Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
DFDV and DDIV Instructions
53 views
Skip to first unread message
Rob Doyle
May 21, 2013, 10:20:21 PM5/21/13
to
So...
I'm enjoying the misery of debugging the DFDV and DDIV instructions on
the KS10 FPGA.
And I was wondering...
Why are these instructions named the way they are? It would seem to me
that these are Quad Word instructions not Double Word instructions.
I don't think the KA10 that I used had any of these.
Rob.
Johnny Billquist
May 22, 2013, 7:56:59 AM5/22/13
to
They operate on double-words. Ie. 72 bits. (DFAD, DFSB, DFMP and DFDV)
And yes, the KA10 lacks hardware for the double precision instructions.
Btw, I can't seem to find a DDIV. Maybe I'm looking at the wrong place.
Johnny
Johnny Billquist
May 22, 2013, 8:16:38 AM5/22/13
to
that DDIV is not. Yeah, DMUL gives a quad result, while DDIV takes a
quad argument.
Note, the DFxxx instructions exists in both the KI and KL, while the
DADD, DSUB, DMUL and DDIV are KL only.
Johnny
Rich Alderson
May 22, 2013, 7:47:55 PM5/22/13
to
Rob Doyle <radi...@gmail.com> writes:
> So...
> I'm enjoying the misery of debugging the DFDV and DDIV instructions on
> the KS10 FPGA.
> And I was wondering...
> Why are these instructions named the way they are? It would seem to me
> that these are Quad Word instructions not Double Word instructions.
They're named for the size of the *inputs*, not the size of the outputs.
> So...
> I'm enjoying the misery of debugging the DFDV and DDIV instructions on
> the KS10 FPGA.
> And I was wondering...
> Why are these instructions named the way they are? It would seem to me
> that these are Quad Word instructions not Double Word instructions.
> I don't think the KA10 that I used had any of these.
--
Rich Alderson ne...@alderson.users.panix.com
the russet leaves of an autumn oak/inspire once again the failed poet/
to take up his pen/and essay to place his meagre words upon the page...
Rob Doyle
May 22, 2013, 9:27:15 PM5/22/13
to
On 5/22/2013 4:47 PM, Rich Alderson wrote:
> Rob Doyle <radi...@gmail.com> writes:
>
>> So...
>
>> I'm enjoying the misery of debugging the DFDV and DDIV instructions on
>> the KS10 FPGA.
>
>> And I was wondering...
>
>> Why are these instructions named the way they are? It would seem to me
>> that these are Quad Word instructions not Double Word instructions.
>
> They're named for the size of the *inputs*, not the size of the outputs.
I guess.
> Rob Doyle <radi...@gmail.com> writes:
>
>> So...
>
>> I'm enjoying the misery of debugging the DFDV and DDIV instructions on
>> the KS10 FPGA.
>
>> And I was wondering...
>
>> Why are these instructions named the way they are? It would seem to me
>> that these are Quad Word instructions not Double Word instructions.
>
> They're named for the size of the *inputs*, not the size of the outputs.
A VAX quadword is 64-bits. Ditto Intel.
At least a Power PC quadword is 128-bits.
I just really found it strange that the 144-bit (141-bit actually)
output of a PDP-10 multiply would be called a double...
Does any know of an arbitrary precision calculator (or one that does at
least 144-bits) that speaks octal?
...now back to trying to remember how non-restoring division worked.
Rob.
glen herrmannsfeldt
May 23, 2013, 2:59:46 AM5/23/13
to
Rob Doyle <radi...@gmail.com> wrote:
(snip)
>>> Why are these instructions named the way they are?
>>> It would seem to me that these are Quad Word instructions
>>> not Double Word instructions.
(snip)
> I just really found it strange that the 144-bit (141-bit actually)
> output of a PDP-10 multiply would be called a double...
> Does any know of an arbitrary precision calculator
> (or one that does at least 144-bits) that speaks octal?
It is usual for multiply to generate a product twice the size
of its operands, and for divide to take a dividend twice the size.
So, single word multiply generates a double word product, and
double word multiply a quad word product. Same for the dividend
in divide.
-- glen
(snip)
>>> Why are these instructions named the way they are?
>>> It would seem to me that these are Quad Word instructions
>>> not Double Word instructions.
> I just really found it strange that the 144-bit (141-bit actually)
> output of a PDP-10 multiply would be called a double...
> Does any know of an arbitrary precision calculator
> (or one that does at least 144-bits) that speaks octal?
of its operands, and for divide to take a dividend twice the size.
So, single word multiply generates a double word product, and
double word multiply a quad word product. Same for the dividend
in divide.
-- glen
Johnny Billquist
May 23, 2013, 3:12:06 AM5/23/13
to
On 2013-05-23 03:27, Rob Doyle wrote:
> On 5/22/2013 4:47 PM, Rich Alderson wrote:
>> Rob Doyle <radi...@gmail.com> writes:
>>
>>> So...
>>
>>> I'm enjoying the misery of debugging the DFDV and DDIV instructions on
>>> the KS10 FPGA.
>>
>>> And I was wondering...
>>
>>> Why are these instructions named the way they are? It would seem to me
>>> that these are Quad Word instructions not Double Word instructions.
>>
>> They're named for the size of the *inputs*, not the size of the outputs.
>
> I guess.
>
> A VAX quadword is 64-bits. Ditto Intel.
>
> At least a Power PC quadword is 128-bits.
>
> I just really found it strange that the 144-bit (141-bit actually)
> output of a PDP-10 multiply would be called a double...
It is not. It's called a quadruple (see
> On 5/22/2013 4:47 PM, Rich Alderson wrote:
>> Rob Doyle <radi...@gmail.com> writes:
>>
>>> So...
>>
>>> I'm enjoying the misery of debugging the DFDV and DDIV instructions on
>>> the KS10 FPGA.
>>
>>> And I was wondering...
>>
>>> Why are these instructions named the way they are? It would seem to me
>>> that these are Quad Word instructions not Double Word instructions.
>>
>> They're named for the size of the *inputs*, not the size of the outputs.
>
> I guess.
>
> A VAX quadword is 64-bits. Ditto Intel.
>
> At least a Power PC quadword is 128-bits.
>
> I just really found it strange that the 144-bit (141-bit actually)
> output of a PDP-10 multiply would be called a double...
http://pdp10.nocrew.org/docs/instruction-set/KL-Only.html). But it has
nothing to do with the number of bits, but that it's multiple of the
word size.
And Rich made it a bit too east for himself. They are called double,
even though one input parameter of one instruction (DDIV) actually takes
a quadruple input as one parameter. But like Glen said, that is pretty
common for integer multiplication and division instructions in various
architectures.
> Does any know of an arbitrary precision calculator (or one that does at
> least 144-bits) that speaks octal?
Johnny
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: b...@softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
0 new messages