viscsum.s

优龙2410linux2.6.8内核源代码

1:	addcc		%g3, %o2, %o2		/*  IEU1	Group		*/
	bcs,a,pn	%xcc, 1f		/*  CTI				*/
	 add		%o2, 1, %o2		/*  IEU0			*/
1:	addcc		%g5, %o2, %o2		/*  IEU1	Group		*/
	bcs,a,pn	%xcc, 1f		/*  CTI				*/
	 add		%o2, 1, %o2		/*  IEU0			*/
1:	addcc		%g7, %o2, %o2		/*  IEU1	Group		*/
	bcs,a,pn	%xcc, 3f		/*  CTI				*/
	 add		%o2, 1, %o2		/*  IEU0			*/
3:	cmp		%o1, 0xc0		/*  IEU1	Group		*/
	blu,pn		%icc, 20f		/*  CTI				*/
	 sllx		%o2, 32, %g5		/*  IEU0			*/
#ifdef __KERNEL__
	VISEntry
#endif
	addcc		%o2, %g5, %o2		/*  IEU1	Group		*/
	sub		%o1, 0xc0, %o1		/*  IEU0			*/
	wr		%g0, ASI_BLK_P, %asi	/*  LSU		Group		*/
	membar		#StoreLoad		/*  LSU		Group		*/
	srlx		%o2, 32, %o2		/*  IEU0	Group		*/
	bcs,a,pn	%xcc, 1f		/*  CTI				*/
	 add		%o2, 1, %o2		/*  IEU1			*/
1:	andcc		%o1, 0x80, %g0		/*  IEU1	Group		*/
	bne,pn		%icc, 7f		/*  CTI				*/
	 andcc		%o1, 0x40, %g0		/*  IEU1	Group		*/
	be,pn		%icc, 6f		/*  CTI				*/
	 fzero		%f12			/*  FPA				*/
	fzero		%f14			/*  FPA		Group		*/
	ldda		[%o0 + 0x000] %asi, %f16
	ldda		[%o0 + 0x040] %asi, %f32
	ldda		[%o0 + 0x080] %asi, %f48
	START_THE_TRICK(f12,f16,f18,f20,f22,f24,f26)
	ba,a,pt		%xcc, 3f
6:	sub		%o0, 0x40, %o0		/*  IEU0	Group		*/
	fzero		%f28			/*  FPA				*/
	fzero		%f30			/*  FPA		Group		*/
	ldda		[%o0 + 0x040] %asi, %f32
	ldda		[%o0 + 0x080] %asi, %f48
	ldda		[%o0 + 0x0c0] %asi, %f0
	START_THE_TRICK(f28,f32,f34,f36,f38,f40,f42)
	ba,a,pt		%xcc, 4f
7:	bne,pt		%icc, 8f		/*  CTI				*/
	 fzero		%f44			/*  FPA				*/
	add		%o0, 0x40, %o0		/*  IEU0	Group		*/
	fzero		%f60			/*  FPA				*/
	fzero		%f62			/*  FPA		Group		*/
	ldda		[%o0 - 0x040] %asi, %f0
	ldda		[%o0 + 0x000] %asi, %f16
	ldda		[%o0 + 0x040] %asi, %f32
	START_THE_TRICK(f60,f0,f2,f4,f6,f8,f10)
	ba,a,pt		%xcc, 2f
8:	add		%o0, 0x80, %o0		/*  IEU0	Group		*/
	fzero		%f46			/*  FPA				*/
	ldda		[%o0 - 0x080] %asi, %f48
	ldda		[%o0 - 0x040] %asi, %f0
	ldda		[%o0 + 0x000] %asi, %f16
	START_THE_TRICK(f44,f48,f50,f52,f54,f56,f58)
1:	DO_THE_TRICK(f44,f46,f48,f50,f52,f54,f56,f58,f60,f62,f0,f2,f4,f6,f8,f10,f12,f14)
	ldda		[%o0 + 0x040] %asi, %f32
2:	DO_THE_TRICK(f60,f62,f0,f2,f4,f6,f8,f10,f12,f14,f16,f18,f20,f22,f24,f26,f28,f30)
	ldda		[%o0 + 0x080] %asi, %f48
3:	DO_THE_TRICK(f12,f14,f16,f18,f20,f22,f24,f26,f28,f30,f32,f34,f36,f38,f40,f42,f44,f46)
	ldda		[%o0 + 0x0c0] %asi, %f0
4:	DO_THE_TRICK(f28,f30,f32,f34,f36,f38,f40,f42,f44,f46,f48,f50,f52,f54,f56,f58,f60,f62)
	add		%o0, 0x100, %o0		/*  IEU0	Group		*/
	subcc		%o1, 0x100, %o1		/*  IEU1			*/
	bgeu,a,pt	%icc, 1b		/*  CTI				*/
	 ldda		[%o0 + 0x000] %asi, %f16
	membar		#Sync			/*  LSU		Group		*/
	DO_THE_TRICK(f44,f46,f48,f50,f52,f54,f56,f58,f60,f62,f0,f2,f4,f6,f8,f10,f12,f14)
