Re: asm/uaccess.h reimplementation, patch. [was: Re: 2.1 kernel bloat revisited]

Harald Koenig (koenig@tat.physik.uni-tuebingen.de)
Sun, 30 Mar 1997 18:25:50 +0200 (MET DST)

> where mem_mask is depending on what 'mode' we are: user or kernel. This
> restricts memory configurations to 2^N user-space virtual memory sizes ...
> Dave, is this acceptable on Sparc? [the trick is to make the user-space
> size 2^N, not physical memory ... thus we can have 3.5G physical memory
> setups too: 512M user-space virtual memory, 3.5G physical memory]

limiting user-space virtual memory to 512M (esp. for 3.5G physical memory boxes)
doesn't seem to make any sense.
and more important this is another artificial limit for user-space size
which we shouldn't introduce just for the sake of performace. there will be
more and more "number crunching" PPros around with lots of memory used
for large size (memory footprint) computations

Harald

--
All SCSI disks will from now on                     ___       _____
be required to send an email notice                0--,|    /OOOOOOO\
24 hours prior to complete hardware failure!      <_/  /  /OOOOOOOOOOO\
                                                    \  \/OOOOOOOOOOOOOOO\
                                                      \ OOOOOOOOOOOOOOOOO|//
Harald Koenig,                                         \/\/\/\/\/\/\/\/\/
Inst.f.Theoret.Astrophysik                              //  /     \\  \
koenig@tat.physik.uni-tuebingen.de                     ^^^^^       ^^^^^