From 1da177e4c3f41524e886b7f1b8a0c1fc7321cac2 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sat, 16 Apr 2005 15:20:36 -0700 Subject: Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip! --- arch/ia64/ia32/Makefile | 12 + arch/ia64/ia32/binfmt_elf32.c | 294 +++++ arch/ia64/ia32/elfcore32.h | 138 +++ arch/ia64/ia32/ia32_entry.S | 500 ++++++++ arch/ia64/ia32/ia32_ioctl.c | 48 + arch/ia64/ia32/ia32_ldt.c | 147 +++ arch/ia64/ia32/ia32_signal.c | 1036 ++++++++++++++++ arch/ia64/ia32/ia32_support.c | 264 ++++ arch/ia64/ia32/ia32_traps.c | 156 +++ arch/ia64/ia32/ia32priv.h | 544 ++++++++ arch/ia64/ia32/sys_ia32.c | 2747 +++++++++++++++++++++++++++++++++++++++++ 11 files changed, 5886 insertions(+) create mode 100644 arch/ia64/ia32/Makefile create mode 100644 arch/ia64/ia32/binfmt_elf32.c create mode 100644 arch/ia64/ia32/elfcore32.h create mode 100644 arch/ia64/ia32/ia32_entry.S create mode 100644 arch/ia64/ia32/ia32_ioctl.c create mode 100644 arch/ia64/ia32/ia32_ldt.c create mode 100644 arch/ia64/ia32/ia32_signal.c create mode 100644 arch/ia64/ia32/ia32_support.c create mode 100644 arch/ia64/ia32/ia32_traps.c create mode 100644 arch/ia64/ia32/ia32priv.h create mode 100644 arch/ia64/ia32/sys_ia32.c (limited to 'arch/ia64/ia32') diff --git a/arch/ia64/ia32/Makefile b/arch/ia64/ia32/Makefile new file mode 100644 index 000000000000..2ed90da81166 --- /dev/null +++ b/arch/ia64/ia32/Makefile @@ -0,0 +1,12 @@ +# +# Makefile for the ia32 kernel emulation subsystem. +# + +obj-y := ia32_entry.o sys_ia32.o ia32_ioctl.o ia32_signal.o \ + ia32_support.o ia32_traps.o binfmt_elf32.o ia32_ldt.o + +CFLAGS_ia32_ioctl.o += -Ifs/ + +# Don't let GCC uses f16-f31 so that save_ia32_fpstate_live() and +# restore_ia32_fpstate_live() can be sure the live register contain user-level state. +CFLAGS_ia32_signal.o += -mfixed-range=f16-f31 diff --git a/arch/ia64/ia32/binfmt_elf32.c b/arch/ia64/ia32/binfmt_elf32.c new file mode 100644 index 000000000000..31de70b7c67f --- /dev/null +++ b/arch/ia64/ia32/binfmt_elf32.c @@ -0,0 +1,294 @@ +/* + * IA-32 ELF support. + * + * Copyright (C) 1999 Arun Sharma + * Copyright (C) 2001 Hewlett-Packard Co + * David Mosberger-Tang + * + * 06/16/00 A. Mallick initialize csd/ssd/tssd/cflg for ia32_load_state + * 04/13/01 D. Mosberger dropped saving tssd in ar.k1---it's not needed + * 09/14/01 D. Mosberger fixed memory management for gdt/tss page + */ +#include + +#include +#include +#include + +#include +#include + +#include "ia32priv.h" +#include "elfcore32.h" + +/* Override some function names */ +#undef start_thread +#define start_thread ia32_start_thread +#define elf_format elf32_format +#define init_elf_binfmt init_elf32_binfmt +#define exit_elf_binfmt exit_elf32_binfmt + +#undef CLOCKS_PER_SEC +#define CLOCKS_PER_SEC IA32_CLOCKS_PER_SEC + +extern void ia64_elf32_init (struct pt_regs *regs); + +static void elf32_set_personality (void); + +#define setup_arg_pages(bprm,tos,exec) ia32_setup_arg_pages(bprm,exec) +#define elf_map elf32_map + +#undef SET_PERSONALITY +#define SET_PERSONALITY(ex, ibcs2) elf32_set_personality() + +#define elf_read_implies_exec(ex, have_pt_gnu_stack) (!(have_pt_gnu_stack)) + +/* Ugly but avoids duplication */ +#include "../../../fs/binfmt_elf.c" + +extern struct page *ia32_shared_page[]; +extern unsigned long *ia32_gdt; +extern struct page *ia32_gate_page; + +struct page * +ia32_install_shared_page (struct vm_area_struct *vma, unsigned long address, int *type) +{ + struct page *pg = ia32_shared_page[smp_processor_id()]; + get_page(pg); + if (type) + *type = VM_FAULT_MINOR; + return pg; +} + +struct page * +ia32_install_gate_page (struct vm_area_struct *vma, unsigned long address, int *type) +{ + struct page *pg = ia32_gate_page; + get_page(pg); + if (type) + *type = VM_FAULT_MINOR; + return pg; +} + + +static struct vm_operations_struct ia32_shared_page_vm_ops = { + .nopage = ia32_install_shared_page +}; + +static struct vm_operations_struct ia32_gate_page_vm_ops = { + .nopage = ia32_install_gate_page +}; + +void +ia64_elf32_init (struct pt_regs *regs) +{ + struct vm_area_struct *vma; + + /* + * Map GDT below 4GB, where the processor can find it. We need to map + * it with privilege level 3 because the IVE uses non-privileged accesses to these + * tables. IA-32 segmentation is used to protect against IA-32 accesses to them. + */ + vma = kmem_cache_alloc(vm_area_cachep, SLAB_KERNEL); + if (vma) { + memset(vma, 0, sizeof(*vma)); + vma->vm_mm = current->mm; + vma->vm_start = IA32_GDT_OFFSET; + vma->vm_end = vma->vm_start + PAGE_SIZE; + vma->vm_page_prot = PAGE_SHARED; + vma->vm_flags = VM_READ|VM_MAYREAD|VM_RESERVED; + vma->vm_ops = &ia32_shared_page_vm_ops; + down_write(¤t->mm->mmap_sem); + { + if (insert_vm_struct(current->mm, vma)) { + kmem_cache_free(vm_area_cachep, vma); + up_write(¤t->mm->mmap_sem); + BUG(); + } + } + up_write(¤t->mm->mmap_sem); + } + + /* + * When user stack is not executable, push sigreturn code to stack makes + * segmentation fault raised when returning to kernel. So now sigreturn + * code is locked in specific gate page, which is pointed by pretcode + * when setup_frame_ia32 + */ + vma = kmem_cache_alloc(vm_area_cachep, SLAB_KERNEL); + if (vma) { + memset(vma, 0, sizeof(*vma)); + vma->vm_mm = current->mm; + vma->vm_start = IA32_GATE_OFFSET; + vma->vm_end = vma->vm_start + PAGE_SIZE; + vma->vm_page_prot = PAGE_COPY_EXEC; + vma->vm_flags = VM_READ | VM_MAYREAD | VM_EXEC + | VM_MAYEXEC | VM_RESERVED; + vma->vm_ops = &ia32_gate_page_vm_ops; + down_write(¤t->mm->mmap_sem); + { + if (insert_vm_struct(current->mm, vma)) { + kmem_cache_free(vm_area_cachep, vma); + up_write(¤t->mm->mmap_sem); + BUG(); + } + } + up_write(¤t->mm->mmap_sem); + } + + /* + * Install LDT as anonymous memory. This gives us all-zero segment descriptors + * until a task modifies them via modify_ldt(). + */ + vma = kmem_cache_alloc(vm_area_cachep, SLAB_KERNEL); + if (vma) { + memset(vma, 0, sizeof(*vma)); + vma->vm_mm = current->mm; + vma->vm_start = IA32_LDT_OFFSET; + vma->vm_end = vma->vm_start + PAGE_ALIGN(IA32_LDT_ENTRIES*IA32_LDT_ENTRY_SIZE); + vma->vm_page_prot = PAGE_SHARED; + vma->vm_flags = VM_READ|VM_WRITE|VM_MAYREAD|VM_MAYWRITE; + down_write(¤t->mm->mmap_sem); + { + if (insert_vm_struct(current->mm, vma)) { + kmem_cache_free(vm_area_cachep, vma); + up_write(¤t->mm->mmap_sem); + BUG(); + } + } + up_write(¤t->mm->mmap_sem); + } + + ia64_psr(regs)->ac = 0; /* turn off alignment checking */ + regs->loadrs = 0; + /* + * According to the ABI %edx points to an `atexit' handler. Since we don't have + * one we'll set it to 0 and initialize all the other registers just to make + * things more deterministic, ala the i386 implementation. + */ + regs->r8 = 0; /* %eax */ + regs->r11 = 0; /* %ebx */ + regs->r9 = 0; /* %ecx */ + regs->r10 = 0; /* %edx */ + regs->r13 = 0; /* %ebp */ + regs->r14 = 0; /* %esi */ + regs->r15 = 0; /* %edi */ + + current->thread.eflag = IA32_EFLAG; + current->thread.fsr = IA32_FSR_DEFAULT; + current->thread.fcr = IA32_FCR_DEFAULT; + current->thread.fir = 0; + current->thread.fdr = 0; + + /* + * Setup GDTD. Note: GDTD is the descrambled version of the pseudo-descriptor + * format defined by Figure 3-11 "Pseudo-Descriptor Format" in the IA-32 + * architecture manual. Also note that the only fields that are not ignored are + * `base', `limit', 'G', `P' (must be 1) and `S' (must be 0). + */ + regs->r31 = IA32_SEG_UNSCRAMBLE(IA32_SEG_DESCRIPTOR(IA32_GDT_OFFSET, IA32_PAGE_SIZE - 1, + 0, 0, 0, 1, 0, 0, 0)); + /* Setup the segment selectors */ + regs->r16 = (__USER_DS << 16) | __USER_DS; /* ES == DS, GS, FS are zero */ + regs->r17 = (__USER_DS << 16) | __USER_CS; /* SS, CS; ia32_load_state() sets TSS and LDT */ + + ia32_load_segment_descriptors(current); + ia32_load_state(current); +} + +int +ia32_setup_arg_pages (struct linux_binprm *bprm, int executable_stack) +{ + unsigned long stack_base; + struct vm_area_struct *mpnt; + struct mm_struct *mm = current->mm; + int i, ret; + + stack_base = IA32_STACK_TOP - MAX_ARG_PAGES*PAGE_SIZE; + mm->arg_start = bprm->p + stack_base; + + bprm->p += stack_base; + if (bprm->loader) + bprm->loader += stack_base; + bprm->exec += stack_base; + + mpnt = kmem_cache_alloc(vm_area_cachep, SLAB_KERNEL); + if (!mpnt) + return -ENOMEM; + + if (security_vm_enough_memory((IA32_STACK_TOP - (PAGE_MASK & (unsigned long) bprm->p)) + >> PAGE_SHIFT)) { + kmem_cache_free(vm_area_cachep, mpnt); + return -ENOMEM; + } + + memset(mpnt, 0, sizeof(*mpnt)); + + down_write(¤t->mm->mmap_sem); + { + mpnt->vm_mm = current->mm; + mpnt->vm_start = PAGE_MASK & (unsigned long) bprm->p; + mpnt->vm_end = IA32_STACK_TOP; + if (executable_stack == EXSTACK_ENABLE_X) + mpnt->vm_flags = VM_STACK_FLAGS | VM_EXEC; + else if (executable_stack == EXSTACK_DISABLE_X) + mpnt->vm_flags = VM_STACK_FLAGS & ~VM_EXEC; + else + mpnt->vm_flags = VM_STACK_FLAGS; + mpnt->vm_page_prot = (mpnt->vm_flags & VM_EXEC)? + PAGE_COPY_EXEC: PAGE_COPY; + if ((ret = insert_vm_struct(current->mm, mpnt))) { + up_write(¤t->mm->mmap_sem); + kmem_cache_free(vm_area_cachep, mpnt); + return ret; + } + current->mm->stack_vm = current->mm->total_vm = vma_pages(mpnt); + } + + for (i = 0 ; i < MAX_ARG_PAGES ; i++) { + struct page *page = bprm->page[i]; + if (page) { + bprm->page[i] = NULL; + install_arg_page(mpnt, page, stack_base); + } + stack_base += PAGE_SIZE; + } + up_write(¤t->mm->mmap_sem); + + /* Can't do it in ia64_elf32_init(). Needs to be done before calls to + elf32_map() */ + current->thread.ppl = ia32_init_pp_list(); + + return 0; +} + +static void +elf32_set_personality (void) +{ + set_personality(PER_LINUX32); + current->thread.map_base = IA32_PAGE_OFFSET/3; + current->thread.task_size = IA32_PAGE_OFFSET; /* use what Linux/x86 uses... */ + set_fs(USER_DS); /* set addr limit for new TASK_SIZE */ +} + +static unsigned long +elf32_map (struct file *filep, unsigned long addr, struct elf_phdr *eppnt, int prot, int type) +{ + unsigned long pgoff = (eppnt->p_vaddr) & ~IA32_PAGE_MASK; + + return ia32_do_mmap(filep, (addr & IA32_PAGE_MASK), eppnt->p_filesz + pgoff, prot, type, + eppnt->p_offset - pgoff); +} + +#define cpu_uses_ia32el() (local_cpu_data->family > 0x1f) + +static int __init check_elf32_binfmt(void) +{ + if (cpu_uses_ia32el()) { + printk("Please use IA-32 EL for executing IA-32 binaries\n"); + return unregister_binfmt(&elf_format); + } + return 0; +} + +module_init(check_elf32_binfmt) diff --git a/arch/ia64/ia32/elfcore32.h b/arch/ia64/ia32/elfcore32.h new file mode 100644 index 000000000000..b73b8b6b10c1 --- /dev/null +++ b/arch/ia64/ia32/elfcore32.h @@ -0,0 +1,138 @@ +/* + * IA-32 ELF core dump support. + * + * Copyright (C) 2003 Arun Sharma + * + * Derived from the x86_64 version + */ +#ifndef _ELFCORE32_H_ +#define _ELFCORE32_H_ + +#include +#include + +#define USE_ELF_CORE_DUMP 1 + +/* Override elfcore.h */ +#define _LINUX_ELFCORE_H 1 +typedef unsigned int elf_greg_t; + +#define ELF_NGREG (sizeof (struct user_regs_struct32) / sizeof(elf_greg_t)) +typedef elf_greg_t elf_gregset_t[ELF_NGREG]; + +typedef struct ia32_user_i387_struct elf_fpregset_t; +typedef struct ia32_user_fxsr_struct elf_fpxregset_t; + +struct elf_siginfo +{ + int si_signo; /* signal number */ + int si_code; /* extra code */ + int si_errno; /* errno */ +}; + +#define jiffies_to_timeval(a,b) do { (b)->tv_usec = 0; (b)->tv_sec = (a)/HZ; }while(0) + +struct elf_prstatus +{ + struct elf_siginfo pr_info; /* Info associated with signal */ + short pr_cursig; /* Current signal */ + unsigned int pr_sigpend; /* Set of pending signals */ + unsigned int pr_sighold; /* Set of held signals */ + pid_t pr_pid; + pid_t pr_ppid; + pid_t pr_pgrp; + pid_t pr_sid; + struct compat_timeval pr_utime; /* User time */ + struct compat_timeval pr_stime; /* System time */ + struct compat_timeval pr_cutime; /* Cumulative user time */ + struct compat_timeval pr_cstime; /* Cumulative system time */ + elf_gregset_t pr_reg; /* GP registers */ + int pr_fpvalid; /* True if math co-processor being used. */ +}; + +#define ELF_PRARGSZ (80) /* Number of chars for args */ + +struct elf_prpsinfo +{ + char pr_state; /* numeric process state */ + char pr_sname; /* char for pr_state */ + char pr_zomb; /* zombie */ + char pr_nice; /* nice val */ + unsigned int pr_flag; /* flags */ + __u16 pr_uid; + __u16 pr_gid; + pid_t pr_pid, pr_ppid, pr_pgrp, pr_sid; + /* Lots missing */ + char pr_fname[16]; /* filename of executable */ + char pr_psargs[ELF_PRARGSZ]; /* initial part of arg list */ +}; + +#define ELF_CORE_COPY_REGS(pr_reg, regs) \ + pr_reg[0] = regs->r11; \ + pr_reg[1] = regs->r9; \ + pr_reg[2] = regs->r10; \ + pr_reg[3] = regs->r14; \ + pr_reg[4] = regs->r15; \ + pr_reg[5] = regs->r13; \ + pr_reg[6] = regs->r8; \ + pr_reg[7] = regs->r16 & 0xffff; \ + pr_reg[8] = (regs->r16 >> 16) & 0xffff; \ + pr_reg[9] = (regs->r16 >> 32) & 0xffff; \ + pr_reg[10] = (regs->r16 >> 48) & 0xffff; \ + pr_reg[11] = regs->r1; \ + pr_reg[12] = regs->cr_iip; \ + pr_reg[13] = regs->r17 & 0xffff; \ + pr_reg[14] = ia64_getreg(_IA64_REG_AR_EFLAG); \ + pr_reg[15] = regs->r12; \ + pr_reg[16] = (regs->r17 >> 16) & 0xffff; + +static inline void elf_core_copy_regs(elf_gregset_t *elfregs, + struct pt_regs *regs) +{ + ELF_CORE_COPY_REGS((*elfregs), regs) +} + +static inline int elf_core_copy_task_regs(struct task_struct *t, + elf_gregset_t* elfregs) +{ + struct pt_regs *pp = ia64_task_regs(t); + ELF_CORE_COPY_REGS((*elfregs), pp); + return 1; +} + +static inline int +elf_core_copy_task_fpregs(struct task_struct *tsk, struct pt_regs *regs, elf_fpregset_t *fpu) +{ + struct ia32_user_i387_struct *fpstate = (void*)fpu; + mm_segment_t old_fs; + + if (!tsk_used_math(tsk)) + return 0; + + old_fs = get_fs(); + set_fs(KERNEL_DS); + save_ia32_fpstate(tsk, (struct ia32_user_i387_struct __user *) fpstate); + set_fs(old_fs); + + return 1; +} + +#define ELF_CORE_COPY_XFPREGS 1 +static inline int +elf_core_copy_task_xfpregs(struct task_struct *tsk, elf_fpxregset_t *xfpu) +{ + struct ia32_user_fxsr_struct *fpxstate = (void*) xfpu; + mm_segment_t old_fs; + + if (!tsk_used_math(tsk)) + return 0; + + old_fs = get_fs(); + set_fs(KERNEL_DS); + save_ia32_fpxstate(tsk, (struct ia32_user_fxsr_struct __user *) fpxstate); + set_fs(old_fs); + + return 1; +} + +#endif /* _ELFCORE32_H_ */ diff --git a/arch/ia64/ia32/ia32_entry.S b/arch/ia64/ia32/ia32_entry.S new file mode 100644 index 000000000000..829a6d80711c --- /dev/null +++ b/arch/ia64/ia32/ia32_entry.S @@ -0,0 +1,500 @@ +#include +#include +#include +#include +#include + +#include "../kernel/minstate.h" + + /* + * execve() is special because in case of success, we need to + * setup a null register window frame (in case an IA-32 process + * is exec'ing an IA-64 program). + */ +ENTRY(ia32_execve) + .prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(3) + alloc loc1=ar.pfs,3,2,4,0 + mov loc0=rp + .body + zxt4 out0=in0 // filename + ;; // stop bit between alloc and call + zxt4 out1=in1 // argv + zxt4 out2=in2 // envp + add out3=16,sp // regs + br.call.sptk.few rp=sys32_execve +1: cmp.ge p6,p0=r8,r0 + mov ar.pfs=loc1 // restore ar.pfs + ;; +(p6) mov ar.pfs=r0 // clear ar.pfs in case of success + sxt4 r8=r8 // return 64-bit result + mov rp=loc0 + br.ret.sptk.few rp +END(ia32_execve) + +ENTRY(ia32_clone) + .prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(5) + alloc r16=ar.pfs,5,2,6,0 + DO_SAVE_SWITCH_STACK + mov loc0=rp + mov loc1=r16 // save ar.pfs across do_fork + .body + zxt4 out1=in1 // newsp + mov out3=16 // stacksize (compensates for 16-byte scratch area) + adds out2=IA64_SWITCH_STACK_SIZE+16,sp // out2 = ®s + mov out0=in0 // out0 = clone_flags + zxt4 out4=in2 // out4 = parent_tidptr + zxt4 out5=in4 // out5 = child_tidptr + br.call.sptk.many rp=do_fork +.ret0: .restore sp + adds sp=IA64_SWITCH_STACK_SIZE,sp // pop the switch stack + mov ar.pfs=loc1 + mov rp=loc0 + br.ret.sptk.many rp +END(ia32_clone) + +ENTRY(sys32_rt_sigsuspend) + .prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(8) + alloc loc1=ar.pfs,8,2,3,0 // preserve all eight input regs + mov loc0=rp + mov out0=in0 // mask + mov out1=in1 // sigsetsize + mov out2=sp // out2 = &sigscratch + .fframe 16 + adds sp=-16,sp // allocate dummy "sigscratch" + ;; + .body + br.call.sptk.many rp=ia32_rt_sigsuspend +1: .restore sp + adds sp=16,sp + mov rp=loc0 + mov ar.pfs=loc1 + br.ret.sptk.many rp +END(sys32_rt_sigsuspend) + +ENTRY(sys32_sigsuspend) + .prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(8) + alloc loc1=ar.pfs,8,2,3,0 // preserve all eight input regs + mov loc0=rp + mov out0=in2 // mask (first two args are ignored) + ;; + mov out1=sp // out1 = &sigscratch + .fframe 16 + adds sp=-16,sp // allocate dummy "sigscratch" + .body + br.call.sptk.many rp=ia32_sigsuspend +1: .restore sp + adds sp=16,sp + mov rp=loc0 + mov ar.pfs=loc1 + br.ret.sptk.many rp +END(sys32_sigsuspend) + +GLOBAL_ENTRY(ia32_ret_from_clone) + PT_REGS_UNWIND_INFO(0) +{ /* + * Some versions of gas generate bad unwind info if the first instruction of a + * procedure doesn't go into