	END_THE_TRICK(f60,f62,f0,f2,f4,f6,f8,f10,f12,f14,f16,f18,f20,f22,f24,f26,f28,f30)
#ifdef __KERNEL__
	ldub		[%g6 + TI_CURRENT_DS], %g7
#endif
	and		%o1, 0x3f, %o1		/*  IEU0	Group		*/
#ifdef __KERNEL__
	VISExit
	wr		%g7, %g0, %asi
#endif
20:	andcc		%o1, 0xf0, %g1		/*  IEU1	Group		*/
	be,pn		%icc, 23f		/*  CTI				*/
	 and		%o1, 0xf, %o3		/*  IEU0			*/
#ifdef __KERNEL__
22:	sll		%g1, 1, %o4		/*  IEU0	Group		*/
	sethi		%hi(23f), %g7		/*  IEU1			*/
	sub		%g7, %o4, %g7		/*  IEU0	Group		*/
	jmpl		%g7 + %lo(23f), %g0	/*  CTI		Group brk forced*/
	 add		%o0, %g1, %o0		/*  IEU0			*/
#else
22:	rd		%pc, %g7		/*  LSU		Group+4bubbles	*/
	sll		%g1, 1, %o4		/*  IEU0	Group		*/
	sub		%g7, %o4, %g7		/*  IEU0	Group (regdep)	*/
	jmpl		%g7 + (23f - 22b), %g0	/*  CTI		Group brk forced*/
	 add		%o0, %g1, %o0		/*  IEU0			*/
#endif
	CSUM_LASTCHUNK(0xe0)
	CSUM_LASTCHUNK(0xd0)
	CSUM_LASTCHUNK(0xc0)
	CSUM_LASTCHUNK(0xb0)
	CSUM_LASTCHUNK(0xa0)
	CSUM_LASTCHUNK(0x90)
	CSUM_LASTCHUNK(0x80)
	CSUM_LASTCHUNK(0x70)
	CSUM_LASTCHUNK(0x60)
	CSUM_LASTCHUNK(0x50)
	CSUM_LASTCHUNK(0x40)
	CSUM_LASTCHUNK(0x30)
	CSUM_LASTCHUNK(0x20)
	CSUM_LASTCHUNK(0x10)
	CSUM_LASTCHUNK(0x00)
23:	brnz,pn		%o3, 26f		/*  CTI+IEU1	Group		*/
24:	 sllx		%o2, 32, %g1		/*  IEU0			*/
25:	addcc		%o2, %g1, %o0		/*  IEU1	Group		*/
	srlx		%o0, 32, %o0		/*  IEU0	Group (regdep)	*/
	bcs,a,pn	%xcc, 1f		/*  CTI				*/
	 add		%o0, 1, %o0		/*  IEU1			*/
1:	retl					/*  CTI		Group brk forced*/
	 srl		%o0, 0, %o0		/*  IEU0			*/
26:	andcc		%o1, 8, %g0		/*  IEU1	Group		*/
	be,pn		%icc, 1f		/*  CTI				*/
	 ldx		[%o0], %g3		/*  Load			*/
	add		%o0, 8, %o0		/*  IEU0	Group		*/
	addcc		%g3, %o2, %o2		/*  IEU1	Group		*/
	bcs,a,pn	%xcc, 1f		/*  CTI				*/
	 add		%o2, 1, %o2		/*  IEU0			*/
1:	andcc		%o1, 4, %g0		/*  IEU1	Group		*/
	be,a,pn		%icc, 1f		/*  CTI				*/
	 clr		%g2			/*  IEU0			*/
	ld		[%o0], %g2		/*  Load			*/
	add		%o0, 4, %o0		/*  IEU0	Group		*/
	sllx		%g2, 32, %g2		/*  IEU0	Group		*/
1:	andcc		%o1, 2, %g0		/*  IEU1			*/
	be,a,pn		%icc, 1f		/*  CTI				*/
	 clr		%o4			/*  IEU0	Group		*/
	lduh		[%o0], %o4		/*  Load			*/
	add		%o0, 2, %o0		/*  IEU1			*/
	sll		%o4, 16, %o4		/*  IEU0	Group		*/
1:	andcc		%o1, 1, %g0		/*  IEU1			*/
	be,a,pn		%icc, 1f		/*  CTI				*/
	 clr		%o5			/*  IEU0	Group		*/
	ldub		[%o0], %o5		/*  Load			*/
	sll		%o5, 8, %o5		/*  IEU0	Group		*/
1:	or		%g2, %o4, %o4		/*  IEU1			*/
	or		%o5, %o4, %o4		/*  IEU0	Group (regdep)	*/
	addcc		%o4, %o2, %o2		/*  IEU1	Group (regdep)	*/
	bcs,a,pn	%xcc, 1f		/*  CTI				*/
	 add		%o2, 1, %o2		/*  IEU0			*/
1:	ba,pt		%xcc, 25b		/*  CTI		Group		*/
	 sllx		%o2, 32, %g1		/*  IEU0			*/
21:	srl		%o2, 0, %o2		/*  IEU0	Group		*/
