Searched +hist:1.125 +hist:2.8 (Results 1 - 25 of 33) sorted by relevance
| /src/doc/ | ||
| H A D | CHANGES.prev | 1.125 Tue Jan 27 18:20:56 GMT 2015 tron Postfix 2.11.3 will ship with NetBSD 7.0. 1.112 Sun Aug 12 22:20:32 GMT 2012 tron Postfix 2.8.12 will be in NetBSD 6.0. 1.108 Tue Jun 19 10:05:53 GMT 2012 tron Postfix 2.8.11 will be in NetBSD 6.0. 1.102 Sun Feb 19 18:52:06 GMT 2012 tron Postfix 2.8.8 will be part of the NetBSD 6.0 release. 1.125 Tue Jan 27 18:20:56 GMT 2015 tron Postfix 2.11.3 will ship with NetBSD 7.0. 1.112 Sun Aug 12 22:20:32 GMT 2012 tron Postfix 2.8.12 will be in NetBSD 6.0. 1.108 Tue Jun 19 10:05:53 GMT 2012 tron Postfix 2.8.11 will be in NetBSD 6.0. 1.102 Sun Feb 19 18:52:06 GMT 2012 tron Postfix 2.8.8 will be part of the NetBSD 6.0 release. |
| /src/sys/compat/ultrix/ | ||
| H A D | ultrix_misc.c | 1.125 Fri Aug 10 21:44:59 GMT 2018 pgoyette branches: 1.125.4; Allow syscall_establish() to install new syscalls when the existing entry-point is either sys_nomodule or sys_nosys. Update the makesyscalls.sh script to create a const array of bits to allow syscall_disestablish() to properly restore the original entry-point. Update all the initializers of struct emul to initialize the pointer to the bit array struct emul. XXX Regen of all files created by makesyscalls.sh will come soon, XXX followed by a kernel version bump (since struct emul is being XXX modified). This commit should address PR kern/45781 and also removes the need for the work-around for that PR in file sys/arch/usermode/modules/syscallemu/syscallemu.c 1.125 Fri Aug 10 21:44:59 GMT 2018 pgoyette branches: 1.125.4; Allow syscall_establish() to install new syscalls when the existing entry-point is either sys_nomodule or sys_nosys. Update the makesyscalls.sh script to create a const array of bits to allow syscall_disestablish() to properly restore the original entry-point. Update all the initializers of struct emul to initialize the pointer to the bit array struct emul. XXX Regen of all files created by makesyscalls.sh will come soon, XXX followed by a kernel version bump (since struct emul is being XXX modified). This commit should address PR kern/45781 and also removes the need for the work-around for that PR in file sys/arch/usermode/modules/syscallemu/syscallemu.c 1.30 Sun Apr 06 23:26:53 GMT 1997 jonathan Add changes to make vic-2.8 Ultrix binaries work on NetBSD with COMPAT_ULTRIX: * Add IPmulticast setsockopt emulation. * Add Ultrix shmsys emulation (untested). * tidy up use of stackgap. |
| /src/sys/arch/i386/conf/ | ||
| H A D | Makefile.i386 | 1.125 Fri Nov 22 15:23:36 GMT 2002 fvdl New interrupt code. The basic idea behind it is to hide the differences in interrupt controllers in struct pic, and try to keep as much common code as possible. At the lowest (asm) level, this is done with CPP macros. The main structure is now struct intrsource, describing an established interrupt line, of any kind (soft/hard local apic/legacy apic/IO apic). For quick masking, there may be a maximum of 32 sources per CPU. Sources can be assigned to any CPU in the MP case, though currently they all go to the boot CPU. 1.90 Sun Apr 12 23:47:43 GMT 1998 tv Add -Wno-main conditional on compiler being gcc 2.8 or egcs. (This adds a HAVE_GCC28 check-variable that can now be used to add other gcc-2.8 flags in cases where they may be useful, or to remove gcc 2.7.2 "bug workaround" flags.) 1.90 Sun Apr 12 23:47:43 GMT 1998 tv Add -Wno-main conditional on compiler being gcc 2.8 or egcs. (This adds a HAVE_GCC28 check-variable that can now be used to add other gcc-2.8 flags in cases where they may be useful, or to remove gcc 2.7.2 "bug workaround" flags.) |
| H A D | INSTALL | 1.302 Tue Jan 02 18:05:19 GMT 2007 dsl Comment out 'bnx', a network card that requires a 120kb download isn't going to be the only way to get data onto a system. This should get the INSTALL image way back under 2.8M again. Someone does need to sort out an ACPI install kernel though.... 1.125 Thu Sep 23 16:09:13 GMT 1999 tron Add Realtek 8129/8139 driver to install kernel as suggested by Patrick Welche in PR install/8477. |
| /src/usr.bin/xinstall/ | ||
| H A D | xinstall.c | 1.125 Tue May 31 06:55:02 GMT 2016 pgoyette Move __MKTEMP_OK up earlier so it has an actual impact. Reduces the number of warnings during build. XXX There are still some other warnings remaining to be resolved. Fixes PR bin/48195 although we really should go back someday and fix this correctly (by replacing all uses of mktemp(3)!) 1.37 Tue Jul 06 14:45:31 GMT 1999 christos pacify gcc-2.8 uninitialized variable warnings, and only use timespecs in struct stat on BSD4_4 systems. |
| /src/sys/ufs/ffs/ | ||
