qf/quakeforge
0
0
Fork 0
mirror of https://git.code.sf.net/p/quake/quakeforge synced 2024-11-27 06:34:11 +00:00
Code Issues Projects Releases Packages Wiki Activity
12400 commits 17 branches 20 tags 29 MiB
Commit graph

86 commits

Author SHA1 Message Date
Bill Currie
fc8cc0b703 [zone] Be more informative when catching memory errors 
The registered error handler is given a chance to report the error, and
the corrupted id is printed.
 2022-02-12 01:57:03 +09:00
Bill Currie
2fcec6e5cb [zone] Move heap check to Z_TagMalloc 
Since Z_Malloc uses Z_TagMalloc to do the work, this ensures the check
is always run.

Also, add the check to Z_Realloc when it needs to adjust an existing
block.
 2022-02-01 14:56:47 +09:00
Bill Currie
1be220559a [zone] Add function to retrieve memory usage 2021-12-27 20:12:03 +09:00
Bill Currie
6a7f78485a [zone] Use memccpy instead of strncpy 
Probably the best way to tell gcc I know what I'm doing.
 2021-12-24 06:45:13 +09:00
Bill Currie
2b332cfe4a [zone] Clean up some magic numbers 2021-12-24 06:45:13 +09:00
Bill Currie
eee25d21ba [zone] Clean out unnecessary memsets 
As the cache blocks are always filled in with a memcpy, there's no need
to zero out the whole allocation, and the header is initialized anyway.
 2021-12-24 06:45:13 +09:00
Bill Currie
ff40563fc0 [zone] Squeeze cache_system_t to fit into 64 bytes 
The cache system pointers are now indices into an array of
cache_system_t blocks, allowing them to be 32 bits instead of 64, thus
allowing cache_system_t to fit into a single CPU cache line. This still
gives and effective 38 bits (256GB) of addressing for cache/hunk. This
does mean that the cache functions cannot work with more than 256GB, but
should that become a problem, cache and working hunking hunk can be
separate, and it should be possible to have multiple cache systems.
 2021-12-24 06:45:13 +09:00
Bill Currie
39c020908c [zone] Return the pointer to the global hunk 
In trying to reduce unnecessary memsets when loading files, I found that
Hunk_RawAllocName already had nonnull on it, so quakefs needed to know
the hunk it was to use. It seemed much better to to go this way (first
step in what is likely to be a lengthy process) than backtracking a
little and removing the nonnull attribute.
 2021-12-24 06:45:13 +09:00
Bill Currie
fe98a513bc [util] Add a function to check hunk pointers 
Its only real utility is to check that a pointer is not pointing into
freed space.
 2021-07-29 15:27:48 +09:00
Bill Currie
8f376a48f8 [util] Add raw versions of hunk alloc and free 
They do not clear memory and thus are good for situations where speed is
more critical.
 2021-07-29 11:44:10 +09:00
Bill Currie
54604d9aa2 [util] Make hunk (optionally) thread-safe 
For now, the functions check for a null hunk pointer and use the global
hunk (initialized via Memory_Init) if necessary. However, Hunk_Init is
available (and used by Memory_Init) to create a hunk from any arbitrary
memory block. So long as that block is 64-byte aligned, allocations
within the hunk will remain 64-byte aligned.
 2021-07-29 11:43:27 +09:00
Bill Currie
c02fcee58a [util] Make zone functions 64-bit clean 
This allows nq and qw clients/servers to use over 2GB of memory if
necessary.
 2021-07-26 15:43:57 +09:00
Bill Currie
5f93c115ff [util] Make developer flag names easier to manage 
They're now an enum, and the flag part of the name is all lowercase, but
now the flag definitions and names list will never get out of sync.
 2021-03-29 22:38:47 +09:00
Bill Currie
5d9bf32d3c Merge branch 'vulkan' 2021-03-23 12:31:58 +09:00
Bill Currie
5d1d85f72e [util] Resurrect Hunk_Print and fix some errors 
It turned out that Hunk_HighAlloc was not creating a hunk header
(ancient bug by me), and I cleaned up a bunch of name-size issues, along
the way.
 2021-02-03 13:21:08 +09:00
Bill Currie
0bfb60775e [util] Ensure hunk allocs are cache alligned 
This doesn't seem to make much difference in the vulkan renderer, but it
certainly doesn't hurt.
 2021-02-03 13:19:19 +09:00
Bill Currie
4f8a06ddd3 [build] Fix a pile of gcc 10 issues 
gcc got stricter about array accesses, complicating progs macros, and
much better at detecting buffer overflows.
 2020-12-28 18:58:51 +09:00
Bill Currie
ab04a1915e [build] Fix a pile of gcc 10 issues 
gcc got stricter about array accesses, complicating progs macros, and
much better at detecting buffer overflows.
 2020-12-21 14:14:29 +09:00
Bill Currie
e2d1a0f7d2 [ruamoko] Tag object memory with the class name 
This makes it a lot easier to see what's leaking.
 2020-03-31 16:43:10 +09:00
Bill Currie
4b503c7f72 Store the requested size in the block. 
This necessitated disabling the id2 padding, but it's only commented out
incase there's more growth. Now the (compiler) error in -addObjectNoRetain
is caught ealier.
 2012-11-19 21:50:42 +09:00
Bill Currie
8ee98e1ef5 Rename memblock_s's size to block_size. 
In order to improve the pointer checking, I plan on adding a size field to
represent the requested size.
 2012-11-19 21:35:27 +09:00
Bill Currie
49ad301d3d Add improved memory checking to progs. 
With pr_boundscheck >= 2, pointer access will be checked against allocated
blocks (qfvalgrind?:). Currently extremely basic, but it seems to work.
 2012-11-19 20:03:21 +09:00
Bill Currie
23a38738fc Massive whitespace cleanup. 
Lots of trailing whitespace and otherwise blank lines.
 2012-05-22 08:23:22 +09:00
Bill Currie
bc1b483525 Nuke the rcsid stuff. 
It's pretty useless in git.
 2012-04-22 10:56:32 +09:00
Bill Currie
be9f05ea14 Catch progs zone errors and give a stack dump. 2012-02-06 12:48:22 +09:00
Bill Currie
e3f76de898 Avoid a division by zero. 
Doing a cache profile on an unused cache wasn't nice.
 2012-01-23 16:46:52 +09:00
Bill Currie
cc5140e3a1 Move min/max/bound and field_offset into better locations. 
min/max bound into mathlib.h (they /are/ math functions, after all) and
field_offset into qtypes.h.
 2011-12-24 10:04:33 +09:00
Bill Currie
87390f53b0 Z_Realloc() should zero-fill the expandded part of the mem 
Patch from Ozkan Sezer (sezero). Thank you.
 2011-12-19 09:45:46 +09:00
Bill Currie
d3771dfeba Resolve some inlining issues in recent gcc. 2011-05-04 09:31:18 +09:00
Bill Currie
7c113a12db Make zone errors a little more useful. 2011-03-20 15:50:09 +09:00
Bill Currie
4432bc0bef Make the offset in zone dumps more useful. 2011-03-08 22:44:56 +09:00
Bill Currie
9b9a60aa89 Make the -mem suggestion a little more realistic. 
16M hasn't been enough for a long time.
 2010-12-13 19:11:48 +09:00
Bill Currie
185e21e4e4 Tweak some developer cvar usage. 2010-11-27 16:53:17 +09:00
Bill Currie
8c4fe2f844 Rename Sys_DPrintf to Sys_MaskPrintf. 
We now have finer runtime control over what gets printed. Need to do a
SYS_DEV audit, creating new masks as apropriate.
 2010-11-26 16:19:26 +09:00
Bill Currie
d6ceaef7d2 Whee, 64bit cpus and sizeof. 2010-11-24 20:42:07 +09:00
Bill Currie
3878b76cc5 Start using autorelease. 
Much more testing is required (oh, but for qc-valgrind), but there is now
a ~945kB block of free data in the menu progs heap :).

