OSDN Git Service

LoongArch: Add misc common routines
authorHuacai Chen <chenhuacai@loongson.cn>
Tue, 31 May 2022 10:04:11 +0000 (18:04 +0800)
committerHuacai Chen <chenhuacai@loongson.cn>
Fri, 3 Jun 2022 12:09:28 +0000 (20:09 +0800)
Add some misc common routines for LoongArch, including: asm-offsets
routines, futex functions, i/o memory access functions, frame-buffer
functions, procfs information display, etc.

Reviewed-by: WANG Xuerui <git@xen0n.name>
Reviewed-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
arch/loongarch/include/asm/asm-offsets.h [new file with mode: 0644]
arch/loongarch/include/asm/fb.h [new file with mode: 0644]
arch/loongarch/include/asm/futex.h [new file with mode: 0644]
arch/loongarch/include/asm/io.h [new file with mode: 0644]
arch/loongarch/kernel/asm-offsets.c [new file with mode: 0644]
arch/loongarch/kernel/io.c [new file with mode: 0644]
arch/loongarch/kernel/proc.c [new file with mode: 0644]

diff --git a/arch/loongarch/include/asm/asm-offsets.h b/arch/loongarch/include/asm/asm-offsets.h
new file mode 100644 (file)
index 0000000..d9ad88d
--- /dev/null
@@ -0,0 +1,5 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (C) 2020-2022 Loongson Technology Corporation Limited
+ */
+#include <generated/asm-offsets.h>
diff --git a/arch/loongarch/include/asm/fb.h b/arch/loongarch/include/asm/fb.h
new file mode 100644 (file)
index 0000000..3116bde
--- /dev/null
@@ -0,0 +1,23 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (C) 2020-2022 Loongson Technology Corporation Limited
+ */
+#ifndef _ASM_FB_H_
+#define _ASM_FB_H_
+
+#include <linux/fb.h>
+#include <linux/fs.h>
+#include <asm/page.h>
+
+static inline void fb_pgprotect(struct file *file, struct vm_area_struct *vma,
+                               unsigned long off)
+{
+       vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
+}
+
+static inline int fb_is_primary_device(struct fb_info *info)
+{
+       return 0;
+}
+
+#endif /* _ASM_FB_H_ */
diff --git a/arch/loongarch/include/asm/futex.h b/arch/loongarch/include/asm/futex.h
new file mode 100644 (file)
index 0000000..b27d55f
--- /dev/null
@@ -0,0 +1,107 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (C) 2020-2022 Loongson Technology Corporation Limited
+ */
+#ifndef _ASM_FUTEX_H
+#define _ASM_FUTEX_H
+
+#include <linux/futex.h>
+#include <linux/uaccess.h>
+#include <asm/barrier.h>
+#include <asm/compiler.h>
+#include <asm/errno.h>
+
+#define __futex_atomic_op(insn, ret, oldval, uaddr, oparg)             \
+{                                                                      \
+       __asm__ __volatile__(                                           \
+       "1:     ll.w    %1, %4 # __futex_atomic_op\n"           \
+       "       " insn  "                               \n"     \
+       "2:     sc.w    $t0, %2                         \n"     \
+       "       beq     $t0, $zero, 1b                  \n"     \
+       "3:                                             \n"     \
+       "       .section .fixup,\"ax\"                  \n"     \
+       "4:     li.w    %0, %6                          \n"     \
+       "       b       3b                              \n"     \
+       "       .previous                               \n"     \
+       "       .section __ex_table,\"a\"               \n"     \
+       "       "__UA_ADDR "\t1b, 4b                    \n"     \
+       "       "__UA_ADDR "\t2b, 4b                    \n"     \
+       "       .previous                               \n"     \
+       : "=r" (ret), "=&r" (oldval),                           \
+         "=ZC" (*uaddr)                                        \
+       : "0" (0), "ZC" (*uaddr), "Jr" (oparg),                 \
+         "i" (-EFAULT)                                         \
+       : "memory", "t0");                                      \
