powerpc/64: vmlinux support building with PCREL addresing

PC-Relative or PCREL addressing is an extension to the ELF ABI which
uses Power ISA v3.1 PC-relative instructions to calculate addresses,
rather than the traditional TOC scheme.

Add an option to build vmlinux using pcrel addressing. Modules continue
to use TOC addressing.

- TOC address helpers and r2 are poisoned with -1 when running vmlinux.
  r2 could be used for something useful once things are ironed out.

- Assembly must call C functions with @notoc annotation, or the linker
  complains aobut a missing nop after the call. This is done with the
  CFUNC macro introduced earlier.

- Boot: with the exception of prom_init, the execution branches to the
  kernel virtual address early in boot, before any addresses are
  generated, which ensures 34-bit pcrel addressing does not miss the
  high PAGE_OFFSET bits. TOC relative addressing has a similar
  requirement. prom_init does not go to the virtual address and its
  addresses should not carry over to the post-prom kernel.

- Ftrace trampolines are converted from TOC addressing to pcrel
  addressing, including module ftrace trampolines that currently use the
  kernel TOC to find ftrace target functions.

- BPF function prologue and function calling generation are converted
  from TOC to pcrel.

- copypage_64.S has an interesting problem, prefixed instructions have
  alignment restrictions so the linker can add padding, which makes the
  assembler treat the difference between two local labels as
  non-constant even if alignment is arranged so padding is not required.
  This may need toolchain help to solve nicely, for now move the prefix
  instruction out of the alternate patch section to work around it.

This reduces kernel text size by about 6%.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230408021752.862660-6-npiggin@gmail.com

8 files changed, 121 insertions, 25 deletions

diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index d24a59a98c0c..9f14d95b8b32 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -185,7 +185,9 @@ int main(void)
 				 offsetof(struct task_struct, thread_info));
 	OFFSET(PACASAVEDMSR, paca_struct, saved_msr);
 	OFFSET(PACAR1, paca_struct, saved_r1);
+#ifndef CONFIG_PPC_KERNEL_PCREL
 	OFFSET(PACATOC, paca_struct, kernel_toc);
+#endif
 	OFFSET(PACAKBASE, paca_struct, kernelbase);
 	OFFSET(PACAKMSR, paca_struct, kernel_msr);
 #ifdef CONFIG_PPC_BOOK3S_64
diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S
index 6e04e30ae44c..f132d8704263 100644
--- a/arch/powerpc/kernel/head_64.S
+++ b/arch/powerpc/kernel/head_64.S
@@ -330,6 +330,12 @@ _GLOBAL(fsl_secondary_thread_init)
  */
 _GLOBAL(generic_secondary_smp_init)
 	FIXUP_ENDIAN
+
+	li	r13,0
+
+	/* Poison TOC */
+	li	r2,-1
+
 	mr	r24,r3
 	mr	r25,r4
 
@@ -528,6 +534,9 @@ __start_initialization_multiplatform:
 	/* Zero r13 (paca) so early program check / mce don't use it */
 	li	r13,0
 
+	/* Poison TOC */
+	li	r2,-1
+
 	/*
 	 * Are we booted from a PROM Of-type client-interface ?
 	 */
@@ -921,6 +930,10 @@ SYM_FUNC_END(enable_64b_mode)
  * this.
  */
 _GLOBAL(relative_toc)
+#ifdef CONFIG_PPC_KERNEL_PCREL
+	tdnei	r2,-1
+	blr
+#else
 	mflr	r0
 	bcl	20,31,$+4
 0:	mflr	r11
@@ -931,6 +944,7 @@ _GLOBAL(relative_toc)
 
