mirrors/emacs
1
Fork 0
mirror of git://git.sv.gnu.org/emacs.git synced 2026-01-08 04:30:45 -08:00
Code Issues Projects Releases Wiki Activity
5040 commits 243 branches 194 tags 713 MiB
Commit graph

1214 commits

Author SHA1 Message Date
Richard Brooksby
e59678ecf9 Purging half of the spare memory, rather than just enough to get under the limit, so that there aren’t many small unmaps, and many scans for spare pages. 
Copied from Perforce
 Change: 184326
 ServerID: perforce.ravenbrook.com
 2014-02-11 18:26:00 +00:00
Richard Brooksby
9e057d1bcc Fixing boundary cases for unmapping page table pages when there aren’t a nice whole number of descriptors on each page. 
Copied from Perforce
 Change: 184325
 ServerID: perforce.ravenbrook.com
 2014-02-11 17:18:21 +00:00
Richard Brooksby
f1d5750639 Adding a cache of free pages per zone to the arena, so that we recycle single pages fast, especially after each reclaim. 
Copied from Perforce
 Change: 184324
 ServerID: perforce.ravenbrook.com
 2014-02-11 15:23:32 +00:00
Richard Brooksby
9bc635b24d Further tidying up of the page descriptor and naming. 
Copied from Perforce
 Change: 184323
 ServerID: perforce.ravenbrook.com
 2014-02-11 14:14:16 +00:00
Richard Brooksby
f9e9805b39 Encoding the page state in the bottom two bits of the first field, rather than using up a whole word, in order to make room for a free list pointer. 
Copied from Perforce
 Change: 184322
 ServerID: perforce.ravenbrook.com
 2014-02-11 13:41:57 +00:00
Richard Brooksby
f8ac78ba0f Improving commentary on page table unmapping code in response to review. see <https://info.ravenbrook.com/mail/2014/01/31/17-26-12/0/>. 
Copied from Perforce
 Change: 184291
 ServerID: perforce.ravenbrook.com
 2014-02-04 16:56:36 +00:00
Richard Brooksby
66a05e1f05 Changes from review comments made in <https://info.ravenbrook.com/mail/2014/01/31/13-40-25/0/>. 
ArenaSpareCommitExceeded changed to ArenaPurgeSpare so that the arena can properly reduce the amount of spare memory when setting the commit limit etc.
pagesMarkAllocated promptly unmaps page table pages in its error path.
Miscellaneous clarifications and extra checking.

Copied from Perforce
 Change: 184267
 ServerID: perforce.ravenbrook.com
 2014-01-31 15:21:21 +00:00
Richard Brooksby
7526f5abe8 Keeping a single ring of spare chunks in the vm arena, in lru order, so that spare pages are unmapped closer to lru. 
Copied from Perforce
 Change: 184251
 ServerID: perforce.ravenbrook.com
 2014-01-31 09:36:42 +00:00
Richard Brooksby
2b08236430 Using a more robust page table unmapping method that catches pages previously left mapped in the error page of pagesmarkallocated. 
Test run passes.

Copied from Perforce
 Change: 184250
 ServerID: perforce.ravenbrook.com
 2014-01-31 02:05:11 +00:00
Richard Brooksby
43602ffd58 Attempting to correct the unmapping of the page table. still doesn’t quite work. 
Copied from Perforce
 Change: 184249
 ServerID: perforce.ravenbrook.com
 2014-01-31 00:03:32 +00:00
Richard Brooksby
ce329f192e Abolishing the spare pages bit table. 
Not quite working yet. This code fails to unmap unused page table pages, and so triggers an assertion at chunk finish.

Copied from Perforce
 Change: 184246
 ServerID: perforce.ravenbrook.com
 2014-01-30 19:58:36 +00:00
Richard Brooksby
0296f9c04a Adding spare pages to a ring within each chunk, so that they can be found without a bit table scan. 
Spare pages are returned to the OS using the ring in roughly LRU order.
Only enough spare pages are returned to maintain hysteresis, rather than all of them.
Nearly ready to strip out the noSparePages bit table.

