XOR is a simple logic-gate operation. SUB would have to be an ALU operation.
A one-bit adder (which is subtraction in reverse) makes signals pass through two gates.
See https://en.wikipedia.org/wiki/Adder_(electronics)
You need the 2 gates for adding/subtracting because you care about carry. So if you're adding/subtracting 8 bits, 16 bits, or more, you're connecting multiples of these together, and that carry has to ripple through all the rest of the gates one-by-one. It can't be paralellized without extra circuitry, which increases your costs in other ways.
Without the AND gate needed for carry, all the XORs can fire off at the same time. If you added the extra circuitry for a parallelizable add/subtract to make it as fast as XOR, your actual parallel XOR would consume less power.
That's all true, but on any modern x86 processor both the single pair of gates for the xor and the 10 or so for a carry-bypass 64 bit wide subtraction both happen with a single clock cycle of latency so from a programmer's perspective they're the same in that sense. There's still an energy difference but its tiny compared to what even the register file and bypass network for the operation use, let along the OoO structures.
It's still the same number of clock cycles, though, isn't it? You're using some extra circuitry during the SUB, but during the XOR, that circuitry is just sitting idle anyway, so it's still six of one/half a dozen of the other.
XOR can do everything in 1 cycle (which is hopefully far, far less than the clock). SUB-if done the simple way-has to take n cycles where n is the number of bits subtracted.
What do you mean by cycles? A ripple-carry adder needs to wait for the carry bits to ripple through yes, but there's no clock cycle involved.
On some of IBM's smaller processors, such as channel controllers and the CSP used in the midrange line prior to the System/38, the xor instruction had a special feature when used with identical source and destination - It would inhibit parity and/or ECC error checking on the read cycle, which meant that xor could be used to clear a register or memory location that had been stored with bad parity without taking a machine check or processor check.
[delayed]
"Bonus bonus chatter: The xor trick doesn’t work for Itanium because mathematical operations don’t reset the NaT bit. Fortunately, Itanium also has a dedicated zero register, so you don’t need this trick. You can just move zero into your desired destination."
Will remember for the next time I write asm for Itanium!
Quite a few architectures have a dedicated 0 register.
Yep. The XOR trick - relying on special use of opcode rather than special register - is probably related to limited number of (general purpose) registers in typical '70 era CPU design (8080, 6502, Z80, 8086).
Unfortunately, 6502 can't XOR the accumulator with itself. I don't recall if the Z80 can, and loading an immediate 0 would be most efficient on those anyway.
XOR A absolutely works on Z80 and it's of course faster and shorter than loading a zero value with LD A,0. LD A,0 is encoded to 2 bytes while XOR A is encoded as a single opcode. XOR A has the additional benefit to also clear all the flags to 0. Sub A will clear the accumulator, but it will always set the N flag on Z80.
Ah, thanks, I couldn't recall off the top of my head.
You're absolutely right, I stand corrected.
The 6502 gets by doing immediate load: 2 clock cycles, 2 bytes (frequently followed by single byte register transfer instruction). Out of curiosity I did a quick scan of the MOS 1.20 rom of the BBC micro:
LDY #0 (a0 00): 38 hits
LDX #0 (a2 00): 28 hits
LDA #0 (a9 00): 48 hitsThe Z80 can do either LD A,0 or SUB A or XOR A, but the LD is slower due to the extra memory cycle to load the second byte of the instruction.
And [as mentioned in the article] even modern x86 implementations have a zero register. So you have this weird special opcode that (when called with identical source and destination) only triggers register renaming
A move on SPARC is technically an OR of the source with the zero register. "move %l0, %l1" is assembled as "or %g0, %l0, %l1". So if you want to zero a register you OR %g0 with itself.
Indeed!!
MIPS - $zero
RISC-V - x0
SPARC - %g0
ARM64 - XZR
PowerPC: "r0 occasionally" (with certain instructions like addi, though this might be better considered an edge case of encoding)
indeed. riscv for instance. also, afaik, xor’ing is faster. i would assume that someone like mr. raymond would know…
> afaik, xor’ing is faster
Even tiny tiny CPUs can do sub in one cycle, so I doubt that. On super-scalar CPUs xor and sub are normally issued to the same execution units so it wouldn't make a difference there either.
On superscalars running xor trick as is would be significantly slower because it implies a data dependency where there isn't one. But all OOO x86's optimize it away internally.
Which part of "mathematical operations don’t reset the NaT bit" did you not understand?
The obvious answer is that XOR is faster. To do a subtract, you have to propagate the carry bit from the least-significant bit to the most-significant bit. In XOR you don't have to do that because the output of every bit is independent of the other adjacent bits.