	cmp		%o1, 0			/*  IEU1			*/
	be,pn		%icc, 24b		/*  CTI				*/
	 andcc		%o1, 4, %g0		/*  IEU1	Group		*/
	be,a,pn		%icc, 1f		/*  CTI				*/
	 clr		%g2			/*  IEU0			*/
	lduh		[%o0], %g3		/*  Load			*/
	lduh		[%o0+2], %g2		/*  Load	Group		*/
	add		%o0, 4, %o0		/*  IEU0	Group		*/
	sllx		%g3, 48, %g3		/*  IEU0	Group		*/
	sllx		%g2, 32, %g2		/*  IEU0	Group		*/
	or		%g3, %g2, %g2		/*  IEU0	Group		*/
1:	andcc		%o1, 2, %g0		/*  IEU1			*/
	be,a,pn		%icc, 1f		/*  CTI				*/
	 clr		%o4			/*  IEU0	Group		*/
	lduh		[%o0], %o4		/*  Load			*/
	add		%o0, 2, %o0		/*  IEU1			*/
	sll		%o4, 16, %o4		/*  IEU0	Group		*/
1:	andcc		%o1, 1, %g0		/*  IEU1			*/
	be,a,pn		%icc, 1f		/*  CTI				*/
	 clr		%o5			/*  IEU0	Group		*/
	ldub		[%o0], %o5		/*  Load			*/
	sll		%o5, 8, %o5		/*  IEU0	Group		*/
1:	or		%g2, %o4, %o4		/*  IEU1			*/
	or		%o5, %o4, %o4		/*  IEU0	Group (regdep)	*/
	addcc		%o4, %o2, %o2		/*  IEU1	Group (regdep)	*/
	bcs,a,pn	%xcc, 1f		/*  CTI				*/
	 add		%o2, 1, %o2		/*  IEU0			*/
1:	ba,pt		%xcc, 25b		/*  CTI		Group		*/
	 sllx		%o2, 32, %g1		/*  IEU0			*/

	/* When buff is byte aligned and len is large, we backoff to
	 * this really slow handling.  The issue is that we cannot do
	 * the VIS stuff when buff is byte aligned as unaligned.c will
	 * not fix it up.
	 */
csump_really_slow:
	mov	%o0, %o3
	mov	%o1, %o4
	cmp	%o1, 0
	ble,pn	%icc, 9f
	 mov	0, %o0
	andcc	%o3, 1, %o5
	be,pt	%icc, 1f
	 sra	%o4, 1, %g3
	add	%o1, -1, %o4
	ldub	[%o3], %o0
	add	%o3, 1, %o3
	sra	%o4, 1, %g3
1:
	cmp	%g3, 0
	be,pt	%icc, 3f
	 and	%o4, 1, %g2
	and	%o3, 2, %g2
	brz,a,pt %g2, 1f
	 sra	%g3, 1, %g3
	add	%g3, -1, %g3
	add	%o4, -2, %o4
	lduh	[%o3], %g2
	add	%o3, 2, %o3
	add	%o0, %g2, %o0
	sra	%g3, 1, %g3
1:
	cmp	%g3, 0
	be,pt	%icc, 2f
	 and	%o4, 2, %g2
1:
	ld	[%o3], %g2
	addcc	%o0, %g2, %o0
	addx	%o0, %g0, %o0
	addcc	%g3, -1, %g3
	bne,pt	%icc, 1b
	 add	%o3, 4, %o3
	srl	%o0, 16, %o1
	sethi	%hi(64512), %g2
	or	%g2, 1023, %g2
	and	%o0, %g2, %g3
	add	%g3, %o1, %g3
	srl	%g3, 16, %o0
	and	%g3, %g2, %g2
	add	%g2, %o0, %g3
	sll	%g3, 16, %g3
	srl	%g3, 16, %o0
	and	%o4, 2, %g2
2:
	cmp	%g2, 0
	be,pt	%icc, 3f
	 and	%o4, 1, %g2
	lduh	[%o3], %g2
	add	%o3, 2, %o3
	add	%o0, %g2, %o0
	and	%o4, 1, %g2
3:
	cmp	%g2, 0
	be,pt	%icc, 1f
	 srl	%o0, 16, %o1
	ldub	[%o3], %g2
	sll	%g2, 8, %g2
	add	%o0, %g2, %o0
	srl	%o0, 16, %o1
1:
	sethi	%hi(64512), %g2
	or	%g2, 1023, %g2
	cmp	%o5, 0
	and	%o0, %g2, %g3
	add	%g3, %o1, %g3
	srl	%g3, 16, %o0
	and	%g3, %g2, %g2
	add	%g2, %o0, %g3
	sll	%g3, 16, %g3
	srl	%g3, 16, %o0
	srl	%g3, 24, %g3
	and	%o0, 255, %g2
	sll	%g2, 8, %g2
	bne,pt	%icc, 1f
	 or	%g3, %g2, %g2
9:
	mov	%o0, %g2
1:
	addcc	%g2, %o2, %g2
	addx	%g2, %g0, %g2
	retl
	 srl	%g2, 0, %o0