WebSVN – Kolibri OS – Blame – /contrib/sdk/sources/SDL-1.2.2_newlib/src/hermes/x86p_32.asm

Rev	Author	Line No.	Line
8210	maxcodehac	1	;
		2	; x86 format converters for HERMES
		3	; Some routines Copyright (c) 1998 Christian Nentwich (brn@eleet.mcb.at)
		4	; This source code is licensed under the GNU LGPL
		5	;
		6	; Please refer to the file COPYING.LIB contained in the distribution for
		7	; licensing conditions
		8	;
		9	; Most routines are (c) Glenn Fiedler (ptc@gaffer.org), used with permission
		10	;
		11
		12	BITS 32
		13
9172	turbocat	14	%include "common.inc"
8210	maxcodehac	15
9172	turbocat	16	SDL_FUNC _ConvertX86p32_32BGR888
		17	SDL_FUNC _ConvertX86p32_32RGBA888
		18	SDL_FUNC _ConvertX86p32_32BGRA888
		19	SDL_FUNC _ConvertX86p32_24RGB888
		20	SDL_FUNC _ConvertX86p32_24BGR888
		21	SDL_FUNC _ConvertX86p32_16RGB565
		22	SDL_FUNC _ConvertX86p32_16BGR565
		23	SDL_FUNC _ConvertX86p32_16RGB555
		24	SDL_FUNC _ConvertX86p32_16BGR555
		25	SDL_FUNC _ConvertX86p32_8RGB332
		26
9202	turbocat	27	SDL_FUNC ConvertX86p32_32BGR888
		28	SDL_FUNC ConvertX86p32_32RGBA888
		29	SDL_FUNC ConvertX86p32_32BGRA888
		30	SDL_FUNC ConvertX86p32_24RGB888
		31	SDL_FUNC ConvertX86p32_24BGR888
		32	SDL_FUNC ConvertX86p32_16RGB565
		33	SDL_FUNC ConvertX86p32_16BGR565
		34	SDL_FUNC ConvertX86p32_16RGB555
		35	SDL_FUNC ConvertX86p32_16BGR555
		36	SDL_FUNC ConvertX86p32_8RGB332
		37
		38
8210	maxcodehac	39	SECTION .text
		40
		41	;; _Convert_*
		42	;; Paramters:
		43	;; ESI = source
		44	;; EDI = dest
		45	;; ECX = amount (NOT 0!!! (the _ConvertX86 routine checks for that though))
		46	;; Destroys:
		47	;; EAX, EBX, EDX
		48
		49
9202	turbocat	50	ConvertX86p32_32BGR888:
8210	maxcodehac	51	_ConvertX86p32_32BGR888:
		52
		53	; check short
		54	cmp ecx,BYTE 32
		55	ja .L3
		56
9172	turbocat	57	.L1: ; short loop
8210	maxcodehac	58	mov edx,[esi]
		59	bswap edx
		60	ror edx,8
		61	mov [edi],edx
		62	add esi,BYTE 4
		63	add edi,BYTE 4
		64	dec ecx
		65	jnz .L1
9172	turbocat	66	.L2:
		67	retn
8210	maxcodehac	68
9172	turbocat	69	.L3: ; save ebp
8210	maxcodehac	70	push ebp
		71
		72	; unroll four times
		73	mov ebp,ecx
		74	shr ebp,2
		75
		76	; save count
		77	push ecx
		78
9172	turbocat	79	.L4: mov eax,[esi]
8210	maxcodehac	80	mov ebx,[esi+4]
		81
		82	bswap eax
		83
		84	bswap ebx
		85
		86	ror eax,8
		87	mov ecx,[esi+8]
		88
		89	ror ebx,8
		90	mov edx,[esi+12]
		91
		92	bswap ecx
		93
		94	bswap edx
		95
		96	ror ecx,8
		97	mov [edi+0],eax
		98
		99	ror edx,8
		100	mov [edi+4],ebx
		101
		102	mov [edi+8],ecx
		103	mov [edi+12],edx
		104
		105	add esi,BYTE 16
		106	add edi,BYTE 16
		107