+}
+
+static inline int
+arch_futex_atomic_op_inuser(int op, int oparg, int *oval, u32 __user *uaddr)
+{
+       int oldval = 0, ret = 0;
+
+       pagefault_disable();
+
+       switch (op) {
+       case FUTEX_OP_SET:
+               __futex_atomic_op("move $t0, %z5", ret, oldval, uaddr, oparg);
+               break;
+       case FUTEX_OP_ADD:
+               __futex_atomic_op("add.w $t0, %1, %z5", ret, oldval, uaddr, oparg);
+               break;
+       case FUTEX_OP_OR:
+               __futex_atomic_op("or   $t0, %1, %z5", ret, oldval, uaddr, oparg);
+               break;
+       case FUTEX_OP_ANDN:
+               __futex_atomic_op("and  $t0, %1, %z5", ret, oldval, uaddr, ~oparg);
+               break;
+       case FUTEX_OP_XOR:
+               __futex_atomic_op("xor  $t0, %1, %z5", ret, oldval, uaddr, oparg);
+               break;
+       default:
+               ret = -ENOSYS;
+       }
+
+       pagefault_enable();
+
+       if (!ret)
+               *oval = oldval;
+
+       return ret;
+}
+
+static inline int
+futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr, u32 oldval, u32 newval)
+{
+       int ret = 0;
+       u32 val = 0;
+
+       if (!access_ok(uaddr, sizeof(u32)))
+               return -EFAULT;
+
+       __asm__ __volatile__(
+       "# futex_atomic_cmpxchg_inatomic                        \n"
+       "1:     ll.w    %1, %3                                  \n"
+       "       bne     %1, %z4, 3f                             \n"
+       "       or      $t0, %z5, $zero                         \n"
+       "2:     sc.w    $t0, %2                                 \n"
+       "       beq     $zero, $t0, 1b                          \n"
+       "3:                                                     \n"
+       "       .section .fixup,\"ax\"                          \n"
+       "4:     li.d    %0, %6                                  \n"
+       "       b       3b                                      \n"
+       "       .previous                                       \n"
+       "       .section __ex_table,\"a\"                       \n"
+       "       "__UA_ADDR "\t1b, 4b                            \n"
+       "       "__UA_ADDR "\t2b, 4b                            \n"
+       "       .previous                                       \n"
+       : "+r" (ret), "=&r" (val), "=" GCC_OFF_SMALL_ASM() (*uaddr)
+       : GCC_OFF_SMALL_ASM() (*uaddr), "Jr" (oldval), "Jr" (newval),
+         "i" (-EFAULT)
+       : "memory", "t0");
+
+       *uval = val;
+
+       return ret;
+}
+
+#endif /* _ASM_FUTEX_H */
diff --git a/arch/loongarch/include/asm/io.h b/arch/loongarch/include/asm/io.h
new file mode 100644 (file)
index 0000000..8845997
--- /dev/null
@@ -0,0 +1,129 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (C) 2020-2022 Loongson Technology Corporation Limited
+ */
+#ifndef _ASM_IO_H
+#define _ASM_IO_H
+
+#define ARCH_HAS_IOREMAP_WC
+
+#include <linux/compiler.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+
+#include <asm/addrspace.h>
+#include <asm/bug.h>
+#include <asm/byteorder.h>
+#include <asm/cpu.h>
+#include <asm/page.h>
+#include <asm/pgtable-bits.h>
+#include <asm/string.h>
+
+/*
+ * On LoongArch, I/O ports mappring is following:
+ *
+ *              |         ....          |
+ *              |-----------------------|
+ *              | pci io ports(64K~32M) |
+ *              |-----------------------|
+ *              | isa io ports(0  ~16K) |
+ * PCI_IOBASE ->|-----------------------|
+ *              |         ....          |
+ */
+#define PCI_IOBASE     ((void __iomem *)(vm_map_base + (2 * PAGE_SIZE)))
+#define PCI_IOSIZE     SZ_32M
+#define ISA_IOSIZE     SZ_16K
+#define IO_SPACE_LIMIT (PCI_IOSIZE - 1)
+
+/*
+ * Change "struct page" to physical address.
