mirrors/emacs
1
Fork 0
mirror of git://git.sv.gnu.org/emacs.git synced 2025-12-24 14:30:43 -08:00
Code Issues Projects Releases Wiki Activity
3605 commits 242 branches 194 tags 7.7 GiB
Commit graph

535 commits

Author SHA1 Message Date
Richard Kistruck
bbd83d25bc Mps br/padding poolamc.c: change amcbufferfill, so that for large requests (> 8 arenaaligns) it gives only requested size; pad the rest. see job001811. the number 8 is provisional... 
Copied from Perforce
 Change: 167759
 ServerID: perforce.ravenbrook.com
 2009-04-08 16:04:09 +01:00
Richard Kistruck
e2b54a424f Mps br/padding exp-167727 
Copied from Perforce
 Change: 167727
 ServerID: perforce.ravenbrook.com
 2009-04-02 18:36:50 +01:00
Richard Kistruck
ed1f97ac9f Mps br/padding: more diagnostic for how padding survives a collection. 
More diag:
  - new traceSetSignalEmergency;
  - new AMCFix_amcSegCreateNailboard and AMCHeaderFix_amcSegCreateNailboard
    (note: the other, unreported, route to becoming boarded is because of a (mutator) buffer);
  - in traces: show SegBase and zone;
   - in TraceStart: show genZoneSets;
   - in amcReclaimNailed: show cbpip and cbpad.
 zcoll.c:
   - sizemethod 1 to Make command: occasionally makes >1MiB objects;
   - declare root_table all MPS_RANK_AMBIG.

Copied from Perforce
 Change: 167726
 ServerID: perforce.ravenbrook.com
 2009-04-02 18:29:10 +01:00
Richard Kistruck
0160f03ac1 Mps br/padding exp-167703 
Copied from Perforce
 Change: 167703
 ServerID: perforce.ravenbrook.com
 2009-04-01 19:05:29 +01:00
Richard Kistruck
fa6fd160ad Mps br/padding diag.c: show some tracestart diag as well (because, res, com, alignment) 
Copied from Perforce
 Change: 167702
 ServerID: perforce.ravenbrook.com
 2009-04-01 19:03:27 +01:00
Richard Kistruck
bc99db41f9 Mps br/padding: exp-167699 
Copied from Perforce
 Change: 167699
 ServerID: perforce.ravenbrook.com
 2009-04-01 18:04:30 +01:00
Richard Kistruck
9318d363ef Mps br/padding poolamc.c: (char) casts needed for vc6.0 compiler on platform.w3i3mv 
Copied from Perforce
 Change: 167698
 ServerID: perforce.ravenbrook.com
 2009-04-01 17:52:55 +01:00
Richard Kistruck
7d6430312d Mps br/padding poolamc.c: (char) casts needed for vc6.0 compiler on platform.w3i3mv 
Copied from Perforce
 Change: 167697
 ServerID: perforce.ravenbrook.com
 2009-04-01 17:48:49 +01:00
Richard Kistruck
492c2519ef Mps br/padding poolamc.c: diags for condemn and reclaim of big segs (>= 1mib) 
Copied from Perforce
 Change: 167696
 ServerID: perforce.ravenbrook.com
 2009-04-01 17:36:58 +01:00
Richard Kistruck
6d25fd294f Mps br/padding poolamc.c: diags for condemn and reclaim of big segs (> 1mib) 
Copied from Perforce
 Change: 167695
 ServerID: perforce.ravenbrook.com
 2009-04-01 17:34:31 +01:00
Richard Kistruck
e71890f15b Mps master zmess.c, zcoll.c (integ from br/timing): oops, need to (re-)initialise myroot array for each script, before declaring it a root. 
Copied from Perforce
 Change: 167402
 ServerID: perforce.ravenbrook.com
 2009-02-17 16:21:25 +00:00
Richard Kistruck
d25b1ab8e2 Mps br/timing zcoll.c: oops, need to (re-)initialise myroot array for each script, before declaring it a root (integ from zmess.c). 
Copied from Perforce
 Change: 167400
 ServerID: perforce.ravenbrook.com
 2009-02-17 16:15:59 +00:00
Richard Kistruck
a0a8d095d0 Mps br/timing zmess.c: oops, need to (re-)initialise myroot array for each script, before declaring it a root. 
Copied from Perforce
 Change: 167397
 ServerID: perforce.ravenbrook.com
 2009-02-17 14:47:40 +00:00
Richard Kistruck
d5379b26ef Mps master (integ from br/timing): 
Messages:
  - _gc_start and _gc now always ControlAlloc'd, new lifecycle, much better code.  Fixes job001989.
  - New test zmess, which tests generating and getting MPS messages.  Tests job001989, and also job001658 (subsumes fin1658a.c).
  - New file traceanc.c, for ancillary code extracted from trace.c.
  - Diag (.variety.di) at ArenaDestroy, say if message queue non-empty.
Randomize() and rnd() now much better seeding, improved (more random) generator, thorough verification, faster.
New test zcoll (incomplete), which tests MPS collections, their timing, and feedback about them.
MPS_ALERT_COLLECTION_START/STOP renamed as MPS_ALERT_COLLECTION_BEGIN/END

