d0p1/ack
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity
ack/mach/powerpc/libem/fef8.s

62 lines
1.5 KiB
ArmAsm
Raw Permalink Normal View History

In PowerPC libem, use the new features of our assembler. The new features are the hi16/lo16 and ha16/lo16 syntax for relocations, and the extended mnemonics like "blr". Use ha16/lo16 to load some double floats with 2 instructions (lis/lfd) instead of 3 (lis/ori/lfd). Use the extended names for branches, comparisons, and bit rotations, so I can more easily read the code. The new names often encode the same machine instructions as the old names, except in a few places where I changed the instructions. Stop using andi. when we don't need to set cr0. In inn.s, I change andi. to extrwi to extract the same bits. In los.s and sts.s, I change "andi. r3, r3, ~3" to "clrrwi r3, r3, 2". This avoids setting cr0 and also stops clearing the high 16 bits of r3. In csa.s, los.s, sts.s, I change some comparisons and right shifts from signed to unsigned (cmplw, cmplwi, srwi), because the sizes are unsigned. In inn.s, the right shift can be signed (sraw) or unsigned (srw), but I use srw because we don't need the carry bit. In fef8.s, I save an instruction by using rlwinm instead of addis/andc to rlwinm to clear a field. The code no longer kills r7. In both fef8.s and fif8.s, I remove the list of killed registers. Also remove some whitespace from ends of lines.
2017-01-23 22:16:39 +00:00
.sect .text; .sect .rom; .sect .data; .sect .bss
Rewrite fef 8 in powerpc assembly. In EM, fef splits a float into exponent and fraction. The old C code, given an infinite float, got stuck in an infinite loop. The new assembly code doesn't loop; it extracts the IEEE exponent.
2016-09-29 19:52:54 +00:00
.sect .text
! Split a double-precision float into fraction and exponent, like
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
! frexp(3) in C, http://en.cppreference.com/w/c/numeric/math/frexp
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
!
! Stack: ( double -- fraction exponent )
Rewrite fef 8 in powerpc assembly. In EM, fef splits a float into exponent and fraction. The old C code, given an infinite float, got stuck in an infinite loop. The new assembly code doesn't loop; it extracts the IEEE exponent.
2016-09-29 19:52:54 +00:00
.define .fef8
.fef8:
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
lwz r3, 0(sp) ! r3 = high word (bits 0..31)
lwz r4, 4(sp) ! r4 = low word (bits 32..63)
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
! IEEE double = sign * 1.fraction * 2**(exponent - 1023)
Rewrite fef 8 in powerpc assembly. In EM, fef splits a float into exponent and fraction. The old C code, given an infinite float, got stuck in an infinite loop. The new assembly code doesn't loop; it extracts the IEEE exponent.
2016-09-29 19:52:54 +00:00
! sign exponent fraction
! 0 1..11 12..63
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
!
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
! IEEE exponent = 1022 in [0.5, 1) or (-1, -0.5].
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
extrwi. r6, r3, 11, 1 ! r6 = IEEE exponent
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
beq 3f ! jump if zero or denormalized
In PowerPC libem, use the new features of our assembler. The new features are the hi16/lo16 and ha16/lo16 syntax for relocations, and the extended mnemonics like "blr". Use ha16/lo16 to load some double floats with 2 instructions (lis/lfd) instead of 3 (lis/ori/lfd). Use the extended names for branches, comparisons, and bit rotations, so I can more easily read the code. The new names often encode the same machine instructions as the old names, except in a few places where I changed the instructions. Stop using andi. when we don't need to set cr0. In inn.s, I change andi. to extrwi to extract the same bits. In los.s and sts.s, I change "andi. r3, r3, ~3" to "clrrwi r3, r3, 2". This avoids setting cr0 and also stops clearing the high 16 bits of r3. In csa.s, los.s, sts.s, I change some comparisons and right shifts from signed to unsigned (cmplw, cmplwi, srwi), because the sizes are unsigned. In inn.s, the right shift can be signed (sraw) or unsigned (srw), but I use srw because we don't need the carry bit. In fef8.s, I save an instruction by using rlwinm instead of addis/andc to rlwinm to clear a field. The code no longer kills r7. In both fef8.s and fif8.s, I remove the list of killed registers. Also remove some whitespace from ends of lines.
