]> git.ipfire.org Git - thirdparty/kernel/stable.git/commit
RISC-V: Don't enable all interrupts in trap_init()
authorAnup Patel <anup.patel@wdc.com>
Sun, 2 Feb 2020 11:02:02 +0000 (16:32 +0530)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 5 Mar 2020 15:45:18 +0000 (16:45 +0100)
commitb41d6924d34d6003e9e2f040f6f80c3ee0e1b155
tree53e8d9236f01d53b4b26cccd151afcccb9841a63
parentdc374cec2383c3a462ac0f6c95cea4007795e94f
RISC-V: Don't enable all interrupts in trap_init()

commit 6a1ce99dc4bde564e4a072936f9d41f4a439140e upstream.

Historically, we have been enabling all interrupts for each
HART in trap_init(). Ideally, we should only enable M-mode
interrupts for M-mode kernel and S-mode interrupts for S-mode
kernel in trap_init().

Currently, we get suprious S-mode interrupts on Kendryte K210
board running M-mode NO-MMU kernel because we are enabling all
interrupts in trap_init(). To fix this, we only enable software
and external interrupt in trap_init(). In future, trap_init()
will only enable software interrupt and PLIC driver will enable
external interrupt using CPU notifiers.

Fixes: a4c3733d32a7 ("riscv: abstract out CSR names for supervisor vs machine mode")
Signed-off-by: Anup Patel <anup.patel@wdc.com>
Reviewed-by: Atish Patra <atish.patra@wdc.com>
Tested-by: Palmer Dabbelt <palmerdabbelt@google.com> [QMEU virt machine with SMP]
[Palmer: Move the Fixes up to a newer commit]
Reviewed-by: Palmer Dabbelt <palmerdabbelt@google.com>
Signed-off-by: Palmer Dabbelt <palmerdabbelt@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
arch/riscv/kernel/traps.c