[PATCH 5/7] powerpc/entry: add irqentry_state and generic entry support

Christophe Leroy christophe.leroy at csgroup.eu
Sat Oct 12 20:35:53 AEDT 2024



Le 12/10/2024 à 05:56, Luming Yu a écrit :
> generic irq entry support via generic irqentry is added for powerpc.
> There may be duplciate calls and missing callbacks requires further
> work.
> 
> Signed-off-by: Luming Yu <luming.yu at shingroup.cn>

This patch doesn't apply.

Applying: powerpc/entry: add irqentry_state and generic entry support
error: patch failed: arch/powerpc/kernel/interrupt.c:3
error: arch/powerpc/kernel/interrupt.c: patch does not apply
Patch failed at 0005 powerpc/entry: add irqentry_state and generic entry 
support

I tried on v6.11, v6.12-rc2 and next-20241011

> ---
>   arch/powerpc/include/asm/entry-common.h | 32 ++++++++++++++++
>   arch/powerpc/kernel/interrupt.c         | 51 +++++--------------------
>   arch/powerpc/kernel/signal.c            |  7 ++++
>   arch/powerpc/kernel/syscall.c           |  2 -
>   4 files changed, 49 insertions(+), 43 deletions(-)
> 
> diff --git a/arch/powerpc/include/asm/entry-common.h b/arch/powerpc/include/asm/entry-common.h
> index 51f1eb767696..faa829e15b5d 100644
> --- a/arch/powerpc/include/asm/entry-common.h
> +++ b/arch/powerpc/include/asm/entry-common.h
> @@ -3,6 +3,7 @@
>   #define ARCH_POWERPC_ENTRY_COMMON_H
>   
>   #include <linux/user-return-notifier.h>
> +#include <asm/switch_to.h>
>   
>   static inline void arch_exit_to_user_mode_prepare(struct pt_regs *regs,
>   						  unsigned long ti_work)
> @@ -13,4 +14,35 @@ static inline void arch_exit_to_user_mode_prepare(struct pt_regs *regs,
>   
>   #define arch_exit_to_user_mode_prepare arch_exit_to_user_mode_prepare
>   
> +static inline void arch_exit_to_user_mode_work(struct pt_regs *regs,
> +						unsigned long ti_work)
> +{
> +
> +	if (IS_ENABLED(CONFIG_PPC_BOOK3S_64) && IS_ENABLED(CONFIG_PPC_FPU)) {
> +		if (IS_ENABLED(CONFIG_PPC_TRANSACTIONAL_MEM) &&
> +				unlikely((ti_work & _TIF_RESTORE_TM))) {
> +			restore_tm_state(regs);
> +		} else {
> +			unsigned long mathflags = MSR_FP;
> +
> +			if (cpu_has_feature(CPU_FTR_VSX))
> +				mathflags |= MSR_VEC | MSR_VSX;
> +			else if (cpu_has_feature(CPU_FTR_ALTIVEC))
> +				mathflags |= MSR_VEC;
> +
> +			/*
> +			 * If userspace MSR has all available FP bits set,
> +			 * then they are live and no need to restore. If not,
> +			 * it means the regs were given up and restore_math
> +			 * may decide to restore them (to avoid taking an FP
> +			 * fault).
> +			 */
> +			if ((regs->msr & mathflags) != mathflags)
> +				restore_math(regs);
> +		}
> +	}
> +}
> +
> +#define arch_exit_to_user_mode_work arch_exit_to_user_mode_work
> +
>   #endif
> diff --git a/arch/powerpc/kernel/interrupt.c b/arch/powerpc/kernel/interrupt.c
> index 609ba48034de..42af9217136d 100644
> --- a/arch/powerpc/kernel/interrupt.c
> +++ b/arch/powerpc/kernel/interrupt.c
> @@ -3,6 +3,7 @@
>   #include <linux/context_tracking.h>
>   #include <linux/err.h>
>   #include <linux/compat.h>
> +#include <linux/entry-common.h>
>   #include <linux/sched/debug.h> /* for show_regs */
>   
>   #include <asm/kup.h>
> @@ -183,47 +184,11 @@ interrupt_exit_user_prepare_main(unsigned long ret, struct pt_regs *regs)
>   
>   again:
>   	ti_flags = read_thread_flags();
