Ove Kaaven wrote:
>
> On Thu, 27 Jul 2000, Marcus Meissner wrote:
>
> > Does not help. The variables are transferred into registers correctly, they
> > are just way too many for them used for egcs-1*
>
> Really uncool.
>
> Log:
> Make a global asm helper function to handle 32-bit RMCB calls,
> so we don't need those complex constraints either.
>
> Index: wine/msdos/dpmi.c
> ===================================================================
> RCS file: /home/wine/wine/msdos/dpmi.c,v
> retrieving revision 1.39
> diff -u -r1.39 dpmi.c
> --- wine/msdos/dpmi.c 2000/07/26 18:04:54 1.39
> +++ wine/msdos/dpmi.c 2000/07/27 23:23:11
> @@ -183,6 +181,66 @@
[...]
> INT_SetRealModeContext((REALMODECALL *)PTR_SEG_OFF_TO_LIN( rmcb->regs_sel,
>rmcb->regs_ofs ), context);
> - ss = SELECTOR_AllocBlock( DOSMEM_MemoryBase(0) +
>(DWORD)(SS_reg(context)<<4), 0x10000, SEGMENT_DATA, FALSE, FALSE );
> + ss = SELECTOR_AllocBlock( DOSMEM_MemoryBase() +
>(DWORD)(SS_reg(context)<<4), 0x10000, SEGMENT_DATA, FALSE, FALSE );
> + esp = ESP_reg(context);
This compiles for me if I put the 0 back as the argument for
DOSMEM_MemoryBase() above (line 263). I'm assuming this change was
unintentional.
Thanks Marcus and Ove for fixing this.
--
James Juran
[EMAIL PROTECTED]
