Merge tag 'mips_5.2' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux

Pull MIPS updates from Paul Burton:

- A set of memblock initialization improvements thanks to Serge Semin,
tidying up after our conversion from bootmem to memblock back in
v4.20.

- Our eBPF JIT the previously supported only MIPS64r2 through MIPS64r5
is improved to also support MIPS64r6. Support for MIPS32 systems is
introduced, with the caveat that it only works for programs that
don't use 64 bit registers or operations - those will bail out & need
to be interpreted.

- Improvements to the allocation & configuration of our exception
vector that should fix issues seen on some platforms using recent
versions of U-Boot.

- Some minor improvements to code generated for jump labels, along with
enabling them by default for generic kernels.

* tag 'mips_5.2' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux: (27 commits)
mips: Manually call fdt_init_reserved_mem() method
mips: Make sure dt memory regions are valid
mips: Perform early low memory test
mips: Dump memblock regions for debugging
mips: Add reserve-nomap memory type support
mips: Use memblock to reserve the __nosave memory range
mips: Discard post-CMA-init foreach loop
mips: Reserve memory for the kernel image resources
MIPS: Remove duplicate EBase configuration
MIPS: Sync icache for whole exception vector
MIPS: Always allocate exception vector for MIPSr2+
MIPS: Use memblock_phys_alloc() for exception vector
mips: Combine memblock init and memory reservation loops
mips: Discard rudiments from bootmem_init
mips: Make sure kernel .bss exists in boot mem pool
mips: vdso: drop unnecessary cc-ldoption
Revert "MIPS: ralink: fix cpu clock of mt7621 and add dt clk devices"
MIPS: generic: Enable CONFIG_JUMP_LABEL
MIPS: jump_label: Use compact branches for >= r6
MIPS: jump_label: Remove redundant nops
...

arch/mips/Kconfig

···
 	select HAVE_ARCH_SECCOMP_FILTER
 	select HAVE_ARCH_TRACEHOOK
 	select HAVE_ARCH_TRANSPARENT_HUGEPAGE if CPU_SUPPORTS_HUGEPAGES && 64BIT
-	select HAVE_CBPF_JIT if (!64BIT && !CPU_MICROMIPS)
-	select HAVE_EBPF_JIT if (64BIT && !CPU_MICROMIPS)
+	select HAVE_EBPF_JIT if (!CPU_MICROMIPS)
 	select HAVE_CONTEXT_TRACKING
 	select HAVE_COPY_THREAD_TLS
 	select HAVE_C_RECORDMCOUNT
···
 	select BCM47XX_SPROM
 	select BCM47XX_SSB if !BCM47XX_BCMA
 	help
-	 Support for BCM47XX based boards
+	  Support for BCM47XX based boards
 
 config BCM63XX
 	bool "Broadcom BCM63XX based boards"
···
 	select MIPS_L1_CACHE_SHIFT_4
 	select CLKDEV_LOOKUP
 	help
-	 Support for BCM63XX based boards
+	  Support for BCM63XX based boards
 
 config MIPS_COBALT
 	bool "Cobalt Server"
···
 	select SYS_SUPPORTS_64BIT_KERNEL
 	select SYS_SUPPORTS_100HZ
 	help
-	 This a family of machines based on the MIPS R4030 chipset which was
-	 used by several vendors to build RISC/os and Windows NT workstations.
-	 Members include the Acer PICA, MIPS Magnum 4000, MIPS Millennium and
-	 Olivetti M700-10 workstations.
+	  This a family of machines based on the MIPS R4030 chipset which was
+	  used by several vendors to build RISC/os and Windows NT workstations.
+	  Members include the Acer PICA, MIPS Magnum 4000, MIPS Millennium and
+	  Olivetti M700-10 workstations.
 
 config MACH_INGENIC
 	bool "Ingenic SoC based machines"
···
 	bool "NXP STB220 board"
 	select SOC_PNX833X
 	help
-	 Support for NXP Semiconductors STB220 Development Board.
+	  Support for NXP Semiconductors STB220 Development Board.
 
 config NXP_STB225
 	bool "NXP 225 board"
 	select SOC_PNX833X
 	select SOC_PNX8335
 	help
-	 Support for NXP Semiconductors STB225 Development Board.
+	  Support for NXP Semiconductors STB225 Development Board.
 
 config PMC_MSP
 	bool "PMC-Sierra MSP chipsets"
···
 	select SYS_SUPPORTS_64BIT_KERNEL
 	select SYS_SUPPORTS_BIG_ENDIAN
 	select MIPS_L1_CACHE_SHIFT_7
-      help
-        This is the SGI Indigo2 with R10000 processor.  To compile a Linux
-        kernel that runs on these, say Y here.
+	help
+	  This is the SGI Indigo2 with R10000 processor.  To compile a Linux
+	  kernel that runs on these, say Y here.
 
 config SGI_IP32
 	bool "SGI IP32 (O2)"
···
 config SYS_SUPPORTS_RELOCATABLE
 	bool
 	help
-	 Selected if the platform supports relocating the kernel.
-	 The platform must provide plat_get_fdt() if it selects CONFIG_USE_OF
-	 to allow access to command line and entropy sources.
+	  Selected if the platform supports relocating the kernel.
+	  The platform must provide plat_get_fdt() if it selects CONFIG_USE_OF
+	  to allow access to command line and entropy sources.
 
 config MIPS_CBPF_JIT
 	def_bool y
···
 # Set to y for ptrace access to watch registers.
 #
 config HARDWARE_WATCHPOINTS
-       bool
-       default y if CPU_MIPSR1 || CPU_MIPSR2 || CPU_MIPSR6
+	bool
+	default y if CPU_MIPSR1 || CPU_MIPSR2 || CPU_MIPSR6
 
 menu "Kernel type"
 
···
 	bool "4kB"
 	depends on !CPU_LOONGSON2 && !CPU_LOONGSON3
 	help
-	 This option select the standard 4kB Linux page size.  On some
-	 R3000-family processors this is the only available page size.  Using
-	 4kB page size will minimize memory consumption and is therefore
-	 recommended for low memory systems.
+	  This option select the standard 4kB Linux page size.  On some
+	  R3000-family processors this is the only available page size.  Using
+	  4kB page size will minimize memory consumption and is therefore
+	  recommended for low memory systems.
 
 config PAGE_SIZE_8KB
 	bool "8kB"
···
 	depends on CPU_SB1 && CPU_SB1_PASS_2
 	default y
 
-
 choice
 	prompt "SmartMIPS or microMIPS ASE support"
 
···
 	bool "Randomize the address of the kernel image"
 	depends on RELOCATABLE
 	---help---
-	   Randomizes the physical and virtual address at which the
-	   kernel image is loaded, as a security feature that
-	   deters exploit attempts relying on knowledge of the location
-	   of kernel internals.
+	  Randomizes the physical and virtual address at which the
+	  kernel image is loaded, as a security feature that
+	  deters exploit attempts relying on knowledge of the location
+	  of kernel internals.
 
-	   Entropy is generated using any coprocessor 0 registers available.
+	  Entropy is generated using any coprocessor 0 registers available.
 
-	   The kernel will be offset by up to RANDOMIZE_BASE_MAX_OFFSET.
+	  The kernel will be offset by up to RANDOMIZE_BASE_MAX_OFFSET.
 
-	   If unsure, say N.
+	  If unsure, say N.
 
 config RANDOMIZE_BASE_MAX_OFFSET
 	hex "Maximum kASLR offset" if EXPERT
···
 	prompt "Timer frequency"
 	default HZ_250
 	help
-	 Allows the configuration of the timer frequency.
+	  Allows the configuration of the timer frequency.
 
 	config HZ_24
 		bool "24 HZ" if SYS_SUPPORTS_24HZ || SYS_SUPPORTS_ARBIT_HZ
···
 	default 15
 
 config ARCH_MMAP_RND_COMPAT_BITS_MIN
-       default 8
+	default 8
 
 config ARCH_MMAP_RND_COMPAT_BITS_MAX
-       default 15
+	default 15
 
 config I8253
 	bool

arch/mips/bcm47xx/Kconfig

···
 	select SSB_DRIVER_GPIO
 	default y
 	help
-	 Add support for old Broadcom BCM47xx boards with Sonics Silicon Backplane support.
+	  Add support for old Broadcom BCM47xx boards with Sonics Silicon Backplane support.
 
-	 This will generate an image with support for SSB and MIPS32 R1 instruction set.
+	  This will generate an image with support for SSB and MIPS32 R1 instruction set.
 
 config BCM47XX_BCMA
 	bool "BCMA Support for Broadcom BCM47XX"
···
 	select BCMA_DRIVER_GPIO
 	default y
 	help
-	 Add support for new Broadcom BCM47xx boards with Broadcom specific Advanced Microcontroller Bus.
+	  Add support for new Broadcom BCM47xx boards with Broadcom specific Advanced Microcontroller Bus.
 
-	 This will generate an image with support for BCMA and MIPS32 R2 instruction set.
+	  This will generate an image with support for BCMA and MIPS32 R2 instruction set.
 
 endif

arch/mips/bcm63xx/boards/Kconfig

···
 	default BOARD_BCM963XX
 
 config BOARD_BCM963XX
-       bool "Generic Broadcom 963xx boards"
+	bool "Generic Broadcom 963xx boards"
 	select SSB
 
 endchoice

arch/mips/configs/generic_defconfig

···
 CONFIG_HIGHMEM=y
 CONFIG_NR_CPUS=16
 CONFIG_MIPS_O32_FP64_SUPPORT=y
+CONFIG_JUMP_LABEL=y
 CONFIG_MODULES=y
 CONFIG_MODULE_UNLOAD=y
 CONFIG_TRIM_UNUSED_KSYMS=y

arch/mips/include/asm/bootinfo.h

···
 #define BOOT_MEM_ROM_DATA	2
 #define BOOT_MEM_RESERVED	3
 #define BOOT_MEM_INIT_RAM	4
+#define BOOT_MEM_NOMAP		5
 
 /*
  * A memory map that's built upon what was determined

arch/mips/include/asm/jump_label.h

···
 #ifndef __ASSEMBLY__
 
 #include <linux/types.h>
+#include <asm/isa-rev.h>
 
 #define JUMP_LABEL_NOP_SIZE 4
 
···
 #endif
 
 #ifdef CONFIG_CPU_MICROMIPS
-#define B_INSN "b32"
+# define B_INSN "b32"
+# define J_INSN "j32"
+#elif MIPS_ISA_REV >= 6
+# define B_INSN "bc"
+# define J_INSN "bc"
 #else
-#define B_INSN "b"
+# define B_INSN "b"
+# define J_INSN "j"
 #endif
 
 static __always_inline bool arch_static_branch(struct static_key *key, bool branch)
 {
 	asm_volatile_goto("1:\t" B_INSN " 2f\n\t"
-		"2:\tnop\n\t"
+		"2:\t.insn\n\t"
 		".pushsection __jump_table,  \"aw\"\n\t"
 		WORD_INSN " 1b, %l[l_yes], %0\n\t"
 		".popsection\n\t"
···
 
 static __always_inline bool arch_static_branch_jump(struct static_key *key, bool branch)
 {
-	asm_volatile_goto("1:\tj %l[l_yes]\n\t"
-		"nop\n\t"
+	asm_volatile_goto("1:\t" J_INSN " %l[l_yes]\n\t"
 		".pushsection __jump_table,  \"aw\"\n\t"
 		WORD_INSN " 1b, %l[l_yes], %0\n\t"
 		".popsection\n\t"

arch/mips/include/asm/uasm.h

···
 Ip_u2u1s3(_daddiu);
 Ip_u3u1u2(_daddu);
 Ip_u1u2(_ddivu);
+Ip_u3u1u2(_ddivu_r6);
 Ip_u1(_di);
 Ip_u2u1msbu3(_dins);
 Ip_u2u1msbu3(_dinsm);
 Ip_u2u1msbu3(_dinsu);
 Ip_u1u2(_divu);
+Ip_u3u1u2(_divu_r6);
 Ip_u1u2u3(_dmfc0);
+Ip_u3u1u2(_dmodu);
 Ip_u1u2u3(_dmtc0);
 Ip_u1u2(_dmultu);
+Ip_u3u1u2(_dmulu);
 Ip_u2u1u3(_drotr);
 Ip_u2u1u3(_drotr32);
 Ip_u2u1(_dsbh);
···
 Ip_u1u2u3(_mfhc0);
 Ip_u1(_mfhi);
 Ip_u1(_mflo);
+Ip_u3u1u2(_modu);
 Ip_u3u1u2(_movn);
 Ip_u3u1u2(_movz);
 Ip_u1u2u3(_mtc0);
···
 Ip_u1(_mtlo);
 Ip_u3u1u2(_mul);
 Ip_u1u2(_multu);
+Ip_u3u1u2(_mulu);
 Ip_u3u1u2(_nor);
 Ip_u3u1u2(_or);
 Ip_u2u1u3(_ori);
···
 Ip_u2s3u1(_sc);
 Ip_u2s3u1(_scd);
 Ip_u2s3u1(_sd);
+Ip_u3u1u2(_seleqz);
+Ip_u3u1u2(_selnez);
 Ip_u2s3u1(_sh);
 Ip_u2u1u3(_sll);
 Ip_u3u2u1(_sllv);

arch/mips/include/uapi/asm/inst.h

···
 	spec3_unused_op, spec4_unused_op, slt_op, sltu_op,
 	dadd_op, daddu_op, dsub_op, dsubu_op,
 	tge_op, tgeu_op, tlt_op, tltu_op,
-	teq_op, spec5_unused_op, tne_op, spec6_unused_op,
-	dsll_op, spec7_unused_op, dsrl_op, dsra_op,
-	dsll32_op, spec8_unused_op, dsrl32_op, dsra32_op
+	teq_op, seleqz_op, tne_op, selnez_op,
+	dsll_op, spec5_unused_op, dsrl_op, dsra_op,
+	dsll32_op, spec6_unused_op, dsrl32_op, dsra32_op
 };
 
