diff options
Diffstat (limited to 'arch/arm')
43 files changed, 988 insertions, 658 deletions
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 51c8df561077..fea5ae8ecd0d 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -3,8 +3,8 @@ config ARM bool default y select ARCH_CLOCKSOURCE_DATA - select ARCH_DISCARD_MEMBLOCK if !HAVE_ARCH_PFN_VALID - select ARCH_HAS_DEBUG_VIRTUAL + select ARCH_DISCARD_MEMBLOCK if !HAVE_ARCH_PFN_VALID && !KEXEC + select ARCH_HAS_DEBUG_VIRTUAL if MMU select ARCH_HAS_DEVMEM_IS_ALLOWED select ARCH_HAS_ELF_RANDOMIZE select ARCH_HAS_SET_MEMORY @@ -99,6 +99,7 @@ config ARM select OLD_SIGACTION select OLD_SIGSUSPEND3 select PERF_USE_VMALLOC + select REFCOUNT_FULL select RTC_LIB select SYS_SUPPORTS_APM_EMULATION # Above selects are sorted alphabetically; please add new ones @@ -1524,12 +1525,10 @@ config THUMB2_KERNEL bool "Compile the kernel in Thumb-2 mode" if !CPU_THUMBONLY depends on (CPU_V7 || CPU_V7M) && !CPU_V6 && !CPU_V6K default y if CPU_THUMBONLY - select ARM_ASM_UNIFIED select ARM_UNWIND help By enabling this option, the kernel will be compiled in - Thumb-2 mode. A compiler/assembler that understand the unified - ARM-Thumb syntax is needed. + Thumb-2 mode. If unsure, say N. @@ -1564,9 +1563,6 @@ config THUMB2_AVOID_R_ARM_THM_JUMP11 Unless you are sure your tools don't have this problem, say Y. -config ARM_ASM_UNIFIED - bool - config ARM_PATCH_IDIV bool "Runtime patch udiv/sdiv instructions into __aeabi_{u}idiv()" depends on CPU_32v7 && !XIP_KERNEL diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug index 17685e19aed8..78a647080ebc 100644 --- a/arch/arm/Kconfig.debug +++ b/arch/arm/Kconfig.debug @@ -3,10 +3,14 @@ menu "Kernel hacking" source "lib/Kconfig.debug" -config ARM_PTDUMP +config ARM_PTDUMP_CORE + def_bool n + +config ARM_PTDUMP_DEBUGFS bool "Export kernel pagetable layout to userspace via debugfs" depends on DEBUG_KERNEL depends on MMU + select ARM_PTDUMP_CORE select DEBUG_FS ---help--- Say Y here if you want to show the kernel pagetable layout in a @@ -16,6 +20,33 @@ config ARM_PTDUMP kernel. If in doubt, say "N" +config DEBUG_WX + bool "Warn on W+X mappings at boot" + select ARM_PTDUMP_CORE + ---help--- + Generate a warning if any W+X mappings are found at boot. + + This is useful for discovering cases where the kernel is leaving + W+X mappings after applying NX, as such mappings are a security risk. + + Look for a message in dmesg output like this: + + arm/mm: Checked W+X mappings: passed, no W+X pages found. + + or like this, if the check failed: + + arm/mm: Checked W+X mappings: FAILED, <N> W+X pages found. + + Note that even if the check fails, your kernel is possibly + still fine, as W+X mappings are not a security hole in + themselves, what they do is that they make the exploitation + of other unfixed kernel bugs easier. + + There is no runtime or memory usage effect of this option + once the kernel has booted up - it's a one time check. + + If in doubt, say "Y". + # RMK wants arm kernels compiled with frame pointers or stack unwinding. # If you know what you are doing and are willing to live without stack # traces, you can get a slightly smaller kernel by setting this option to diff --git a/arch/arm/Makefile b/arch/arm/Makefile index 80351e505fd5..e83f5161fdd8 100644 --- a/arch/arm/Makefile +++ b/arch/arm/Makefile @@ -115,9 +115,11 @@ ifeq ($(CONFIG_ARM_UNWIND),y) CFLAGS_ABI +=-funwind-tables endif +# Accept old syntax despite ".syntax unified" +AFLAGS_NOWARN :=$(call as-option,-Wa$(comma)-mno-warn-deprecated,-Wa$(comma)-W) + ifeq ($(CONFIG_THUMB2_KERNEL),y) AFLAGS_AUTOIT :=$(call as-option,-Wa$(comma)-mimplicit-it=always,-Wa$(comma)-mauto-it) -AFLAGS_NOWARN :=$(call as-option,-Wa$(comma)-mno-warn-deprecated,-Wa$(comma)-W) CFLAGS_ISA :=-mthumb $(AFLAGS_AUTOIT) $(AFLAGS_NOWARN) AFLAGS_ISA :=$(CFLAGS_ISA) -Wa$(comma)-mthumb # Work around buggy relocation from gas if requested: @@ -125,7 +127,7 @@ ifeq ($(CONFIG_THUMB2_AVOID_R_ARM_THM_JUMP11),y) KBUILD_CFLAGS_MODULE +=-fno-optimize-sibling-calls endif else -CFLAGS_ISA :=$(call cc-option,-marm,) +CFLAGS_ISA :=$(call cc-option,-marm,) $(AFLAGS_NOWARN) AFLAGS_ISA :=$(CFLAGS_ISA) endif diff --git a/arch/arm/boot/compressed/string.c b/arch/arm/boot/compressed/string.c index 309e1bbad75d..13c90abc68d6 100644 --- a/arch/arm/boot/compressed/string.c +++ b/arch/arm/boot/compressed/string.c @@ -130,8 +130,3 @@ void *memset(void *s, int c, size_t count) *xs++ = c; return s; } - -void __memzero(void *s, size_t count) -{ - memset(s, 0, count); -} diff --git a/arch/arm/boot/compressed/vmlinux.lds.S b/arch/arm/boot/compressed/vmlinux.lds.S index e6bf6774c4bb..2b963d8e76dd 100644 --- a/arch/arm/boot/compressed/vmlinux.lds.S +++ b/arch/arm/boot/compressed/vmlinux.lds.S @@ -56,6 +56,7 @@ SECTIONS .rodata : { *(.rodata) *(.rodata.*) + *(.data.rel.ro) } .piggydata : { *(.piggydata) @@ -101,6 +102,12 @@ SECTIONS * this symbol allows further debug in the near future. */ .image_end (NOLOAD) : { + /* + * EFI requires that the image is aligned to 512 bytes, and appended + * DTB requires that we know where the end of the image is. Ensure + * that both are satisfied by ensuring that there are no additional + * sections emitted into the decompressor image. + */ _edata_real = .; } @@ -128,3 +135,4 @@ SECTIONS .stab.indexstr 0 : { *(.stab.indexstr) } .comment 0 : { *(.comment) } } +ASSERT(_edata_real == _edata, "error: zImage file size is incorrect"); diff --git a/arch/arm/common/bL_switcher_dummy_if.c b/arch/arm/common/bL_switcher_dummy_if.c index 4c10c6452678..f4dc1714a79e 100644 --- a/arch/arm/common/bL_switcher_dummy_if.c +++ b/arch/arm/common/bL_switcher_dummy_if.c @@ -57,3 +57,7 @@ static struct miscdevice bL_switcher_device = { &bL_switcher_fops }; module_misc_device(bL_switcher_device); + +MODULE_AUTHOR("Nicolas Pitre <nico@linaro.org>"); +MODULE_LICENSE("GPL v2"); +MODULE_DESCRIPTION("big.LITTLE switcher dummy user interface"); diff --git a/arch/arm/common/sa1111.c b/arch/arm/common/sa1111.c index 4ecd5120fce7..a2c878769eaf 100644 --- a/arch/arm/common/sa1111.c +++ b/arch/arm/common/sa1111.c @@ -108,6 +108,7 @@ struct sa1111 { spinlock_t lock; void __iomem *base; struct sa1111_platform_data *pdata; + struct irq_domain *irqdomain; struct gpio_chip gc; #ifdef CONFIG_PM void *saved_state; @@ -125,7 +126,7 @@ struct sa1111_dev_info { unsigned long skpcr_mask; bool dma; unsigned int devid; - unsigned int irq[6]; + unsigned int hwirq[6]; }; static struct sa1111_dev_info sa1111_devices[] = { @@ -134,7 +135,7 @@ static struct sa1111_dev_info sa1111_devices[] = { .skpcr_mask = SKPCR_UCLKEN, .dma = true, .devid = SA1111_DEVID_USB, - .irq = { + .hwirq = { IRQ_USBPWR, IRQ_HCIM, IRQ_HCIBUFFACC, @@ -148,7 +149,7 @@ static struct sa1111_dev_info sa1111_devices[] = { .skpcr_mask = SKPCR_I2SCLKEN | SKPCR_L3CLKEN, .dma = true, .devid = SA1111_DEVID_SAC, - .irq = { + .hwirq = { AUDXMTDMADONEA, AUDXMTDMADONEB, AUDRCVDMADONEA, @@ -164,7 +165,7 @@ static struct sa1111_dev_info sa1111_devices[] = { .offset = SA1111_KBD, .skpcr_mask = SKPCR_PTCLKEN, .devid = SA1111_DEVID_PS2_KBD, - .irq = { + .hwirq = { IRQ_TPRXINT, IRQ_TPTXINT }, @@ -173,7 +174,7 @@ static struct sa1111_dev_info sa1111_devices[] = { .offset = SA1111_MSE, .skpcr_mask = SKPCR_PMCLKEN, .devid = SA1111_DEVID_PS2_MSE, - .irq = { + .hwirq = { IRQ_MSRXINT, IRQ_MSTXINT }, @@ -182,7 +183,7 @@ static struct sa1111_dev_info sa1111_devices[] = { .offset = 0x1800, .skpcr_mask = 0, .devid = SA1111_DEVID_PCMCIA, - .irq = { + .hwirq = { IRQ_S0_READY_NINT, IRQ_S0_CD_VALID, IRQ_S0_BVD1_STSCHG, @@ -193,6 +194,19 @@ static struct sa1111_dev_info sa1111_devices[] = { }, }; +static int sa1111_map_irq(struct sa1111 *sachip, irq_hw_number_t hwirq) +{ + return irq_create_mapping(sachip->irqdomain, hwirq); +} + +static void sa1111_handle_irqdomain(struct irq_domain *irqdomain, int irq) +{ + struct irq_desc *d = irq_to_desc(irq_linear_revmap(irqdomain, irq)); + + if (d) + generic_handle_irq_desc(d); +} + /* * SA1111 interrupt support. Since clearing an IRQ while there are * active IRQs causes the interrupt output to pulse, the upper levels @@ -202,49 +216,45 @@ static void sa1111_irq_handler(struct irq_desc *desc) { unsigned int stat0, stat1, i; struct sa1111 *sachip = irq_desc_get_handler_data(desc); + struct irq_domain *irqdomain; void __iomem *mapbase = sachip->base + SA1111_INTC; - stat0 = sa1111_readl(mapbase + SA1111_INTSTATCLR0); - stat1 = sa1111_readl(mapbase + SA1111_INTSTATCLR1); + stat0 = readl_relaxed(mapbase + SA1111_INTSTATCLR0); + stat1 = readl_relaxed(mapbase + SA1111_INTSTATCLR1); - sa1111_writel(stat0, mapbase + SA1111_INTSTATCLR0); + writel_relaxed(stat0, mapbase + SA1111_INTSTATCLR0); desc->irq_data.chip->irq_ack(&desc->irq_data); - sa1111_writel(stat1, mapbase + SA1111_INTSTATCLR1); + writel_relaxed(stat1, mapbase + SA1111_INTSTATCLR1); if (stat0 == 0 && stat1 == 0) { do_bad_IRQ(desc); return; } + irqdomain = sachip->irqdomain; + for (i = 0; stat0; i++, stat0 >>= 1) if (stat0 & 1) - generic_handle_irq(i + sachip->irq_base); + sa1111_handle_irqdomain(irqdomain, i); for (i = 32; stat1; i++, stat1 >>= 1) if (stat1 & 1) - generic_handle_irq(i + sachip->irq_base); + sa1111_handle_irqdomain(irqdomain, i); /* For level-based interrupts */ desc->irq_data.chip->irq_unmask(&desc->irq_data); } -#define SA1111_IRQMASK_LO(x) (1 << (x - sachip->irq_base)) -#define SA1111_IRQMASK_HI(x) (1 << (x - sachip->irq_base - 32)) - static u32 sa1111_irqmask(struct irq_data *d) { - struct sa1111 *sachip = irq_data_get_irq_chip_data(d); - - return BIT((d->irq - sachip->irq_base) & 31); + return BIT(irqd_to_hwirq(d) & 31); } static int sa1111_irqbank(struct irq_data *d) { - struct sa1111 *sachip = irq_data_get_irq_chip_data(d); - - return ((d->irq - sachip->irq_base) / 32) * 4; + return (irqd_to_hwirq(d) / 32) * 4; } static void sa1111_ack_irq(struct irq_data *d) @@ -257,9 +267,9 @@ static void sa1111_mask_irq(struct irq_data *d) void __iomem *mapbase = sachip->base + SA1111_INTC + sa1111_irqbank(d); u32 ie; - ie = sa1111_readl(mapbase + SA1111_INTEN0); + ie = readl_relaxed(mapbase + SA1111_INTEN0); ie &= ~sa1111_irqmask(d); - sa1111_writel(ie, mapbase + SA1111_INTEN0); + writel(ie, mapbase + SA1111_INTEN0); } static void sa1111_unmask_irq(struct irq_data *d) @@ -268,9 +278,9 @@ static void sa1111_unmask_irq(struct irq_data *d) void __iomem *mapbase = sachip->base + SA1111_INTC + sa1111_irqbank(d); u32 ie; - ie = sa1111_readl(mapbase + SA1111_INTEN0); + ie = readl_relaxed(mapbase + SA1111_INTEN0); ie |= sa1111_irqmask(d); - sa1111_writel(ie, mapbase + SA1111_INTEN0); + writel_relaxed(ie, mapbase + SA1111_INTEN0); } /* @@ -287,11 +297,11 @@ static int sa1111_retrigger_irq(struct irq_data *d) u32 ip, mask = sa1111_irqmask(d); int i; - ip = sa1111_readl(mapbase + SA1111_INTPOL0); + ip = readl_relaxed(mapbase + SA1111_INTPOL0); for (i = 0; i < 8; i++) { - sa1111_writel(ip ^ mask, mapbase + SA1111_INTPOL0); - sa1111_writel(ip, mapbase + SA1111_INTPOL0); - if (sa1111_readl(mapbase + SA1111_INTSTATCLR0) & mask) + writel_relaxed(ip ^ mask, mapbase + SA1111_INTPOL0); + writel_relaxed(ip, mapbase + SA1111_INTPOL0); + if (readl_relaxed(mapbase + SA1111_INTSTATCLR0) & mask) break; } @@ -313,13 +323,13 @@ static int sa1111_type_irq(struct irq_data *d, unsigned int flags) if ((!