the first slot of a bundle. This is a workaround. + */ + nop.m 0 + nop.i 0 + /* + * We need to call schedule_tail() to complete the scheduling process. + * Called by ia64_switch_to after do_fork()->copy_thread(). r8 contains the + * address of the previously executing task. + */ + br.call.sptk.many rp=ia64_invoke_schedule_tail +} +.ret1: + adds r2=TI_FLAGS+IA64_TASK_SIZE,r13 + ;; + ld4 r2=[r2] + ;; + mov r8=0 + and r2=_TIF_SYSCALL_TRACEAUDIT,r2 + ;; + cmp.ne p6,p0=r2,r0 +(p6) br.cond.spnt .ia32_strace_check_retval + ;; // prevent RAW on r8 +END(ia32_ret_from_clone) + // fall thrugh +GLOBAL_ENTRY(ia32_ret_from_syscall) + PT_REGS_UNWIND_INFO(0) + + cmp.ge p6,p7=r8,r0 // syscall executed successfully? + adds r2=IA64_PT_REGS_R8_OFFSET+16,sp // r2 = &pt_regs.r8 + ;; + alloc r3=ar.pfs,0,0,0,0 // drop the syscall argument frame + st8 [r2]=r8 // store return value in slot for r8 + br.cond.sptk.many ia64_leave_kernel +END(ia32_ret_from_syscall) + + // + // Invoke a system call, but do some tracing before and after the call. + // We MUST preserve the current register frame throughout this routine + // because some system calls (such as ia64_execve) directly + // manipulate ar.pfs. + // + // Input: + // r8 = syscall number + // b6 = syscall entry point + // +GLOBAL_ENTRY(ia32_trace_syscall) + PT_REGS_UNWIND_INFO(0) + mov r3=-38 + adds r2=IA64_PT_REGS_R8_OFFSET+16,sp + ;; + st8 [r2]=r3 // initialize return code to -ENOSYS + br.call.sptk.few rp=syscall_trace_enter // give parent a chance to catch syscall args +.ret2: // Need to reload arguments (they may be changed by the tracing process) + adds r2=IA64_PT_REGS_R1_OFFSET+16,sp // r2 = &pt_regs.r1 + adds r3=IA64_PT_REGS_R13_OFFSET+16,sp // r3 = &pt_regs.r13 + mov r15=IA32_NR_syscalls + ;; + ld4 r8=[r2],IA64_PT_REGS_R9_OFFSET-IA64_PT_REGS_R1_OFFSET + movl r16=ia32_syscall_table + ;; + ld4 r33=[r2],8 // r9 == ecx + ld4 r37=[r3],16 // r13 == ebp + cmp.ltu.unc p6,p7=r8,r15 + ;; + ld4 r34=[r2],8 // r10 == edx + ld4 r36=[r3],8 // r15 == edi +(p6) shladd r16=r8,3,r16 // force ni_syscall if not valid syscall number + ;; + ld8 r16=[r16] + ;; + ld4 r32=[r2],8 // r11 == ebx + mov b6=r16 + ld4 r35=[r3],8 // r14 == esi + br.call.sptk.few rp=b6 // do the syscall +.ia32_strace_check_retval: + cmp.lt p6,p0=r8,r0 // syscall failed? + adds r2=IA64_PT_REGS_R8_OFFSET+16,sp // r2 = &pt_regs.r8 + ;; + st8.spill [r2]=r8 // store return value in slot for r8 + br.call.sptk.few rp=syscall_trace_leave // give parent a chance to catch return value +.ret4: alloc r2=ar.pfs,0,0,0,0 // drop the syscall argument frame + br.cond.sptk.many ia64_leave_kernel +END(ia32_trace_syscall) + +GLOBAL_ENTRY(sys32_vfork) + alloc r16=ar.pfs,2,2,4,0;; + mov out0=IA64_CLONE_VFORK|IA64_CLONE_VM|SIGCHLD // out0 = clone_flags + br.cond.sptk.few .fork1 // do the work +END(sys32_vfork) + +GLOBAL_ENTRY(sys32_fork) + .prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(2) + alloc r16=ar.pfs,2,2,4,0 + mov out0=SIGCHLD // out0 = clone_flags + ;; +.fork1: + mov loc0=rp + mov loc1=r16 // save ar.pfs across do_fork + DO_SAVE_SWITCH_STACK + + .body + + mov out1=0 + mov out3=0 + adds out2=IA64_SWITCH_STACK_SIZE+16,sp // out2 = ®s + br.call.sptk.few rp=do_fork +.ret5: .restore sp + adds sp=IA64_SWITCH_STACK_SIZE,sp // pop the switch stack + mov ar.pfs=loc1 + mov rp=loc0 + br.ret.sptk.many rp +END(sys32_fork) + + .rodata + .align 8 + .globl ia32_syscall_table +ia32_syscall_table: + data8 sys_ni_syscall /* 0 - old "setup(" system call*/ + data8 sys_exit + data8 sys32_fork + data8 sys_read + data8 sys_write + data8 sys32_open /* 5 */ + data8 sys_close + data8 sys32_waitpid + data8 sys_creat + data8 sys_link + data8 sys_unlink /* 10 */ + data8 ia32_execve + data8 sys_chdir + data8 compat_sys_time + data8 sys_mknod + data8 sys_chmod /* 15 */ + data8 sys_lchown /* 16-bit version */ + data8 sys_ni_syscall /* old break syscall holder */ + data8 sys_ni_syscall + data8 sys32_lseek + data8 sys_getpid /* 20 */ + data8 compat_sys_mount + data8 sys_oldumount + data8 sys_setuid /* 16-bit version */ + data8 sys_getuid /* 16-bit version */ + data8 compat_sys_stime /* 25 */ + data8 sys32_ptrace + data8 sys32_alarm + data8 sys_ni_syscall + data8 sys32_pause + data8 compat_sys_utime /* 30 */ + data8 sys_ni_syscall /* old stty syscall holder */ + data8 sys_ni_syscall /* old gtty syscall holder */ + data8 sys_access + data8 sys_nice + data8 sys_ni_syscall /* 35 */ /* old ftime syscall holder */ + data8 sys_sync + data8 sys_kill + data8 sys_rename + data8 sys_mkdir + data8 sys_rmdir /* 40 */ + data8 sys_dup + data8 sys32_pipe + data8 compat_sys_times + data8 sys_ni_syscall /* old prof syscall holder */ + data8 sys32_brk /* 45 */ + data8 sys_setgid /* 16-bit version */ + data8 sys_getgid /* 16-bit version */ + data8 sys32_signal + data8 sys_geteuid /* 16-bit version */ + data8 sys_getegid /* 16-bit version */ /* 50 */ + data8 sys_acct + data8 sys_umount /* recycled never used phys( */ + data8 sys_ni_syscall /* old lock syscall holder */ + data8 compat_sys_ioctl + data8 compat_sys_fcntl /* 55 */ + data8 sys_ni_syscall /* old mpx syscall holder */ + data8 sys_setpgid + data8 sys_ni_syscall /* old ulimit syscall holder */ + data8 sys_ni_syscall + data8 sys_umask /* 60 */ + data8 sys_chroot + data8 sys_ustat + data8 sys_dup2 + data8 sys_getppid + data8 sys_getpgrp /* 65 */ + data8 sys_setsid + data8 sys32_sigaction + data8 sys_ni_syscall + data8 sys_ni_syscall + data8 sys_setreuid /* 16-bit version */ /* 70 */ + data8 sys_setregid /* 16-bit version */ + data8 sys32_sigsuspend + data8 compat_sys_sigpending + data8 sys_sethostname + data8 compat_sys_setrlimit /* 75 */ + data8 compat_sys_old_getrlimit + data8 compat_sys_getrusage + data8 sys32_gettimeofday + data8 sys32_settimeofday + data8 sys32_getgroups16 /* 80 */ + data8 sys32_setgroups16 + data8 sys32_old_select + data8 sys_symlink + data8 sys_ni_syscall + data8 sys_readlink /* 85 */ + data8 sys_uselib + data8 sys_swapon + data8 sys_reboot + data8 sys32_readdir + data8 sys32_mmap /* 90 */ + data8 sys32_munmap + data8 sys_truncate + data8 sys_ftruncate + data8 sys_fchmod + data8 sys_fchown /* 16-bit version */ /* 95 */ + data8 sys_getpriority + data8 sys_setpriority + data8 sys_ni_syscall /* old profil syscall holder */ + data8 compat_sys_statfs + data8 compat_sys_fstatfs /* 100 */ + data8 sys_ni_syscall /* ioperm */ + data8 compat_sys_socketcall + data8 sys_syslog + data8 compat_sys_setitimer + data8 compat_sys_getitimer /* 105 */ + data8 compat_sys_newstat + data8 compat_sys_newlstat + data8 compat_sys_newfstat + data8 sys_ni_syscall + data8 sys_ni_syscall /* iopl */ /* 110 */ + data8 sys_vhangup + data8 sys_ni_syscall /* used to be sys_idle */ + data8 sys_ni_syscall + data8 compat_sys_wait4 + data8 sys_swapoff /* 115 */ + data8 sys32_sysinfo + data8 sys32_ipc + data8 sys_fsync + data8 sys32_sigreturn + data8 ia32_clone /* 120 */ + data8 sys_setdomainname + data8 sys32_newuname + data8 sys32_modify_ldt + data8 sys_ni_syscall /* adjtimex */ + data8 sys32_mprotect /* 125 */ + data8 compat_sys_sigprocmask + data8 sys_ni_syscall /* create_module */ + data8 sys_ni_syscall /* init_module */ + data8 sys_ni_syscall /* delete_module */ + data8 sys_ni_syscall /* get_kernel_syms */ /* 130 */ + data8 sys_quotactl + data8 sys_getpgid + data8 sys_fchdir + data8 sys_ni_syscall /* sys_bdflush */ + data8 sys_sysfs /* 135 */ + data8 sys32_personality + data8 sys_ni_syscall /* for afs_syscall */ + data8 sys_setfsuid /* 16-bit version */ + data8 sys_setfsgid /* 16-bit version */ + data8 sys_llseek /* 140 */ + data8 compat_sys_getdents + data8 compat_sys_select + data8 sys_flock + data8 sys32_msync + data8 compat_sys_readv /* 145 */ + data8 compat_sys_writev + data8 sys_getsid + data8 sys_fdatasync + data8 sys32_sysctl + data8 sys_mlock /* 150 */ + data8 sys_munlock + data8 sys_mlockall + data8 sys_munlockall + data8 sys_sched_setparam + data8 sys_sched_getparam /* 155 */ + data8 sys_sched_setscheduler + data8 sys_sched_getscheduler + data8 sys_sched_yield + data8 sys_sched_get_priority_max + data8 sys_sched_get_priority_min /* 160 */ + data8 sys32_sched_rr_get_interval + data8 compat_sys_nanosleep + data8 sys32_mremap + data8 sys_setresuid /* 16-bit version */ + data8 sys32_getresuid16 /* 16-bit version */ /* 165 */ + data8 sys_ni_syscall /* vm86 */ + data8 sys_ni_syscall /* sys_query_module */ + data8 sys_poll + data8 sys_ni_syscall /* nfsservctl */ + data8 sys_setresgid /* 170 */ + data8 sys32_getresgid16 + data8 sys_prctl + data8 sys32_rt_sigreturn + data8 sys32_rt_sigaction + data8 sys32_rt_sigprocmask /* 175 */ + data8 sys_rt_sigpending + data8 compat_sys_rt_sigtimedwait + data8 sys32_rt_sigqueueinfo + data8 sys32_rt_sigsuspend + data8 sys32_pread /* 180 */ + data8 sys32_pwrite + data8 sys_chown /* 16-bit version */ + data8 sys_getcwd + data8 sys_capget + data8 sys_capset /* 185 */ + data8 sys32_sigaltstack + data8 sys32_sendfile + data8 sys_ni_syscall /* streams1 */ + data8 sys_ni_syscall /* streams2 */ + data8 sys32_vfork /* 190 */ + data8 compat_sys_getrlimit + data8 sys32_mmap2 + data8 sys32_truncate64 + data8 sys32_ftruncate64 + data8 sys32_stat64 /* 195 */ + data8 sys32_lstat64 + data8 sys32_fstat64 + data8 sys_lchown + data8 sys_getuid + data8 sys_getgid /* 200 */ + data8 sys_geteuid + data8 sys_getegid + data8 sys_setreuid + data8 sys_setregid + data8 sys_getgroups /* 205 */ + data8 sys_setgroups + data8 sys_fchown + data8 sys_setresuid + data8 sys_getresuid + data8 sys_setresgid /* 210 */ + data8 sys_getresgid + data8 sys_chown + data8 sys_setuid + data8 sys_setgid + data8 sys_setfsuid /* 215 */ + data8 sys_setfsgid + data8 sys_pivot_root + data8 sys_mincore + data8 sys_madvise + data8 compat_sys_getdents64 /* 220 */ + data8 compat_sys_fcntl64 + data8 sys_ni_syscall /* reserved for TUX */ + data8 sys_ni_syscall /* reserved for Security */ + data8 sys_gettid + data8 sys_readahead /* 225 */ + data8 sys_setxattr + data8 sys_lsetxattr + data8 sys_fsetxattr + data8 sys_getxattr + data8 sys_lgetxattr /* 230 */ + data8 sys_fgetxattr + data8 sys_listxattr + data8 sys_llistxattr + data8 sys_flistxattr + data8 sys_removexattr /* 235 */ + data8 sys_lremovexattr + data8 sys_fremovexattr + data8 sys_tkill + data8 sys_sendfile64 + data8 compat_sys_futex /* 240 */ + data8 compat_sys_sched_setaffinity + data8 compat_sys_sched_getaffinity + data8 sys32_set_thread_area + data8 sys32_get_thread_area + data8 compat_sys_io_setup /* 245 */ + data8 sys_io_destroy + data8 compat_sys_io_getevents + data8 compat_sys_io_submit + data8 sys_io_cancel + data8 sys_fadvise64 /* 250 */ + data8 sys_ni_syscall + data8 sys_exit_group + data8 sys_lookup_dcookie + data8 sys_epoll_create + data8 sys32_epoll_ctl /* 255 */ + data8 sys32_epoll_wait + data8 sys_remap_file_pages + data8 sys_set_tid_address + data8 sys32_timer_create + data8 compat_sys_timer_settime /* 260 */ + data8 compat_sys_timer_gettime + data8 sys_timer_getoverrun + data8 sys_timer_delete + data8 compat_sys_clock_settime + data8 compat_sys_clock_gettime /* 265 */ + data8 compat_sys_clock_getres + data8 compat_sys_clock_nanosleep + data8 compat_sys_statfs64 + data8 compat_sys_fstatfs64 + data8 sys_tgkill /* 270 */ + data8 compat_sys_utimes + data8 sys32_fadvise64_64 + data8 sys_ni_syscall + data8 sys_ni_syscall + data8 sys_ni_syscall /* 275 */ + data8 sys_ni_syscall + data8 compat_sys_mq_open + data8 sys_mq_unlink + data8 compat_sys_mq_timedsend + data8 compat_sys_mq_timedreceive /* 280 */ + data8 compat_sys_mq_notify + data8 compat_sys_mq_getsetattr + data8 sys_ni_syscall /* reserved for kexec */ + data8 compat_sys_waitid + + // guard against failures to increase IA32_NR_syscalls + .org ia32_syscall_table + 8*IA32_NR_syscalls diff --git a/arch/ia64/ia32/ia32_ioctl.c b/arch/ia64/ia32/ia32_ioctl.c new file mode 100644 index 000000000000..9845dabe2613 --- /dev/null +++ b/arch/ia64/ia32/ia32_ioctl.c @@ -0,0 +1,48 @@ +/* + * IA32 Architecture-specific ioctl shim code + * + * Copyright (C) 2000 VA Linux Co + * Copyright (C) 2000 Don Dugger + * Copyright (C) 2001-2003 Hewlett-Packard Co + * David Mosberger-Tang + */ + +#include /* argh, msdos_fs.h isn't self-contained... */ +#include +#include "ia32priv.h" + +#define INCLUDES +#include "compat_ioctl.c" +#include + +#define IOCTL_NR(a) ((a) & ~(_IOC_SIZEMASK << _IOC_SIZESHIFT)) + +#define DO_IOCTL(fd, cmd, arg) ({ \ + int _ret; \ + mm_segment_t _old_fs = get_fs(); \ + \ + set_fs(KERNEL_DS); \ + _ret = sys_ioctl(fd, cmd, (unsigned long)arg); \ + set_fs(_old_fs); \ + _ret; \ +}) + +#define CODE +#include "compat_ioctl.c" + +typedef int (* ioctl32_handler_t)(unsigned int, unsigned int, unsigned long, struct file *); + +#define COMPATIBLE_IOCTL(cmd) HANDLE_IOCTL((cmd),sys_ioctl) +#define HANDLE_IOCTL(cmd,handler) { (cmd), (ioctl32_handler_t)(handler), NULL }, +#define IOCTL_TABLE_START \ + struct ioctl_trans ioctl_start[] = { +#define IOCTL_TABLE_END \ + }; + +IOCTL_TABLE_START +#define DECLARES +#include "compat_ioctl.c" +#include +IOCTL_TABLE_END + +int ioctl_table_size = ARRAY_SIZE(ioctl_start); diff --git a/arch/ia64/ia32/ia32_ldt.c b/arch/ia64/ia32/ia32_ldt.c new file mode 100644 index 000000000000..a152738c7d0d --- /dev/null +++ b/arch/ia64/ia32/ia32_ldt.c @@ -0,0 +1,147 @@ +/* + * Copyright (C) 2001, 2004 Hewlett-Packard Co + * David Mosberger-Tang + * + * Adapted from arch/i386/kernel/ldt.c + */ + +#include +#include +#include +#include +#include +#include +#include + +#include + +#include "ia32priv.h" + +/* + * read_ldt() is not really atomic - this is not a problem since synchronization of reads + * and writes done to the LDT has to be assured by user-space anyway. Writes are atomic, + * to protect the security checks done on new descriptors. + */ +static int +read_ldt (void __user *ptr, unsigned long bytecount) +{ + unsigned long bytes_left, n; + char __user *src, *dst; + char buf[256]; /* temporary buffer (don't overflow kernel stack!) */ + + if (bytecount > IA32_LDT_ENTRIES*IA32_LDT_ENTRY_SIZE) + bytecount = IA32_LDT_ENTRIES*IA32_LDT_ENTRY_SIZE; + + bytes_left = bytecount; + + src = (void __user *) IA32_LDT_OFFSET; + dst = ptr; + + while (bytes_left) { + n = sizeof(buf); + if (n > bytes_left) + n = bytes_left; + + /* + * We know we're reading valid memory, but we still must guard against + * running out of memory. + */ + if (__copy_from_user(buf, src, n)) + return -EFAULT; + + if (copy_to_user(dst, buf, n)) + return -EFAULT; + + src += n; + dst += n; + bytes_left -= n; + } + return bytecount; +} + +static int +read_default_ldt (void __user * ptr, unsigned long bytecount) +{ + unsigned long size; + int err; + + /* XXX fix me: should return equivalent of default_ldt[0] */ + err = 0; + size = 8; + if (size > bytecount) + size = bytecount; + + err = size; + if (clear_user(ptr, size)) + err = -EFAULT; + + return err; +} + +static int +write_ldt (void __user * ptr, unsigned long bytecount, int oldmode) +{ + struct ia32_user_desc ldt_info; + __u64 entry; + int ret; + + if (bytecount != sizeof(ldt_info)) + return -EINVAL; + if (copy_from_user(&ldt_info, ptr, sizeof(ldt_info))) + return -EFAULT; + + if (ldt_info.entry_number >= IA32_LDT_ENTRIES) + return -EINVAL; + if (ldt_info.contents == 3) { + if (oldmode) + return -EINVAL; + if (ldt_info.seg_not_present == 0) + return -EINVAL; + } + + if (ldt_info.base_addr == 0 && ldt_info.limit == 0 + && (oldmode || (ldt_info.contents == 0 && ldt_info.read_exec_only == 1 + && ldt_info.seg_32bit == 0 && ldt_info.limit_in_pages == 0 + && ldt_info.seg_not_present == 1 && ldt_info.useable == 0))) + /* allow LDTs to be cleared by the user */ + entry = 0; + else + /* we must set the "Accessed" bit as IVE doesn't emulate it */ + entry = IA32_SEG_DESCRIPTOR(ldt_info.base_addr, ldt_info.limit, + (((ldt_info.read_exec_only ^ 1) << 1) + | (ldt_info.contents << 2)) | 1, + 1, 3, ldt_info.seg_not_present ^ 1, + (oldmode ? 0 : ldt_info.useable), + ldt_info.seg_32bit, + ldt_info.limit_in_pages); + /* + * Install the new entry. We know we're accessing valid (mapped) user-level + * memory, but we still need to guard against out-of-memory, hence we must use + * put_user(). + */ + ret = __put_user(entry, (__u64 __user *) IA32_LDT_OFFSET + ldt_info.entry_number); + ia32_load_segment_descriptors(current); + return ret; +} + +asmlinkage int +sys32_modify_ldt (int func, unsigned int ptr, unsigned int bytecount) +{ + int ret = -ENOSYS; + + switch (func) { + case 0: + ret = read_ldt(compat_ptr(ptr), bytecount); + break; + case 1: + ret = write_ldt(compat_ptr(ptr), bytecount, 1); + break; + case 2: + ret = read_default_ldt(compat_ptr(ptr), bytecount); + break; + case 0x11: + ret = write_ldt(compat_ptr(ptr), bytecount, 0); + break; + } + return ret; +} diff --git a/arch/ia64/ia32/ia32_signal.c b/arch/ia64/ia32/ia32_signal.c new file mode 100644 index 000000000000..19b02adce68c --- /dev/null +++ b/arch/ia64/ia32/ia32_signal.c @@ -0,0 +1,1036 @@ +/* + * IA32 Architecture-specific signal handling support. + * + * Copyright (C) 1999, 2001-2002, 2005 Hewlett-Packard Co + * David Mosberger-Tang + * Copyright (C) 1999 Arun Sharma + * Copyright (C) 2000 VA Linux Co + * Copyright (C) 2000 Don Dugger + * + * Derived from i386 and Alpha versions. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#include "ia32priv.h" + +#include "../kernel/sigframe.h" + +#define A(__x) ((unsigned long)(__x)) + +#define DEBUG_SIG 0 +#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) + +#define __IA32_NR_sigreturn 119 +#define __IA32_NR_rt_sigreturn 173 + +struct sigframe_ia32 +{ + int pretcode; + int sig; + struct sigcontext_ia32 sc; + struct _fpstate_ia32 fpstate; + unsigned int extramask[_COMPAT_NSIG_WORDS-1]; + char retcode[8]; +}; + +struct rt_sigframe_ia32 +{ + int pretcode; + int sig; + int pinfo; + int puc; + compat_siginfo_t info; + struct ucontext_ia32 uc; + struct _fpstate_ia32 fpstate; + char retcode[8]; +}; + +int +copy_siginfo_from_user32 (siginfo_t *to, compat_siginfo_t __user *from) +{ + unsigned long tmp; + int err; + + if (!access_ok(VERIFY_READ, from, sizeof(compat_siginfo_t))) + return -EFAULT; + + err = __get_user(to->si_signo, &from->si_signo); + err |= __get_user(to->si_errno, &from->si_errno); + err |= __get_user(to->si_code, &from->si_code); + + if (to->si_code < 0) + err |= __copy_from_user(&to->_sifields._pad, &from->_sifields._pad, SI_PAD_SIZE); + else { + switch (to->si_code >> 16) { + case __SI_CHLD >> 16: + err |= __get_user(to->si_utime, &from->si_utime); + err |= __get_user(to->si_stime, &from->si_stime); + err |= __get_user(to->si_status, &from->si_status); + default: + err |= __get_user(to->si_pid, &from->si_pid); + err |= __get_user(to->si_uid, &from->si_uid); + break; + case __SI_FAULT >> 16: + err |= __get_user(tmp, &from->si_addr); + to->si_addr = (void __user *) tmp; + break; + case __SI_POLL >> 16: + err |= __get_user(to->si_band, &from->si_band); + err |= __get_user(to->si_fd, &from->si_fd); + break; + case __SI_RT >> 16: /* This is not generated by the kernel as of now. */ + case __SI_MESGQ >> 16: + err |= __get_user(to->si_pid, &from->si_pid); + err |= __get_user(to->si_uid, &from->si_uid); + err |= __get_user(to->si_int, &from->si_int); + break; + } + } + return err; +} + +int +copy_siginfo_to_user32 (compat_siginfo_t __user *to, siginfo_t *from) +{ + unsigned int addr; + int err; + + if (!access_ok(VERIFY_WRITE, to, sizeof(compat_siginfo_t))) + return -EFAULT; + + /* If you change siginfo_t structure, please be sure + this code is fixed accordingly. + It should never copy any pad contained in the structure + to avoid security leaks, but must copy the generic + 3 ints plus the relevant union member. + This routine must convert siginfo from 64bit to 32bit as well + at the same time. */ + err = __put_user(from->si_signo, &to->si_signo); + err |= __put_user(from->si_errno, &to->si_errno); + err |= __put_user((short)from->si_code, &to->si_code); + if (from->si_code < 0) + err |= __copy_to_user(&to->_sifields._pad, &from->_sifields._pad, SI_PAD_SIZE); + else { + switch (from->si_code >> 16) { + case __SI_CHLD >> 16: + err |= __put_user(from->si_utime, &to->si_utime); + err |= __put_user(from->si_stime, &to->si_stime); + err |= __put_user(from->si_status, &to->si_status); + default: + err |= __put_user(from->si_pid, &to->si_pid); + err |= __put_user(from->si_uid, &to->si_uid); + break; + case __SI_FAULT >> 16: + /* avoid type-checking warnings by copying _pad[0] in lieu of si_addr... */ + err |= __put_user(from->_sifields._pad[0], &to->si_addr); + break; + case __SI_POLL >> 16: + err |= __put_user(from->si_band, &to->si_band); + err |= __put_user(from->si_fd, &to->si_fd); + break; + case __SI_TIMER >> 16: + err |= __put_user(from->si_tid, &to->si_tid); + err |= __put_user(from->si_overrun, &to->si_overrun); + addr = (unsigned long) from->si_ptr; + err |= __put_user(addr, &to->si_ptr); + break; + case __SI_RT >> 16: /* Not generated by the kernel as of now. */ + case __SI_MESGQ >> 16: + err |= __put_user(from->si_uid, &to->si_uid); + err |= __put_user(from->si_pid, &to->si_pid); + addr = (unsigned long) from->si_ptr; + err |= __put_user(addr, &to->si_ptr); + break; + } + } + return err; +} + + +/* + * SAVE and RESTORE of ia32 fpstate info, from ia64 current state + * Used in exception handler to pass the fpstate to the user, and restore + * the fpstate while returning from the exception handler. + * + * fpstate info and their mapping to IA64 regs: + * fpstate REG(BITS) Attribute Comments + * cw ar.fcr(0:12) with bits 7 and 6 not used + * sw ar.fsr(0:15) + * tag ar.fsr(16:31) with odd numbered bits not used + * (read returns 0, writes ignored) + * ipoff ar.fir(0:31) + * cssel ar.fir(32:47) + * dataoff ar.fdr(0:31) + * datasel ar.fdr(32:47) + * + * _st[(0+TOS)%8] f8 + * _st[(1+TOS)%8] f9 + * _st[(2+TOS)%8] f10 + * _st[(3+TOS)%8] f11 (f8..f11 from ptregs) + * : : : (f12..f15 from live reg) + * : : : + * _st[(7+TOS)%8] f15 TOS=sw.top(bits11:13) + * + * status Same as sw RO + * magic 0 as X86_FXSR_MAGIC in ia32 + * mxcsr Bits(7:15)=ar.fcr(39:47) + * Bits(0:5) =ar.fsr(32:37) with bit 6 reserved + * _xmm[0..7] f16..f31 (live registers) + * with _xmm[0] + * Bit(64:127)=f17(0:63) + * Bit(0:63)=f16(0:63) + * All other fields unused... + */ + +static int +save_ia32_fpstate_live (struct _fpstate_ia32 __user *save) +{ + struct task_struct *tsk = current; + struct pt_regs *ptp; + struct _fpreg_ia32 *fpregp; + char buf[32]; + unsigned long fsr, fcr, fir, fdr; + unsigned long new_fsr; + unsigned long num128[2]; + unsigned long mxcsr=0; + int fp_tos, fr8_st_map; + + if (!access_ok(VERIFY_WRITE, save, sizeof(*save))) + return -EFAULT; + + /* Read in fsr, fcr, fir, fdr and copy onto fpstate */ + fsr = ia64_getreg(_IA64_REG_AR_FSR); + fcr = ia64_getreg(_IA64_REG_AR_FCR); + fir = ia64_getreg(_IA64_REG_AR_FIR); + fdr = ia64_getreg(_IA64_REG_AR_FDR); + + /* + * We need to clear the exception state before calling the signal handler. Clear + * the bits 15, bits 0-7 in fp status word. Similar to the functionality of fnclex + * instruction. + */ + new_fsr = fsr & ~0x80ff; + ia64_setreg(_IA64_REG_AR_FSR, new_fsr); + + __put_user(fcr & 0xffff, &save->cw); + __put_user(fsr & 0xffff, &save->sw); + __put_user((fsr>>16) & 0xffff, &save->tag); + __put_user(fir, &save->ipoff); + __put_user((fir>>32) & 0xffff, &save->cssel); + __put_user(fdr, &save->dataoff); + __put_user((fdr>>32) & 0xffff, &save->datasel); + __put_user(fsr & 0xffff, &save->status); + + mxcsr = ((fcr>>32) & 0xff80) | ((fsr>>32) & 0x3f); + __put_user(mxcsr & 0xffff, &save->mxcsr); + __put_user( 0, &save->magic); //#define X86_FXSR_MAGIC 0x0000 + + /* + * save f8..f11 from pt_regs + * save f12..f15 from live register set + */ + /* + * Find the location where f8 has to go in fp reg stack. This depends on + * TOP(11:13) field of sw. Other f reg continue sequentially from where f8 maps + * to. + */ + fp_tos = (fsr>>11)&0x7; + fr8_st_map = (8-fp_tos)&0x7; + ptp = ia64_task_regs(tsk); + fpregp = (struct _fpreg_ia32 *)(((unsigned long)buf + 15) & ~15); + ia64f2ia32f(fpregp, &ptp->f8); + copy_to_user(&save->_st[(0+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); + ia64f2ia32f(fpregp, &ptp->f9); + copy_to_user(&save->_st[(1+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); + ia64f2ia32f(fpregp, &ptp->f10); + copy_to_user(&save->_st[(2+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); + ia64f2ia32f(fpregp, &ptp->f11); + copy_to_user(&save->_st[(3+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); + + ia64_stfe(fpregp, 12); + copy_to_user(&save->_st[(4+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); + ia64_stfe(fpregp, 13); + copy_to_user(&save->_st[(5+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); + ia64_stfe(fpregp, 14); + copy_to_user(&save->_st[(6+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); + ia64_stfe(fpregp, 15); + copy_to_user(&save->_st[(7+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); + + ia64_stf8(&num128[0], 16); + ia64_stf8(&num128[1], 17); + copy_to_user(&save->_xmm[0], num128, sizeof(struct _xmmreg_ia32)); + + ia64_stf8(&num128[0], 18); + ia64_stf8(&num128[1], 19); + copy_to_user(&save->_xmm[1], num128, sizeof(struct _xmmreg_ia32)); + + ia64_stf8(&num128[0], 20); + ia64_stf8(&num128[1], 21); + copy_to_user(&save->_xmm[2], num128, sizeof(struct _xmmreg_ia32)); + + ia64_stf8(&num128[0], 22); + ia64_stf8(&num128[1], 23); + copy_to_user(&save->_xmm[3], num128, sizeof(struct _xmmreg_ia32)); + + ia64_stf8(&num128[0], 24); + ia64_stf8(&num128[1], 25); + copy_to_user(&save->_xmm[4], num128, sizeof(struct _xmmreg_ia32)); + + ia64_stf8(&num128[0], 26); + ia64_stf8(&num128[1], 27); + copy_to_user(&save->_xmm[5], num128, sizeof(struct _xmmreg_ia32)); + + ia64_stf8(&num128[0], 28); + ia64_stf8(&num128[1], 29); + copy_to_user(&save->_xmm[6], num128, sizeof(struct _xmmreg_ia32)); + + ia64_stf8(&num128[0], 30); + ia64_stf8(&num128[1], 31); + copy_to_user(&save->_xmm[7], num128, sizeof(struct _xmmreg_ia32)); + return 0; +} + +static int +restore_ia32_fpstate_live (struct _fpstate_ia32 __user *save) +{ + struct task_struct *tsk = current; + struct pt_regs *ptp; + unsigned int lo, hi; + unsigned long num128[2]; + unsigned long num64, mxcsr; + struct _fpreg_ia32 *fpregp; + char buf[32]; + unsigned long fsr, fcr, fir, fdr; + int fp_tos, fr8_st_map; + + if (!access_ok(VERIFY_READ, save, sizeof(*save))) + return(-EFAULT); + + /* + * Updating fsr, fcr, fir, fdr. + * Just a bit more complicated than save. + * - Need to make sure that we don't write any value other than the + * specific fpstate info + * - Need to make sure that the untouched part of frs, fdr, fir, fcr + * should remain same while writing. + * So, we do a read, change specific fields and write. + */ + fsr = ia64_getreg(_IA64_REG_AR_FSR); + fcr = ia64_getreg(_IA64_REG_AR_FCR); + fir = ia64_getreg(_IA64_REG_AR_FIR); + fdr = ia64_getreg(_IA64_REG_AR_FDR); + + __get_user(mxcsr, (unsigned int __user *)&save->mxcsr); + /* setting bits 0..5 8..12 with cw and 39..47 from mxcsr */ + __get_user(lo, (unsigned int __user *)&save->cw); + num64 = mxcsr & 0xff10; + num64 = (num64 << 32) | (lo & 0x1f3f); + fcr = (fcr & (~0xff1000001f3fUL)) | num64; + + /* setting bits 0..31 with sw and tag and 32..37 from mxcsr */ + __get_user(lo, (unsigned int __user *)&save->sw); + /* set bits 15,7 (fsw.b, fsw.es) to reflect the current error status */ + if ( !(lo & 0x7f) ) + lo &= (~0x8080); + __get_user(hi, (unsigned int __user *)&save->tag); + num64 = mxcsr & 0x3f; + num64 = (num64 << 16) | (hi & 0xffff); + num64 = (num64 << 16) | (lo & 0xffff); + fsr = (fsr & (~0x3fffffffffUL)) | num64; + + /* setting bits 0..47 with cssel and ipoff */ + __get_user(lo, (unsigned int __user *)&save->ipoff); + __get_user(hi, (unsigned int __user *)&save->cssel); + num64 = hi & 0xffff; + num64 = (num64 << 32) | lo; + fir = (fir & (~0xffffffffffffUL)) | num64; + + /* setting bits 0..47 with datasel and dataoff */ + __get_user(lo, (unsigned int __user *)&save->dataoff); + __get_user(hi, (unsigned int __user *)&save->datasel); + num64 = hi & 0xffff; + num64 = (num64 << 32) | lo; + fdr = (fdr & (~0xffffffffffffUL)) | num64; + + ia64_setreg(_IA64_REG_AR_FSR, fsr); + ia64_setreg(_IA64_REG_AR_FCR, fcr); + ia64_setreg(_IA64_REG_AR_FIR, fir); + ia64_setreg(_IA64_REG_AR_FDR, fdr); + + /* + * restore f8..f11 onto pt_regs + * restore f12..f15 onto live registers + */ + /* + * Find the location where f8 has to go in fp reg stack. This depends on + * TOP(11:13) field of sw. Other f reg continue sequentially from where f8 maps + * to. + */ + fp_tos = (fsr>>11)&0x7; + fr8_st_map = (8-fp_tos)&0x7; + fpregp = (struct _fpreg_ia32 *)(((unsigned long)buf + 15) & ~15); + + ptp = ia64_task_regs(tsk); + copy_from_user(fpregp, &save->_st[(0+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); + ia32f2ia64f(&ptp->f8, fpregp); + copy_from_user(fpregp, &save->_st[(1+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); + ia32f2ia64f(&ptp->f9, fpregp); + copy_from_user(fpregp, &save->_st[(2+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); + ia32f2ia64f(&ptp->f10, fpregp); + copy_from_user(fpregp, &save->_st[(3+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); + ia32f2ia64f(&ptp->f11, fpregp); + + copy_from_user(fpregp, &save->_st[(4+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); + ia64_ldfe(12, fpregp); + copy_from_user(fpregp, &save->_st[(5+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); + ia64_ldfe(13, fpregp); + copy_from_user(fpregp, &save->_st[(6+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); + ia64_ldfe(14, fpregp); + copy_from_user(fpregp, &save->_st[(7+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); + ia64_ldfe(15, fpregp); + + copy_from_user(num128, &save->_xmm[0], sizeof(struct _xmmreg_ia32)); + ia64_ldf8(16, &num128[0]); + ia64_ldf8(17, &num128[1]); + + copy_from_user(num128, &save->_xmm[1], sizeof(struct _xmmreg_ia32)); + ia64_ldf8(18, &num128[0]); + ia64_ldf8(19, &num128[1]); + + copy_from_user(num128, &save->_xmm[2], sizeof(struct _xmmreg_ia32)); + ia64_ldf8(20, &num128[0]); + ia64_ldf8(21, &num128[1]); + + copy_from_user(num128, &save->_xmm[3], sizeof(struct _xmmreg_ia32)); + ia64_ldf8(22, &num128[0]); + ia64_ldf8(23, &num128[1]); + + copy_from_user(num128, &save->_xmm[4], sizeof(struct _xmmreg_ia32)); + ia64_ldf8(24, &num128[0]); + ia64_ldf8(25, &num128[1]); + + copy_from_user(num128, &save->_xmm[5], sizeof(struct _xmmreg_ia32)); + ia64_ldf8(26, &num128[0]); + ia64_ldf8(27, &num128[1]); + + copy_from_user(num128, &save->_xmm[6], sizeof(struct _xmmreg_ia32)); + ia64_ldf8(28, &num128[0]); + ia64_ldf8(29, &num128[1]); + + copy_from_user(num128, &save->_xmm[7], sizeof(struct _xmmreg_ia32)); + ia64_ldf8(30, &num128[0]); + ia64_ldf8(31, &num128[1]); + return 0; +} + +static inline void +sigact_set_handler (struct k_sigaction *sa, unsigned int handler, unsigned int restorer) +{ + if (handler + 1 <= 2) + /* SIG_DFL, SIG_IGN, or SIG_ERR: must sign-extend to 64-bits */ + sa->sa.sa_handler = (__sighandler_t) A((int) handler); + else + sa->sa.sa_handler = (__sighandler_t) (((unsigned long) restorer << 32) | handler); +} + +long +__ia32_rt_sigsuspend (compat_sigset_t *sset, unsigned int sigsetsize, struct sigscratch *scr) +{ + extern long ia64_do_signal (sigset_t *oldset, struct sigscratch *scr, long in_syscall); + sigset_t oldset, set; + + scr->scratch_unat = 0; /* avoid leaking kernel bits to user level */ + memset(&set, 0, sizeof(&set)); + + if (memcpy(&set.sig, &sset->sig, sigsetsize)) + return -EFAULT; + + sigdelsetmask(&set, ~_BLOCKABLE); + + spin_lock_irq(¤t->sighand->siglock); + { + oldset = current->blocked; + current->blocked = set; + recalc_sigpending(); + } + spin_unlock_irq(¤t->sighand->siglock); + + /* + * The return below usually returns to the signal handler. We need to pre-set the + * correct error code here to ensure that the right values get saved in sigcontext + * by ia64_do_signal. + */ + scr->pt.r8 = -EINTR; + while (1) { + current->state = TASK_INTERRUPTIBLE; + schedule(); + if (ia64_do_signal(&oldset, scr, 1)) + return -EINTR; + } +} + +asmlinkage long +ia32_rt_sigsuspend (compat_sigset_t __user *uset, unsigned int sigsetsize, struct sigscratch *scr) +{ + compat_sigset_t set; + + if (sigsetsize > sizeof(compat_sigset_t)) + return -EINVAL; + + if (copy_from_user(&set.sig, &uset->sig, sigsetsize)) + return -EFAULT; + + return __ia32_rt_sigsuspend(&set, sigsetsize, scr); +} + +asmlinkage long +ia32_sigsuspend (unsigned int mask, struct sigscratch *scr) +{ + return __ia32_rt_sigsuspend((compat_sigset_t *) &mask, sizeof(mask), scr); +} + +asmlinkage long +sys32_signal (int sig, unsigned int handler) +{ + struct k_sigaction new_sa, old_sa; + int ret; + + sigact_set_handler(&new_sa, handler, 0); + new_sa.sa.sa_flags = SA_ONESHOT | SA_NOMASK; + + ret = do_sigaction(sig, &new_sa, &old_sa); + + return ret ? ret : IA32_SA_HANDLER(&old_sa); +} + +asmlinkage long +sys32_rt_sigaction (int sig, struct sigaction32 __user *act, + struct sigaction32 __user *oact, unsigned int sigsetsize) +{ + struct k_sigaction new_ka, old_ka; + unsigned int handler, restorer; + int ret; + + /* XXX: Don't preclude handling different sized sigset_t's. */ + if (sigsetsize != sizeof(compat_sigset_t)) + return -EINVAL; + + if (act) { + ret = get_user(handler, &act->sa_handler); + ret |= get_user(new_ka.sa.sa_flags, &act->sa_flags); + ret |= get_user(restorer, &act->sa_restorer); + ret |= copy_from_user(&new_ka.sa.sa_mask, &act->sa_mask, sizeof(compat_sigset_t)); + if (ret) + return -EFAULT; + + sigact_set_handler(&new_ka, handler, restorer); + } + + ret = do_sigaction(sig, act ? &new_ka : NULL, oact ? &old_ka : NULL); + + if (!ret && oact) { + ret = put_user(IA32_SA_HANDLER(&old_ka), &oact->sa_handler); + ret |= put_user(old_ka.sa.sa_flags, &oact->sa_flags); + ret |= put_user(IA32_SA_RESTORER(&old_ka), &oact->sa_restorer); + ret |= copy_to_user(&oact->sa_mask, &old_ka.sa.sa_mask, sizeof(compat_sigset_t)); + } + return ret; +} + + +asmlinkage long +sys32_rt_sigprocmask (int how, compat_sigset_t __user *set, compat_sigset_t __user *oset, + unsigned int sigsetsize) +{ + mm_segment_t old_fs = get_fs(); + sigset_t s; + long ret; + + if (sigsetsize > sizeof(s)) + return -EINVAL; + + if (set) { + memset(&s, 0, sizeof(s)); + if (copy_from_user(&s.sig, set, sigsetsize)) + return -EFAULT; + } + set_fs(KERNEL_DS); + ret = sys_rt_sigprocmask(how, + set ? (sigset_t __user *) &s : NULL, + oset ? (sigset_t __user *) &s : NULL, sizeof(s)); + set_fs(old_fs); + if (ret) + return ret; + if (oset) { + if (copy_to_user(oset, &s.sig, sigsetsize)) + return -EFAULT; + } + return 0; +} + +asmlinkage long +sys32_rt_sigqueueinfo (int pid, int sig, compat_siginfo_t __user *uinfo) +{ + mm_segment_t old_fs = get_fs(); + siginfo_t info; + int ret; + + if (copy_siginfo_from_user32(&info, uinfo)) + return -EFAULT; + set_fs(KERNEL_DS); + ret = sys_rt_sigqueueinfo(pid, sig, (siginfo_t __user *) &info); + set_fs(old_fs); + return ret; +} + +asmlinkage long +sys32_sigaction (int sig, struct old_sigaction32 __user *act, struct old_sigaction32 __user *oact) +{ + struct k_sigaction new_ka, old_ka; + unsigned int handler, restorer; + int ret; + + if (act) { + compat_old_sigset_t mask; + + ret = get_user(handler, &act->sa_handler); + ret |= get_user(new_ka.sa.sa_flags, &act->sa_flags); + ret |= get_user(restorer, &act->sa_restorer); + ret |= get_user(mask, &act->sa_mask); + if (ret) + return ret; + + sigact_set_handler(&new_ka, handler, restorer); + siginitset(&new_ka.sa.sa_mask, mask); + } + + ret = do_sigaction(sig, act ? &new_ka : NULL, oact ? &old_ka : NULL); + + if (!ret && oact) { + ret = put_user(IA32_SA_HANDLER(&old_ka), &oact->sa_handler); + ret |= put_user(old_ka.sa.sa_flags, &oact->sa_flags); + ret |= put_user(IA32_SA_RESTORER(&old_ka), &oact->sa_restorer); + ret |= put_user(old_ka.sa.sa_mask.sig[0], &oact->sa_mask); + } + + return ret; +} + +static int +setup_sigcontext_ia32 (struct sigcontext_ia32 __user *sc, struct _fpstate_ia32 __user *fpstate, + struct pt_regs *regs, unsigned long mask) +{ + int err = 0; + unsigned long flag; + + if (!access_ok(VERIFY_WRITE, sc, sizeof(*sc))) + return -EFAULT; + + err |= __put_user((regs->r16 >> 32) & 0xffff, (unsigned int __user *)&sc->fs); + err |= __put_user((regs->r16 >> 48) & 0xffff, (unsigned int __user *)&sc->gs); + err |= __put_user((regs->r16 >> 16) & 0xffff, (unsigned int __user *)&sc->es); + err |= __put_user(regs->r16 & 0xffff, (unsigned int __user *)&sc->ds); + err |= __put_user(regs->r15, &sc->edi); + err |= __put_user(regs->r14, &sc->esi); + err |= __put_user(regs->r13, &sc->ebp); + err |= __put_user(regs->r12, &sc->esp); + err |= __put_user(regs->r11, &sc->ebx); + err |= __put_user(regs->r10, &sc->edx); + err |= __put_user(regs->r9, &sc->ecx); + err |= __put_user(regs->r8, &sc->eax); +#if 0 + err |= __put_user(current->tss.trap_no, &sc->trapno); + err |= __put_user(current->tss.error_code, &sc->err); +#endif + err |= __put_user(regs->cr_iip, &sc->eip); + err |= __put_user(regs->r17 & 0xffff, (unsigned int __user *)&sc->cs); + /* + * `eflags' is in an ar register for this context + */ + flag = ia64_getreg(_IA64_REG_AR_EFLAG); + err |= __put_user((unsigned int)flag, &sc->eflags); + err |= __put_user(regs->r12, &sc->esp_at_signal); + err |= __put_user((regs->r17 >> 16) & 0xffff, (unsigned int __user *)&sc->ss); + + if ( save_ia32_fpstate_live(fpstate) < 0 ) + err = -EFAULT; + else + err |= __put_user((u32)(u64)fpstate, &sc->fpstate); + +#if 0 + tmp = save_i387(fpstate); + if (tmp < 0) + err = 1; + else + err |= __put_user(tmp ? fpstate : NULL, &sc->fpstate); + + /* non-iBCS2 extensions.. */ +#endif + err |= __put_user(mask, &sc->oldmask); +#if 0 + err |= __put_user(current->tss.cr2, &sc->cr2); +#endif + return err; +} + +static int +restore_sigcontext_ia32 (struct pt_regs *regs, struct sigcontext_ia32 __user *sc, int *peax) +{ + unsigned int err = 0; + + /* Always make any pending restarted system calls return -EINTR */ + current_thread_info()->restart_block.fn = do_no_restart_syscall; + + if (!access_ok(VERIFY_READ, sc, sizeof(*sc))) + return(-EFAULT); + +#define COPY(ia64x, ia32x) err |= __get_user(regs->ia64x, &sc->ia32x) + +#define copyseg_gs(tmp) (regs->r16 |= (unsigned long) (tmp) << 48) +#define copyseg_fs(tmp) (regs->r16 |= (unsigned long) (tmp) << 32) +#define copyseg_cs(tmp) (regs->r17 |= tmp) +#define copyseg_ss(tmp) (regs->r17 |= (unsigned long) (tmp) << 16) +#define copyseg_es(tmp) (regs->r16 |= (unsigned long) (tmp) << 16) +#define copyseg_ds(tmp) (regs->r16 |= tmp) + +#define COPY_SEG(seg) \ + { \ + unsigned short tmp; \ + err |= __get_user(tmp, &sc->seg); \ + copyseg_##seg(tmp); \ + } +#define COPY_SEG_STRICT(seg) \ + { \ + unsigned short tmp; \ + err |= __get_user(tmp, &sc->seg); \ + copyseg_##seg(tmp|3); \ + } + + /* To make COPY_SEGs easier, we zero r16, r17 */ + regs->r16 = 0; + regs->r17 = 0; + + COPY_SEG(gs); + COPY_SEG(fs); + COPY_SEG(es); + COPY_SEG(ds); + COPY(r15, edi); + COPY(r14, esi); + COPY(r13, ebp); + COPY(r12, esp); + COPY(r11, ebx); + COPY(r10, edx); + COPY(r9, ecx); + COPY(cr_iip, eip); + COPY_SEG_STRICT(cs); + COPY_SEG_STRICT(ss); + ia32_load_segment_descriptors(current); + { + unsigned int tmpflags; + unsigned long flag; + + /* + * IA32 `eflags' is not part of `pt_regs', it's in an ar register which + * is part of the thread context. Fortunately, we are executing in the + * IA32 process's context. + */ + err |= __get_user(tmpflags, &sc->eflags); + flag = ia64_getreg(_IA64_REG_AR_EFLAG); + flag &= ~0x40DD5; + flag |= (tmpflags & 0x40DD5); + ia64_setreg(_IA64_REG_AR_EFLAG, flag); + + regs->r1 = -1; /* disable syscall checks, r1 is orig_eax */ + } + + { + struct _fpstate_ia32 __user *buf = NULL; + u32 fpstate_ptr; + err |= get_user(fpstate_ptr, &(sc->fpstate)); + buf = compat_ptr(fpstate_ptr); + if (buf) { + err |= restore_ia32_fpstate_live(buf); + } + } + +#if 0 + { + struct _fpstate * buf; + err |= __get_user(buf, &sc->fpstate); + if (buf) { + if (!access_ok(VERIFY_READ, buf, sizeof(*buf))) + goto badframe; + err |= restore_i387(buf); + } + } +#endif + + err |= __get_user(*peax, &sc->eax); + return err; + +#if 0 + badframe: + return 1; +#endif +} + +/* + * Determine which stack to use.. + */ +static inline void __user * +get_sigframe (struct k_sigaction *ka, struct pt_regs * regs, size_t frame_size) +{ + unsigned long esp; + + /* Default to using normal stack (truncate off sign-extension of bit 31: */ + esp = (unsigned int) regs->r12; + + /* This is the X/Open sanctioned signal stack switching. */ + if (ka->sa.sa_flags & SA_ONSTACK) { + if (!on_sig_stack(esp)) + esp = current->sas_ss_sp + current->sas_ss_size; + } + /* Legacy stack switching not supported */ + + return (void __user *)((esp - frame_size) & -8ul); +} + +static int +setup_frame_ia32 (int sig, struct k_sigaction *ka, sigset_t *set, struct pt_regs * regs) +{ + struct exec_domain *ed = current_thread_info()->exec_domain; + struct sigframe_ia32 __user *frame; + int err = 0; + + frame = get_sigframe(ka, regs, sizeof(*frame)); + + if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) + goto give_sigsegv; + + err |= __put_user((ed && ed->signal_invmap && sig < 32 + ? (int)(ed->signal_invmap[sig]) : sig), &frame->sig); + + err |= setup_sigcontext_ia32(&frame->sc, &frame->fpstate, regs, set->sig[0]); + + if (_COMPAT_NSIG_WORDS > 1) + err |= __copy_to_user(frame->extramask, (char *) &set->sig + 4, + sizeof(frame->extramask)); + + /* Set up to return from userspace. If provided, use a stub + already in userspace. */ + if (ka->sa.sa_flags & SA_RESTORER) { + unsigned int restorer = IA32_SA_RESTORER(ka); + err |= __put_user(restorer, &frame->pretcode); + } else { + /* Pointing to restorer in ia32 gate page */ + err |= __put_user(IA32_GATE_OFFSET, &frame->pretcode); + } + + /* This is popl %eax ; movl $,%eax ; int $0x80 + * and there for historical reasons only. + * See arch/i386/kernel/signal.c + */ + + err |= __put_user(0xb858, (short __user *)(frame->retcode+0)); + err |= __put_user(__IA32_NR_sigreturn, (int __user *)(frame->retcode+2)); + err |= __put_user(0x80cd, (short __user *)(frame->retcode+6)); + + if (err) + goto give_sigsegv; + + /* Set up registers for signal handler */ + regs->r12 = (unsigned long) frame; + regs->cr_iip = IA32_SA_HANDLER(ka); + + set_fs(USER_DS); + +#if 0 + regs->eflags &= ~TF_MASK; +#endif + +#if 0 + printk("SIG deliver (%s:%d): sig=%d sp=%p pc=%lx ra=%x\n", + current->comm, current->pid, sig, (void *) frame, regs->cr_iip, frame->pretcode); +#endif + + return 1; + + give_sigsegv: + force_sigsegv(sig, current); + return 0; +} + +static int +setup_rt_frame_ia32 (int sig, struct k_sigaction *ka, siginfo_t *info, + sigset_t *set, struct pt_regs * regs) +{ + struct exec_domain *ed = current_thread_info()->exec_domain; + compat_uptr_t pinfo, puc; + struct rt_sigframe_ia32 __user *frame; + int err = 0; + + frame = get_sigframe(ka, regs, sizeof(*frame)); + + if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) + goto give_sigsegv; + + err |= __put_user((ed && ed->signal_invmap + && sig < 32 ? ed->signal_invmap[sig] : sig), &frame->sig); + + pinfo = (long __user) &frame->info; + puc = (long __user) &frame->uc; + err |= __put_user(pinfo, &frame->pinfo); + err |= __put_user(puc, &frame->puc); + err |= copy_siginfo_to_user32(&frame->info, info); + + /* Create the ucontext. */ + err |= __put_user(0, &frame->uc.uc_flags); + err |= __put_user(0, &frame->uc.uc_link); + err |= __put_user(current->sas_ss_sp, &frame->uc.uc_stack.ss_sp); + err |= __put_user(sas_ss_flags(regs->r12), &frame->uc.uc_stack.ss_flags); + err |= __put_user(current->sas_ss_size, &frame->uc.uc_stack.ss_size); + err |= setup_sigcontext_ia32(&frame->uc.uc_mcontext, &frame->fpstate, regs, set->sig[0]); + err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); + if (err) + goto give_sigsegv; + + /* Set up to return from userspace. If provided, use a stub + already in userspace. */ + if (ka->sa.sa_flags & SA_RESTORER) { + unsigned int restorer = IA32_SA_RESTORER(ka); + err |= __put_user(restorer, &frame->pretcode); + } else { + /* Pointing to rt_restorer in ia32 gate page */ + err |= __put_user(IA32_GATE_OFFSET + 8, &frame->pretcode); + } + + /* This is movl $,%eax ; int $0x80 + * and there for historical reasons only. + * See arch/i386/kernel/signal.c + */ + + err |= __put_user(0xb8, (char __user *)(frame->retcode+0)); + err |= __put_user(__IA32_NR_rt_sigreturn, (int __user *)(frame->retcode+1)); + err |= __put_user(0x80cd, (short __user *)(frame->retcode+5)); + + if (err) + goto give_sigsegv; + + /* Set up registers for signal handler */ + regs->r12 = (unsigned long) frame; + regs->cr_iip = IA32_SA_HANDLER(ka); + + set_fs(USER_DS); + +#if 0 + regs->eflags &= ~TF_MASK; +#endif + +#if 0 + printk("SIG deliver (%s:%d): sp=%p pc=%lx ra=%x\n", + current->comm, current->pid, (void *) frame, regs->cr_iip, frame->pretcode); +#endif + + return 1; + +give_sigsegv: + force_sigsegv(sig, current); + return 0; +} + +int +ia32_setup_frame1 (int sig, struct k_sigaction *ka, siginfo_t *info, + sigset_t *set, struct pt_regs *regs) +{ + /* Set up the stack frame */ + if (ka->sa.sa_flags & SA_SIGINFO) + return setup_rt_frame_ia32(sig, ka, info, set, regs); + else + return setup_frame_ia32(sig, ka, set, regs); +} + +asmlinkage long +sys32_sigreturn (int arg0, int arg1, int arg2, int arg3, int arg4, int arg5, + int arg6, int arg7, struct pt_regs regs) +{ + unsigned long esp = (unsigned int) regs.r12; + struct sigframe_ia32 __user *frame = (struct sigframe_ia32 __user *)(esp - 8); + sigset_t set; + int eax; + + if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) + goto badframe; + + if (__get_user(set.sig[0], &frame->sc.oldmask) + || (_COMPAT_NSIG_WORDS > 1 && __copy_from_user((char *) &set.sig + 4, &frame->extramask, + sizeof(frame->extramask)))) + goto badframe; + + sigdelsetmask(&set, ~_BLOCKABLE); + spin_lock_irq(¤t->sighand->siglock); + current->blocked = set; + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + + if (restore_sigcontext_ia32(®s, &frame->sc, &eax)) + goto badframe; + return eax; + + badframe: + force_sig(SIGSEGV, current); + return 0; +} + +asmlinkage long +sys32_rt_sigreturn (int arg0, int arg1, int arg2, int arg3, int arg4, + int arg5, int arg6, int arg7, struct pt_regs regs) +{ + unsigned long esp = (unsigned int) regs.r12; + struct rt_sigframe_ia32 __user *frame = (struct rt_sigframe_ia32 __user *)(esp - 4); + sigset_t set; + int eax; + + if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) + goto badframe; + if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) + goto badframe; + + sigdelsetmask(&set, ~_BLOCKABLE); + spin_lock_irq(¤t->sighand->siglock); + current->blocked = set; + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + + if (restore_sigcontext_ia32(®s, &frame->uc.uc_mcontext, &eax)) + goto badframe; + + /* It is more difficult to avoid calling this function than to + call it and ignore errors. */ + do_sigaltstack((stack_t __user *) &frame->uc.uc_stack, NULL, esp); + + return eax; + + badframe: + force_sig(SIGSEGV, current); + return 0; +} diff --git a/arch/ia64/ia32/ia32_support.c b/arch/ia64/ia32/ia32_support.c new file mode 100644 index 000000000000..4f630043b3ae --- /dev/null +++ b/arch/ia64/ia32/ia32_support.c @@ -0,0 +1,264 @@ +/* + * IA32 helper functions + * + * Copyright (C) 1999 Arun Sharma + * Copyright (C) 2000 Asit K. Mallick + * Copyright (C) 2001-2002 Hewlett-Packard Co + * David Mosberger-Tang + * + * 06/16/00 A. Mallick added csd/ssd/tssd for ia32 thread context + * 02/19/01 D. Mosberger dropped tssd; it's not needed + * 09/14/01 D. Mosberger fixed memory management for gdt/tss page + * 09/29/01 D. Mosberger added ia32_load_segment_descriptors() + */ + +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#include "ia32priv.h" + +extern void die_if_kernel (char *str, struct pt_regs *regs, long err); + +struct exec_domain ia32_exec_domain; +struct page *ia32_shared_page[NR_CPUS]; +unsigned long *ia32_boot_gdt; +unsigned long *cpu_gdt_table[NR_CPUS]; +struct page *ia32_gate_page; + +static unsigned long +load_desc (u16 selector) +{ + unsigned long *table, limit, index; + + if (!selector) + return 0; + if (selector & IA32_SEGSEL_TI) { + table = (unsigned long *) IA32_LDT_OFFSET; + limit = IA32_LDT_ENTRIES; + } else { + table = cpu_gdt_table[smp_processor_id()]; + limit = IA32_PAGE_SIZE / sizeof(ia32_boot_gdt[0]); + } + index = selector >> IA32_SEGSEL_INDEX_SHIFT; + if (index >= limit) + return 0; + return IA32_SEG_UNSCRAMBLE(table[index]); +} + +void +ia32_load_segment_descriptors (struct task_struct *task) +{ + struct pt_regs *regs = ia64_task_regs(task); + + /* Setup the segment descriptors */ + regs->r24 = load_desc(regs->r16 >> 16); /* ESD */ + regs->r27 = load_desc(regs->r16 >> 0); /* DSD */ + regs->r28 = load_desc(regs->r16 >> 32); /* FSD */ + regs->r29 = load_desc(regs->r16 >> 48); /* GSD */ + regs->ar_csd = load_desc(regs->r17 >> 0); /* CSD */ + regs->ar_ssd = load_desc(regs->r17 >> 16); /* SSD */ +} + +int +ia32_clone_tls (struct task_struct *child, struct pt_regs *childregs) +{ + struct desc_struct *desc; + struct ia32_user_desc info; + int idx; + + if (copy_from_user(&info, (void __user *)(childregs->r14 & 0xffffffff), sizeof(info))) + return -EFAULT; + if (LDT_empty(&info)) + return -EINVAL; + + idx = info.entry_number; + if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX) + return -EINVAL; + + desc = child->thread.tls_array + idx - GDT_ENTRY_TLS_MIN; + desc->a = LDT_entry_a(&info); + desc->b = LDT_entry_b(&info); + + /* XXX: can this be done in a cleaner way ? */ + load_TLS(&child->thread, smp_processor_id()); + ia32_load_segment_descriptors(child); + load_TLS(¤t->thread, smp_processor_id()); + + return 0; +} + +void +ia32_save_state (struct task_struct *t) +{ + t->thread.eflag = ia64_getreg(_IA64_REG_AR_EFLAG); + t->thread.fsr = ia64_getreg(_IA64_REG_AR_FSR); + t->thread.fcr = ia64_getreg(_IA64_REG_AR_FCR); + t->thread.fir = ia64_getreg(_IA64_REG_AR_FIR); + t->thread.fdr = ia64_getreg(_IA64_REG_AR_FDR); + ia64_set_kr(IA64_KR_IO_BASE, t->thread.old_iob); + ia64_set_kr(IA64_KR_TSSD, t->thread.old_k1); +} + +void +ia32_load_state (struct task_struct *t) +{ + unsigned long eflag, fsr, fcr, fir, fdr, tssd; + struct pt_regs *regs = ia64_task_regs(t); + + eflag = t->thread.eflag; + fsr = t->thread.fsr; + fcr = t->thread.fcr; + fir = t->thread.fir; + fdr = t->thread.fdr; + tssd = load_desc(_TSS); /* TSSD */ + + ia64_setreg(_IA64_REG_AR_EFLAG, eflag); + ia64_setreg(_IA64_REG_AR_FSR, fsr); + ia64_setreg(_IA64_REG_AR_FCR, fcr); + ia64_setreg(_IA64_REG_AR_FIR, fir); + ia64_setreg(_IA64_REG_AR_FDR, fdr); + current->thread.old_iob = ia64_get_kr(IA64_KR_IO_BASE); + current->thread.old_k1 = ia64_get_kr(IA64_KR_TSSD); + ia64_set_kr(IA64_KR_IO_BASE, IA32_IOBASE); + ia64_set_kr(IA64_KR_TSSD, tssd); + + regs->r17 = (_TSS << 48) | (_LDT << 32) | (__u32) regs->r17; + regs->r30 = load_desc(_LDT); /* LDTD */ + load_TLS(&t->thread, smp_processor_id()); +} + +/* + * Setup IA32 GDT and TSS + */ +void +ia32_gdt_init (void) +{ + int cpu = smp_processor_id(); + + ia32_shared_page[cpu] = alloc_page(GFP_KERNEL); + if (!ia32_shared_page[cpu]) + panic("failed to allocate ia32_shared_page[%d]\n", cpu); + + cpu_gdt_table[cpu] = page_address(ia32_shared_page[cpu]); + + /* Copy from the boot cpu's GDT */ + memcpy(cpu_gdt_table[cpu], ia32_boot_gdt, PAGE_SIZE); +} + + +/* + * Setup IA32 GDT and TSS + */ +static void +ia32_boot_gdt_init (void) +{ + unsigned long ldt_size; + + ia32_shared_page[0] = alloc_page(GFP_KERNEL); + if (!ia32_shared_page[0]) + panic("failed to allocate ia32_shared_page[0]\n"); + + ia32_boot_gdt = page_address(ia32_shared_page[0]); + cpu_gdt_table[0] = ia32_boot_gdt; + + /* CS descriptor in IA-32 (scrambled) format */ + ia32_boot_gdt[__USER_CS >> 3] + = IA32_SEG_DESCRIPTOR(0, (IA32_GATE_END-1) >> IA32_PAGE_SHIFT, + 0xb, 1, 3, 1, 1, 1, 1); + + /* DS descriptor in IA-32 (scrambled) format */ + ia32_boot_gdt[__USER_DS >> 3] + = IA32_SEG_DESCRIPTOR(0, (IA32_GATE_END-1) >> IA32_PAGE_SHIFT, + 0x3, 1, 3, 1, 1, 1, 1); + + ldt_size = PAGE_ALIGN(IA32_LDT_ENTRIES*IA32_LDT_ENTRY_SIZE); + ia32_boot_gdt[TSS_ENTRY] = IA32_SEG_DESCRIPTOR(IA32_TSS_OFFSET, 235, + 0xb, 0, 3, 1, 1, 1, 0); + ia32_boot_gdt[LDT_ENTRY] = IA32_SEG_DESCRIPTOR(IA32_LDT_OFFSET, ldt_size - 1, + 0x2, 0, 3, 1, 1, 1, 0); +} + +static void +ia32_gate_page_init(void) +{ + unsigned long *sr; + + ia32_gate_page = alloc_page(GFP_KERNEL); + sr = page_address(ia32_gate_page); + /* This is popl %eax ; movl $,%eax ; int $0x80 */ + *sr++ = 0xb858 | (__IA32_NR_sigreturn << 16) | (0x80cdUL << 48); + + /* This is movl $,%eax ; int $0x80 */ + *sr = 0xb8 | (__IA32_NR_rt_sigreturn << 8) | (0x80cdUL << 40); +} + +void +ia32_mem_init(void) +{ + ia32_boot_gdt_init(); + ia32_gate_page_init(); +} + +/* + * Handle bad IA32 interrupt via syscall + */ +void +ia32_bad_interrupt (unsigned long int_num, struct pt_regs *regs) +{ + siginfo_t siginfo; + + die_if_kernel("Bad IA-32 interrupt", regs, int_num); + + siginfo.si_signo = SIGTRAP; + siginfo.si_errno = int_num; /* XXX is it OK to abuse si_errno like this? */ + siginfo.si_flags = 0; + siginfo.si_isr = 0; + siginfo.si_addr = NULL; + siginfo.si_imm = 0; + siginfo.si_code = TRAP_BRKPT; + force_sig_info(SIGTRAP, &siginfo, current); +} + +void +ia32_cpu_init (void) +{ + /* initialize global ia32 state - CR0 and CR4 */ + ia64_setreg(_IA64_REG_AR_CFLAG, (((ulong) IA32_CR4 << 32) | IA32_CR0)); +} + +static int __init +ia32_init (void) +{ + ia32_exec_domain.name = "Linux/x86"; + ia32_exec_domain.handler = NULL; + ia32_exec_domain.pers_low = PER_LINUX32; + ia32_exec_domain.pers_high = PER_LINUX32; + ia32_exec_domain.signal_map = default_exec_domain.signal_map; + ia32_exec_domain.signal_invmap = default_exec_domain.signal_invmap; + register_exec_domain(&ia32_exec_domain); + +#if PAGE_SHIFT > IA32_PAGE_SHIFT + { + extern kmem_cache_t *partial_page_cachep; + + partial_page_cachep = kmem_cache_create("partial_page_cache", + sizeof(struct partial_page), 0, 0, + NULL, NULL); + if (!partial_page_cachep) + panic("Cannot create partial page SLAB cache"); + } +#endif + return 0; +} + +__initcall(ia32_init); diff --git a/arch/ia64/ia32/ia32_traps.c b/arch/ia64/ia32/ia32_traps.c new file mode 100644 index 000000000000..e486042672f1 --- /dev/null +++ b/arch/ia64/ia32/ia32_traps.c @@ -0,0 +1,156 @@ +/* + * IA-32 exception handlers + * + * Copyright (C) 2000 Asit K. Mallick + * Copyright (C) 2001-2002 Hewlett-Packard Co + * David Mosberger-Tang + * + * 06/16/00 A. Mallick added siginfo for most cases (close to IA32) + * 09/29/00 D. Mosberger added ia32_intercept() + */ + +#include +#include + +#include "ia32priv.h" + +#include +#include + +int +ia32_intercept (struct pt_regs *regs, unsigned long isr) +{ + switch ((isr >> 16) & 0xff) { + case 0: /* Instruction intercept fault */ + case 4: /* Locked Data reference fault */ + case 1: /* Gate intercept trap */ + return -1; + + case 2: /* System flag trap */ + if (((isr >> 14) & 0x3) >= 2) { + /* MOV SS, POP SS instructions */ + ia64_psr(regs)->id = 1; + return 0; + } else + return -1; + } + return -1; +} + +int +ia32_exception (struct pt_regs *regs, unsigned long isr) +{ + struct siginfo siginfo; + + /* initialize these fields to avoid leaking kernel bits to user space: */ + siginfo.si_errno = 0; + siginfo.si_flags = 0; + siginfo.si_isr = 0; + siginfo.si_imm = 0; + switch ((isr >> 16) & 0xff) { + case 1: + case 2: + siginfo.si_signo = SIGTRAP; + if (isr == 0) + siginfo.si_code = TRAP_TRACE; + else if (isr & 0x4) + siginfo.si_code = TRAP_BRANCH; + else + siginfo.si_code = TRAP_BRKPT; + break; + + case 3: + siginfo.si_signo = SIGTRAP; + siginfo.si_code = TRAP_BRKPT; + break; + + case 0: /* Divide fault */ + siginfo.si_signo = SIGFPE; + siginfo.si_code = FPE_INTDIV; + break; + + case 4: /* Overflow */ + case 5: /* Bounds fault */ + siginfo.si_signo = SIGFPE; + siginfo.si_code = 0; + break; + + case 6: /* Invalid Op-code */ + siginfo.si_signo = SIGILL; + siginfo.si_code = ILL_ILLOPN; + break; + + case 7: /* FP DNA */ + case 8: /* Double Fault */ + case 9: /* Invalid TSS */ + case 11: /* Segment not present */ + case 12: /* Stack fault */ + case 13: /* General Protection Fault */ + siginfo.si_signo = SIGSEGV; + siginfo.si_code = 0; + break; + + case 16: /* Pending FP error */ + { + unsigned long fsr, fcr; + + fsr = ia64_getreg(_IA64_REG_AR_FSR); + fcr = ia64_getreg(_IA64_REG_AR_FCR); + + siginfo.si_signo = SIGFPE; + /* + * (~cwd & swd) will mask out exceptions that are not set to unmasked + * status. 0x3f is the exception bits in these regs, 0x200 is the + * C1 reg you need in case of a stack fault, 0x040 is the stack + * fault bit. We should only be taking one exception at a time, + * so if this combination doesn't produce any single exception, + * then we have a bad program that isn't synchronizing its FPU usage + * and it will suffer the consequences since we won't be able to + * fully reproduce the context of the exception + */ + siginfo.si_isr = isr; + siginfo.si_flags = __ISR_VALID; + switch(((~fcr) & (fsr & 0x3f)) | (fsr & 0x240)) { + case 0x000: + default: + siginfo.si_code = 0; + break; + case 0x001: /* Invalid Op */ + case 0x040: /* Stack Fault */ + case 0x240: /* Stack Fault | Direction */ + siginfo.si_code = FPE_FLTINV; + break; + case 0x002: /* Denormalize */ + case 0x010: /* Underflow */ + siginfo.si_code = FPE_FLTUND; + break; + case 0x004: /* Zero Divide */ + siginfo.si_code = FPE_FLTDIV; + break; + case 0x008: /* Overflow */ + siginfo.si_code = FPE_FLTOVF; + break; + case 0x020: /* Precision */ + siginfo.si_code = FPE_FLTRES; + break; + } + + break; + } + + case 17: /* Alignment check */ + siginfo.si_signo = SIGSEGV; + siginfo.si_code = BUS_ADRALN; + break; + + case 19: /* SSE Numeric error */ + siginfo.si_signo = SIGFPE; + siginfo.si_code = 0; + break; + + default: + return -1; + } + force_sig_info(siginfo.si_signo, &siginfo, current); + return 0; +} diff --git a/arch/ia64/ia32/ia32priv.h b/arch/ia64/ia32/ia32priv.h new file mode 100644 index 000000000000..b2de948bdaea --- /dev/null +++ b/arch/ia64/ia32/ia32priv.h @@ -0,0 +1,544 @@ +#ifndef _ASM_IA64_IA32_PRIV_H +#define _ASM_IA64_IA32_PRIV_H + +#include + +#include + +#ifdef CONFIG_IA32_SUPPORT + +#include +#include +#include + +#include + +/* + * 32 bit structures for IA32 support. + */ + +#define IA32_PAGE_SIZE (1UL << IA32_PAGE_SHIFT) +#define IA32_PAGE_MASK (~(IA32_PAGE_SIZE - 1)) +#define IA32_PAGE_ALIGN(addr) (((addr) + IA32_PAGE_SIZE - 1) & IA32_PAGE_MASK) +#define IA32_CLOCKS_PER_SEC 100 /* Cast in stone for IA32 Linux */ + +/* + * partially mapped pages provide precise accounting of which 4k sub pages + * are mapped and which ones are not, thereby improving IA-32 compatibility. + */ +struct partial_page { + struct partial_page *next; /* linked list, sorted by address */ + struct rb_node pp_rb; + /* 64K is the largest "normal" page supported by ia64 ABI. So 4K*32 + * should suffice.*/ + unsigned int bitmap; + unsigned int base; +}; + +struct partial_page_list { + struct partial_page *pp_head; /* list head, points to the lowest + * addressed partial page */ + struct rb_root ppl_rb; + struct partial_page *pp_hint; /* pp_hint->next is the last + * accessed partial page */ + atomic_t pp_count; /* reference count */ +}; + +#if PAGE_SHIFT > IA32_PAGE_SHIFT +struct partial_page_list* ia32_init_pp_list (void); +#else +# define ia32_init_pp_list() 0 +#endif + +/* sigcontext.h */ +/* + * As documented in the iBCS2 standard.. + * + * The first part of "struct _fpstate" is just the + * normal i387 hardware setup, the extra "status" + * word is used to save the coprocessor status word + * before entering the handler. + */ +struct _fpreg_ia32 { + unsigned short significand[4]; + unsigned short exponent; +}; + +struct _fpxreg_ia32 { + unsigned short significand[4]; + unsigned short exponent; + unsigned short padding[3]; +}; + +struct _xmmreg_ia32 { + unsigned int element[4]; +}; + + +struct _fpstate_ia32 { + unsigned int cw, + sw, + tag, + ipoff, + cssel, + dataoff, + datasel; + struct _fpreg_ia32 _st[8]; + unsigned short status; + unsigned short magic; /* 0xffff = regular FPU data only */ + + /* FXSR FPU environment */ + unsigned int _fxsr_env[6]; /* FXSR FPU env is ignored */ + unsigned int mxcsr; + unsigned int reserved; + struct _fpxreg_ia32 _fxsr_st[8]; /* FXSR FPU reg data is ignored */ + struct _xmmreg_ia32 _xmm[8]; + unsigned int padding[56]; +}; + +struct sigcontext_ia32 { + unsigned short gs, __gsh; + unsigned short fs, __fsh; + unsigned short es, __esh; + unsigned short ds, __dsh; + unsigned int edi; + unsigned int esi; + unsigned int ebp; + unsigned int esp; + unsigned int ebx; + unsigned int edx; + unsigned int ecx; + unsigned int eax; + unsigned int trapno; + unsigned int err; + unsigned int eip; + unsigned short cs, __csh; + unsigned int eflags; + unsigned int esp_at_signal; + unsigned short ss, __ssh; + unsigned int fpstate; /* really (struct _fpstate_ia32 *) */ + unsigned int oldmask; + unsigned int cr2; +}; + +/* user.h */ +/* + * IA32 (Pentium III/4) FXSR, SSE support + * + * Provide support for the GDB 5.0+ PTRACE_{GET|SET}FPXREGS requests for + * interacting with the FXSR-format floating point environment. Floating + * point data can be accessed in the regular format in the usual manner, + * and both the standard and SIMD floating point data can be accessed via + * the new ptrace requests. In either case, changes to the FPU environment + * will be reflected in the task's state as expected. + */ +struct ia32_user_i387_struct { + int cwd; + int swd; + int twd; + int fip; + int fcs; + int foo; + int fos; + /* 8*10 bytes for each FP-reg = 80 bytes */ + struct _fpreg_ia32 st_space[8]; +}; + +struct ia32_user_fxsr_struct { + unsigned short cwd; + unsigned short swd; + unsigned short twd; + unsigned short fop; + int fip; + int fcs; + int foo; + int fos; + int mxcsr; + int reserved; + int st_space[32]; /* 8*16 bytes for each FP-reg = 128 bytes */ + int xmm_space[32]; /* 8*16 bytes for each XMM-reg = 128 bytes */ + int padding[56]; +}; + +/* signal.h */ +#define IA32_SET_SA_HANDLER(ka,handler,restorer) \ + ((ka)->sa.sa_handler = (__sighandler_t) \ + (((unsigned long)(restorer) << 32) \ + | ((handler) & 0xffffffff))) +#define IA32_SA_HANDLER(ka) ((unsigned long) (ka)->sa.sa_handler & 0xffffffff) +#define IA32_SA_RESTORER(ka) ((unsigned long) (ka)->sa.sa_handler >> 32) + +#define __IA32_NR_sigreturn 119 +#define __IA32_NR_rt_sigreturn 173 + +struct sigaction32 { + unsigned int sa_handler; /* Really a pointer, but need to deal with 32 bits */ + unsigned int sa_flags; + unsigned int sa_restorer; /* Another 32 bit pointer */ + compat_sigset_t sa_mask; /* A 32 bit mask */ +}; + +struct old_sigaction32 { + unsigned int sa_handler; /* Really a pointer, but need to deal + with 32 bits */ + compat_old_sigset_t sa_mask; /* A 32 bit mask */ + unsigned int sa_flags; + unsigned int sa_restorer; /* Another 32 bit pointer */ +}; + +typedef struct sigaltstack_ia32 { + unsigned int ss_sp; + int ss_flags; + unsigned int ss_size; +} stack_ia32_t; + +struct ucontext_ia32 { + unsigned int uc_flags; + unsigned int uc_link; + stack_ia32_t uc_stack; + struct sigcontext_ia32 uc_mcontext; + sigset_t uc_sigmask; /* mask last for extensibility */ +}; + +struct stat64 { + unsigned long long st_dev; + unsigned char __pad0[4]; + unsigned int __st_ino; + unsigned int st_mode; + unsigned int st_nlink; + unsigned int st_uid; + unsigned int st_gid; + unsigned long long st_rdev; + unsigned char __pad3[4]; + unsigned int st_size_lo; + unsigned int st_size_hi; + unsigned int st_blksize; + unsigned int st_blocks; /* Number 512-byte blocks allocated. */ + unsigned int __pad4; /* future possible st_blocks high bits */ + unsigned int st_atime; + unsigned int st_atime_nsec; + unsigned int st_mtime; + unsigned int st_mtime_nsec; + unsigned int st_ctime; + unsigned int st_ctime_nsec; + unsigned int st_ino_lo; + unsigned int st_ino_hi; +}; + +typedef struct compat_siginfo { + int si_signo; + int si_errno; + int si_code; + + union { + int _pad[((128/sizeof(int)) - 3)]; + + /* kill() */ + struct { + unsigned int _pid; /* sender's pid */ + unsigned int _uid; /* sender's uid */ + } _kill; + + /* POSIX.1b timers */ + struct { + timer_t _tid; /* timer id */ + int _overrun; /* overrun count */ + char _pad[sizeof(unsigned int) - sizeof(int)]; + compat_sigval_t _sigval; /* same as below */ + int _sys_private; /* not to be passed to user */ + } _timer; + + /* POSIX.1b signals */ + struct { + unsigned int _pid; /* sender's pid */ + unsigned int _uid; /* sender's uid */ + compat_sigval_t _sigval; + } _rt; + + /* SIGCHLD */ + struct { + unsigned int _pid; /* which child */ + unsigned int _uid; /* sender's uid */ + int _status; /* exit code */ + compat_clock_t _utime; + compat_clock_t _stime; + } _sigchld; + + /* SIGILL, SIGFPE, SIGSEGV, SIGBUS */ + struct { + unsigned int _addr; /* faulting insn/memory ref. */ + } _sigfault; + + /* SIGPOLL */ + struct { + int _band; /* POLL_IN, POLL_OUT, POLL_MSG */ + int _fd; + } _sigpoll; + } _sifields; +} compat_siginfo_t; + +struct old_linux32_dirent { + u32 d_ino; + u32 d_offset; + u16 d_namlen; + char d_name[1]; +}; + +/* + * IA-32 ELF specific definitions for IA-64. + */ + +#define _ASM_IA64_ELF_H /* Don't include elf.h */ + +#include +#include + +/* + * This is used to ensure we don't load something for the wrong architecture. + */ +#define elf_check_arch(x) ((x)->e_machine == EM_386) + +/* + * These are used to set parameters in the core dumps. + */ +#define ELF_CLASS ELFCLASS32 +#define ELF_DATA ELFDATA2LSB +#define ELF_ARCH EM_386 + +#define IA32_PAGE_OFFSET 0xc0000000 +#define IA32_STACK_TOP IA32_PAGE_OFFSET +#define IA32_GATE_OFFSET IA32_PAGE_OFFSET +#define IA32_GATE_END IA32_PAGE_OFFSET + PAGE_SIZE + +/* + * The system segments (GDT, TSS, LDT) have to be mapped below 4GB so the IA-32 engine can + * access them. + */ +#define IA32_GDT_OFFSET (IA32_PAGE_OFFSET + PAGE_SIZE) +#define IA32_TSS_OFFSET (IA32_PAGE_OFFSET + 2*PAGE_SIZE) +#define IA32_LDT_OFFSET (IA32_PAGE_OFFSET + 3*PAGE_SIZE) + +#define ELF_EXEC_PAGESIZE IA32_PAGE_SIZE + +/* + * This is the location that an ET_DYN program is loaded if exec'ed. + * Typical use of this is to invoke "./ld.so someprog" to test out a + * new version of the loader. We need to make sure that it is out of + * the way of the program that it will "exec", and that there is + * sufficient room for the brk. + */ +#define ELF_ET_DYN_BASE (IA32_PAGE_OFFSET/3 + 0x1000000) + +void ia64_elf32_init(struct pt_regs *regs); +#define ELF_PLAT_INIT(_r, load_addr) ia64_elf32_init(_r) + +#define elf_addr_t u32 + +/* This macro yields a bitmask that programs can use to figure out + what instruction set this CPU supports. */ +#define ELF_HWCAP 0 + +/* This macro yields a string that ld.so will use to load + implementation specific libraries for optimization. Not terribly + relevant until we have real hardware to play with... */ +#define ELF_PLATFORM NULL + +#ifdef __KERNEL__ +# define SET_PERSONALITY(EX,IBCS2) \ + (current->personality = (IBCS2) ? PER_SVR4 : PER_LINUX) +#endif + +#define IA32_EFLAG 0x200 + +/* + * IA-32 ELF specific definitions for IA-64. + */ + +#define __USER_CS 0x23 +#define __USER_DS 0x2B + +/* + * The per-cpu GDT has 32 entries: see + */ +#define GDT_ENTRIES 32 + +#define GDT_SIZE (GDT_ENTRIES * 8) + +#define TSS_ENTRY 14 +#define LDT_ENTRY (TSS_ENTRY + 1) + +#define IA32_SEGSEL_RPL (0x3 << 0) +#define IA32_SEGSEL_TI (0x1 << 2) +#define IA32_SEGSEL_INDEX_SHIFT 3 + +#define _TSS ((unsigned long) TSS_ENTRY << IA32_SEGSEL_INDEX_SHIFT) +#define _LDT ((unsigned long) LDT_ENTRY << IA32_SEGSEL_INDEX_SHIFT) + +#define IA32_SEG_BASE 16 +#define IA32_SEG_TYPE 40 +#define IA32_SEG_SYS 44 +#define IA32_SEG_DPL 45 +#define IA32_SEG_P 47 +#define IA32_SEG_HIGH_LIMIT 48 +#define IA32_SEG_AVL 52 +#define IA32_SEG_DB 54 +#define IA32_SEG_G 55 +#define IA32_SEG_HIGH_BASE 56 + +#define IA32_SEG_DESCRIPTOR(base, limit, segtype, nonsysseg, dpl, segpresent, avl, segdb, gran) \ + (((limit) & 