/netbsd-current/external/lgpl3/gmp/dist/mpn/x86/pentium/mmx/ |
H A D | rshift.asm | 44 C It takes 6 mmx instructions to process 2 limbs, making 1.5 cycles/limb, 125 C eax counter, limbs, negative 277 C eax counter, limbs, negative 316 C eax 0 to 3 representing respectively 3 to 0 limbs remaining 338 C eax 2 or 3 representing respectively 1 or 0 limbs remaining 415 C No extra limbs, destination was aligned. 431 C No extra limbs, destination was unaligned. 449 C or two limbs should be written.
|
H A D | lshift.asm | 114 C eax counter, limbs, negative 265 C eax counter, limbs 305 C eax -4 to -1 representing respectively 0 to 3 limbs remaining 327 C eax -4 or -3 representing respectively 0 or 1 limbs remaining 407 C No extra src limbs, destination was aligned. 423 C No extra src limbs, destination was unaligned.
|
/netbsd-current/external/lgpl3/gmp/dist/mini-gmp/tests/ |
H A D | Makefile | 33 t-reuse t-aorsmul t-limbs t-cong t-pprime_p t-lucm \
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/sparc32/ |
H A D | add_n.asm | 71 C Add blocks of 8 limbs until less than 8 limbs remain 109 C Add blocks of 2 limbs until less than 2 limbs remain 179 C Add blocks of 8 limbs until less than 8 limbs remain
|
H A D | lshift.asm | 48 andcc %o2,4-1,%g4 C number of limbs in first loop 50 be L(0) C if multiple of 4 limbs, skip first loop
|
H A D | rshift.asm | 45 andcc %o2,4-1,%g4 C number of limbs in first loop 47 be L(0) C if multiple of 4 limbs, skip first loop
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/x86/k6/k62mmx/ |
H A D | rshift.asm | 183 C Four src limbs are held at the start, and four more will be read. 184 C Four dst limbs will be written. This schedule seems necessary for 225 C Now have the four limbs in mm2 (low) and mm0 (high), and %eax is 0 226 C to 3 representing respectively 3 to 0 further limbs. 231 C Two or three extra limbs: rshift mm2, OR it with lshifted mm0, mm0 283 C no further extra limbs
|
H A D | lshift.asm | 176 C Four src limbs are held at the start, and four more will be read. 177 C Four dst limbs will be written. This schedule seems necessary for 217 C Now have four limbs in mm2 (prev) and mm0 (next), plus eax mod 4. 221 C limbs that must be read. 227 C Two more limbs: lshift mm2, OR it with rshifted mm0, mm0 becomes
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/x86/pentium4/sse2/ |
H A D | mod_34lsub1.asm | 42 C more small sizes. 2 limbs measures about 20 cycles, but 3 limbs jumps to 125 C ecx is -2, -1 or 0 representing 0, 1 or 2 more limbs, respectively
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/powerpc32/ |
H A D | addlsh1_n.asm | 72 adde s1, s1, u0 C add limbs with cy, set cy 81 adde s0, s0, u0 C add limbs with cy, set cy
|
H A D | sublsh1_n.asm | 71 subfe s1, s1, u0 C add limbs with cy, set cy 80 subfe s0, s0, u0 C add limbs with cy, set cy
|
H A D | mod_34lsub1.asm | 53 C also a bdz or bdnz for each and a pointer update say every three limbs 73 lis r7, 0 C no carry if just three limbs
|
H A D | lshift.asm | 51 cmpwi cr0, r5, 30 C more than 30 limbs? 55 bgt L(BIG) C branch if more than 12 limbs
|
H A D | lshiftc.asm | 51 cmpwi cr0, r5, 30 C more than 30 limbs? 55 bgt L(BIG) C branch if more than 12 limbs
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/powerpc64/mode32/ |
H A D | sub_n.asm | 55 subfe r7, r0, r8 C subtract limbs with cy, set cy 63 subfe r7, r10, r9 C subtract limbs with cy, set cy
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/x86/k7/mmx/ |
H A D | copyd.asm | 112 C ecx counter, limbs 123 C now %ecx is -4 to -1 representing respectively 0 to 3 limbs remaining
|
H A D | copyi.asm | 126 C ecx counter, negative, limbs 137 C now %ecx is 0 to 3 representing respectively 3 to 0 limbs remaining
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/x86/pentium/ |
H A D | copyd.asm | 68 C ecx counter, limbs 102 C ecx -7 to 0, representing respectively 0 to 7 limbs remaining
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/x86/k6/ |
H A D | aors_n.asm | 68 C loop control, which with 4 limbs/loop means an extra 0.25 c/l. 211 C two or three more limbs 302 C now %esi is 0 to 3 representing respectively 1 to 4 limbs more 310 C two or three more limbs
|
H A D | aorsmul_1.asm | 58 dnl Unrolling to 16 limbs/loop makes the unrolled loop fit exactly in a 256 146 C where b=2^32 is the size of a limb, x*y is the two carry limbs, and u and 147 C v are the two limbs it's added to (being the low of the next mul, and a 250 C The unrolled loop uses a "two carry limbs" scheme. At the top of the loop 256 C registers at the point of doing the mul for the initial two carry limbs.
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/alpha/ |
H A D | sub_n.asm | 53 blt r19,$Lend2 C if less than 4 limbs, goto 2nd loop 74 blt r19,$Lend1 C if less than 4 limbs remain, jump 75 C 1st loop handles groups of 4 limbs in a software pipeline 139 C 2nd loop handles remaining 1-3 limbs
|
H A D | com.asm | 43 C 2.0 c/l. In general, a pattern like this unrolled to N limbs per loop 56 C The main loop handles an odd count of limbs, being two limbs loaded before 60 C An even number of limbs is handled by an explicit dst[0]=~src[0] in the
|
H A D | add_n.asm | 53 blt r19,$Lend2 C if less than 4 limbs, goto 2nd loop 74 blt r19,$Lend1 C if less than 4 limbs remain, jump 75 C 1st loop handles groups of 4 limbs in a software pipeline 139 C 2nd loop handles remaining 1-3 limbs
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/x86/k6/mmx/ |
H A D | dive_1.asm | 149 C ecx counter, limbs, negative 207 jz L(even_two) C if only two limbs 219 C ecx counter, limbs, negative 225 C mm0 scratch, source limbs
|
/netbsd-current/external/lgpl3/gmp/dist/mpn/x86/k7/ |
H A D | mod_34lsub1.asm | 106 C ecx counter, limbs 133 C ecx is -2, -1 or 0 representing 0, 1 or 2 more limbs, respectively
|