(flags & IRQ_TYPE_EDGE_RISING) ^ !(flags & IRQ_TYPE_EDGE_FALLING)) == 0) return -EINVAL; - ip = sa1111_readl(mapbase + SA1111_INTPOL0); + ip = readl_relaxed(mapbase + SA1111_INTPOL0); if (flags & IRQ_TYPE_EDGE_RISING) ip &= ~mask; else ip |= mask; - sa1111_writel(ip, mapbase + SA1111_INTPOL0); - sa1111_writel(ip, mapbase + SA1111_WAKEPOL0); + writel_relaxed(ip, mapbase + SA1111_INTPOL0); + writel_relaxed(ip, mapbase + SA1111_WAKEPOL0); return 0; } @@ -330,12 +340,12 @@ static int sa1111_wake_irq(struct irq_data *d, unsigned int on) void __iomem *mapbase = sachip->base + SA1111_INTC + sa1111_irqbank(d); u32 we, mask = sa1111_irqmask(d); - we = sa1111_readl(mapbase + SA1111_WAKEEN0); + we = readl_relaxed(mapbase + SA1111_WAKEEN0); if (on) we |= mask; else we &= ~mask; - sa1111_writel(we, mapbase + SA1111_WAKEEN0); + writel_relaxed(we, mapbase + SA1111_WAKEEN0); return 0; } @@ -350,10 +360,30 @@ static struct irq_chip sa1111_irq_chip = { .irq_set_wake = sa1111_wake_irq, }; +static int sa1111_irqdomain_map(struct irq_domain *d, unsigned int irq, + irq_hw_number_t hwirq) +{ + struct sa1111 *sachip = d->host_data; + + /* Disallow unavailable interrupts */ + if (hwirq > SSPROR && hwirq < AUDXMTDMADONEA) + return -EINVAL; + + irq_set_chip_data(irq, sachip); + irq_set_chip_and_handler(irq, &sa1111_irq_chip, handle_edge_irq); + irq_clear_status_flags(irq, IRQ_NOREQUEST | IRQ_NOPROBE); + + return 0; +} + +static const struct irq_domain_ops sa1111_irqdomain_ops = { + .map = sa1111_irqdomain_map, + .xlate = irq_domain_xlate_twocell, +}; + static int sa1111_setup_irq(struct sa1111 *sachip, unsigned irq_base) { void __iomem *irqbase = sachip->base + SA1111_INTC; - unsigned i, irq; int ret; /* @@ -373,38 +403,40 @@ static int sa1111_setup_irq(struct sa1111 *sachip, unsigned irq_base) sachip->irq_base = ret; /* disable all IRQs */ - sa1111_writel(0, irqbase + SA1111_INTEN0); - sa1111_writel(0, irqbase + SA1111_INTEN1); - sa1111_writel(0, irqbase + SA1111_WAKEEN0); - sa1111_writel(0, irqbase + SA1111_WAKEEN1); + writel_relaxed(0, irqbase + SA1111_INTEN0); + writel_relaxed(0, irqbase + SA1111_INTEN1); + writel_relaxed(0, irqbase + SA1111_WAKEEN0); + writel_relaxed(0, irqbase + SA1111_WAKEEN1); /* * detect on rising edge. Note: Feb 2001 Errata for SA1111 * specifies that S0ReadyInt and S1ReadyInt should be '1'. */ - sa1111_writel(0, irqbase + SA1111_INTPOL0); - sa1111_writel(BIT(IRQ_S0_READY_NINT & 31) | - BIT(IRQ_S1_READY_NINT & 31), - irqbase + SA1111_INTPOL1); + writel_relaxed(0, irqbase + SA1111_INTPOL0); + writel_relaxed(BIT(IRQ_S0_READY_NINT & 31) | + BIT(IRQ_S1_READY_NINT & 31), + irqbase + SA1111_INTPOL1); /* clear all IRQs */ - sa1111_writel(~0, irqbase + SA1111_INTSTATCLR0); - sa1111_writel(~0, irqbase + SA1111_INTSTATCLR1); - - for (i = IRQ_GPAIN0; i <= SSPROR; i++) { - irq = sachip->irq_base + i; - irq_set_chip_and_handler(irq, &sa1111_irq_chip, handle_edge_irq); - irq_set_chip_data(irq, sachip); - irq_clear_status_flags(irq, IRQ_NOREQUEST | IRQ_NOPROBE); - } + writel_relaxed(~0, irqbase + SA1111_INTSTATCLR0); + writel_relaxed(~0, irqbase + SA1111_INTSTATCLR1); - for (i = AUDXMTDMADONEA; i <= IRQ_S1_BVD1_STSCHG; i++) { - irq = sachip->irq_base + i; - irq_set_chip_and_handler(irq, &sa1111_irq_chip, handle_edge_irq); - irq_set_chip_data(irq, sachip); - irq_clear_status_flags(irq, IRQ_NOREQUEST | IRQ_NOPROBE); + sachip->irqdomain = irq_domain_add_linear(NULL, SA1111_IRQ_NR, + &sa1111_irqdomain_ops, + sachip); + if (!sachip->irqdomain) { + irq_free_descs(sachip->irq_base, SA1111_IRQ_NR); + return -ENOMEM; } + irq_domain_associate_many(sachip->irqdomain, + sachip->irq_base + IRQ_GPAIN0, + IRQ_GPAIN0, SSPROR + 1 - IRQ_GPAIN0); + irq_domain_associate_many(sachip->irqdomain, + sachip->irq_base + AUDXMTDMADONEA, + AUDXMTDMADONEA, + IRQ_S1_BVD1_STSCHG + 1 - AUDXMTDMADONEA); + /* * Register SA1111 interrupt */ @@ -420,20 +452,22 @@ static int sa1111_setup_irq(struct sa1111 *sachip, unsigned irq_base) static void sa1111_remove_irq(struct sa1111 *sachip) { + struct irq_domain *domain = sachip->irqdomain; void __iomem *irqbase = sachip->base + SA1111_INTC; + int i; /* disable all IRQs */ - sa1111_writel(0, irqbase + SA1111_INTEN0); - sa1111_writel(0, irqbase + SA1111_INTEN1); - sa1111_writel(0, irqbase + SA1111_WAKEEN0); - sa1111_writel(0, irqbase + SA1111_WAKEEN1); + writel_relaxed(0, irqbase + SA1111_INTEN0); + writel_relaxed(0, irqbase + SA1111_INTEN1); + writel_relaxed(0, irqbase + SA1111_WAKEEN0); + writel_relaxed(0, irqbase + SA1111_WAKEEN1); - if (sachip->irq != NO_IRQ) { - irq_set_chained_handler_and_data(sachip->irq, NULL, NULL); - irq_free_descs(sachip->irq_base, SA1111_IRQ_NR); + irq_set_chained_handler_and_data(sachip->irq, NULL, NULL); + for (i = 0; i < SA1111_IRQ_NR; i++) + irq_dispose_mapping(irq_find_mapping(domain, i)); + irq_domain_remove(domain); - release_mem_region(sachip->phys + SA1111_INTC, 512); - } + release_mem_region(sachip->phys + SA1111_INTC, 512); } enum { @@ -572,7 +606,7 @@ static int sa1111_gpio_to_irq(struct gpio_chip *gc, unsigned offset) { struct sa1111 *sachip = gc_to_sa1111(gc); - return sachip->irq_base + offset; + return sa1111_map_irq(sachip, offset); } static int sa1111_setup_gpios(struct sa1111 *sachip) @@ -618,11 +652,11 @@ static void sa1111_wake(struct sa1111 *sachip) /* * Turn VCO on, and disable PLL Bypass. */ - r = sa1111_readl(sachip->base + SA1111_SKCR); + r = readl_relaxed(sachip->base + SA1111_SKCR); r &= ~SKCR_VCO_OFF; - sa1111_writel(r, sachip->base + SA1111_SKCR); + writel_relaxed(r, sachip->base + SA1111_SKCR); r |= SKCR_PLL_BYPASS | SKCR_OE_EN; - sa1111_writel(r, sachip->base + SA1111_SKCR); + writel_relaxed(r, sachip->base + SA1111_SKCR); /* * Wait lock time. SA1111 manual _doesn't_ @@ -634,7 +668,7 @@ static void sa1111_wake(struct sa1111 *sachip) * Enable RCLK. We also ensure that RDYEN is set. */ r |= SKCR_RCLKEN | SKCR_RDYEN; - sa1111_writel(r, sachip->base + SA1111_SKCR); + writel_relaxed(r, sachip->base + SA1111_SKCR); /* * Wait 14 RCLK cycles for the chip to finish coming out @@ -645,7 +679,7 @@ static void sa1111_wake(struct sa1111 *sachip) /* * Ensure all clocks are initially off. */ - sa1111_writel(0, sachip->base + SA1111_SKPCR); + writel_relaxed(0, sachip->base + SA1111_SKPCR); spin_unlock_irqrestore(&sachip->lock, flags); } @@ -675,7 +709,7 @@ sa1111_configure_smc(struct sa1111 *sachip, int sdram, unsigned int drac, if (cas_latency == 3) smcr |= SMCR_CLAT; - sa1111_writel(smcr, sachip->base + SA1111_SMCR); + writel_relaxed(smcr, sachip->base + SA1111_SMCR); /* * Now clear the bits in the DMA mask to work around the SA1111 @@ -723,8 +757,8 @@ sa1111_init_one_child(struct sa1111 *sachip, struct resource *parent, dev->mapbase = sachip->base + info->offset; dev->skpcr_mask = info->skpcr_mask; - for (i = 0; i < ARRAY_SIZE(info->irq); i++) - dev->irq[i] = sachip->irq_base + info->irq[i]; + for (i = 0; i < ARRAY_SIZE(info->hwirq); i++) + dev->hwirq[i] = info->hwirq[i]; /* * If the parent device has a DMA mask associated with it, and @@ -814,7 +848,7 @@ static int __sa1111_probe(struct device *me, struct resource *mem, int irq) /* * Probe for the chip. Only touch the SBI registers. */ - id = sa1111_readl(sachip->base + SA1111_SKID); + id = readl_relaxed(sachip->base + SA1111_SKID); if ((id & SKID_ID_MASK) != SKID_SA1111_ID) { printk(KERN_DEBUG "SA1111 not detected: ID = %08lx\n", id); ret = -ENODEV; @@ -833,11 +867,9 @@ static int __sa1111_probe(struct device *me, struct resource *mem, int irq) * The interrupt controller must be initialised before any * other device to ensure that the interrupts are available. */ - if (sachip->irq != NO_IRQ) { - ret = sa1111_setup_irq(sachip, pd->irq_base); - if (ret) - goto err_clk; - } + ret = sa1111_setup_irq(sachip, pd->irq_base); + if (ret) + goto err_clk; /* Setup the GPIOs - should really be done after the IRQ setup */ ret = sa1111_setup_gpios(sachip); @@ -864,8 +896,8 @@ static int __sa1111_probe(struct device *me, struct resource *mem, int irq) * DMA. It can otherwise be held firmly in the off position. * (currently, we always enable it.) */ - val = sa1111_readl(sachip->base + SA1111_SKPCR); - sa1111_writel(val | SKPCR_DCLKEN, sachip->base + SA1111_SKPCR); + val = readl_relaxed(sachip->base + SA1111_SKPCR); + writel_relaxed(val | SKPCR_DCLKEN, sachip->base + SA1111_SKPCR); /* * Enable the SA1110 memory bus request and grant signals. @@ -962,31 +994,31 @@ static int sa1111_suspend_noirq(struct device *dev) * Save state. */ base = sachip->base; - save->skcr = sa1111_readl(base + SA1111_SKCR); - save->skpcr = sa1111_readl(base + SA1111_SKPCR); - save->skcdr = sa1111_readl(base + SA1111_SKCDR); - save->skaud = sa1111_readl(base + SA1111_SKAUD); - save->skpwm0 = sa1111_readl(base + SA1111_SKPWM0); - save->skpwm1 = sa1111_readl(base + SA1111_SKPWM1); + save->skcr = readl_relaxed(base + SA1111_SKCR); + save->skpcr = readl_relaxed(base + SA1111_SKPCR); + save->skcdr = readl_relaxed(base + SA1111_SKCDR); + save->skaud = readl_relaxed(base + SA1111_SKAUD); + save->skpwm0 = readl_relaxed(base + SA1111_SKPWM0); + save->skpwm1 = readl_relaxed(base + SA1111_SKPWM1); - sa1111_writel(0, sachip->base + SA1111_SKPWM0); - sa1111_writel(0, sachip->base + SA1111_SKPWM1); + writel_relaxed(0, sachip->base + SA1111_SKPWM0); + writel_relaxed(0, sachip->base + SA1111_SKPWM1); base = sachip->base + SA1111_INTC; - save->intpol0 = sa1111_readl(base + SA1111_INTPOL0); - save->intpol1 = sa1111_readl(base + SA1111_INTPOL1); - save->inten0 = sa1111_readl(base + SA1111_INTEN0); - save->inten1 = sa1111_readl(base + SA1111_INTEN1); - save->wakepol0 = sa1111_readl(base + SA1111_WAKEPOL0); - save->wakepol1 = sa1111_readl(base + SA1111_WAKEPOL1); - save->wakeen0 = sa1111_readl(base + SA1111_WAKEEN0); - save->wakeen1 = sa1111_readl(base + SA1111_WAKEEN1); + save->intpol0 = readl_relaxed(base + SA1111_INTPOL0); + save->intpol1 = readl_relaxed(base + SA1111_INTPOL1); + save->inten0 = readl_relaxed(base + SA1111_INTEN0); + save->inten1 = readl_relaxed(base + SA1111_INTEN1); + save->wakepol0 = readl_relaxed(base + SA1111_WAKEPOL0); + save->wakepol1 = readl_relaxed(base + SA1111_WAKEPOL1); + save->wakeen0 = readl_relaxed(base + SA1111_WAKEEN0); + save->wakeen1 = readl_relaxed(base + SA1111_WAKEEN1); /* * Disable. */ - val = sa1111_readl(sachip->base + SA1111_SKCR); - sa1111_writel(val | SKCR_SLEEP, sachip->base + SA1111_SKCR); + val = readl_relaxed(sachip->base + SA1111_SKCR); + writel_relaxed(val | SKCR_SLEEP, sachip->base + SA1111_SKCR); clk_disable(sachip->clk); @@ -1023,7 +1055,7 @@ static int sa1111_resume_noirq(struct device *dev) * Ensure that the SA1111 is still here. * FIXME: shouldn't do this here. */ - id = sa1111_readl(sachip->base + SA1111_SKID); + id = readl_relaxed(sachip->base + SA1111_SKID); if ((id & SKID_ID_MASK) != SKID_SA1111_ID) { __sa1111_remove(sachip); dev_set_drvdata(dev, NULL); @@ -1047,26 +1079,26 @@ static int sa1111_resume_noirq(struct device *dev) */ spin_lock_irqsave(&sachip->lock, flags); - sa1111_writel(0, sachip->base + SA1111_INTC + SA1111_INTEN0); - sa1111_writel(0, sachip->base + SA1111_INTC + SA1111_INTEN1); + writel_relaxed(0, sachip->base + SA1111_INTC + SA1111_INTEN0); + writel_relaxed(0, sachip->base + SA1111_INTC + SA1111_INTEN1); base = sachip->base; - sa1111_writel(save->skcr, base + SA1111_SKCR); - sa1111_writel(save->skpcr, base + SA1111_SKPCR); - sa1111_writel(save->skcdr, base + SA1111_SKCDR); - sa1111_writel(save->skaud, base + SA1111_SKAUD); - sa1111_writel(save->skpwm0, base + SA1111_SKPWM0); - sa1111_writel(save->skpwm1, base + SA1111_SKPWM1); + writel_relaxed(save->skcr, base + SA1111_SKCR); + writel_relaxed(save->skpcr, base + SA1111_SKPCR); + writel_relaxed(save->skcdr, base + SA1111_SKCDR); + writel_relaxed(save->skaud, base + SA1111_SKAUD); + writel_relaxed(save->skpwm0, base + SA1111_SKPWM0); + writel_relaxed(save->skpwm1, base + SA1111_SKPWM1); base = sachip->base + SA1111_INTC; - sa1111_writel(save->intpol0, base + SA1111_INTPOL0); - sa1111_writel(save->intpol1, base + SA1111_INTPOL1); - sa1111_writel(save->inten0, base + SA1111_INTEN0); - sa1111_writel(save->inten1, base + SA1111_INTEN1); - sa1111_writel(save->wakepol0, base + SA1111_WAKEPOL0); - sa1111_writel(save->wakepol1, base + SA1111_WAKEPOL1); - sa1111_writel(save->wakeen0, base + SA1111_WAKEEN0); - sa1111_writel(save->wakeen1, base + SA1111_WAKEEN1); + writel_relaxed(save->intpol0, base + SA1111_INTPOL0); + writel_relaxed(save->intpol1, base + SA1111_INTPOL1); + writel_relaxed(save->inten0, base + SA1111_INTEN0); + writel_relaxed(save->inten1, base + SA1111_INTEN1); + writel_relaxed(save->wakepol0, base + SA1111_WAKEPOL0); + writel_relaxed(save->wakepol1, base + SA1111_WAKEPOL1); + writel_relaxed(save->wakeen0, base + SA1111_WAKEEN0); + writel_relaxed(save->wakeen1, base + SA1111_WAKEEN1); spin_unlock_irqrestore(&sachip->lock, flags); @@ -1153,7 +1185,7 @@ static unsigned int __sa1111_pll_clock(struct sa1111 *sachip) { unsigned int skcdr, fbdiv, ipdiv, opdiv; - skcdr = sa1111_readl(sachip->base + SA1111_SKCDR); + skcdr = readl_relaxed(sachip->base + SA1111_SKCDR); fbdiv = (skcdr & 0x007f) + 2; ipdiv = ((skcdr & 0x0f80) >> 7) + 2; @@ -1195,13 +1227,13 @@ void sa1111_select_audio_mode(struct sa1111_dev *sadev, int mode) spin_lock_irqsave(&sachip->lock, flags); - val = sa1111_readl(sachip->base + SA1111_SKCR); + val = readl_relaxed(sachip->base + SA1111_SKCR); if (mode == SA1111_AUDIO_I2S) { val &= ~SKCR_SELAC; } else { val |= SKCR_SELAC; } - sa1111_writel(val, sachip->base + SA1111_SKCR); + writel_relaxed(val, sachip->base + SA1111_SKCR); spin_unlock_irqrestore(&sachip->lock, flags); } @@ -1226,7 +1258,7 @@ int sa1111_set_audio_rate(struct sa1111_dev *sadev, int rate) if (div > 128) div = 128; - sa1111_writel(div - 1, sachip->base + SA1111_SKAUD); + writel_relaxed(div - 1, sachip->base + SA1111_SKAUD); return 0; } @@ -1244,7 +1276,7 @@ int sa1111_get_audio_rate(struct sa1111_dev *sadev) if (sadev->devid != SA1111_DEVID_SAC) return -EINVAL; - div = sa1111_readl(sachip->base + SA1111_SKAUD) + 1; + div = readl_relaxed(sachip->base + SA1111_SKAUD) + 1; return __sa1111_pll_clock(sachip) / (256 * div); } @@ -1261,10 +1293,10 @@ void sa1111_set_io_dir(struct sa1111_dev *sadev, #define MODIFY_BITS(port, mask, dir) \ if (mask) { \ - val = sa1111_readl(port); \ + val = readl_relaxed(port); \ val &= ~(mask); \ val |= (dir) & (mask); \ - sa1111_writel(val, port); \ + writel_relaxed(val, port); \ } spin_lock_irqsave(&sachip->lock, flags); @@ -1329,8 +1361,8 @@ int sa1111_enable_device(struct sa1111_dev *sadev) if (ret == 0) { spin_lock_irqsave(&sachip->lock, flags); - val = sa1111_readl(sachip->base + SA1111_SKPCR); - sa1111_writel(val | sadev->skpcr_mask, sachip->base + SA1111_SKPCR); + val = readl_relaxed(sachip->base + SA1111_SKPCR); + writel_relaxed(val | sadev->skpcr_mask, sachip->base + SA1111_SKPCR); spin_unlock_irqrestore(&sachip->lock, flags); } return ret; @@ -1348,8 +1380,8 @@ void sa1111_disable_device(struct sa1111_dev *sadev) unsigned int val; spin_lock_irqsave(&sachip->lock, flags); - val = sa1111_readl(sachip->base + SA1111_SKPCR); - sa1111_writel(val & ~sadev->skpcr_mask, sachip->base + SA1111_SKPCR); + val = readl_relaxed(sachip->base + SA1111_SKPCR); + writel_relaxed(val & ~sadev->skpcr_mask, sachip->base + SA1111_SKPCR); spin_unlock_irqrestore(&sachip->lock, flags); if (sachip->pdata && sachip->pdata->disable) @@ -1359,9 +1391,10 @@ EXPORT_SYMBOL(sa1111_disable_device); int sa1111_get_irq(struct sa1111_dev *sadev, unsigned num) { - if (num >= ARRAY_SIZE(sadev->irq)) + struct sa1111 *sachip = sa1111_chip_driver(sadev); + if (num >= ARRAY_SIZE(sadev->hwirq)) return -EINVAL; - return sadev->irq[num]; + return sa1111_map_irq(sachip, sadev->hwirq[num]); } EXPORT_SYMBOL_GPL(sa1111_get_irq); @@ -1379,36 +1412,6 @@ static int sa1111_match(struct device *_dev, struct device_driver *_drv) return !!(dev->devid & drv->devid); } -static int sa1111_bus_suspend(struct device *dev, pm_message_t state) -{ - struct sa1111_dev *sadev = to_sa1111_device(dev); - struct sa1111_driver *drv = SA1111_DRV(dev->driver); - int ret = 0; - - if (drv && drv->suspend) - ret = drv->suspend(sadev, state); - return ret; -} - -static int sa1111_bus_resume(struct device *dev) -{ - struct sa1111_dev *sadev = to_sa1111_device(dev); - struct sa1111_driver *drv = SA1111_DRV(dev->driver); - int ret = 0; - - if (drv && drv->resume) - ret = drv->resume(sadev); - return ret; -} - -static void sa1111_bus_shutdown(struct device *dev) -{ - struct sa1111_driver *drv = SA1111_DRV(dev->driver); - - if (drv && drv->shutdown) - drv->shutdown(to_sa1111_device(dev)); -} - static int sa1111_bus_probe(struct device *dev) { struct sa1111_dev *sadev = to_sa1111_device(dev); @@ -1436,9 +1439,6 @@ struct bus_type sa1111_bus_type = { .match = sa1111_match, .probe = sa1111_bus_probe, .remove = sa1111_bus_remove, - .suspend = sa1111_bus_suspend, - .resume = sa1111_bus_resume, - .shutdown = sa1111_bus_shutdown, }; EXPORT_SYMBOL(sa1111_bus_type); diff --git a/arch/arm/include/asm/exception.h b/arch/arm/include/asm/exception.h index a7273ad9587a..58e039a851af 100644 --- a/arch/arm/include/asm/exception.h +++ b/arch/arm/include/asm/exception.h @@ -10,11 +10,10 @@ #include <linux/interrupt.h> -#define __exception __attribute__((section(".exception.text"))) #ifdef CONFIG_FUNCTION_GRAPH_TRACER #define __exception_irq_entry __irq_entry #else -#define __exception_irq_entry __exception +#define __exception_irq_entry #endif #endif /* __ASM_ARM_EXCEPTION_H */ diff --git a/arch/arm/include/asm/glue-cache.h b/arch/arm/include/asm/glue-cache.h index 01c3d92624e5..8d1f498e5dd8 100644 --- a/arch/arm/include/asm/glue-cache.h +++ b/arch/arm/include/asm/glue-cache.h @@ -117,6 +117,10 @@ # endif #endif +#if defined(CONFIG_CACHE_B15_RAC) +# define MULTI_CACHE 1 +#endif + #if defined(CONFIG_CPU_V7M) # define MULTI_CACHE 1 #endif diff --git a/arch/arm/include/asm/hardware/cache-b15-rac.h b/arch/arm/include/asm/hardware/cache-b15-rac.h new file mode 100644 index 000000000000..3d43ec06fd35 --- /dev/null +++ b/arch/arm/include/asm/hardware/cache-b15-rac.h @@ -0,0 +1,10 @@ +#ifndef __ASM_ARM_HARDWARE_CACHE_B15_RAC_H +#define __ASM_ARM_HARDWARE_CACHE_B15_RAC_H + +#ifndef __ASSEMBLY__ + +void b15_flush_kern_cache_all(void); + +#endif + +#endif diff --git a/arch/arm/include/asm/hardware/sa1111.h b/arch/arm/include/asm/hardware/sa1111.h index 0bbf163d1ed3..798e520e8a49 100644 --- a/arch/arm/include/asm/hardware/sa1111.h +++ b/arch/arm/include/asm/hardware/sa1111.h @@ -16,33 +16,6 @@ #include <mach/bitfield.h> /* - * The SA1111 is always located at virtual 0xf4000000, and is always - * "native" endian. - */ - -#define SA1111_VBASE 0xf4000000 - -/* Don't use these! */ -#define SA1111_p2v( x ) ((x) - SA1111_BASE + SA1111_VBASE) -#define SA1111_v2p( x ) ((x) - SA1111_VBASE + SA1111_BASE) - -#ifndef __ASSEMBLY__ -#define _SA1111(x) ((x) + sa1111->resource.start) -#endif - -#define sa1111_writel(val,addr) __raw_writel(val, addr) -#define sa1111_readl(addr) __raw_readl(addr) - -/* - * 26 bits of the SA-1110 address bus are available to the SA-1111. - * Use these when feeding target addresses to the DMA engines. - */ - -#define SA1111_ADDR_WIDTH (26) -#define SA1111_ADDR_MASK ((1<<SA1111_ADDR_WIDTH)-1) -#define SA1111_DMA_ADDR(x) ((x)&SA1111_ADDR_MASK) - -/* * Don't ask the (SAC) DMA engines to move less than this amount. */ @@ -417,7 +390,7 @@ struct sa1111_dev { struct resource res; void __iomem *mapbase; unsigned int skpcr_mask; - unsigned int irq[6]; + unsigned int hwirq[6]; u64 dma_mask; }; @@ -431,9 +404,6 @@ struct sa1111_driver { unsigned int devid; int (*probe)(struct sa1111_dev *); int (*remove)(struct sa1111_dev *); - int (*suspend)(struct sa1111_dev *, pm_message_t); - int (*resume)(struct sa1111_dev *); - void (*shutdown)(struct sa1111_dev *); }; #define SA1111_DRV(_d) container_of((_d), struct sa1111_driver, drv) diff --git a/arch/arm/include/asm/memory.h b/arch/arm/include/asm/memory.h index 1f54e4e98c1e..496667703693 100644 --- a/arch/arm/include/asm/memory.h +++ b/arch/arm/include/asm/memory.h @@ -88,6 +88,7 @@ #else /* CONFIG_MMU */ #ifndef __ASSEMBLY__ +extern unsigned long setup_vectors_base(void); extern unsigned long vectors_base; #define VECTORS_BASE vectors_base #endif diff --git a/arch/arm/include/asm/ptdump.h b/arch/arm/include/asm/ptdump.h new file mode 100644 index 000000000000..3ebf9718288d --- /dev/null +++ b/arch/arm/include/asm/ptdump.h @@ -0,0 +1,43 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright (C) 2014 ARM Ltd. */ +#ifndef __ASM_PTDUMP_H +#define __ASM_PTDUMP_H + +#ifdef CONFIG_ARM_PTDUMP_CORE + +#include <linux/mm_types.h> +#include <linux/seq_file.h> + +struct addr_marker { + unsigned long start_address; + char *name; +}; + +struct ptdump_info { + struct mm_struct *mm; + const struct addr_marker *markers; + unsigned long base_addr; +}; + +void ptdump_walk_pgd(struct seq_file *s, struct ptdump_info *info); +#ifdef CONFIG_ARM_PTDUMP_DEBUGFS +int ptdump_debugfs_register(struct ptdump_info *info, const char *name); +#else +static inline int ptdump_debugfs_register(struct ptdump_info *info, + const char *name) +{ + return 0; +} +#endif /* CONFIG_ARM_PTDUMP_DEBUGFS */ + +void ptdump_check_wx(void); + +#endif /* CONFIG_ARM_PTDUMP_CORE */ + +#ifdef CONFIG_DEBUG_WX +#define debug_checkwx() ptdump_check_wx() +#else +#define debug_checkwx() do { } while (0) +#endif + +#endif /* __ASM_PTDUMP_H */ diff --git a/arch/arm/include/asm/sections.h b/arch/arm/include/asm/sections.h index 63dfe1f10335..4ceb4f757d4d 100644 --- a/arch/arm/include/asm/sections.h +++ b/arch/arm/include/asm/sections.h @@ -6,4 +6,25 @@ extern char _exiprom[]; +extern char __idmap_text_start[]; +extern char __idmap_text_end[]; +extern char __entry_text_start[]; +extern char __entry_text_end[]; +extern char __hyp_idmap_text_start[]; +extern char __hyp_idmap_text_end[]; + +static inline bool in_entry_text(unsigned long addr) +{ + return memory_contains(__entry_text_start, __entry_text_end, + (void *)addr, 1); +} + +static inline bool in_idmap_text(unsigned long addr) +{ + void *a = (void *)addr; + return memory_contains(__idmap_text_start, __idmap_text_end, a, 1) || + memory_contains(__hyp_idmap_text_start, __hyp_idmap_text_end, + a, 1); +} + #endif /* _ASM_ARM_SECTIONS_H */ diff --git a/arch/arm/include/asm/string.h b/arch/arm/include/asm/string.h index f54a3136aac6..111a1d8a41dd 100644 --- a/arch/arm/include/asm/string.h +++ b/arch/arm/include/asm/string.h @@ -39,18 +39,4 @@ static inline void *memset64(uint64_t *p, uint64_t v, __kernel_size_t n) return __memset64(p, v, n * 8, v >> 32); } -extern void __memzero(void *ptr, __kernel_size_t n); - -#define memset(p,v,n) \ - ({ \ - void *__p = (p); size_t __n = n; \ - if ((__n) != 0) { \ - if (__builtin_constant_p((v)) && (v) == 0) \ - __memzero((__p),(__n)); \ - else \ - memset((__p),(v),(__n)); \ - } \ - (__p); \ - }) - #endif diff --git a/arch/arm/include/asm/traps.h b/arch/arm/include/asm/traps.h index f9a6c5fc3fd1..a00288d75ee6 100644 --- a/arch/arm/include/asm/traps.h +++ b/arch/arm/include/asm/traps.h @@ -28,18 +28,6 @@ static inline int __in_irqentry_text(unsigned long ptr) ptr < (unsigned long)&__irqentry_text_end; } -static inline int in_exception_text(unsigned long ptr) -{ - extern char __exception_text_start[]; - extern char __exception_text_end[]; - int in; - - in = ptr >= (unsigned long)&__exception_text_start && - ptr < (unsigned long)&__exception_text_end; - - return in ? : __in_irqentry_text(ptr); -} - extern void __init early_trap_init(void *); extern void dump_backtrace_entry(unsigned long where, unsigned long from, unsigned long frame); extern void ptrace_break(struct task_struct *tsk, struct pt_regs *regs); diff --git a/arch/arm/include/asm/unified.h b/arch/arm/include/asm/unified.h index a91ae499614c..2c3b952be63e 100644 --- a/arch/arm/include/asm/unified.h +++ b/arch/arm/include/asm/unified.h @@ -20,8 +20,10 @@ #ifndef __ASM_UNIFIED_H #define __ASM_UNIFIED_H -#if defined(__ASSEMBLY__) && defined(CONFIG_ARM_ASM_UNIFIED) +#if defined(__ASSEMBLY__) .syntax unified +#else +__asm__(".syntax unified"); #endif #ifdef CONFIG_CPU_V7M @@ -64,77 +66,4 @@ #endif /* CONFIG_THUMB2_KERNEL */ -#ifndef CONFIG_ARM_ASM_UNIFIED - -/* - * If the unified assembly syntax isn't used (in ARM mode), these - * macros expand to an empty string - */ -#ifdef __ASSEMBLY__ - .macro it, cond - .endm - .macro itt, cond - .endm - .macro ite, cond - .endm - .macro ittt, cond - .endm - .macro itte, cond - .endm - .macro itet, cond - .endm - .macro itee, cond - .endm - .macro itttt, cond - .endm - .macro ittte, cond - .endm - .macro ittet, cond - .endm - .macro ittee, cond - .endm - .macro itett, cond - .endm - .macro itete, cond - .endm - .macro iteet, cond - .endm - .macro iteee, cond - .endm -#else /* !__ASSEMBLY__ */ -__asm__( -" .macro it, cond\n" -" .endm\n" -" .macro itt, cond\n" -" .endm\n" -" .macro ite, cond\n" -" .endm\n" -" .macro ittt, cond\n" -" .endm\n" -" .macro itte, cond\n" -" .endm\n" -" .macro itet, cond\n" -" .endm\n" -" .macro itee, cond\n" -" .endm\n" -" .macro itttt, cond\n" -" .endm\n" -" .macro ittte, cond\n" -" .endm\n" -" .macro ittet, cond\n" -" .endm\n" -" .macro ittee, cond\n" -" .endm\n" -" .macro itett, cond\n" -" .endm\n" -" .macro itete, cond\n" -" .endm\n" -" .macro iteet, cond\n" -" .endm\n" -" .macro iteee, cond\n" -" .endm\n"); -#endif /* __ASSEMBLY__ */ - -#endif /* CONFIG_ARM_ASM_UNIFIED */ - #endif /* !__ASM_UNIFIED_H */ diff --git a/arch/arm/kernel/armksyms.c b/arch/arm/kernel/armksyms.c index 5266fd9ad6b4..783fbb4de5f9 100644 --- a/arch/arm/kernel/armksyms.c +++ b/arch/arm/kernel/armksyms.c @@ -92,7 +92,6 @@ EXPORT_SYMBOL(__memset64); EXPORT_SYMBOL(memcpy); EXPORT_SYMBOL(memmove); EXPORT_SYMBOL(memchr); -EXPORT_SYMBOL(__memzero); EXPORT_SYMBOL(mmioset); EXPORT_SYMBOL(mmiocpy); diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index fbc707626b3e..1752033b0070 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -82,11 +82,7 @@ #endif .endm -#ifdef CONFIG_KPROBES - .section .kprobes.text,"ax",%progbits -#else - .text -#endif + .section .entry.text,"ax",%progbits /* * Invalid mode handlers diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index e655dcd0a933..3c4f88701f22 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -37,6 +37,7 @@ saved_pc .req lr #define TRACE(x...) #endif + .section .entry.text,"ax",%progbits .align 5 #if !(IS_ENABLED(CONFIG_TRACE_IRQFLAGS) || IS_ENABLED(CONFIG_CONTEXT_TRACKING)) /* diff --git a/arch/arm/kernel/head-common.S b/arch/arm/kernel/head-common.S index 21dde771a7dd..6e0375e7db05 100644 --- a/arch/arm/kernel/head-common.S +++ b/arch/arm/kernel/head-common.S @@ -105,8 +105,9 @@ __mmap_switched: ARM( ldmia r4!, {r0, r1, sp} ) THUMB( ldmia r4!, {r0, r1, r3} ) THUMB( mov sp, r3 ) - sub r1, r1, r0 - bl __memzero @ clear .bss + sub r2, r1, r0 + mov r1, #0 + bl memset @ clear .bss ldmia r4, {r0, r1, r2, r3} str r9, [r0] @ Save processor ID diff --git a/arch/arm/kernel/hw_breakpoint.c b/arch/arm/kernel/hw_breakpoint.c index af2a7f1e3103..629e25152c0d 100644 --- a/arch/arm/kernel/hw_breakpoint.c +++ b/arch/arm/kernel/hw_breakpoint.c @@ -44,17 +44,17 @@ static DEFINE_PER_CPU(struct perf_event *, bp_on_reg[ARM_MAX_BRP]); static DEFINE_PER_CPU(struct perf_event *, wp_on_reg[ARM_MAX_WRP]); /* Number of BRP/WRP registers on this CPU. */ -static int core_num_brps; -static int core_num_wrps; +static int core_num_brps __ro_after_init; +static int core_num_wrps __ro_after_init; /* Debug architecture version. */ -static u8 debug_arch; +static u8 debug_arch __ro_after_init; /* Does debug architecture support OS Save and Restore? */ -static bool has_ossr; +static bool has_ossr __ro_after_init; /* Maximum supported watchpoint length. */ -static u8 max_watchpoint_len; +static u8 max_watchpoint_len __ro_after_init; #define READ_WB_REG_CASE(OP2, M, VAL) \ case ((OP2 << 4) + M): \ diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c index b4fbf00ee4ad..2da087926ebe 100644 --- a/arch/arm/kernel/smp.c +++ b/arch/arm/kernel/smp.c @@ -379,6 +379,9 @@ asmlinkage void secondary_start_kernel(void) cpu_init(); +#ifndef CONFIG_MMU + setup_vectors_base(); +#endif pr_debug("CPU%u: Booted secondary processor\n", cpu); preempt_disable(); diff --git a/arch/arm/kernel/stacktrace.c b/arch/arm/kernel/stacktrace.c index 65228bf4c6df..a56e7c856ab5 100644 --- a/arch/arm/kernel/stacktrace.c +++ b/arch/arm/kernel/stacktrace.c @@ -3,6 +3,7 @@ #include <linux/sched/debug.h> #include <linux/stacktrace.h> +#include <asm/sections.h> #include <asm/stacktrace.h> #include <asm/traps.h> @@ -63,7 +64,6 @@ EXPORT_SYMBOL(walk_stackframe); #ifdef CONFIG_STACKTRACE struct stack_trace_data { struct stack_trace *trace; - unsigned long last_pc; unsigned int no_sched_functions; unsigned int skip; }; @@ -87,16 +87,7 @@ static int save_trace(struct stackframe *frame, void *d) if (trace->nr_entries >= trace->max_entries) return 1; - /* - * in_exception_text() is designed to test if the PC is one of - * the functions which has an exception stack above it, but - * unfortunately what is in frame->pc is the return LR value, - * not the saved PC value. So, we need to track the previous - * frame PC value when doing this. - */ - addr = data->last_pc; - data->last_pc = frame->pc; - if (!in_exception_text(addr)) + if (!in_entry_text(frame->pc)) return 0; regs = (struct pt_regs *)frame->sp; @@ -114,7 +105,6 @@ static noinline void __save_stack_trace(struct task_struct *tsk, struct stackframe frame; data.trace = trace; - data.last_pc = ULONG_MAX; data.skip = trace->skip; data.no_sched_functions = nosched; diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index 5cf04888c581..e344bdd2e5ac 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -72,7 +72,7 @@ void dump_backtrace_entry(unsigned long where, unsigned long from, unsigned long printk("Function entered at [<%08lx>] from [<%08lx>]\n", where, from); #endif - if (in_exception_text(where)) + if (in_entry_text(from)) dump_mem("", "Exception stack", frame + 4, frame + 4 + sizeof(struct pt_regs)); } @@ -433,7 +433,7 @@ static int call_undef_hook(struct pt_regs *regs, unsigned int instr) return fn ? fn(regs, instr) : 1; } -asmlinkage void __exception do_undefinstr(struct pt_regs *regs) +asmlinkage void do_undefinstr(struct pt_regs *regs) { unsigned int instr; siginfo_t info; diff --git a/arch/arm/kernel/vmlinux-xip.lds.S b/arch/arm/kernel/vmlinux-xip.lds.S index ec4b3f94ad80..12b87591eb7c 100644 --- a/arch/arm/kernel/vmlinux-xip.lds.S +++ b/arch/arm/kernel/vmlinux-xip.lds.S @@ -96,9 +96,9 @@ SECTIONS .text : { /* Real text segment */ _stext = .; /* Text and read-only data */ IDMAP_TEXT - __exception_text_start = .; - *(.exception.text) - __exception_text_end = .; + __entry_text_start = .; + *(.entry.text) + __entry_text_end = .; IRQENTRY_TEXT TEXT_TEXT SCHED_TEXT diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S index ee53f6518872..84a1ae3ce46e 100644 --- a/arch/arm/kernel/vmlinux.lds.S +++ b/arch/arm/kernel/vmlinux.lds.S @@ -105,9 +105,9 @@ SECTIONS .text : { /* Real text segment */ _stext = .; /* Text and read-only data */ IDMAP_TEXT - __exception_text_start = .; - *(.exception.text) - __exception_text_end = .; + __entry_text_start = .; + *(.entry.text) + __entry_text_end = .; IRQENTRY_TEXT SOFTIRQENTRY_TEXT TEXT_TEXT diff --git a/arch/arm/lib/Makefile b/arch/arm/lib/Makefile index 4cb0b9624d8f..ad25fd1872c7 100644 --- a/arch/arm/lib/Makefile +++ b/arch/arm/lib/Makefile @@ -8,7 +8,7 @@ lib-y := backtrace.o changebit.o csumipv6.o csumpartial.o \ csumpartialcopy.o csumpartialcopyuser.o clearbit.o \ delay.o delay-loop.o findbit.o memchr.o