|Previous||Table of Contents||Next|
The key to understanding Listing 26.1 is understanding the effect of ANDing the rotated CPU data with the contents of the Bit Mask register. The CPU data is the pattern for the character to be drawn, with bits equal to 1 indicating where character pixels are to appear. The Data Rotate register is set to rotate the CPU data to pixel-align it, since without rotation characters could only be drawn on byte boundaries.
|As I pointed out in Chapter 25, the CPU is perfectly capable of rotating the data itself, and its often the case that thats more efficient. The problem with using the Data Rotate register is that the OUT that sets that register is time-consuming, especially for proportional text, which requires a different rotation for each character. Also, if the code performs full-byte accesses to display memorythat is, if it combines pieces of two adjacent characters into one bytewhenever possible for efficiency, the CPU generally has to do extra work to prepare the data so the VGAs rotator can handle it.|
At the same time that the Data Rotate register is set, the Bit Mask register is set to allow the CPU to modify only that portion of the display memory byte accessed that the pixel-aligned character falls in, so that other characters and/or graphics data wont be wiped out. The result of ANDing the rotated CPU data byte with the contents of the Bit Mask register is a bit mask that allows only the bits equal to 1 in the original character pattern (rotated and masked to provide pixel alignment) to be modified by the CPU; all other bits come straight from the latches. The latches should have previously been loaded from the target address, so the effect of the ultimate synthesized bit mask value is to allow the CPU to modify only those pixels in display memory that correspond to the 1 bits in that part of the pixel-aligned character that falls in the currently addressed byte. The color of the pixels set by the CPU is determined by the contents of the Set/Reset register.
Whew. It sounds complex, but given an understanding of what the data rotator, set/reset, and the bit mask do, its not that bad. One good way to make sense of it is to refer to the original text-drawing program in Listing 25.1 back in Chapter 25, and then see how Listing 26.1 differs from that program.
Its worth noting that the results generated by Listing 26.1 could have been accomplished without write mode 3. Write mode 0 could have been used instead, but at a significant performance cost. Instead of letting write mode 3 rotate the CPU data and AND it with the contents of the Bit Mask register, the CPU could simply have rotated the CPU data directly and ANDed it with the value destined for the Bit Mask register and then set the Bit Mask register to the resulting value. Additionally, enable set/reset could have been forced on for all planes, emulating what write mode 3 does to provide pixel colors.
The write mode 3 approach used in Listing 26.1 can be efficiently extended to drawing large blocks of text. For example, suppose that we were to draw a line of 8-pixel-wide bit-mapped text 40 characters long. We could then set up the bit mask and data rotation as appropriate for the left portion of each bit-aligned character (the portion of each character to the left of the byte boundary) and then draw the left portions only of all 40 characters in write mode 3. Then the bit mask could be set up for the right portion of each character, and the right portions of all 40 characters could be drawn. The VGAs fast rotator would be used to do all rotation, and the only OUTs required would be those required to set the bit mask and data rotation. This technique could well outperform single-character bit-mapped text drivers such as the one in Listing 26.1 by a significant margin. Listing 26.2 illustrates one implementation of such an approach. Incidentally, note the use of the 8×14 ROM font in Listing 26.2, rather than the 8×8 ROM font used in Listing 26.1. There is also an 8×16 font stored in ROM, along with the tables used to alter the 8×14 and 8×16 ROM fonts into 9×14 and 9×16 fonts.