		108	dec ebp
		109	jnz .L4
		110
		111	; check tail
		112	pop ecx
		113	and ecx,BYTE 11b
		114	jz .L6
		115
9172	turbocat	116	.L5: ; tail loop
8210	maxcodehac	117	mov edx,[esi]
		118	bswap edx
		119	ror edx,8
		120	mov [edi],edx
		121	add esi,BYTE 4
		122	add edi,BYTE 4
		123	dec ecx
		124	jnz .L5
		125
9172	turbocat	126	.L6: pop ebp
		127	retn
8210	maxcodehac	128
		129
		130
9202	turbocat	131	ConvertX86p32_32RGBA888:
8210	maxcodehac	132	_ConvertX86p32_32RGBA888:
		133
		134	; check short
		135	cmp ecx,BYTE 32
		136	ja .L3
		137
9172	turbocat	138	.L1: ; short loop
8210	maxcodehac	139	mov edx,[esi]
		140	rol edx,8
		141	mov [edi],edx
		142	add esi,BYTE 4
		143	add edi,BYTE 4
		144	dec ecx
		145	jnz .L1
9172	turbocat	146	.L2:
		147	retn
8210	maxcodehac	148
9172	turbocat	149	.L3: ; save ebp
8210	maxcodehac	150	push ebp
		151
		152	; unroll four times
		153	mov ebp,ecx
		154	shr ebp,2
		155
		156	; save count
		157	push ecx
		158
9172	turbocat	159	.L4: mov eax,[esi]
8210	maxcodehac	160	mov ebx,[esi+4]
		161
		162	rol eax,8
		163	mov ecx,[esi+8]
		164
		165	rol ebx,8
		166	mov edx,[esi+12]
		167
		168	rol ecx,8
		169	mov [edi+0],eax
		170
		171	rol edx,8
		172	mov [edi+4],ebx
		173
		174	mov [edi+8],ecx
		175	mov [edi+12],edx
		176
		177	add esi,BYTE 16
		178	add edi,BYTE 16
		179
		180	dec ebp
		181	jnz .L4
		182
		183	; check tail
		184	pop ecx
		185	and ecx,BYTE 11b
		186	jz .L6
		187
9172	turbocat	188	.L5: ; tail loop
8210	maxcodehac	189	mov edx,[esi]
		190	rol edx,8
		191	mov [edi],edx
		192	add esi,BYTE 4
		193	add edi,BYTE 4
		194	dec ecx
		195	jnz .L5
		196
9172	turbocat	197	.L6: pop ebp
		198	retn
8210	maxcodehac	199
		200
		201
9202	turbocat	202	ConvertX86p32_32BGRA888:
8210	maxcodehac	203	_ConvertX86p32_32BGRA888:
		204
		205	; check short
		206	cmp ecx,BYTE 32
		207	ja .L3
		208
9172	turbocat	209	.L1: ; short loop
8210	maxcodehac	210	mov edx,[esi]
		211	bswap edx
		212	mov [edi],edx
		213	add esi,BYTE 4
		214	add edi,BYTE 4
		215	dec ecx
		216	jnz .L1
9172	turbocat	217	.L2:
		218	retn
8210	maxcodehac	219
9172	turbocat	220	.L3: ; save ebp
8210	maxcodehac	221	push ebp
		222
		223	; unroll four times
		224	mov ebp,ecx
		225	shr ebp,2
		226
		227	; save count
		228	push ecx
		229
9172	turbocat	230	.L4: mov eax,[esi]
8210	maxcodehac	231	mov ebx,[esi+4]
		232
		233	mov ecx,[esi+8]
		234	mov edx,[esi+12]
		235
		236	bswap eax
		237
		238	bswap ebx
		239
		240	bswap ecx
		241
		242	bswap edx
		243
		244	mov [edi+0],eax
		245	mov [edi+4],ebx
		246
		247	mov [edi+8],ecx
		248	mov [edi+12],edx
		249
		250	add esi,BYTE 16
		251	add edi,BYTE 16
		252
		253	dec ebp