 /*

arch/mips/kernel/entry.S

···
 	local_irq_disable
 	lw	t0, TI_PRE_COUNT($28)
 	bnez	t0, restore_all
-need_resched:
 	LONG_L	t0, TI_FLAGS($28)
 	andi	t1, t0, _TIF_NEED_RESCHED
 	beqz	t1, restore_all
 	LONG_L	t0, PT_STATUS(sp)		# Interrupts off?
 	andi	t0, 1
 	beqz	t0, restore_all
-	jal	preempt_schedule_irq
-	b	need_resched
+	PTR_LA	ra, restore_all
+	j	preempt_schedule_irq
 #endif
 
 FEXPORT(ret_from_kernel_thread)

arch/mips/kernel/jump_label.c

···
 {
 	union mips_instruction *insn_p;
 	union mips_instruction insn;
+	long offset;
 
 	insn_p = (union mips_instruction *)msk_isa16_mode(e->code);
-
-	/* Jump only works within an aligned region its delay slot is in. */
-	BUG_ON((e->target & ~J_RANGE_MASK) != ((e->code + 4) & ~J_RANGE_MASK));
 
 	/* Target must have the right alignment and ISA must be preserved. */
 	BUG_ON((e->target & J_ALIGN_MASK) != J_ISA_BIT);
 
 	if (type == JUMP_LABEL_JMP) {
-		insn.j_format.opcode = J_ISA_BIT ? mm_j32_op : j_op;
-		insn.j_format.target = e->target >> J_RANGE_SHIFT;
+		if (!IS_ENABLED(CONFIG_CPU_MICROMIPS) && MIPS_ISA_REV >= 6) {
+			offset = e->target - ((unsigned long)insn_p + 4);
+			offset >>= 2;
+
+			/*
+			 * The branch offset must fit in the instruction's 26
+			 * bit field.
+			 */
+			WARN_ON((offset >= BIT(25)) ||
+				(offset < -(long)BIT(25)));
+
+			insn.j_format.opcode = bc6_op;
+			insn.j_format.target = offset;
+		} else {
+			/*
+			 * Jump only works within an aligned region its delay
+			 * slot is in.
+			 */
+			WARN_ON((e->target & ~J_RANGE_MASK) !=
+				((e->code + 4) & ~J_RANGE_MASK));
+
+			insn.j_format.opcode = J_ISA_BIT ? mm_j32_op : j_op;
+			insn.j_format.target = e->target >> J_RANGE_SHIFT;
+		}
 	} else {
 		insn.word = 0; /* nop */
 	}

arch/mips/kernel/prom.c

···
 #ifdef CONFIG_USE_OF
 void __init early_init_dt_add_memory_arch(u64 base, u64 size)
 {
-	return add_memory_region(base, size, BOOT_MEM_RAM);
+	if (base >= PHYS_ADDR_MAX) {
+		pr_warn("Trying to add an invalid memory region, skipped\n");
+		return;
+	}
+
+	/* Truncate the passed memory region instead of type casting */
+	if (base + size - 1 >= PHYS_ADDR_MAX || base + size < base) {
+		pr_warn("Truncate memory region %llx @ %llx to size %llx\n",
+			size, base, PHYS_ADDR_MAX - base);
+		size = PHYS_ADDR_MAX - base;
+	}
+
+	add_memory_region(base, size, BOOT_MEM_RAM);
 }
 
 int __init early_init_dt_reserve_memory_arch(phys_addr_t base,
 					phys_addr_t size, bool nomap)
 {
-	add_memory_region(base, size, BOOT_MEM_RESERVED);
+	add_memory_region(base, size,
+			  nomap ? BOOT_MEM_NOMAP : BOOT_MEM_RESERVED);
+
 	return 0;
 }
 

arch/mips/kernel/setup.c

···
 #include <linux/dma-contiguous.h>
 #include <linux/decompress/generic.h>
 #include <linux/of_fdt.h>
+#include <linux/of_reserved_mem.h>
 
 #include <asm/addrspace.h>
 #include <asm/bootinfo.h>
···
 				in_ram = true;
 			break;
 		case BOOT_MEM_RESERVED:
+		case BOOT_MEM_NOMAP:
 			if ((start >= start_ && start < end_) ||
 			    (start < start_ && start + size >= start_))
 				free = false;
···
 			break;
 		case BOOT_MEM_RESERVED:
 			printk(KERN_CONT "(reserved)\n");
+			break;
+		case BOOT_MEM_NOMAP:
+			printk(KERN_CONT "(nomap)\n");
 			break;
 		default:
 			printk(KERN_CONT "type %lu\n", boot_mem_map.map[i].type);
···
 
 static void __init bootmem_init(void)
 {
-	unsigned long reserved_end;
 	phys_addr_t ramstart = PHYS_ADDR_MAX;
 	int i;
 
···
 	 * will reserve the area used for the initrd.
 	 */
 	init_initrd();
-	reserved_end = (unsigned long) PFN_UP(__pa_symbol(&_end));
 
-	memblock_reserve(PHYS_OFFSET,
-			 (reserved_end << PAGE_SHIFT) - PHYS_OFFSET);
+	/* Reserve memory occupied by kernel. */
+	memblock_reserve(__pa_symbol(&_text),
+			__pa_symbol(&_end) - __pa_symbol(&_text));
 
 	/*
 	 * max_low_pfn is not a number of pages. The number of pages
···
 	min_low_pfn = ~0UL;
 	max_low_pfn = 0;
 
-	/*
-	 * Find the highest page frame number we have available
-	 * and the lowest used RAM address
-	 */
+	/* Find the highest and lowest page frame numbers we have available. */
 	for (i = 0; i < boot_mem_map.nr_map; i++) {
 		unsigned long start, end;
 
···
 			max_low_pfn = end;
 		if (start < min_low_pfn)
 			min_low_pfn = start;
-		if (end <= reserved_end)
-			continue;
-#ifdef CONFIG_BLK_DEV_INITRD
-		/* Skip zones before initrd and initrd itself */
-		if (initrd_end && end <= (unsigned long)PFN_UP(__pa(initrd_end)))
-			continue;
-#endif
 	}
 
 	if (min_low_pfn >= max_low_pfn)
···
 		max_low_pfn = PFN_DOWN(HIGHMEM_START);
 	}
 
+	/* Install all valid RAM ranges to the memblock memory region */
 	for (i = 0; i < boot_mem_map.nr_map; i++) {
 		unsigned long start, end;
 
···
 		end = PFN_DOWN(boot_mem_map.map[i].addr
 				+ boot_mem_map.map[i].size);
 
-		if (start <= min_low_pfn)
+		if (start < min_low_pfn)
 			start = min_low_pfn;
-		if (start >= end)
-			continue;
-
 #ifndef CONFIG_HIGHMEM
+		/* Ignore highmem regions if highmem is unsupported */
 		if (end > max_low_pfn)
 			end = max_low_pfn;
-
-		/*
-		 * ... finally, is the area going away?
-		 */
+#endif
 		if (end <= start)
 			continue;
-#endif
 
 		memblock_add_node(PFN_PHYS(start), PFN_PHYS(end - start), 0);
-	}
 
-	/*
-	 * Register fully available low RAM pages with the bootmem allocator.
-	 */
-	for (i = 0; i < boot_mem_map.nr_map; i++) {
-		unsigned long start, end, size;
-
-		start = PFN_UP(boot_mem_map.map[i].addr);
-		end   = PFN_DOWN(boot_mem_map.map[i].addr
-				    + boot_mem_map.map[i].size);
-
-		/*
-		 * Reserve usable memory.
-		 */
+		/* Reserve any memory except the ordinary RAM ranges. */
 		switch (boot_mem_map.map[i].type) {
 		case BOOT_MEM_RAM:
 			break;
-		case BOOT_MEM_INIT_RAM:
-			memory_present(0, start, end);
+		case BOOT_MEM_NOMAP: /* Discard the range from the system. */
+			memblock_remove(PFN_PHYS(start), PFN_PHYS(end - start));
 			continue;
-		default:
-			/* Not usable memory */
-			if (start > min_low_pfn && end < max_low_pfn)
-				memblock_reserve(boot_mem_map.map[i].addr,
-						boot_mem_map.map[i].size);
-
-			continue;
+		default: /* Reserve the rest of the memory types at boot time */
+			memblock_reserve(PFN_PHYS(start), PFN_PHYS(end - start));
+			break;
 		}
 
 		/*
-		 * We are rounding up the start address of usable memory
-		 * and at the end of the usable range downwards.
+		 * In any case the added to the memblock memory regions
+		 * (highmem/lowmem, available/reserved, etc) are considered
+		 * as present, so inform sparsemem about them.
 		 */
-		if (start >= max_low_pfn)
-			continue;
-		if (start < reserved_end)
-			start = reserved_end;
-		if (end > max_low_pfn)
-			end = max_low_pfn;
-
-		/*
-		 * ... finally, is the area going away?
-		 */
-		if (end <= start)
-			continue;
-		size = end - start;
-
-		/* Register lowmem ranges */
 		memory_present(0, start, end);
 	}
-
-#ifdef CONFIG_RELOCATABLE
-	/*
-	 * The kernel reserves all memory below its _end symbol as bootmem,
-	 * but the kernel may now be at a much higher address. The memory
-	 * between the original and new locations may be returned to the system.
-	 */
-	if (__pa_symbol(_text) > __pa_symbol(VMLINUX_LOAD_ADDRESS)) {
-		unsigned long offset;
-		extern void show_kernel_relocation(const char *level);
-
-		offset = __pa_symbol(_text) - __pa_symbol(VMLINUX_LOAD_ADDRESS);
-		memblock_free(__pa_symbol(VMLINUX_LOAD_ADDRESS), offset);
-
-#if defined(CONFIG_DEBUG_KERNEL) && defined(CONFIG_DEBUG_INFO)
-		/*
-		 * This information is necessary when debugging the kernel
-		 * But is a security vulnerability otherwise!
-		 */
-		show_kernel_relocation(KERN_INFO);
-#endif
-	}
-#endif
 
 	/*
 	 * Reserve initrd memory if needed.
···
  */
 static void __init arch_mem_init(char **cmdline_p)
 {
-	struct memblock_region *reg;
 	extern void plat_mem_setup(void);
 
 	/*
···
 	arch_mem_addpart(PFN_UP(__pa_symbol(&__init_begin)) << PAGE_SHIFT,
 			 PFN_DOWN(__pa_symbol(&__init_end)) << PAGE_SHIFT,
 			 BOOT_MEM_INIT_RAM);
+	arch_mem_addpart(PFN_DOWN(__pa_symbol(&__bss_start)) << PAGE_SHIFT,
+			 PFN_UP(__pa_symbol(&__bss_stop)) << PAGE_SHIFT,
+			 BOOT_MEM_RAM);
 
 	pr_info("Determined physical RAM map:\n");
 	print_memory_map();
···
 	plat_swiotlb_setup();
 
 	dma_contiguous_reserve(PFN_PHYS(max_low_pfn));
-	/* Tell bootmem about cma reserved memblock section */
-	for_each_memblock(reserved, reg)
-		if (reg->size != 0)
-			memblock_reserve(reg->base, reg->size);
 
-	reserve_bootmem_region(__pa_symbol(&__nosave_begin),
-			__pa_symbol(&__nosave_end)); /* Reserve for hibernation */
+	/* Reserve for hibernation. */
+	memblock_reserve(__pa_symbol(&__nosave_begin),
+		__pa_symbol(&__nosave_end) - __pa_symbol(&__nosave_begin));
+
+	fdt_init_reserved_mem();
+
+	memblock_dump_all();
+
+	early_memtest(PFN_PHYS(min_low_pfn), PFN_PHYS(max_low_pfn));
 }
 
 static void __init resource_init(void)
···
 			res->flags |= IORESOURCE_SYSRAM;
 			break;
 		case BOOT_MEM_RESERVED:
+		case BOOT_MEM_NOMAP:
 		default:
 			res->name = "reserved";
 		}

arch/mips/kernel/traps.c

···
 
 static void configure_exception_vector(void)
 {
-	if (cpu_has_veic || cpu_has_vint) {
+	if (cpu_has_mips_r2_r6) {
 		unsigned long sr = set_c0_status(ST0_BEV);
 		/* If available, use WG to set top bits of EBASE */
 		if (cpu_has_ebase_wg) {
···
 		}
 		write_c0_ebase(ebase);
 		write_c0_status(sr);
+	}
+	if (cpu_has_veic || cpu_has_vint) {
 		/* Setting vector spacing enables EI/VI mode  */
 		change_c0_intctl(0x3e0, VECTORSPACING);
 	}
···
 	 *  o read IntCtl.IPFDC to determine the fast debug channel interrupt
 	 */
 	if (cpu_has_mips_r2_r6) {
-		/*
-		 * We shouldn't trust a secondary core has a sane EBASE register
-		 * so use the one calculated by the boot CPU.
-		 */
-		if (!is_boot_cpu) {
-			/* If available, use WG to set top bits of EBASE */
-			if (cpu_has_ebase_wg) {
-#ifdef CONFIG_64BIT
-				write_c0_ebase_64(ebase | MIPS_EBASE_WG);
-#else
-				write_c0_ebase(ebase | MIPS_EBASE_WG);
-#endif
-			}
-			write_c0_ebase(ebase);
-		}
-
 		cp0_compare_irq_shift = CAUSEB_TI - CAUSEB_IP;
 		cp0_compare_irq = (read_c0_intctl() >> INTCTLB_IPTI) & 7;
 		cp0_perfcount_irq = (read_c0_intctl() >> INTCTLB_IPPCI) & 7;
···
 	extern char except_vec3_generic;
 	extern char except_vec4;
 	extern char except_vec3_r4000;
-	unsigned long i;
+	unsigned long i, vec_size;
+	phys_addr_t ebase_pa;
 
 	check_wait();
 
-	if (cpu_has_veic || cpu_has_vint) {
-		unsigned long size = 0x200 + VECTORSPACING*64;
-		phys_addr_t ebase_pa;
+	if (!cpu_has_mips_r2_r6) {
+		ebase = CAC_BASE;
+		ebase_pa = virt_to_phys((void *)ebase);
+		vec_size = 0x400;
 