2017-01-23 22:16:39 +00:00
cmpwi r6, 2047
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
addi r5, r6, -1022 ! r5 = our exponent
beq 2f ! jump if infinity or NaN
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
! fall through if normalized
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
! Put fraction in [0.5, 1) or (-1, -0.5].
1: li r6, 1022
insrwi r3, r6, 11, 1 ! IEEE exponent = 1022
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
! fall through
Rewrite fef 8 in powerpc assembly. In EM, fef splits a float into exponent and fraction. The old C code, given an infinite float, got stuck in an infinite loop. The new assembly code doesn't loop; it extracts the IEEE exponent.
2016-09-29 19:52:54 +00:00
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
2: stw r3, 0(sp)
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
stw r4, 4(sp) ! push fraction
stwu r5, -4(sp) ! push exponent
blr
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
! Got denormalized number or zero, probably zero.
! If zero, then exponent must also be zero.
3: extrwi r6, r3, 20, 12
Rewrite fef 8 in powerpc assembly. In EM, fef splits a float into exponent and fraction. The old C code, given an infinite float, got stuck in an infinite loop. The new assembly code doesn't loop; it extracts the IEEE exponent.
2016-09-29 19:52:54 +00:00
or. r6, r6, r4 ! r6 = high|low fraction
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
bne 4f ! jump if not zero
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
li r5, 0 ! exponent = 0
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
b 2b
Rewrite fef 8 in powerpc assembly. In EM, fef splits a float into exponent and fraction. The old C code, given an infinite float, got stuck in an infinite loop. The new assembly code doesn't loop; it extracts the IEEE exponent.
2016-09-29 19:52:54 +00:00
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
! Got denormalized number = 0.fraction * 2**-1022
4: lfd f0, 0(sp)
lis r6, ha16[.fs_2_64]
lfs f1, lo16[.fs_2_64](r6)
Rewrite fef 8 in powerpc assembly. In EM, fef splits a float into exponent and fraction. The old C code, given an infinite float, got stuck in an infinite loop. The new assembly code doesn't loop; it extracts the IEEE exponent.
2016-09-29 19:52:54 +00:00
fmul f0, f0, f1 ! multiply it by 2**64
stfd f0, 0(sp)
lwz r3, 0(sp)
lwz r4, 4(sp)
In PowerPC libem, use the new features of our assembler. The new features are the hi16/lo16 and ha16/lo16 syntax for relocations, and the extended mnemonics like "blr". Use ha16/lo16 to load some double floats with 2 instructions (lis/lfd) instead of 3 (lis/ori/lfd). Use the extended names for branches, comparisons, and bit rotations, so I can more easily read the code. The new names often encode the same machine instructions as the old names, except in a few places where I changed the instructions. Stop using andi. when we don't need to set cr0. In inn.s, I change andi. to extrwi to extract the same bits. In los.s and sts.s, I change "andi. r3, r3, ~3" to "clrrwi r3, r3, 2". This avoids setting cr0 and also stops clearing the high 16 bits of r3. In csa.s, los.s, sts.s, I change some comparisons and right shifts from signed to unsigned (cmplw, cmplwi, srwi), because the sizes are unsigned. In inn.s, the right shift can be signed (sraw) or unsigned (srw), but I use srw because we don't need the carry bit. In fef8.s, I save an instruction by using rlwinm instead of addis/andc to rlwinm to clear a field. The code no longer kills r7. In both fef8.s and fif8.s, I remove the list of killed registers. Also remove some whitespace from ends of lines.
2017-01-23 22:16:39 +00:00
extrwi r6, r3, 11, 1 ! r6 = IEEE exponent
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
addi r5, r6, -1022 - 64 ! r5 = our exponent
b 1b
Rewrite fef 8 in powerpc assembly. In EM, fef splits a float into exponent and fraction. The old C code, given an infinite float, got stuck in an infinite loop. The new assembly code doesn't loop; it extracts the IEEE exponent.
2016-09-29 19:52:54 +00:00
.sect .rom
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
.fs_2_64:
!float 1.84467440737095516e+19 sz 4
.data1 0137,0200,00,00
Reference in a new issue Copy permalink