+ */
+#define page_to_phys(page)     ((phys_addr_t)page_to_pfn(page) << PAGE_SHIFT)
+
+extern void __init __iomem *early_ioremap(u64 phys_addr, unsigned long size);
+extern void __init early_iounmap(void __iomem *addr, unsigned long size);
+
+#define early_memremap early_ioremap
+#define early_memunmap early_iounmap
+
+static inline void __iomem *ioremap_prot(phys_addr_t offset, unsigned long size,
+                                        unsigned long prot_val)
+{
+       if (prot_val == _CACHE_CC)
+               return (void __iomem *)(unsigned long)(CACHE_BASE + offset);
+       else
+               return (void __iomem *)(unsigned long)(UNCACHE_BASE + offset);
+}
+
+/*
+ * ioremap -   map bus memory into CPU space
+ * @offset:    bus address of the memory
+ * @size:      size of the resource to map
+ *
+ * ioremap performs a platform specific sequence of operations to
+ * make bus memory CPU accessible via the readb/readw/readl/writeb/
+ * writew/writel functions and the other mmio helpers. The returned
+ * address is not guaranteed to be usable directly as a virtual
+ * address.
+ */
+#define ioremap(offset, size)                                  \
+       ioremap_prot((offset), (size), _CACHE_SUC)
+
+/*
+ * ioremap_wc - map bus memory into CPU space
+ * @offset:     bus address of the memory
+ * @size:       size of the resource to map
+ *
+ * ioremap_wc performs a platform specific sequence of operations to
+ * make bus memory CPU accessible via the readb/readw/readl/writeb/
+ * writew/writel functions and the other mmio helpers. The returned
+ * address is not guaranteed to be usable directly as a virtual
+ * address.
+ *
+ * This version of ioremap ensures that the memory is marked uncachable
+ * but accelerated by means of write-combining feature. It is specifically
+ * useful for PCIe prefetchable windows, which may vastly improve a
+ * communications performance. If it was determined on boot stage, what
+ * CPU CCA doesn't support WUC, the method shall fall-back to the
+ * _CACHE_SUC option (see cpu_probe() method).
+ */
+#define ioremap_wc(offset, size)                               \
+       ioremap_prot((offset), (size), _CACHE_WUC)
+
+/*
+ * ioremap_cache -  map bus memory into CPU space
+ * @offset:        bus address of the memory
+ * @size:          size of the resource to map
+ *
+ * ioremap_cache performs a platform specific sequence of operations to
+ * make bus memory CPU accessible via the readb/readw/readl/writeb/
+ * writew/writel functions and the other mmio helpers. The returned
+ * address is not guaranteed to be usable directly as a virtual
+ * address.
+ *
+ * This version of ioremap ensures that the memory is marked cachable by
+ * the CPU.  Also enables full write-combining.         Useful for some
+ * memory-like regions on I/O busses.
+ */
+#define ioremap_cache(offset, size)                            \
+       ioremap_prot((offset), (size), _CACHE_CC)
+
+static inline void iounmap(const volatile void __iomem *addr)
+{
+}
+
+#define mmiowb() asm volatile ("dbar 0" ::: "memory")
+
+/*
+ * String version of I/O memory access operations.
+ */
+extern void __memset_io(volatile void __iomem *dst, int c, size_t count);
+extern void __memcpy_toio(volatile void __iomem *to, const void *from, size_t count);
+extern void __memcpy_fromio(void *to, const volatile void __iomem *from, size_t count);
+#define memset_io(c, v, l)     __memset_io((c), (v), (l))
+#define memcpy_fromio(a, c, l) __memcpy_fromio((a), (c), (l))
+#define memcpy_toio(c, a, l)   __memcpy_toio((c), (a), (l))
+
+#include <asm-generic/io.h>
+
+#endif /* _ASM_IO_H */
diff --git a/arch/loongarch/kernel/asm-offsets.c b/arch/loongarch/kernel/asm-offsets.c
new file mode 100644 (file)
index 0000000..3531e3c
--- /dev/null
@@ -0,0 +1,254 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * asm-offsets.c: Calculate pt_regs and task_struct offsets.