Copied from Perforce
 Change: 167392
 ServerID: perforce.ravenbrook.com
 2009-02-17 13:01:39 +00:00
Richard Kistruck
0110292e21 Mps br/timing: rename mps_alert_collection_*. the correct term for when a new mps operation begins is "begin", and when it ends is "end". (whereas several starts/stops may happen during a single mps operation). 
Copied from Perforce
 Change: 167386
 ServerID: perforce.ravenbrook.com
 2009-02-16 17:11:05 +00:00
Richard Kistruck
f8b818a156 Mps br/timing finaltest.c: (comment only) warn that this file has been hacked to be a bad client 
Copied from Perforce
 Change: 167384
 ServerID: perforce.ravenbrook.com
 2009-02-16 16:41:20 +00:00
Richard Kistruck
8a2e1280de Mps br/timing poolamc.c: (revert temporary hack @167343 =) make amcnailgetmark() faithfully report presence of mark. (this removes the temporary support for ambiguous interior pointers). 
Copied from Perforce
 Change: 167382
 ServerID: perforce.ravenbrook.com

GitHub-reference: https://github.com/Ravenbrook/mps/issues/4
 2009-02-16 16:33:38 +00:00
Richard Kistruck
b7907a053d Mps br/timing zcoll.c: catalog example is broken because of 
unmanaged ambiguous interior pointers (when compiled with 
moderate optimization), so comment it out for now.

Copied from Perforce
 Change: 167381
 ServerID: perforce.ravenbrook.com
 2009-02-16 16:26:39 +00:00
Richard Kistruck
ce6846ffa8 Mps br/timing zcoll.c: (fix compiler warning for w3i3m9) vc9 warns that sscanf is unsafe; disable this warning for zcoll.c. see also job001934. 
Copied from Perforce
 Change: 167379
 ServerID: perforce.ravenbrook.com
 2009-02-16 14:57:54 +00:00
Richard Kistruck
2a1485e45d Mps br/timing testlib.c: (fix compiler warning) msvc thinks "local variable 'r' may be used without having been initialized" 
Copied from Perforce
 Change: 167378
 ServerID: perforce.ravenbrook.com
 2009-02-16 14:45:11 +00:00
Richard Kistruck
c5dc7885d8 Mps br/timing zcoll: show gc timing, by printing mps_message_clock value (currently formatted for centiseconds, ie. clock() on unix) 
Copied from Perforce
 Change: 167351
 ServerID: perforce.ravenbrook.com
 2009-02-06 22:14:35 +00:00
Richard Kistruck
edb3054ba3 Mps br/timing notes: disassembly of gcc -o1 output showing ambiguous interior pointers. 
Copied from Perforce
 Change: 167350
 ServerID: perforce.ravenbrook.com
 2009-02-06 17:42:27 +00:00
Richard Kistruck
3477d797d5 Mps br/timing zcoll.c: (comments only) briefly describe script commands 
Copied from Perforce
 Change: 167349
 ServerID: perforce.ravenbrook.com
 2009-02-06 17:38:34 +00:00
Richard Kistruck
f32af8a65a mps br/timing poolamc.c: [temporary code] in order to allow ambig interior pointers to preserve objects in place, make amcnailgetmark() always return true. this works: ambig interior pointers are sufficent to preserve in place the object they point into. 
Copied from Perforce
 Change: 167343
 ServerID: perforce.ravenbrook.com
 2009-02-05 15:51:30 +00:00
Richard Kistruck
41a67fc94f Mps br/timing poolamc.c: [back out previous broken code] revert to original poolamc.c 
Copied from Perforce
 Change: 167342
 ServerID: perforce.ravenbrook.com
 2009-02-05 15:48:58 +00:00
Richard Kistruck
18bc05a2f4 Mps br/timing poolamc.c: [broken code] attempt to never use nailboards, aiming to fall back to nailing whole seg, thereby allowing ambig interior pointers to presrve objects in place. does not work. 
Copied from Perforce
 Change: 167341
 ServerID: perforce.ravenbrook.com
 2009-02-05 15:47:48 +00:00
Richard Kistruck
21ee127448 Mps br/timing zcoll catalog client: 
1. CatalogCheck(), which verifies the tree.
2. Don't set empty object slots to NULL -- make_dylan_vector() has already set them to a DYLAN_INT(0), which means dylan's fmt_scan ignores them.
3. fflush(stdout) more.