| H A D | ffs_alloc.c | 1.125 Sun Feb 21 13:55:58 GMT 2010 mlelstv branches: 1.125.2; 1.125.4; 1.125.6; For the UVM_PAGE_TRKOWN test do not require that the relevant pages must exist. 1.125 Sun Feb 21 13:55:58 GMT 2010 mlelstv branches: 1.125.2; 1.125.4; 1.125.6; For the UVM_PAGE_TRKOWN test do not require that the relevant pages must exist. 1.125 Sun Feb 21 13:55:58 GMT 2010 mlelstv branches: 1.125.2; 1.125.4; 1.125.6; For the UVM_PAGE_TRKOWN test do not require that the relevant pages must exist. 1.125 Sun Feb 21 13:55:58 GMT 2010 mlelstv branches: 1.125.2; 1.125.4; 1.125.6; For the UVM_PAGE_TRKOWN test do not require that the relevant pages must exist. 1.50 Thu Sep 06 02:16:01 GMT 2001 lukem branches: 1.50.2; Incorporate the enhanced ffs_dirpref() by Grigoriy Orlov, as found in FreeBSD (three commits; the initial work, man page updates, and a fix to ffs_reload()), with the following differences: - Be consistent between newfs(8) and tunefs(8) as to the options which set and control the tuning parameters for this work (avgfilesize & avgfpdir) - Use u_int16_t instead of u_int8_t to keep track of the number of contiguous directories (suggested by Chuck Silvers) - Work within our FFS_EI framework - Ensure that fs->fs_maxclusters and fs->fs_contigdirs don't point to the same area of memory The new algorithm has a marked performance increase, especially when performing tasks such as untarring pkgsrc.tar.gz, etc. The original FreeBSD commit messages are attached: ===== mckusick 2001/04/10 01:39:00 PDT Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>. His description of the problem and solution follow. My own tests show speedups on typical filesystem intensive workloads of 5% to 12% which is very impressive considering the small amount of code change involved. ------ One day I noticed that some file operations run much faster on small file systems then on big ones. I've looked at the ffs algorithms, thought about them, and redesigned the dirpref algorithm. First I want to describe the results of my tests. These results are old and I have improved the algorithm after these tests were done. Nevertheless they show how big the perfomance speedup may be. I have done two file/directory intensive tests on a two OpenBSD systems with old and new dirpref algorithm. The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports". The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release. It contains 6596 directories and 13868 files. The test systems are: 1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for test is at wd1. Size of test file system is 8 Gb, number of cg=991, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=35 2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system at wd0, file system for test is at wd1. Size of test file system is 40 Gb, number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50 You can get more info about the test systems and methods at: http://www.ptci.ru/gluk/dirpref/old/dirpref.html Test Results tar -xzf ports.tar.gz rm -rf ports mode old dirpref new dirpref speedup old dirprefnew dirpref speedup First system normal 667 472 1.41 477 331 1.44 async 285 144 1.98 130 14 9.29 sync 768 616 1.25 477 334 1.43 softdep 413 252 1.64 241 38 6.34 Second system normal 329 81 4.06 263.5 93.5 2.81 async 302 25.7 11.75 112 2.26 49.56 sync 281 57.0 4.93 263 90.5 2.9 softdep 341 40.6 8.4 284 4.76 59.66 "old dirpref" and "new dirpref" columns give a test time in seconds. speedup - speed increasement in times, ie. old dirpref / new dirpref. ------ Algorithm description The old dirpref algorithm is described in comments: /* * Find a cylinder to place a directory. * * The policy implemented by this algorithm is to select from * among those cylinder groups with above the average number of * free inodes, the one with the smallest number of directories. */ A new directory is allocated in a different cylinder groups than its parent directory resulting in a directory tree that is spreaded across all the cylinder groups. This spreading out results in a non-optimal access to the directories and files. When we have a small filesystem it is not a problem but when the filesystem is big then perfomance degradation becomes very apparent. What I mean by a big file system ? 1. A big filesystem is a filesystem which occupy 20-30 or more percent of total drive space, i.e. first and last cylinder are physically located relatively far from each other. 