+ *
+ * Copyright (C) 2020-2022 Loongson Technology Corporation Limited
+ */
+#include <linux/types.h>
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <linux/kbuild.h>
+#include <linux/suspend.h>
+#include <asm/cpu-info.h>
+#include <asm/ptrace.h>
+#include <asm/processor.h>
+
+void output_ptreg_defines(void)
+{
+       COMMENT("LoongArch pt_regs offsets.");
+       OFFSET(PT_R0, pt_regs, regs[0]);
+       OFFSET(PT_R1, pt_regs, regs[1]);
+       OFFSET(PT_R2, pt_regs, regs[2]);
+       OFFSET(PT_R3, pt_regs, regs[3]);
+       OFFSET(PT_R4, pt_regs, regs[4]);
+       OFFSET(PT_R5, pt_regs, regs[5]);
+       OFFSET(PT_R6, pt_regs, regs[6]);
+       OFFSET(PT_R7, pt_regs, regs[7]);
+       OFFSET(PT_R8, pt_regs, regs[8]);
+       OFFSET(PT_R9, pt_regs, regs[9]);
+       OFFSET(PT_R10, pt_regs, regs[10]);
+       OFFSET(PT_R11, pt_regs, regs[11]);
+       OFFSET(PT_R12, pt_regs, regs[12]);
+       OFFSET(PT_R13, pt_regs, regs[13]);
+       OFFSET(PT_R14, pt_regs, regs[14]);
+       OFFSET(PT_R15, pt_regs, regs[15]);
+       OFFSET(PT_R16, pt_regs, regs[16]);
+       OFFSET(PT_R17, pt_regs, regs[17]);
+       OFFSET(PT_R18, pt_regs, regs[18]);
+       OFFSET(PT_R19, pt_regs, regs[19]);
+       OFFSET(PT_R20, pt_regs, regs[20]);
+       OFFSET(PT_R21, pt_regs, regs[21]);
+       OFFSET(PT_R22, pt_regs, regs[22]);
+       OFFSET(PT_R23, pt_regs, regs[23]);
+       OFFSET(PT_R24, pt_regs, regs[24]);
+       OFFSET(PT_R25, pt_regs, regs[25]);
+       OFFSET(PT_R26, pt_regs, regs[26]);
+       OFFSET(PT_R27, pt_regs, regs[27]);
+       OFFSET(PT_R28, pt_regs, regs[28]);
+       OFFSET(PT_R29, pt_regs, regs[29]);
+       OFFSET(PT_R30, pt_regs, regs[30]);
+       OFFSET(PT_R31, pt_regs, regs[31]);
+       OFFSET(PT_CRMD, pt_regs, csr_crmd);
+       OFFSET(PT_PRMD, pt_regs, csr_prmd);
+       OFFSET(PT_EUEN, pt_regs, csr_euen);
+       OFFSET(PT_ECFG, pt_regs, csr_ecfg);
+       OFFSET(PT_ESTAT, pt_regs, csr_estat);
+       OFFSET(PT_ERA, pt_regs, csr_era);
+       OFFSET(PT_BVADDR, pt_regs, csr_badvaddr);
+       OFFSET(PT_ORIG_A0, pt_regs, orig_a0);
+       DEFINE(PT_SIZE, sizeof(struct pt_regs));
+       BLANK();
+}
+
+void output_task_defines(void)
+{
+       COMMENT("LoongArch task_struct offsets.");
+       OFFSET(TASK_STATE, task_struct, __state);
+       OFFSET(TASK_THREAD_INFO, task_struct, stack);
+       OFFSET(TASK_FLAGS, task_struct, flags);
+       OFFSET(TASK_MM, task_struct, mm);
+       OFFSET(TASK_PID, task_struct, pid);
+       DEFINE(TASK_STRUCT_SIZE, sizeof(struct task_struct));
+       BLANK();
+}
+
+void output_thread_info_defines(void)
+{
+       COMMENT("LoongArch thread_info offsets.");
+       OFFSET(TI_TASK, thread_info, task);
+       OFFSET(TI_FLAGS, thread_info, flags);
+       OFFSET(TI_TP_VALUE, thread_info, tp_value);
+       OFFSET(TI_CPU, thread_info, cpu);
+       OFFSET(TI_PRE_COUNT, thread_info, preempt_count);
+       OFFSET(TI_REGS, thread_info, regs);
+       DEFINE(_THREAD_SIZE, THREAD_SIZE);
+       DEFINE(_THREAD_MASK, THREAD_MASK);
+       DEFINE(_IRQ_STACK_SIZE, IRQ_STACK_SIZE);
+       DEFINE(_IRQ_STACK_START, IRQ_STACK_START);
+       BLANK();
+}
+
+void output_thread_defines(void)
+{
+       COMMENT("LoongArch specific thread_struct offsets.");