memcpy.o \ - memmove.o memset.o memzero.o setbit.o \ + memmove.o memset.o setbit.o \ strchr.o strrchr.o \ testchangebit.o testclearbit.o testsetbit.o \ ashldi3.o ashrdi3.o lshrdi3.o muldi3.o \ diff --git a/arch/arm/lib/memzero.S b/arch/arm/lib/memzero.S deleted file mode 100644 index 0eded952e089..000000000000 --- a/arch/arm/lib/memzero.S +++ /dev/null @@ -1,137 +0,0 @@ -/* - * linux/arch/arm/lib/memzero.S - * - * Copyright (C) 1995-2000 Russell King - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ -#include <linux/linkage.h> -#include <asm/assembler.h> -#include <asm/unwind.h> - - .text - .align 5 - .word 0 -/* - * Align the pointer in r0. r3 contains the number of bytes that we are - * mis-aligned by, and r1 is the number of bytes. If r1 < 4, then we - * don't bother; we use byte stores instead. - */ -UNWIND( .fnstart ) -1: subs r1, r1, #4 @ 1 do we have enough - blt 5f @ 1 bytes to align with? - cmp r3, #2 @ 1 - strltb r2, [r0], #1 @ 1 - strleb r2, [r0], #1 @ 1 - strb r2, [r0], #1 @ 1 - add r1, r1, r3 @ 1 (r1 = r1 - (4 - r3)) -/* - * The pointer is now aligned and the length is adjusted. Try doing the - * memzero again. - */ - -ENTRY(__memzero) - mov r2, #0 @ 1 - ands r3, r0, #3 @ 1 unaligned? - bne 1b @ 1 -/* - * r3 = 0, and we know that the pointer in r0 is aligned to a word boundary. - */ - cmp r1, #16 @ 1 we can skip this chunk if we - blt 4f @ 1 have < 16 bytes - -#if ! CALGN(1)+0 - -/* - * We need an extra register for this loop - save the return address and - * use the LR - */ - str lr, [sp, #-4]! @ 1 -UNWIND( .fnend ) -UNWIND( .fnstart ) -UNWIND( .save {lr} ) - mov ip, r2 @ 1 - mov lr, r2 @ 1 - -3: subs r1, r1, #64 @ 1 write 32 bytes out per loop - stmgeia r0!, {r2, r3, ip, lr} @ 4 - stmgeia r0!, {r2, r3, ip, lr} @ 4 - stmgeia r0!, {r2, r3, ip, lr} @ 4 - stmgeia r0!, {r2, r3, ip, lr} @ 4 - bgt 3b @ 1 - ldmeqfd sp!, {pc} @ 1/2 quick exit -/* - * No need to correct the count; we're only testing bits from now on - */ - tst r1, #32 @ 1 - stmneia r0!, {r2, r3, ip, lr} @ 4 - stmneia r0!, {r2, r3, ip, lr} @ 4 - tst r1, #16 @ 1 16 bytes or more? - stmneia r0!, {r2, r3, ip, lr} @ 4 - ldr lr, [sp], #4 @ 1 -UNWIND( .fnend ) - -#else - -/* - * This version aligns the destination pointer in order to write - * whole cache lines at once. - */ - - stmfd sp!, {r4-r7, lr} -UNWIND( .fnend ) -UNWIND( .fnstart ) -UNWIND( .save {r4-r7, lr} ) - mov r4, r2 - mov r5, r2 - mov r6, r2 - mov r7, r2 - mov ip, r2 - mov lr, r2 - - cmp r1, #96 - andgts ip, r0, #31 - ble 3f - - rsb ip, ip, #32 - sub r1, r1, ip - movs ip, ip, lsl #(32 - 4) - stmcsia r0!, {r4, r5, r6, r7} - stmmiia r0!, {r4, r5} - movs ip, ip, lsl #2 - strcs r2, [r0], #4 - -3: subs r1, r1, #64 - stmgeia r0!, {r2-r7, ip, lr} - stmgeia r0!, {r2-r7, ip, lr} - bgt 3b - ldmeqfd sp!, {r4-r7, pc} - - tst r1, #32 - stmneia r0!, {r2-r7, ip, lr} - tst r1, #16 - stmneia r0!, {r4-r7} - ldmfd sp!, {r4-r7, lr} -UNWIND( .fnend ) - -#endif - -UNWIND( .fnstart ) -4: tst r1, #8 @ 1 8 bytes or more? - stmneia r0!, {r2, r3} @ 2 - tst r1, #4 @ 1 4 bytes or more? - strne r2, [r0], #4 @ 1 -/* - * When we get here, we've got less than 4 bytes to zero. We - * may have an unaligned pointer as well. - */ -5: tst r1, #2 @ 1 2 bytes or more? - strneb r2, [r0], #1 @ 1 - strneb r2, [r0], #1 @ 1 - tst r1, #1 @ 1 a byte left over - strneb r2, [r0], #1 @ 1 - ret lr @ 1 -UNWIND( .fnend ) -ENDPROC(__memzero) diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index fd9077a74fce..7f14acf67caf 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -909,6 +909,14 @@ config OUTER_CACHE_SYNC The outer cache has a outer_cache_fns.sync function pointer that can be used to drain the write buffer of the outer cache. +config CACHE_B15_RAC + bool "Enable the Broadcom Brahma-B15 read-ahead cache controller" + depends on ARCH_BRCMSTB + default y + help + This option enables the Broadcom Brahma-B15 read-ahead cache + controller. If disabled, the read-ahead cache remains off. + config CACHE_FEROCEON_L2 bool "Enable the Feroceon L2 cache controller" depends on ARCH_MV78XX0 || ARCH_MVEBU diff --git a/arch/arm/mm/Makefile b/arch/arm/mm/Makefile index 01bcc33f59e3..9dbb84923e12 100644 --- a/arch/arm/mm/Makefile +++ b/arch/arm/mm/Makefile @@ -13,7 +13,8 @@ obj-y += nommu.o obj-$(CONFIG_ARM_MPU) += pmsa-v7.o endif -obj-$(CONFIG_ARM_PTDUMP) += dump.o +obj-$(CONFIG_ARM_PTDUMP_CORE) += dump.o +obj-$(CONFIG_ARM_PTDUMP_DEBUGFS) += ptdump_debugfs.o obj-$(CONFIG_MODULES) += proc-syms.o obj-$(CONFIG_DEBUG_VIRTUAL) += physaddr.o @@ -103,6 +104,7 @@ AFLAGS_proc-v6.o :=-Wa,-march=armv6 AFLAGS_proc-v7.o :=-Wa,-march=armv7-a obj-$(CONFIG_OUTER_CACHE) += l2c-common.o +obj-$(CONFIG_CACHE_B15_RAC) += cache-b15-rac.o obj-$(CONFIG_CACHE_FEROCEON_L2) += cache-feroceon-l2.o obj-$(CONFIG_CACHE_L2X0) += cache-l2x0.o l2c-l2x0-resume.o obj-$(CONFIG_CACHE_L2X0_PMU) += cache-l2x0-pmu.o diff --git a/arch/arm/mm/cache-b15-rac.c b/arch/arm/mm/cache-b15-rac.c new file mode 100644 index 000000000000..d9586ba2e63c --- /dev/null +++ b/arch/arm/mm/cache-b15-rac.c @@ -0,0 +1,356 @@ +/* + * Broadcom Brahma-B15 CPU read-ahead cache management functions + * + * Copyright (C) 2015-2016 Broadcom + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/err.h> +#include <linux/spinlock.h> +#include <linux/io.h> +#include <linux/bitops.h> +#include <linux/of_address.h> +#include <linux/notifier.h> +#include <linux/cpu.h> +#include <linux/syscore_ops.h> +#include <linux/reboot.h> + +#include <asm/cacheflush.h> +#include <asm/hardware/cache-b15-rac.h> + +extern void v7_flush_kern_cache_all(void); + +/* RAC register offsets, relative to the HIF_CPU_BIUCTRL register base */ +#define RAC_CONFIG0_REG (0x78) +#define RACENPREF_MASK (0x3) +#define RACPREFINST_SHIFT (0) +#define RACENINST_SHIFT (2) +#define RACPREFDATA_SHIFT (4) +#define RACENDATA_SHIFT (6) +#define RAC_CPU_SHIFT (8) +#define RACCFG_MASK (0xff) +#define RAC_CONFIG1_REG (0x7c) +#define RAC_FLUSH_REG (0x80) +#define FLUSH_RAC (1 << 0) + +/* Bitmask to enable instruction and data prefetching with a 256-bytes stride */ +#define RAC_DATA_INST_EN_MASK (1 << RACPREFINST_SHIFT | \ + RACENPREF_MASK << RACENINST_SHIFT | \ + 1 << RACPREFDATA_SHIFT | \ + RACENPREF_MASK << RACENDATA_SHIFT) + +#define RAC_ENABLED 0 +/* Special state where we want to bypass the spinlock and call directly + * into the v7 cache maintenance operations during suspend/resume + */ +#define RAC_SUSPENDED 1 + +static void __iomem *b15_rac_base; +static DEFINE_SPINLOCK(rac_lock); + +static u32 rac_config0_reg; + +/* Initialization flag to avoid checking for b15_rac_base, and to prevent + * multi-platform kernels from crashing here as well. + */ +static unsigned long b15_rac_flags; + +static inline u32 __b15_rac_disable(void) +{ + u32 val = __raw_readl(b15_rac_base + RAC_CONFIG0_REG); + __raw_writel(0, b15_rac_base + RAC_CONFIG0_REG); + dmb(); + return val; +} + +static inline void __b15_rac_flush(void) +{ + u32 reg; + + __raw_writel(FLUSH_RAC, b15_rac_base + RAC_FLUSH_REG); + do { + /* This dmb() is required to force the Bus Interface Unit + * to clean oustanding writes, and forces an idle cycle + * to be inserted. + */ + dmb(); + reg = __raw_readl(b15_rac_base + RAC_FLUSH_REG); + } while (reg & FLUSH_RAC); +} + +static inline u32 b15_rac_disable_and_flush(void) +{ + u32 reg; + + reg = __b15_rac_disable(); + __b15_rac_flush(); + return reg; +} + +static inline void __b15_rac_enable(u32 val) +{ + __raw_writel(val, b15_rac_base + RAC_CONFIG0_REG); + /* dsb() is required here to be consistent with __flush_icache_all() */ + dsb(); +} + +#define BUILD_RAC_CACHE_OP(name, bar) \ +void b15_flush_##name(void) \ +{ \ + unsigned int do_flush; \ + u32 val = 0; \ + \ + if (test_bit(RAC_SUSPENDED, &b15_rac_flags)) { \ + v7_flush_##name(); \ + bar; \ + return; \ + } \ + \ + spin_lock(&rac_lock); \ + do_flush = test_bit(RAC_ENABLED, &b15_rac_flags); \ + if (do_flush) \ + val = b15_rac_disable_and_flush(); \ + v7_flush_##name(); \ + if (!do_flush) \ + bar; \ + else \ + __b15_rac_enable(val); \ + spin_unlock(&rac_lock); \ +} + +#define nobarrier + +/* The readahead cache present in the Brahma-B15 CPU is a special piece of + * hardware after the integrated L2 cache of the B15 CPU complex whose purpose + * is to prefetch instruction and/or data with a line size of either 64 bytes + * or 256 bytes. The rationale is that the data-bus of the CPU interface is + * optimized for 256-bytes transactions, and enabling the readahead cache + * provides a significant performance boost we want it enabled (typically + * twice the performance for a memcpy benchmark application). + * + * The readahead cache is transparent for Modified Virtual Addresses + * cache maintenance operations: ICIMVAU, DCIMVAC, DCCMVAC, DCCMVAU and + * DCCIMVAC. + * + * It is however not transparent for the following cache maintenance + * operations: DCISW, DCCSW, DCCISW, ICIALLUIS and ICIALLU which is precisely + * what we are patching here with our BUILD_RAC_CACHE_OP here. + */ +BUILD_RAC_CACHE_OP(kern_cache_all, nobarrier); + +static void b15_rac_enable(void) +{ + unsigned int cpu; + u32 enable = 0; + + for_each_possible_cpu(cpu) + enable |= (RAC_DATA_INST_EN_MASK << (cpu * RAC_CPU_SHIFT)); + + b15_rac_disable_and_flush(); + __b15_rac_enable(enable); +} + +static int b15_rac_reboot_notifier(struct notifier_block *nb, + unsigned long action, + void *data) +{ + /* During kexec, we are not yet migrated on the boot CPU, so we need to + * make sure we are SMP safe here. Once the RAC is disabled, flag it as + * suspended such that the hotplug notifier returns early. + */ + if (action == SYS_RESTART) { + spin_lock(&rac_lock); + b15_rac_disable_and_flush(); + clear_bit(RAC_ENABLED, &b15_rac_flags); + set_bit(RAC_SUSPENDED, &b15_rac_flags); + spin_unlock(&rac_lock); + } + + return NOTIFY_DONE; +} + +static struct notifier_block b15_rac_reboot_nb = { + .notifier_call = b15_rac_reboot_notifier, +}; + +/* The CPU hotplug case is the most interesting one, we basically need to make + * sure that the RAC is disabled for the entire system prior to having a CPU + * die, in particular prior to this dying CPU having exited the coherency + * domain. + * + * Once this CPU is marked dead, we can safely re-enable the RAC for the + * remaining CPUs in the system which are still online. + * + * Offlining a CPU is the problematic case, onlining a CPU is not much of an + * issue since the CPU and its cache-level hierarchy will start filling with + * the RAC disabled, so L1 and L2 only. + * + * In this function, we should NOT have to verify any unsafe setting/condition + * b15_rac_base: + * + * It is protected by the RAC_ENABLED flag which is cleared by default, and + * being cleared when initial procedure is done. b15_rac_base had