-		ebase = (unsigned long)
-			memblock_alloc(size, 1 << fls(size));
-		if (!ebase)
+		memblock_reserve(ebase_pa, vec_size);
+	} else {
+		if (cpu_has_veic || cpu_has_vint)
+			vec_size = 0x200 + VECTORSPACING*64;
+		else
+			vec_size = PAGE_SIZE;
+
+		ebase_pa = memblock_phys_alloc(vec_size, 1 << fls(vec_size));
+		if (!ebase_pa)
 			panic("%s: Failed to allocate %lu bytes align=0x%x\n",
-			      __func__, size, 1 << fls(size));
+			      __func__, vec_size, 1 << fls(vec_size));
 
 		/*
 		 * Try to ensure ebase resides in KSeg0 if possible.
···
 		 * EVA is special though as it allows segments to be rearranged
 		 * and to become uncached during cache error handling.
 		 */
-		ebase_pa = __pa(ebase);
 		if (!IS_ENABLED(CONFIG_EVA) && !WARN_ON(ebase_pa >= 0x20000000))
 			ebase = CKSEG0ADDR(ebase_pa);
-	} else {
-		ebase = CAC_BASE;
-
-		if (cpu_has_mips_r2_r6) {
-			if (cpu_has_ebase_wg) {
-#ifdef CONFIG_64BIT
-				ebase = (read_c0_ebase_64() & ~0xfff);
-#else
-				ebase = (read_c0_ebase() & ~0xfff);
-#endif
-			} else {
-				ebase += (read_c0_ebase() & 0x3ffff000);
-			}
-		}
+		else
+			ebase = (unsigned long)phys_to_virt(ebase_pa);
 	}
 
 	if (cpu_has_mmips) {
···
 	else
 		set_handler(0x080, &except_vec3_generic, 0x80);
 
-	local_flush_icache_range(ebase, ebase + 0x400);
+	local_flush_icache_range(ebase, ebase + vec_size);
 
 	sort_extable(__start___dbe_table, __stop___dbe_table);
 

arch/mips/kvm/emulate.c

···
 	unsigned long pc = vcpu->arch.pc;
 	int index;
 
-	get_random_bytes(&index, sizeof(index));
-	index &= (KVM_MIPS_GUEST_TLB_SIZE - 1);
-
+	index = prandom_u32_max(KVM_MIPS_GUEST_TLB_SIZE);
 	tlb = &vcpu->arch.guest_tlb[index];
 
 	kvm_mips_invalidate_guest_tlb(vcpu, tlb);

arch/mips/mm/uasm-mips.c

···
 	[insn_daddiu]	= {M(daddiu_op, 0, 0, 0, 0, 0), RS | RT | SIMM},
 	[insn_daddu]	= {M(spec_op, 0, 0, 0, 0, daddu_op), RS | RT | RD},
 	[insn_ddivu]	= {M(spec_op, 0, 0, 0, 0, ddivu_op), RS | RT},
+	[insn_ddivu_r6]	= {M(spec_op, 0, 0, 0, ddivu_ddivu6_op, ddivu_op),
+				RS | RT | RD},
 	[insn_di]	= {M(cop0_op, mfmc0_op, 0, 12, 0, 0), RT},
 	[insn_dins]	= {M(spec3_op, 0, 0, 0, 0, dins_op), RS | RT | RD | RE},
 	[insn_dinsm]	= {M(spec3_op, 0, 0, 0, 0, dinsm_op), RS | RT | RD | RE},
 	[insn_dinsu]	= {M(spec3_op, 0, 0, 0, 0, dinsu_op), RS | RT | RD | RE},
 	[insn_divu]	= {M(spec_op, 0, 0, 0, 0, divu_op), RS | RT},
+	[insn_divu_r6]	= {M(spec_op, 0, 0, 0, divu_divu6_op, divu_op),
+				RS | RT | RD},
 	[insn_dmfc0]	= {M(cop0_op, dmfc_op, 0, 0, 0, 0), RT | RD | SET},
+	[insn_dmodu]	= {M(spec_op, 0, 0, 0, ddivu_dmodu_op, ddivu_op),
+				RS | RT | RD},
 	[insn_dmtc0]	= {M(cop0_op, dmtc_op, 0, 0, 0, 0), RT | RD | SET},
 	[insn_dmultu]	= {M(spec_op, 0, 0, 0, 0, dmultu_op), RS | RT},
+	[insn_dmulu]	= {M(spec_op, 0, 0, 0, dmult_dmul_op, dmultu_op),
+				RS | RT | RD},
 	[insn_drotr]	= {M(spec_op, 1, 0, 0, 0, dsrl_op), RT | RD | RE},
 	[insn_drotr32]	= {M(spec_op, 1, 0, 0, 0, dsrl32_op), RT | RD | RE},
 	[insn_dsbh]	= {M(spec3_op, 0, 0, 0, dsbh_op, dbshfl_op), RT | RD},
···
 	[insn_mfhc0]	= {M(cop0_op, mfhc0_op, 0, 0, 0, 0),  RT | RD | SET},
 	[insn_mfhi]	= {M(spec_op, 0, 0, 0, 0, mfhi_op), RD},
 	[insn_mflo]	= {M(spec_op, 0, 0, 0, 0, mflo_op), RD},
+	[insn_modu]	= {M(spec_op, 0, 0, 0, divu_modu_op, divu_op),
+				RS | RT | RD},
 	[insn_movn]	= {M(spec_op, 0, 0, 0, 0, movn_op), RS | RT | RD},
 	[insn_movz]	= {M(spec_op, 0, 0, 0, 0, movz_op), RS | RT | RD},
 	[insn_mtc0]	= {M(cop0_op, mtc_op, 0, 0, 0, 0),  RT | RD | SET},
 	[insn_mthc0]	= {M(cop0_op, mthc0_op, 0, 0, 0, 0),  RT | RD | SET},
 	[insn_mthi]	= {M(spec_op, 0, 0, 0, 0, mthi_op), RS},
 	[insn_mtlo]	= {M(spec_op, 0, 0, 0, 0, mtlo_op), RS},
+	[insn_mulu]	= {M(spec_op, 0, 0, 0, multu_mulu_op, multu_op),
+				RS | RT | RD},
 #ifndef CONFIG_CPU_MIPSR6
 	[insn_mul]	= {M(spec2_op, 0, 0, 0, 0, mul_op), RS | RT | RD},
 #else
···
 	[insn_scd]	= {M6(spec3_op, 0, 0, 0, scd6_op),  RS | RT | SIMM9},
 #endif
 	[insn_sd]	= {M(sd_op, 0, 0, 0, 0, 0),  RS | RT | SIMM},
+	[insn_seleqz]	= {M(spec_op, 0, 0, 0, 0, seleqz_op), RS | RT | RD},
+	[insn_selnez]	= {M(spec_op, 0, 0, 0, 0, selnez_op), RS | RT | RD},
 	[insn_sh]	= {M(sh_op, 0, 0, 0, 0, 0),  RS | RT | SIMM},
 	[insn_sll]	= {M(spec_op, 0, 0, 0, 0, sll_op),  RT | RD | RE},
 	[insn_sllv]	= {M(spec_op, 0, 0, 0, 0, sllv_op),  RS | RT | RD},

arch/mips/mm/uasm.c

···
 	insn_beq, insn_beql, insn_bgez, insn_bgezl, insn_bgtz, insn_blez,
 	insn_bltz, insn_bltzl, insn_bne, insn_break, insn_cache, insn_cfc1,
 	insn_cfcmsa, insn_ctc1, insn_ctcmsa, insn_daddiu, insn_daddu, insn_ddivu,
-	insn_di, insn_dins, insn_dinsm, insn_dinsu, insn_divu, insn_dmfc0,
-	insn_dmtc0, insn_dmultu, insn_drotr, insn_drotr32, insn_dsbh, insn_dshd,
-	insn_dsll, insn_dsll32, insn_dsllv, insn_dsra, insn_dsra32, insn_dsrav,
-	insn_dsrl, insn_dsrl32, insn_dsrlv, insn_dsubu, insn_eret, insn_ext,
-	insn_ins, insn_j, insn_jal, insn_jalr, insn_jr, insn_lb, insn_lbu,
-	insn_ld, insn_lddir, insn_ldpte, insn_ldx, insn_lh, insn_lhu,
-	insn_ll, insn_lld, insn_lui, insn_lw, insn_lwu, insn_lwx, insn_mfc0,
-	insn_mfhc0, insn_mfhi, insn_mflo, insn_movn, insn_movz, insn_mtc0,
-	insn_mthc0, insn_mthi, insn_mtlo, insn_mul, insn_multu, insn_nor,
-	insn_or, insn_ori, insn_pref, insn_rfe, insn_rotr, insn_sb,
-	insn_sc, insn_scd, insn_sd, insn_sh, insn_sll, insn_sllv,
-	insn_slt, insn_slti, insn_sltiu, insn_sltu, insn_sra, insn_srav,
-	insn_srl, insn_srlv, insn_subu, insn_sw, insn_sync, insn_syscall,
-	insn_tlbp, insn_tlbr, insn_tlbwi, insn_tlbwr, insn_wait, insn_wsbh,
-	insn_xor, insn_xori, insn_yield,
+	insn_ddivu_r6, insn_di, insn_dins, insn_dinsm, insn_dinsu, insn_divu,
+	insn_divu_r6, insn_dmfc0, insn_dmodu, insn_dmtc0, insn_dmultu,
+	insn_dmulu, insn_drotr, insn_drotr32, insn_dsbh, insn_dshd, insn_dsll,
+	insn_dsll32, insn_dsllv, insn_dsra, insn_dsra32, insn_dsrav, insn_dsrl,
+	insn_dsrl32, insn_dsrlv, insn_dsubu, insn_eret, insn_ext, insn_ins,
+	insn_j, insn_jal, insn_jalr, insn_jr, insn_lb, insn_lbu, insn_ld,
+	insn_lddir, insn_ldpte, insn_ldx, insn_lh, insn_lhu, insn_ll, insn_lld,
+	insn_lui, insn_lw, insn_lwu, insn_lwx, insn_mfc0, insn_mfhc0, insn_mfhi,
+	insn_mflo, insn_modu, insn_movn, insn_movz, insn_mtc0, insn_mthc0,
+	insn_mthi, insn_mtlo, insn_mul, insn_multu, insn_mulu, insn_nor,
+	insn_or, insn_ori, insn_pref, insn_rfe, insn_rotr, insn_sb, insn_sc,
+	insn_scd, insn_seleqz, insn_selnez, insn_sd, insn_sh, insn_sll,
+	insn_sllv, insn_slt, insn_slti, insn_sltiu, insn_sltu, insn_sra,
+	insn_srav, insn_srl, insn_srlv, insn_subu, insn_sw, insn_sync,
+	insn_syscall, insn_tlbp, insn_tlbr, insn_tlbwi, insn_tlbwr, insn_wait,
+	insn_wsbh, insn_xor, insn_xori, insn_yield,
 	insn_invalid /* insn_invalid must be last */
 };
 
···
 I_u1u2(_ctc1)
 I_u2u1(_ctcmsa)
 I_u1u2(_ddivu)
+I_u3u1u2(_ddivu_r6)
 I_u1u2u3(_dmfc0)
+I_u3u1u2(_dmodu)
 I_u1u2u3(_dmtc0)
 I_u1u2(_dmultu)
+I_u3u1u2(_dmulu)
 I_u2u1s3(_daddiu)
 I_u3u1u2(_daddu)
 I_u1(_di);
 I_u1u2(_divu)
+I_u3u1u2(_divu_r6)
 I_u2u1(_dsbh);
 I_u2u1(_dshd);
 I_u2u1u3(_dsll)
···
 I_u2s3u1(_lwu)
 I_u1u2u3(_mfc0)
 I_u1u2u3(_mfhc0)
+I_u3u1u2(_modu)
 I_u3u1u2(_movn)
 I_u3u1u2(_movz)
 I_u1(_mfhi)
···
 I_u1(_mtlo)
 I_u3u1u2(_mul)
 I_u1u2(_multu)
+I_u3u1u2(_mulu)
 I_u3u1u2(_nor)
 I_u3u1u2(_or)
 I_u2u1u3(_ori)
···
 I_u2s3u1(_sc)
 I_u2s3u1(_scd)
 I_u2s3u1(_sd)
+I_u3u1u2(_seleqz)
+I_u3u1u2(_selnez)
 I_u2s3u1(_sh)
 I_u2u1u3(_sll)
 I_u3u2u1(_sllv)

arch/mips/net/Makefile

···
 # MIPS networking code
 
-obj-$(CONFIG_MIPS_CBPF_JIT) += bpf_jit.o bpf_jit_asm.o
 obj-$(CONFIG_MIPS_EBPF_JIT) += ebpf_jit.o