LISTING 26.2 L26-2.ASM
; Program to illustrate high-speed text-drawing operation of ; write mode 3 of the VGA. ; Draws a string of 8×14 characters at arbitrary locations ; without disturbing the background, using VGAs 8×14 ROM font. ; Designed for use with modes 0Dh, 0Eh, 0Fh, 10h, and 12h. ; Runs only on VGAs (in Models 50 & up and IBM Display Adapter ; and 100% compatibles). ; Assembled with MASM ; By Michael Abrash ; stack segment para stack STACK db 512 dup(?) stack ends ; VGA_VIDEO_SEGMENT equ 0a000h ;VGA display memory segment SCREEN_WIDTH_IN_BYTES equ 044ah ;offset of BIOS variable FONT_CHARACTER_SIZE equ 14 ;# bytes in each font char ; ; VGA register equates. ; SC_INDEX equ 3c4h ;SC index register SC_MAP_MASK equ 2 ;SC map mask register index GC_INDEX equ 3ceh ;GC index register GC_SET_RESET equ 0 ;GC set/reset register index GC_ENABLE_SET_RESET equ 1 ;GC enable set/reset register index GC_ROTATE equ 3 ;GC data rotate/logical function ; register index GC_MODE equ 5 ;GC Mode register GC_BIT_MASK equ 8 ;GC bit mask register index ; dseg segment para common DATA TEST_TEXT_ROW equ 69 ;row to display test text at TEST_TEXT_COL equ 17 ;column to display test text at TEST_TEXT_COLOR equ 0fh ;high intensity white TestString label byte db Hello, world!,0 ;test string to print. FontPointer dd ? ;font offset dseg ends ; cseg segment para public CODE assume cs:cseg, ds:dseg start proc near mov ax,dseg mov ds,ax ; ; Select 640×480 graphics mode. ; mov ax,012h int 10h ; ; Set the screen to all blue, using the readability of VGA registers ; to preserve reserved bits. ; mov dx,GC_INDEX mov al,GC_SET_RESET out dx,al inc dx in al,dx and al,0f0h or al,1 ;blue plane only set, others reset out dx,al dec dx mov al,GC_ENABLE_SET_RESET out dx,al inc dx in al,dx and al,0f0h or al,0fh ;enable set/reset for all planes out dx,al mov dx,VGA_VIDEO_SEGMENT mov es,dx ;point to display memory mov di,0 mov cx,8000h ;fill all 32k words mov ax,0ffffh ;because of set/reset, the value ; written actually doesnt matter rep stosw ;fill with blue ; ; Set driver to use the 8×14 font. ; mov ah,11h ;VGA BIOS character generator function, mov al,30h ; return info subfunction mov bh,2 ;get 8×14 font pointer int 10h call SelectFont ; ; Print the test string. ; mov si,offset TestString mov bx,TEST_TEXT_ROW mov cx,TEST_TEXT_COL mov ah,TEST_TEXT_COLOR call DrawString ; ; Wait for a key, then set to text mode & end. ; mov ah,1 int 21h ;wait for a key mov ax,3 int 10h ;restore text mode ; ; Exit to DOS. ; mov ah,4ch int 21h Start endp ; ; Subroutine to draw a text string left-to-right in a linear ; graphics mode (0Dh, 0Eh, 0Fh, 010h, 012h) with 8-dot-wide ; characters. Background around the pixels that make up the ; characters is preserved. ; Font used should be pointed to by FontPointer. ; ; Input: ; AH = color to draw string in ; BX = row to draw string on ; CX = column to start string at ; DS:SI = string to draw ; ; Forces ALU function to move. ; Forces write mode 3. ; DrawString proc near push ax push bx push cx push dx push si push di push bp push ds ; ; Set up set/reset to produce character color, using the readability ; of VGA register to preserve the setting of reserved bits 7-4. ; mov dx,GC_INDEX mov al,GC_SET_RESET out dx,al inc dx in al,dx and al,0f0h and ah,0fh or al,ah out dx,al ; ; Select write mode 3, using the readability of VGA registers ; to leave bits other than the write mode bits unchanged. ; mov dx,GC_INDEX mov al,GC_MODE out dx,al inc dx in al,dx or al,3 out dx,al mov dx,VGA_VIDEO_SEGMENT mov es,dx ;point to display memory ; ; Calculate screen address of byte character starts in. ; push ds ;point to BIOS data segment sub dx,dx mov ds,dx mov di,ds:[SCREEN_WIDTH_IN_BYTES] ;retrieve BIOS ; screen width pop ds mov ax,bx ;row mul di ;calculate offset of start of row push di ;set aside screen width mov di,cx ;set aside the column and cl,0111b ;keep only the column in-byte address shr di,1 shr di,1 shr di,1 ;divide column by 8 to make a byte address add di,ax ;and point to byte ; ; Set up the GC rotation. In write mode 3, this is the rotation ; of CPU data before it is ANDed with the Bit Mask register to ; form the bit mask. Force the ALU function to move. Uses the ; readability of VGA registers to leave reserved bits unchanged. ; mov dx,GC_INDEX mov al,GC_ROTATE out dx,al inc dx in al,dx and al,0e0h or al,cl out dx,al ; ; Set up BH as bit mask for left half, BL as rotation for right half. ; mov bx,0ffffh shr bh,cl neg cl add cl,8 shl bl,cl ; ; Draw all characters, left portion first, then right portion in the ; succeeding byte, using the data rotation to position the character ; across the byte boundary and then using write mode 3 to combine the ; character data with the bit mask to allow the set/reset value (the ; character color) through only for the proper portion (where the ; font bits for the character are 1) of the character for each byte. ; Wherever the font bits for the character are 0, the background ; color is