Re: [PATCH] Speed-up of libsvn_diff using token counts

On Sun, May 29, 2011 at 6:00 PM, Bert Huijben <bert_at_qqmail.nl> wrote:
>
>
>> -----Original Message-----
>> From: Morten Kloster [mailto:morklo_at_gmail.com]
>> Sent: zondag 29 mei 2011 17:35
>> To: Julian Foad
>> Cc: Mark Phippard; dev_at_subversion.apache.org
>> Subject: Re: [PATCH] Speed-up of libsvn_diff using token counts
>>
>> On Fri, May 27, 2011 at 7:57 PM, Julian Foad <julian.foad_at_wandisco.com>
>> wrote:
>> > Morten Kloster wrote:
>> >> On Fri, May 27, 2011 at 4:55 PM, Julian Foad <julian.foad_at_wandisco.com>
>> wrote:
>> >> > Morten Kloster wrote:
>> >> >> I haven't changed the index/count types yet. What's the right type
>> >> >> to use to get signed 32 bit on 32-bit machines and signed 64 bit
>> >> >> on 64-bit machines?
>> >> >
>> >> > "int"?
>> >>
>> >> Is int guaranteed to correspond to (or be larger than) single-process
>> >> addressable space?
>> >
>> > No.  Some 64-bit platforms use 32-bit ints by default and 64-bit
>> > pointers.
>> >
>> > But do you really need to guarantee that svn's text-diff will cope with
>> > more than 2 billion lines on such a system?  Personally I don't think
>> > so.  If you think that is important, you can use "long int".
>
> long int isn't guaranteed to be 64 bit either. (E.g. on Windows 64 long int
> is just 32 bit)
>
> You need something like intptr_t if you want an integer type of pointer
> size.
>
> But threating more than 4 billion lines of text as just every other textfile
> doesn't seem necessary to me.
>
>        Bert
>
>

And I forgot the attachment, as well... Ok, trying again, using intptr_t this
time.

Morten