arch/mips/net/bpf_jit.c

···
-/*
- * Just-In-Time compiler for BPF filters on MIPS
- *
- * Copyright (c) 2014 Imagination Technologies Ltd.
- * Author: Markos Chandras <[e-mail]>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License as published by the
- * Free Software Foundation; version 2 of the License.
- */
-
-#include <linux/bitops.h>
-#include <linux/compiler.h>
-#include <linux/errno.h>
-#include <linux/filter.h>
-#include <linux/if_vlan.h>
-#include <linux/moduleloader.h>
-#include <linux/netdevice.h>
-#include <linux/string.h>
-#include <linux/slab.h>
-#include <linux/types.h>
-#include <asm/asm.h>
-#include <asm/bitops.h>
-#include <asm/cacheflush.h>
-#include <asm/cpu-features.h>
-#include <asm/uasm.h>
-
-#include "bpf_jit.h"
-
-/* ABI
- * r_skb_hl	SKB header length
- * r_data	SKB data pointer
- * r_off	Offset
- * r_A		BPF register A
- * r_X		BPF register X
- * r_skb	*skb
- * r_M		*scratch memory
- * r_skb_len	SKB length
- *
- * On entry (*bpf_func)(*skb, *filter)
- * a0 = MIPS_R_A0 = skb;
- * a1 = MIPS_R_A1 = filter;
- *
- * Stack
- * ...
- * M[15]
- * M[14]
- * M[13]
- * ...
- * M[0] <-- r_M
- * saved reg k-1
- * saved reg k-2
- * ...
- * saved reg 0 <-- r_sp
- * <no argument area>
- *
- *                     Packet layout
- *
- * <--------------------- len ------------------------>
- * <--skb-len(r_skb_hl)-->< ----- skb->data_len ------>
- * ----------------------------------------------------
- * |                  skb->data                       |
- * ----------------------------------------------------
- */
-
-#define ptr typeof(unsigned long)
-
-#define SCRATCH_OFF(k)		(4 * (k))
-
-/* JIT flags */
-#define SEEN_CALL		(1 << BPF_MEMWORDS)
-#define SEEN_SREG_SFT		(BPF_MEMWORDS + 1)
-#define SEEN_SREG_BASE		(1 << SEEN_SREG_SFT)
-#define SEEN_SREG(x)		(SEEN_SREG_BASE << (x))
-#define SEEN_OFF		SEEN_SREG(2)
-#define SEEN_A			SEEN_SREG(3)
-#define SEEN_X			SEEN_SREG(4)
-#define SEEN_SKB		SEEN_SREG(5)
-#define SEEN_MEM		SEEN_SREG(6)
-/* SEEN_SK_DATA also implies skb_hl an skb_len */
-#define SEEN_SKB_DATA		(SEEN_SREG(7) | SEEN_SREG(1) | SEEN_SREG(0))
-
-/* Arguments used by JIT */
-#define ARGS_USED_BY_JIT	2 /* only applicable to 64-bit */
-
-#define SBIT(x)			(1 << (x)) /* Signed version of BIT() */
-
-/**
- * struct jit_ctx - JIT context
- * @skf:		The sk_filter
- * @prologue_bytes:	Number of bytes for prologue
- * @idx:		Instruction index
- * @flags:		JIT flags
- * @offsets:		Instruction offsets
- * @target:		Memory location for the compiled filter
- */
-struct jit_ctx {
-	const struct bpf_prog *skf;
-	unsigned int prologue_bytes;
-	u32 idx;
-	u32 flags;
-	u32 *offsets;
-	u32 *target;
-};
-
-
-static inline int optimize_div(u32 *k)
-{
-	/* power of 2 divides can be implemented with right shift */
-	if (!(*k & (*k-1))) {
-		*k = ilog2(*k);
-		return 1;
-	}
-
-	return 0;
-}
-
-static inline void emit_jit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx);
-
-/* Simply emit the instruction if the JIT memory space has been allocated */
-#define emit_instr(ctx, func, ...)			\
-do {							\
-	if ((ctx)->target != NULL) {			\
-		u32 *p = &(ctx)->target[ctx->idx];	\
-		uasm_i_##func(&p, ##__VA_ARGS__);	\
-	}						\
-	(ctx)->idx++;					\
-} while (0)
-
-/*
- * Similar to emit_instr but it must be used when we need to emit
- * 32-bit or 64-bit instructions
- */
-#define emit_long_instr(ctx, func, ...)			\
-do {							\
-	if ((ctx)->target != NULL) {			\
-		u32 *p = &(ctx)->target[ctx->idx];	\
-		UASM_i_##func(&p, ##__VA_ARGS__);	\
-	}						\
-	(ctx)->idx++;					\
-} while (0)
-
-/* Determine if immediate is within the 16-bit signed range */
-static inline bool is_range16(s32 imm)
-{
-	return !(imm >= SBIT(15) || imm < -SBIT(15));
-}
-
-static inline void emit_addu(unsigned int dst, unsigned int src1,
-			     unsigned int src2, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, addu, dst, src1, src2);
-}
-
-static inline void emit_nop(struct jit_ctx *ctx)
-{
-	emit_instr(ctx, nop);
-}
-
-/* Load a u32 immediate to a register */
-static inline void emit_load_imm(unsigned int dst, u32 imm, struct jit_ctx *ctx)
-{
-	if (ctx->target != NULL) {
-		/* addiu can only handle s16 */
-		if (!is_range16(imm)) {
-			u32 *p = &ctx->target[ctx->idx];
-			uasm_i_lui(&p, r_tmp_imm, (s32)imm >> 16);
-			p = &ctx->target[ctx->idx + 1];
-			uasm_i_ori(&p, dst, r_tmp_imm, imm & 0xffff);
-		} else {
-			u32 *p = &ctx->target[ctx->idx];
-			uasm_i_addiu(&p, dst, r_zero, imm);
-		}
-	}
-	ctx->idx++;
-
-	if (!is_range16(imm))
-		ctx->idx++;
-}
-
-static inline void emit_or(unsigned int dst, unsigned int src1,
-			   unsigned int src2, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, or, dst, src1, src2);
-}
-
-static inline void emit_ori(unsigned int dst, unsigned src, u32 imm,
-			    struct jit_ctx *ctx)
-{
-	if (imm >= BIT(16)) {
-		emit_load_imm(r_tmp, imm, ctx);
-		emit_or(dst, src, r_tmp, ctx);
-	} else {
-		emit_instr(ctx, ori, dst, src, imm);
-	}
-}
-
-static inline void emit_daddiu(unsigned int dst, unsigned int src,
-			       int imm, struct jit_ctx *ctx)
-{
-	/*
-	 * Only used for stack, so the imm is relatively small
-	 * and it fits in 15-bits
-	 */
-	emit_instr(ctx, daddiu, dst, src, imm);
-}
-
-static inline void emit_addiu(unsigned int dst, unsigned int src,
-			      u32 imm, struct jit_ctx *ctx)
-{
-	if (!is_range16(imm)) {
-		emit_load_imm(r_tmp, imm, ctx);
-		emit_addu(dst, r_tmp, src, ctx);
-	} else {
-		emit_instr(ctx, addiu, dst, src, imm);
-	}
-}
-
-static inline void emit_and(unsigned int dst, unsigned int src1,
-			    unsigned int src2, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, and, dst, src1, src2);
-}
-
-static inline void emit_andi(unsigned int dst, unsigned int src,
-			     u32 imm, struct jit_ctx *ctx)
-{
-	/* If imm does not fit in u16 then load it to register */
-	if (imm >= BIT(16)) {
-		emit_load_imm(r_tmp, imm, ctx);
-		emit_and(dst, src, r_tmp, ctx);
-	} else {
-		emit_instr(ctx, andi, dst, src, imm);
-	}
-}
-
-static inline void emit_xor(unsigned int dst, unsigned int src1,
-			    unsigned int src2, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, xor, dst, src1, src2);
-}
-
-static inline void emit_xori(ptr dst, ptr src, u32 imm, struct jit_ctx *ctx)
-{
-	/* If imm does not fit in u16 then load it to register */
-	if (imm >= BIT(16)) {
-		emit_load_imm(r_tmp, imm, ctx);
-		emit_xor(dst, src, r_tmp, ctx);
-	} else {
-		emit_instr(ctx, xori, dst, src, imm);
-	}
-}
-
-static inline void emit_stack_offset(int offset, struct jit_ctx *ctx)
-{
-	emit_long_instr(ctx, ADDIU, r_sp, r_sp, offset);
-}
-
-static inline void emit_subu(unsigned int dst, unsigned int src1,
-			     unsigned int src2, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, subu, dst, src1, src2);
-}
-
-static inline void emit_neg(unsigned int reg, struct jit_ctx *ctx)
-{
-	emit_subu(reg, r_zero, reg, ctx);
-}
-
-static inline void emit_sllv(unsigned int dst, unsigned int src,
-			     unsigned int sa, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, sllv, dst, src, sa);
-}
-
-static inline void emit_sll(unsigned int dst, unsigned int src,
-			    unsigned int sa, struct jit_ctx *ctx)
-{
-	/* sa is 5-bits long */
-	if (sa >= BIT(5))
-		/* Shifting >= 32 results in zero */
-		emit_jit_reg_move(dst, r_zero, ctx);
-	else
-		emit_instr(ctx, sll, dst, src, sa);
-}
-
-static inline void emit_srlv(unsigned int dst, unsigned int src,
-			     unsigned int sa, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, srlv, dst, src, sa);
-}
-
-static inline void emit_srl(unsigned int dst, unsigned int src,
-			    unsigned int sa, struct jit_ctx *ctx)
-{
-	/* sa is 5-bits long */
-	if (sa >= BIT(5))
-		/* Shifting >= 32 results in zero */
-		emit_jit_reg_move(dst, r_zero, ctx);
-	else
-		emit_instr(ctx, srl, dst, src, sa);
-}
-
-static inline void emit_slt(unsigned int dst, unsigned int src1,
-			    unsigned int src2, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, slt, dst, src1, src2);
-}
-
-static inline void emit_sltu(unsigned int dst, unsigned int src1,
-			     unsigned int src2, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, sltu, dst, src1, src2);
-}
-
-static inline void emit_sltiu(unsigned dst, unsigned int src,
-			      unsigned int imm, struct jit_ctx *ctx)
-{
-	/* 16 bit immediate */
-	if (!is_range16((s32)imm)) {
-		emit_load_imm(r_tmp, imm, ctx);
-		emit_sltu(dst, src, r_tmp, ctx);
-	} else {
-		emit_instr(ctx, sltiu, dst, src, imm);
-	}
-
-}
-
-/* Store register on the stack */
-static inline void emit_store_stack_reg(ptr reg, ptr base,
-					unsigned int offset,
-					struct jit_ctx *ctx)
-{
-	emit_long_instr(ctx, SW, reg, offset, base);
-}
-
-static inline void emit_store(ptr reg, ptr base, unsigned int offset,
-			      struct jit_ctx *ctx)
-{
-	emit_instr(ctx, sw, reg, offset, base);
-}
-
-static inline void emit_load_stack_reg(ptr reg, ptr base,
-				       unsigned int offset,
-				       struct jit_ctx *ctx)
-{
-	emit_long_instr(ctx, LW, reg, offset, base);
-}
-
-static inline void emit_load(unsigned int reg, unsigned int base,
-			     unsigned int offset, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, lw, reg, offset, base);
-}
-
-static inline void emit_load_byte(unsigned int reg, unsigned int base,
-				  unsigned int offset, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, lb, reg, offset, base);
-}
-
-static inline void emit_half_load(unsigned int reg, unsigned int base,
-				  unsigned int offset, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, lh, reg, offset, base);
-}
-
-static inline void emit_half_load_unsigned(unsigned int reg, unsigned int base,
-					   unsigned int offset, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, lhu, reg, offset, base);
-}
-
-static inline void emit_mul(unsigned int dst, unsigned int src1,
-			    unsigned int src2, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, mul, dst, src1, src2);
-}
-
-static inline void emit_div(unsigned int dst, unsigned int src,
-			    struct jit_ctx *ctx)
-{
-	if (ctx->target != NULL) {
-		u32 *p = &ctx->target[ctx->idx];
-		uasm_i_divu(&p, dst, src);
-		p = &ctx->target[ctx->idx + 1];
-		uasm_i_mflo(&p, dst);
-	}
-	ctx->idx += 2; /* 2 insts */
-}
-
-static inline void emit_mod(unsigned int dst, unsigned int src,
-			    struct jit_ctx *ctx)
-{
-	if (ctx->target != NULL) {
-		u32 *p = &ctx->target[ctx->idx];
-		uasm_i_divu(&p, dst, src);
-		p = &ctx->target[ctx->idx + 1];
-		uasm_i_mfhi(&p, dst);
-	}
-	ctx->idx += 2; /* 2 insts */
-}
-
-static inline void emit_dsll(unsigned int dst, unsigned int src,
-			     unsigned int sa, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, dsll, dst, src, sa);
-}
-
-static inline void emit_dsrl32(unsigned int dst, unsigned int src,
-			       unsigned int sa, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, dsrl32, dst, src, sa);
-}
-
-static inline void emit_wsbh(unsigned int dst, unsigned int src,
-			     struct jit_ctx *ctx)
-{
-	emit_instr(ctx, wsbh, dst, src);
-}
-
-/* load pointer to register */
-static inline void emit_load_ptr(unsigned int dst, unsigned int src,
-				     int imm, struct jit_ctx *ctx)
-{
-	/* src contains the base addr of the 32/64-pointer */
-	emit_long_instr(ctx, LW, dst, imm, src);
-}
-
-/* load a function pointer to register */
-static inline void emit_load_func(unsigned int reg, ptr imm,
-				  struct jit_ctx *ctx)
-{
-	if (IS_ENABLED(CONFIG_64BIT)) {
-		/* At this point imm is always 64-bit */
-		emit_load_imm(r_tmp, (u64)imm >> 32, ctx);
-		emit_dsll(r_tmp_imm, r_tmp, 16, ctx); /* left shift by 16 */
-		emit_ori(r_tmp, r_tmp_imm, (imm >> 16) & 0xffff, ctx);