Copied from Perforce
 Change: 184245
 ServerID: perforce.ravenbrook.com
 2014-01-30 19:58:36 +00:00
Richard Brooksby
10e9a638e4 Fix broken hysteresis in tract mapping. only purge spare pages if it’s possible to free a chunk, rather than speculatively every reclaim. see <https://info.ravenbrook.com/mail/2014/01/25/17-37-52/0/>. 
Copied from Perforce
 Change: 184125
 ServerID: perforce.ravenbrook.com
 2014-01-25 18:00:51 +00:00
Richard Brooksby
56d48bbd1d iterating over segments using the pool and segment rings, rather than scanning the tract table, which may be full of irrelevant tracts, especially when non-gc pools are large. see <https://info.ravenbrook.com/mail/2014/01/25/12-16-09/0/>. 
Copied from Perforce
 Change: 184120
 ServerID: perforce.ravenbrook.com
 2014-01-25 15:33:10 +00:00
Richard Brooksby
893ac1da13 Adding a rash configuration to xcode builds. 
Copied from Perforce
 Change: 184117
 ServerID: perforce.ravenbrook.com
 2014-01-25 15:25:52 +00:00
Richard Brooksby
a788a9efdd Ignoring a.out (default output from cc). 
Copied from Perforce
 Change: 184116
 ServerID: perforce.ravenbrook.com
 2014-01-25 15:24:28 +00:00
Richard Brooksby
f91a987e32 Inclusion order mysteriously prevents profiling under xcode. 
Copied from Perforce
 Change: 184111
 ServerID: perforce.ravenbrook.com
 2014-01-24 16:00:49 +00:00
Richard Brooksby
8ef0ecb155 Merging the "dj" benchmark from the benchmarking development branch. 
Copied from Perforce
 Change: 184106
 ServerID: perforce.ravenbrook.com
 2014-01-24 15:29:28 +00:00
Richard Brooksby
199ef9ca11 Catch-up merge from master. 
Copied from Perforce
 Change: 184103
 ServerID: perforce.ravenbrook.com
 2014-01-24 14:22:27 +00:00
Gareth Rees
1ccc6908a6 Remove obsolete file osxc.h. 
Copied from Perforce
 Change: 183954
 ServerID: perforce.ravenbrook.com
 2014-01-12 21:57:54 +00:00
Gareth Rees
f6160d9720 Test case libcbt.c is obsolete: mpslibcb mechanism was removed in change 179279. 
Copied from Perforce
 Change: 183952
 ServerID: perforce.ravenbrook.com
 2014-01-12 21:29:53 +00:00
Gareth Rees
997f5a914e Clarify decision to preserve/reclaim in amcreclaimnailed, avoiding ternary operator inside the if condition. 
Copied from Perforce
 Change: 183948
 ServerID: perforce.ravenbrook.com
 2014-01-12 11:21:21 +00:00
Gareth Rees
5ddfd0dbcd Remove unused (and unusable) macros arenarootring() and arenatracering(). 
Copied from Perforce
 Change: 183947
 ServerID: perforce.ravenbrook.com
 2014-01-12 11:19:35 +00:00
Gareth Rees
f316e646b3 Ignore build products for new platform lii6ll. 
Ignore telemetry database mpiso*.sql.

Copied from Perforce
 Change: 183942
 ServerID: perforce.ravenbrook.com
 2014-01-11 21:37:19 +00:00
Gareth Rees
3afb84b7b5 Rename spi3.c to spw3i3mv.c as it is windows- and microsoft- specific. 
Copied from Perforce
 Change: 183939
 ServerID: perforce.ravenbrook.com
 2014-01-10 20:51:13 +00:00
Gareth Rees
d917eb1d0c Merge branch/2013-05-17/emergency into the master sources. 
This change refactors MVTBufferFill and makes related improvements to improve the clarity of the code.

Copied from Perforce
 Change: 183935
 ServerID: perforce.ravenbrook.com
 2014-01-10 18:52:08 +00:00
Gareth Rees
c5bce746b1 Change result type of abqpush, abqpop and abqpeek to bool. 
Copied from Perforce
 Change: 183934
 ServerID: perforce.ravenbrook.com
 2014-01-10 18:44:46 +00:00
Gareth Rees
b14f5a51d7 Make improvements (mostly to comments) based on reviews: 
https://info.ravenbrook.com/mail/2014/01/10/15-57-50/0/
https://info.ravenbrook.com/mail/2014/01/10/15-40-47/0/

Copied from Perforce
 Change: 183932
 ServerID: perforce.ravenbrook.com
 2014-01-10 18:18:06 +00:00
Gareth Rees
ea9df0b2b7 Merge custom cet mainline into the master sources. this includes fixes for job003519 (clock values in text telemetry logs on windows have top 32 bits zero), job003536 (cet office can't replicate 100k snowmen in reasonable time), job003651 (stack overflows are not reliably handled). 
Copied from Perforce
 Change: 183926
 ServerID: perforce.ravenbrook.com
 2014-01-10 14:16:08 +00:00
Gareth Rees
2044d82a77 Clarify comment above bufferisreset: by "all pointers into the region" it actually means "the base, init, alloc, and limit pointers". 
Copied from Perforce
 Change: 183637
 ServerID: perforce.ravenbrook.com
 2013-11-04 15:31:58 +00:00
Gareth Rees
73fa242bb1 Improve documentation of thread- and exception-safety precautions. 
Copied from Perforce
 Change: 183635
 ServerID: perforce.ravenbrook.com
 2013-11-04 15:14:47 +00:00
Gareth Rees
ba0e8c1b8b Put testsetsize back to its original value of 200, so that randomsizealigned does not invoke undefined behaviour by right-shifting by more than the word width. 
Copied from Perforce
 Change: 183617
 ServerID: perforce.ravenbrook.com
 2013-10-31 09:39:09 +00:00
Gareth Rees
8f9423a8c3 Add comments and assertions to make it clearer how btfindresrange and btfindresrangehigh work. in particular, explain why the possible unsigned integer underflow in btfindresrangehigh is safe. 
Copied from Perforce
 Change: 183613
 ServerID: perforce.ravenbrook.com
 2013-10-30 15:53:19 +00:00
Gareth Rees
b87ce7680f Don't call mvtsegfree from mvtfinish: we don't know whether or not there was any fragmented (unavailable) space in this segment, and so we can't keep the accounting correct. add assertion in mvtsegfree that detects incorrect accounting. 
Copied from Perforce
 Change: 183610
 ServerID: perforce.ravenbrook.com
 2013-10-30 13:15:09 +00:00
Gareth Rees
73350475ed The next release of the mps will be 1.112.0, so update version.c. 
Update the version creating procedure: version.c is now the only file containing the release number.

