Re: Portable dumper with union type
On Thu, Jan 20, 2000 at 03:17:43AM -0800, Martin Buchholz wrote:
>>>>> "OG" == Olivier Galibert
<galibert(a)pobox.com> writes:
OG> On Wed, Jan 19, 2000 at 04:35:23AM -0500, Olivier Galibert wrote:
>> For now I'd say let's put it on hold, and I am going to review my code
>> for this problem. Most of the (macro-hidden) casts are unnecessay.
OG> And here it is. There are some LISP_TO_VOID and back, but that's
OG> where I thought it was the clearer thing to do. Martin?
I just built with pdump for the first time. To make absolutely sure I
get a core dump, I built using Sun CC (that's the C++ compiler).
I fixed the following errors:
"/net/lasker/xemacs/ws/dev/src/alloc.c", line 4273: Error: Cannot assign void*
to pdump_entry_list_elmt*.
Note that these "errors" are only because you're compiling with a C++
compiler. The code is perfectly valid ISO C code. And frankly, I
tend to consider all the added casts unnecessary noise.
OTOH, being able to get the better warnings a C++ compiler can give
you is a good thing, so I won't veto any patch on this basis.
Fatal error: assertion failed, file
/net/lasker/xemacs/ws/dev/src/lisp.h, line 1292, INTP (obj)
I _think_ that's a side effect of not saving the DEFVAR_LISP. I'm
going to work on it.
Here's my new patch, which supersedes Olivier's patch:
(Oh, I still don't like those LISP_TO_VOIDs, but I don't understand
the code enough to suggest an alternative right now.)
The code does:
if (the lisp object is actually a pointer) then
add <delta> to the pointer
endif
Olivier: you're welcome to apply this patch, but of course the
core
dumps will have to get fixed.
Ok, consider it applied.
(Oh, there are still remaining calls to malloc. Let's get rid of
those.)
(Oh, could you get rid of the shadowed variable warnings? Thanks.)
Ok.
OG.