> -	while (unlikely(ti_flags & (_TIF_USER_WORK_MASK & ~_TIF_RESTORE_TM))) {
> -		local_irq_enable();
> -		if (ti_flags & _TIF_NEED_RESCHED) {
> -			schedule();
> -		} else {
> -			/*
> -			 * SIGPENDING must restore signal handler function
> -			 * argument GPRs, and some non-volatiles (e.g., r1).
> -			 * Restore all for now. This could be made lighter.
> -			 */
> -			if (ti_flags & _TIF_SIGPENDING)
> -				ret |= _TIF_RESTOREALL;
> -			do_notify_resume(regs, ti_flags);
> -		}
> -		local_irq_disable();
> -		ti_flags = read_thread_flags();
> -	}
>   
> -	if (IS_ENABLED(CONFIG_PPC_BOOK3S_64) && IS_ENABLED(CONFIG_PPC_FPU)) {
> -		if (IS_ENABLED(CONFIG_PPC_TRANSACTIONAL_MEM) &&
> -				unlikely((ti_flags & _TIF_RESTORE_TM))) {
> -			restore_tm_state(regs);
> -		} else {
> -			unsigned long mathflags = MSR_FP;
> -
> -			if (cpu_has_feature(CPU_FTR_VSX))
> -				mathflags |= MSR_VEC | MSR_VSX;
> -			else if (cpu_has_feature(CPU_FTR_ALTIVEC))
> -				mathflags |= MSR_VEC;
> -
> -			/*
> -			 * If userspace MSR has all available FP bits set,
> -			 * then they are live and no need to restore. If not,
> -			 * it means the regs were given up and restore_math
> -			 * may decide to restore them (to avoid taking an FP
> -			 * fault).
> -			 */
> -			if ((regs->msr & mathflags) != mathflags)
> -				restore_math(regs);
> -		}
> -	}
> +	if (ti_flags & _TIF_SIGPENDING)
> +		ret |= _TIF_RESTOREALL;
> +	if (unlikely(ti_flags & EXIT_TO_USER_MODE_WORK))
> +		ti_flags = exit_to_user_mode_loop(regs, ti_flags);
>   
>   	check_return_regs_valid(regs);
>   
> @@ -297,11 +262,15 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
>   	}
>   
>   	local_irq_disable();
> -	ret = interrupt_exit_user_prepare_main(ret, regs);
> +	if (ti_flags & _TIF_RESTOREALL)
> +		ret |= _TIF_RESTOREALL;
>   
> +	if (ti_flags & _TIF_SIGPENDING)
> +		ret |= _TIF_RESTOREALL;
>   #ifdef CONFIG_PPC64
>   	regs->exit_result = ret;
>   #endif
> +	syscall_exit_to_user_mode(regs);
>   
>   	return ret;
>   }
> diff --git a/arch/powerpc/kernel/signal.c b/arch/powerpc/kernel/signal.c
> index aa17e62f3754..da21e7fef46a 100644
> --- a/arch/powerpc/kernel/signal.c
> +++ b/arch/powerpc/kernel/signal.c
> @@ -11,6 +11,7 @@
>   #include <linux/uprobes.h>
>   #include <linux/key.h>
>   #include <linux/context_tracking.h>
> +#include <linux/entry-common.h>
>   #include <linux/livepatch.h>
>   #include <linux/syscalls.h>
>   #include <asm/hw_breakpoint.h>
> @@ -368,3 +369,9 @@ void signal_fault(struct task_struct *tsk, struct pt_regs *regs,
>   		printk_ratelimited(regs->msr & MSR_64BIT ? fm64 : fm32, tsk->comm,
>   				   task_pid_nr(tsk), where, ptr, regs->nip, regs->link);
>   }
> +
> +void arch_do_signal_or_restart(struct pt_regs *regs)
> +{
> +	BUG_ON(regs != current->thread.regs);
> +	do_signal(current);
> +}
> diff --git a/arch/powerpc/kernel/syscall.c b/arch/powerpc/kernel/syscall.c
> index e0338bd8d383..97f158d13944 100644
> --- a/arch/powerpc/kernel/syscall.c
> +++ b/arch/powerpc/kernel/syscall.c
> @@ -185,8 +185,6 @@ notrace long system_call_exception(struct pt_regs *regs, unsigned long r0)
>   	 * So the resulting 6 or 7 bits of entropy is seen in SP[9:4] or SP[9:3].
>   	 */
>   	choose_random_kstack_offset(mftb());
> -	/*common entry*/
> -	syscall_exit_to_user_mode(regs);
>   
>   	return ret;
>   }


More information about the Linuxppc-dev mailing list