Copied from Perforce
 Change: 167321
 ServerID: perforce.ravenbrook.com
 2009-02-04 17:40:54 +00:00
Richard Kistruck
70998dd894 Mps br/timing zcoll.c: start to write the "catalog" example client, and find it already gives some non-obvious results. 
Copied from Perforce
 Change: 167227
 ServerID: perforce.ravenbrook.com
 2009-01-27 14:14:11 +00:00
Richard Kistruck
6c3156fa91 Mps br/timing zcoll.c: move arenasize and rootspace into script; 
showStatsAscii uses ascii-graphics to show collections

Copied from Perforce
 Change: 167211
 ServerID: perforce.ravenbrook.com
 2009-01-23 17:31:51 +00:00
Richard Kistruck
ee6d463a74 Mps br/timing zcoll.c: get() reports collection statistics 
also (in zcoll's tiny script language) tidy up Make() command.

Copied from Perforce
 Change: 167196
 ServerID: perforce.ravenbrook.com
 2009-01-21 16:29:40 +00:00
Richard Kistruck
acceb5c41f Mps br/timing zcoll.c: remove temporary rnd()-test code 
Copied from Perforce
 Change: 167193
 ServerID: perforce.ravenbrook.com
 2009-01-21 14:25:43 +00:00
Richard Kistruck
7598c0ff41 Mps br/timing testlib.c: polish comments and code (no functionality change) 
Copied from Perforce
 Change: 167192
 ServerID: perforce.ravenbrook.com
 2009-01-21 14:19:38 +00:00
Richard Kistruck
b27e4ff179 Mps br/timing (fix warnings "... may be used uninitialized"): 
trace.c, tract.c: initialise variable to NULL

Copied from Perforce
 Change: 167191
 ServerID: perforce.ravenbrook.com
 2009-01-21 14:17:56 +00:00
Richard Kistruck
7cc4b77b6a Mps br/timing testlib.c: fix warning ("... may be used uninitialized") 
Copied from Perforce
 Change: 167188
 ServerID: perforce.ravenbrook.com
 2009-01-21 13:20:20 +00:00
Richard Kistruck
cd3c824df3 Mps br/timing rnd(): 
Use a = 48271.  Use fast implementation.  Much improved verification by rnd_verify().

Copied from Perforce
 Change: 167182
 ServerID: perforce.ravenbrook.com
 2009-01-20 18:23:53 +00:00
Richard Kistruck
e486078a79 Mps br.timing: testlib randomize() was wrong (simply discarding the first 0..32000 values is a throughly defective way to seed the generator). correct it to simply use timestamp (modulo m) as the initial seed, and then iterate 10 times. 
rnd(): add notes, and create rnd_verify() to confirm implementation is correct.

Copied from Perforce
 Change: 167181
 ServerID: perforce.ravenbrook.com
 2009-01-20 15:59:32 +00:00
Richard Kistruck
4fc1ea19fb Mps br/timing zcoll.c: show count of objects made & kept. 
Also (TEMPORARY) investigate poor seeding of rnd().  Aha -- 
it's a full-period generator.  Therefore the current 
randomize is a very poor way to seed it.