2. It has a relatively large number of cylinder groups, for example more cylinder groups than 50% of the buffers in the buffer cache. The first results in long access times, while the second results in many buffers being used by metadata operations. Such operations use cylinder group blocks and on-disk inode blocks. The cylinder group block (fs->fs_cblkno) contains struct cg, inode and block bit maps. It is 2k in size for the default filesystem parameters. If new and parent directories are located in different cylinder groups then the system performs more input/output operations and uses more buffers. On filesystems with many cylinder groups, lots of cache buffers are used for metadata operations. My solution for this problem is very simple. I allocate many directories in one cylinder group. I also do some things, so that the new allocation method does not cause excessive fragmentation and all directory inodes will not be located at a location far from its file's inodes and data. The algorithm is: /* * Find a cylinder group to place a directory. * * The policy implemented by this algorithm is to allocate a * directory inode in the same cylinder group as its parent * directory, but also to reserve space for its files inodes * and data. Restrict the number of directories which may be * allocated one after another in the same cylinder group * without intervening allocation of files. * * If we allocate a first level directory then force allocation * in another cylinder group. */ My early versions of dirpref give me a good results for a wide range of file operations and different filesystem capacities except one case: those applications that create their entire directory structure first and only later fill this structure with files. My solution for such and similar cases is to limit a number of directories which may be created one after another in the same cylinder group without intervening file creations. For this purpose, I allocate an array of counters at mount time. This array is linked to the superblock fs->fs_contigdirs[cg]. Each time a directory is created the counter increases and each time a file is created the counter decreases. A 60Gb filesystem with 8mb/cg requires 10kb of memory for the counters array. The maxcontigdirs is a maximum number of directories which may be created without an intervening file creation. I found in my tests that the best performance occurs when I restrict the number of directories in one cylinder group such that all its files may be located in the same cylinder group. There may be some deterioration in performance if all the file inodes are in the same cylinder group as its containing directory, but their data partially resides in a different cylinder group. The maxcontigdirs value is calculated to try to prevent this condition. Since there is no way to know how many files and directories will be allocated later I added two optimization parameters in superblock/tunefs. They are: int32_t fs_avgfilesize; /* expected average file size */ int32_t fs_avgfpdir; /* expected # of files per directory */ These parameters have reasonable defaults but may be tweeked for special uses of a filesystem. They are only necessary in rare cases like better tuning a filesystem being used to store a squid cache. I have been using this algorithm for about 3 months. I have done a lot of testing on filesystems with different capacities, average filesize, average number of files per directory, and so on. I think this algorithm has no negative impact on filesystem perfomance. It works better than the default one in all cases. The new dirpref will greatly improve untarring/removing/coping of big directories, decrease load on cvs servers and much more. The new dirpref doesn't speedup a compilation process, but also doesn't slow it down. Obtained from: Grigoriy Orlov <gluk@ptci.ru> ===== ===== iedowse 2001/04/23 17:37:17 PDT Pre-dirpref versions of fsck may zero out the new superblock fields fs_contigdirs, fs_avgfilesize and fs_avgfpdir. This could cause panics if these fields were zeroed while a filesystem was mounted read-only, and then remounted read-write. Add code to ffs_reload() which copies the fs_contigdirs pointer from the previous superblock, and reinitialises fs_avgf* if necessary. Reviewed by: mckusick ===== ===== nik 2001/04/10 03:36:44 PDT Add information about the new options to newfs and tunefs which set the expected average file size and number of files per directory. Could do with some fleshing out. ===== |
| H A D | ffs_vfsops.c | 1.125 Wed Oct 15 11:29:01 GMT 2003 hannken Add the gating of system calls that cause modifications to the underlying file system. The function vfs_write_suspend stops all new write operations to a file system, allows any file system modifying system calls already in progress to complete, then sync's the file system to disk and returns. The function vfs_write_resume allows the suspended write operations to complete. From FreeBSD with slight modifications. Approved by: Frank van der Linden <fvdl@netbsd.org> 1.85 Thu Sep 06 02:16:02 GMT 2001 lukem branches: 1.85.2; Incorporate the enhanced ffs_dirpref() by Grigoriy Orlov, as found in FreeBSD (three commits; the initial work, man page updates, and a fix to ffs_reload()), with the following differences: - Be consistent between newfs(8) and tunefs(8) as to the options which set and control the tuning parameters for this work (avgfilesize & avgfpdir) - Use u_int16_t instead of u_int8_t to keep track of the number of contiguous directories (suggested