Probably, there are ALU pipeline designs where you don't pay an explicit penalty. But not all, and so XOR is faster.
Surely, someone as awesome as Raymond Chen knows that. The answer is so obvious and basic I must be missing something myself?
> To do a subtract, you have to propagate the carry bit from the least-significant bit to the most-significant bit.
Yes, but that need not scale linearly with the number of bits. https://en.wikipedia.org/wiki/Carry-lookahead_adder:
“A carry-lookahead adder (CLA) or fast adder is a type of electronics adder used in digital logic. A carry-lookahead adder […] can be contrasted with the simpler, but usually slower, ripple-carry adder (RCA), for which the carry bit is calculated alongside the sum bit, and each stage must wait until the previous carry bit has been calculated to begin calculating its own sum bit and carry bit. The carry-lookahead adder calculates one or more carry bits before the sum, which reduces the wait time to calculate the result of the larger-value bits of the adder.
[…]
Already in the mid-1800s, Charles Babbage recognized the performance penalty imposed by the ripple-carry used in his difference engine, and subsequently designed mechanisms for anticipating carriage for his never-built analytical engine.[1][2] Konrad Zuse is thought to have implemented the first carry-lookahead adder in his 1930s binary mechanical computer, the Zuse Z1.”
I think most, if not all, current ALUs implement such adders.
Carry lookahead is definitely faster than ripple carry but it's not free. It requires high-fan-in gates that take up a fair amount of silicon. That silicon saves time though, so as you say almost nobody uses ripple carry any more.
There's a structure called a carry-bypass adder[1] that lets you add two numbers in O(√n) time for only O(n) gates. That or a similar structure is what modern CPUs use and they allow you two add two numbers in a single clock cycle which is all you care about from a software perspective.
There are also tree adders which add in O(log(n)) time but use O(n^2) gates if you really need the speed, but AFAIK nobody actually does need to.
His point is that in x86 there is no performance difference but everyone except his colleague/friend uses xor, while sub actually leaves cleaner flags behind. So he suspects its some kind of social convention selected at random and then propagated via spurious arguments in support (or that it “looks cooler” as a bit of a term of art).
It could also be as a result of most people working in assembly being aware of the properties of logic gates, so they carry the understanding that under the hood it might somehow be better.
GP seems to think it strange that "x86" would actually not have a performance difference here.
I think this might just be due to not realizing just how far back in CPU history this goes.
In a clockless cpu design you'd indeed expect xor to be faster. But in a regular CPU with a clock you either waste a bit of xor performance by making xor and sub both take the same number of ticks, or you speed up the clock enough that the speed difference between xor and sub justifies sub being at least a full tick slower
The former just seems way more practical
Even if they take the same number of ticks, shouldn't xor fundamentally needing less work also mean it can be performed while drawing less power/heating less, which is just as much an improvement in the long run?
I think an even more likely explanation would be that x86 assembly programmers often were, or learned from other-architecture assembly programmers. Maybe there's a place where it makes more sense and it can be so attributed. 6502 and 68k being first places I would look at.
6502 doesn't even have register-to-register ALU operations, there's no alternative to LDA #0.
8080/Z80 is probably where XOR A got a lead over SUB A, but they are also the same number of cycles.
For 68k depending on the size you're interested in then it mostly doesn't matter.
.b and .w -> clr eor sub are all identical
for .l moveq #0 is the winner
> The answer is so obvious
A tangent, but what is Obvious depends on what you know.
Often experts don't explain the things they think are Obvious, but those things are only Obvious to them, because they are the expert.
We should all kind, and explain also the Obvious things those who do not know.
"The proof is left as an exercise for the reader" comes to mind
That comment is not very useful without pointing to realworld CPUs where SUB is more expensive than XOR ;)
E.g. on Z80 and 6502 both have the same cycle count.
With more bits, then SUB is going to be more and more expensive to fit in the same number of clocks as XOR. So with an 8-bit CPU like Z80, it probably makes design sense to have XOR and SUB both take one cycle. But if for instance a CPU uses 128-bit registers, then the propagate-and-carry logic for ADD/SUB might take way much longer than XOR that the designers might not try to fit ADD/SUB into the same single clock cycle as XOR, and so might instead do multi-cycle pipelined ADD/SUB.
A real-world CPU example is the Cray-1, where S-Register Scalar Operations (64-bit) take 3 cycles for ADD/SUB but still only 1 cycle for XOR. [1]
[1] https://ed-thelen.org/comp-hist/CRAY-1-HardRefMan/CRAY-1-HRM...