0xffff) \ + | (((unsigned long) (base) & 0xffffff) << IA32_SEG_BASE) \ + | ((unsigned long) (segtype) << IA32_SEG_TYPE) \ + | ((unsigned long) (nonsysseg) << IA32_SEG_SYS) \ + | ((unsigned long) (dpl) << IA32_SEG_DPL) \ + | ((unsigned long) (segpresent) << IA32_SEG_P) \ + | ((((unsigned long) (limit) >> 16) & 0xf) << IA32_SEG_HIGH_LIMIT) \ + | ((unsigned long) (avl) << IA32_SEG_AVL) \ + | ((unsigned long) (segdb) << IA32_SEG_DB) \ + | ((unsigned long) (gran) << IA32_SEG_G) \ + | ((((unsigned long) (base) >> 24) & 0xff) << IA32_SEG_HIGH_BASE)) + +#define SEG_LIM 32 +#define SEG_TYPE 52 +#define SEG_SYS 56 +#define SEG_DPL 57 +#define SEG_P 59 +#define SEG_AVL 60 +#define SEG_DB 62 +#define SEG_G 63 + +/* Unscramble an IA-32 segment descriptor into the IA-64 format. */ +#define IA32_SEG_UNSCRAMBLE(sd) \ + ( (((sd) >> IA32_SEG_BASE) & 0xffffff) | ((((sd) >> IA32_SEG_HIGH_BASE) & 0xff) << 24) \ + | ((((sd) & 0xffff) | ((((sd) >> IA32_SEG_HIGH_LIMIT) & 0xf) << 16)) << SEG_LIM) \ + | ((((sd) >> IA32_SEG_TYPE) & 0xf) << SEG_TYPE) \ + | ((((sd) >> IA32_SEG_SYS) & 0x1) << SEG_SYS) \ + | ((((sd) >> IA32_SEG_DPL) & 0x3) << SEG_DPL) \ + | ((((sd) >> IA32_SEG_P) & 0x1) << SEG_P) \ + | ((((sd) >> IA32_SEG_AVL) & 0x1) << SEG_AVL) \ + | ((((sd) >> IA32_SEG_DB) & 0x1) << SEG_DB) \ + | ((((sd) >> IA32_SEG_G) & 0x1) << SEG_G)) + +#define IA32_IOBASE 0x2000000000000000UL /* Virtual address for I/O space */ + +#define IA32_CR0 0x80000001 /* Enable PG and PE bits */ +#define IA32_CR4 0x600 /* MMXEX and FXSR on */ + +/* + * IA32 floating point control registers starting values + */ + +#define IA32_FSR_DEFAULT 0x55550000 /* set all tag bits */ +#define IA32_FCR_DEFAULT 0x17800000037fUL /* extended precision, all masks */ + +#define IA32_PTRACE_GETREGS 12 +#define IA32_PTRACE_SETREGS 13 +#define IA32_PTRACE_GETFPREGS 14 +#define IA32_PTRACE_SETFPREGS 15 +#define IA32_PTRACE_GETFPXREGS 18 +#define IA32_PTRACE_SETFPXREGS 19 + +#define ia32_start_thread(regs,new_ip,new_sp) do { \ + set_fs(USER_DS); \ + ia64_psr(regs)->cpl = 3; /* set user mode */ \ + ia64_psr(regs)->ri = 0; /* clear return slot number */ \ + ia64_psr(regs)->is = 1; /* IA-32 instruction set */ \ + regs->cr_iip = new_ip; \ + regs->ar_rsc = 0xc; /* enforced lazy mode, priv. level 3 */ \ + regs->ar_rnat = 0; \ + regs->loadrs = 0; \ + regs->r12 = new_sp; \ +} while (0) + +/* + * Local Descriptor Table (LDT) related declarations. + */ + +#define IA32_LDT_ENTRIES 8192 /* Maximum number of LDT entries supported. */ +#define IA32_LDT_ENTRY_SIZE 8 /* The size of each LDT entry. */ + +#define LDT_entry_a(info) \ + ((((info)->base_addr & 0x0000ffff) << 16) | ((info)->limit & 0x0ffff)) + +#define LDT_entry_b(info) \ + (((info)->base_addr & 0xff000000) | \ + (((info)->base_addr & 0x00ff0000) >> 16) | \ + ((info)->limit & 0xf0000) | \ + (((info)->read_exec_only ^ 1) << 9) | \ + ((info)->contents << 10) | \ + (((info)->seg_not_present ^ 1) << 15) | \ + ((info)->seg_32bit << 22) | \ + ((info)->limit_in_pages << 23) | \ + ((info)->useable << 20) | \ + 0x7100) + +#define LDT_empty(info) ( \ + (info)->base_addr == 0 && \ + (info)->limit == 0 && \ + (info)->contents == 0 && \ + (info)->read_exec_only == 1 && \ + (info)->seg_32bit == 0 && \ + (info)->limit_in_pages == 0 && \ + (info)->seg_not_present == 1 && \ + (info)->useable == 0 ) + +static inline void +load_TLS (struct thread_struct *t, unsigned int cpu) +{ + extern unsigned long *cpu_gdt_table[NR_CPUS]; + + memcpy(cpu_gdt_table[cpu] + GDT_ENTRY_TLS_MIN + 0, &t->tls_array[0], sizeof(long)); + memcpy(cpu_gdt_table[cpu] + GDT_ENTRY_TLS_MIN + 1, &t->tls_array[1], sizeof(long)); + memcpy(cpu_gdt_table[cpu] + GDT_ENTRY_TLS_MIN + 2, &t->tls_array[2], sizeof(long)); +} + +struct ia32_user_desc { + unsigned int entry_number; + unsigned int base_addr; + unsigned int limit; + unsigned int seg_32bit:1; + unsigned int contents:2; + unsigned int read_exec_only:1; + unsigned int limit_in_pages:1; + unsigned int seg_not_present:1; + unsigned int useable:1; +}; + +struct linux_binprm; + +extern void ia32_init_addr_space (struct pt_regs *regs); +extern int ia32_setup_arg_pages (struct linux_binprm *bprm, int exec_stack); +extern unsigned long ia32_do_mmap (struct file *, unsigned long, unsigned long, int, int, loff_t); +extern void ia32_load_segment_descriptors (struct task_struct *task); + +#define ia32f2ia64f(dst,src) \ +do { \ + ia64_ldfe(6,src); \ + ia64_stop(); \ + ia64_stf_spill(dst, 6); \ +} while(0) + +#define ia64f2ia32f(dst,src) \ +do { \ + ia64_ldf_fill(6, src); \ + ia64_stop(); \ + ia64_stfe(dst, 6); \ +} while(0) + +struct user_regs_struct32 { + __u32 ebx, ecx, edx, esi, edi, ebp, eax; + unsigned short ds, __ds, es, __es; + unsigned short fs, __fs, gs, __gs; + __u32 orig_eax, eip; + unsigned short cs, __cs; + __u32 eflags, esp; + unsigned short ss, __ss; +}; + +/* Prototypes for use in elfcore32.h */ +extern int save_ia32_fpstate (struct task_struct *, struct ia32_user_i387_struct __user *); +extern int save_ia32_fpxstate (struct task_struct *, struct ia32_user_fxsr_struct __user *); + +#endif /* !CONFIG_IA32_SUPPORT */ + +#endif /* _ASM_IA64_IA32_PRIV_H */ diff --git a/arch/ia64/ia32/sys_ia32.c b/arch/ia64/ia32/sys_ia32.c new file mode 100644 index 000000000000..247a21c64aea --- /dev/null +++ b/arch/ia64/ia32/sys_ia32.c @@ -0,0 +1,2747 @@ +/* + * sys_ia32.c: Conversion between 32bit and 64bit native syscalls. Derived from sys_sparc32.c. + * + * Copyright (C) 2000 VA Linux Co + * Copyright (C) 2000 Don Dugger + * Copyright (C) 1999 Arun Sharma + * Copyright (C) 1997,1998 Jakub Jelinek (jj@sunsite.mff.cuni.cz) + * Copyright (C) 1997 David S. Miller (davem@caip.rutgers.edu) + * Copyright (C) 2000-2003, 2005 Hewlett-Packard Co + * David Mosberger-Tang + * Copyright (C) 2004 Gordon Jin + * + * These routines maintain argument size conversion between 32bit and 64bit + * environment. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#include "ia32priv.h" + +#include +#include + +#define DEBUG 0 + +#if DEBUG +# define DBG(fmt...) printk(KERN_DEBUG fmt) +#else +# define DBG(fmt...) +#endif + +#define ROUND_UP(x,a) ((__typeof__(x))(((unsigned long)(x) + ((a) - 1)) & ~((a) - 1))) + +#define OFFSET4K(a) ((a) & 0xfff) +#define PAGE_START(addr) ((addr) & PAGE_MASK) +#define MINSIGSTKSZ_IA32 2048 + +#define high2lowuid(uid) ((uid) > 65535 ? 65534 : (uid)) +#define high2lowgid(gid) ((gid) > 65535 ? 65534 : (gid)) + +/* + * Anything that modifies or inspects ia32 user virtual memory must hold this semaphore + * while doing so. + */ +/* XXX make per-mm: */ +static DECLARE_MUTEX(ia32_mmap_sem); + +asmlinkage long +sys32_execve (char __user *name, compat_uptr_t __user *argv, compat_uptr_t __user *envp, + struct pt_regs *regs) +{ + long error; + char *filename; + unsigned long old_map_base, old_task_size, tssd; + + filename = getname(name); + error = PTR_ERR(filename); + if (IS_ERR(filename)) + return error; + + old_map_base = current->thread.map_base; + old_task_size = current->thread.task_size; + tssd = ia64_get_kr(IA64_KR_TSSD); + + /* we may be exec'ing a 64-bit process: reset map base, task-size, and io-base: */ + current->thread.map_base = DEFAULT_MAP_BASE; + current->thread.task_size = DEFAULT_TASK_SIZE; + ia64_set_kr(IA64_KR_IO_BASE, current->thread.old_iob); + ia64_set_kr(IA64_KR_TSSD, current->thread.old_k1); + + error = compat_do_execve(filename, argv, envp, regs); + putname(filename); + + if (error < 0) { + /* oops, execve failed, switch back to old values... */ + ia64_set_kr(IA64_KR_IO_BASE, IA32_IOBASE); + ia64_set_kr(IA64_KR_TSSD, tssd); + current->thread.map_base = old_map_base; + current->thread.task_size = old_task_size; + } + + return error; +} + +int cp_compat_stat(struct kstat *stat, struct compat_stat __user *ubuf) +{ + int err; + + if ((u64) stat->size > MAX_NON_LFS || + !old_valid_dev(stat->dev) || + !old_valid_dev(stat->rdev)) + return -EOVERFLOW; + + if (clear_user(ubuf, sizeof(*ubuf))) + return -EFAULT; + + err = __put_user(old_encode_dev(stat->dev), &ubuf->st_dev); + err |= __put_user(stat->ino, &ubuf->st_ino); + err |= __put_user(stat->mode, &ubuf->st_mode); + err |= __put_user(stat->nlink, &ubuf->st_nlink); + err |= __put_user(high2lowuid(stat->uid), &ubuf->st_uid); + err |= __put_user(high2lowgid(stat->gid), &ubuf->st_gid); + err |= __put_user(old_encode_dev(stat->rdev), &ubuf->st_rdev); + err |= __put_user(stat->size, &ubuf->st_size); + err |= __put_user(stat->atime.tv_sec, &ubuf->st_atime); + err |= __put_user(stat->atime.tv_nsec, &ubuf->st_atime_nsec); + err |= __put_user(stat->mtime.tv_sec, &ubuf->st_mtime); + err |= __put_user(stat->mtime.tv_nsec, &ubuf->st_mtime_nsec); + err |= __put_user(stat->ctime.tv_sec, &ubuf->st_ctime); + err |= __put_user(stat->ctime.tv_nsec, &ubuf->st_ctime_nsec); + err |= __put_user(stat->blksize, &ubuf->st_blksize); + err |= __put_user(stat->blocks, &ubuf->st_blocks); + return err; +} + +#if PAGE_SHIFT > IA32_PAGE_SHIFT + + +static int +get_page_prot (struct vm_area_struct *vma, unsigned long addr) +{ + int prot = 0; + + if (!vma || vma->vm_start > addr) + return 0; + + if (vma->vm_flags & VM_READ) + prot |= PROT_READ; + if (vma->vm_flags & VM_WRITE) + prot |= PROT_WRITE; + if (vma->vm_flags & VM_EXEC) + prot |= PROT_EXEC; + return prot; +} + +/* + * Map a subpage by creating an anonymous page that contains the union of the old page and + * the subpage. + */ +static unsigned long +mmap_subpage (struct file *file, unsigned long start, unsigned long end, int prot, int flags, + loff_t off) +{ + void *page = NULL; + struct inode *inode; + unsigned long ret = 0; + struct vm_area_struct *vma = find_vma(current->mm, start); + int old_prot = get_page_prot(vma, start); + + DBG("mmap_subpage(file=%p,start=0x%lx,end=0x%lx,prot=%x,flags=%x,off=0x%llx)\n", + file, start, end, prot, flags, off); + + + /* Optimize the case where the old mmap and the new mmap are both anonymous */ + if ((old_prot & PROT_WRITE) && (flags & MAP_ANONYMOUS) && !vma->vm_file) { + if (clear_user((void __user *) start, end - start)) { + ret = -EFAULT; + goto out; + } + goto skip_mmap; + } + + page = (void *) get_zeroed_page(GFP_KERNEL); + if (!page) + return -ENOMEM; + + if (old_prot) + copy_from_user(page, (void __user *) PAGE_START(start), PAGE_SIZE); + + down_write(¤t->mm->mmap_sem); + { + ret = do_mmap(NULL, PAGE_START(start), PAGE_SIZE, prot | PROT_WRITE, + flags | MAP_FIXED | MAP_ANONYMOUS, 0); + } + up_write(¤t->mm->mmap_sem); + + if (IS_ERR((void *) ret)) + goto out; + + if (old_prot) { + /* copy back the old page contents. */ + if (offset_in_page(start)) + copy_to_user((void __user *) PAGE_START(start), page, + offset_in_page(start)); + if (offset_in_page(end)) + copy_to_user((void __user *) end, page + offset_in_page(end), + PAGE_SIZE - offset_in_page(end)); + } + + if (!(flags & MAP_ANONYMOUS)) { + /* read the file contents */ + inode = file->f_dentry->d_inode; + if (!inode->i_fop || !file->f_op->read + || ((*file->f_op->read)(file, (char __user *) start, end - start, &off) < 0)) + { + ret = -EINVAL; + goto out; + } + } + + skip_mmap: + if (!(prot & PROT_WRITE)) + ret = sys_mprotect(PAGE_START(start), PAGE_SIZE, prot | old_prot); + out: + if (page) + free_page((unsigned long) page); + return ret; +} + +/* SLAB cache for partial_page structures */ +kmem_cache_t *partial_page_cachep; + +/* + * init partial_page_list. + * return 0 means kmalloc fail. + */ +struct partial_page_list* +ia32_init_pp_list(void) +{ + struct partial_page_list *p; + + if ((p = kmalloc(sizeof(*p), GFP_KERNEL)) == NULL) + return p; + p->pp_head = NULL; + p->ppl_rb = RB_ROOT; + p->pp_hint = NULL; + atomic_set(&p->pp_count, 1); + return p; +} + +/* + * Search for the partial page with @start in partial page list @ppl. + * If finds the partial page, return the found partial page. + * Else, return 0 and provide @pprev, @rb_link, @rb_parent to + * be used by later __ia32_insert_pp(). + */ +static struct partial_page * +__ia32_find_pp(struct partial_page_list *ppl, unsigned int start, + struct partial_page **pprev, struct rb_node ***rb_link, + struct rb_node **rb_parent) +{ + struct partial_page *pp; + struct rb_node **__rb_link, *__rb_parent, *rb_prev; + + pp = ppl->pp_hint; + if (pp && pp->base == start) + return pp; + + __rb_link = &ppl->ppl_rb.rb_node; + rb_prev = __rb_parent = NULL; + + while (*__rb_link) { + __rb_parent = *__rb_link; + pp = rb_entry(__rb_parent, struct partial_page, pp_rb); + + if (pp->base == start) { + ppl->pp_hint = pp; + return pp; + } else if (pp->base < start) { + rb_prev = __rb_parent; + __rb_link = &__rb_parent->rb_right; + } else { + __rb_link = &__rb_parent->rb_left; + } + } + + *rb_link = __rb_link; + *rb_parent = __rb_parent; + *pprev = NULL; + if (rb_prev) + *pprev = rb_entry(rb_prev, struct partial_page, pp_rb); + return NULL; +} + +/* + * insert @pp into @ppl. + */ +static void +__ia32_insert_pp(struct partial_page_list *ppl, struct partial_page *pp, + struct partial_page *prev, struct rb_node **rb_link, + struct rb_node *rb_parent) +{ + /* link list */ + if (prev) { + pp->next = prev->next; + prev->next = pp; + } else { + ppl->pp_head = pp; + if (rb_parent) + pp->next = rb_entry(rb_parent, + struct partial_page, pp_rb); + else + pp->next = NULL; + } + + /* link rb */ + rb_link_node(&pp->pp_rb, rb_parent, rb_link); + rb_insert_color(&pp->pp_rb, &ppl->ppl_rb); + + ppl->pp_hint = pp; +} + +/* + * delete @pp from partial page list @ppl. + */ +static void +__ia32_delete_pp(struct partial_page_list *ppl, struct partial_page *pp, + struct partial_page *prev) +{ + if (prev) { + prev->next = pp->next; + if (ppl->pp_hint == pp) + ppl->pp_hint = prev; + } else { + ppl->pp_head = pp->next; + if (ppl->pp_hint == pp) + ppl->pp_hint = pp->next; + } + rb_erase(&pp->pp_rb, &ppl->ppl_rb); + kmem_cache_free(partial_page_cachep, pp); +} + +static struct partial_page * +__pp_prev(struct partial_page *pp) +{ + struct rb_node *prev = rb_prev(&pp->pp_rb); + if (prev) + return rb_entry(prev, struct partial_page, pp_rb); + else + return NULL; +} + +/* + * Delete partial pages with address between @start and @end. + * @start and @end are page aligned. + */ +static void +__ia32_delete_pp_range(unsigned int start, unsigned int end) +{ + struct partial_page *pp, *prev; + struct rb_node **rb_link, *rb_parent; + + if (start >= end) + return; + + pp = __ia32_find_pp(current->thread.ppl, start, &prev, + &rb_link, &rb_parent); + if (pp) + prev = __pp_prev(pp); + else { + if (prev) + pp = prev->next; + else + pp = current->thread.ppl->pp_head; + } + + while (pp && pp->base < end) { + struct partial_page *tmp = pp->next; + __ia32_delete_pp(current->thread.ppl, pp, prev); + pp = tmp; + } +} + +/* + * Set the range between @start and @end in bitmap. + * @start and @end should be IA32 page aligned and in the same IA64 page. + */ +static int +__ia32_set_pp(unsigned int start, unsigned int end, int flags) +{ + struct partial_page *pp, *prev; + struct rb_node ** rb_link, *rb_parent; + unsigned int pstart, start_bit, end_bit, i; + + pstart = PAGE_START(start); + start_bit = (start % PAGE_SIZE) / IA32_PAGE_SIZE; + end_bit = (end % PAGE_SIZE) / IA32_PAGE_SIZE; + if (end_bit == 0) + end_bit = PAGE_SIZE / IA32_PAGE_SIZE; + pp = __ia32_find_pp(current->thread.ppl, pstart, &prev, + &rb_link, &rb_parent); + if (pp) { + for (i = start_bit; i < end_bit; i++) + set_bit(i, &pp->bitmap); + /* + * Check: if this partial page has been set to a full page, + * then delete it. + */ + if (find_first_zero_bit(&pp->bitmap, sizeof(pp->bitmap)*8) >= + PAGE_SIZE/IA32_PAGE_SIZE) { + __ia32_delete_pp(current->thread.ppl, pp, __pp_prev(pp)); + } + return 0; + } + + /* + * MAP_FIXED may lead to overlapping mmap. + * In this case, the requested mmap area may already mmaped as a full + * page. So check vma before adding a new partial page. + */ + if (flags & MAP_FIXED) { + struct vm_area_struct *vma = find_vma(current->mm, pstart); + if (vma && vma->vm_start <= pstart) + return 0; + } + + /* new a partial_page */ + pp = kmem_cache_alloc(partial_page_cachep, GFP_KERNEL); + if (!pp) + return -ENOMEM; + pp->base = pstart; + pp->bitmap = 0; + for (i=start_bit; ibitmap)); + pp->next = NULL; + __ia32_insert_pp(current->thread.ppl, pp, prev, rb_link, rb_parent); + return 0; +} + +/* + * @start and @end should be IA32 page aligned, but don't need to be in the + * same IA64 page. Split @start and @end to make sure they're in the same IA64 + * page, then call __ia32_set_pp(). + */ +static void +ia32_set_pp(unsigned int start, unsigned int end, int flags) +{ + down_write(¤t->mm->mmap_sem); + if (flags & MAP_FIXED) { + /* + * MAP_FIXED may lead to overlapping mmap. When this happens, + * a series of complete IA64 pages results in deletion of + * old partial pages in that range. + */ + __ia32_delete_pp_range(PAGE_ALIGN(start), PAGE_START(end)); + } + + if (end < PAGE_ALIGN(start)) { + __ia32_set_pp(start, end, flags); + } else { + if (offset_in_page(start)) + __ia32_set_pp(start, PAGE_ALIGN(start), flags); + if (offset_in_page(end)) + __ia32_set_pp(PAGE_START(end), end, flags); + } + up_write(¤t->mm->mmap_sem); +} + +/* + * Unset the range between @start and @end in bitmap. + * @start and @end should be IA32 page aligned and in the same IA64 page. + * After doing that, if the bitmap is 0, then free the page and return 1, + * else return 0; + * If not find the partial page in the list, then + * If the vma exists, then the full page is set to a partial page; + * Else return -ENOMEM. + */ +static int +__ia32_unset_pp(unsigned int start, unsigned int end) +{ + struct partial_page *pp, *prev; + struct rb_node ** rb_link, *rb_parent; + unsigned int pstart, start_bit, end_bit, i; + struct vm_area_struct *vma; + + pstart = PAGE_START(start); + start_bit = (start % PAGE_SIZE) / IA32_PAGE_SIZE; + end_bit = (end % PAGE_SIZE) / IA32_PAGE_SIZE; + if (end_bit == 0) + end_bit = PAGE_SIZE / IA32_PAGE_SIZE; + + pp = __ia32_find_pp(current->thread.ppl, pstart, &prev, + &rb_link, &rb_parent); + if (pp) { + for (i = start_bit; i < end_bit; i++) + clear_bit(i, &pp->bitmap); + if (pp->bitmap == 0) { + __ia32_delete_pp(current->thread.ppl, pp, __pp_prev(pp)); + return 1; + } + return 0; + } + + vma = find_vma(current->mm, pstart); + if (!vma || vma->vm_start > pstart) { + return -ENOMEM; + } + + /* new a partial_page */ + pp = kmem_cache_alloc(partial_page_cachep, GFP_KERNEL); + if (!pp) + return -ENOMEM; + pp->base = pstart; + pp->bitmap = 0; + for (i = 0; i < start_bit; i++) + set_bit(i, &(pp->bitmap)); + for (i = end_bit; i < PAGE_SIZE / IA32_PAGE_SIZE; i++) + set_bit(i, &(pp->bitmap)); + pp->next = NULL; + __ia32_insert_pp(current->thread.ppl, pp, prev, rb_link, rb_parent); + return 0; +} + +/* + * Delete pp between PAGE_ALIGN(start) and PAGE_START(end) by calling + * __ia32_delete_pp_range(). Unset possible partial pages by calling + * __ia32_unset_pp(). + * The returned value see __ia32_unset_pp(). + */ +static int +ia32_unset_pp(unsigned int *startp, unsigned int *endp) +{ + unsigned int start = *startp, end = *endp; + int ret = 0; + + down_write(¤t->mm->mmap_sem); + + __ia32_delete_pp_range(PAGE_ALIGN(start), PAGE_START(end)); + + if (end < PAGE_ALIGN(start)) { + ret = __ia32_unset_pp(start, end); + if (ret == 1) { + *startp = PAGE_START(start); + *endp = PAGE_ALIGN(end); + } + if (ret == 0) { + /* to shortcut sys_munmap() in sys32_munmap() */ + *startp = PAGE_START(start); + *endp = PAGE_START(end); + } + } else { + if (offset_in_page(start)) { + ret = __ia32_unset_pp(start, PAGE_ALIGN(start)); + if (ret == 1) + *startp = PAGE_START(start); + if (ret == 0) + *startp = PAGE_ALIGN(start); + if (ret < 0) + goto out; + } + if (offset_in_page(end)) { + ret = __ia32_unset_pp(PAGE_START(end), end); + if (ret == 1) + *endp = PAGE_ALIGN(end); + if (ret == 0) + *endp = PAGE_START(end); + } + } + + out: + up_write(¤t->mm->mmap_sem); + return ret; +} + +/* + * Compare the range between @start and @end with bitmap in partial page. + * @start and @end should be IA32 page aligned and in the same IA64 page. + */ +static int +__ia32_compare_pp(unsigned int start, unsigned int end) +{ + struct partial_page *pp, *prev; + struct rb_node ** rb_link, *rb_parent; + unsigned int pstart, start_bit, end_bit, size; + unsigned int first_bit, next_zero_bit; /* the first range in bitmap */ + + pstart = PAGE_START(start); + + pp = __ia32_find_pp(current->thread.ppl, pstart, &prev, + &rb_link, &rb_parent); + if (!pp) + return 1; + + start_bit = (start % PAGE_SIZE) / IA32_PAGE_SIZE; + end_bit = (end % PAGE_SIZE) / IA32_PAGE_SIZE; + size = sizeof(pp->bitmap) * 8; + first_bit = find_first_bit(&pp->bitmap, size); + next_zero_bit = find_next_zero_bit(&pp->bitmap, size, first_bit); + if ((start_bit < first_bit) || (end_bit > next_zero_bit)) { + /* exceeds the first range in bitmap */ + return -ENOMEM; + } else if ((start_bit == first_bit) && (end_bit == next_zero_bit)) { + first_bit = find_next_bit(&pp->bitmap, size, next_zero_bit); + if ((next_zero_bit < first_bit) && (first_bit < size)) + return 1; /* has next range */ + else + return 0; /* no next range */ + } else + return 1; +} + +/* + * @start and @end should be IA32 page aligned, but don't need to be in the + * same IA64 page. Split @start and @end to make sure they're in the same IA64 + * page, then call __ia32_compare_pp(). + * + * Take this as example: the range is the 1st and 2nd 4K page. + * Return 0 if they fit bitmap exactly, i.e. bitmap = 00000011; + * Return 1 if the range doesn't cover whole bitmap, e.g. bitmap = 00001111; + * Return -ENOMEM if the range exceeds the bitmap, e.g. bitmap = 00000001 or + * bitmap = 00000101. + */ +static int +ia32_compare_pp(unsigned int *startp, unsigned int *endp) +{ + unsigned int start = *startp, end = *endp; + int retval = 0; + + down_write(¤t->mm->mmap_sem); + + if (end < PAGE_ALIGN(start)) { + retval = __ia32_compare_pp(start, end); + if (retval == 0) { + *startp = PAGE_START(start); + *endp = PAGE_ALIGN(end); + } + } else { + if (offset_in_page(start)) { + retval = __ia32_compare_pp(start, + PAGE_ALIGN(start)); + if (retval == 0) + *startp = PAGE_START(start); + if (retval < 0) + goto out; + } + if (offset_in_page(end)) { + retval = __ia32_compare_pp(PAGE_START(end), end); + if (retval == 0) + *endp = PAGE_ALIGN(end); + } + } + + out: + up_write(¤t->mm->mmap_sem); + return retval; +} + +static void +__ia32_drop_pp_list(struct partial_page_list *ppl) +{ + struct partial_page *pp = ppl->pp_head; + + while (pp) { + struct partial_page *next = pp->next; + kmem_cache_free(partial_page_cachep, pp); + pp = next; + } + + kfree(ppl); +} + +void +ia32_drop_partial_page_list(struct task_struct *task) +{ + struct partial_page_list* ppl = task->thread.ppl; + + if (ppl && atomic_dec_and_test(&ppl->pp_count)) + __ia32_drop_pp_list(ppl); +} + +/* + * Copy current->thread.ppl to ppl (already initialized). + */ +static int +__ia32_copy_pp_list(struct partial_page_list *ppl) +{ + struct partial_page *pp, *tmp, *prev; + struct rb_node **rb_link, *rb_parent; + + ppl->pp_head = NULL; + ppl->pp_hint = NULL; + ppl->ppl_rb = RB_ROOT; + rb_link = &ppl->ppl_rb.rb_node; + rb_parent = NULL; + prev = NULL; + + for (pp = current->thread.ppl->pp_head; pp; pp = pp->next) { + tmp = kmem_cache_alloc(partial_page_cachep, GFP_KERNEL); + if (!tmp) + return -ENOMEM; + *tmp = *pp; + __ia32_insert_pp(ppl, tmp, prev, rb_link, rb_parent); + prev = tmp; + rb_link = &tmp->pp_rb.rb_right; + rb_parent = &tmp->pp_rb; + } + return 0; +} + +int +ia32_copy_partial_page_list(struct task_struct *p, unsigned long clone_flags) +{ + int retval = 0; + + if (clone_flags & CLONE_VM) { + atomic_inc(¤t->thread.ppl->pp_count); + p->thread.ppl = current->thread.ppl; + } else { + p->thread.ppl = ia32_init_pp_list(); + if (!p->thread.ppl) + return -ENOMEM; + down_write(¤t->mm->mmap_sem); + { + retval = __ia32_copy_pp_list(p->thread.ppl); + } + up_write(¤t->mm->mmap_sem); + } + + return retval; +} + +static unsigned long +emulate_mmap (struct file *file, unsigned long start, unsigned long len, int prot, int flags, + loff_t off) +{ + unsigned long tmp, end, pend, pstart, ret, is_congruent, fudge = 0; + struct inode *inode; + loff_t poff; + + end = start + len; + pstart = PAGE_START(start); + pend = PAGE_ALIGN(end); + + if (flags & MAP_FIXED) { + ia32_set_pp((unsigned int)start, (unsigned int)end, flags); + if (start > pstart) { + if (flags & MAP_SHARED) + printk(KERN_INFO + "%s(%d): emulate_mmap() can't share head (addr=0x%lx)\n", + current->comm, current->pid, start); + ret = mmap_subpage(file, start, min(PAGE_ALIGN(start), end), prot, flags, + off); + if (IS_ERR((void *) ret)) + return ret; + pstart += PAGE_SIZE; + if (pstart >= pend) + goto out; /* done */ + } + if (end < pend) { + if (flags & MAP_SHARED) + printk(KERN_INFO + "%s(%d): emulate_mmap() can't share tail (end=0x%lx)\n", + current->comm, current->pid, end); + ret = mmap_subpage(file, max(start, PAGE_START(end)), end, prot, flags, + (off + len) - offset_in_page(end)); + if (IS_ERR((void *) ret)) + return ret; + pend -= PAGE_SIZE; + if (pstart >= pend) + goto out; /* done */ + } + } else { + /* + * If a start address was specified, use it if the entire rounded out area + * is available. + */ + if (start && !pstart) + fudge = 1; /* handle case of mapping to range (0,PAGE_SIZE) */ + tmp = arch_get_unmapped_area(file, pstart - fudge, pend - pstart, 0, flags); + if (tmp != pstart) { + pstart = tmp; + start = pstart + offset_in_page(off); /* make start congruent with off */ + end = start + len; + pend = PAGE_ALIGN(end); + } + } + + poff = off + (pstart - start); /* note: (pstart - start) may be negative */ + is_congruent = (flags & MAP_ANONYMOUS) || (offset_in_page(poff) == 0); + + if ((flags & MAP_SHARED) && !is_congruent) + printk(KERN_INFO "%s(%d): emulate_mmap() can't share contents of incongruent mmap " + "(addr=0x%lx,off=0x%llx)\n", current->comm, current->pid, start, off); + + DBG("mmap_body: mapping [0x%lx-0x%lx) %s with poff 0x%llx\n", pstart, pend, + is_congruent ? "congruent" : "not congruent", poff); + + down_write(¤t->mm->mmap_sem); + { + if (!(flags & MAP_ANONYMOUS) && is_congruent) + ret = do_mmap(file, pstart, pend - pstart, prot, flags | MAP_FIXED, poff); + else + ret = do_mmap(NULL, pstart, pend - pstart, + prot | ((flags & MAP_ANONYMOUS) ? 0 : PROT_WRITE), + flags | MAP_FIXED | MAP_ANONYMOUS, 0); + } + up_write(¤t->mm->mmap_sem); + + if (IS_ERR((void *) ret)) + return ret; + + if (!is_congruent) { + /* read the file contents */ + inode = file->f_dentry->d_inode; + if (!inode->i_fop || !file->f_op->read + || ((*file->f_op->read)(file, (char __user *) pstart, pend - pstart, &poff) + < 0)) + { + sys_munmap(pstart, pend - pstart); + return -EINVAL; + } + if (!(prot & PROT_WRITE) && sys_mprotect(pstart, pend - pstart, prot) < 0) + return -EINVAL; + } + + if (!(flags & MAP_FIXED)) + ia32_set_pp((unsigned int)start, (unsigned int)end, flags); +out: + return start; +} + +#endif /* PAGE_SHIFT > IA32_PAGE_SHIFT */ + +static inline unsigned int +get_prot32 (unsigned int prot) +{ + if (prot & PROT_WRITE) + /* on x86, PROT_WRITE implies PROT_READ which implies PROT_EEC */ + prot |= PROT_READ | PROT_WRITE | PROT_EXEC; + else if (prot & (PROT_READ | PROT_EXEC)) + /* on x86, there is no distinction between PROT_READ and PROT_EXEC */ + prot |= (PROT_READ | PROT_EXEC); + + return prot; +} + +unsigned long +ia32_do_mmap (struct file *file, unsigned long addr, unsigned long len, int prot, int flags, + loff_t offset) +{ + DBG("ia32_do_mmap(file=%p,addr=0x%lx,len=0x%lx,prot=%x,flags=%x,offset=0x%llx)\n", + file, addr, len, prot, flags, offset); + + if (file && (!file->f_op || !file->f_op->mmap)) + return -ENODEV; + + len = IA32_PAGE_ALIGN(len); + if (len == 0) + return addr; + + if (len > IA32_PAGE_OFFSET || addr > IA32_PAGE_OFFSET - len) + { + if (flags & MAP_FIXED) + return -ENOMEM; + else + return -EINVAL; + } + + if (OFFSET4K(offset)) + return -EINVAL; + + prot = get_prot32(prot); + +#if PAGE_SHIFT > IA32_PAGE_SHIFT + down(&ia32_mmap_sem); + { + addr = emulate_mmap(file, addr, len, prot, flags, offset); + } + up(&ia32_mmap_sem); +#else + down_write(¤t->mm->mmap_sem); + { + addr = do_mmap(file, addr, len, prot, flags, offset); + } + up_write(¤t->mm->mmap_sem); +#endif + DBG("ia32_do_mmap: returning 0x%lx\n", addr); + return addr; +} + +/* + * Linux/i386 didn't use to be able to handle more than 4 system call parameters, so these + * system calls used a memory block for parameter passing.. + */ + +struct mmap_arg_struct { + unsigned int addr; + unsigned int len; + unsigned int prot; + unsigned int flags; + unsigned int fd; + unsigned int offset; +}; + +asmlinkage long +sys32_mmap (struct mmap_arg_struct __user *arg) +{ + struct mmap_arg_struct a; + struct file *file = NULL; + unsigned long addr; + int flags; + + if (copy_from_user(&a, arg, sizeof(a))) + return -EFAULT; + + if (OFFSET4K(a.offset)) + return -EINVAL; + + flags = a.flags; + + flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE); + if (!(flags & MAP_ANONYMOUS)) { + file = fget(a.fd); + if (!file) + return -EBADF; + } + + addr = ia32_do_mmap(file, a.addr, a.len, a.prot, flags, a.offset); + + if (file) + fput(file); + return addr; +} + +asmlinkage long +sys32_mmap2 (unsigned int addr, unsigned int len, unsigned int prot, unsigned int flags, + unsigned int fd, unsigned int pgoff) +{ + struct file *file = NULL; + unsigned long retval; + + flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE); + if (!(flags & MAP_ANONYMOUS)) { + file = fget(fd); + if (!file) + return -EBADF; + } + + retval = ia32_do_mmap(file, addr, len, prot, flags, + (unsigned long) pgoff << IA32_PAGE_SHIFT); + + if (file) + fput(file); + return retval; +} + +asmlinkage long +sys32_munmap (unsigned int start, unsigned int len) +{ + unsigned int end = start + len; + long ret; + +#if PAGE_SHIFT <= IA32_PAGE_SHIFT + ret = sys_munmap(start, end - start); +#else + if (OFFSET4K(start)) + return -EINVAL; + + end = IA32_PAGE_ALIGN(end); + if (start >= end) + return -EINVAL; + + ret = ia32_unset_pp(&start, &end); + if (ret < 0) + return ret; + + if (start >= end) + return 0; + + down(&ia32_mmap_sem); + { + ret = sys_munmap(start, end - start); + } + up(&ia32_mmap_sem); +#endif + return ret; +} + +#if PAGE_SHIFT > IA32_PAGE_SHIFT + +/* + * When mprotect()ing a partial page, we set the permission to the union of the old + * settings and the new settings. In other words, it's only possible to make access to a + * partial page less restrictive. + */ +static long +mprotect_subpage (unsigned long address, int new_prot) +{ + int old_prot; + struct vm_area_struct *vma; + + if (new_prot == PROT_NONE) + return 0; /* optimize case where nothing changes... */ + vma = find_vma(current->mm, address); + old_prot = get_page_prot(vma, address); + return sys_mprotect(address, PAGE_SIZE, new_prot | old_prot); +} + +#endif /* PAGE_SHIFT > IA32_PAGE_SHIFT */ + +asmlinkage long +sys32_mprotect (unsigned int start, unsigned int len, int prot) +{ + unsigned int end = start + len; +#if PAGE_SHIFT > IA32_PAGE_SHIFT + long retval = 0; +#endif + + prot = get_prot32(prot); + +#if PAGE_SHIFT <= IA32_PAGE_SHIFT + return sys_mprotect(start, end - start, prot); +#else + if (OFFSET4K(start)) + return -EINVAL; + + end = IA32_PAGE_ALIGN(end); + if (end < start) + return -EINVAL; + + retval = ia32_compare_pp(&start, &end); + + if (retval < 0) + return retval; + + down(&ia32_mmap_sem); + { + if (offset_in_page(start)) { + /* start address is 4KB aligned but not page aligned. */ + retval = mprotect_subpage(PAGE_START(start), prot); + if (retval < 0) + goto out; + + start = PAGE_ALIGN(start); + if (start >= end) + goto out; /* retval is already zero... */ + } + + if (offset_in_page(end)) { + /* end address is 4KB aligned but not page aligned. */ + retval = mprotect_subpage(PAGE_START(end), prot); + if (retval < 0) + goto out; + + end = PAGE_START(end); + } + retval = sys_mprotect(start, end - start, prot); + } + out: + up(&ia32_mmap_sem); + return retval; +#endif +} + +asmlinkage long +sys32_mremap (unsigned int addr, unsigned int old_len, unsigned int new_len, + unsigned int flags, unsigned int new_addr) +{ + long ret; + +#if PAGE_SHIFT <= IA32_PAGE_SHIFT + ret = sys_mremap(addr, old_len, new_len, flags, new_addr); +#else + unsigned int old_end, new_end; + + if (OFFSET4K(addr)) + return -EINVAL; + + old_len = IA32_PAGE_ALIGN(old_len); + new_len = IA32_PAGE_ALIGN(new_len); + old_end = addr + old_len; + new_end = addr + new_len; + + if (!new_len) + return -EINVAL; + + if ((flags & MREMAP_FIXED) && (OFFSET4K(new_addr))) + return -EINVAL; + + if (old_len >= new_len) { + ret = sys32_munmap(addr + new_len, old_len - new_len); + if (ret && old_len != new_len) + return ret; + ret = addr; + if (!