This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Re : I tell in the resolution of the bug of the stdcall of the progress
- To: <egcs at cygnus dot com>
- Subject: Re: Re : I tell in the resolution of the bug of the stdcall of the progress
- From: "Antonio M. O. Neto" <anmendes at cruzeironet dot com dot br>
- Date: Wed, 17 Jun 1998 17:13:13 -0300
Hello,
Mumit Khan wrote:
>
>Does this patch solve the problem with non-virtual methods only?
>
>Regards,
>Mumit
>
No, this patch solve problem with virtual and non-virtual methods.
The only problem that I located, is with virtual methods, that are generated
with pointers of 12 bytes! (Check in the code generated asm). This still
disables the use of virtual methods in Directx and in some other places of
Windows that pointers come back for classes. Windows assumes that the
pointers are of 4 bytes (the correct, I should increase).
I am working to try to solve this problem and with this to turn compiler
EGCS, ideal for programming in C and C++ for Windows (I don't know how to
program in Fortran, I cannot say anything on him).
To purpose, you verified the bug introduced by the patchs in the
EGCS-19980608, he doesn't get to be compiled (the previous, EGCS-19980531
compel perfectly). I placed the message commenting the bug with the title "
Problem with the EGCS-19980608 with--target=i386-pc-mingw32 " in
egcs-bugs@cygnus.com
Regards,
Antonio Mendes of Oliveira Neto-Brazil (anmendes@cruzeironet.com.br)