MIPS: Check TLB before handle_ri_rdhwr() for Loongson-3
authorHuacai Chen <chenhc@lemote.com>
Thu, 16 Mar 2017 13:00:26 +0000 (21:00 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 12 Apr 2017 10:41:14 +0000 (12:41 +0200)
commit3d5e13d891b0e7f4b7eee97aea1c15037a994d68
tree620a551bbbda6e9b6657fd516048a243072ca6ff
parentc7f6633f7657a56bdb8628c129de6c3079eb7d07
MIPS: Check TLB before handle_ri_rdhwr() for Loongson-3

commit 5a34133167dce36666ea054e30a561b7f4413b7f upstream.

Loongson-3's micro TLB (ITLB) is not strictly a subset of JTLB. That
means: when a JTLB entry is replaced by hardware, there may be an old
valid entry exists in ITLB. So, a TLB miss exception may occur while
handle_ri_rdhwr() is running because it try to access EPC's content.
However, handle_ri_rdhwr() doesn't clear EXL, which makes a TLB Refill
exception be treated as a TLB Invalid exception and tlbp may fail. In
this case, if FTLB (which is usually set-associative instead of set-
associative) is enabled, a tlbp failure will cause an invalid tlbwi,
which will hang the whole system.

This patch rename handle_ri_rdhwr_vivt to handle_ri_rdhwr_tlbp and use
it for Loongson-3. It try to solve the same problem described as below,
but more straightforwards.

https://patchwork.linux-mips.org/patch/12591/

I think Loongson-2 has the same problem, but it has no FTLB, so we just
keep it as is.

Signed-off-by: Huacai Chen <chenhc@lemote.com>
Cc: Rui Wang <wangr@lemote.com>
Cc: John Crispin <john@phrozen.org>
Cc: Steven J . Hill <Steven.Hill@caviumnetworks.com>
Cc: Fuxin Zhang <zhangfx@lemote.com>
Cc: Zhangjin Wu <wuzhangjin@gmail.com>
Cc: Huacai Chen <chenhc@lemote.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/15753/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
arch/mips/kernel/genex.S
arch/mips/kernel/traps.c