preserved. ; Does not check for case where character is byte-aligned and ; no rotation and only one write is required. ; ; Draw the left portion of each character in the string. ; pop cx ;get back screen width push si push di push bx ; ; Set the bit mask for the left half of the character. ; mov dx,GC_INDEX mov al,GC_BIT_MASK mov ah,bh out dx,ax LeftHalfLoop: lodsb and al,al jz LeftHalfLoopDone call CharacterUp inc di ;point to next character location jmp LeftHalfLoop LeftHalfLoopDone: pop bx pop di pop si ; ; Draw the right portion of each character in the string. ; inc di ;right portion of each character is across ; byte boundary ; ; Set the bit mask for the right half of the character. ; mov dx,GC_INDEX mov al,GC_BIT_MASK mov ah,bl out dx,ax RightHalfLoop: lodsb and al,al jz RightHalfLoopDone call CharacterUp inc di ;point to next character location jmp RightHalfLoop RightHalfLoopDone: ; pop ds pop bp pop di pop si pop dx pop cx pop bx pop ax ret DrawString endp ; ; Draw a character. ; ; Input: ; AL = character ; CX = screen width ; ES:DI = address to draw character at ; CharacterUp proc near push cx push si push di push ds ; ; Set DS:SI to point to font and ES to point to display memory. ; lds si,[FontPointer] ;point to font ; ; Calculate font address of character. ; mov bl,14 ;14 bytes per character mul bl add si,ax ;offset in font segment of character mov bp,FONT_CHARACTER_SIZE dec cx ; -1 because one byte per char CharacterLoop: lodsb ;get character byte mov ah,es:[di] ;load latches stosb ;write character byte ; ; Point to next line of character in display memory. ; add di,cx ; dec bp jnz CharacterLoop ; pop ds pop di pop si pop cx ret CharacterUp endp ; ; Set the pointer to the font to draw from to ES:BP. ; SelectFont proc near mov word ptr [FontPointer],bp ;save pointer mov word ptr [FontPointer+2],es ret SelectFont endp ; cseg ends end start
In this chapter, Ive tried to give you a feel for how write mode 3 works and what it might be used for, rather than providing polished, optimized, plug-it-in-and-go code. Like the rest of the VGAs write path, write mode 3 is a resource that can be used in a remarkable variety of ways, and I dont want to lock you into thinking of it as useful in just one context. Instead, you should take the time to thoroughly understand what write mode 3 does, and then, when you do VGA programming, think about how write mode 3 can best be applied to the task at hand. Because I focused on illustrating the operation of write mode 3, neither listing in this chapter is the fastest way to accomplish the desired result. For example, Listing 26.2 could be made nearly twice as fast by simply having the CPU rotate, mask, and join the bytes from adjacent characters, then draw the combined bytes to display memory in a single operation.
Similarly, Listing 26.1 is designed to illustrate write mode 3 and its interaction with the rest of the VGA as a contrast to Listing 25.1 in Chapter 25, rather than for maximum speed, and it could be made considerably more efficient. If we were going for performance, wed have the CPU not only rotate the bytes into position, but also do the masking by ANDing in software. Even more significantly, we would have the CPU combine adjacent characters into complete, rotated bytes whenever possible, so that only one drawing operation would be required per byte of display memory modified. By doing this, we would eliminate all per-character OUTs, and would minimize display memory accesses, approximately doubling text-drawing speed.
As a final note, consider that non-transparent text could also be accelerated with write mode 3. The latches could be filled with the background (text box) color, set/reset could be set to the foreground (text) color, and write mode 3 could then be used to turn monochrome text bytes written by the CPU into characters on the screen with just one write per byte. There are complications, such as drawing partial bytes, and rotating the bytes to align the characters, which well revisit later on in Chapter 55, while were working through the details of the X-Sharp library. Nonetheless, the performance benefit of this approach can be a speedup of as much as four timesall thanks to the decidedly quirky but surprisingly powerful and flexible write mode 3.
If you take a quick look, youll see that the code in Listing 26.1 uses the readable register feature of the VGA to preserve reserved bits and bits other than those being modified. Older adapters such as the CGA and EGA had few readable registers, so it was necessary to set all bits in a register whenever that register was modified. Happily, all VGA registers are readable, which makes it possible to change only those bits of immediate interest, and, in general, I highly recommend doing exactly that, since IBM (or clone manufacturers) may well someday use some of those reserved bits or change the meanings of some of the bits that are currently in use.
|Previous||Table of Contents||Next|