2007-05-02 22:41:41

by Ulrich Drepper

[permalink] [raw]
Subject: Why ssse3?

Note the extra 's'. We use "sse" and "sse2", but "ssse3". I assume
it's a typo.


Signed-off-by: Ulrich Drepper <[email protected]>

--- arch/i386/kernel/cpu/proc.c 2007-02-15 11:21:18.000000000 -0800
+++ arch/i386/kernel/cpu/proc.c-new 2007-05-02 15:31:17.000000000 -0700
@@ -46,7 +46,7 @@

/* Intel-defined (#2) */
"pni", NULL, NULL, "monitor", "ds_cpl", "vmx", "smx", "est",
- "tm2", "ssse3", "cid", NULL, NULL, "cx16", "xtpr", NULL,
+ "tm2", "sse3", "cid", NULL, NULL, "cx16", "xtpr", NULL,
NULL, NULL, "dca", NULL, NULL, NULL, NULL, "popcnt",
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,


--
➧ Ulrich Drepper ➧ Red Hat, Inc. ➧ 444 Castro St ➧ Mountain View, CA ❖


Attachments:
signature.asc (251.00 B)
OpenPGP digital signature

2007-05-02 22:45:54

by Ismail Dönmez

[permalink] [raw]
Subject: Re: Why ssse3?

On Thursday 03 May 2007 01:41:22 Ulrich Drepper wrote:
> Note the extra 's'. We use "sse" and "sse2", but "ssse3". I assume
> it's a typo.

This might not be a typo see http://en.wikipedia.org/wiki/SSSE3

Regards,
ismail


Attachments:
(No filename) (223.00 B)
signature.asc (189.00 B)
This is a digitally signed message part.
Download all attachments

2007-05-02 22:46:18

by Andi Kleen

[permalink] [raw]
Subject: Re: Why ssse3?

On Thursday 03 May 2007 00:41:22 Ulrich Drepper wrote:
> Note the extra 's'. We use "sse" and "sse2", but "ssse3". I assume
> it's a typo.

Nope. SSE3 != SSSE3. The additional S means Supplemential.

It's probably because the few changes didn't justify a SSE4

-Andi

2007-05-02 22:56:45

by Ulrich Drepper

[permalink] [raw]
Subject: Re: Why ssse3?

Andi Kleen wrote:
> Nope. SSE3 != SSSE3. The additional S means Supplemential.
>
> It's probably because the few changes didn't justify a SSE4

OK, the problem is that the actual sse3 bit is misnamed. According to
Intel's docs bit 0 of ECX is "sse", the kernel uses "pni". Too bad.

--
➧ Ulrich Drepper ➧ Red Hat, Inc. ➧ 444 Castro St ➧ Mountain View, CA ❖


Attachments:
signature.asc (251.00 B)
OpenPGP digital signature

2007-05-03 02:28:48

by H. Peter Anvin

[permalink] [raw]
Subject: Re: Why ssse3?

Ulrich Drepper wrote:
> Andi Kleen wrote:
>> Nope. SSE3 != SSSE3. The additional S means Supplemential.
>>
>> It's probably because the few changes didn't justify a SSE4
>
> OK, the problem is that the actual sse3 bit is misnamed. According to
> Intel's docs bit 0 of ECX is "sse", the kernel uses "pni". Too bad.

Intel has a nasty habit of renaming things after they are already
deployed in Linux.

-hpa

2007-05-03 08:54:18

by Andi Kleen

[permalink] [raw]
Subject: Re: Why ssse3?

On Thursday 03 May 2007 00:56:26 Ulrich Drepper wrote:
> Andi Kleen wrote:
> > Nope. SSE3 != SSSE3. The additional S means Supplemential.
> >
> > It's probably because the few changes didn't justify a SSE4
>
> OK, the problem is that the actual sse3 bit is misnamed. According to
> Intel's docs bit 0 of ECX is "sse", the kernel uses "pni". Too bad.

PNI (Prescott New Instructions) was the original engineering code name. Unfortunately
it was added too early before the marketing name was known and then it couldn't be
changed anymore.

-Andi

2007-05-03 09:02:34

by Avi Kivity

[permalink] [raw]
Subject: Re: Why ssse3?

Andi Kleen wrote:
> On Thursday 03 May 2007 00:56:26 Ulrich Drepper wrote:
>
>> Andi Kleen wrote:
>>
>>> Nope. SSE3 != SSSE3. The additional S means Supplemential.
>>>
>>> It's probably because the few changes didn't justify a SSE4
>>>
>> OK, the problem is that the actual sse3 bit is misnamed. According to
>> Intel's docs bit 0 of ECX is "sse", the kernel uses "pni". Too bad.
>>
>
> PNI (Prescott New Instructions) was the original engineering code name. Unfortunately
> it was added too early before the marketing name was known and then it couldn't be
> changed anymore.
>

Perhaps sse3 could be added as an alias to pni.


--
error compiling committee.c: too many arguments to function

2007-05-03 09:07:45

by Andi Kleen

[permalink] [raw]
Subject: Re: Why ssse3?

On Thursday 03 May 2007 11:02:31 Avi Kivity wrote:
> Andi Kleen wrote:
> > On Thursday 03 May 2007 00:56:26 Ulrich Drepper wrote:
> >
> >> Andi Kleen wrote:
> >>
> >>> Nope. SSE3 != SSSE3. The additional S means Supplemential.
> >>>
> >>> It's probably because the few changes didn't justify a SSE4
> >>>
> >> OK, the problem is that the actual sse3 bit is misnamed. According to
> >> Intel's docs bit 0 of ECX is "sse", the kernel uses "pni". Too bad.
> >>
> >
> > PNI (Prescott New Instructions) was the original engineering code name. Unfortunately
> > it was added too early before the marketing name was known and then it couldn't be
> > changed anymore.
> >
>
> Perhaps sse3 could be added as an alias to pni.

That was considered at some point, but it would be a ugly special case and is
probably not worth it. Usually these flags are Greek for most people anyways
(and something else for the Greek speaking people @)

-Andi

2007-05-03 16:03:42

by Roland Dreier

[permalink] [raw]
Subject: Re: Why ssse3?

> PNI (Prescott New Instructions) was the original engineering code name. Unfortunately
> it was added too early before the marketing name was known and then it couldn't be
> changed anymore.

... and just to make things more fun, SSE4 is sometimes called Penryn
New Instructions -- PNI all over again.

- R.