-		emit_dsll(r_tmp_imm, r_tmp, 16, ctx); /* left shift by 16 */
-		emit_ori(reg, r_tmp_imm, imm & 0xffff, ctx);
-	} else {
-		emit_load_imm(reg, imm, ctx);
-	}
-}
-
-/* Move to real MIPS register */
-static inline void emit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx)
-{
-	emit_long_instr(ctx, ADDU, dst, src, r_zero);
-}
-
-/* Move to JIT (32-bit) register */
-static inline void emit_jit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx)
-{
-	emit_addu(dst, src, r_zero, ctx);
-}
-
-/* Compute the immediate value for PC-relative branches. */
-static inline u32 b_imm(unsigned int tgt, struct jit_ctx *ctx)
-{
-	if (ctx->target == NULL)
-		return 0;
-
-	/*
-	 * We want a pc-relative branch. We only do forward branches
-	 * so tgt is always after pc. tgt is the instruction offset
-	 * we want to jump to.
-
-	 * Branch on MIPS:
-	 * I: target_offset <- sign_extend(offset)
-	 * I+1: PC += target_offset (delay slot)
-	 *
-	 * ctx->idx currently points to the branch instruction
-	 * but the offset is added to the delay slot so we need
-	 * to subtract 4.
-	 */
-	return ctx->offsets[tgt] -
-		(ctx->idx * 4 - ctx->prologue_bytes) - 4;
-}
-
-static inline void emit_bcond(int cond, unsigned int reg1, unsigned int reg2,
-			     unsigned int imm, struct jit_ctx *ctx)
-{
-	if (ctx->target != NULL) {
-		u32 *p = &ctx->target[ctx->idx];
-
-		switch (cond) {
-		case MIPS_COND_EQ:
-			uasm_i_beq(&p, reg1, reg2, imm);
-			break;
-		case MIPS_COND_NE:
-			uasm_i_bne(&p, reg1, reg2, imm);
-			break;
-		case MIPS_COND_ALL:
-			uasm_i_b(&p, imm);
-			break;
-		default:
-			pr_warn("%s: Unhandled branch conditional: %d\n",
-				__func__, cond);
-		}
-	}
-	ctx->idx++;
-}
-
-static inline void emit_b(unsigned int imm, struct jit_ctx *ctx)
-{
-	emit_bcond(MIPS_COND_ALL, r_zero, r_zero, imm, ctx);
-}
-
-static inline void emit_jalr(unsigned int link, unsigned int reg,
-			     struct jit_ctx *ctx)
-{
-	emit_instr(ctx, jalr, link, reg);
-}
-
-static inline void emit_jr(unsigned int reg, struct jit_ctx *ctx)
-{
-	emit_instr(ctx, jr, reg);
-}
-
-static inline u16 align_sp(unsigned int num)
-{
-	/* Double word alignment for 32-bit, quadword for 64-bit */
-	unsigned int align = IS_ENABLED(CONFIG_64BIT) ? 16 : 8;
-	num = (num + (align - 1)) & -align;
-	return num;
-}
-
-static void save_bpf_jit_regs(struct jit_ctx *ctx, unsigned offset)
-{
-	int i = 0, real_off = 0;
-	u32 sflags, tmp_flags;
-
-	/* Adjust the stack pointer */
-	if (offset)
-		emit_stack_offset(-align_sp(offset), ctx);
-
-	tmp_flags = sflags = ctx->flags >> SEEN_SREG_SFT;
-	/* sflags is essentially a bitmap */
-	while (tmp_flags) {
-		if ((sflags >> i) & 0x1) {
-			emit_store_stack_reg(MIPS_R_S0 + i, r_sp, real_off,
-					     ctx);
-			real_off += SZREG;
-		}
-		i++;
-		tmp_flags >>= 1;
-	}
-
-	/* save return address */
-	if (ctx->flags & SEEN_CALL) {
-		emit_store_stack_reg(r_ra, r_sp, real_off, ctx);
-		real_off += SZREG;
-	}
-
-	/* Setup r_M leaving the alignment gap if necessary */
-	if (ctx->flags & SEEN_MEM) {
-		if (real_off % (SZREG * 2))
-			real_off += SZREG;
-		emit_long_instr(ctx, ADDIU, r_M, r_sp, real_off);
-	}
-}
-
-static void restore_bpf_jit_regs(struct jit_ctx *ctx,
-				 unsigned int offset)
-{
-	int i, real_off = 0;
-	u32 sflags, tmp_flags;
-
-	tmp_flags = sflags = ctx->flags >> SEEN_SREG_SFT;
-	/* sflags is a bitmap */
-	i = 0;
-	while (tmp_flags) {
-		if ((sflags >> i) & 0x1) {
-			emit_load_stack_reg(MIPS_R_S0 + i, r_sp, real_off,
-					    ctx);
-			real_off += SZREG;
-		}
-		i++;
-		tmp_flags >>= 1;
-	}
-
-	/* restore return address */
-	if (ctx->flags & SEEN_CALL)
-		emit_load_stack_reg(r_ra, r_sp, real_off, ctx);
-
-	/* Restore the sp and discard the scrach memory */
-	if (offset)
-		emit_stack_offset(align_sp(offset), ctx);
-}
-
-static unsigned int get_stack_depth(struct jit_ctx *ctx)
-{
-	int sp_off = 0;
-
-
-	/* How may s* regs do we need to preserved? */
-	sp_off += hweight32(ctx->flags >> SEEN_SREG_SFT) * SZREG;
-
-	if (ctx->flags & SEEN_MEM)
-		sp_off += 4 * BPF_MEMWORDS; /* BPF_MEMWORDS are 32-bit */
-
-	if (ctx->flags & SEEN_CALL)
-		sp_off += SZREG; /* Space for our ra register */
-
-	return sp_off;
-}
-
-static void build_prologue(struct jit_ctx *ctx)
-{
-	int sp_off;
-
-	/* Calculate the total offset for the stack pointer */
-	sp_off = get_stack_depth(ctx);
-	save_bpf_jit_regs(ctx, sp_off);
-
-	if (ctx->flags & SEEN_SKB)
-		emit_reg_move(r_skb, MIPS_R_A0, ctx);
-
-	if (ctx->flags & SEEN_SKB_DATA) {
-		/* Load packet length */
-		emit_load(r_skb_len, r_skb, offsetof(struct sk_buff, len),
-			  ctx);
-		emit_load(r_tmp, r_skb, offsetof(struct sk_buff, data_len),
-			  ctx);
-		/* Load the data pointer */
-		emit_load_ptr(r_skb_data, r_skb,
-			      offsetof(struct sk_buff, data), ctx);
-		/* Load the header length */
-		emit_subu(r_skb_hl, r_skb_len, r_tmp, ctx);
-	}
-
-	if (ctx->flags & SEEN_X)
-		emit_jit_reg_move(r_X, r_zero, ctx);
-
-	/*
-	 * Do not leak kernel data to userspace, we only need to clear
-	 * r_A if it is ever used.  In fact if it is never used, we
-	 * will not save/restore it, so clearing it in this case would
-	 * corrupt the state of the caller.
-	 */
-	if (bpf_needs_clear_a(&ctx->skf->insns[0]) &&
-	    (ctx->flags & SEEN_A))
-		emit_jit_reg_move(r_A, r_zero, ctx);
-}
-
-static void build_epilogue(struct jit_ctx *ctx)
-{
-	unsigned int sp_off;
-
-	/* Calculate the total offset for the stack pointer */
-
-	sp_off = get_stack_depth(ctx);
-	restore_bpf_jit_regs(ctx, sp_off);
-
-	/* Return */
-	emit_jr(r_ra, ctx);
-	emit_nop(ctx);
-}
-
-#define CHOOSE_LOAD_FUNC(K, func) \
-	((int)K < 0 ? ((int)K >= SKF_LL_OFF ? func##_negative : func) : \
-	 func##_positive)
-
-static int build_body(struct jit_ctx *ctx)
-{
-	const struct bpf_prog *prog = ctx->skf;
-	const struct sock_filter *inst;
-	unsigned int i, off, condt;
-	u32 k, b_off __maybe_unused;
-	u8 (*sk_load_func)(unsigned long *skb, int offset);
-
-	for (i = 0; i < prog->len; i++) {
-		u16 code;
-
-		inst = &(prog->insns[i]);
-		pr_debug("%s: code->0x%02x, jt->0x%x, jf->0x%x, k->0x%x\n",
-			 __func__, inst->code, inst->jt, inst->jf, inst->k);
-		k = inst->k;
-		code = bpf_anc_helper(inst);
-
-		if (ctx->target == NULL)
-			ctx->offsets[i] = ctx->idx * 4;
-
-		switch (code) {
-		case BPF_LD | BPF_IMM:
-			/* A <- k ==> li r_A, k */
-			ctx->flags |= SEEN_A;
-			emit_load_imm(r_A, k, ctx);
-			break;
-		case BPF_LD | BPF_W | BPF_LEN:
-			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, len) != 4);
-			/* A <- len ==> lw r_A, offset(skb) */
-			ctx->flags |= SEEN_SKB | SEEN_A;
-			off = offsetof(struct sk_buff, len);
-			emit_load(r_A, r_skb, off, ctx);
-			break;
-		case BPF_LD | BPF_MEM:
-			/* A <- M[k] ==> lw r_A, offset(M) */
-			ctx->flags |= SEEN_MEM | SEEN_A;
-			emit_load(r_A, r_M, SCRATCH_OFF(k), ctx);
-			break;
-		case BPF_LD | BPF_W | BPF_ABS:
-			/* A <- P[k:4] */
-			sk_load_func = CHOOSE_LOAD_FUNC(k, sk_load_word);
-			goto load;
-		case BPF_LD | BPF_H | BPF_ABS:
-			/* A <- P[k:2] */
-			sk_load_func = CHOOSE_LOAD_FUNC(k, sk_load_half);
-			goto load;
-		case BPF_LD | BPF_B | BPF_ABS:
-			/* A <- P[k:1] */
-			sk_load_func = CHOOSE_LOAD_FUNC(k, sk_load_byte);
-load:
-			emit_load_imm(r_off, k, ctx);
-load_common:
-			ctx->flags |= SEEN_CALL | SEEN_OFF |
-				SEEN_SKB | SEEN_A | SEEN_SKB_DATA;
-
-			emit_load_func(r_s0, (ptr)sk_load_func, ctx);
-			emit_reg_move(MIPS_R_A0, r_skb, ctx);
-			emit_jalr(MIPS_R_RA, r_s0, ctx);
-			/* Load second argument to delay slot */
-			emit_reg_move(MIPS_R_A1, r_off, ctx);
-			/* Check the error value */
-			emit_bcond(MIPS_COND_EQ, r_ret, 0, b_imm(i + 1, ctx),
-				   ctx);
-			/* Load return register on DS for failures */
-			emit_reg_move(r_ret, r_zero, ctx);
-			/* Return with error */
-			emit_b(b_imm(prog->len, ctx), ctx);
-			emit_nop(ctx);
-			break;
-		case BPF_LD | BPF_W | BPF_IND:
-			/* A <- P[X + k:4] */
-			sk_load_func = sk_load_word;
-			goto load_ind;
-		case BPF_LD | BPF_H | BPF_IND:
-			/* A <- P[X + k:2] */
-			sk_load_func = sk_load_half;
-			goto load_ind;
-		case BPF_LD | BPF_B | BPF_IND:
-			/* A <- P[X + k:1] */
-			sk_load_func = sk_load_byte;
-load_ind:
-			ctx->flags |= SEEN_OFF | SEEN_X;
-			emit_addiu(r_off, r_X, k, ctx);
-			goto load_common;
-		case BPF_LDX | BPF_IMM:
-			/* X <- k */
-			ctx->flags |= SEEN_X;
-			emit_load_imm(r_X, k, ctx);
-			break;
-		case BPF_LDX | BPF_MEM:
-			/* X <- M[k] */
-			ctx->flags |= SEEN_X | SEEN_MEM;
-			emit_load(r_X, r_M, SCRATCH_OFF(k), ctx);
-			break;
-		case BPF_LDX | BPF_W | BPF_LEN:
-			/* X <- len */
-			ctx->flags |= SEEN_X | SEEN_SKB;
-			off = offsetof(struct sk_buff, len);
-			emit_load(r_X, r_skb, off, ctx);
-			break;
-		case BPF_LDX | BPF_B | BPF_MSH:
-			/* X <- 4 * (P[k:1] & 0xf) */
-			ctx->flags |= SEEN_X | SEEN_CALL | SEEN_SKB;
-			/* Load offset to a1 */
-			emit_load_func(r_s0, (ptr)sk_load_byte, ctx);
-			/*
-			 * This may emit two instructions so it may not fit
-			 * in the delay slot. So use a0 in the delay slot.
-			 */
-			emit_load_imm(MIPS_R_A1, k, ctx);
-			emit_jalr(MIPS_R_RA, r_s0, ctx);
-			emit_reg_move(MIPS_R_A0, r_skb, ctx); /* delay slot */
-			/* Check the error value */
-			emit_bcond(MIPS_COND_NE, r_ret, 0,
-				   b_imm(prog->len, ctx), ctx);
-			emit_reg_move(r_ret, r_zero, ctx);
-			/* We are good */
-			/* X <- P[1:K] & 0xf */
-			emit_andi(r_X, r_A, 0xf, ctx);
-			/* X << 2 */
-			emit_b(b_imm(i + 1, ctx), ctx);
-			emit_sll(r_X, r_X, 2, ctx); /* delay slot */
-			break;
-		case BPF_ST:
-			/* M[k] <- A */
-			ctx->flags |= SEEN_MEM | SEEN_A;
-			emit_store(r_A, r_M, SCRATCH_OFF(k), ctx);
-			break;
-		case BPF_STX:
-			/* M[k] <- X */
-			ctx->flags |= SEEN_MEM | SEEN_X;
-			emit_store(r_X, r_M, SCRATCH_OFF(k), ctx);
-			break;
-		case BPF_ALU | BPF_ADD | BPF_K:
-			/* A += K */
-			ctx->flags |= SEEN_A;
-			emit_addiu(r_A, r_A, k, ctx);
-			break;
-		case BPF_ALU | BPF_ADD | BPF_X:
-			/* A += X */
-			ctx->flags |= SEEN_A | SEEN_X;
-			emit_addu(r_A, r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_SUB | BPF_K:
-			/* A -= K */
-			ctx->flags |= SEEN_A;
-			emit_addiu(r_A, r_A, -k, ctx);
-			break;
-		case BPF_ALU | BPF_SUB | BPF_X:
-			/* A -= X */
-			ctx->flags |= SEEN_A | SEEN_X;
-			emit_subu(r_A, r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_MUL | BPF_K:
-			/* A *= K */
-			/* Load K to scratch register before MUL */
-			ctx->flags |= SEEN_A;
-			emit_load_imm(r_s0, k, ctx);
-			emit_mul(r_A, r_A, r_s0, ctx);
-			break;
-		case BPF_ALU | BPF_MUL | BPF_X:
-			/* A *= X */
-			ctx->flags |= SEEN_A | SEEN_X;
-			emit_mul(r_A, r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_DIV | BPF_K:
-			/* A /= k */
-			if (k == 1)
-				break;
-			if (optimize_div(&k)) {
-				ctx->flags |= SEEN_A;
-				emit_srl(r_A, r_A, k, ctx);
-				break;
-			}
-			ctx->flags |= SEEN_A;
-			emit_load_imm(r_s0, k, ctx);
-			emit_div(r_A, r_s0, ctx);
-			break;
-		case BPF_ALU | BPF_MOD | BPF_K:
-			/* A %= k */
-			if (k == 1) {
-				ctx->flags |= SEEN_A;
-				emit_jit_reg_move(r_A, r_zero, ctx);
-			} else {
-				ctx->flags |= SEEN_A;
-				emit_load_imm(r_s0, k, ctx);
-				emit_mod(r_A, r_s0, ctx);
-			}
-			break;
-		case BPF_ALU | BPF_DIV | BPF_X:
-			/* A /= X */
-			ctx->flags |= SEEN_X | SEEN_A;
-			/* Check if r_X is zero */
-			emit_bcond(MIPS_COND_EQ, r_X, r_zero,
-				   b_imm(prog->len, ctx), ctx);
-			emit_load_imm(r_ret, 0, ctx); /* delay slot */
-			emit_div(r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_MOD | BPF_X:
-			/* A %= X */
-			ctx->flags |= SEEN_X | SEEN_A;
-			/* Check if r_X is zero */