by Chuck Silvers) - Work within our FFS_EI framework - Ensure that fs->fs_maxclusters and fs->fs_contigdirs don't point to the same area of memory The new algorithm has a marked performance increase, especially when performing tasks such as untarring pkgsrc.tar.gz, etc. The original FreeBSD commit messages are attached: ===== mckusick 2001/04/10 01:39:00 PDT Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>. His description of the problem and solution follow. My own tests show speedups on typical filesystem intensive workloads of 5% to 12% which is very impressive considering the small amount of code change involved. ------ One day I noticed that some file operations run much faster on small file systems then on big ones. I've looked at the ffs algorithms, thought about them, and redesigned the dirpref algorithm. First I want to describe the results of my tests. These results are old and I have improved the algorithm after these tests were done. Nevertheless they show how big the perfomance speedup may be. I have done two file/directory intensive tests on a two OpenBSD systems with old and new dirpref algorithm. The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports". The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release. It contains 6596 directories and 13868 files. The test systems are: 1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for test is at wd1. Size of test file system is 8 Gb, number of cg=991, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=35 2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system at wd0, file system for test is at wd1. Size of test file system is 40 Gb, number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50 You can get more info about the test systems and methods at: http://www.ptci.ru/gluk/dirpref/old/dirpref.html Test Results tar -xzf ports.tar.gz rm -rf ports mode old dirpref new dirpref speedup old dirprefnew dirpref speedup First system normal 667 472 1.41 477 331 1.44 async 285 144 1.98 130 14 9.29 sync 768 616 1.25 477 334 1.43 softdep 413 252 1.64 241 38 6.34 Second system normal 329 81 4.06 263.5 93.5 2.81 async 302 25.7 11.75 112 2.26 49.56 sync 281 57.0 4.93 263 90.5 2.9 softdep 341 40.6 8.4 284 4.76 59.66 "old dirpref" and "new dirpref" columns give a test time in seconds. speedup - speed increasement in times, ie. old dirpref / new dirpref. ------ Algorithm description The old dirpref algorithm is described in comments: /* * Find a cylinder to place a directory. * * The policy implemented by this algorithm is to select from * among those cylinder groups with above the average number of * free inodes, the one with the smallest number of directories. */ A new directory is allocated in a different cylinder groups than its parent directory resulting in a directory tree that is spreaded across all the cylinder groups. This spreading out results in a non-optimal access to the directories and files. When we have a small filesystem it is not a problem but when the filesystem is big then perfomance degradation becomes very apparent. What I mean by a big file system ? 1. A big filesystem is a filesystem which occupy 20-30 or more percent of total drive space, i.e. first and last cylinder are physically located relatively far from each other. 2. It has a relatively large number of cylinder groups, for example more cylinder groups than 50% of the buffers in the buffer cache. The first results in long access times, while the second results in many buffers being used by metadata operations. Such operations use cylinder group blocks and on-disk inode blocks. The cylinder group block (fs->fs_cblkno) contains struct cg, inode and block bit maps. It is 2k in size for the default filesystem parameters. If new and parent directories are located in different cylinder groups then the system performs more input/output operations and uses more buffers. On filesystems with many cylinder groups, lots of cache buffers are used for metadata operations. My solution for this problem is very simple. I allocate many directories in one cylinder group. I also do some things, so that the new allocation method does not cause excessive fragmentation and all directory inodes will not be located at a location far from its file's inodes and data. The algorithm is: /* * Find a cylinder group to place a directory. * * The policy implemented by this algorithm is to allocate a * directory inode in the same cylinder group as its parent * directory, but also to reserve space for its files inodes * and data. Restrict the number of directories which may be * allocated one after another in the same cylinder group * without intervening allocation of files. * * If we allocate a first level directory then force allocation * in another cylinder group. */ My early versions of dirpref give me a good results for a wide range of file operations and different filesystem