 .balign 8
 p_toc:	.8byte	.TOC. - 0b
+#endif
 
 /*
  * This is where the main kernel code starts.
diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c
index c9535f2760b5..6f7d4edaa0bc 100644
--- a/arch/powerpc/kernel/irq.c
+++ b/arch/powerpc/kernel/irq.c
@@ -206,7 +206,11 @@ static __always_inline void call_do_softirq(const void *sp)
 	asm volatile (
 		 PPC_STLU "	%%r1, %[offset](%[sp])	;"
 		"mr		%%r1, %[sp]		;"
+#ifdef CONFIG_PPC_KERNEL_PCREL
+		"bl		%[callee]@notoc		;"
+#else
 		"bl		%[callee]		;"
+#endif
 		 PPC_LL "	%%r1, 0(%%r1)		;"
 		 : // Outputs
 		 : // Inputs
@@ -259,7 +263,11 @@ static __always_inline void call_do_irq(struct pt_regs *regs, void *sp)
 		 PPC_STLU "	%%r1, %[offset](%[sp])	;"
 		"mr		%%r4, %%r1		;"
 		"mr		%%r1, %[sp]		;"
+#ifdef CONFIG_PPC_KERNEL_PCREL
+		"bl		%[callee]@notoc		;"
+#else
 		"bl		%[callee]		;"
+#endif
 		 PPC_LL "	%%r1, 0(%%r1)		;"
 		 : // Outputs
 		   "+r" (r3)
diff --git a/arch/powerpc/kernel/module_64.c b/arch/powerpc/kernel/module_64.c
index 2ac78d207f77..0fc04abac3bd 100644
--- a/arch/powerpc/kernel/module_64.c
+++ b/arch/powerpc/kernel/module_64.c
@@ -101,17 +101,18 @@ static unsigned long stub_func_addr(func_desc_t func)
 /* Like PPC32, we need little trampolines to do > 24-bit jumps (into
    the kernel itself).  But on PPC64, these need to be used for every
    jump, actually, to reset r2 (TOC+0x8000). */
-struct ppc64_stub_entry
-{
-	/* 28 byte jump instruction sequence (7 instructions). We only
-	 * need 6 instructions on ABIv2 but we always allocate 7 so
-	 * so we don't have to modify the trampoline load instruction. */
+struct ppc64_stub_entry {
+	/*
+	 * 28 byte jump instruction sequence (7 instructions) that can
+	 * hold ppc64_stub_insns or stub_insns. Must be 8-byte aligned
+	 * with PCREL kernels that use prefix instructions in the stub.
+	 */
 	u32 jump[7];
 	/* Used by ftrace to identify stubs */
 	u32 magic;
 	/* Data for the above code */
 	func_desc_t funcdata;
-};
+} __aligned(8);
 
 /*
  * PPC64 uses 24 bit jumps, but we need to jump into other modules or
@@ -333,11 +334,21 @@ int module_frob_arch_sections(Elf64_Ehdr *hdr,
 #ifdef CONFIG_MPROFILE_KERNEL
 
 static u32 stub_insns[] = {
+#ifdef CONFIG_PPC_KERNEL_PCREL
+	PPC_RAW_LD(_R12, _R13, offsetof(struct paca_struct, kernelbase)),
+	PPC_RAW_NOP(), /* align the prefix insn */
+	/* paddi r12,r12,addr */
+	PPC_PREFIX_MLS | __PPC_PRFX_R(0),
+	PPC_INST_PADDI | ___PPC_RT(_R12) | ___PPC_RA(_R12),
+	PPC_RAW_MTCTR(_R12),
+	PPC_RAW_BCTR(),
+#else
 	PPC_RAW_LD(_R12, _R13, offsetof(struct paca_struct, kernel_toc)),
 	PPC_RAW_ADDIS(_R12, _R12, 0),
 	PPC_RAW_ADDI(_R12, _R12, 0),
 	PPC_RAW_MTCTR(_R12),
 	PPC_RAW_BCTR(),
+#endif
 };
 
 /*
@@ -358,18 +369,37 @@ static inline int create_ftrace_stub(struct ppc64_stub_entry *entry,
 {
 	long reladdr;
 