Copied from Perforce
 Change: 183548
 ServerID: perforce.ravenbrook.com
 2013-10-10 14:29:31 +01:00
Gareth Rees
5824d8a73f Improve wording of comment. 
Copied from Perforce
 Change: 183526
 ServerID: perforce.ravenbrook.com
 2013-10-04 11:50:19 +01:00
Gareth Rees
d2c28e7deb Fix broken mps_key_pool_debug_options keyword argument. 
Copied from Perforce
 Change: 183524
 ServerID: perforce.ravenbrook.com
 2013-10-04 11:05:56 +01:00
Gareth Rees
e522a0b273 Document that you can specify mps_key_rank when creating an allocation point on an ams pool. make mps_key_rank default to mps_rank_exact(). 
Copied from Perforce
 Change: 183515
 ServerID: perforce.ravenbrook.com
 2013-10-02 15:44:34 +01:00
Gareth Rees
692489dd2b Make keyword mps_key_ams_support_ambiguous optional, with default value false. update documentation accordingly. 
Copied from Perforce
 Change: 183513
 ServerID: perforce.ravenbrook.com
 2013-10-02 15:14:57 +01:00
Gareth Rees
fab773659d Use macro nonnegative so that we can keep asserting that values are >= 0 without provoking "unsigned comparison >= 0 is always true" warnings from gcc. 
Copied from Perforce
 Change: 183505
 ServerID: perforce.ravenbrook.com
 2013-10-01 14:27:52 +01:00
Gareth Rees
544a160559 Merge branch/2013-08-21/lii6ll/... into the master sources, adding new supported platform lii6ll (linux, x86-64, clang/llvm). 
Copied from Perforce
 Change: 183503
 ServerID: perforce.ravenbrook.com
 2013-10-01 14:04:36 +01:00
Gareth Rees
51a97f1c37 Avoid comparison of unsigned >= 0 so that we can compile without warnings under gcc -wextra. 
Copied from Perforce
 Change: 183501
 ServerID: perforce.ravenbrook.com
 2013-10-01 11:10:59 +01:00
Gareth Rees
af06b04a77 Raise an error if variety is something other than rash, hot, or cool. 
Copied from Perforce
 Change: 183499
 ServerID: perforce.ravenbrook.com
 2013-10-01 10:53:09 +01:00
Gareth Rees
949578bac4 Avoid warnings when compiling with -wwrite-strings: 
* Use const char * in places where we want to assign string constants.
Avoid warnings when compiling with -Wextra:
* Avoid if statement with empty body in amcsshe.c
* Avoid comparison of unsigned >= 0.

Copied from Perforce
 Change: 183498
 ServerID: perforce.ravenbrook.com
 2013-10-01 10:52:45 +01:00
Gareth Rees
a3c7ed0cab Turn on -wextra -wwrite-strings when compiling with gcc or clang. 
Add missing initializer to mps_args_none.
Change "char *" to "const char *" where necessary.

Copied from Perforce
 Change: 183494
 ServerID: perforce.ravenbrook.com
 2013-09-30 17:18:52 +01:00
Gareth Rees
b59ba75674 New supported platform lii6ll (linux, x86-64, clang/llvm). 
Copied from Perforce
 Change: 183414
 ServerID: perforce.ravenbrook.com
 2013-09-11 10:27:22 +01:00
Gareth Rees
1d39ec5b87 Make the mps compilable under clang/llvm 3.0 with the -ansi -pedantic options: 
Avoid if with empty body by adding NOOP.
Avoid extended offsetof in expansion of PageOfTract, linkOfMessage, linkOfRing, and Pool2AMST macros.

Copied from Perforce
 Change: 183413
 ServerID: perforce.ravenbrook.com
 2013-09-11 10:21:12 +01:00
Gareth Rees
7bc31bced8 Don't detect the compiler as gcc when it's actually clang. emit an error instead. 
Copied from Perforce
 Change: 183334
 ServerID: perforce.ravenbrook.com
 2013-08-20 17:10:36 +01:00
Richard Brooksby
35dd801372 Fixing up build for test platform xci3gc, broken by os x threads development. 
Enabling amcssth and awlutth tests for all platforms, since I believe these were fixed by OS X threads development.

Copied from Perforce
 Change: 183118
 ServerID: perforce.ravenbrook.com
 2013-07-17 16:49:31 +01:00
Richard Brooksby
086ad82475 Removing false comment about multiple thread registration, but see job003559. 
Copied from Perforce
 Change: 183114
 ServerID: perforce.ravenbrook.com
 2013-07-17 14:38:05 +01:00