This patch allows the LoongArch BPF JIT to handle recoverable memory
access errors generated by BPF_PROBE_MEM* instructions.

When a BPF program performs memory access operations, the instructions
it executes may trigger ADEM exceptions. The kernel’s built-in BPF
exception table mechanism (EX_TYPE_BPF) will generate corresponding
exception fixup entries in the JIT compilation phase; however, the
architecture-specific trap handling function needs to proactively call
the common fixup routine to achieve exception recovery.

do_ade(): fix EX_TYPE_BPF memory access exceptions for BPF programs,
ensure safe execution.

Signed-off-by: Chenghao Duan <duanchenghao@kylinos.cn>
---
 arch/loongarch/kernel/traps.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/arch/loongarch/kernel/traps.c b/arch/loongarch/kernel/traps.c
index da5926fead4a..4cf72e0af6a3 100644
--- a/arch/loongarch/kernel/traps.c
+++ b/arch/loongarch/kernel/traps.c
@@ -534,7 +534,13 @@ asmlinkage void noinstr do_fpe(struct pt_regs *regs, unsigned long fcsr)
 
 asmlinkage void noinstr do_ade(struct pt_regs *regs)
 {
-	irqentry_state_t state = irqentry_enter(regs);
+	irqentry_state_t state;
+	unsigned int esubcode = FIELD_GET(CSR_ESTAT_ESUBCODE, regs->csr_estat);
+
+	if ((esubcode == EXSUBCODE_ADEM) && fixup_exception(regs))
+		return;
+
+	state = irqentry_enter(regs);
 
 	die_if_kernel("Kernel ade access", regs);
 	force_sig_fault(SIGBUS, BUS_ADRERR, (void __user *)regs->csr_badvaddr);
-- 
2.25.1

Hi, Chenghao,

On Fri, Dec 12, 2025 at 5:11 PM Chenghao Duan <duanchenghao@kylinos.cn> wrote:
>
> This patch allows the LoongArch BPF JIT to handle recoverable memory
> access errors generated by BPF_PROBE_MEM* instructions.
>
> When a BPF program performs memory access operations, the instructions
> it executes may trigger ADEM exceptions. The kernel’s built-in BPF
> exception table mechanism (EX_TYPE_BPF) will generate corresponding
> exception fixup entries in the JIT compilation phase; however, the
> architecture-specific trap handling function needs to proactively call
> the common fixup routine to achieve exception recovery.
>
> do_ade(): fix EX_TYPE_BPF memory access exceptions for BPF programs,
> ensure safe execution.
>
> Signed-off-by: Chenghao Duan <duanchenghao@kylinos.cn>
> ---
>  arch/loongarch/kernel/traps.c | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
>
> diff --git a/arch/loongarch/kernel/traps.c b/arch/loongarch/kernel/traps.c
> index da5926fead4a..4cf72e0af6a3 100644
> --- a/arch/loongarch/kernel/traps.c
> +++ b/arch/loongarch/kernel/traps.c
> @@ -534,7 +534,13 @@ asmlinkage void noinstr do_fpe(struct pt_regs *regs, unsigned long fcsr)
>
>  asmlinkage void noinstr do_ade(struct pt_regs *regs)
>  {
> -       irqentry_state_t state = irqentry_enter(regs);
> +       irqentry_state_t state;
> +       unsigned int esubcode = FIELD_GET(CSR_ESTAT_ESUBCODE, regs->csr_estat);
> +
> +       if ((esubcode == EXSUBCODE_ADEM) && fixup_exception(regs))
> +               return;
No chance for ADEF? And I don't think ixup_exception() can be done out
of irqentry_enter().

This patch is needed by BPF but not part of BPF, so I think the
subject should be:
LoongArch: Enable exception fixup for specific ADE subcode

Huacai

> +
> +       state = irqentry_enter(regs);
>
>         die_if_kernel("Kernel ade access", regs);
>         force_sig_fault(SIGBUS, BUS_ADRERR, (void __user *)regs->csr_badvaddr);
> --
> 2.25.1
>

On Sat, Dec 13, 2025 at 09:16:00PM +0800, Huacai Chen wrote:
> Hi, Chenghao,
> 
> On Fri, Dec 12, 2025 at 5:11 PM Chenghao Duan <duanchenghao@kylinos.cn> wrote:
> >
> > This patch allows the LoongArch BPF JIT to handle recoverable memory
> > access errors generated by BPF_PROBE_MEM* instructions.
> >
> > When a BPF program performs memory access operations, the instructions
> > it executes may trigger ADEM exceptions. The kernel’s built-in BPF
> > exception table mechanism (EX_TYPE_BPF) will generate corresponding
> > exception fixup entries in the JIT compilation phase; however, the
> > architecture-specific trap handling function needs to proactively call
> > the common fixup routine to achieve exception recovery.
> >
> > do_ade(): fix EX_TYPE_BPF memory access exceptions for BPF programs,
> > ensure safe execution.
> >
> > Signed-off-by: Chenghao Duan <duanchenghao@kylinos.cn>
> > ---
> >  arch/loongarch/kernel/traps.c | 8 +++++++-
> >  1 file changed, 7 insertions(+), 1 deletion(-)
> >
> > diff --git a/arch/loongarch/kernel/traps.c b/arch/loongarch/kernel/traps.c
> > index da5926fead4a..4cf72e0af6a3 100644
> > --- a/arch/loongarch/kernel/traps.c
> > +++ b/arch/loongarch/kernel/traps.c
> > @@ -534,7 +534,13 @@ asmlinkage void noinstr do_fpe(struct pt_regs *regs, unsigned long fcsr)
> >
> >  asmlinkage void noinstr do_ade(struct pt_regs *regs)
> >  {
> > -       irqentry_state_t state = irqentry_enter(regs);
> > +       irqentry_state_t state;
> > +       unsigned int esubcode = FIELD_GET(CSR_ESTAT_ESUBCODE, regs->csr_estat);
> > +
> > +       if ((esubcode == EXSUBCODE_ADEM) && fixup_exception(regs))
> > +               return;
> No chance for ADEF? And I don't think ixup_exception() can be done out
> of irqentry_enter().

At present, exception fixes are only applied to memory-type BPF programs,
with no handling implemented for ADEF.

In the next version, fixup_exception will be processed internally within
irqentry_enter()/irqentry_exit, and the title will be revised with
reference to your suggestions.

Chenghao

> 
> This patch is needed by BPF but not part of BPF, so I think the
> subject should be:
> LoongArch: Enable exception fixup for specific ADE subcode
> 
> Huacai
> 
> > +
> > +       state = irqentry_enter(regs);
> >
> >         die_if_kernel("Kernel ade access", regs);
> >         force_sig_fault(SIGBUS, BUS_ADRERR, (void __user *)regs->csr_badvaddr);
> > --
> > 2.25.1
> >