Also, correct the printed size of the memory block to not include the
block header size.
 2010-11-24 17:01:18 +09:00
Bill Currie
204c509a43 sezero's "Bad old_size value in Z_Realloc()" patch. 2010-11-21 10:06:56 +09:00
Bill Currie
669771681a split up and modernize configure.ac 
Most of the guts of configure.ac have been moved to config.d and are then
brought in by m4_include. This will make maintaining configure.ac much easier.

Also drop use of PROGRAM and VERSION, using PACKAGE_NAME, PACKAGE_VERSION, and
on occasion, PACKAGE_STRING instead, and clean out some old files we no longer
need.
 2010-08-19 15:01:43 +09:00
Bill Currie
997102fea8 audit the usage of "only" 
There are still a few iffy places (notably around certain prepositions), but
the relevant sentences are now much easier to read.
 2010-01-13 06:42:26 +00:00
Bill Currie
542c1bb0d9 new cache scheme now works properly. next step: shuffle blocks to the top of the heap before locking them 2007-03-30 07:41:17 +00:00
Bill Currie
4cab5b90e6 new cache allocation scheme. still a couple warts, but at least now there's something to work with, and more importantly, there won't be cache movement anywhere near as often. 2007-03-28 13:09:49 +00:00
Bill Currie
e5b972529b take a chainsaw to the existing code, cleaning out anything that's obviously unnecessary. preparation for the cache re-write 2007-03-28 09:52:01 +00:00
Bill Currie
7add93e3ce forgot to make Cache_ReadLock visible 2007-03-23 14:22:59 +00:00
Bill Currie
818756ace5 add a debug function 2007-03-23 12:36:55 +00:00
Bill Currie
81a57bb3fa patch from phrosty for vc2005 support. includes fixes for a bunch of gcc-isms that crept in over the years. 2007-03-22 23:20:57 +00:00
Bill Currie
a3ceb050a2 be a little more verbose 2007-03-18 12:53:59 +00:00
Bill Currie
99c0954b47 the big dso visibility patch :). Sure, we have to have unique names for static builds, but with controlled visibitly we should get faster program loads (although this isn't C++, so it's not as bad) and complex plugins are cleaner. 2007-03-10 12:00:59 +00:00
Bill Currie
3fb03fc2be hah, should have been using __attribute__((used)) all that time (rather 
than __attribute__((unused))). fixes the missing console in -x11
 2005-08-04 15:27:09 +00:00
Bill Currie
23f71485d1 more tree recovering. This includes Despair's anisotropy support work. 2005-01-02 14:23:20 +00:00
Bill Currie
2488465bc9 resurrect Z_Print. 
This is an imperfect revision of history.
 2004-11-14 00:49:00 +00:00