On Tue, Jun 13, 2023 at 9:01 PM Clément Léger <cleger@xxxxxxxxxxxx> wrote: > > > > On 29/05/2023 10:45, guoren@xxxxxxxxxx wrote: > > From: Guo Ren <guoren@xxxxxxxxxxxxxxxxx> > > > > Add independent irq stacks for percpu to prevent kernel stack overflows. > > It is also compatible with VMAP_STACK by arch_alloc_vmap_stack. > > > > Tested-by: Jisheng Zhang <jszhang@xxxxxxxxxx> > > Signed-off-by: Guo Ren <guoren@xxxxxxxxxxxxxxxxx> > > Signed-off-by: Guo Ren <guoren@xxxxxxxxxx> > > Cc: Conor Dooley <conor.dooley@xxxxxxxxxxxxx> > > --- > > arch/riscv/Kconfig | 7 ++++++ > > arch/riscv/include/asm/irq_stack.h | 32 +++++++++++++++++++++++++ > > arch/riscv/include/asm/thread_info.h | 2 ++ > > arch/riscv/kernel/irq.c | 33 ++++++++++++++++++++++++++ > > arch/riscv/kernel/traps.c | 35 ++++++++++++++++++++++++++-- > > 5 files changed, 107 insertions(+), 2 deletions(-) > > create mode 100644 arch/riscv/include/asm/irq_stack.h > > > > diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig > > index fa256f2e23c1..44b4c9690f94 100644 > > --- a/arch/riscv/Kconfig > > +++ b/arch/riscv/Kconfig > > @@ -588,6 +588,13 @@ config FPU > > > > If you don't know what to do here, say Y. > > > > +config IRQ_STACKS > > + bool "Independent irq stacks" if EXPERT > > + default y > > + select HAVE_IRQ_EXIT_ON_IRQ_STACK > > + help > > + Add independent irq stacks for percpu to prevent kernel stack overflows. > > + > > endmenu # "Platform type" > > > > menu "Kernel features" > > diff --git a/arch/riscv/include/asm/irq_stack.h b/arch/riscv/include/asm/irq_stack.h > > new file mode 100644 > > index 000000000000..b0dcee9a3fa2 > > --- /dev/null > > +++ b/arch/riscv/include/asm/irq_stack.h > > @@ -0,0 +1,32 @@ > > +/* SPDX-License-Identifier: GPL-2.0 */ > > + > > +#ifndef _ASM_RISCV_IRQ_STACK_H > > +#define _ASM_RISCV_IRQ_STACK_H > > + > > +#include <linux/bug.h> > > +#include <linux/gfp.h> > > +#include <linux/kconfig.h> > > +#include <linux/vmalloc.h> > > +#include <linux/pgtable.h> > > +#include <asm/thread_info.h> > > + > > +DECLARE_PER_CPU(ulong *, irq_stack_ptr); > > + > > +#ifdef CONFIG_VMAP_STACK > > +/* > > + * To ensure that VMAP'd stack overflow detection works correctly, all VMAP'd > > + * stacks need to have the same alignment. > > + */ > > +static inline unsigned long *arch_alloc_vmap_stack(size_t stack_size, int node) > > +{ > > + void *p; > > + > > + BUILD_BUG_ON(!IS_ENABLED(CONFIG_VMAP_STACK)); > > Hi Guo, > > Since this function is already guarded with #ifdef CONFIG_VMAP_STACK, I > guess this BUILD_BUG_ON() is unnecessary). Yes, my carelessness. I would remove it in the next version. > > Clément -- Best Regards Guo Ren