capacities except one case: those applications that create their entire directory structure first and only later fill this structure with files. My solution for such and similar cases is to limit a number of directories which may be created one after another in the same cylinder group without intervening file creations. For this purpose, I allocate an array of counters at mount time. This array is linked to the superblock fs->fs_contigdirs[cg]. Each time a directory is created the counter increases and each time a file is created the counter decreases. A 60Gb filesystem with 8mb/cg requires 10kb of memory for the counters array. The maxcontigdirs is a maximum number of directories which may be created without an intervening file creation. I found in my tests that the best performance occurs when I restrict the number of directories in one cylinder group such that all its files may be located in the same cylinder group. There may be some deterioration in performance if all the file inodes are in the same cylinder group as its containing directory, but their data partially resides in a different cylinder group. The maxcontigdirs value is calculated to try to prevent this condition. Since there is no way to know how many files and directories will be allocated later I added two optimization parameters in superblock/tunefs. They are: int32_t fs_avgfilesize; /* expected average file size */ int32_t fs_avgfpdir; /* expected # of files per directory */ These parameters have reasonable defaults but may be tweeked for special uses of a filesystem. They are only necessary in rare cases like better tuning a filesystem being used to store a squid cache. I have been using this algorithm for about 3 months. I have done a lot of testing on filesystems with different capacities, average filesize, average number of files per directory, and so on. I think this algorithm has no negative impact on filesystem perfomance. It works better than the default one in all cases. The new dirpref will greatly improve untarring/removing/coping of big directories, decrease load on cvs servers and much more. The new dirpref doesn't speedup a compilation process, but also doesn't slow it down. Obtained from: Grigoriy Orlov <gluk@ptci.ru> ===== ===== iedowse 2001/04/23 17:37:17 PDT Pre-dirpref versions of fsck may zero out the new superblock fields fs_contigdirs, fs_avgfilesize and fs_avgfpdir. This could cause panics if these fields were zeroed while a filesystem was mounted read-only, and then remounted read-write. Add code to ffs_reload() which copies the fs_contigdirs pointer from the previous superblock, and reinitialises fs_avgf* if necessary. Reviewed by: mckusick ===== ===== nik 2001/04/10 03:36:44 PDT Add information about the new options to newfs and tunefs which set the expected average file size and number of files per directory. Could do with some fleshing out. ===== |
| /src/sys/netinet/ | ||
| H A D | in_pcb.c | 1.138 Tue May 03 18:28:45 GMT 2011 dyoung Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive. 1.125 Mon May 05 17:11:17 GMT 2008 ad branches: 1.125.2; 1.125.6; - Convert hashinit() to use kmem_alloc(). The hash tables can be large and it's better to not have them in kmem_map. - Convert a couple of minor items along the way to kmem_alloc(). - Fix some memory leaks. 1.125 Mon May 05 17:11:17 GMT 2008 ad branches: 1.125.2; 1.125.6; - Convert hashinit() to use kmem_alloc(). The hash tables can be large and it's better to not have them in kmem_map. - Convert a couple of minor items along the way to kmem_alloc(). - Fix some memory leaks. 1.125 Mon May 05 17:11:17 GMT 2008 ad branches: 1.125.2; 1.125.6; - Convert hashinit() to use kmem_alloc(). The hash tables can be large and it's better to not have them in kmem_map. - Convert a couple of minor items along the way to kmem_alloc(). - Fix some memory leaks. |
| H A D | tcp_usrreq.c | 1.159 Tue May 03 18:28:45 GMT 2011 dyoung branches: 1.159.2; Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive. 1.125 Fri Oct 13 15:39:19 GMT 2006 elad Introduce KAUTH_REQ_NETWORK_SOCKET_CANSEE. Since we're not gonna be having credentials on sockets, at least not anytime soon, this is a way to check if we can "look" at a socket. Later on when (and if) we do have socket credentials, the interface usage remains the same because we pass the socket. This also fixes sysctl for inet/inet6 pcblist. |
| H A D | tcp_var.h | 1.166 Tue May 03 18:28:45 GMT 2011 dyoung Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive. 1.125 Tue Apr 05 01:07:17 GMT 2005 kurahone Added sysctl tunable limits for the number of maximum SACK holes per connection and per system. Idea taken from FreeBSD. |
| H A D | udp_usrreq.c | 1.180 Tue May 03 18:28:45 GMT 2011 dyoung Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive. 1.125 Wed Dec 15 04:25:20 GMT 2004 thorpej Don't perform checksums on loopback interfaces. They can be reenabled with the net.inet.*.do_loopback_cksum sysctl. Approved by: groo |