		254	jnz .L4
		255
		256	; check tail
		257	pop ecx
		258	and ecx,BYTE 11b
		259	jz .L6
		260
9172	turbocat	261	.L5: ; tail loop
8210	maxcodehac	262	mov edx,[esi]
		263	bswap edx
		264	mov [edi],edx
		265	add esi,BYTE 4
		266	add edi,BYTE 4
		267	dec ecx
		268	jnz .L5
		269
9172	turbocat	270	.L6: pop ebp
		271	retn
8210	maxcodehac	272
		273
		274
		275
		276	;; 32 bit RGB 888 to 24 BIT RGB 888
		277
9202	turbocat	278	ConvertX86p32_24RGB888:
8210	maxcodehac	279	_ConvertX86p32_24RGB888:
		280
		281	; check short
		282	cmp ecx,BYTE 32
		283	ja .L3
		284
9172	turbocat	285	.L1: ; short loop
8210	maxcodehac	286	mov al,[esi]
		287	mov bl,[esi+1]
		288	mov dl,[esi+2]
		289	mov [edi],al
		290	mov [edi+1],bl
		291	mov [edi+2],dl
		292	add esi,BYTE 4
		293	add edi,BYTE 3
		294	dec ecx
		295	jnz .L1
9172	turbocat	296	.L2:
		297	retn
8210	maxcodehac	298
9172	turbocat	299	.L3: ; head
8210	maxcodehac	300	mov edx,edi
		301	and edx,BYTE 11b
		302	jz .L4
		303	mov al,[esi]
		304	mov bl,[esi+1]
		305	mov dl,[esi+2]
		306	mov [edi],al
		307	mov [edi+1],bl
		308	mov [edi+2],dl
		309	add esi,BYTE 4
		310	add edi,BYTE 3
		311	dec ecx
		312	jmp SHORT .L3
		313
9172	turbocat	314	.L4: ; unroll 4 times
8210	maxcodehac	315	push ebp
		316	mov ebp,ecx
		317	shr ebp,2
		318
		319	; save count
		320	push ecx
		321
9172	turbocat	322	.L5: mov eax,[esi] ; first dword eax = [A][R][G][B]
8210	maxcodehac	323	mov ebx,[esi+4] ; second dword ebx = [a][r][g][b]
		324
		325	shl eax,8 ; eax = [R][G][B][.]
		326	mov ecx,[esi+12] ; third dword ecx = [a][r][g][b]
		327
		328	shl ebx,8 ; ebx = [r][g][b][.]
		329	mov al,[esi+4] ; eax = [R][G][B][b]
		330
		331	ror eax,8 ; eax = [b][R][G][B] (done)
		332	mov bh,[esi+8+1] ; ebx = [r][g][G][.]
		333
		334	mov [edi],eax
		335	add edi,BYTE 3*4
		336
		337	shl ecx,8 ; ecx = [r][g][b][.]
		338	mov bl,[esi+8+0] ; ebx = [r][g][G][B]
		339
		340	rol ebx,16 ; ebx = [G][B][r][g] (done)
		341	mov cl,[esi+8+2] ; ecx = [r][g][b][R] (done)
		342
		343	mov [edi+4-3*4],ebx
		344	add esi,BYTE 4*4
		345
		346	mov [edi+8-3*4],ecx
		347	dec ebp
		348
		349	jnz .L5
		350
		351	; check tail
		352	pop ecx
		353	and ecx,BYTE 11b
		354	jz .L7
		355
9172	turbocat	356	.L6: ; tail loop
8210	maxcodehac	357	mov al,[esi]
		358	mov bl,[esi+1]
		359	mov dl,[esi+2]
		360	mov [edi],al
		361	mov [edi+1],bl
		362	mov [edi+2],dl
		363	add esi,BYTE 4
		364	add edi,BYTE 3
		365	dec ecx
		366	jnz .L6
		367
9172	turbocat	368	.L7: pop ebp
		369	retn
8210	maxcodehac	370
		371
		372
		373
		374	;; 32 bit RGB 888 to 24 bit BGR 888
		375
9202	turbocat	376	ConvertX86p32_24BGR888:
8210	maxcodehac	377	_ConvertX86p32_24BGR888:
		378
		379	; check short
		380	cmp ecx,BYTE 32
		381	ja .L3
		382
9172	turbocat	383	.L1: ; short loop
8210	maxcodehac	384	mov dl,[esi]
		385	mov bl,[esi+1]
		386	mov al,[esi+2]
		387	mov [edi],al
		388	mov [edi+1],bl
		389	mov [edi+2],dl
		390	add esi,BYTE 4
		391	add edi,BYTE 3
		392	dec ecx
		393	jnz .L1
9172	turbocat	394	.L2:
		395	retn
8210	maxcodehac	396
9172	turbocat	397	.L3: ; head
8210	maxcodehac	398	mov edx,edi
		399	and edx,BYTE 11b
		400	jz .L4
		401	mov dl,[esi]
		402	mov bl,[esi+1]
		403	mov al,[esi+2]
		404	mov [edi],al
		405	mov [edi+1],bl
		406	mov [edi+2],dl
		407	add esi,BYTE 4
		408	add edi,BYTE 3
		409	dec ecx
		410	jmp SHORT .L3
		411
9172	turbocat	412	.L4: ; unroll 4 times
8210	maxcodehac	413	push ebp
		414	mov ebp,ecx
		415	shr ebp,2
		416
		417	; save count
		418	push ecx
		419
9172	turbocat	420	.L5:
8210	maxcodehac	421	mov eax,[esi] ; first dword eax = [A][R][G][B]
		422	mov ebx,[esi+4] ; second dword ebx = [a][r][g][b]
9172	turbocat	423
8210	maxcodehac	424	bswap eax ; eax = [B][G][R][A]
		425
		426	bswap ebx ; ebx = [b][g][r][a]
		427
		428	mov al,[esi+4+2] ; eax = [B][G][R][r]
		429	mov bh,[esi+4+4+1] ; ebx = [b][g][G][a]
		430
		431	ror eax,8 ; eax = [r][B][G][R] (done)
		432	mov bl,[esi+4+4+2] ; ebx = [b][g][G][R]
		433
		434	ror ebx,16 ; ebx = [G][R][b][g] (done)
		435	mov [edi],eax
		436
		437	mov [edi+4],ebx
		438	mov ecx,[esi+12] ; third dword ecx = [a][r][g][b]
		439
		440	bswap ecx ; ecx = [b][g][r][a]
		441
		442	mov cl,[esi+8] ; ecx = [b][g][r][B] (done)
		443	add esi,BYTE 4*4
		444
		445	mov [edi+8],ecx
		446	add edi,BYTE 3*4
		447
		448	dec ebp
		449	jnz .L5
		450
		451	; check tail
		452	pop ecx
		453	and ecx,BYTE 11b
		454	jz .L7
		455
9172	turbocat	456	.L6: ; tail loop
8210	maxcodehac	457	mov dl,[esi]
		458	mov bl,[esi+1]
		459	mov al,[esi+2]
		460	mov [edi],al
		461	mov [edi+1],bl
		462	mov [edi+2],dl
		463	add esi,BYTE 4
		464	add edi,BYTE 3
		465	dec ecx
		466	jnz .L6
		467
9172	turbocat	468	.L7:
8210	maxcodehac	469	pop ebp
9172	turbocat	470	retn
8210	maxcodehac	471
		472
		473
		474
		475	;; 32 bit RGB 888 to 16 BIT RGB 565
9202	turbocat	476	ConvertX86p32_16RGB565:
8210	maxcodehac	477	_ConvertX86p32_16RGB565:
		478	; check short
		479	cmp ecx,BYTE 16
		480	ja .L3
		481
9172	turbocat	482	.L1: ; short loop
8210	maxcodehac	483	mov bl,[esi+0] ; blue
		484	mov al,[esi+1] ; green
		485	mov ah,[esi+2] ; red
		486	shr ah,3
		487	and al,11111100b
		488	shl eax,3
		489	shr bl,3
		490	add al,bl
		491	mov [edi+0],al
		492	mov [edi+1],ah
		493	add esi,BYTE 4
		494	add edi,BYTE 2
		495	dec ecx