Copied from Perforce
 Change: 167152
 ServerID: perforce.ravenbrook.com
 2009-01-13 16:30:27 +00:00
Richard Kistruck
a899c472e1 Mps br/timing zcoll.c: (fix unused variable) 
Copied from Perforce
 Change: 167018
 ServerID: perforce.ravenbrook.com
 2008-12-22 17:33:05 +00:00
Richard Kistruck
262e9ae530 Mps br/timing zcoll.c: (new test) hack in the basics; build it 
Copied from Perforce
 Change: 167017
 ServerID: perforce.ravenbrook.com
 2008-12-22 17:31:14 +00:00
Richard Kistruck
228bdebf81 Mps br/timing zmess.c: (clarity tweaks) 
Copied from Perforce
 Change: 167016
 ServerID: perforce.ravenbrook.com
 2008-12-22 17:30:49 +00:00
Richard Kistruck
3ba57bdcf5 Mps br/timing zmess.c: copyright date 
Copied from Perforce
 Change: 167015
 ServerID: perforce.ravenbrook.com
 2008-12-22 16:51:55 +00:00
Richard Kistruck
6bd96ccc41 Mps br/timing zcoll.c: (new test) create by integrating from zmess.c 
Copied from Perforce
 Change: 167014
 ServerID: perforce.ravenbrook.com
 2008-12-22 16:28:32 +00:00
Richard Kistruck
3260800a3a Mps br/timing traceanc.c review and polish: 
.whybuf.nul: the last char in the array must be NUL (even if there is another NUL earlier in the buffer).
Message Delete functions set sig to invalid, and call MessageFinish.
TraceIdMessagesDestroy uses message Delete functions.

Copied from Perforce
 Change: 167003
 ServerID: perforce.ravenbrook.com
 2008-12-22 14:09:37 +00:00
Richard Kistruck
9457df93db Mps br/timing: zmess.c subsumes and replaces fin1658a.c 
Also improve comments in zmess.c.

Copied from Perforce
 Change: 166994
 ServerID: perforce.ravenbrook.com
 2008-12-19 14:47:54 +00:00
Richard Kistruck
b0790cf19a Mps br/timing: rename z001989a.c as zmess.c 
Copied from Perforce
 Change: 166993
 ServerID: perforce.ravenbrook.com
 2008-12-19 14:27:48 +00:00
Richard Kistruck
538d4990e7 Mps br/timing: rename z001989a.c as zmessco.c 
Copied from Perforce
 Change: 166991
 ServerID: perforce.ravenbrook.com
 2008-12-19 14:08:14 +00:00
Richard Kistruck
179962bb9c Mps br/timing traceanc.c: timca_remote: undo testing-only additions 
to TraceIdMessagesCreate.

Copied from Perforce
 Change: 166960
 ServerID: perforce.ravenbrook.com
 2008-12-17 17:27:01 +00:00
Richard Kistruck
eb1cca20e4 Mps br/timing z001989a.c: timca_remote -- traceidmessagescreate alloc remote control 
In low memory situations, ControlAlloc may be unable to allocate 
memory for GC messages.  This needs to work flawlessly, but is 
hard to test.  To simulate it for testing purposes, do macro-magic 
in traceanc.c, before the definition of TraceIdMessagesCreate.
Test all designed behaviour of TraceIdMessagesCreate failing.
See <design/message-gc#lifecycle>.

Copied from Perforce
 Change: 166959
 ServerID: perforce.ravenbrook.com
 2008-12-17 17:17:55 +00:00
Richard Kistruck
bf3c873639 Mps br/timing (revert previous local-only change @166956) 
Copied from Perforce
 Change: 166957
 ServerID: perforce.ravenbrook.com
 2008-12-16 18:01:18 +00:00
Richard Kistruck
bace1c7bc1 Mps br/timing (local only - must revert) hack trace.c so it does not 
pre-allocate GC messages after the first trace, and verify that 
everything still works, droppedMessages is incremented, and reported 
as non-zero (in diagnostic varieties only) by global.c on 
ArenaDestroy.  Test passes.

Copied from Perforce
 Change: 166956
 ServerID: perforce.ravenbrook.com
 2008-12-16 17:59:50 +00:00