| /src/sys/kern/ | ||
| H A D | sys_pipe.c | 1.125 Sun Dec 13 20:02:23 GMT 2009 dsl Another, better, fix for PR/26567. Only sleep once within each pipe_read/pipe_write call. If there is no data/space available after we wakeup return ERESTART so then the 'fd' number is validated again. A simple broadcast of the cvs is then enough to evict the correct threads when close() is called from an active thread. 1.92 Fri Dec 28 13:11:16 GMT 2007 ad Pull up 1.87.2.8. |
| /src/sbin/newfs/ | ||
| H A D | mkfs.c | 1.125 Tue Jun 16 23:18:55 GMT 2015 christos fix error messages containing \n 1.55 Thu Sep 06 02:16:01 GMT 2001 lukem Incorporate the enhanced ffs_dirpref() by Grigoriy Orlov, as found in FreeBSD (three commits; the initial work, man page updates, and a fix to ffs_reload()), with the following differences: - Be consistent between newfs(8) and tunefs(8) as to the options which set and control the tuning parameters for this work (avgfilesize & avgfpdir) - Use u_int16_t instead of u_int8_t to keep track of the number of contiguous directories (suggested by Chuck Silvers) - Work within our FFS_EI framework - Ensure that fs->fs_maxclusters and fs->fs_contigdirs don't point to the same area of memory The new algorithm has a marked performance increase, especially when performing tasks such as untarring pkgsrc.tar.gz, etc. The original FreeBSD commit messages are attached: ===== mckusick 2001/04/10 01:39:00 PDT Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>. His description of the problem and solution follow. My own tests show speedups on typical filesystem intensive workloads of 5% to 12% which is very impressive considering the small amount of code change involved. ------ One day I noticed that some file operations run much faster on small file systems then on big ones. I've looked at the ffs algorithms, thought about them, and redesigned the dirpref algorithm. First I want to describe the results of my tests. These results are old and I have improved the algorithm after these tests were done. Nevertheless they show how big the perfomance speedup may be. I have done two file/directory intensive tests on a two OpenBSD systems with old and new dirpref algorithm. The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports". The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release. It contains 6596 directories and 13868 files. The test systems are: 1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for test is at wd1. Size of test file system is 8 Gb, number of cg=991, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=35 2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system at wd0, file system for test is at wd1. Size of test file system is 40 Gb, number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50 You can get more info about the test systems and methods at: http://www.ptci.ru/gluk/dirpref/old/dirpref.html Test Results tar -xzf ports.tar.gz rm -rf ports mode old dirpref new dirpref speedup old dirprefnew dirpref speedup First system normal 667 472 1.41 477 331 1.44 async 285 144 1.98 130 14 9.29 sync 768 616 1.25 477 334 1.43 softdep 413 252 1.64 241 38 6.34 Second system normal 329 81 4.06 263.5 93.5 2.81 async 302 25.7 11.75 112 2.26 49.56 sync 281 57.0 4.93 263 90.5 2.9 softdep 341 40.6 8.4 284 4.76 59.66 "old dirpref" and "new dirpref" columns give a test time in seconds. speedup - speed increasement in times, ie. old dirpref / new dirpref. ------ Algorithm description The old dirpref algorithm is described in comments: /* * Find a cylinder to place a directory. * * The policy implemented by this algorithm is to select from * among those cylinder groups with above the average number of * free inodes, the one with the smallest number of directories. */ A new directory is allocated in a different cylinder groups than its parent directory resulting in a directory tree that is spreaded across all the cylinder groups. This spreading out results in a non-optimal access to the directories and files. When we have a small filesystem it is not a problem but when the filesystem is big then perfomance degradation becomes very apparent. What I mean by a big file system ? 1. A big filesystem is a filesystem which occupy 20-30 or more percent of total drive space, i.e. first and last cylinder are physically located relatively far from each other. 