The 6502 doesn't support XOR A or SUB A, and in fact doesn't have a SUB opcode at all, only SBC (subtract with carry, requiring an extra opcode to set the carry flag beforehand).
I was handwaving over the details, SBC is identical to SUB when the carry flag is clear, so it's understandable why the 6502 designers didn't waste an instruction slot.
EOR and SBC still have the same cycle counts though.
Sure, in some contexts you would know that the carry flag was set or clear (depending on what you needed), and it was common to take advantage of that and not add an explicit clc or sec, although you better comment the assumption/dependency on the preceding code.
However the 6502 doesn't support reg-reg ALU operations, only reg-mem, so there simply is no xor a,a or sbc a,a support. You'd either have to do the explicit lda #0, or maybe use txa/tya if there was a free zero to be had.
Cortex A8 vsub reads the second source register a cycle earlier than veor, so that can add one cycle latency
Not scalar, but still sub vs xor. Though you’d use vmov immediate for zeroing anyway.
Harvard Mark I? Not sure why people think programming started with Z80.
The article is about x86, and x86 assembly is mostly a superset of 8080 (which is why machine language numbers registers as AX/CX/DX/BX, matching roughly the function of A/BC/DE/HL on the 8080—in particular with respect to BX and HL being last).
My WW2-era assembly is a bit rusty, but I don't think the Harvard Mark 1 had bitwise logical operations?
XOR is faster when you do that alone in an FPGA or in an ASIC.
When you do XOR together with many other operations in an ALU (arithmetic-logical unit), the speed is determined by the slowest operation, so the speed of any faster operation does not matter.
This means that in almost all CPUs XOR and addition and subtraction have the same speed, despite the fact that XOR could be done faster.
In a modern pipelined CPU, the clock frequency is normally chosen so that a 64-bit addition can be done in 1 clock cycle, when including all the overheads caused by registers, multiplexers and other circuitry outside the ALU stages.
Operations more complex than 64-bit addition/subtraction have a latency greater than 1 clock cycle, even if one such operation can be initiated every clock cycle in one of the execution pipelines.
The operations less complex than 64-bit addition/subtraction, like XOR, are still executed in 1 clock cycle, so they do not have any speed advantage.
There have existed so-called superpipelined CPUs, where the clock frequency is increased, so that even addition/subtraction has a latency of 2 or more clock cycles.
Only in superpipelined CPUs it would be possible to have a XOR instruction that is faster than subtraction, but I do not know if this has ever been implemented in a real superpipelined CPU, because it could complicate the execution pipeline for negligible performance improvements.
Initially superpipelining was promoted by DEC as a supposedly better alternative to the superscalar processors promoted by IBM. However, later superpipelining was abandoned, because the superscalar approach provides better energy efficiency for the same performance. (I.e. even if for a few years it was thought that a Speed Demon beats a Brainiac, eventually it was proven that a Brainiac beats a Speed Demon, like shown in the Apple CPUs)
While mainstream CPUs do not use superpipelining, there have been some relatively recent IBM POWER CPUs that were superpipelined, but for a different reason than originally proposed. Those POWER CPUs were intended for having good performance only in multi-threaded workloads when using SMT, and not in single-thread applications. So by running simultaneous threads on the same ALU the multi-cycle latency of addition/subtraction was masked. This technique allowed IBM a simpler implementation of a CPU intended to run at 5 GHz or more, by degrading only the single-thread performance, without affecting the SMT performance. Because this would not have provided any advantage when using SMT, I assume that in those POWER CPUs XOR was not made faster than subtraction, even if this would have theoretically been possible.
As TFA says, on x86 `sub eax, eax` encodes to the same number of bytes and executes in the same number of cycles.
On modern ones, x86 has quite a history and the idiom might carry on from an even older machine.
Edit: Looked at comments, seems like x86 and the major 8bit cpu's had the same speed, pondering in this might be a remnant from the 4-bit ALU times.
> seems like x86 and the major 8bit cpu's had the same speed, pondering in this might be a remnant from the 4-bit ALU times.
I think that era of CPUs used a single circuit capable of doing add, sub, xor etc. They'd have 8 of them and the signals propagate through them in a row. I think this page explains the situation on the 6502: https://c74project.com/card-b-alu-cu/
And this one for the ARM 1: https://daveshacks.blogspot.com/2015/12/inside-alu-of-armv1-...
But I'm a software engineer speculating about how hardware works. You might want to ask a hardware engineer instead.
Nope.
In any ALU the speed is determined by the slowest operation, so XOR is never faster. It does not matter which is the width of the ALU, all that matters is that an ALU does many kinds of operations, including XOR and subtraction, where the operation done by an ALU is selected by some control bits.
