On Thu, Dec 2, 2010 at 8:37 AM, Mark Kettenis <mark.kettenis@xs4all.nl>
wrote:
quoted text
>> Date: Wed, 1 Dec 2010 23:05:41 -0500 (EST)
>> From: Ted Unangst <ted.unangst@gmail.com>
>>
>> reminder that "i tested this and it works" responses are more helpful than
>> "yo this is awesome" responses. :)
>
> Unless Theo withdraws hos objection to decoupling MAXDSIZ and BRKSIZ,
> there isn't much point in testing this :(.

I'm not aware of any objection other than "this needs testing on more
than i386".

quoted text
>> Index: arch/i386/include/vmparam.h
>> ===================================================================
>> RCS file: /cvs/src/sys/arch/i386/include/vmparam.h,v
>> retrieving revision 1.43
>> diff -u -r1.43 vmparam.h
>> --- arch/i386/include/vmparam.h       16 Jun 2009 16:42:41 -0000      1.43
>> +++ arch/i386/include/vmparam.h       30 Jun 2010 02:17:34 -0000
>> @@ -63,7 +63,10 @@
>>  #define      DFLDSIZ         (64*1024*1024)          /* initial data size
limit */
quoted text
>>  #endif
>>  #ifndef MAXDSIZ
>> -#define      MAXDSIZ         (1024*1024*1024)        /* max data size */
>> +#define      MAXDSIZ         (2U*1024*1024*1024 - 4096)      /* max data
size */
quoted text
>
> what's the reason for the - 4096 here?

I am worried about signed bugs.  I didn't find any, but I do not want
the the maxdsiz to end up in a signed int and turn negative (possibly
in some crappy app code).  it's more paranoia than any justified
reason.