-			emit_bcond(MIPS_COND_EQ, r_X, r_zero,
-				   b_imm(prog->len, ctx), ctx);
-			emit_load_imm(r_ret, 0, ctx); /* delay slot */
-			emit_mod(r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_OR | BPF_K:
-			/* A |= K */
-			ctx->flags |= SEEN_A;
-			emit_ori(r_A, r_A, k, ctx);
-			break;
-		case BPF_ALU | BPF_OR | BPF_X:
-			/* A |= X */
-			ctx->flags |= SEEN_A;
-			emit_ori(r_A, r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_XOR | BPF_K:
-			/* A ^= k */
-			ctx->flags |= SEEN_A;
-			emit_xori(r_A, r_A, k, ctx);
-			break;
-		case BPF_ANC | SKF_AD_ALU_XOR_X:
-		case BPF_ALU | BPF_XOR | BPF_X:
-			/* A ^= X */
-			ctx->flags |= SEEN_A;
-			emit_xor(r_A, r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_AND | BPF_K:
-			/* A &= K */
-			ctx->flags |= SEEN_A;
-			emit_andi(r_A, r_A, k, ctx);
-			break;
-		case BPF_ALU | BPF_AND | BPF_X:
-			/* A &= X */
-			ctx->flags |= SEEN_A | SEEN_X;
-			emit_and(r_A, r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_LSH | BPF_K:
-			/* A <<= K */
-			ctx->flags |= SEEN_A;
-			emit_sll(r_A, r_A, k, ctx);
-			break;
-		case BPF_ALU | BPF_LSH | BPF_X:
-			/* A <<= X */
-			ctx->flags |= SEEN_A | SEEN_X;
-			emit_sllv(r_A, r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_RSH | BPF_K:
-			/* A >>= K */
-			ctx->flags |= SEEN_A;
-			emit_srl(r_A, r_A, k, ctx);
-			break;
-		case BPF_ALU | BPF_RSH | BPF_X:
-			ctx->flags |= SEEN_A | SEEN_X;
-			emit_srlv(r_A, r_A, r_X, ctx);
-			break;
-		case BPF_ALU | BPF_NEG:
-			/* A = -A */
-			ctx->flags |= SEEN_A;
-			emit_neg(r_A, ctx);
-			break;
-		case BPF_JMP | BPF_JA:
-			/* pc += K */
-			emit_b(b_imm(i + k + 1, ctx), ctx);
-			emit_nop(ctx);
-			break;
-		case BPF_JMP | BPF_JEQ | BPF_K:
-			/* pc += ( A == K ) ? pc->jt : pc->jf */
-			condt = MIPS_COND_EQ | MIPS_COND_K;
-			goto jmp_cmp;
-		case BPF_JMP | BPF_JEQ | BPF_X:
-			ctx->flags |= SEEN_X;
-			/* pc += ( A == X ) ? pc->jt : pc->jf */
-			condt = MIPS_COND_EQ | MIPS_COND_X;
-			goto jmp_cmp;
-		case BPF_JMP | BPF_JGE | BPF_K:
-			/* pc += ( A >= K ) ? pc->jt : pc->jf */
-			condt = MIPS_COND_GE | MIPS_COND_K;
-			goto jmp_cmp;
-		case BPF_JMP | BPF_JGE | BPF_X:
-			ctx->flags |= SEEN_X;
-			/* pc += ( A >= X ) ? pc->jt : pc->jf */
-			condt = MIPS_COND_GE | MIPS_COND_X;
-			goto jmp_cmp;
-		case BPF_JMP | BPF_JGT | BPF_K:
-			/* pc += ( A > K ) ? pc->jt : pc->jf */
-			condt = MIPS_COND_GT | MIPS_COND_K;
-			goto jmp_cmp;
-		case BPF_JMP | BPF_JGT | BPF_X:
-			ctx->flags |= SEEN_X;
-			/* pc += ( A > X ) ? pc->jt : pc->jf */
-			condt = MIPS_COND_GT | MIPS_COND_X;
-jmp_cmp:
-			/* Greater or Equal */
-			if ((condt & MIPS_COND_GE) ||
-			    (condt & MIPS_COND_GT)) {
-				if (condt & MIPS_COND_K) { /* K */
-					ctx->flags |= SEEN_A;
-					emit_sltiu(r_s0, r_A, k, ctx);
-				} else { /* X */
-					ctx->flags |= SEEN_A |
-						SEEN_X;
-					emit_sltu(r_s0, r_A, r_X, ctx);
-				}
-				/* A < (K|X) ? r_scrach = 1 */
-				b_off = b_imm(i + inst->jf + 1, ctx);
-				emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off,
-					   ctx);
-				emit_nop(ctx);
-				/* A > (K|X) ? scratch = 0 */
-				if (condt & MIPS_COND_GT) {
-					/* Checking for equality */
-					ctx->flags |= SEEN_A | SEEN_X;
-					if (condt & MIPS_COND_K)
-						emit_load_imm(r_s0, k, ctx);
-					else
-						emit_jit_reg_move(r_s0, r_X,
-								  ctx);
-					b_off = b_imm(i + inst->jf + 1, ctx);
-					emit_bcond(MIPS_COND_EQ, r_A, r_s0,
-						   b_off, ctx);
-					emit_nop(ctx);
-					/* Finally, A > K|X */
-					b_off = b_imm(i + inst->jt + 1, ctx);
-					emit_b(b_off, ctx);
-					emit_nop(ctx);
-				} else {
-					/* A >= (K|X) so jump */
-					b_off = b_imm(i + inst->jt + 1, ctx);
-					emit_b(b_off, ctx);
-					emit_nop(ctx);
-				}
-			} else {
-				/* A == K|X */
-				if (condt & MIPS_COND_K) { /* K */
-					ctx->flags |= SEEN_A;
-					emit_load_imm(r_s0, k, ctx);
-					/* jump true */
-					b_off = b_imm(i + inst->jt + 1, ctx);
-					emit_bcond(MIPS_COND_EQ, r_A, r_s0,
-						   b_off, ctx);
-					emit_nop(ctx);
-					/* jump false */
-					b_off = b_imm(i + inst->jf + 1,
-						      ctx);
-					emit_bcond(MIPS_COND_NE, r_A, r_s0,
-						   b_off, ctx);
-					emit_nop(ctx);
-				} else { /* X */
-					/* jump true */
-					ctx->flags |= SEEN_A | SEEN_X;
-					b_off = b_imm(i + inst->jt + 1,
-						      ctx);
-					emit_bcond(MIPS_COND_EQ, r_A, r_X,
-						   b_off, ctx);
-					emit_nop(ctx);
-					/* jump false */
-					b_off = b_imm(i + inst->jf + 1, ctx);
-					emit_bcond(MIPS_COND_NE, r_A, r_X,
-						   b_off, ctx);
-					emit_nop(ctx);
-				}
-			}
-			break;
-		case BPF_JMP | BPF_JSET | BPF_K:
-			ctx->flags |= SEEN_A;
-			/* pc += (A & K) ? pc -> jt : pc -> jf */
-			emit_load_imm(r_s1, k, ctx);
-			emit_and(r_s0, r_A, r_s1, ctx);
-			/* jump true */
-			b_off = b_imm(i + inst->jt + 1, ctx);
-			emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off, ctx);
-			emit_nop(ctx);
-			/* jump false */
-			b_off = b_imm(i + inst->jf + 1, ctx);
-			emit_b(b_off, ctx);
-			emit_nop(ctx);
-			break;
-		case BPF_JMP | BPF_JSET | BPF_X:
-			ctx->flags |= SEEN_X | SEEN_A;
-			/* pc += (A & X) ? pc -> jt : pc -> jf */
-			emit_and(r_s0, r_A, r_X, ctx);
-			/* jump true */
-			b_off = b_imm(i + inst->jt + 1, ctx);
-			emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off, ctx);
-			emit_nop(ctx);
-			/* jump false */
-			b_off = b_imm(i + inst->jf + 1, ctx);
-			emit_b(b_off, ctx);
-			emit_nop(ctx);
-			break;
-		case BPF_RET | BPF_A:
-			ctx->flags |= SEEN_A;
-			if (i != prog->len - 1)
-				/*
-				 * If this is not the last instruction
-				 * then jump to the epilogue
-				 */
-				emit_b(b_imm(prog->len, ctx), ctx);
-			emit_reg_move(r_ret, r_A, ctx); /* delay slot */
-			break;
-		case BPF_RET | BPF_K:
-			/*
-			 * It can emit two instructions so it does not fit on
-			 * the delay slot.
-			 */
-			emit_load_imm(r_ret, k, ctx);
-			if (i != prog->len - 1) {
-				/*
-				 * If this is not the last instruction
-				 * then jump to the epilogue
-				 */
-				emit_b(b_imm(prog->len, ctx), ctx);
-				emit_nop(ctx);
-			}
-			break;
-		case BPF_MISC | BPF_TAX:
-			/* X = A */
-			ctx->flags |= SEEN_X | SEEN_A;
-			emit_jit_reg_move(r_X, r_A, ctx);
-			break;
-		case BPF_MISC | BPF_TXA:
-			/* A = X */
-			ctx->flags |= SEEN_A | SEEN_X;
-			emit_jit_reg_move(r_A, r_X, ctx);
-			break;
-		/* AUX */
-		case BPF_ANC | SKF_AD_PROTOCOL:
-			/* A = ntohs(skb->protocol */
-			ctx->flags |= SEEN_SKB | SEEN_OFF | SEEN_A;
-			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff,
-						  protocol) != 2);
-			off = offsetof(struct sk_buff, protocol);
-			emit_half_load(r_A, r_skb, off, ctx);
-#ifdef CONFIG_CPU_LITTLE_ENDIAN
-			/* This needs little endian fixup */
-			if (cpu_has_wsbh) {
-				/* R2 and later have the wsbh instruction */
-				emit_wsbh(r_A, r_A, ctx);
-			} else {
-				/* Get first byte */
-				emit_andi(r_tmp_imm, r_A, 0xff, ctx);
-				/* Shift it */
-				emit_sll(r_tmp, r_tmp_imm, 8, ctx);
-				/* Get second byte */
-				emit_srl(r_tmp_imm, r_A, 8, ctx);
-				emit_andi(r_tmp_imm, r_tmp_imm, 0xff, ctx);
-				/* Put everyting together in r_A */
-				emit_or(r_A, r_tmp, r_tmp_imm, ctx);
-			}
-#endif
-			break;
-		case BPF_ANC | SKF_AD_CPU:
-			ctx->flags |= SEEN_A | SEEN_OFF;
-			/* A = current_thread_info()->cpu */
-			BUILD_BUG_ON(FIELD_SIZEOF(struct thread_info,
-						  cpu) != 4);
-			off = offsetof(struct thread_info, cpu);
-			/* $28/gp points to the thread_info struct */
-			emit_load(r_A, 28, off, ctx);
-			break;
-		case BPF_ANC | SKF_AD_IFINDEX:
-			/* A = skb->dev->ifindex */
-		case BPF_ANC | SKF_AD_HATYPE:
-			/* A = skb->dev->type */
-			ctx->flags |= SEEN_SKB | SEEN_A;
-			off = offsetof(struct sk_buff, dev);
-			/* Load *dev pointer */
-			emit_load_ptr(r_s0, r_skb, off, ctx);
-			/* error (0) in the delay slot */
-			emit_bcond(MIPS_COND_EQ, r_s0, r_zero,
-				   b_imm(prog->len, ctx), ctx);
-			emit_reg_move(r_ret, r_zero, ctx);
-			if (code == (BPF_ANC | SKF_AD_IFINDEX)) {
-				BUILD_BUG_ON(FIELD_SIZEOF(struct net_device, ifindex) != 4);
-				off = offsetof(struct net_device, ifindex);
-				emit_load(r_A, r_s0, off, ctx);
-			} else { /* (code == (BPF_ANC | SKF_AD_HATYPE) */
-				BUILD_BUG_ON(FIELD_SIZEOF(struct net_device, type) != 2);
-				off = offsetof(struct net_device, type);
-				emit_half_load_unsigned(r_A, r_s0, off, ctx);
-			}
-			break;
-		case BPF_ANC | SKF_AD_MARK:
-			ctx->flags |= SEEN_SKB | SEEN_A;
-			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, mark) != 4);
-			off = offsetof(struct sk_buff, mark);
-			emit_load(r_A, r_skb, off, ctx);
-			break;
-		case BPF_ANC | SKF_AD_RXHASH:
-			ctx->flags |= SEEN_SKB | SEEN_A;
-			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, hash) != 4);
-			off = offsetof(struct sk_buff, hash);
-			emit_load(r_A, r_skb, off, ctx);
-			break;
-		case BPF_ANC | SKF_AD_VLAN_TAG:
-			ctx->flags |= SEEN_SKB | SEEN_A;
-			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff,
-						  vlan_tci) != 2);
-			off = offsetof(struct sk_buff, vlan_tci);
-			emit_half_load_unsigned(r_A, r_skb, off, ctx);
-			break;
-		case BPF_ANC | SKF_AD_VLAN_TAG_PRESENT:
-			ctx->flags |= SEEN_SKB | SEEN_A;
-			emit_load_byte(r_A, r_skb, PKT_VLAN_PRESENT_OFFSET(), ctx);
-			if (PKT_VLAN_PRESENT_BIT)
-				emit_srl(r_A, r_A, PKT_VLAN_PRESENT_BIT, ctx);
-			if (PKT_VLAN_PRESENT_BIT < 7)
-				emit_andi(r_A, r_A, 1, ctx);
-			break;
-		case BPF_ANC | SKF_AD_PKTTYPE:
-			ctx->flags |= SEEN_SKB;
-
-			emit_load_byte(r_tmp, r_skb, PKT_TYPE_OFFSET(), ctx);
-			/* Keep only the last 3 bits */
-			emit_andi(r_A, r_tmp, PKT_TYPE_MAX, ctx);
-#ifdef __BIG_ENDIAN_BITFIELD
-			/* Get the actual packet type to the lower 3 bits */
-			emit_srl(r_A, r_A, 5, ctx);
-#endif
-			break;
-		case BPF_ANC | SKF_AD_QUEUE:
-			ctx->flags |= SEEN_SKB | SEEN_A;
-			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff,
-						  queue_mapping) != 2);
-			BUILD_BUG_ON(offsetof(struct sk_buff,
-					      queue_mapping) > 0xff);
-			off = offsetof(struct sk_buff, queue_mapping);
-			emit_half_load_unsigned(r_A, r_skb, off, ctx);
-			break;
-		default:
-			pr_debug("%s: Unhandled opcode: 0x%02x\n", __FILE__,
-				 inst->code);
-			return -1;
-		}
-	}
-
-	/* compute offsets only during the first pass */
-	if (ctx->target == NULL)
-		ctx->offsets[i] = ctx->idx * 4;
-
-	return 0;
-}
-
-void bpf_jit_compile(struct bpf_prog *fp)
-{
-	struct jit_ctx ctx;
-	unsigned int alloc_size, tmp_idx;
-
-	if (!bpf_jit_enable)
-		return;
-
-	memset(&ctx, 0, sizeof(ctx));
-
-	ctx.offsets = kcalloc(fp->len + 1, sizeof(*ctx.offsets), GFP_KERNEL);
-	if (ctx.offsets == NULL)
-		return;
-
-	ctx.skf = fp;
-
-	if (build_body(&ctx))
-		goto out;
-
-	tmp_idx = ctx.idx;
-	build_prologue(&ctx);
-	ctx.prologue_bytes = (ctx.idx - tmp_idx) * 4;
-	/* just to complete the ctx.idx count */
-	build_epilogue(&ctx);
-
-	alloc_size = 4 * ctx.idx;
-	ctx.target = module_alloc(alloc_size);
-	if (ctx.target == NULL)
-		goto out;
-
-	/* Clean it */
-	memset(ctx.target, 0, alloc_size);
-
-	ctx.idx = 0;
-
-	/* Generate the actual JIT code */
-	build_prologue(&ctx);
-	build_body(&ctx);
-	build_epilogue(&ctx);
-
-	/* Update the icache */
-	flush_icache_range((ptr)ctx.target, (ptr)(ctx.target + ctx.idx));
-
-	if (bpf_jit_enable > 1)
-		/* Dump JIT code */
-		bpf_jit_dump(fp->len, alloc_size, 2, ctx.target);
-
-	fp->bpf_func = (void *)ctx.target;
-	fp->jited = 1;
-
-out:
-	kfree(ctx.offsets);
-}
-
-void bpf_jit_free(struct bpf_prog *fp)
-{
-	if (fp->jited)
-		module_memfree(fp->bpf_func);
-
-	bpf_prog_unlock_free(fp);
-}