been set at + * that time. + * + * RAC_ENABLED: + * There is a small timing windows, in b15_rac_init(), between + * cpuhp_setup_state_*() + * ... + * set RAC_ENABLED + * However, there is no hotplug activity based on the Linux booting procedure. + * + * Since we have to disable RAC for all cores, we keep RAC on as long as as + * possible (disable it as late as possible) to gain the cache benefit. + * + * Thus, dying/dead states are chosen here + * + * We are choosing not do disable the RAC on a per-CPU basis, here, if we did + * we would want to consider disabling it as early as possible to benefit the + * other active CPUs. + */ + +/* Running on the dying CPU */ +static int b15_rac_dying_cpu(unsigned int cpu) +{ + /* During kexec/reboot, the RAC is disabled via the reboot notifier + * return early here. + */ + if (test_bit(RAC_SUSPENDED, &b15_rac_flags)) + return 0; + + spin_lock(&rac_lock); + + /* Indicate that we are starting a hotplug procedure */ + __clear_bit(RAC_ENABLED, &b15_rac_flags); + + /* Disable the readahead cache and save its value to a global */ + rac_config0_reg = b15_rac_disable_and_flush(); + + spin_unlock(&rac_lock); + + return 0; +} + +/* Running on a non-dying CPU */ +static int b15_rac_dead_cpu(unsigned int cpu) +{ + /* During kexec/reboot, the RAC is disabled via the reboot notifier + * return early here. + */ + if (test_bit(RAC_SUSPENDED, &b15_rac_flags)) + return 0; + + spin_lock(&rac_lock); + + /* And enable it */ + __b15_rac_enable(rac_config0_reg); + __set_bit(RAC_ENABLED, &b15_rac_flags); + + spin_unlock(&rac_lock); + + return 0; +} + +static int b15_rac_suspend(void) +{ + /* Suspend the read-ahead cache oeprations, forcing our cache + * implementation to fallback to the regular ARMv7 calls. + * + * We are guaranteed to be running on the boot CPU at this point and + * with every other CPU quiesced, so setting RAC_SUSPENDED is not racy + * here. + */ + rac_config0_reg = b15_rac_disable_and_flush(); + set_bit(RAC_SUSPENDED, &b15_rac_flags); + + return 0; +} + +static void b15_rac_resume(void) +{ + /* Coming out of a S3 suspend/resume cycle, the read-ahead cache + * register RAC_CONFIG0_REG will be restored to its default value, make + * sure we re-enable it and set the enable flag, we are also guaranteed + * to run on the boot CPU, so not racy again. + */ + __b15_rac_enable(rac_config0_reg); + clear_bit(RAC_SUSPENDED, &b15_rac_flags); +} + +static struct syscore_ops b15_rac_syscore_ops = { + .suspend = b15_rac_suspend, + .resume = b15_rac_resume, +}; + +static int __init b15_rac_init(void) +{ + struct device_node *dn; + int ret = 0, cpu; + u32 reg, en_mask = 0; + + dn = of_find_compatible_node(NULL, NULL, "brcm,brcmstb-cpu-biu-ctrl"); + if (!dn) + return -ENODEV; + + if (WARN(num_possible_cpus() > 4, "RAC only supports 4 CPUs\n")) + goto out; + + b15_rac_base = of_iomap(dn, 0); + if (!b15_rac_base) { + pr_err("failed to remap BIU control base\n"); + ret = -ENOMEM; + goto out; + } + + ret = register_reboot_notifier(&b15_rac_reboot_nb); + if (ret) { + pr_err("failed to register reboot notifier\n"); + iounmap(b15_rac_base); + goto out; + } + + if (IS_ENABLED(CONFIG_HOTPLUG_CPU)) { + ret = cpuhp_setup_state_nocalls(CPUHP_AP_ARM_CACHE_B15_RAC_DEAD, + "arm/cache-b15-rac:dead", + NULL, b15_rac_dead_cpu); + if (ret) + goto out_unmap; + + ret = cpuhp_setup_state_nocalls(CPUHP_AP_ARM_CACHE_B15_RAC_DYING, + "arm/cache-b15-rac:dying", + NULL, b15_rac_dying_cpu); + if (ret) + goto out_cpu_dead; + } + + if (IS_ENABLED(CONFIG_PM_SLEEP)) + register_syscore_ops(&b15_rac_syscore_ops); + + spin_lock(&rac_lock); + reg = __raw_readl(b15_rac_base + RAC_CONFIG0_REG); + for_each_possible_cpu(cpu) + en_mask |= ((1 << RACPREFDATA_SHIFT) << (cpu * RAC_CPU_SHIFT)); + WARN(reg & en_mask, "Read-ahead cache not previously disabled\n"); + + b15_rac_enable(); + set_bit(RAC_ENABLED, &b15_rac_flags); + spin_unlock(&rac_lock); + + pr_info("Broadcom Brahma-B15 readahead cache at: 0x%p\n", + b15_rac_base + RAC_CONFIG0_REG); + + goto out; + +out_cpu_dead: + cpuhp_remove_state_nocalls(CPUHP_AP_ARM_CACHE_B15_RAC_DYING); +out_unmap: + unregister_reboot_notifier(&b15_rac_reboot_nb); + iounmap(b15_rac_base); +out: + of_node_put(dn); + return ret; +} +arch_initcall(b15_rac_init); diff --git a/arch/arm/mm/cache-v7.S b/arch/arm/mm/cache-v7.S index de78109d002d..215df435bfb9 100644 --- a/arch/arm/mm/cache-v7.S +++ b/arch/arm/mm/cache-v7.S @@ -15,6 +15,7 @@ #include <asm/assembler.h> #include <asm/errno.h> #include <asm/unwind.h> +#include <asm/hardware/cache-b15-rac.h> #include "proc-macros.S" @@ -446,3 +447,23 @@ ENDPROC(v7_dma_unmap_area) @ define struct cpu_cache_fns (see <asm/cacheflush.h> and proc-macros.S) define_cache_functions v7 + + /* The Broadcom Brahma-B15 read-ahead cache requires some modifications + * to the v7_cache_fns, we only override the ones we need + */ +#ifndef CONFIG_CACHE_B15_RAC + globl_equ b15_flush_kern_cache_all, v7_flush_kern_cache_all +#endif + globl_equ b15_flush_icache_all, v7_flush_icache_all + globl_equ b15_flush_kern_cache_louis, v7_flush_kern_cache_louis + globl_equ b15_flush_user_cache_all, v7_flush_user_cache_all + globl_equ b15_flush_user_cache_range, v7_flush_user_cache_range + globl_equ b15_coherent_kern_range, v7_coherent_kern_range + globl_equ b15_coherent_user_range, v7_coherent_user_range + globl_equ b15_flush_kern_dcache_area, v7_flush_kern_dcache_area + + globl_equ b15_dma_map_area, v7_dma_map_area + globl_equ b15_dma_unmap_area, v7_dma_unmap_area + globl_equ b15_dma_flush_range, v7_dma_flush_range + + define_cache_functions b15 diff --git a/arch/arm/mm/dump.c b/arch/arm/mm/dump.c index fc3b44028cfb..084779c5c893 100644 --- a/arch/arm/mm/dump.c +++ b/arch/arm/mm/dump.c @@ -21,11 +21,7 @@ #include <asm/fixmap.h> #include <asm/memory.h> #include <asm/pgtable.h> - -struct addr_marker { - unsigned long start_address; - const char *name; -}; +#include <asm/ptdump.h> static struct addr_marker address_markers[] = { { MODULES_VADDR, "Modules" }, @@ -38,12 +34,26 @@ static struct addr_marker address_markers[] = { { -1, NULL }, }; +#define pt_dump_seq_printf(m, fmt, args...) \ +({ \ + if (m) \ + seq_printf(m, fmt, ##args); \ +}) + +#define pt_dump_seq_puts(m, fmt) \ +({ \ + if (m) \ + seq_printf(m, fmt); \ +}) + struct pg_state { struct seq_file *seq; const struct addr_marker *marker; unsigned long start_address; unsigned level; u64 current_prot; + bool check_wx; + unsigned long wx_pages; const char *current_domain; }; @@ -52,6 +62,8 @@ struct prot_bits { u64 val; const char *set; const char *clear; + bool ro_bit; + bool nx_bit; }; static const struct prot_bits pte_bits[] = { @@ -65,11 +77,13 @@ static const struct prot_bits pte_bits[] = { .val = L_PTE_RDONLY, .set = "ro", .clear = "RW", + .ro_bit = true, }, { .mask = L_PTE_XN, .val = L_PTE_XN, .set = "NX", .clear = "x ", + .nx_bit = true, }, { .mask = L_PTE_SHARED, .val = L_PTE_SHARED, @@ -133,11 +147,13 @@ static const struct prot_bits section_bits[] = { .val = L_PMD_SECT_RDONLY | PMD_SECT_AP2, .set = "ro", .clear = "RW", + .ro_bit = true, #elif __LINUX_ARM_ARCH__ >= 6 { .mask = PMD_SECT_APX | PMD_SECT_AP_READ | PMD_SECT_AP_WRITE, .val = PMD_SECT_APX | PMD_SECT_AP_WRITE, .set = " ro", + .ro_bit = true, }, { .mask = PMD_SECT_APX | PMD_SECT_AP_READ | PMD_SECT_AP_WRITE, .val = PMD_SECT_AP_WRITE, @@ -156,6 +172,7 @@ static const struct prot_bits section_bits[] = { .mask = PMD_SECT_AP_READ | PMD_SECT_AP_WRITE, .val = 0, .set = " ro", + .ro_bit = true, }, { .mask = PMD_SECT_AP_READ | PMD_SECT_AP_WRITE, .val = PMD_SECT_AP_WRITE, @@ -174,6 +191,7 @@ static const struct prot_bits section_bits[] = { .val = PMD_SECT_XN, .set = "NX", .clear = "x ", + .nx_bit = true, }, { .mask = PMD_SECT_S, .val = PMD_SECT_S, @@ -186,6 +204,8 @@ struct pg_level { const struct prot_bits *bits; size_t num; u64 mask; + const struct prot_bits *ro_bit; + const struct prot_bits *nx_bit; }; static struct pg_level pg_level[] = { @@ -214,10 +234,27 @@ static void dump_prot(struct pg_state *st, const struct prot_bits *bits, size_t s = bits->clear; if (s) - seq_printf(st->seq, " %s", s); + pt_dump_seq_printf(st->seq, " %s", s); } } +static void note_prot_wx(struct pg_state *st, unsigned long addr) +{ + if (!st->check_wx) + return; + if ((st->current_prot & pg_level[st->level].ro_bit->mask) == + pg_level[st->level].ro_bit->val) + return; + if ((st->current_prot & pg_level[st->level].nx_bit->mask) == + pg_level[st->level].nx_bit->val) + return; + + WARN_ONCE(1, "arm/mm: Found insecure W+X mapping at address %pS\n", + (void *)st->start_address); + + st->wx_pages += (addr - st->start_address) / PAGE_SIZE; +} + static void note_page(struct pg_state *st, unsigned long addr, unsigned int level, u64 val, const char *domain) { @@ -228,7 +265,7 @@ static void note_page(struct pg_state *st, unsigned long addr, st->level = level; st->current_prot = prot; st->current_domain = domain; - seq_printf(st->seq, "---[ %s ]---\n", st->marker->name); + pt_dump_seq_printf(st->seq, "---[ %s ]---\n", st->marker->name); } else if (prot != st->current_prot || level != st->level || domain != st->current_domain || addr >= st->marker[1].start_address) { @@ -236,7 +273,8 @@ static void note_page(struct pg_state *st, unsigned long addr, unsigned long delta; if (st->current_prot) { - seq_printf(st->seq, "0x%08lx-0x%08lx ", + note_prot_wx(st, addr); + pt_dump_seq_printf(st->seq, "0x%08lx-0x%08lx ", st->start_address, addr); delta = (addr - st->start_address) >> 10; @@ -244,17 +282,19 @@ static void note_page(struct pg_state *st, unsigned long addr, delta >>= 10; unit++; } - seq_printf(st->seq, "%9lu%c", delta, *unit); + pt_dump_seq_printf(st->seq, "%9lu%c", delta, *unit); if (st->current_domain) - seq_printf(st->seq, " %s", st->current_domain); + pt_dump_seq_printf(st->seq, " %s", + st->current_domain); if (pg_level[st->level].bits) dump_prot(st, pg_level[st->level].bits, pg_level[st->level].num); - seq_printf(st->seq, "\n"); + pt_dump_seq_printf(st->seq, "\n"); } if (addr >= st->marker[1].start_address) { st->marker++; - seq_printf(st->seq, "---[ %s ]---\n", st->marker->name); + pt_dump_seq_printf(st->seq, "---[ %s ]---\n", + st->marker->name); } st->start_address = addr; st->current_prot = prot; @@ -335,61 +375,82 @@ static void walk_pud(struct pg_state *st, pgd_t *pgd, unsigned long start) } } -static void walk_pgd(struct seq_file *m) +static void walk_pgd(struct pg_state *st, struct mm_struct *mm, + unsigned long start) { - pgd_t *pgd = swapper_pg_dir; - struct pg_state st; - unsigned long addr; + pgd_t *pgd = pgd_offset(mm, 0UL); unsigned i; - - memset(&st, 0, sizeof(st)); - st.seq = m; - st.marker = address_markers; + unsigned long addr; for (i = 0; i < PTRS_PER_PGD; i++, pgd++) { - addr = i * PGDIR_SIZE; + addr = start + i * PGDIR_SIZE; if (!pgd_none(*pgd)) { - walk_pud(&st, pgd, addr); + walk_pud(st, pgd, addr); } else { - note_page(&st, addr, 1, pgd_val(*pgd), NULL); + note_page(st, addr, 1, pgd_val(*pgd), NULL); } } - - note_page(&st, 0, 0, 0, NULL); } -static int ptdump_show(struct seq_file *m, void *v) +void ptdump_walk_pgd(struct seq_file *m, struct ptdump_info *info) { - walk_pgd(m); - return 0; -} + struct pg_state st = { + .seq = m, + .marker = info->markers, + .check_wx = false, + }; -static int ptdump_open(struct inode *inode, struct file *file) -{ - return single_open(file, ptdump_show, NULL); + walk_pgd(&st, info->mm, info->base_addr); + note_page(&st, 0, 0, 0, NULL); } -static const struct file_operations ptdump_fops = { - .open = ptdump_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -static int ptdump_init(void) +static void ptdump_initialize(void) { - struct dentry *pe; unsigned i, j; for (i = 0; i < ARRAY_SIZE(pg_level); i++) if (pg_level[i].bits) - for (j = 0; j < pg_level[i].num; j++) + for (j = 0; j < pg_level[i].num; j++) { pg_level[i].mask |= pg_level[i].bits[j].mask; + if (pg_level[i].bits[j].ro_bit) + pg_level[i].ro_bit = &pg_level[i].bits[j]; + if (pg_level[i].bits[j].nx_bit) + pg_level[i].nx_bit = &pg_level[i].bits[j]; + } address_markers[2].start_address = VMALLOC_START; +} + +static struct ptdump_info kernel_ptdump_info = { + .mm = &init_mm, + .markers = address_markers, + .base_addr = 0, +}; - pe = debugfs_create_file("kernel_page_tables", 0400, NULL, NULL, - &ptdump_fops); - return pe ? 