[PATCH v2] powerpc/32: Add support for out-of-line static calls

[Peter Zijlstra]

On Tue, Aug 31, 2021 at 01:12:26PM +0000, Christophe Leroy wrote:
> diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
> index 36b72d972568..a0fe69d8ec83 100644
> --- a/arch/powerpc/Kconfig
> +++ b/arch/powerpc/Kconfig
> @@ -247,6 +247,7 @@ config PPC
>  	select HAVE_SOFTIRQ_ON_OWN_STACK
>  	select HAVE_STACKPROTECTOR		if PPC32 && $(cc-option,-mstack-protector-guard=tls -mstack-protector-guard-reg=r2)
>  	select HAVE_STACKPROTECTOR		if PPC64 && $(cc-option,-mstack-protector-guard=tls -mstack-protector-guard-reg=r13)
> +	select HAVE_STATIC_CALL			if PPC32
>  	select HAVE_SYSCALL_TRACEPOINTS
>  	select HAVE_VIRT_CPU_ACCOUNTING
>  	select HUGETLB_PAGE_SIZE_VARIABLE	if PPC_BOOK3S_64 && HUGETLB_PAGE
> diff --git a/arch/powerpc/include/asm/static_call.h b/arch/powerpc/include/asm/static_call.h
> new file mode 100644
> index 000000000000..2402c6d32439
> --- /dev/null
> +++ b/arch/powerpc/include/asm/static_call.h
> @@ -0,0 +1,25 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +#ifndef _ASM_POWERPC_STATIC_CALL_H
> +#define _ASM_POWERPC_STATIC_CALL_H
> +
> +#define __POWERPC_SCT(name, inst)					\
> +	asm(".pushsection .text, \"ax\"				\n"	\
> +	    ".align 5						\n"	\
> +	    ".globl " STATIC_CALL_TRAMP_STR(name) "		\n"	\
> +	    STATIC_CALL_TRAMP_STR(name) ":			\n"	\
> +	    inst "						\n"	\
> +	    "	lis	12,1f at ha				\n"	\
> +	    "	lwz	12,1f at l(12)				\n"	\
> +	    "	mtctr	12					\n"	\
> +	    "	bctr						\n"	\
> +	    "1:	.long 0						\n"	\
> +	    "	nop						\n"	\
> +	    "	nop						\n"	\
> +	    ".type " STATIC_CALL_TRAMP_STR(name) ", @function	\n"	\
> +	    ".size " STATIC_CALL_TRAMP_STR(name) ", . - " STATIC_CALL_TRAMP_STR(name) " \n" \
> +	    ".popsection					\n")
> +
> +#define ARCH_DEFINE_STATIC_CALL_TRAMP(name, func)	__POWERPC_SCT(name, "b " #func)
> +#define ARCH_DEFINE_STATIC_CALL_NULL_TRAMP(name)	__POWERPC_SCT(name, "blr")
> +
> +#endif /* _ASM_POWERPC_STATIC_CALL_H */

> diff --git a/arch/powerpc/kernel/static_call.c b/arch/powerpc/kernel/static_call.c
> new file mode 100644
> index 000000000000..e5e78205ccb4
> --- /dev/null
> +++ b/arch/powerpc/kernel/static_call.c
> @@ -0,0 +1,36 @@
> +// SPDX-License-Identifier: GPL-2.0
> +#include <linux/memory.h>
> +#include <linux/static_call.h>
> +
> +#include <asm/code-patching.h>
> +
> +void arch_static_call_transform(void *site, void *tramp, void *func, bool tail)
> +{
> +	int err;
> +	unsigned long target = (long)func;
> +	bool is_short = is_offset_in_branch_range((long)target - (long)tramp);
> +
> +	if (!tramp)
> +		return;
> +
> +	mutex_lock(&text_mutex);
> +
> +	if (func && !is_short) {
> +		err = patch_instruction(tramp + 20, ppc_inst(target));
> +		if (err)
> +			goto out;
> +	}
> +
> +	if (!func)
> +		err = patch_instruction(tramp, ppc_inst(PPC_RAW_BLR()));
> +	else if (is_short)
> +		err = patch_branch(tramp, target, 0);
> +	else
> +		err = patch_instruction(tramp, ppc_inst(PPC_RAW_NOP()));
> +out:
> +	mutex_unlock(&text_mutex);
> +
> +	if (err)
> +		panic("%s: patching failed %pS at %pS\n", __func__, func, tramp);
> +}
> +EXPORT_SYMBOL_GPL(arch_static_call_transform);

Yes, this should work nicely!

Since you have the two nop's at the end, you could frob in an
optimization for __static_call_return0 without too much issue.

Replace the two nops with (excuse my ppc asm):

	li r3, 0
	blr

and augment arch_static_call_transform() with something like:

	if (func == &__static_call_return0)
		err = patch_branch(tramp, tramp+24, 0);