2. It has a relatively large number of cylinder groups, for example more cylinder groups than 50% of the buffers in the buffer cache. The first results in long access times, while the second results in many buffers being used by metadata operations. Such operations use cylinder group blocks and on-disk inode blocks. The cylinder group block (fs->fs_cblkno) contains struct cg, inode and block bit maps. It is 2k in size for the default filesystem parameters. If new and parent directories are located in different cylinder groups then the system performs more input/output operations and uses more buffers. On filesystems with many cylinder groups, lots of cache buffers are used for metadata operations. My solution for this problem is very simple. I allocate many directories in one cylinder group. I also do some things, so that the new allocation method does not cause excessive fragmentation and all directory inodes will not be located at a location far from its file's inodes and data. The algorithm is: /* * Find a cylinder group to place a directory. * * The policy implemented by this algorithm is to allocate a * directory inode in the same cylinder group as its parent * directory, but also to reserve space for its files inodes * and data. Restrict the number of directories which may be * allocated one after another in the same cylinder group * without intervening allocation of files. * * If we allocate a first level directory then force allocation * in another cylinder group. */ My early versions of dirpref give me a good results for a wide range of file operations and different filesystem capacities except one case: those applications that create their entire directory structure first and only later fill this structure with files. My solution for such and similar cases is to limit a number of directories which may be created one after another in the same cylinder group without intervening file creations. For this purpose, I allocate an array of counters at mount time. This array is linked to the superblock fs->fs_contigdirs[cg]. Each time a directory is created the counter increases and each time a file is created the counter decreases. A 60Gb filesystem with 8mb/cg requires 10kb of memory for the counters array. The maxcontigdirs is a maximum number of directories which may be created without an intervening file creation. I found in my tests that the best performance occurs when I restrict the number of directories in one cylinder group such that all its files may be located in the same cylinder group. There may be some deterioration in performance if all the file inodes are in the same cylinder group as its containing directory, but their data partially resides in a different cylinder group. The maxcontigdirs value is calculated to try to prevent this condition. Since there is no way to know how many files and directories will be allocated later I added two optimization parameters in superblock/tunefs. They are: int32_t fs_avgfilesize; /* expected average file size */ int32_t fs_avgfpdir; /* expected # of files per directory */ These parameters have reasonable defaults but may be tweeked for special uses of a filesystem. They are only necessary in rare cases like better tuning a filesystem being used to store a squid cache. I have been using this algorithm for about 3 months. I have done a lot of testing on filesystems with different capacities, average filesize, average number of files per directory, and so on. I think this algorithm has no negative impact on filesystem perfomance. It works better than the default one in all cases. The new dirpref will greatly improve untarring/removing/coping of big directories, decrease load on cvs servers and much more. The new dirpref doesn't speedup a compilation process, but also doesn't slow it down. Obtained from: Grigoriy Orlov <gluk@ptci.ru> ===== ===== iedowse 2001/04/23 17:37:17 PDT Pre-dirpref versions of fsck may zero out the new superblock fields fs_contigdirs, fs_avgfilesize and fs_avgfpdir. This could cause panics if these fields were zeroed while a filesystem was mounted read-only, and then remounted read-write. Add code to ffs_reload() which copies the fs_contigdirs pointer from the previous superblock, and reinitialises fs_avgf* if necessary. Reviewed by: mckusick ===== ===== nik 2001/04/10 03:36:44 PDT Add information about the new options to newfs and tunefs which set the expected average file size and number of files per directory. Could do with some fleshing out. ===== |
| /src/usr.bin/indent/ | ||
| H A D | io.c | 1.125 Fri Nov 19 19:55:15 GMT 2021 rillig indent: replace ps.procname with ps.is_function_definition Omly the first character of ps.procname was ever read, and it was only compared to '\0'. Using a bool for this means simpler code, less memory and fewer wasted CPU cycles due to the removed strncpy. No functional change. 1.9 Sat Dec 19 17:00:08 GMT 1998 christos branches: 1.9.2; 1.9.10; char -> unsigned char, braces for gcc-2.8.1 |
| H A D | lexi.c | 1.125 Sun Oct 31 19:57:44 GMT 2021 rillig indent: replace simple cases of keyword_kind with lexer_symbol The remaining keyword kinds 'tag' and 'type' require a bit more thought, so do them in a separate step. No functional change. 1.8 Sat Dec 19 17:00:08 GMT 1998 christos char -> unsigned char, braces for gcc-2.8.1 |
| /src/usr.bin/make/ | ||
