diff options
| author | Henry Stiles <henry.stiles@artifex.com> | 2006-09-19 19:21:29 +0000 |
|---|---|---|
| committer | Henry Stiles <henry.stiles@artifex.com> | 2006-09-19 19:21:29 +0000 |
| commit | 14787aeb0ea80d00e6d3589e530938b13dc81dd6 (patch) | |
| tree | 2412feb197ebaa7c605c7b699ec2edfdbd46a67f /pl/plalloc.c | |
| parent | 272b2ea600287a74db342c84ff3140dbc29d8513 (diff) | |
Warning: These changes have not been completely tested or are
incomplete, so the system may be unstable.
gs 8.55 merge regression;
the new path code does not update the current point when using the low
level gx_path_add_point(), we now use gs_moveto() and avoid the mess
of manually updating current point. In fact this rectangle path
procedure and all of xl should be updated to use exclusively gs_*
procs and not gx_* to avoid this type of problem in the future. Time
does not allow that now.
C30[67] p 1, 2, 6
C318 missing data and no current point errors throughout.
similar problems in C318, C326, C407, and C409.
pxparse.c, gdevdflt.c: replace return_error() with just an error so
the -Z# is useful.
git-svn-id: http://svn.ghostscript.com/ghostpcl/trunk/ghostpcl@2587 06663e23-700e-0410-b217-a244a6096597
Diffstat (limited to 'pl/plalloc.c')
| -rw-r--r-- | pl/plalloc.c | 5 |
1 files changed, 0 insertions, 5 deletions
diff --git a/pl/plalloc.c b/pl/plalloc.c index 54be139ae..287cf6cfc 100644 --- a/pl/plalloc.c +++ b/pl/plalloc.c @@ -207,11 +207,6 @@ pl_alloc(gs_memory_t *mem, uint size, gs_memory_type_ptr_t type, client_name_t c { uint minsize, newsize; - /* NB apparently there is code floating around that does 0 size - mallocs. sigh. */ - if ( size == 0 ) - return 0; - /* use 2 starting machine words for size and type - assumes malloc() returns on max boundary and first 2 words will hold two longs. Doesn't check for overflow - malloc will fail for |