rwx usr lib libnsl so 1 FF340000 16K r x usr lib libc_psr so 1 FF350000 16K r x usr lib libmp so 2 FF364000 8K rwx usr lib libmp so 2 FF380000 40K r x usr lib libsocket so 1 FF39A000 8K rwx usr lib libsocket so 1 FF3A0000 8K r x usr lib libdl so 1 FF3B0000 8K rwx anon FF3C0000 152K r x usr lib ld so 1 FF3F6000 8K rwx usr lib ld so 1 FFBFC000 16K rw stack total 1880K Example 2 Displaying memory allocation and mapping types The x option can be used to provide information about the memory allocation and mapping types per mapping The amount of resident non shared anonymous and locked memory is shown for each mapping continued on next page 12 Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page exampleS pmap x 102908 102908 sh Address Kbytes 00010000 88 00036000 8 00038000 16 FF260000 16 FF272000 16 FF280000 664 FF336000 32 FF360000 16 FF380000 24 FF396000 8 FF3A0000 8 FF3B0000 8 FF3C0000 152 FF3F6000 8 FFBFEOOO 8 total Kb 1072 RSS 88 8 16 16 16 624 32 16 24 Anon Locked
46. t licenci s Sun reconna t les efforts de pionniers de Xerox pour la recherche et le d veloppement du concept des interfaces d utilisation visuelle ou graphique pour l industrie de l informatique Sun d tient une licence non exclusive de Xerox sur l interface d utilisation graphique Xerox cette licence couvrant galement les licenci s de Sun qui mettent en place l interface d utilisation graphique OPEN LOOK et qui en outre se conforment aux licences crites de Sun CETTE PUBLICATION EST FOURNIE EN L ETAT ET AUCUNE GARANTIE EXPRESSE OU IMPLICITE N EST ACCORDEE Y COMPRIS DES GARANTIES CONCERNANT LA VALEUR MARCHANDE L APTITUDE DE LA PUBLICATION A REPONDRE A UNE UTILISATION PARTICULIERE OU LE FAIT QU ELLE NE SOIT PAS CONTREFAISANTE DE PRODUIT DE TIERS CE DENI DE GARANTIE NE S APPLIQUERAIT PAS DANS LA MESURE OU IL SERAIT TENU JURIDIQUEMENT NUL ET NON AVENU amp a Please Adobe PostScript Recycle Supporting Multiple Page Sizes in the Solaris Operating System Appendix This article is a quick reference to the Solaris Operating System manual sections for the commands and library interfaces referenced the SunBluePrints article Supporting Multiple Page Sizes in the Solaris Operating System Reference this article for information about using the following new tools to perform tasks described in the referenced article m cc 1 on page 2 m cpustat 1M on page 5 m pmap 1 on page 8 m ppgsz 1
47. tion minimizes the probe effect when seeking specific data This example yields statistics for only the dtlb prot and syscall 32 traps on CPUs 12 through 15 example trapstat e dtlb prot syscall 32 c 12 15 vet name cpul2 cpul3 cpul4 cpul5 pa Gg Ae A ME IE RE EP LR LP GE a SD 6c dtlb prot 817 754 1018 560 108 syscall 32 1426 1647 2186 1142 vct name cpul2 cpul3 cpul4 cpul5 PRE ae ee ER RER IE AE SE E ee ee ee ee AURER 6c dtlb prot 1085 996 800 707 108 syscall 32 2578 2167 1638 1452 continued on next page trapstat 1M 29 30 continued from preceding page Example 6 Using trapstat with a Higher Sampling Rate The following example uses the r option to specify a sam pling rate of 1000 samples per second and filter only for the level 10 trap Additionally specifying the P option yields parsable output Notice the timestamp difference between the level 10 events 9 998 000 nanoseconds and 10 007 000 nanoseconds These level 10 events correspond to the system clock which by default ticks at 100 hertz that is every 10 000 000 nanoseconds example trapstat e level 10 P r 1000 1070400 0 4a level 10 0 2048600 0 4a level 10 0 3030400 0 4a level 10 1 4035800 0 4a level 10 0 5027200 0 4a level 10 0 6027200 0 4a level 10 0 7027400 0 4a level 10 0 8028200 0 4a level 10 0 9026400 0 4a level 10 0 10029600 0 4a level 10 0 11028600 0 4a level 10 0 12024000 0 4a level 10 0 13028400 0 4a leve
48. tmost column The CPUs are del ineated with a solid line If the parsable option P is specified in addition to the t option the format of the data is as follows Field Contents af Timestamp nanoseconds since start 2 CPU ID 3 Mode k denotes kernel u denotes user 4 I TLB misses 5 Percentage of time in I TLB miss handler 6 I TSB misses 7 Percentage of time in I TSB miss handler 8 D TLB misses 9 Percentage of time in D TLB miss handler 10 D TSB misses 11 Percentage of time in D TSB miss handler continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page Each field is separated with whitespace If the format is modified it will be modified by adding potentially new fields beginning with field 12 extant fields will remain unchanged Enables TLB statistics with page size information As with the t option a table is displayed with four principal columns of data itlb miss itsb miss dtlb miss and dtsb miss The columns contain both absolute number of the corresponding event and percentage of CPU time spent processing the event the the The percentage of CPU time is given only in terms of a single CPU The rows of the table correspond to CPUs with each CPU consuming two sets of rows one set user level events denoted with u and one set kernel level events denoted with k Each set for for in