+       OFFSET(THREAD_REG01, task_struct, thread.reg01);
+       OFFSET(THREAD_REG03, task_struct, thread.reg03);
+       OFFSET(THREAD_REG22, task_struct, thread.reg22);
+       OFFSET(THREAD_REG23, task_struct, thread.reg23);
+       OFFSET(THREAD_REG24, task_struct, thread.reg24);
+       OFFSET(THREAD_REG25, task_struct, thread.reg25);
+       OFFSET(THREAD_REG26, task_struct, thread.reg26);
+       OFFSET(THREAD_REG27, task_struct, thread.reg27);
+       OFFSET(THREAD_REG28, task_struct, thread.reg28);
+       OFFSET(THREAD_REG29, task_struct, thread.reg29);
+       OFFSET(THREAD_REG30, task_struct, thread.reg30);
+       OFFSET(THREAD_REG31, task_struct, thread.reg31);
+       OFFSET(THREAD_CSRCRMD, task_struct,
+              thread.csr_crmd);
+       OFFSET(THREAD_CSRPRMD, task_struct,
+              thread.csr_prmd);
+       OFFSET(THREAD_CSREUEN, task_struct,
+              thread.csr_euen);
+       OFFSET(THREAD_CSRECFG, task_struct,
+              thread.csr_ecfg);
+
+       OFFSET(THREAD_SCR0, task_struct, thread.scr0);
+       OFFSET(THREAD_SCR1, task_struct, thread.scr1);
+       OFFSET(THREAD_SCR2, task_struct, thread.scr2);
+       OFFSET(THREAD_SCR3, task_struct, thread.scr3);
+
+       OFFSET(THREAD_EFLAGS, task_struct, thread.eflags);
+
+       OFFSET(THREAD_FPU, task_struct, thread.fpu);
+
+       OFFSET(THREAD_BVADDR, task_struct, \
+              thread.csr_badvaddr);
+       OFFSET(THREAD_ECODE, task_struct, \
+              thread.error_code);
+       OFFSET(THREAD_TRAPNO, task_struct, thread.trap_nr);
+       BLANK();
+}
+
+void output_thread_fpu_defines(void)
+{
+       OFFSET(THREAD_FPR0, loongarch_fpu, fpr[0]);
+       OFFSET(THREAD_FPR1, loongarch_fpu, fpr[1]);
+       OFFSET(THREAD_FPR2, loongarch_fpu, fpr[2]);
+       OFFSET(THREAD_FPR3, loongarch_fpu, fpr[3]);
+       OFFSET(THREAD_FPR4, loongarch_fpu, fpr[4]);
+       OFFSET(THREAD_FPR5, loongarch_fpu, fpr[5]);
+       OFFSET(THREAD_FPR6, loongarch_fpu, fpr[6]);
+       OFFSET(THREAD_FPR7, loongarch_fpu, fpr[7]);
+       OFFSET(THREAD_FPR8, loongarch_fpu, fpr[8]);
+       OFFSET(THREAD_FPR9, loongarch_fpu, fpr[9]);
+       OFFSET(THREAD_FPR10, loongarch_fpu, fpr[10]);
+       OFFSET(THREAD_FPR11, loongarch_fpu, fpr[11]);
+       OFFSET(THREAD_FPR12, loongarch_fpu, fpr[12]);
+       OFFSET(THREAD_FPR13, loongarch_fpu, fpr[13]);
+       OFFSET(THREAD_FPR14, loongarch_fpu, fpr[14]);
+       OFFSET(THREAD_FPR15, loongarch_fpu, fpr[15]);
+       OFFSET(THREAD_FPR16, loongarch_fpu, fpr[16]);
+       OFFSET(THREAD_FPR17, loongarch_fpu, fpr[17]);
+       OFFSET(THREAD_FPR18, loongarch_fpu, fpr[18]);
+       OFFSET(THREAD_FPR19, loongarch_fpu, fpr[19]);
+       OFFSET(THREAD_FPR20, loongarch_fpu, fpr[20]);
+       OFFSET(THREAD_FPR21, loongarch_fpu, fpr[21]);
+       OFFSET(THREAD_FPR22, loongarch_fpu, fpr[22]);
+       OFFSET(THREAD_FPR23, loongarch_fpu, fpr[23]);
+       OFFSET(THREAD_FPR24, loongarch_fpu, fpr[24]);
+       OFFSET(THREAD_FPR25, loongarch_fpu, fpr[25]);
+       OFFSET(THREAD_FPR26, loongarch_fpu, fpr[26]);
+       OFFSET(THREAD_FPR27, loongarch_fpu, fpr[27]);
+       OFFSET(THREAD_FPR28, loongarch_fpu, fpr[28]);
+       OFFSET(THREAD_FPR29, loongarch_fpu, fpr[29]);