I have explained in another comment that the only CPUs where XOR can be faster than subtraction are the so-called superpipelined CPUs. Superpipelined CPUs have been made only after 1990 and there were very few such CPUs. Even if in superpipelined CPUs it is possible for XOR to be faster than subtraction, it is very unlikely that this feature has been implemented in anyone of the few superpipelined CPU models that have ever been made, because it would not have been worthwhile.
For general-purpose computers, there have never been "4-bit ALU times".
The first monolithic general-purpose processor was Intel 8008 (i.e. the monolithic version of Datapoint 2200), with an 8-bit ISA.
Intel claims that Intel 4004 was the first "microprocessor" (in order to move its priority earlier by one year), but that was not a processor for a general-purpose computer, but a calculator IC. Its only historical relevance for the history of personal computers is that the Intel team which designed 4004 gained a lot of experience with it and they established a logic design methodology with PMOS transistors, which they used for designing the Intel 8008 processor.
Intel 4004, its successors and similar 4-bit processors introduced later by Rockwell, TI and others, were suitable only for calculators or for industrial controllers, never for general-purpose computers.
The first computers with monolithic processors, a.k.a. microcomputers, used 8-bit processors, and then 16-bit processors, and so on.
For cost reduction, it is possible for an 8-bit ISA to use a 4-bit ALU or even just a serial 1-bit ALU, but this is transparent for the programmer and for general-purpose computers there never were 4-bit instruction sets.
From TFA:
The predominance of these idioms as a way to zero out a register led Intel to add special xor r, r-detection and sub r, r-detection in the instruction decoding front-end and rename the destination to an internal zero register, bypassing the execution of the instruction entirely.
I'm not actually aware of any CPUs that preform a XOR faster than a SUB. And more importantly, they have identical timings on the 8086, which is where this pattern comes from.
I had a similar reaction when learning 8086 assembly and finding the correct way to do `if x==y` was a CMP instruction which performed a subtraction and set only the flags. (The book had a section with all the branch instructions to use for a variety of comparison operators.) I think I spent a few minutes experimenting with XOR to see if I could fashion a compare-two-values-and-branch macro that avoided any subtraction.
Comparing for equality can use either SUB or XOR: it sets the zero flag if (and only if) the two values are equal. That's why JE/JNE (jump if equal/not equal) is an alias for JZ/JNZ (jump if zero/not zero).
There's also the TEST instruction, which does a logical AND but without storing the result (like CMP does for SUB). This can be used to test specific bits.
Testing a single register for zero can be done in several ways, in addition to CMP with 0:
TEST AX,AX
AND AX,AX
OR AX,AX
INC AX followed by DEC AX (or the other way around)
Also any arithmetic operation sets those flags, so you may not even need an explicit test. MOV doesn't set flags however, at least on x86 -- it does on some other architectures.
From TFA:
> It encodes to the same number of bytes, executes in the same number of cycles.
Those aren't the only resources. I could imagine XOR takes less energy because using it might activate less circuitry than SUB.
I'm not aware of any stories in the historical record of "real programmers" optimizing for power use, only for speed or code size.
For a few years I worked in the team that wrote software for an embedded audio DSP. The power draw to do something was normally more important than the speed. Eg when decoding MP3 or SBC you probably had enough MIPS to keep up with the stream rate, so the main thing the customers cared about was battery life. Mostly the techniques to optimize for speed were the same as those for power. But I remember being told that add/sub used less power than multiply even though both were single cycle. And that for loops with fewer than 16 instructions used less power because there was a simple 16 instruction program memory cache that saved the energy required to fetch instructions from RAM or ROM. (The RAM and ROM access was generally single cycle too).
Nowadays, I expect optimizations that minimize energy consumption are an important target for LLM hosts.
The operation is slightly more complex yes, but has there ever been an x86 CPU where SUB or XOR takes more than a single CPU cycle?
I wonder if you could measure the difference in power consumption.
I mean, not for zeroing because we know from the TFA that it's special-cased anyway. But maybe if you test on different registers?
Yea, that’s what immediately went through my head, too. XOR is ALWAYS going to be single cycle because it’s bit-parallel.
I would be surprised if modern CPUs didn't decode "xor eax, eax" into a set of micro-ops that simply moves from an externally invisible dedicated 0 register. These days the x86 ISA is more of an API contract than an actual representation of what the hardware internals do.
From TFA:
The predominance of these idioms as a way to zero out a register led Intel to add special xor r, r-detection and sub r, r-detection in the instruction decoding front-end and rename the destination to an internal zero register, bypassing the execution of the instruction entirely. You can imagine that the instruction, in some sense, “takes zero cycles to execute”."rename the destination to an internal zero register"
That would be quite late then, 1997 Pentium 2 for general population.
