Re: [tortoisesvn] r20779 committed - * use SSE2 to push FastLowerCaseConversion execution...

On 28.01.2011 21:42, Stefan Fuhrmann wrote:
>> On Fri, Jan 28, 2011 at 01:57,<tortoisesvn at googlecode dot com> wrote:
>>> Revision: 20779
>>> Author: stefan.fuhrmann.1974
>>> Date: Thu Jan 27 16:54:13 2011
>>> Log: * use SSE2 to push FastLowerCaseConversion execution
>>> * skip unnecessary path un-escaping
>>> http://code.google.com/p/tortoisesvn/source/detail?r=20779
>>
>> You can't use SSE2 unconditionally!
> I was about to add that on your request but you bet my by a few minutes ;)
>
>> Believe me: there are still a *lot* of PCs in use (even new ones!)
>> that use processors that don't have or support SSE2.
>
> Out of curiosity: what CPUs are that? AMD Geode?

Not really. In my company, we released our software with code that uses
SSE2. Soon afterwards we got a lot of complains that the software
crashed. We found out that their processors didn't support SSE2. Even
though they bought the PCs half a year ago. They were AMD processors.
Then after a lot of digging and countless hours we found that their
supplier used cheap processors which had SSE disabled! They used
defective processors and 'repaired' them by disabling the defective
parts, which in these cases was the SSE extensions.

Since that supplier is one of the bigger suppliers in south Germany, we
had to ship another version of our software which doesn't use SSE anymore.

Stefan

-- 
        ___
   oo  // \\      "De Chelonian Mobile"
  (_,\/ \_/ \     TortoiseSVN
    \ \_/_\_/>    The coolest Interface to (Sub)Version Control
    /_/   \_\     http://tortoisesvn.net
------------------------------------------------------
http://tortoisesvn.tigris.org/ds/viewMessage.do?dsForumId=757&dsMessageId=2701247
To unsubscribe from this discussion, e-mail: [dev-unsubscribe_at_tortoisesvn.tigris.org].