+       OFFSET(THREAD_FPR30, loongarch_fpu, fpr[30]);
+       OFFSET(THREAD_FPR31, loongarch_fpu, fpr[31]);
+
+       OFFSET(THREAD_FCSR, loongarch_fpu, fcsr);
+       OFFSET(THREAD_FCC,  loongarch_fpu, fcc);
+       OFFSET(THREAD_VCSR, loongarch_fpu, vcsr);
+       BLANK();
+}
+
+void output_mm_defines(void)
+{
+       COMMENT("Size of struct page");
+       DEFINE(STRUCT_PAGE_SIZE, sizeof(struct page));
+       BLANK();
+       COMMENT("Linux mm_struct offsets.");
+       OFFSET(MM_USERS, mm_struct, mm_users);
+       OFFSET(MM_PGD, mm_struct, pgd);
+       OFFSET(MM_CONTEXT, mm_struct, context);
+       BLANK();
+       DEFINE(_PGD_T_SIZE, sizeof(pgd_t));
+       DEFINE(_PMD_T_SIZE, sizeof(pmd_t));
+       DEFINE(_PTE_T_SIZE, sizeof(pte_t));
+       BLANK();
+       DEFINE(_PGD_T_LOG2, PGD_T_LOG2);
+#ifndef __PAGETABLE_PMD_FOLDED
+       DEFINE(_PMD_T_LOG2, PMD_T_LOG2);
+#endif
+       DEFINE(_PTE_T_LOG2, PTE_T_LOG2);
+       BLANK();
+       DEFINE(_PGD_ORDER, PGD_ORDER);
+#ifndef __PAGETABLE_PMD_FOLDED
+       DEFINE(_PMD_ORDER, PMD_ORDER);
+#endif
+       DEFINE(_PTE_ORDER, PTE_ORDER);
+       BLANK();
+       DEFINE(_PMD_SHIFT, PMD_SHIFT);
+       DEFINE(_PGDIR_SHIFT, PGDIR_SHIFT);
+       BLANK();
+       DEFINE(_PTRS_PER_PGD, PTRS_PER_PGD);
+       DEFINE(_PTRS_PER_PMD, PTRS_PER_PMD);
+       DEFINE(_PTRS_PER_PTE, PTRS_PER_PTE);
+       BLANK();
+       DEFINE(_PAGE_SHIFT, PAGE_SHIFT);
+       DEFINE(_PAGE_SIZE, PAGE_SIZE);
+       BLANK();
+}
+
+void output_sc_defines(void)
+{
+       COMMENT("Linux sigcontext offsets.");
+       OFFSET(SC_REGS, sigcontext, sc_regs);
+       OFFSET(SC_PC, sigcontext, sc_pc);
+       BLANK();
+}
+
+void output_signal_defines(void)
+{
+       COMMENT("Linux signal numbers.");
+       DEFINE(_SIGHUP, SIGHUP);
+       DEFINE(_SIGINT, SIGINT);
+       DEFINE(_SIGQUIT, SIGQUIT);
+       DEFINE(_SIGILL, SIGILL);
+       DEFINE(_SIGTRAP, SIGTRAP);
+       DEFINE(_SIGIOT, SIGIOT);
+       DEFINE(_SIGABRT, SIGABRT);
+       DEFINE(_SIGFPE, SIGFPE);
+       DEFINE(_SIGKILL, SIGKILL);
+       DEFINE(_SIGBUS, SIGBUS);
+       DEFINE(_SIGSEGV, SIGSEGV);
+       DEFINE(_SIGSYS, SIGSYS);
+       DEFINE(_SIGPIPE, SIGPIPE);
+       DEFINE(_SIGALRM, SIGALRM);
+       DEFINE(_SIGTERM, SIGTERM);
+       DEFINE(_SIGUSR1, SIGUSR1);
+       DEFINE(_SIGUSR2, SIGUSR2);
+       DEFINE(_SIGCHLD, SIGCHLD);
+       DEFINE(_SIGPWR, SIGPWR);
+       DEFINE(_SIGWINCH, SIGWINCH);
+       DEFINE(_SIGURG, SIGURG);
+       DEFINE(_SIGIO, SIGIO);
+       DEFINE(_SIGSTOP, SIGSTOP);
+       DEFINE(_SIGTSTP, SIGTSTP);
+       DEFINE(_SIGCONT, SIGCONT);
+       DEFINE(_SIGTTIN, SIGTTIN);
+       DEFINE(_SIGTTOU, SIGTTOU);
+       DEFINE(_SIGVTALRM, SIGVTALRM);
+       DEFINE(_SIGPROF, SIGPROF);
+       DEFINE(_SIGXCPU, SIGXCPU);
+       DEFINE(_SIGXFSZ, SIGXFSZ);
+       BLANK();
+}
diff --git a/arch/loongarch/kernel/io.c b/arch/loongarch/kernel/io.c
new file mode 100644 (file)
index 0000000..cb85bda
--- /dev/null
@@ -0,0 +1,94 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020-2022 Loongson Technology Corporation Limited
+ */
+#include <linux/export.h>
+#include <linux/types.h>
+#include <linux/io.h>
+
+/*
+ * Copy data from IO memory space to "real" memory space.