Zero micro ops to be precise, that’s handled entirely at the register rename stage with no data movement.
It's like 0.5 cycles vs 0.9 cycles. So both are 1 cycle, considering synchronization.
But energy consumption could be different for this hypothetical 0.5 and 0.9.
Energy consumption wasn't really a concern when the idiom developed. I don't think people really cared about the energy consumption of instructions until well into the x86-64 era.
Not sure why this is being downvoted, but it’s absolutely correct. For most of the history of computing, people were happy that it worked at all. Being concerned about energy efficiency is a recent byproduct of mobile devices and, even more recently, giant amounts of compute adding up to gigawatts.
The non-obvious bit is why there isn't an even faster and shorter "mov <register>,0" instructions - the processors started short-circuiting xor <register>,<register> much later.
In x86, a basic immediate instruction with a 1 Byte immediate value is encoded like this:
<op> (1 Byte opcode), <Registers> (1 Byte), <immediate value> (1 Byte)
While xor eax, eax only uses 2 bytes. Since there are only 8 registers, meaning they can be encoded with 3 bits, you can pack two values into the <Registers> field (ModR/M).
Making mov eax, 0 only take two bytes would require significant changes of the ISA to allow immediate values in the ModR/M byte (or similar) but there would be little benefit since zeroing can already be done in 2 bytes and I doubt that other cases are even close to frequent enough for this to be any significant benefit overall. An actual improvement would be if there was a dedicated 1 Byte set-rax-to-0 instruction, but obviously that comes at a tradeoff where we have to encode another operation differently (probably with more bytes) again (and you can't zero anything else with it).
A number of the RISC processors have a special zero register, giving you a "mov reg, zero" instruction.
Of course many of the RISC processors also have fixed length instructions, with small literal values being encoded as part of the instruction, so "mov reg, #0" and "mov reg, zero" would both be same length.
Right, like a “set reg to zero” instruction. One byte. Just encodes the operation and the reg to zero. I’m surprised we didn’t have it on those old processors. Maybe the thinking was that it was already there: xor reg,reg.
One byte instructions, with 8 registers as in the 8086, waste 8 opcodes which is 3% of the total. There are just five: "INC reg", "DEC reg", "PUSH reg", "POP reg", "XCHG AX, reg" (which is 7 wasted opcodes instead of 8, because "XCHG AX, AX" doubles as NOP).
One-byte INC/DEC was dropped with x86-64, and PUSH/POP are almost obsolete in APX due to its addition of PUSH2/POP2, leaving only the least useful of the five in the most recent incantation of the instruction set.
I’m not sure I understand what you mean by “waste 8 opcodes.”
They occupy 8 of the possible 256 byte values. Together, those five cases used about 15% of the space.
Though I was forgetting one important case: MOV r,imm also used one-byte opcodes with the register index embedded. And it came in byte and word variants, so it used a further 16 opcodes bytes for a total of 56 one byte opcodes with register encoding.
There are only 256 1-byte opcodes or prefixes available, if you take 8 of these to zero registers, they won't be available for other instruction, and unless you consider zeroing to be so important that they really need their 1-byte opcodes, it is redundant since you can use the 2-byte "xor reg,reg" instead, hence the "waste'.
In addition, you would need 16 opcodes, not 8, if you also wanted to cover 8 bit registers (AH/AL,...).
Special shout-out to the undocumented SALC instruction, which puts the carry flag into AL. If you know that the carry will be 0, it is a nice sizecoding trick to zero AL in 1 byte.
Traditionally in x86, only the first byte is the opcode used to select the instruction, and any further bytes contain only operands. Thus, since there exist 256 possible values for the initial byte, there are at most 256 possible opcodes to represent different instructions.
So if you add a 1-byte instruction for each register to zero its value, that consumes 8 of the possible 256 opcodes, since there are 8 registers. Traditional x86 did have several groups of 1-byte instructions for common operations, but most of them were later replaced with multibyte encodings to free up space for other instructions.
special mov 0 instruction times 8 registers. The opcode space, especially 1 byte opcode space, is precious so encoding redundant operations is wasteful.
Instruction slots are extremely valuable in 8-bit instruction sets. The Z80 has some free slots left in the ED-prefixed instruction subset, but being prefix-instructions means they could at best run at half speed of one-byte instructions (8 vs 4 clock cycles).
> The obvious answer is that XOR is faster.
It used to be not only faster but also smaller. And back then this mattered.