49. turn contains as many rows as there are page sizes supported see getpagesizes 3C For each row the percentage of CPU time is totalled and displayed in the right most column The two sets are delineated with a dashed line CPUs are delineated with a solid line If the parsable option P is specified in addition to the T option the format of the data is as follows Field Contents 1 Timestamp nanoseconds since start 2 CPU ID 3 Mode k denotes kernel u denotes user 4 Page size in decimal 5 I TLB misses 6 Percentage of time in I TLB miss handler 7 I TSB misses 8 Percentage of time in I TSB miss handler 9 D TLB misses 10 Percentage of time in D TLB miss handler 11 D TSB misses 12 Percentage of time in D TSB miss handler Each field is separated with whitespace If the format is modified it will be modified by adding potentially new fields beginning with field 13 extant fields will remain unchanged continued on next page trapstat 1M 25 26 continued from preceding page EXAMPLES Example 1 Using trapstat When run without options types and CPUs At most terminal width if as in six CPUs multiple tables example trapstat vct name 24 cleanwin 41 level 1 44 level 4 45 level 5 47 level 7 49 level 9 4a level 10 4d level 13 4e level 14 60 int vec 64 itlb miss 68 dtlb miss 6c dtlb prot pill user 32 pill user 64 pill user 32 cln pill user 64 c
50. ultiple Page Sizes in the Solaris Operating System Appendix March 2004 cpustat 1M NAME cpustat monitor system behavior using CPU performance counters SYNOPSIS cpustat c eventspec c eventspec ntD interval count cpustat h DESCRIPTION The cpustat utility allows CPU performance counters to be used to monitor the overall behavior of the CPUs in the system If interval is specified cpustat samples activity every interval seconds repeating forever If a count is speci fied the statistics are repeated count times If neither are specified an interval of five seconds is used and there is no limit to the number of samples that will be taken OPTIONS The following options are supported c eventspec Specifies a set of events for the CPU performance counters to monitor The syntax of these event specif ication can be determined using the h option to cause the usage message to be generated The semantics of these event specifications can be determined by reading the CPU manufacturers documentation for the events See cpc_strtoevent 3CPC for description of the syntax Multiple c options may be specified in which case the command cycles between the different event set tings on each sample D Enables debug mode h Prints an extensive help message on how to use the utility and how to program the processor dependent counters f Omits all header output useful if cpustat is the beginn
51. y be lost unless the sampling interval is kept short enough The events most prone to wrap are those that count processor clock cycles If such an event is of interest sampling should occur frequently so that less than 4 billion clock cycles can occur between samples The output of cpustat is designed to be readily parseable by nawk 1 and perl 1 thereby allowing performance tools to be composed by embedding cpustat in scripts Alternatively tools may be constructed directly using the same APIs that cpustat is built upon using the facilities of libcpc 3LIB See cpc 3CPC The cpustat utility only monitors the CPUs that are accessi ble to it in the current processor set Thus several instances of the utility can be running on the CPUs in dif ferent processor sets See psrset 1M for more information about processor sets continued on next page Supporting Multiple Page Sizes in the Solaris Operating System Appendix March 2004 continued from preceding page Because cpustat uses LWPs bound to CPUs the utility may have to be terminated before the configuration of the relevant processor can be changed WARNINGS By running the cpustat command the super user will forcibly invalidate all existing performance counter context This may in turn cause all invocations of the cputrack command and other users of performance counter context to exit prematurely with unspecified errors If cpustat is invoked on a s
52. ystem that has CPU performance counters but on which the packages containing the kernel support for those counters is not installed the following message appears cpustat CPU performance counters are inaccessible on this machine This error message implies that cpc_access has failed and is documented in cpc_access 3CPC Review this documentation for more information about the problem and possible solu tions ATTRIBUTES See attributes 5 for descriptions of the following attri butes ATTRIBUTE TYPE ATTRIBUTE VALUE Availability SUNWcpcu Interface Stability Evolving SEE ALSO cputrack 1 nawk 1 perl 1 iostat 1M prstat 1M psrset 1M vmstat 1M cpc 3CPC cpc_access 3CPC cpc_strtoevent 3CPC gethrtime 3C libcpc 3LIB attri butes 5 Sun Microelectronics UltraSPARC I amp II User s Manual January 1997 STP1031 http www sun com sparc Intel Architecture Software Developer s Manual Volume 3 System Programmers Guide 243192 http developer intel com cpustat 1M 7 pmap 1 NAME pmap display information about the address space of a pro cess SYNOPSIS usr bin pmap rslF pid core usr bin pmap x aslF pid core usr bin pmap S alF pid core DESCRIPTION The pmap utility prints information about the address space of a process OPTIONS The following options are supported a Prints anonymous and swap reservat
Download Pdf Manuals
Related Search