+ */
+void __memcpy_fromio(void *to, const volatile void __iomem *from, size_t count)
+{
+       while (count && !IS_ALIGNED((unsigned long)from, 8)) {
+               *(u8 *)to = __raw_readb(from);
+               from++;
+               to++;
+               count--;
+       }
+
+       while (count >= 8) {
+               *(u64 *)to = __raw_readq(from);
+               from += 8;
+               to += 8;
+               count -= 8;
+       }
+
+       while (count) {
+               *(u8 *)to = __raw_readb(from);
+               from++;
+               to++;
+               count--;
+       }
+}
+EXPORT_SYMBOL(__memcpy_fromio);
+
+/*
+ * Copy data from "real" memory space to IO memory space.
+ */
+void __memcpy_toio(volatile void __iomem *to, const void *from, size_t count)
+{
+       while (count && !IS_ALIGNED((unsigned long)to, 8)) {
+               __raw_writeb(*(u8 *)from, to);
+               from++;
+               to++;
+               count--;
+       }
+
+       while (count >= 8) {
+               __raw_writeq(*(u64 *)from, to);
+               from += 8;
+               to += 8;
+               count -= 8;
+       }
+
+       while (count) {
+               __raw_writeb(*(u8 *)from, to);
+               from++;
+               to++;
+               count--;
+       }
+}
+EXPORT_SYMBOL(__memcpy_toio);
+
+/*
+ * "memset" on IO memory space.
+ */
+void __memset_io(volatile void __iomem *dst, int c, size_t count)
+{
+       u64 qc = (u8)c;
+
+       qc |= qc << 8;
+       qc |= qc << 16;
+       qc |= qc << 32;
+
+       while (count && !IS_ALIGNED((unsigned long)dst, 8)) {
+               __raw_writeb(c, dst);
+               dst++;
+               count--;
+       }
+
+       while (count >= 8) {
+               __raw_writeq(qc, dst);
+               dst += 8;
+               count -= 8;
+       }
+
+       while (count) {
+               __raw_writeb(c, dst);
+               dst++;
+               count--;
+       }
+}
+EXPORT_SYMBOL(__memset_io);
diff --git a/arch/loongarch/kernel/proc.c b/arch/loongarch/kernel/proc.c
new file mode 100644 (file)
index 0000000..d25592a
--- /dev/null
@@ -0,0 +1,122 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020-2022 Loongson Technology Corporation Limited
+ */
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/seq_file.h>
+#include <asm/bootinfo.h>
+#include <asm/cpu.h>
+#include <asm/cpu-features.h>
+#include <asm/idle.h>
+#include <asm/processor.h>
+#include <asm/time.h>
+
+/*
+ * No lock; only written during early bootup by CPU 0.
