On 02/23/10 20:27, Manuel Bouyer wrote:
On Mon, Feb 22, 2010 at 10:25:35AM +0100, Jean-Yves Migeon wrote:
Second small patch [2] contains fixes regarding UVM:
- RLIMIT_RSS which is now set to VM_MAXUSER_ADDRESS (shouldn't it be set to
MIN(VM_MAXUSER_ADDRESS, ptoa(uvmexp.free) instead?),
I think it should be MIN(VM_MAXUSER_ADDRESS, ptoa(uvmexp.free).
Otherwise it's a change in behavior not related to i386PAE.
Just checked, uvmexp elements are "int", so IMHO
ctob((rlim_t)uvmexp.free) is less error-prone (again, risk of overflow
with free > 4GB).
Not that it could happen on a daily basis :o
- ptoa => ctob use (to avoid vaddr_t casts, which might be bad for systems
with total RAM> 4GB)
- a small fix inside AGP heuristics to avoid masking high order bits
Compile tested for ALL, GENERIC, XEN3 kernels for i386, and GENERIC, XEN3*
for amd64.
Looks good.
Committed, thanks
In case someone wonders, if PAE is added as default option for ALL, more
fixes will be needed in some drivers under sys/dev/ (paddr_t vs unsigned
long issues). Question is: should it be added to ALL for additional checks?
I think so. Such issues should be fixed in MI code anyway. paddr_t !=
unsigned long is something we want to support.
Will do.
Thanks all for the review.
Cheers,
--
Jean-Yves Migeon
[email protected]
