{ x86 surface clear routines for HERMES Copyright (c) 1998 Christian Nentwich (c.nentwich@cs.ucl.ac.uk) This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version with the following modification: As a special exception, the copyright holders of this library give you permission to link this library with independent modules to produce an executable, regardless of the license terms of these independent modules,and to copy and distribute the resulting executable under terms of your choice, provided that you also meet, for each linked independent module, the terms and conditions of the license of that module. An independent module is a module which is not derived from or based on this library. If you modify this library, you may extend this exception to your version of the library, but you are not obligated to do so. If you do not wish to do so, delete this exception statement from your version. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA (04/10/99) Modified ClearX86_8 } {$ASMMODE att} { -------------------------------------------------------------------------- HermesClearInterface (ebp+..) 0: char8 *dest 4: int32 value 8: unsigned int width (already checked to be >0!) 12: unsigned int height (already checked to be >0!) 16: int add } procedure ClearX86_32(hci: PHermesClearInterface); cdecl; assembler; asm pushl %ebp movl 8(%ebp),%ebp movl (%ebp),%edi // destination movl 4(%ebp),%eax // pixel value movl 12(%ebp),%edx // height .balign 4 .L_y: movl 8(%ebp),%ecx rep stosl addl 16(%ebp),%edi decl %edx jnz .L_y popl %ebp end; procedure ClearX86_24(hci: PHermesClearInterface); cdecl; assembler; asm end; procedure ClearX86_16(hci: PHermesClearInterface); cdecl; assembler; asm pushl %ebp movl 8(%ebp),%ebp movl (%ebp),%edi // destination movl 4(%ebp),%eax // pixel value movl 12(%ebp),%edx // height movl %eax,%ebx shll $16,%eax // Duplicate pixel value andl $0x0ffff,%ebx orl %ebx,%eax .L_y: movl 8(%ebp),%ecx testl $3,%edi // Check if destination is aligned mod 4 jz .L_aligned movw %ax,(%edi) // otherwise write one pixel addl $2,%edi decl %ecx jz .L_endline .L_aligned: shrl $1,%ecx rep stosl jnc .L_endline movw %ax,(%edi) addl $2,%edi .L_endline: addl 16(%ebp),%edi decl %edx jnz .L_y popl %ebp end; procedure ClearX86_8(hci: PHermesClearInterface); cdecl; nostackframe; assembler; asm pushl %ebp movl %esp,%ebp movl 8(%ebp),%ebp movl 4(%ebp),%eax // pixel value movl 12(%ebp),%edx // height movb %al,%ah movl (%ebp),%edi // destination movl %eax,%ecx shll $16,%eax // Put the byte pixel value in all four bytes andl $0x0ffff,%ecx // of eax movl 8(%ebp),%ebx orl %ecx,%eax cmpl $5,%ebx // removes need for extra checks later jbe .L_short_y .balign 4 .L_y: testl $3,%edi jz .L_aligned movl %edi,%ecx negl %ecx andl $3,%ecx subl %ecx,%ebx rep stosb .L_aligned: movl %ebx,%ecx shrl $2,%ecx andl $3,%ebx rep stosl movl %ebx,%ecx rep stosb addl 16(%ebp),%edi decl %edx movl 8(%ebp),%ebx jnz .L_y popl %ebp ret // Short loop .balign 4 .L_short_y: movl %ebx,%ecx rep stosb addl 16(%ebp),%edi decl %edx jnz .L_short_y popl %ebp end; { ClearX86_8 version 2, Im not sure wheather this is faster or not... too many jumps could confuse the CPU branch guessing } procedure ClearX86_8_2(hci: PHermesClearInterface); cdecl; nostackframe; assembler; asm pushl %ebp movl %esp,%ebp movl 8(%ebp),%ebp movl 4(%ebp),%eax // pixel value movl 12(%ebp),%edx // height movb %al,%ah movl (%ebp),%edi // destination movl %eax,%ecx shll $16,%eax // Put the byte pixel value in all four bytes andl $0x0ffff,%ecx // of eax movl 8(%ebp),%ebx orl %ecx,%eax cmpl $5,%ebx // removes need for extra checks in main loop jbe .L_short_y .balign 4 .L_y: testl $3,%edi jz .L_aligned movl %edi,%ecx negl %ecx andl $3,%ecx movb %al,(%edi) subl %ecx,%ebx incl %edi decl %ecx jz .L_aligned movb %al,(%edi) incl %edi decl %ecx jz .L_aligned movb %al,(%edi) incl %edi .L_aligned: movl %ebx,%ecx shrl $2,%ecx andl $3,%ebx rep stosl jz .L_endline // ebx movb %al,(%edi) // Write remaining (1,2 or 3) pixels incl %edi decl %ebx jz .L_endline movb %al,(%edi) incl %edi decl %ebx jz .L_endline movb %al,(%edi) incl %edi decl %ebx jz .L_endline movb %al,(%edi) incl %edi .L_endline: addl 16(%ebp),%edi decl %edx movl 8(%ebp),%ebx jnz .L_y popl %ebp ret // Short loop .balign 4 .L_short_y: movl %ebx,%ecx rep stosb addl 16(%ebp),%edi decl %edx jnz .L_short_y popl %ebp end;