		496	jnz .L1
		497
		498	.L2: ; End of short loop
9172	turbocat	499	retn
8210	maxcodehac	500
		501
9172	turbocat	502	.L3: ; head
8210	maxcodehac	503	mov ebx,edi
		504	and ebx,BYTE 11b
		505	jz .L4
		506
		507	mov bl,[esi+0] ; blue
		508	mov al,[esi+1] ; green
		509	mov ah,[esi+2] ; red
		510	shr ah,3
		511	and al,11111100b
		512	shl eax,3
		513	shr bl,3
		514	add al,bl
		515	mov [edi+0],al
		516	mov [edi+1],ah
		517	add esi,BYTE 4
		518	add edi,BYTE 2
		519	dec ecx
		520
		521	.L4:
		522	; save count
		523	push ecx
		524
		525	; unroll twice
		526	shr ecx,1
		527
		528	; point arrays to end
		529	lea esi,[esi+ecx*8]
		530	lea edi,[edi+ecx*4]
		531
		532	; negative counter
		533	neg ecx
		534	jmp SHORT .L6
		535
		536	.L5:
		537	mov [edi+ecx*4-4],eax
		538	.L6:
		539	mov eax,[esi+ecx*8]
		540
		541	shr ah,2
		542	mov ebx,[esi+ecx*8+4]
		543
		544	shr eax,3
		545	mov edx,[esi+ecx*8+4]
		546
		547	shr bh,2
		548	mov dl,[esi+ecx*8+2]
		549
		550	shl ebx,13
		551	and eax,000007FFh
		552
		553	shl edx,8
		554	and ebx,07FF0000h
		555
		556	and edx,0F800F800h
		557	add eax,ebx
		558
		559	add eax,edx
		560	inc ecx
		561
		562	jnz .L5
		563
		564	mov [edi+ecx*4-4],eax
		565
		566	; tail
		567	pop ecx
		568	test cl,1
		569	jz .L7
		570
		571	mov bl,[esi+0] ; blue
		572	mov al,[esi+1] ; green
		573	mov ah,[esi+2] ; red
		574	shr ah,3
		575	and al,11111100b
		576	shl eax,3
		577	shr bl,3
		578	add al,bl
		579	mov [edi+0],al
		580	mov [edi+1],ah
		581	add esi,BYTE 4
		582	add edi,BYTE 2
		583
		584	.L7:
9172	turbocat	585	retn
8210	maxcodehac	586
		587
		588
		589
		590	;; 32 bit RGB 888 to 16 BIT BGR 565
		591
9202	turbocat	592	ConvertX86p32_16BGR565:
8210	maxcodehac	593	_ConvertX86p32_16BGR565:
		594
		595	; check short
		596	cmp ecx,BYTE 16
		597	ja .L3
		598
9172	turbocat	599	.L1: ; short loop
8210	maxcodehac	600	mov ah,[esi+0] ; blue
		601	mov al,[esi+1] ; green
		602	mov bl,[esi+2] ; red
		603	shr ah,3
		604	and al,11111100b
		605	shl eax,3
		606	shr bl,3
		607	add al,bl
		608	mov [edi+0],al
		609	mov [edi+1],ah
		610	add esi,BYTE 4
		611	add edi,BYTE 2
		612	dec ecx
		613	jnz .L1
9172	turbocat	614	.L2:
		615	retn
8210	maxcodehac	616
9172	turbocat	617	.L3: ; head
8210	maxcodehac	618	mov ebx,edi
		619	and ebx,BYTE 11b
		620	jz .L4
		621	mov ah,[esi+0] ; blue
		622	mov al,[esi+1] ; green
		623	mov bl,[esi+2] ; red
		624	shr ah,3
		625	and al,11111100b
		626	shl eax,3
		627	shr bl,3
		628	add al,bl
		629	mov [edi+0],al
		630	mov [edi+1],ah
		631	add esi,BYTE 4
		632	add edi,BYTE 2
		633	dec ecx
		634
9172	turbocat	635	.L4: ; save count
8210	maxcodehac	636	push ecx
		637