0 : -ENOMEM; +void ptdump_check_wx(void) +{ + struct pg_state st = { + .seq = NULL, + .marker = (struct addr_marker[]) { + { 0, NULL}, + { -1, NULL}, + }, + .check_wx = true, + }; + + walk_pgd(&st, &init_mm, 0); + note_page(&st, 0, 0, 0, NULL); + if (st.wx_pages) + pr_warn("Checked W+X mappings: FAILED, %lu W+X pages found\n", + st.wx_pages); + else + pr_info("Checked W+X mappings: passed, no W+X pages found\n"); +} + +static int ptdump_init(void) +{ + ptdump_initialize(); + return ptdump_debugfs_register(&kernel_ptdump_info, + "kernel_page_tables"); } __initcall(ptdump_init); diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c index 42f585379e19..b75eada23d0a 100644 --- a/arch/arm/mm/fault.c +++ b/arch/arm/mm/fault.c @@ -21,7 +21,6 @@ #include <linux/highmem.h> #include <linux/perf_event.h> -#include <asm/exception.h> #include <asm/pgtable.h> #include <asm/system_misc.h> #include <asm/system_info.h> @@ -545,7 +544,7 @@ hook_fault_code(int nr, int (*fn)(unsigned long, unsigned int, struct pt_regs *) /* * Dispatch a data abort to the relevant handler. */ -asmlinkage void __exception +asmlinkage void do_DataAbort(unsigned long addr, unsigned int fsr, struct pt_regs *regs) { const struct fsr_info *inf = fsr_info + fsr_fs(fsr); @@ -578,7 +577,7 @@ hook_ifault_code(int nr, int (*fn)(unsigned long, unsigned int, struct pt_regs * ifsr_info[nr].name = name; } -asmlinkage void __exception +asmlinkage void do_PrefetchAbort(unsigned long addr, unsigned int ifsr, struct pt_regs *regs) { const struct fsr_info *inf = ifsr_info + fsr_fs(ifsr); diff --git a/arch/arm/mm/idmap.c b/arch/arm/mm/idmap.c index 10bfba85eb96..1d1edd064199 100644 --- a/arch/arm/mm/idmap.c +++ b/arch/arm/mm/idmap.c @@ -16,8 +16,8 @@ * are not supported on any CPU using the idmap tables as its current * page tables. */ -pgd_t *idmap_pgd; -long long arch_phys_to_idmap_offset; +pgd_t *idmap_pgd __ro_after_init; +long long arch_phys_to_idmap_offset __ro_after_init; #ifdef CONFIG_ARM_LPAE static void idmap_add_pmd(pud_t *pud, unsigned long addr, unsigned long end, diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c index a1f11a7ee81b..bd6f4513539a 100644 --- a/arch/arm/mm/init.c +++ b/arch/arm/mm/init.c @@ -36,6 +36,7 @@ #include <asm/system_info.h> #include <asm/tlb.h> #include <asm/fixmap.h> +#include <asm/ptdump.h> #include <asm/mach/arch.h> #include <asm/mach/map.h> @@ -738,6 +739,7 @@ static int __mark_rodata_ro(void *unused) void mark_rodata_ro(void) { stop_machine(__mark_rodata_ro, NULL, NULL); + debug_checkwx(); } void set_kernel_text_rw(void) diff --git a/arch/arm/mm/nommu.c b/arch/arm/mm/nommu.c index e4370810f4f1..7c087961b7ce 100644 --- a/arch/arm/mm/nommu.c +++ b/arch/arm/mm/nommu.c @@ -31,7 +31,7 @@ struct mpu_rgn_info mpu_rgn_info; #ifdef CONFIG_CPU_CP15 #ifdef CONFIG_CPU_HIGH_VECTOR -static unsigned long __init setup_vectors_base(void) +unsigned long setup_vectors_base(void) { unsigned long reg = get_cr(); @@ -57,7 +57,7 @@ static inline bool security_extensions_enabled(void) return 0; } -static unsigned long __init setup_vectors_base(void) +unsigned long setup_vectors_base(void) { unsigned long base = 0, reg = get_cr(); diff --git a/arch/arm/mm/pmsa-v7.c b/arch/arm/mm/pmsa-v7.c index 976df60ac426..e2853bfff74e 100644 --- a/arch/arm/mm/pmsa-v7.c +++ b/arch/arm/mm/pmsa-v7.c @@ -6,6 +6,7 @@ #include <linux/bitops.h> #include <linux/memblock.h> +#include <linux/string.h> #include <asm/cacheflush.h> #include <asm/cp15.h> @@ -296,6 +297,7 @@ void __init adjust_lowmem_bounds_mpu(void) } } + memset(mem, 0, sizeof(mem)); num = allocate_region(mem_start, specified_mem_size, mem_max_regions, mem); for (i = 0; i < num; i++) { @@ -433,7 +435,7 @@ void __init mpu_setup(void) /* Background */ err |= mpu_setup_region(region++, 0, 32, - MPU_ACR_XN | MPU_RGN_STRONGLY_ORDERED | MPU_AP_PL1RW_PL0NA, + MPU_ACR_XN | MPU_RGN_STRONGLY_ORDERED | MPU_AP_PL1RW_PL0RW, 0, false); #ifdef CONFIG_XIP_KERNEL diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S index 01d64c0b2563..d55d493f9a1e 100644 --- a/arch/arm/mm/proc-v7.S +++ b/arch/arm/mm/proc-v7.S @@ -567,7 +567,7 @@ __v7_setup_stack: /* * Standard v7 proc info content */ -.macro __v7_proc name, initfunc, mm_mmuflags = 0, io_mmuflags = 0, hwcaps = 0, proc_fns = v7_processor_functions +.macro __v7_proc name, initfunc, mm_mmuflags = 0, io_mmuflags = 0, hwcaps = 0, proc_fns = v7_processor_functions, cache_fns = v7_cache_fns ALT_SMP(.long PMD_TYPE_SECT | PMD_SECT_AP_WRITE | PMD_SECT_AP_READ | \ PMD_SECT_AF | PMD_FLAGS_SMP | \mm_mmuflags) ALT_UP(.long PMD_TYPE_SECT | PMD_SECT_AP_WRITE | PMD_SECT_AP_READ | \ @@ -583,7 +583,7 @@ __v7_setup_stack: .long \proc_fns .long v7wbi_tlb_fns .long v6_user_fns - .long v7_cache_fns + .long \cache_fns .endm #ifndef CONFIG_ARM_LPAE @@ -678,7 +678,7 @@ __v7_ca15mp_proc_info: __v7_b15mp_proc_info: .long 0x420f00f0 .long 0xff0ffff0 - __v7_proc __v7_b15mp_proc_info, __v7_b15mp_setup + __v7_proc __v7_b15mp_proc_info, __v7_b15mp_setup, cache_fns = b15_cache_fns .size __v7_b15mp_proc_info, . - __v7_b15mp_proc_info /* diff --git a/arch/arm/mm/ptdump_debugfs.c b/arch/arm/mm/ptdump_debugfs.c new file mode 100644 index 000000000000..be8d87be4b93 --- /dev/null +++ b/arch/arm/mm/ptdump_debugfs.c @@ -0,0 +1,34 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/debugfs.h> +#include <linux/seq_file.h> + +#include <asm/ptdump.h> + +static int ptdump_show(struct seq_file *m, void *v) +{ + struct ptdump_info *info = m->private; + + ptdump_walk_pgd(m, info); + return 0; +} + +static int ptdump_open(struct inode *inode, struct file *file) +{ + return single_open(file, ptdump_show, inode->i_private); +} + +static const struct file_operations ptdump_fops = { + .open = ptdump_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +int ptdump_debugfs_register(struct ptdump_info *info, const char *name) +{ + struct dentry *pe; + + pe = debugfs_create_file(name, 0400, NULL, info, &ptdump_fops); + return pe ? 0 : -ENOMEM; + +} diff --git a/arch/arm/net/bpf_jit_32.c b/arch/arm/net/bpf_jit_32.c index c199990e12b6..323a4df59a6c 100644 --- a/arch/arm/net/bpf_jit_32.c +++ b/arch/arm/net/bpf_jit_32.c @@ -27,14 +27,58 @@ int bpf_jit_enable __read_mostly; +/* + * eBPF prog stack layout: + * + * high + * original ARM_SP => +-----+ + * | | callee saved registers + * +-----+ <= (BPF_FP + SCRATCH_SIZE) + * | ... | eBPF JIT scratch space + * eBPF fp register => +-----+ + * (BPF_FP) | ... | eBPF prog stack + * +-----+ + * |RSVD | JIT scratchpad + * current ARM_SP => +-----+ <= (BPF_FP - STACK_SIZE + SCRATCH_SIZE) + * | | + * | ... | Function call stack + * | | + * +-----+ + * low + * + * The callee saved registers depends on whether frame pointers are enabled. + * With frame pointers (to be compliant with the ABI): + * + * high + * original ARM_SP => +------------------+ \ + * | pc | | + * current ARM_FP => +------------------+ } callee saved registers + * |r4-r8,r10,fp,ip,lr| | + * +------------------+ / + * low + * + * Without frame pointers: + * + * high + * original ARM_SP => +------------------+ + * | r4-r8,r10,fp,lr | callee saved registers + * current ARM_FP => +------------------+ + * low + * + * When popping registers off the stack at the end of a BPF function, we + * reference them via the current ARM_FP register. + */ +#define CALLEE_MASK (1 << ARM_R4 | 1 << ARM_R5 | 1 << ARM_R6 | \ + 1 << ARM_R7 | 1 << ARM_R8 | 1 << ARM_R10 | \ + 1 << ARM_FP) +#define CALLEE_PUSH_MASK (CALLEE_MASK | 1 << ARM_LR) +#define CALLEE_POP_MASK (CALLEE_MASK | 1 << ARM_PC) + #define STACK_OFFSET(k) (k) #define TMP_REG_1 (MAX_BPF_JIT_REG + 0) /* TEMP Register 1 */ #define TMP_REG_2 (MAX_BPF_JIT_REG + 1) /* TEMP Register 2 */ #define TCALL_CNT (MAX_BPF_JIT_REG + 2) /* Tail Call Count */ -/* Flags used for JIT optimization */ -#define SEEN_CALL (1 << 0) - #define FLAG_IMM_OVERFLOW (1 << 0) /* @@ -95,7 +139,6 @@ static const u8 bpf2a32[][2] = { * idx : index of current last JITed instruction. * prologue_bytes : bytes used in prologue. * epilogue_offset : offset of epilogue starting. - * seen : bit mask used for JIT optimization. * offsets : array of eBPF instruction offsets in * JITed code. * target : final JITed code. @@ -110,7 +153,6 @@ struct jit_ctx { unsigned int idx; unsigned int prologue_bytes; unsigned int epilogue_offset; - u32 seen; u32 flags; u32 *offsets; u32 *target; @@ -179,8 +221,13 @@ static void jit_fill_hole(void *area, unsigned int size) *ptr++ = __opcode_to_mem_arm(ARM_INST_UDF); } -/* Stack must be multiples of 16 Bytes */ -#define STACK_ALIGN(sz) (((sz) + 3) & ~3) +#if defined(CONFIG_AEABI) && (__LINUX_ARM_ARCH__ >= 5) +/* EABI requires the stack to be aligned to 64-bit boundaries */ +#define STACK_ALIGNMENT 8 +#else +/* Stack must be aligned to 32-bit boundaries */ +#define STACK_ALIGNMENT 4 +#endif /* Stack space for BPF_REG_2, BPF_REG_3, BPF_REG_4, * BPF_REG_5, BPF_REG_7, BPF_REG_8, BPF_REG_9, @@ -194,7 +241,7 @@ static void jit_fill_hole(void *area, unsigned int size) + SCRATCH_SIZE + \ + 4 /* extra for skb_copy_bits buffer */) -#define STACK_SIZE STACK_ALIGN(_STACK_SIZE) +#define STACK_SIZE ALIGN(_STACK_SIZE, STACK_ALIGNMENT) /* Get the offset of eBPF REGISTERs stored on scratch space. */ #define