Say you had a computer running at 33 Mhz, you had 33 million cycles per second to do your stuff. A 60 Hz game? 33 million / 60 and suddenly you only have about 500 000 cycles per frame. 200 scanlines? Suddenly you're left with only 2500 cycles per scanline to do your stuff. And 2500 cycles really isn't that much.
So every cycle counted back then. We'd use the official doc and see how many cycles each instruction would take. And we'd then verify by code that this was correct too. And memory mattered too.
XOR was both faster and smaller (less bytes) then a MOV ..., 0.
Full stop.
And when those CPU first began having cache, the cache were really tiny at first: literally caching ridiculously low number of CPU instructions. We could actually count the size of the cache manually (for example by filling with a few NOP instructions then modifying them to, say, add one, and checking which result we got at the end).
XOR, due to being smaller, allowed to put more instructions in the cache too.
Now people may lament that it persisted way long after our x86 CPUs weren't even real x86 CPUs anymore and that is another topic.
But there's a reason XOR was used and people should deal with it.
We zero with XOR EAX,EAX and that's it.
The context was comparison to SUB EAX,EAX, not to a MOV.
Because he is explicitly talking about x86 - maybe you missed that.
XOR and SUB have had identical cycle counts and latencies since the 8088. That's because you can "look ahead" when doing carries in binary. It's just a matter of how much floorspace on the chip you want to use.
https://en.wikipedia.org/wiki/Carry-lookahead_adder
The only minor difference between the two on x86, really, is SUB sets OF and CF according to the result while XOR always clears them.
A carry lookahead adder makes your circuit depth logarithmic in the width of the inputs vs linear for a ripple carry adder, but that is still asymptotically worse than XORs constant depth.
(But this does not discount the fact that basically all CPUs treat them both as one cycle)
OF/CF/AF are always cleared anyway by SUB r,r. So there's absolutely no difference.
Relatedly, there's a steganographic opportunity to hide info in machine code by using "XOR rax,rax" for a "zero" and "SUB rax,rax" for a "one" in your executable. Shouldn't be too hard to add a compiler feature to allow you to specify the string you want encoded into its output.
That could be a style metric, too. Time spent reversing MS-DOS viruses in my youth showed me assembler programmers very clearly have styles to their code. It's too weak for definitive attribution but it was interesting to see "rhymes" between, for example, the viruses written by The Dark Avenger.
You can do better. X86 has both "op [mem], reg" and "op reg, [mem]" variants of most instructions, where "[mem]" can be a register too. So you have two ways to encode "xor eax, eax", differing by which of the operands is in the "possible memory operand" slot, the source or the destination.
This sounds like a Paged Out article ;)
https://www.cs.columbia.edu/~angelos/Papers/hydan.pdf
Here's some more prior art
Back when I was in university, one of the units touching Assembly[0] required students to use subtraction to zero out the register instead of using the move instruction (which also worked), as it used fewer cycles.
I looked it up afterwards and xor was also a valid instruction in that architecture to zero out a register, and used even fewer cycles than the subtraction method; but it was not listed in the subset of the assembly language instructions we were allowed to use for that unit. I suspect that it was deemed a bit off-topic, since you would need to explain what the mathematical XOR operation was (if you didn't already learn about it in other units), when the unit was about something else entirely- but everyone knows what subtraction is, and that subtracting a number by itself leads to zero.
[0] Not x86, I do not recall the exact architecture.
It amazes me how entertaining Raymond's writing on most mundane aspects of computing often is.
> but xor took a slightly lead due to some fluke, perhaps because it felt more “clever”.
Absolutely. But I can also imagine that it feels more like something that should be more efficient, because it's "a bit hack" rather than arithmetic. After all, it avoids all the "data dependencies" (carries, never mind the ALU is clocked to allow time for that regardless)!
I imagine that a similar feeling is behind XOR swap.
> Once an instruction has an edge, even if only extremely slight, that’s enough to tip the scales and rally everyone to that side.
Network effects are much older than social media, then....
I ran into this rabbithole while writing an x86-64 asm rewriter.
xor was the default zeroing idiom.I onkly did sub reg,reg when I actually want its flags result. Otherwise the main rule is: do not touch either form unless flags liveness makes the rewrite obviously safe. Had about 40 such idioms for the passes.
Once an instruction has an edge, even if only extremely slight, that’s enough to tip the scales and rally everyone to that side.
This is a hypothesis about why the chirality of life on earth is what it is, but I don't think there's enough evidence to state that this (or any competing hypothesis) is definitely the correct explanation.
Well "definitely correct" has no real place in probabilistic arguments almost by ipso factum absurdum :-)
The chirality argument made is more akin to dynamic systems balance; yes, you can balance a pencil on its point .. but given a bit of random tilt one way or the other it's going to tend to keep going and end near flat on the table.
