forked from luck/tmp_suning_uos_patched
6373467482
Commit 775b089aef ("MIPS: tlbex: Remove cpu_has_local_ebase") removed
generating tlb refill handlers for every CPU, which was needed for
generating per node exception handlers on IP27. Instead of resurrecting
(and fixing) refill handler generation, we simply copy all exception
vectors from the boot node to the other nodes. Also remove the config
option since the memory tradeoff for expection handler replication
is just 8k per node.
Signed-off-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
Signed-off-by: Paul Burton <paulburton@kernel.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Paul Burton <paul.burton@mips.com>
Cc: James Hogan <jhogan@kernel.org>
Cc: linux-mips@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
41 lines
1.2 KiB
Plaintext
41 lines
1.2 KiB
Plaintext
# SPDX-License-Identifier: GPL-2.0
|
|
choice
|
|
prompt "Node addressing mode"
|
|
depends on SGI_IP27
|
|
default SGI_SN_M_MODE
|
|
|
|
config SGI_SN_M_MODE
|
|
bool "IP27 M-Mode"
|
|
help
|
|
The nodes of Origin, Onyx, Fuel and Tezro systems can be configured
|
|
in either N-Modes which allows for more nodes or M-Mode which allows
|
|
for more memory. Your hardware is almost certainly running in
|
|
M-Mode, so choose M-mode here.
|
|
|
|
config SGI_SN_N_MODE
|
|
bool "IP27 N-Mode"
|
|
help
|
|
The nodes of Origin, Onyx, Fuel and Tezro systems can be configured
|
|
in either N-Modes which allows for more nodes or M-Mode which allows
|
|
for more memory. Your hardware is almost certainly running in
|
|
M-Mode, so choose M-mode here.
|
|
|
|
endchoice
|
|
|
|
config MAPPED_KERNEL
|
|
bool "Mapped kernel support"
|
|
depends on SGI_IP27
|
|
help
|
|
Change the way a Linux kernel is loaded into memory on a MIPS64
|
|
machine. This is required in order to support text replication on
|
|
NUMA. If you need to understand it, read the source code.
|
|
|
|
config REPLICATE_KTEXT
|
|
bool "Kernel text replication support"
|
|
depends on SGI_IP27
|
|
select MAPPED_KERNEL
|
|
help
|
|
Say Y here to enable replicating the kernel text across multiple
|
|
nodes in a NUMA cluster. This trades memory for speed.
|
|
|