		638	; unroll twice
		639	shr ecx,1
		640
		641	; point arrays to end
		642	lea esi,[esi+ecx*8]
		643	lea edi,[edi+ecx*4]
		644
		645	; negative count
		646	neg ecx
		647	jmp SHORT .L6
		648
9172	turbocat	649	.L5:
8210	maxcodehac	650	mov [edi+ecx*4-4],eax
9172	turbocat	651	.L6:
8210	maxcodehac	652	mov edx,[esi+ecx*8+4]
		653
		654	mov bh,[esi+ecx*8+4]
		655	mov ah,[esi+ecx*8]
		656
		657	shr bh,3
		658	mov al,[esi+ecx*8+1]
		659
		660	shr ah,3
		661	mov bl,[esi+ecx*8+5]
		662
		663	shl eax,3
		664	mov dl,[esi+ecx*8+2]
		665
		666	shl ebx,19
		667	and eax,0000FFE0h
		668
		669	shr edx,3
		670	and ebx,0FFE00000h
		671
		672	and edx,001F001Fh
		673	add eax,ebx
		674
		675	add eax,edx
		676	inc ecx
		677
		678	jnz .L5
		679
		680	mov [edi+ecx*4-4],eax
		681
		682	; tail
		683	pop ecx
		684	and ecx,BYTE 1
		685	jz .L7
		686	mov ah,[esi+0] ; blue
		687	mov al,[esi+1] ; green
		688	mov bl,[esi+2] ; red
		689	shr ah,3
		690	and al,11111100b
		691	shl eax,3
		692	shr bl,3
		693	add al,bl
		694	mov [edi+0],al
		695	mov [edi+1],ah
		696	add esi,BYTE 4
		697	add edi,BYTE 2
		698
9172	turbocat	699	.L7:
		700	retn
8210	maxcodehac	701
		702
		703
		704
		705	;; 32 BIT RGB TO 16 BIT RGB 555
9202	turbocat	706	ConvertX86p32_16RGB555:
8210	maxcodehac	707	_ConvertX86p32_16RGB555:
		708
		709	; check short
		710	cmp ecx,BYTE 16
		711	ja .L3
		712
9172	turbocat	713	.L1: ; short loop
8210	maxcodehac	714	mov bl,[esi+0] ; blue
		715	mov al,[esi+1] ; green
		716	mov ah,[esi+2] ; red
		717	shr ah,3
		718	and al,11111000b
		719	shl eax,2
		720	shr bl,3
		721	add al,bl
		722	mov [edi+0],al
		723	mov [edi+1],ah
		724	add esi,BYTE 4
		725	add edi,BYTE 2
		726	dec ecx
		727	jnz .L1
9172	turbocat	728	.L2:
		729	retn
8210	maxcodehac	730
9172	turbocat	731	.L3: ; head
8210	maxcodehac	732	mov ebx,edi
		733	and ebx,BYTE 11b
		734	jz .L4
		735	mov bl,[esi+0] ; blue
		736	mov al,[esi+1] ; green
		737	mov ah,[esi+2] ; red
		738	shr ah,3
		739	and al,11111000b
		740	shl eax,2
		741	shr bl,3
		742	add al,bl
		743	mov [edi+0],al
		744	mov [edi+1],ah
		745	add esi,BYTE 4
		746	add edi,BYTE 2
		747	dec ecx
		748
9172	turbocat	749	.L4: ; save count
8210	maxcodehac	750	push ecx
		751
		752	; unroll twice
		753	shr ecx,1
		754
		755	; point arrays to end
		756	lea esi,[esi+ecx*8]
		757	lea edi,[edi+ecx*4]
		758
		759	; negative counter
		760	neg ecx
		761	jmp SHORT .L6
		762
9172	turbocat	763	.L5:
8210	maxcodehac	764	mov [edi+ecx*4-4],eax
9172	turbocat	765	.L6:
8210	maxcodehac	766	mov eax,[esi+ecx*8]
		767
		768	shr ah,3
		769	mov ebx,[esi+ecx*8+4]
		770
		771	shr eax,3
		772	mov edx,[esi+ecx*8+4]