You still need to explain why this case creates a positive feedback loop rather than a negative one. I mean left/right fuel intakes in cars and male/female ratios somehow tend to balance at 50/50.
Regarding gender ratios: https://en.wikipedia.org/wiki/Fisher's_principle
There's exceptions, but they tend to be colonial animals in the broadest sense e.g. how clownfish males are famously able to become female but each group has one breeding male and one breeding female at any given time*, or bees where the males (drones) are functionally flying sperm and there's only one fertile female in any given colony; or some reptiles which have a temperature-dependent sex determination that may have been 50/50 before we started causing rapid climate change but in many cases isn't now: https://en.wikipedia.org/wiki/Temperature-dependent_sex_dete...
* Wolves, despite being where nomenclature of "alpha" comes from, are not this. The researcher who coined the term realised they made a mistake and what he thought of as the "alpha" pair were simply the parents of the others in that specific situation: https://davemech.org/wolf-news-and-information/
As someone with a right side fuel intake, that’s certainly isn’t true in the US. Left side fuel intake dominates completely and when the 8 pump station I prefer is busy, I only ever see left hand intake cars being fueled from the “wrong” side.
products of an asymmetric reaction performed without enantiomeric control can selectively catalyse the formation of more products with the same handedness -- this is called autocatalysis. so the first full reaction might produce a left-handed product (by chance) but that left-handed product will then cause future products to be preferentially left-handed. see the [Soai reaction](https://en.wikipedia.org/wiki/Soai_reaction?wprov=sfla1) for an example of this.
as mentioned by others this is conjectural but it is a popular (if somewhat unfalsifiable) explanation for homochirality
Wrt amino acids and sugars I personally don't have to explain as a good many others have already.
eg: For one, Isaac Asimov in the 1970s wrote at length on this in his role as a non fiction science writer with a Chemistry Phd
> male/female ratios somehow tend to balance at 50/50.
This is different to the case of actual right handed dominance in humans and to L- Vs R- dominance in chirality ...
( Men and women aren't actual mirror images of each other ... )
> left/right fuel intakes in cars
Are I believe chosen by intelligent humans who are deliberately trying to keep the lines at gas stations balanced.
It should be noted that XOR is just (bitwise) subtraction modulo 2.
There are many kinds of SUB instructions in the x86-64 ISA, which do subtraction modulo 2^64, modulo 2^32, modulo 2^16 or modulo 2^8.
To produce a null result, any kind of subtraction can be used, and XOR is just a particular case of subtraction, it is not a different kind of operation.
Unlike for bigger moduli, when operations are done modulo 2 addition and subtraction are the same, so XOR can be used for either addition modulo 2 or subtraction modulo 2.
> XOR is just a particular case of subtraction, it is not a different kind of operation.
It's different in that there's no carry propagation.
That is not a property specific to XOR.
Whenever you do addition/subtraction modulo some power of two, the carry does not propagate over the boundaries that correspond to the size of the modulus.
For instance, you can make the 128-bit register XMM1 to be zero in one of the following ways:
PXOR XMM1, XMM1 ; Subtraction modulo 2^1
PSUBB XMM1, XMM1 ; Subtraction modulo 2^8
PSUBW XMM1, XMM1 ; Subtraction modulo 2^16
PSUBD XMM1, XMM1 ; Subtraction modulo 2^32
PSUBQ XMM1, XMM1 ; Subtraction modulo 2^64
For XOR, i.e. subtraction modulo 2^1, the size of a chunk is just 1 bit, so the propagation of the carry inside the chunk happens to do nothing.
There are no special rules for XOR, its behavior is the same as for any other subtraction, any behavior that seems special is caused by the facts that the numbers 1 (size in bits of the integer residue) and 0 (number of carry propagations inside a number having the size of the residue) are somewhat more special numbers than the other cardinal numbers.
When you do not do those 5 operations inside a single ALU, but with separate adders, the shorter is the number of bits over which the carry must propagate, the faster is the logic device. But when a single ALU does all 5, the speed of the ALU is a little slower than the slowest of those 5 (a little slower because there are additional control gates for selecting the desired operation).
The other bitwise operations are also just particular cases of more general vector operations. Each of the 3 most important bitwise operations is the 1-bit limit of 2 operations which are distinct for numbers with sizes greater than 1 bit, but which are equivalent for 1-bit numbers. While XOR is just addition or subtraction of 1-bit numbers, AND is just minimum or multiplication of 1-bit numbers, and OR is just maximum of 1-bit numbers or the 1-bit version of the function that gives the probability for 1 of 2 events to happen (i.e. difference between sum and product).