+ */
+static RAW_NOTIFIER_HEAD(proc_cpuinfo_chain);
+
+int __ref register_proc_cpuinfo_notifier(struct notifier_block *nb)
+{
+       return raw_notifier_chain_register(&proc_cpuinfo_chain, nb);
+}
+
+int proc_cpuinfo_notifier_call_chain(unsigned long val, void *v)
+{
+       return raw_notifier_call_chain(&proc_cpuinfo_chain, val, v);
+}
+
+static int show_cpuinfo(struct seq_file *m, void *v)
+{
+       unsigned long n = (unsigned long) v - 1;
+       unsigned int version = cpu_data[n].processor_id & 0xff;
+       unsigned int fp_version = cpu_data[n].fpu_vers;
+       struct proc_cpuinfo_notifier_args proc_cpuinfo_notifier_args;
+
+       /*
+        * For the first processor also print the system type
+        */
+       if (n == 0)
+               seq_printf(m, "system type\t\t: %s\n\n", get_system_type());
+
+       seq_printf(m, "processor\t\t: %ld\n", n);
+       seq_printf(m, "package\t\t\t: %d\n", cpu_data[n].package);
+       seq_printf(m, "core\t\t\t: %d\n", cpu_data[n].core);
+       seq_printf(m, "CPU Family\t\t: %s\n", __cpu_family[n]);
+       seq_printf(m, "Model Name\t\t: %s\n", __cpu_full_name[n]);
+       seq_printf(m, "CPU Revision\t\t: 0x%02x\n", version);
+       seq_printf(m, "FPU Revision\t\t: 0x%02x\n", fp_version);
+       seq_printf(m, "CPU MHz\t\t\t: %llu.%02llu\n",
+                     cpu_clock_freq / 1000000, (cpu_clock_freq / 10000) % 100);
+       seq_printf(m, "BogoMIPS\t\t: %llu.%02llu\n",
+                     (lpj_fine * cpu_clock_freq / const_clock_freq) / (500000/HZ),
+                     ((lpj_fine * cpu_clock_freq / const_clock_freq) / (5000/HZ)) % 100);
+       seq_printf(m, "TLB Entries\t\t: %d\n", cpu_data[n].tlbsize);
+       seq_printf(m, "Address Sizes\t\t: %d bits physical, %d bits virtual\n",
+                     cpu_pabits + 1, cpu_vabits + 1);
+
+       seq_printf(m, "ISA\t\t\t:");
+       if (cpu_has_loongarch32)
+               seq_printf(m, " loongarch32");
+       if (cpu_has_loongarch64)
+               seq_printf(m, " loongarch64");
+       seq_printf(m, "\n");
+
+       seq_printf(m, "Features\t\t:");
+       if (cpu_has_cpucfg)     seq_printf(m, " cpucfg");
+       if (cpu_has_lam)        seq_printf(m, " lam");
+       if (cpu_has_ual)        seq_printf(m, " ual");
+       if (cpu_has_fpu)        seq_printf(m, " fpu");
+       if (cpu_has_lsx)        seq_printf(m, " lsx");
+       if (cpu_has_lasx)       seq_printf(m, " lasx");
+       if (cpu_has_complex)    seq_printf(m, " complex");
+       if (cpu_has_crypto)     seq_printf(m, " crypto");
+       if (cpu_has_lvz)        seq_printf(m, " lvz");
+       if (cpu_has_lbt_x86)    seq_printf(m, " lbt_x86");
+       if (cpu_has_lbt_arm)    seq_printf(m, " lbt_arm");
+       if (cpu_has_lbt_mips)   seq_printf(m, " lbt_mips");
+       seq_printf(m, "\n");
+
+       seq_printf(m, "Hardware Watchpoint\t: %s",
+                     cpu_has_watch ? "yes, " : "no\n");
+       if (cpu_has_watch) {
+               seq_printf(m, "iwatch count: %d, dwatch count: %d\n",
+                     cpu_data[n].watch_ireg_count, cpu_data[n].watch_dreg_count);
+       }
+
+       proc_cpuinfo_notifier_args.m = m;
+       proc_cpuinfo_notifier_args.n = n;
+
+       raw_notifier_call_chain(&proc_cpuinfo_chain, 0,
+                               &proc_cpuinfo_notifier_args);
+
+       seq_printf(m, "\n");
+
+       return 0;
+}
+
+static void *c_start(struct seq_file *m, loff_t *pos)
+{
+       unsigned long i = *pos;
+
+       return i < NR_CPUS ? (void *)(i + 1) : NULL;
+}
+
+static void *c_next(struct seq_file *m, void *v, loff_t *pos)
+{
+       ++*pos;
+       return c_start(m, pos);
+}
+
+static void c_stop(struct seq_file *m, void *v)
+{
+}
+
+const struct seq_operations cpuinfo_op = {
+       .start  = c_start,
+       .next   = c_next,
+       .stop   = c_stop,
+       .show   = show_cpuinfo,
+};