[AD] Allegro 3.9.38 and gcc 3.0

[ Thread Index | Date Index | More lists.liballeg.org/allegro-developers Archives ]


Hi All!

When I talked about uploading Allegro 3.12 patches I meant the following:
Source Forge have plenty of space, why not uploading the patches so if any
user needs it the rest can indicate which file to download?

Now about 3.9.38:

1) I'm attaching a small set of patches. They fixes it:
 a) gcc 3.x ignores volatile modifier on return values and gives a warning.
 b) It doesn't like array[0]={...}, but accepts array[]={...}.
 c) -mpentium is deprecated. I changed it by -march=pentium but I don't know

    if it was the original idea.
 d) The following is invalid: condition ? object_type_1 : object_type2

2) I don't agree with the concept of "debug" library used by the makefile.
You can't use an optimized library and debug an unoptimized one. You'll be
simply debugging another code. And in fact I experimented it while testing.

3) I think I hit another bug in gcc 3.x (I had problems with the C++ streams

last week). The library can't be compiled with -fomit-frame-pointer because
the ustrzcpy() function is wrongly generated (gcc clobbers EBP and then uses

it to get the value of another parameter ... Kbum ;-).
  As this function is used during initialization no Allegro program worked.
  Now related to (2), when I recompiled for debug the programs stopped
crashing. I think here the bug is in gcc, but in other cases could be in the

code (I saw it many times).

SET

P.S. I think the patches are in DOS format, be careful if they are applied
in UNIX.

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: set@xxxxxxxxxx set@xxxxxxxxxx
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013


Attachment: all.3938.diff.gz
Description: GNU Zip compressed data



Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/