And in practice it is very likely that XOR and the variously sized vector ADDs and SUBs are implemented exactly by the same ALU circuitry, parameterized by a bitmasks of the carry lines to enable (none for XOR, all except the vector size boundaries for the vector operations).
It might be because XOR is rarely (in terms of static count, dynamically it surely appears a lot in some hot loops) used for anything else, so it is easier to spot and identify as "special" if you are writing manual assembly.
And helps with SMT
Edit: this is apparently not the case, see @tliltocatl's comment down the thread
What's SMT in this context?
Simultaneous Multi-Threading (hyper-threading as Intel calls it). I'm not a cpu guy, but I think the ALU used for subtraction would be a more valuable resource to leave available to the other thread than whatever implements a xor. Hence you prefer to use the xor for zeroing and conserve the ALU for other threads to use.
I don't think that's how it works.
- Normally ALU implements all "light" operations (i. e. add/sub/and/or/xor) in a single block, separating them would result in far more interconnect overhead. Often, CPUs have specialized adder-only units for address generation, but never a xor-specialized block.
- All CPUs that implement hyper-threading also optimize a XOR EAX,EAX into MOV EAX,ZERO/SET FLAGS (where ZERO is an invisible zero register just like on Itanium and RISCs). This helps register renaming and eliminates a spurious dependency.
- The XOR trick is about as old as 8086 if not older.
Right. Keeping down the number of slots the scheduler and bypass network need to worry about is an important design pressure.
By the time you get to a CPU complex enough to be to have SMT it is likely to detect these “clear register” patterns and special case them.
XOR would also be handled by the ALU, the L is for logic.
Most CPU use the same ALU for xor and sub.
XOR appears a lot in any code touching encryption.
PS. What is static vs dynamic count?
Static count - how many times an instruction appears in a binary (or assembly source).
Dynamic count - how many times an opcode gets executed.
I. e. an instruction that doesn't appear often in code, but comes up in some hot loops (like encryption) would have low static and high dynamic.
The hw implementation of xor is simpler than sub, so it should consume slightly less energy. Wondering how much energy was saved in the whole world by using xor instead of sub.
I doubt any of that is measurable, since all ALU operations are usually implemented with the same logic (e.g. see https://www.righto.com/2013/09/the-z-80-has-4-bit-alu-heres-...)
I guess everything what was saved was burned by the first useless image created per AI
My favorite (admittedly not super useful) trick in this domain is that sbb eax, eax breaks the dependency on the previous value of eax (just like xor and sub) and only depends on the carry flag. arm64 is less obtuse and just gives you csetm (special case of csinv) for this purpose.
I vaguely remember we used the XOR trick on processors other than Intel, so it may not be Intel-specific.
In principle, sub requires 4 steps:
1. Move both operands to the ALU
2. Invert second operand (twos complement convert)
3. Add (which internally is just XOR plus carry propagate)
4. Move result to proper result register.
This is absolutely not how modern processors do it in practice; there are many shortcuts, but at least with pure XOR you don't need twos complement conversion or carry propagation.
Source: Wrote microcode at work a million years ago when designing a GPU.
Looking at some random 1989 Zenith 386SX bios written in assembly so purely programmer preferences:
8 'sub al, al', 14 'sub ah, ah', 3 'sub ax, ax'
26 'xor al, al', 43 'xor ah, ah', 3 'xor ax, ax'
edit: checked a 2010 bios and not a single 'sub x, x'
Could be used to express 1 bit of information in some non-obvious convention.
Back in the stone ages XOR ing was just 1 byte of opcode. Habbits stick. In effect XORing is no longer faster since a long time.
The article’s point is about why XOR is preferred over SUB, both being one byte.
MOV is right out.
The XOR trick is implemented as a (malloc from register file) on modern processors, implemented in the decoder and it won't even issue a uOp to the execution pipelines.
Its basically free today. Of course, mov RAX, 0 is also free and does the same thing. But CPUs have limited decoder lengths per clock tick, so the more instructions you fit in a given size, the more parallel a modern CPU can potentially execute.
So.... definitely still use XOR trick today. But really, let the compiler handle it. Its pretty good at keeping track of these things in practice.
-----------
I'm not sure if "sub" is hard-coded to be recognized in the decoder as a zero'd out allocation from the register file. There's only certain instructions that have been guaranteed to do this by Intel/AMD.
Depending on what's stone-age for you, a SUB with a register was also only one byte, and was the same cost as XOR, at least in the Intel/Zilog lineage all the way back to the 70s ;)