arch/mips/net/bpf_jit_asm.S

···
-/*
- * bpf_jib_asm.S: Packet/header access helper functions for MIPS/MIPS64 BPF
- * compiler.
- *
- * Copyright (C) 2015 Imagination Technologies Ltd.
- * Author: Markos Chandras <[e-mail]>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License as published by the
- * Free Software Foundation; version 2 of the License.
- */
-
-#include <asm/asm.h>
-#include <asm/isa-rev.h>
-#include <asm/regdef.h>
-#include "bpf_jit.h"
-
-/* ABI
- *
- * r_skb_hl	skb header length
- * r_skb_data	skb data
- * r_off(a1)	offset register
- * r_A		BPF register A
- * r_X		PF register X
- * r_skb(a0)	*skb
- * r_M		*scratch memory
- * r_skb_le	skb length
- * r_s0		Scratch register 0
- * r_s1		Scratch register 1
- *
- * On entry:
- * a0: *skb
- * a1: offset (imm or imm + X)
- *
- * All non-BPF-ABI registers are free for use. On return, we only
- * care about r_ret. The BPF-ABI registers are assumed to remain
- * unmodified during the entire filter operation.
- */
-
-#define skb	a0
-#define offset	a1
-#define SKF_LL_OFF  (-0x200000) /* Can't include linux/filter.h in assembly */
-
-	/* We know better :) so prevent assembler reordering etc */
-	.set 	noreorder
-
-#define is_offset_negative(TYPE)				\
-	/* If offset is negative we have more work to do */	\
-	slti	t0, offset, 0;					\
-	bgtz	t0, bpf_slow_path_##TYPE##_neg;			\
-	/* Be careful what follows in DS. */
-
-#define is_offset_in_header(SIZE, TYPE)				\
-	/* Reading from header? */				\
-	addiu	$r_s0, $r_skb_hl, -SIZE;			\
-	slt	t0, $r_s0, offset;				\
-	bgtz	t0, bpf_slow_path_##TYPE;			\
-
-LEAF(sk_load_word)
-	is_offset_negative(word)
-FEXPORT(sk_load_word_positive)
-	is_offset_in_header(4, word)
-	/* Offset within header boundaries */
-	PTR_ADDU t1, $r_skb_data, offset
-	.set	reorder
-	lw	$r_A, 0(t1)
-	.set	noreorder
-#ifdef CONFIG_CPU_LITTLE_ENDIAN
-# if MIPS_ISA_REV >= 2
-	wsbh	t0, $r_A
-	rotr	$r_A, t0, 16
-# else
-	sll	t0, $r_A, 24
-	srl	t1, $r_A, 24
-	srl	t2, $r_A, 8
-	or	t0, t0, t1
-	andi	t2, t2, 0xff00
-	andi	t1, $r_A, 0xff00
-	or	t0, t0, t2
-	sll	t1, t1, 8
-	or	$r_A, t0, t1
-# endif
-#endif
-	jr	$r_ra
-	 move	$r_ret, zero
-	END(sk_load_word)
-
-LEAF(sk_load_half)
-	is_offset_negative(half)
-FEXPORT(sk_load_half_positive)
-	is_offset_in_header(2, half)
-	/* Offset within header boundaries */
-	PTR_ADDU t1, $r_skb_data, offset
-	lhu	$r_A, 0(t1)
-#ifdef CONFIG_CPU_LITTLE_ENDIAN
-# if MIPS_ISA_REV >= 2
-	wsbh	$r_A, $r_A
-# else
-	sll	t0, $r_A, 8
-	srl	t1, $r_A, 8
-	andi	t0, t0, 0xff00
-	or	$r_A, t0, t1
-# endif
-#endif
-	jr	$r_ra
-	 move	$r_ret, zero
-	END(sk_load_half)
-
-LEAF(sk_load_byte)
-	is_offset_negative(byte)
-FEXPORT(sk_load_byte_positive)
-	is_offset_in_header(1, byte)
-	/* Offset within header boundaries */
-	PTR_ADDU t1, $r_skb_data, offset
-	lbu	$r_A, 0(t1)
-	jr	$r_ra
-	 move	$r_ret, zero
-	END(sk_load_byte)
-
-/*
- * call skb_copy_bits:
- * (prototype in linux/skbuff.h)
- *
- * int skb_copy_bits(sk_buff *skb, int offset, void *to, int len)
- *
- * o32 mandates we leave 4 spaces for argument registers in case
- * the callee needs to use them. Even though we don't care about
- * the argument registers ourselves, we need to allocate that space
- * to remain ABI compliant since the callee may want to use that space.
- * We also allocate 2 more spaces for $r_ra and our return register (*to).
- *
- * n64 is a bit different. The *caller* will allocate the space to preserve
- * the arguments. So in 64-bit kernels, we allocate the 4-arg space for no
- * good reason but it does not matter that much really.
- *
- * (void *to) is returned in r_s0
- *
- */
-#ifdef CONFIG_CPU_LITTLE_ENDIAN
-#define DS_OFFSET(SIZE) (4 * SZREG)
-#else
-#define DS_OFFSET(SIZE) ((4 * SZREG) + (4 - SIZE))
-#endif
-#define bpf_slow_path_common(SIZE)				\
-	/* Quick check. Are we within reasonable boundaries? */ \
-	LONG_ADDIU	$r_s1, $r_skb_len, -SIZE;		\
-	sltu		$r_s0, offset, $r_s1;			\
-	beqz		$r_s0, fault;				\
-	/* Load 4th argument in DS */				\
-	 LONG_ADDIU	a3, zero, SIZE;				\
-	PTR_ADDIU	$r_sp, $r_sp, -(6 * SZREG);		\
-	PTR_LA		t0, skb_copy_bits;			\
-	PTR_S		$r_ra, (5 * SZREG)($r_sp);		\
-	/* Assign low slot to a2 */				\
-	PTR_ADDIU	a2, $r_sp, DS_OFFSET(SIZE);		\
-	jalr		t0;					\
-	/* Reset our destination slot (DS but it's ok) */	\
-	 INT_S		zero, (4 * SZREG)($r_sp);		\
-	/*							\
-	 * skb_copy_bits returns 0 on success and -EFAULT	\
-	 * on error. Our data live in a2. Do not bother with	\
-	 * our data if an error has been returned.		\
-	 */							\
-	/* Restore our frame */					\
-	PTR_L		$r_ra, (5 * SZREG)($r_sp);		\
-	INT_L		$r_s0, (4 * SZREG)($r_sp);		\
-	bltz		v0, fault;				\
-	 PTR_ADDIU	$r_sp, $r_sp, 6 * SZREG;		\
-	move		$r_ret, zero;				\
-
-NESTED(bpf_slow_path_word, (6 * SZREG), $r_sp)
-	bpf_slow_path_common(4)
-#ifdef CONFIG_CPU_LITTLE_ENDIAN
-# if MIPS_ISA_REV >= 2
-	wsbh	t0, $r_s0
-	jr	$r_ra
-	 rotr	$r_A, t0, 16
-# else
-	sll	t0, $r_s0, 24
-	srl	t1, $r_s0, 24
-	srl	t2, $r_s0, 8
-	or	t0, t0, t1
-	andi	t2, t2, 0xff00
-	andi	t1, $r_s0, 0xff00
-	or	t0, t0, t2
-	sll	t1, t1, 8
-	jr	$r_ra
-	 or	$r_A, t0, t1
-# endif
-#else
-	jr	$r_ra
-	 move	$r_A, $r_s0
-#endif
-
-	END(bpf_slow_path_word)
-
-NESTED(bpf_slow_path_half, (6 * SZREG), $r_sp)
-	bpf_slow_path_common(2)
-#ifdef CONFIG_CPU_LITTLE_ENDIAN
-# if MIPS_ISA_REV >= 2
-	jr	$r_ra
-	 wsbh	$r_A, $r_s0
-# else
-	sll	t0, $r_s0, 8
-	andi	t1, $r_s0, 0xff00
-	andi	t0, t0, 0xff00
-	srl	t1, t1, 8
-	jr	$r_ra
-	 or	$r_A, t0, t1
-# endif
-#else
-	jr	$r_ra
-	 move	$r_A, $r_s0
-#endif
-
-	END(bpf_slow_path_half)
-
-NESTED(bpf_slow_path_byte, (6 * SZREG), $r_sp)
-	bpf_slow_path_common(1)
-	jr	$r_ra
-	 move	$r_A, $r_s0
-
-	END(bpf_slow_path_byte)
-
-/*
- * Negative entry points
- */
-	.macro bpf_is_end_of_data
-	li	t0, SKF_LL_OFF
-	/* Reading link layer data? */
-	slt	t1, offset, t0
-	bgtz	t1, fault
-	/* Be careful what follows in DS. */
-	.endm
-/*
- * call skb_copy_bits:
- * (prototype in linux/filter.h)
- *
- * void *bpf_internal_load_pointer_neg_helper(const struct sk_buff *skb,
- *                                            int k, unsigned int size)
- *
- * see above (bpf_slow_path_common) for ABI restrictions
- */
-#define bpf_negative_common(SIZE)					\
-	PTR_ADDIU	$r_sp, $r_sp, -(6 * SZREG);			\
-	PTR_LA		t0, bpf_internal_load_pointer_neg_helper;	\
-	PTR_S		$r_ra, (5 * SZREG)($r_sp);			\
-	jalr		t0;						\
-	 li		a2, SIZE;					\
-	PTR_L		$r_ra, (5 * SZREG)($r_sp);			\
-	/* Check return pointer */					\
-	beqz		v0, fault;					\
-	 PTR_ADDIU	$r_sp, $r_sp, 6 * SZREG;			\
-	/* Preserve our pointer */					\
-	move		$r_s0, v0;					\
-	/* Set return value */						\
-	move		$r_ret, zero;					\
-
-bpf_slow_path_word_neg:
-	bpf_is_end_of_data
-NESTED(sk_load_word_negative, (6 * SZREG), $r_sp)
-	bpf_negative_common(4)
-	jr	$r_ra
-	 lw	$r_A, 0($r_s0)
-	END(sk_load_word_negative)
-
-bpf_slow_path_half_neg:
-	bpf_is_end_of_data
-NESTED(sk_load_half_negative, (6 * SZREG), $r_sp)
-	bpf_negative_common(2)
-	jr	$r_ra
-	 lhu	$r_A, 0($r_s0)
-	END(sk_load_half_negative)
-
-bpf_slow_path_byte_neg:
-	bpf_is_end_of_data
-NESTED(sk_load_byte_negative, (6 * SZREG), $r_sp)
-	bpf_negative_common(1)
-	jr	$r_ra
-	 lbu	$r_A, 0($r_s0)
-	END(sk_load_byte_negative)
-
-fault:
-	jr	$r_ra
-	 addiu $r_ret, zero, 1

arch/mips/net/ebpf_jit.c

···
 #include <asm/byteorder.h>
 #include <asm/cacheflush.h>
 #include <asm/cpu-features.h>
+#include <asm/isa-rev.h>
 #include <asm/uasm.h>
 