-	memcpy(entry->jump, stub_insns, sizeof(stub_insns));
-
-	/* Stub uses address relative to kernel toc (from the paca) */
-	reladdr = addr - kernel_toc_addr();
-	if (reladdr > 0x7FFFFFFF || reladdr < -(0x80000000L)) {
-		pr_err("%s: Address of %ps out of range of kernel_toc.\n",
-							me->name, (void *)addr);
+	if ((unsigned long)entry->jump % 8 != 0) {
+		pr_err("%s: Address of stub entry is not 8-byte aligned\n", me->name);
 		return 0;
 	}
 
-	entry->jump[1] |= PPC_HA(reladdr);
-	entry->jump[2] |= PPC_LO(reladdr);
+	BUILD_BUG_ON(sizeof(stub_insns) > sizeof(entry->jump));
+	memcpy(entry->jump, stub_insns, sizeof(stub_insns));
+
+	if (IS_ENABLED(CONFIG_PPC_KERNEL_PCREL)) {
+		/* Stub uses address relative to kernel base (from the paca) */
+		reladdr = addr - local_paca->kernelbase;
+		if (reladdr > 0x1FFFFFFFFL || reladdr < -0x200000000L) {
+			pr_err("%s: Address of %ps out of range of 34-bit relative address.\n",
+				me->name, (void *)addr);
+			return 0;
+		}
+
+		entry->jump[2] |= IMM_H18(reladdr);
+		entry->jump[3] |= IMM_L(reladdr);
+	} else {
+		/* Stub uses address relative to kernel toc (from the paca) */
+		reladdr = addr - kernel_toc_addr();
+		if (reladdr > 0x7FFFFFFF || reladdr < -(0x80000000L)) {
+			pr_err("%s: Address of %ps out of range of kernel_toc.\n",
+				me->name, (void *)addr);
+			return 0;
+		}
+
+		entry->jump[1] |= PPC_HA(reladdr);
+		entry->jump[2] |= PPC_LO(reladdr);
+	}
 
 	/* Even though we don't use funcdata in the stub, it's needed elsewhere. */
 	entry->funcdata = func_desc(addr);
diff --git a/arch/powerpc/kernel/paca.c b/arch/powerpc/kernel/paca.c
index be8db402e963..cda4e00b67c1 100644
--- a/arch/powerpc/kernel/paca.c
+++ b/arch/powerpc/kernel/paca.c
@@ -191,7 +191,9 @@ void __init initialise_paca(struct paca_struct *new_paca, int cpu)
 #endif
 	new_paca->lock_token = 0x8000;
 	new_paca->paca_index = cpu;
+#ifndef CONFIG_PPC_KERNEL_PCREL
 	new_paca->kernel_toc = kernel_toc_addr();
+#endif
 	new_paca->kernelbase = (unsigned long) _stext;
 	/* Only set MSR:IR/DR when MMU is initialized */
 	new_paca->kernel_msr = MSR_KERNEL & ~(MSR_IR | MSR_DR);
diff --git a/arch/powerpc/kernel/trace/ftrace.c b/arch/powerpc/kernel/trace/ftrace.c
index 72864fb7a6cc..a47f30373423 100644
--- a/arch/powerpc/kernel/trace/ftrace.c
+++ b/arch/powerpc/kernel/trace/ftrace.c
@@ -727,6 +727,15 @@ int __init ftrace_dyn_arch_init(void)
 {
 	int i;
 	unsigned int *tramp[] = { ftrace_tramp_text, ftrace_tramp_init };
+#ifdef CONFIG_PPC_KERNEL_PCREL
+	u32 stub_insns[] = {
+		/* pla r12,addr */
+		PPC_PREFIX_MLS | __PPC_PRFX_R(1),
+		PPC_INST_PADDI | ___PPC_RT(_R12),
+		PPC_RAW_MTCTR(_R12),
+		PPC_RAW_BCTR()
+	};
+#else
 	u32 stub_insns[] = {
 		PPC_RAW_LD(_R12, _R13, PACATOC),
 		PPC_RAW_ADDIS(_R12, _R12, 0),
@@ -734,6 +743,8 @@ int __init ftrace_dyn_arch_init(void)
 		PPC_RAW_MTCTR(_R12),
 		PPC_RAW_BCTR()
 	};
+#endif
+
 	unsigned long addr;
 	long reladdr;
 
