| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
| |
Note that mingetty does no longer work with serial console, you have to
switch to another getty like getty_ps. This commit also includes a
fix for a setitimer bug which did prevent getty_ps from working on
older kernels.
|
| |
|
|
| |
come. Stay frightened :-)
|
| |
|
|
| |
is used in __wake_up.
|
| | |
|
| |
|
|
|
|
| |
translations, use a kernel pmd table that points into the kptbl[].
This is to make the generic part of vmalloc()/vfree() find pgd/pmd/pte
that it expects.
|
| |
|
|
|
|
|
|
| |
are stashed in an array of page tables, starting from kptbl[]. The fast
tlbmiss handler quickly checks to see if the faulting address is in the
vmalloc range, and if so, it uses the translations in the kptbl to update
the tlbs. Still to do: tlb invalid faults in the vmalloc range needs to
be handled properly.
|
| |
|
|
|
|
|
| |
distinct from invalid-pmd-table for use during pmd allocation failure (so as
to differentiate pgd_bad and pgd_none). The bad page table must have invalid
pte entries to catch any user references to the range of virtual addresses
it covers.
|
| |
|
|
| |
to their rightful locations.
|
| |
|
|
|
| |
initialization. Make sure to allocate only one page for the page table
(prevent memory leaks), since only one page is freed up.
|
| |
|
|
| |
boundaries.
|
| |
|
|
| |
for the mapped kernel. Compute the "pfn" field properly.
|
| |
|
|
|
| |
nasids that the master nasid found for it. Sanitize the replication
procedures.
|
| |
|
|
|
| |
tell it where to get the kernel data from (ie, which nasid holds the
kernel data). Remove debug cruft from head.S.
|
| |
|
|
| |
from the mapped kernel code.
|
| |
|
|
|
| |
IRIX, to get the same kind of loader behavior. Read comments in
mapped_kernel.h for more details.
|
| | |
|
| | |
|
| |
|
|
| |
mapped kernel text/data for tlb translation to work properly.
|
| | |
|
| |
|
|
|
|
| |
are placed in the cksseg area, and such behavior is tuned thru a make
option (off by default). Basis for future kernel text replication and
partitioning work/study.
|
| |
|
|
|
| |
was not getting generated (coding problem or cpp/as problem, no point
tracking this down).
|
| |
|
|
|
|
|
|
|
|
| |
kernel mode for the first time from user mode (on intr/exception/
syscall etc). If so, we need to set the sp to point to the kernel
stack. On UP kernels, the kernel stack pointer is stored in
the global variable "kernelsp". For SMP kernel, the physical
address of the current task structure is stuffed into the
watchlo/watchhi registers, so on first entry into the kernel,
we need to munge this value properly to setup the sp register.
|
| | |
|
| |
|
|
|
|
|
|
| |
PROM is getting the bit set before entry into kernel).
Fix per cpu frequency reporting.
Fix bug so that we do not clear information about mips4 availability.
Have the slaves flush their cache/tlb and set status based on what
the master processor did.
|
| |
|
|
| |
master.
|
| | |
|
| |
|
|
|
|
| |
only the slave on the same node as the master gets launched, investigating
why the slaves on the other nodes are not ... Slave processors use a
static stack surrently while booting up.
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
So, _all_ page tables in the system _must_ be 2 pages, including
empty_bad_page_table. And, the invalid page table _must_ be initialized
to all invalid entries, so that even if the fast handler picks something
up from an unallocated pmd/ptetable, the processor takes a mm fault.
|
| | |
|
| |
|
|
|
|
|
|
|
| |
- IOC3 driver does no longer use GFP_DMA which given the _very_ small
number of available GFP_DMA pages might have deadlocked the system.
- First cut of Origin support. Last minute change: Do no longer use
ARC memory / MD hub memory configuration information but klconfig.h
stuff. Simpler, faster, shorter.
- Zillions of MIPS64 fixes.
|
| | |
|
| |
|