		773
		774	shr bh,3
		775	mov dl,[esi+ecx*8+2]
		776
		777	shl ebx,13
		778	and eax,000007FFh
		779
		780	shl edx,7
		781	and ebx,07FF0000h
		782
		783	and edx,07C007C00h
		784	add eax,ebx
		785
		786	add eax,edx
		787	inc ecx
		788
		789	jnz .L5
		790
		791	mov [edi+ecx*4-4],eax
		792
		793	; tail
		794	pop ecx
		795	and ecx,BYTE 1
		796	jz .L7
		797	mov bl,[esi+0] ; blue
		798	mov al,[esi+1] ; green
		799	mov ah,[esi+2] ; red
		800	shr ah,3
		801	and al,11111000b
		802	shl eax,2
		803	shr bl,3
		804	add al,bl
		805	mov [edi+0],al
		806	mov [edi+1],ah
		807	add esi,BYTE 4
		808	add edi,BYTE 2
		809
9172	turbocat	810	.L7:
		811	retn
8210	maxcodehac	812
		813
		814
		815
		816	;; 32 BIT RGB TO 16 BIT BGR 555
9202	turbocat	817	ConvertX86p32_16BGR555:
8210	maxcodehac	818	_ConvertX86p32_16BGR555:
		819
		820	; check short
		821	cmp ecx,BYTE 16
		822	ja .L3
		823
		824
9172	turbocat	825	.L1: ; short loop
8210	maxcodehac	826	mov ah,[esi+0] ; blue
		827	mov al,[esi+1] ; green
		828	mov bl,[esi+2] ; red
		829	shr ah,3
		830	and al,11111000b
		831	shl eax,2
		832	shr bl,3
		833	add al,bl
		834	mov [edi+0],al
		835	mov [edi+1],ah
		836	add esi,BYTE 4
		837	add edi,BYTE 2
		838	dec ecx
		839	jnz .L1
9172	turbocat	840	.L2:
		841	retn
8210	maxcodehac	842
9172	turbocat	843	.L3: ; head
8210	maxcodehac	844	mov ebx,edi
		845	and ebx,BYTE 11b
		846	jz .L4
		847	mov ah,[esi+0] ; blue
		848	mov al,[esi+1] ; green
		849	mov bl,[esi+2] ; red
		850	shr ah,3
		851	and al,11111000b
		852	shl eax,2
		853	shr bl,3
		854	add al,bl
		855	mov [edi+0],al
		856	mov [edi+1],ah
		857	add esi,BYTE 4
		858	add edi,BYTE 2
		859	dec ecx
		860
9172	turbocat	861	.L4: ; save count
8210	maxcodehac	862	push ecx
		863
		864	; unroll twice
		865	shr ecx,1
		866
		867	; point arrays to end
		868	lea esi,[esi+ecx*8]
		869	lea edi,[edi+ecx*4]
		870
		871	; negative counter
		872	neg ecx
		873	jmp SHORT .L6
		874
9172	turbocat	875	.L5:
8210	maxcodehac	876	mov [edi+ecx*4-4],eax
9172	turbocat	877	.L6:
8210	maxcodehac	878	mov edx,[esi+ecx*8+4]
		879
		880	mov bh,[esi+ecx*8+4]
		881	mov ah,[esi+ecx*8]
		882
		883	shr bh,3
		884	mov al,[esi+ecx*8+1]
		885
		886	shr ah,3
		887	mov bl,[esi+ecx*8+5]
		888
		889	shl eax,2
		890	mov dl,[esi+ecx*8+2]
		891
		892	shl ebx,18
		893	and eax,00007FE0h
		894
		895	shr edx,3
		896	and ebx,07FE00000h
		897
		898	and edx,001F001Fh
		899	add eax,ebx
		900
		901	add eax,edx
		902	inc ecx
		903
		904	jnz .L5
		905
		906	mov [edi+ecx*4-4],eax
		907
		908	; tail
		909	pop ecx
		910	and ecx,BYTE 1
		911	jz .L7
		912	mov ah,[esi+0] ; blue
		913	mov al,[esi+1] ; green