@@ -742,19 +753,36 @@ int __init ftrace_dyn_arch_init(void)
 	else
 		addr = ppc_global_function_entry((void *)ftrace_caller);
 
-	reladdr = addr - kernel_toc_addr();
+	if (IS_ENABLED(CONFIG_PPC_KERNEL_PCREL)) {
+		for (i = 0; i < 2; i++) {
+			reladdr = addr - (unsigned long)tramp[i];
 
-	if (reladdr >= SZ_2G || reladdr < -(long)SZ_2G) {
-		pr_err("Address of %ps out of range of kernel_toc.\n",
+			if (reladdr >= (long)SZ_8G || reladdr < -(long)SZ_8G) {
+				pr_err("Address of %ps out of range of pcrel address.\n",
+					(void *)addr);
+				return -1;
+			}
+
+			memcpy(tramp[i], stub_insns, sizeof(stub_insns));
+			tramp[i][0] |= IMM_H18(reladdr);
+			tramp[i][1] |= IMM_L(reladdr);
+			add_ftrace_tramp((unsigned long)tramp[i]);
+		}
+	} else {
+		reladdr = addr - kernel_toc_addr();
+
+		if (reladdr >= (long)SZ_2G || reladdr < -(long)SZ_2G) {
+			pr_err("Address of %ps out of range of kernel_toc.\n",
 				(void *)addr);
-		return -1;
-	}
+			return -1;
+		}
 
-	for (i = 0; i < 2; i++) {
-		memcpy(tramp[i], stub_insns, sizeof(stub_insns));
-		tramp[i][1] |= PPC_HA(reladdr);
-		tramp[i][2] |= PPC_LO(reladdr);
-		add_ftrace_tramp((unsigned long)tramp[i]);
+		for (i = 0; i < 2; i++) {
+			memcpy(tramp[i], stub_insns, sizeof(stub_insns));
+			tramp[i][1] |= PPC_HA(reladdr);
+			tramp[i][2] |= PPC_LO(reladdr);
+			add_ftrace_tramp((unsigned long)tramp[i]);
+		}
 	}
 
 	return 0;
diff --git a/arch/powerpc/kernel/vector.S b/arch/powerpc/kernel/vector.S
index ffe5d90abe17..fcc0ad6d9c7b 100644
--- a/arch/powerpc/kernel/vector.S
+++ b/arch/powerpc/kernel/vector.S
@@ -177,10 +177,16 @@ fpone:
 fphalf:
 	.quad	0x3fe0000000000000	/* 0.5 */
 
+#ifdef CONFIG_PPC_KERNEL_PCREL
+#define LDCONST(fr, name)		\
+	pla	r11,name@pcrel;		\
+	lfd	fr,0(r11)
+#else
 #define LDCONST(fr, name)		\
 	addis	r11,r2,name@toc@ha;	\
 	lfd	fr,name@toc@l(r11)
 #endif
+#endif
 	.text
 /*
  * Internal routine to enable floating point and set FPSCR to 0.
diff --git a/arch/powerpc/kernel/vmlinux.lds.S b/arch/powerpc/kernel/vmlinux.lds.S
index ee86753e444e..13614f0b269c 100644
--- a/arch/powerpc/kernel/vmlinux.lds.S
+++ b/arch/powerpc/kernel/vmlinux.lds.S
@@ -169,12 +169,18 @@ SECTIONS
 	}
 
 #else /* CONFIG_PPC32 */
+#ifndef CONFIG_PPC_KERNEL_PCREL
 	.toc1 : AT(ADDR(.toc1) - LOAD_OFFSET) {
 		*(.toc1)
 	}
+#endif
 
 	.got : AT(ADDR(.got) - LOAD_OFFSET) ALIGN(256) {
+#ifdef CONFIG_PPC_KERNEL_PCREL
+		*(.got)
+#else
 		*(.got .toc)
+#endif
 	}
 
 	SOFT_MASK_TABLE(8)