| H A D | for.c | 1.125 Thu Dec 31 03:19:00 GMT 2020 rillig make(1): clean up ForReadMore After the previous clean up in for.c 1.123 from 2020-12-30, GCC 5.5 did not inline the function SubstVarLong anymore since it was now called from 2 places. GCC didn't notice that the function call was essentially the same since in differed only in the end character. By combining the cases for ${V} and $(V), the code becomes even shorter than before, while still being understandable. 1.117 Sun Dec 13 21:27:45 GMT 2020 rillig make(1): replace %zu with %u in printf calls This is needed to compile bmake with GCC 2.8.1 on SunOS 5.9. To support ancient systems like this, the whole code of usr.bin/make is supposed to use only ISO C90 features, except for filemon, which is not used on these systems. |
| H A D | meta.c | 1.160 Sun Dec 13 21:27:45 GMT 2020 rillig make(1): replace %zu with %u in printf calls This is needed to compile bmake with GCC 2.8.1 on SunOS 5.9. To support ancient systems like this, the whole code of usr.bin/make is supposed to use only ISO C90 features, except for filemon, which is not used on these systems. 1.125 Sun Oct 18 13:02:10 GMT 2020 rillig make(1): rename Lst_Init to Lst_New For the other types such as HashTable and Buffer, the Init function does not allocate the memory for the structure itself, it only fills it. |
| /src/sys/netinet6/ | ||
| H A D | in6_pcb.c | 1.125 Fri May 30 01:39:03 GMT 2014 christos Introduce 2 new variables: ipsec_enabled and ipsec_used. Ipsec enabled is controlled by sysctl and determines if is allowed. ipsec_used is set automatically based on ipsec being enabled, and rules existing. 1.113 Tue May 03 18:28:45 GMT 2011 dyoung Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive. |
| H A D | udp6_usrreq.c | 1.125 Tue Nov 15 20:50:28 GMT 2016 mlelstv Enforce alignment requirements that are violated in some cases. For machines that don't need strict alignment (i386,amd64,vax,m68k) this is a no-op. Fixes PR kern/50766 but should be improved. 1.89 Tue May 03 18:28:45 GMT 2011 dyoung Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive. |
| H A D | ip6_input.c | 1.130 Tue May 03 18:28:45 GMT 2011 dyoung Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive. 1.125 Wed Mar 18 17:06:52 GMT 2009 cegger bcopy -> memcpy |
| H A D | nd6_nbr.c | 1.125 Mon Jul 25 04:21:20 GMT 2016 ozaki-r Make DAD of ARP/NDP MP-safe with coarse-grained locks The change also prevents arp_dad_timer/nd6_dad_timer from running if arp_dad_stop/nd6_dad_stop is called, which makes sure that callout_reset won't be called during callout_halt. 1.74 Thu May 17 00:53:26 GMT 2007 dyoung Fix the memory leak reported in kern/36337. Thanks Matthias Scheler for the heads-up. My fix is based on the following patches from FreeBSD, however, I extracted the code into a subroutine, nd6_llinfo_release_pkts(): http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/netinet6/nd6.c.diff?r1=1.48.2.18;r2=1.48.2.19 http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/netinet6/nd6_nbr.c.diff?r1=1.29.2.8;r2=1.29.2.9 |
| H A D | raw_ip6.c | 1.125 Mon Jul 07 17:13:56 GMT 2014 rtr * sprinkle KASSERT(solocked(so)); in all pr_stat() functions. * fix remaining inconsistent struct socket parameter names. 1.108 Tue May 03 18:28:45 GMT 2011 dyoung branches: 1.108.4; 1.108.8; Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive. |
| H A D | nd6.c | 1.125 Tue Apr 15 03:57:04 GMT 2008 thorpej branches: 1.125.2; Make ip6 and icmp6 stats per-cpu. 1.125 Tue Apr 15 03:57:04 GMT 2008 thorpej branches: 1.125.2; Make ip6 and icmp6 stats per-cpu. 1.115 Thu May 17 00:53:26 GMT 2007 dyoung Fix the memory leak reported in kern/36337. Thanks Matthias Scheler for the heads-up. My fix is based on the following patches from FreeBSD, however, I extracted the code into a subroutine, nd6_llinfo_release_pkts(): http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/netinet6/nd6.c.diff?r1=1.48.2.18;r2=1.48.2.19 http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/netinet6/nd6_nbr.c.diff?r1=1.29.2.8;r2=1.29.2.9 |
| /src/usr.bin/ftp/ | ||
| H A D | fetch.c | 1.125 Thu Sep 28 00:29:23 GMT 2000 lukem explicitly use SOCK_STREAM with socket() instead of res->ai_socktype, because it appears that linux with glibc doesn't set the latter correctly after one of getaddrinfo() or getnameinfo(). 1.51 Mon Mar 15 08:52:17 GMT 1999 christos Add a few more variables that end up in registers in gcc-2.8.1 |
| /src/ | ||
| H A D | Makefile | 1.125 Sun Jun 10 13:15:29 GMT 2001 mrg clarify some variable documentation; from cagney 1.58 Fri Jul 24 16:48:47 GMT 1998 tv Fix the rebuild of libgcc: - If USE_EGCS is set, rebuild egcs's libgcc and install it (unless DESTDIR is set and system compiler is not gcc 2.8, in which case print a warning message and do nothing). - Do not rebuild gcc 2.7's libgcc. egcs can build this fine. |
Completed in 432 milliseconds