		914	mov bl,[esi+2] ; red
		915	shr ah,3
		916	and al,11111000b
		917	shl eax,2
		918	shr bl,3
		919	add al,bl
		920	mov [edi+0],al
		921	mov [edi+1],ah
		922	add esi,BYTE 4
		923	add edi,BYTE 2
		924
9172	turbocat	925	.L7:
		926	retn
8210	maxcodehac	927
		928
		929
		930
		931
		932	;; FROM 32 BIT RGB to 8 BIT RGB (rrrgggbbb)
		933	;; This routine writes FOUR pixels at once (dword) and then, if they exist
		934	;; the trailing three pixels
9202	turbocat	935
		936	ConvertX86p32_8RGB332:
8210	maxcodehac	937	_ConvertX86p32_8RGB332:
		938
9172	turbocat	939	.L_ALIGNED:
8210	maxcodehac	940	push ecx
		941
		942	shr ecx,2 ; We will draw 4 pixels at once
		943	jnz .L1
		944
		945	jmp .L2 ; short jump out of range :(
		946
		947	.L1:
		948	mov eax,[esi] ; first pair of pixels
		949	mov edx,[esi+4]
		950
		951	shr dl,6
		952	mov ebx,eax
		953
		954	shr al,6
		955	and ah,0e0h
		956
		957	shr ebx,16
		958	and dh,0e0h
		959
		960	shr ah,3
		961	and bl,0e0h
		962
		963	shr dh,3
		964
		965	or al,bl
		966
		967	mov ebx,edx
		968	or al,ah
		969
		970	shr ebx,16
		971	or dl,dh
		972
		973	and bl,0e0h
		974
		975	or dl,bl
		976
		977	mov ah,dl
		978
		979
		980
		981	mov ebx,[esi+8] ; second pair of pixels
		982
		983	mov edx,ebx
		984	and bh,0e0h
		985
		986	shr bl,6
		987	and edx,0e00000h
		988
		989	shr edx,16
		990
		991	shr bh,3
		992
		993	ror eax,16
		994	or bl,dl
		995
		996	mov edx,[esi+12]
		997	or bl,bh
		998
		999	mov al,bl
		1000
		1001	mov ebx,edx
		1002	and dh,0e0h
		1003
		1004	shr dl,6
		1005	and ebx,0e00000h
		1006
		1007	shr dh,3
		1008	mov ah,dl
		1009
		1010	shr ebx,16
		1011	or ah,dh
		1012
		1013	or ah,bl
		1014
		1015	rol eax,16
		1016	add esi,BYTE 16
		1017
		1018	mov [edi],eax
		1019	add edi,BYTE 4
		1020
		1021	dec ecx
		1022	jz .L2 ; L1 out of range for short jump :(
		1023
		1024	jmp .L1
		1025	.L2:
		1026
		1027	pop ecx
		1028	and ecx,BYTE 3 ; mask out number of pixels to draw
		1029
		1030	jz .L4 ; Nothing to do anymore
		1031
		1032	.L3:
		1033	mov eax,[esi] ; single pixel conversion for trailing pixels
		1034
		1035	mov ebx,eax
		1036
		1037	shr al,6
		1038	and ah,0e0h
		1039
		1040	shr ebx,16
		1041
		1042	shr ah,3
		1043	and bl,0e0h
		1044
		1045	or al,ah
		1046	or al,bl
		1047
		1048	mov [edi],al
		1049
		1050	inc edi
		1051	add esi,BYTE 4
		1052
		1053	dec ecx
		1054	jnz .L3
		1055
		1056	.L4:
9172	turbocat	1057	retn
		1058
		1059	%ifidn __OUTPUT_FORMAT__,elf32
		1060	section .note.GNU-stack noalloc noexec nowrite progbits
		1061	%endif

Subversion Repositories Kolibri OS

(root)/contrib/sdk/sources/SDL-1.2.2_newlib/src/hermes/x86p_32.asm – Rev 9202