 /* Registers used by JIT */
···
 }
 
 /* Simply emit the instruction if the JIT memory space has been allocated */
-#define emit_instr(ctx, func, ...)			\
-do {							\
-	if ((ctx)->target != NULL) {			\
-		u32 *p = &(ctx)->target[ctx->idx];	\
-		uasm_i_##func(&p, ##__VA_ARGS__);	\
-	}						\
-	(ctx)->idx++;					\
+#define emit_instr_long(ctx, func64, func32, ...)		\
+do {								\
+	if ((ctx)->target != NULL) {				\
+		u32 *p = &(ctx)->target[ctx->idx];		\
+		if (IS_ENABLED(CONFIG_64BIT))			\
+			uasm_i_##func64(&p, ##__VA_ARGS__);	\
+		else						\
+			uasm_i_##func32(&p, ##__VA_ARGS__);	\
+	}							\
+	(ctx)->idx++;						\
 } while (0)
+
+#define emit_instr(ctx, func, ...)				\
+	emit_instr_long(ctx, func, func, ##__VA_ARGS__)
 
 static unsigned int j_target(struct jit_ctx *ctx, int target_idx)
 {
···
 		 * If RA we are doing a function call and may need
 		 * extra 8-byte tmp area.
 		 */
-		stack_adjust += 16;
+		stack_adjust += 2 * sizeof(long);
 	if (ctx->flags & EBPF_SAVE_S0)
-		stack_adjust += 8;
+		stack_adjust += sizeof(long);
 	if (ctx->flags & EBPF_SAVE_S1)
-		stack_adjust += 8;
+		stack_adjust += sizeof(long);
 	if (ctx->flags & EBPF_SAVE_S2)
-		stack_adjust += 8;
+		stack_adjust += sizeof(long);
 	if (ctx->flags & EBPF_SAVE_S3)
-		stack_adjust += 8;
+		stack_adjust += sizeof(long);
 	if (ctx->flags & EBPF_SAVE_S4)
-		stack_adjust += 8;
+		stack_adjust += sizeof(long);
 
 	BUILD_BUG_ON(MAX_BPF_STACK & 7);
 	locals_size = (ctx->flags & EBPF_SEEN_FP) ? MAX_BPF_STACK : 0;
···
 	 * On tail call we skip this instruction, and the TCC is
 	 * passed in $v1 from the caller.
 	 */
-	emit_instr(ctx, daddiu, MIPS_R_V1, MIPS_R_ZERO, MAX_TAIL_CALL_CNT);
+	emit_instr(ctx, addiu, MIPS_R_V1, MIPS_R_ZERO, MAX_TAIL_CALL_CNT);
 	if (stack_adjust)
-		emit_instr(ctx, daddiu, MIPS_R_SP, MIPS_R_SP, -stack_adjust);
+		emit_instr_long(ctx, daddiu, addiu,
+					MIPS_R_SP, MIPS_R_SP, -stack_adjust);
 	else
 		return 0;
 
-	store_offset = stack_adjust - 8;
+	store_offset = stack_adjust - sizeof(long);
 
 	if (ctx->flags & EBPF_SAVE_RA) {
-		emit_instr(ctx, sd, MIPS_R_RA, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, sd, sw,
+					MIPS_R_RA, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S0) {
-		emit_instr(ctx, sd, MIPS_R_S0, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, sd, sw,
+					MIPS_R_S0, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S1) {
-		emit_instr(ctx, sd, MIPS_R_S1, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, sd, sw,
+					MIPS_R_S1, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S2) {
-		emit_instr(ctx, sd, MIPS_R_S2, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, sd, sw,
+					MIPS_R_S2, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S3) {
-		emit_instr(ctx, sd, MIPS_R_S3, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, sd, sw,
+					MIPS_R_S3, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S4) {
-		emit_instr(ctx, sd, MIPS_R_S4, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, sd, sw,
+					MIPS_R_S4, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 
 	if ((ctx->flags & EBPF_SEEN_TC) && !(ctx->flags & EBPF_TCC_IN_V1))
-		emit_instr(ctx, daddu, MIPS_R_S4, MIPS_R_V1, MIPS_R_ZERO);
+		emit_instr_long(ctx, daddu, addu,
+					MIPS_R_S4, MIPS_R_V1, MIPS_R_ZERO);
 
 	return 0;
 }
···
 {
 	const struct bpf_prog *prog = ctx->skf;
 	int stack_adjust = ctx->stack_size;
-	int store_offset = stack_adjust - 8;
+	int store_offset = stack_adjust - sizeof(long);
 	enum reg_val_type td;
 	int r0 = MIPS_R_V0;
 
···
 	}
 
 	if (ctx->flags & EBPF_SAVE_RA) {
-		emit_instr(ctx, ld, MIPS_R_RA, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, ld, lw,
+					MIPS_R_RA, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S0) {
-		emit_instr(ctx, ld, MIPS_R_S0, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, ld, lw,
+					MIPS_R_S0, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S1) {
-		emit_instr(ctx, ld, MIPS_R_S1, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, ld, lw,
+					MIPS_R_S1, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S2) {
-		emit_instr(ctx, ld, MIPS_R_S2, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, ld, lw,
+				MIPS_R_S2, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S3) {
-		emit_instr(ctx, ld, MIPS_R_S3, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, ld, lw,
+					MIPS_R_S3, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	if (ctx->flags & EBPF_SAVE_S4) {
-		emit_instr(ctx, ld, MIPS_R_S4, store_offset, MIPS_R_SP);
-		store_offset -= 8;
+		emit_instr_long(ctx, ld, lw,
+					MIPS_R_S4, store_offset, MIPS_R_SP);
+		store_offset -= sizeof(long);
 	}
 	emit_instr(ctx, jr, dest_reg);
 
 	if (stack_adjust)
-		emit_instr(ctx, daddiu, MIPS_R_SP, MIPS_R_SP, stack_adjust);
+		emit_instr_long(ctx, daddiu, addiu,
+					MIPS_R_SP, MIPS_R_SP, stack_adjust);
 	else
 		emit_instr(ctx, nop);
 
···
 	s64 t64s;
 	int bpf_op = BPF_OP(insn->code);
 
+	if (IS_ENABLED(CONFIG_32BIT) && ((BPF_CLASS(insn->code) == BPF_ALU64)
+						|| (bpf_op == BPF_DW)))
+		return -EINVAL;
+
 	switch (insn->code) {
 	case BPF_ALU64 | BPF_ADD | BPF_K: /* ALU64_IMM */
 	case BPF_ALU64 | BPF_SUB | BPF_K: /* ALU64_IMM */
···
 		if (insn->imm == 1) /* Mult by 1 is a nop */
 			break;
 		gen_imm_to_reg(insn, MIPS_R_AT, ctx);
-		emit_instr(ctx, dmultu, MIPS_R_AT, dst);
-		emit_instr(ctx, mflo, dst);
+		if (MIPS_ISA_REV >= 6) {
+			emit_instr(ctx, dmulu, dst, dst, MIPS_R_AT);
+		} else {
+			emit_instr(ctx, dmultu, MIPS_R_AT, dst);
+			emit_instr(ctx, mflo, dst);
+		}
 		break;
 	case BPF_ALU64 | BPF_NEG | BPF_K: /* ALU64_IMM */
 		dst = ebpf_to_mips_reg(ctx, insn, dst_reg);
···
 		if (insn->imm == 1) /* Mult by 1 is a nop */
 			break;
 		gen_imm_to_reg(insn, MIPS_R_AT, ctx);
-		emit_instr(ctx, multu, dst, MIPS_R_AT);
-		emit_instr(ctx, mflo, dst);
+		if (MIPS_ISA_REV >= 6) {
+			emit_instr(ctx, mulu, dst, dst, MIPS_R_AT);
+		} else {
+			emit_instr(ctx, multu, dst, MIPS_R_AT);
+			emit_instr(ctx, mflo, dst);
+		}
 		break;
 	case BPF_ALU | BPF_NEG | BPF_K: /* ALU_IMM */
 		dst = ebpf_to_mips_reg(ctx, insn, dst_reg);
···
 			break;
 		}
 		gen_imm_to_reg(insn, MIPS_R_AT, ctx);
+		if (MIPS_ISA_REV >= 6) {
+			if (bpf_op == BPF_DIV)
+				emit_instr(ctx, divu_r6, dst, dst, MIPS_R_AT);
+			else
+				emit_instr(ctx, modu, dst, dst, MIPS_R_AT);
+			break;
+		}
 		emit_instr(ctx, divu, dst, MIPS_R_AT);
 		if (bpf_op == BPF_DIV)
 			emit_instr(ctx, mflo, dst);
···
 			break;
 		}
 		gen_imm_to_reg(insn, MIPS_R_AT, ctx);
+		if (MIPS_ISA_REV >= 6) {
+			if (bpf_op == BPF_DIV)
+				emit_instr(ctx, ddivu_r6, dst, dst, MIPS_R_AT);
+			else
+				emit_instr(ctx, modu, dst, dst, MIPS_R_AT);
+			break;
+		}
 		emit_instr(ctx, ddivu, dst, MIPS_R_AT);
 		if (bpf_op == BPF_DIV)
 			emit_instr(ctx, mflo, dst);
···
 			emit_instr(ctx, and, dst, dst, src);
 			break;
 		case BPF_MUL:
-			emit_instr(ctx, dmultu, dst, src);
-			emit_instr(ctx, mflo, dst);
+			if (MIPS_ISA_REV >= 6) {
+				emit_instr(ctx, dmulu, dst, dst, src);
+			} else {
+				emit_instr(ctx, dmultu, dst, src);
+				emit_instr(ctx, mflo, dst);
+			}
 			break;
 		case BPF_DIV:
 		case BPF_MOD:
+			if (MIPS_ISA_REV >= 6) {
+				if (bpf_op == BPF_DIV)
+					emit_instr(ctx, ddivu_r6,
+							dst, dst, src);
+				else
+					emit_instr(ctx, modu, dst, dst, src);
+				break;
+			}
 			emit_instr(ctx, ddivu, dst, src);
 			if (bpf_op == BPF_DIV)
 				emit_instr(ctx, mflo, dst);
···
 			break;
 		case BPF_DIV:
 		case BPF_MOD:
+			if (MIPS_ISA_REV >= 6) {
+				if (bpf_op == BPF_DIV)
+					emit_instr(ctx, divu_r6, dst, dst, src);
+				else
+					emit_instr(ctx, modu, dst, dst, src);
+				break;
+			}
 			emit_instr(ctx, divu, dst, src);
 			if (bpf_op == BPF_DIV)
 				emit_instr(ctx, mflo, dst);
···
 			emit_instr(ctx, dsubu, MIPS_R_T8, dst, src);
 			emit_instr(ctx, sltu, MIPS_R_AT, dst, src);
 			/* SP known to be non-zero, movz becomes boolean not */
-			emit_instr(ctx, movz, MIPS_R_T9, MIPS_R_SP, MIPS_R_T8);
-			emit_instr(ctx, movn, MIPS_R_T9, MIPS_R_ZERO, MIPS_R_T8);
+			if (MIPS_ISA_REV >= 6) {
+				emit_instr(ctx, seleqz, MIPS_R_T9,
+						MIPS_R_SP, MIPS_R_T8);
+			} else {
+				emit_instr(ctx, movz, MIPS_R_T9,
+						MIPS_R_SP, MIPS_R_T8);
+				emit_instr(ctx, movn, MIPS_R_T9,
+						MIPS_R_ZERO, MIPS_R_T8);
+			}
 			emit_instr(ctx, or, MIPS_R_AT, MIPS_R_T9, MIPS_R_AT);
 			cmp_eq = bpf_op == BPF_JGT;
 			dst = MIPS_R_AT;
···
 
 	case BPF_JMP | BPF_CALL:
 		ctx->flags |= EBPF_SAVE_RA;
-		t64s = (s64)insn->imm + (s64)__bpf_call_base;
+		t64s = (s64)insn->imm + (long)__bpf_call_base;
 		emit_const_to_reg(ctx, MIPS_R_T9, (u64)t64s);
 		emit_instr(ctx, jalr, MIPS_R_RA, MIPS_R_T9);
 		/* delay slot */
···
 		if (src < 0)
 			return src;
 		if (BPF_MODE(insn->code) == BPF_XADD) {
+			/*
+			 * If mem_off does not fit within the 9 bit ll/sc
+			 * instruction immediate field, use a temp reg.
+			 */
+			if (MIPS_ISA_REV >= 6 &&
+			    (mem_off >= BIT(8) || mem_off < -BIT(8))) {
+				emit_instr(ctx, daddiu, MIPS_R_T6,
+						dst, mem_off);
+				mem_off = 0;
+				dst = MIPS_R_T6;
+			}
 			switch (BPF_SIZE(insn->code)) {
 			case BPF_W:
 				if (get_reg_val_type(ctx, this_idx, insn->src_reg) == REG_32BIT) {
···
 	unsigned int image_size;
 	u8 *image_ptr;
 
-	if (!prog->jit_requested || !cpu_has_mips64r2)
+	if (!prog->jit_requested || MIPS_ISA_REV < 2)
 		return prog;
 
 	tmp = bpf_jit_blind_constants(prog);

arch/mips/pic32/Kconfig

···
 	  Select the devicetree.
 
 config DTB_PIC32_NONE
-       bool "None"
+	bool "None"
 
 config DTB_PIC32_MZDA_SK
-       bool "PIC32MZDA Starter Kit"
-       depends on PIC32MZDA
-       select BUILTIN_DTB
+	bool "PIC32MZDA Starter Kit"
+	depends on PIC32MZDA
+	select BUILTIN_DTB
 
 endchoice
 

arch/mips/vdso/Makefile

···
 VDSO_LDFLAGS := \
 	-Wl,-Bsymbolic -Wl,--no-undefined -Wl,-soname=linux-vdso.so.1 \
 	$(addprefix -Wl$(comma),$(filter -E%,$(KBUILD_CFLAGS))) \
-	-nostdlib -shared \
-	$(call cc-ldoption, -Wl$(comma)--hash-style=sysv) \
-	$(call cc-ldoption, -Wl$(comma)--build-id)
+	-nostdlib -shared -Wl,--hash-style=sysv -Wl,--build-id
 
 GCOV_PROFILE := n
 UBSAN_SANITIZE := n