STACK_VAR(off) (STACK_SIZE-off-4) @@ -285,16 +332,19 @@ static inline void emit_mov_i(const u8 rd, u32 val, struct jit_ctx *ctx) emit_mov_i_no8m(rd, val, ctx); } -static inline void emit_blx_r(u8 tgt_reg, struct jit_ctx *ctx) +static void emit_bx_r(u8 tgt_reg, struct jit_ctx *ctx) { - ctx->seen |= SEEN_CALL; -#if __LINUX_ARM_ARCH__ < 5 - emit(ARM_MOV_R(ARM_LR, ARM_PC), ctx); - if (elf_hwcap & HWCAP_THUMB) emit(ARM_BX(tgt_reg), ctx); else emit(ARM_MOV_R(ARM_PC, tgt_reg), ctx); +} + +static inline void emit_blx_r(u8 tgt_reg, struct jit_ctx *ctx) +{ +#if __LINUX_ARM_ARCH__ < 5 + emit(ARM_MOV_R(ARM_LR, ARM_PC), ctx); + emit_bx_r(tgt_reg, ctx); #else emit(ARM_BLX_R(tgt_reg), ctx); #endif @@ -354,7 +404,6 @@ static inline void emit_udivmod(u8 rd, u8 rm, u8 rn, struct jit_ctx *ctx, u8 op) } /* Call appropriate function */ - ctx->seen |= SEEN_CALL; emit_mov_i(ARM_IP, op == BPF_DIV ? (u32)jit_udiv32 : (u32)jit_mod32, ctx); emit_blx_r(ARM_IP, ctx); @@ -620,8 +669,6 @@ static inline void emit_a32_lsh_r64(const u8 dst[], const u8 src[], bool dstk, /* Do LSH operation */ emit(ARM_SUB_I(ARM_IP, rt, 32), ctx); emit(ARM_RSB_I(tmp2[0], rt, 32), ctx); - /* As we are using ARM_LR */ - ctx->seen |= SEEN_CALL; emit(ARM_MOV_SR(ARM_LR, rm, SRTYPE_ASL, rt), ctx); emit(ARM_ORR_SR(ARM_LR, ARM_LR, rd, SRTYPE_ASL, ARM_IP), ctx); emit(ARM_ORR_SR(ARM_IP, ARM_LR, rd, SRTYPE_LSR, tmp2[0]), ctx); @@ -656,8 +703,6 @@ static inline void emit_a32_arsh_r64(const u8 dst[], const u8 src[], bool dstk, /* Do the ARSH operation */ emit(ARM_RSB_I(ARM_IP, rt, 32), ctx); emit(ARM_SUBS_I(tmp2[0], rt, 32), ctx); - /* As we are using ARM_LR */ - ctx->seen |= SEEN_CALL; emit(ARM_MOV_SR(ARM_LR, rd, SRTYPE_LSR, rt), ctx); emit(ARM_ORR_SR(ARM_LR, ARM_LR, rm, SRTYPE_ASL, ARM_IP), ctx); _emit(ARM_COND_MI, ARM_B(0), ctx); @@ -692,8 +737,6 @@ static inline void emit_a32_lsr_r64(const u8 dst[], const u8 src[], bool dstk, /* Do LSH operation */ emit(ARM_RSB_I(ARM_IP, rt, 32), ctx); emit(ARM_SUBS_I(tmp2[0], rt, 32), ctx); - /* As we are using ARM_LR */ - ctx->seen |= SEEN_CALL; emit(ARM_MOV_SR(ARM_LR, rd, SRTYPE_LSR, rt), ctx); emit(ARM_ORR_SR(ARM_LR, ARM_LR, rm, SRTYPE_ASL, ARM_IP), ctx); emit(ARM_ORR_SR(ARM_LR, ARM_LR, rm, SRTYPE_LSR, tmp2[0]), ctx); @@ -828,8 +871,6 @@ static inline void emit_a32_mul_r64(const u8 dst[], const u8 src[], bool dstk, /* Do Multiplication */ emit(ARM_MUL(ARM_IP, rd, rn), ctx); emit(ARM_MUL(ARM_LR, rm, rt), ctx); - /* As we are using ARM_LR */ - ctx->seen |= SEEN_CALL; emit(ARM_ADD_R(ARM_LR, ARM_IP, ARM_LR), ctx); emit(ARM_UMULL(ARM_IP, rm, rd, rt), ctx); @@ -872,33 +913,53 @@ static inline void emit_str_r(const u8 dst, const u8 src, bool dstk, } /* dst = *(size*)(src + off) */ -static inline void emit_ldx_r(const u8 dst, const u8 src, bool dstk, - const s32 off, struct jit_ctx *ctx, const u8 sz){ +static inline void emit_ldx_r(const u8 dst[], const u8 src, bool dstk, + s32 off, struct jit_ctx *ctx, const u8 sz){ const u8 *tmp = bpf2a32[TMP_REG_1]; - u8 rd = dstk ? tmp[1] : dst; + const u8 *rd = dstk ? tmp : dst; u8 rm = src; + s32 off_max; - if (off) { + if (sz == BPF_H) + off_max = 0xff; + else + off_max = 0xfff; + + if (off < 0 || off > off_max) { emit_a32_mov_i(tmp[0], off, false, ctx); emit(ARM_ADD_R(tmp[0], tmp[0], src), ctx); rm = tmp[0]; + off = 0; + } else if (rd[1] == rm) { + emit(ARM_MOV_R(tmp[0], rm), ctx); + rm = tmp[0]; } switch (sz) { - case BPF_W: - /* Load a Word */ - emit(ARM_LDR_I(rd, rm, 0), ctx); + case BPF_B: + /* Load a Byte */ + emit(ARM_LDRB_I(rd[1], rm, off), ctx); + emit_a32_mov_i(dst[0], 0, dstk, ctx); break; case BPF_H: /* Load a HalfWord */ - emit(ARM_LDRH_I(rd, rm, 0), ctx); + emit(ARM_LDRH_I(rd[1], rm, off), ctx); + emit_a32_mov_i(dst[0], 0, dstk, ctx); break; - case BPF_B: - /* Load a Byte */ - emit(ARM_LDRB_I(rd, rm, 0), ctx); + case BPF_W: + /* Load a Word */ + emit(ARM_LDR_I(rd[1], rm, off), ctx); + emit_a32_mov_i(dst[0], 0, dstk, ctx); + break; + case BPF_DW: + /* Load a Double Word */ + emit(ARM_LDR_I(rd[1], rm, off), ctx); + emit(ARM_LDR_I(rd[0], rm, off + 4), ctx); break; } if (dstk) - emit(ARM_STR_I(rd, ARM_SP, STACK_VAR(dst)), ctx); + emit(ARM_STR_I(rd[1], ARM_SP, STACK_VAR(dst[1])), ctx); + if (dstk && sz == BPF_DW) + emit(ARM_STR_I(rd[0], ARM_SP, STACK_VAR(dst[0])), ctx); } /* Arithmatic Operation */ @@ -906,7 +967,6 @@ static inline void emit_ar_r(const u8 rd, const u8 rt, const u8 rm, const u8 rn, struct jit_ctx *ctx, u8 op) { switch (op) { case BPF_JSET: - ctx->seen |= SEEN_CALL; emit(ARM_AND_R(ARM_IP, rt, rn), ctx); emit(ARM_AND_R(ARM_LR, rd, rm), ctx); emit(ARM_ORRS_R(ARM_IP, ARM_LR, ARM_IP), ctx); @@ -945,7 +1005,7 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) const u8 *tcc = bpf2a32[TCALL_CNT]; const int idx0 = ctx->idx; #define cur_offset (ctx->idx - idx0) -#define jmp_offset (out_offset - (cur_offset)) +#define jmp_offset (out_offset - (cur_offset) - 2) u32 off, lo, hi; /* if (index >= array->map.max_entries) @@ -956,7 +1016,7 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) emit_a32_mov_i(tmp[1], off, false, ctx); emit(ARM_LDR_I(tmp2[1], ARM_SP, STACK_VAR(r2[1])), ctx); emit(ARM_LDR_R(tmp[1], tmp2[1], tmp[1]), ctx); - /* index (64 bit) */ + /* index is 32-bit for arrays */ emit(ARM_LDR_I(tmp2[1], ARM_SP, STACK_VAR(r3[1])), ctx); /* index >= array->map.max_entries */ emit(ARM_CMP_R(tmp2[1], tmp[1]), ctx); @@ -997,7 +1057,7 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) emit_a32_mov_i(tmp2[1], off, false, ctx); emit(ARM_LDR_R(tmp[1], tmp[1], tmp2[1]), ctx); emit(ARM_ADD_I(tmp[1], tmp[1], ctx->prologue_bytes), ctx); - emit(ARM_BX(tmp[1]), ctx); + emit_bx_r(tmp[1], ctx); /* out: */ if (out_offset == -1) @@ -1070,54 +1130,22 @@ static void build_prologue(struct jit_ctx *ctx) const u8 r2 = bpf2a32[BPF_REG_1][1]; const u8 r3 = bpf2a32[BPF_REG_1][0]; const u8 r4 = bpf2a32[BPF_REG_6][1]; - const u8 r5 = bpf2a32[BPF_REG_6][0]; - const u8 r6 = bpf2a32[TMP_REG_1][1]; - const u8 r7 = bpf2a32[TMP_REG_1][0]; - const u8 r8 = bpf2a32[TMP_REG_2][1]; - const u8 r10 = bpf2a32[TMP_REG_2][0]; const u8 fplo = bpf2a32[BPF_REG_FP][1]; const u8 fphi = bpf2a32[BPF_REG_FP][0]; - const u8 sp = ARM_SP; const u8 *tcc = bpf2a32[TCALL_CNT]; - u16 reg_set = 0; - - /* - * eBPF prog stack layout - * - * high - * original ARM_SP => +-----+ eBPF prologue - * |FP/LR| - * current ARM_FP => +-----+ - * | ... | callee saved registers - * eBPF fp register => +-----+ <= (BPF_FP) - * | ... | eBPF JIT scratch space - * | | eBPF prog stack - * +-----+ - * |RSVD | JIT scratchpad - * current A64_SP => +-----+ <= (BPF_FP - STACK_SIZE) - * | | - * | ... | Function call stack - * | | - * +-----+ - * low - */ - /* Save callee saved registers. */ - reg_set |= (1<<r4) | (1<<r5) | (1<<r6) | (1<<r7) | (1<<r8) | (1<<r10); #ifdef CONFIG_FRAME_POINTER - reg_set |= (1<<ARM_FP) | (1<<ARM_IP) | (1<<ARM_LR) | (1<<ARM_PC); - emit(ARM_MOV_R(ARM_IP, sp), ctx); + u16 reg_set = CALLEE_PUSH_MASK | 1 << ARM_IP | 1 << ARM_PC; + emit(ARM_MOV_R(ARM_IP, ARM_SP), ctx); emit(ARM_PUSH(reg_set), ctx); emit(ARM_SUB_I(ARM_FP, ARM_IP, 4), ctx); #else - /* Check if call instruction exists in BPF body */ - if (ctx->seen & SEEN_CALL) - reg_set |= (1<<ARM_LR); - emit(ARM_PUSH(reg_set), ctx); + emit(ARM_PUSH(CALLEE_PUSH_MASK), ctx); + emit(ARM_MOV_R(ARM_FP, ARM_SP), ctx); #endif /* Save frame pointer for later */ - emit(ARM_SUB_I(ARM_IP, sp, SCRATCH_SIZE), ctx); + emit(ARM_SUB_I(ARM_IP, ARM_SP, SCRATCH_SIZE), ctx); ctx->stack_size = imm8m(STACK_SIZE); @@ -1140,33 +1168,19 @@ static void build_prologue(struct jit_ctx *ctx) /* end of prologue */ } +/* restore callee saved registers. */ static void build_epilogue(struct jit_ctx *ctx) { - const u8 r4 = bpf2a32[BPF_REG_6][1]; - const u8 r5 = bpf2a32[BPF_REG_6][0]; - const u8 r6 = bpf2a32[TMP_REG_1][1]; - const u8 r7 = bpf2a32[TMP_REG_1][0]; - const u8 r8 = bpf2a32[TMP_REG_2][1]; - const u8 r10 = bpf2a32[TMP_REG_2][0]; - u16 reg_set = 0; - - /* unwind function call stack */ - emit(ARM_ADD_I(ARM_SP, ARM_SP, ctx->stack_size), ctx); - - /* restore callee saved registers. */ - reg_set |= (1<<r4) | (1<<r5) | (1<<r6) | (1<<r7) | (1<<r8) | (1<<r10); #ifdef CONFIG_FRAME_POINTER - /* the first instruction of the prologue was: mov ip, sp */ - reg_set |= (1<<ARM_FP) | (1<<ARM_SP) | (1<<ARM_PC); + /* When using frame pointers, some additional registers need to + * be loaded. */ + u16 reg_set = CALLEE_POP_MASK | 1 << ARM_SP; + emit(ARM_SUB_I(ARM_SP, ARM_FP, hweight16(reg_set) * 4), ctx); emit(ARM_LDM(ARM_SP, reg_set), ctx); #else - if (ctx->seen & SEEN_CALL) - reg_set |= (1<<ARM_PC); /* Restore callee saved registers. */ - emit(ARM_POP(reg_set), ctx); - /* Return back to the callee function */ - if (!(ctx->seen & SEEN_CALL)) - emit(ARM_BX(ARM_LR), ctx); + emit(ARM_MOV_R(ARM_SP, ARM_FP), ctx); + emit(ARM_POP(CALLEE_POP_MASK), ctx); #endif } @@ -1394,8 +1408,6 @@ static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx) emit_rev32(rt, rt, ctx); goto emit_bswap_uxt; case 64: - /* Because of the usage of ARM_LR */ - ctx->seen |= SEEN_CALL; emit_rev32(ARM_LR, rt, ctx); emit_rev32(rt, rd, ctx); emit(ARM_MOV_R(rd, ARM_LR), ctx); @@ -1448,22 +1460,7 @@ exit: rn = sstk ? tmp2[1] : src_lo; if (sstk) emit(ARM_LDR_I(rn, ARM_SP, STACK_VAR(src_lo)), ctx); - switch (BPF_SIZE(code)) { - case BPF_W: - /* Load a Word */ - case BPF_H: - /* Load a Half-Word */ - case BPF_B: - /* Load a Byte */ - emit_ldx_r(dst_lo, rn, dstk, off, ctx, BPF_SIZE(code)); - emit_a32_mov_i(dst_hi, 0, dstk, ctx); - break; - case BPF_DW: - /* Load a double word */ - emit_ldx_r(dst_lo, rn, dstk, off, ctx, BPF_W); - emit_ldx_r(dst_hi, rn, dstk, off+4, ctx, BPF_W); - break; - } + emit_ldx_r(dst, rn, dstk, off, ctx, BPF_SIZE(code)); break; /* R0 = ntohx(*(size *)(((struct sk_buff *)R6)->data + imm)) */ case BPF_LD | BPF_ABS | BPF_W: diff --git a/arch/arm/probes/kprobes/core.c b/arch/arm/probes/kprobes/core.c index 52d1cd14fda4..e90cc8a08186 100644 --- a/arch/arm/probes/kprobes/core.c +++ b/arch/arm/probes/kprobes/core.c @@ -32,6 +32,7 @@ #include <linux/percpu.h> #include <linux/bug.h> #include <asm/patch.h> +#include <asm/sections.h> #include "../decode-arm.h" #include "../decode-thumb.h" @@ -64,9 +65,6 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p) int is; const struct decode_checker **checkers; - if (in_exception_text(addr)) - return -EINVAL; - #ifdef CONFIG_THUMB2_KERNEL thumb = true; addr &= ~1; /* Bit 0 would normally be set to indicate Thumb code */ @@ -680,3 +678,13 @@ int __init arch_init_kprobes() #endif return 0; } + +bool arch_within_kprobe_blacklist(unsigned long addr) +{ + void *a = (void *)addr; + + return __in_irqentry_text(addr) || + in_entry_text(addr) || + in_idmap_text(addr) || + memory_contains(__kprobes_text_start, __kprobes_text_end, a, 1); +} |