ceriel
d33da47a2a
Do not use '#endif/#else xxx'; it is not allowed for ANSI C
1991-12-17 15:28:58 +00:00
ceriel
53aa761bcd
made more acceptable for ANSI C
1991-11-26 14:48:27 +00:00
ceriel
f9c4a5e37a
do not do lintlib installation on 'make install', also make sure
...
that every proto.make has a lintlib entry
1991-10-09 19:02:55 +00:00
ceriel
c2caead0c8
New installation mechanism
1991-08-26 16:50:25 +00:00
ceriel
df06ee22f2
Changed mechanism for defining size_type
1991-07-12 10:16:11 +00:00
ceriel
b6cd19f7ba
Several fixes
1990-11-01 09:33:51 +00:00
ceriel
e508fbd614
Fix: did not work with ALIGNMENT < MIN_SIZE
1990-03-21 16:45:39 +00:00
eck
c78b5a2bf8
fixed wrong assertion
1990-03-21 13:33:14 +00:00
ceriel
6457b9560b
several fixes in the CHECK part
1990-01-18 17:11:41 +00:00
ceriel
19320b2f5d
publicdata is extern for development version
1989-12-18 13:18:43 +00:00
ceriel
29e74f51f2
bug fix: the in_store test did not work when NON_STANDARD was defined,
...
because in this case, the bit tested resided in user area
1989-11-30 17:29:00 +00:00
ceriel
d717a5cac9
Corrected assertion
1989-11-30 15:12:11 +00:00
ceriel
5b8c6f2943
changed commented #defines to #undefs
1989-11-30 14:25:40 +00:00
ceriel
cc56ae65dd
if malloc can return 0 on a 0 argument, free should also accept 0
1989-10-31 09:44:21 +00:00
ceriel
0eca1bf583
corrected assertions
1989-10-24 12:58:23 +00:00
ceriel
30260ca8e7
made more ANSI conformant
1989-08-09 16:25:50 +00:00
ceriel
e4df2e1d58
fixed: arguments to fputs were the wrong way around
1989-08-09 16:22:59 +00:00
ceriel
4d98ce5fa5
Added some assertions
1989-07-31 12:55:33 +00:00
ceriel
ebf61935c5
fixed a problem with the store: we sometimes lost some memory
1989-07-17 15:13:09 +00:00
ceriel
da3000aa58
Avoid I/O buffering when using the checking malloc
1989-06-19 11:10:58 +00:00
ceriel
068ffbc391
make realloc behave like most Unix realloc versions when handed a
...
null pointer
1989-03-29 12:58:18 +00:00
ceriel
73ac1d53ec
Made easier to use other compiler
1989-02-16 16:33:29 +00:00
ceriel
b3702e14a3
Try to allocate to the next power of 2, instead of just aligning
...
with GRABSIZE; This way, malloc(1000000) followed by free() followed by
another malloc(1000000) will result in the same block being allocated,
because first_present will now find it
1989-02-10 09:18:19 +00:00
ceriel
472bc43825
try another SBRK
1988-07-11 14:06:11 +00:00
ceriel
7b873fea90
improved; delayed frees did not work anyway
1988-07-05 11:50:11 +00:00
ceriel
734e402508
better check if a block was already freed
1988-01-14 11:35:07 +00:00
ceriel
1f2fa05492
first free-ing and then reallocing did not work properly.
...
Now, free does nothing, except remembering that a block was freed.
malloc, realloc and free check for this.
1987-08-14 11:52:17 +00:00
ceriel
b0df114710
Initial revision
1987-04-13 10:32:47 +00:00
ceriel
7e9733350e
Added Rcs Id
1987-03-10 09:24:02 +00:00
ceriel
ea8e8f95b4
Added copyright notice
1987-03-09 15:15:03 +00:00
ceriel
12869eaa07
Initial revision
1987-01-06 11:25:09 +00:00