(flags & MREMAP_FIXED) || (new_addr == addr)) + return ret; + old_len = new_len; + } + + addr = PAGE_START(addr); + old_len = PAGE_ALIGN(old_end) - addr; + new_len = PAGE_ALIGN(new_end) - addr; + + down(&ia32_mmap_sem); + { + ret = sys_mremap(addr, old_len, new_len, flags, new_addr); + } + up(&ia32_mmap_sem); + + if ((ret >= 0) && (old_len < new_len)) { + /* mremap expanded successfully */ + ia32_set_pp(old_end, new_end, flags); + } +#endif + return ret; +} + +asmlinkage long +sys32_pipe (int __user *fd) +{ + int retval; + int fds[2]; + + retval = do_pipe(fds); + if (retval) + goto out; + if (copy_to_user(fd, fds, sizeof(fds))) + retval = -EFAULT; + out: + return retval; +} + +static inline long +get_tv32 (struct timeval *o, struct compat_timeval __user *i) +{ + return (!access_ok(VERIFY_READ, i, sizeof(*i)) || + (__get_user(o->tv_sec, &i->tv_sec) | __get_user(o->tv_usec, &i->tv_usec))); +} + +static inline long +put_tv32 (struct compat_timeval __user *o, struct timeval *i) +{ + return (!access_ok(VERIFY_WRITE, o, sizeof(*o)) || + (__put_user(i->tv_sec, &o->tv_sec) | __put_user(i->tv_usec, &o->tv_usec))); +} + +asmlinkage unsigned long +sys32_alarm (unsigned int seconds) +{ + struct itimerval it_new, it_old; + unsigned int oldalarm; + + it_new.it_interval.tv_sec = it_new.it_interval.tv_usec = 0; + it_new.it_value.tv_sec = seconds; + it_new.it_value.tv_usec = 0; + do_setitimer(ITIMER_REAL, &it_new, &it_old); + oldalarm = it_old.it_value.tv_sec; + /* ehhh.. We can't return 0 if we have an alarm pending.. */ + /* And we'd better return too much than too little anyway */ + if (it_old.it_value.tv_usec) + oldalarm++; + return oldalarm; +} + +/* Translations due to time_t size differences. Which affects all + sorts of things, like timeval and itimerval. */ + +extern struct timezone sys_tz; + +asmlinkage long +sys32_gettimeofday (struct compat_timeval __user *tv, struct timezone __user *tz) +{ + if (tv) { + struct timeval ktv; + do_gettimeofday(&ktv); + if (put_tv32(tv, &ktv)) + return -EFAULT; + } + if (tz) { + if (copy_to_user(tz, &sys_tz, sizeof(sys_tz))) + return -EFAULT; + } + return 0; +} + +asmlinkage long +sys32_settimeofday (struct compat_timeval __user *tv, struct timezone __user *tz) +{ + struct timeval ktv; + struct timespec kts; + struct timezone ktz; + + if (tv) { + if (get_tv32(&ktv, tv)) + return -EFAULT; + kts.tv_sec = ktv.tv_sec; + kts.tv_nsec = ktv.tv_usec * 1000; + } + if (tz) { + if (copy_from_user(&ktz, tz, sizeof(ktz))) + return -EFAULT; + } + + return do_sys_settimeofday(tv ? &kts : NULL, tz ? &ktz : NULL); +} + +struct getdents32_callback { + struct compat_dirent __user *current_dir; + struct compat_dirent __user *previous; + int count; + int error; +}; + +struct readdir32_callback { + struct old_linux32_dirent __user * dirent; + int count; +}; + +static int +filldir32 (void *__buf, const char *name, int namlen, loff_t offset, ino_t ino, + unsigned int d_type) +{ + struct compat_dirent __user * dirent; + struct getdents32_callback * buf = (struct getdents32_callback *) __buf; + int reclen = ROUND_UP(offsetof(struct compat_dirent, d_name) + namlen + 1, 4); + + buf->error = -EINVAL; /* only used if we fail.. */ + if (reclen > buf->count) + return -EINVAL; + buf->error = -EFAULT; /* only used if we fail.. */ + dirent = buf->previous; + if (dirent) + if (put_user(offset, &dirent->d_off)) + return -EFAULT; + dirent = buf->current_dir; + buf->previous = dirent; + if (put_user(ino, &dirent->d_ino) + || put_user(reclen, &dirent->d_reclen) + || copy_to_user(dirent->d_name, name, namlen) + || put_user(0, dirent->d_name + namlen)) + return -EFAULT; + dirent = (struct compat_dirent __user *) ((char __user *) dirent + reclen); + buf->current_dir = dirent; + buf->count -= reclen; + return 0; +} + +asmlinkage long +sys32_getdents (unsigned int fd, struct compat_dirent __user *dirent, unsigned int count) +{ + struct file * file; + struct compat_dirent __user * lastdirent; + struct getdents32_callback buf; + int error; + + error = -EBADF; + file = fget(fd); + if (!file) + goto out; + + buf.current_dir = dirent; + buf.previous = NULL; + buf.count = count; + buf.error = 0; + + error = vfs_readdir(file, filldir32, &buf); + if (error < 0) + goto out_putf; + error = buf.error; + lastdirent = buf.previous; + if (lastdirent) { + error = -EINVAL; + if (put_user(file->f_pos, &lastdirent->d_off)) + goto out_putf; + error = count - buf.count; + } + +out_putf: + fput(file); +out: + return error; +} + +static int +fillonedir32 (void * __buf, const char * name, int namlen, loff_t offset, ino_t ino, + unsigned int d_type) +{ + struct readdir32_callback * buf = (struct readdir32_callback *) __buf; + struct old_linux32_dirent __user * dirent; + + if (buf->count) + return -EINVAL; + buf->count++; + dirent = buf->dirent; + if (put_user(ino, &dirent->d_ino) + || put_user(offset, &dirent->d_offset) + || put_user(namlen, &dirent->d_namlen) + || copy_to_user(dirent->d_name, name, namlen) + || put_user(0, dirent->d_name + namlen)) + return -EFAULT; + return 0; +} + +asmlinkage long +sys32_readdir (unsigned int fd, void __user *dirent, unsigned int count) +{ + int error; + struct file * file; + struct readdir32_callback buf; + + error = -EBADF; + file = fget(fd); + if (!file) + goto out; + + buf.count = 0; + buf.dirent = dirent; + + error = vfs_readdir(file, fillonedir32, &buf); + if (error >= 0) + error = buf.count; + fput(file); +out: + return error; +} + +struct sel_arg_struct { + unsigned int n; + unsigned int inp; + unsigned int outp; + unsigned int exp; + unsigned int tvp; +}; + +asmlinkage long +sys32_old_select (struct sel_arg_struct __user *arg) +{ + struct sel_arg_struct a; + + if (copy_from_user(&a, arg, sizeof(a))) + return -EFAULT; + return compat_sys_select(a.n, compat_ptr(a.inp), compat_ptr(a.outp), + compat_ptr(a.exp), compat_ptr(a.tvp)); +} + +#define SEMOP 1 +#define SEMGET 2 +#define SEMCTL 3 +#define SEMTIMEDOP 4 +#define MSGSND 11 +#define MSGRCV 12 +#define MSGGET 13 +#define MSGCTL 14 +#define SHMAT 21 +#define SHMDT 22 +#define SHMGET 23 +#define SHMCTL 24 + +asmlinkage long +sys32_ipc(u32 call, int first, int second, int third, u32 ptr, u32 fifth) +{ + int version; + + version = call >> 16; /* hack for backward compatibility */ + call &= 0xffff; + + switch (call) { + case SEMTIMEDOP: + if (fifth) + return compat_sys_semtimedop(first, compat_ptr(ptr), + second, compat_ptr(fifth)); + /* else fall through for normal semop() */ + case SEMOP: + /* struct sembuf is the same on 32 and 64bit :)) */ + return sys_semtimedop(first, compat_ptr(ptr), second, + NULL); + case SEMGET: + return sys_semget(first, second, third); + case SEMCTL: + return compat_sys_semctl(first, second, third, compat_ptr(ptr)); + + case MSGSND: + return compat_sys_msgsnd(first, second, third, compat_ptr(ptr)); + case MSGRCV: + return compat_sys_msgrcv(first, second, fifth, third, version, compat_ptr(ptr)); + case MSGGET: + return sys_msgget((key_t) first, second); + case MSGCTL: + return compat_sys_msgctl(first, second, compat_ptr(ptr)); + + case SHMAT: + return compat_sys_shmat(first, second, third, version, compat_ptr(ptr)); + break; + case SHMDT: + return sys_shmdt(compat_ptr(ptr)); + case SHMGET: + return sys_shmget(first, (unsigned)second, third); + case SHMCTL: + return compat_sys_shmctl(first, second, compat_ptr(ptr)); + + default: + return -ENOSYS; + } + return -EINVAL; +} + +asmlinkage long +compat_sys_wait4 (compat_pid_t pid, compat_uint_t * stat_addr, int options, + struct compat_rusage *ru); + +asmlinkage long +sys32_waitpid (int pid, unsigned int *stat_addr, int options) +{ + return compat_sys_wait4(pid, stat_addr, options, NULL); +} + +static unsigned int +ia32_peek (struct task_struct *child, unsigned long addr, unsigned int *val) +{ + size_t copied; + unsigned int ret; + + copied = access_process_vm(child, addr, val, sizeof(*val), 0); + return (copied != sizeof(ret)) ? -EIO : 0; +} + +static unsigned int +ia32_poke (struct task_struct *child, unsigned long addr, unsigned int val) +{ + + if (access_process_vm(child, addr, &val, sizeof(val), 1) != sizeof(val)) + return -EIO; + return 0; +} + +/* + * The order in which registers are stored in the ptrace regs structure + */ +#define PT_EBX 0 +#define PT_ECX 1 +#define PT_EDX 2 +#define PT_ESI 3 +#define PT_EDI 4 +#define PT_EBP 5 +#define PT_EAX 6 +#define PT_DS 7 +#define PT_ES 8 +#define PT_FS 9 +#define PT_GS 10 +#define PT_ORIG_EAX 11 +#define PT_EIP 12 +#define PT_CS 13 +#define PT_EFL 14 +#define PT_UESP 15 +#define PT_SS 16 + +static unsigned int +getreg (struct task_struct *child, int regno) +{ + struct pt_regs *child_regs; + + child_regs = ia64_task_regs(child); + switch (regno / sizeof(int)) { + case PT_EBX: return child_regs->r11; + case PT_ECX: return child_regs->r9; + case PT_EDX: return child_regs->r10; + case PT_ESI: return child_regs->r14; + case PT_EDI: return child_regs->r15; + case PT_EBP: return child_regs->r13; + case PT_EAX: return child_regs->r8; + case PT_ORIG_EAX: return child_regs->r1; /* see dispatch_to_ia32_handler() */ + case PT_EIP: return child_regs->cr_iip; + case PT_UESP: return child_regs->r12; + case PT_EFL: return child->thread.eflag; + case PT_DS: case PT_ES: case PT_FS: case PT_GS: case PT_SS: + return __USER_DS; + case PT_CS: return __USER_CS; + default: + printk(KERN_ERR "ia32.getreg(): unknown register %d\n", regno); + break; + } + return 0; +} + +static void +putreg (struct task_struct *child, int regno, unsigned int value) +{ + struct pt_regs *child_regs; + + child_regs = ia64_task_regs(child); + switch (regno / sizeof(int)) { + case PT_EBX: child_regs->r11 = value; break; + case PT_ECX: child_regs->r9 = value; break; + case PT_EDX: child_regs->r10 = value; break; + case PT_ESI: child_regs->r14 = value; break; + case PT_EDI: child_regs->r15 = value; break; + case PT_EBP: child_regs->r13 = value; break; + case PT_EAX: child_regs->r8 = value; break; + case PT_ORIG_EAX: child_regs->r1 = value; break; + case PT_EIP: child_regs->cr_iip = value; break; + case PT_UESP: child_regs->r12 = value; break; + case PT_EFL: child->thread.eflag = value; break; + case PT_DS: case PT_ES: case PT_FS: case PT_GS: case PT_SS: + if (value != __USER_DS) + printk(KERN_ERR + "ia32.putreg: attempt to set invalid segment register %d = %x\n", + regno, value); + break; + case PT_CS: + if (value != __USER_CS) + printk(KERN_ERR + "ia32.putreg: attempt to to set invalid segment register %d = %x\n", + regno, value); + break; + default: + printk(KERN_ERR "ia32.putreg: unknown register %d\n", regno); + break; + } +} + +static void +put_fpreg (int regno, struct _fpreg_ia32 __user *reg, struct pt_regs *ptp, + struct switch_stack *swp, int tos) +{ + struct _fpreg_ia32 *f; + char buf[32]; + + f = (struct _fpreg_ia32 *)(((unsigned long)buf + 15) & ~15); + if ((regno += tos) >= 8) + regno -= 8; + switch (regno) { + case 0: + ia64f2ia32f(f, &ptp->f8); + break; + case 1: + ia64f2ia32f(f, &ptp->f9); + break; + case 2: + ia64f2ia32f(f, &ptp->f10); + break; + case 3: + ia64f2ia32f(f, &ptp->f11); + break; + case 4: + case 5: + case 6: + case 7: + ia64f2ia32f(f, &swp->f12 + (regno - 4)); + break; + } + copy_to_user(reg, f, sizeof(*reg)); +} + +static void +get_fpreg (int regno, struct _fpreg_ia32 __user *reg, struct pt_regs *ptp, + struct switch_stack *swp, int tos) +{ + + if ((regno += tos) >= 8) + regno -= 8; + switch (regno) { + case 0: + copy_from_user(&ptp->f8, reg, sizeof(*reg)); + break; + case 1: + copy_from_user(&ptp->f9, reg, sizeof(*reg)); + break; + case 2: + copy_from_user(&ptp->f10, reg, sizeof(*reg)); + break; + case 3: + copy_from_user(&ptp->f11, reg, sizeof(*reg)); + break; + case 4: + case 5: + case 6: + case 7: + copy_from_user(&swp->f12 + (regno - 4), reg, sizeof(*reg)); + break; + } + return; +} + +int +save_ia32_fpstate (struct task_struct *tsk, struct ia32_user_i387_struct __user *save) +{ + struct switch_stack *swp; + struct pt_regs *ptp; + int i, tos; + + if (!access_ok(VERIFY_WRITE, save, sizeof(*save))) + return -EFAULT; + + __put_user(tsk->thread.fcr & 0xffff, &save->cwd); + __put_user(tsk->thread.fsr & 0xffff, &save->swd); + __put_user((tsk->thread.fsr>>16) & 0xffff, &save->twd); + __put_user(tsk->thread.fir, &save->fip); + __put_user((tsk->thread.fir>>32) & 0xffff, &save->fcs); + __put_user(tsk->thread.fdr, &save->foo); + __put_user((tsk->thread.fdr>>32) & 0xffff, &save->fos); + + /* + * Stack frames start with 16-bytes of temp space + */ + swp = (struct switch_stack *)(tsk->thread.ksp + 16); + ptp = ia64_task_regs(tsk); + tos = (tsk->thread.fsr >> 11) & 7; + for (i = 0; i < 8; i++) + put_fpreg(i, &save->st_space[i], ptp, swp, tos); + return 0; +} + +static int +restore_ia32_fpstate (struct task_struct *tsk, struct ia32_user_i387_struct __user *save) +{ + struct switch_stack *swp; + struct pt_regs *ptp; + int i, tos; + unsigned int fsrlo, fsrhi, num32; + + if (!access_ok(VERIFY_READ, save, sizeof(*save))) + return(-EFAULT); + + __get_user(num32, (unsigned int __user *)&save->cwd); + tsk->thread.fcr = (tsk->thread.fcr & (~0x1f3f)) | (num32 & 0x1f3f); + __get_user(fsrlo, (unsigned int __user *)&save->swd); + __get_user(fsrhi, (unsigned int __user *)&save->twd); + num32 = (fsrhi << 16) | fsrlo; + tsk->thread.fsr = (tsk->thread.fsr & (~0xffffffff)) | num32; + __get_user(num32, (unsigned int __user *)&save->fip); + tsk->thread.fir = (tsk->thread.fir & (~0xffffffff)) | num32; + __get_user(num32, (unsigned int __user *)&save->foo); + tsk->thread.fdr = (tsk->thread.fdr & (~0xffffffff)) | num32; + + /* + * Stack frames start with 16-bytes of temp space + */ + swp = (struct switch_stack *)(tsk->thread.ksp + 16); + ptp = ia64_task_regs(tsk); + tos = (tsk->thread.fsr >> 11) & 7; + for (i = 0; i < 8; i++) + get_fpreg(i, &save->st_space[i], ptp, swp, tos); + return 0; +} + +int +save_ia32_fpxstate (struct task_struct *tsk, struct ia32_user_fxsr_struct __user *save) +{ + struct switch_stack *swp; + struct pt_regs *ptp; + int i, tos; + unsigned long mxcsr=0; + unsigned long num128[2]; + + if (!access_ok(VERIFY_WRITE, save, sizeof(*save))) + return -EFAULT; + + __put_user(tsk->thread.fcr & 0xffff, &save->cwd); + __put_user(tsk->thread.fsr & 0xffff, &save->swd); + __put_user((tsk->thread.fsr>>16) & 0xffff, &save->twd); + __put_user(tsk->thread.fir, &save->fip); + __put_user((tsk->thread.fir>>32) & 0xffff, &save->fcs); + __put_user(tsk->thread.fdr, &save->foo); + __put_user((tsk->thread.fdr>>32) & 0xffff, &save->fos); + + /* + * Stack frames start with 16-bytes of temp space + */ + swp = (struct switch_stack *)(tsk->thread.ksp + 16); + ptp = ia64_task_regs(tsk); + tos = (tsk->thread.fsr >> 11) & 7; + for (i = 0; i < 8; i++) + put_fpreg(i, (struct _fpreg_ia32 __user *)&save->st_space[4*i], ptp, swp, tos); + + mxcsr = ((tsk->thread.fcr>>32) & 0xff80) | ((tsk->thread.fsr>>32) & 0x3f); + __put_user(mxcsr & 0xffff, &save->mxcsr); + for (i = 0; i < 8; i++) { + memcpy(&(num128[0]), &(swp->f16) + i*2, sizeof(unsigned long)); + memcpy(&(num128[1]), &(swp->f17) + i*2, sizeof(unsigned long)); + copy_to_user(&save->xmm_space[0] + 4*i, num128, sizeof(struct _xmmreg_ia32)); + } + return 0; +} + +static int +restore_ia32_fpxstate (struct task_struct *tsk, struct ia32_user_fxsr_struct __user *save) +{ + struct switch_stack *swp; + struct pt_regs *ptp; + int i, tos; + unsigned int fsrlo, fsrhi, num32; + int mxcsr; + unsigned long num64; + unsigned long num128[2]; + + if (!access_ok(VERIFY_READ, save, sizeof(*save))) + return(-EFAULT); + + __get_user(num32, (unsigned int __user *)&save->cwd); + tsk->thread.fcr = (tsk->thread.fcr & (~0x1f3f)) | (num32 & 0x1f3f); + __get_user(fsrlo, (unsigned int __user *)&save->swd); + __get_user(fsrhi, (unsigned int __user *)&save->twd); + num32 = (fsrhi << 16) | fsrlo; + tsk->thread.fsr = (tsk->thread.fsr & (~0xffffffff)) | num32; + __get_user(num32, (unsigned int __user *)&save->fip); + tsk->thread.fir = (tsk->thread.fir & (~0xffffffff)) | num32; + __get_user(num32, (unsigned int __user *)&save->foo); + tsk->thread.fdr = (tsk->thread.fdr & (~0xffffffff)) | num32; + + /* + * Stack frames start with 16-bytes of temp space + */ + swp = (struct switch_stack *)(tsk->thread.ksp + 16); + ptp = ia64_task_regs(tsk); + tos = (tsk->thread.fsr >> 11) & 7; + for (i = 0; i < 8; i++) + get_fpreg(i, (struct _fpreg_ia32 __user *)&save->st_space[4*i], ptp, swp, tos); + + __get_user(mxcsr, (unsigned int __user *)&save->mxcsr); + num64 = mxcsr & 0xff10; + tsk->thread.fcr = (tsk->thread.fcr & (~0xff1000000000UL)) | (num64<<32); + num64 = mxcsr & 0x3f; + tsk->thread.fsr = (tsk->thread.fsr & (~0x3f00000000UL)) | (num64<<32); + + for (i = 0; i < 8; i++) { + copy_from_user(num128, &save->xmm_space[0] + 4*i, sizeof(struct _xmmreg_ia32)); + memcpy(&(swp->f16) + i*2, &(num128[0]), sizeof(unsigned long)); + memcpy(&(swp->f17) + i*2, &(num128[1]), sizeof(unsigned long)); + } + return 0; +} + +asmlinkage long +sys32_ptrace (int request, pid_t pid, unsigned int addr, unsigned int data) +{ + struct task_struct *child; + unsigned int value, tmp; + long i, ret; + + lock_kernel(); + if (request == PTRACE_TRACEME) { + ret = sys_ptrace(request, pid, addr, data); + goto out; + } + + ret = -ESRCH; + read_lock(&tasklist_lock); + child = find_task_by_pid(pid); + if (child) + get_task_struct(child); + read_unlock(&tasklist_lock); + if (!child) + goto out; + ret = -EPERM; + if (pid == 1) /* no messing around with init! */ + goto out_tsk; + + if (request == PTRACE_ATTACH) { + ret = sys_ptrace(request, pid, addr, data); + goto out_tsk; + } + + ret = ptrace_check_attach(child, request == PTRACE_KILL); + if (ret < 0) + goto out_tsk; + + switch (request) { + case PTRACE_PEEKTEXT: + case PTRACE_PEEKDATA: /* read word at location addr */ + ret = ia32_peek(child, addr, &value); + if (ret == 0) + ret = put_user(value, (unsigned int __user *) compat_ptr(data)); + else + ret = -EIO; + goto out_tsk; + + case PTRACE_POKETEXT: + case PTRACE_POKEDATA: /* write the word at location addr */ + ret = ia32_poke(child, addr, data); + goto out_tsk; + + case PTRACE_PEEKUSR: /* read word at addr in USER area */ + ret = -EIO; + if ((addr & 3) || addr > 17*sizeof(int)) + break; + + tmp = getreg(child, addr); + if (!put_user(tmp, (unsigned int __user *) compat_ptr(data))) + ret = 0; + break; + + case PTRACE_POKEUSR: /* write word at addr in USER area */ + ret = -EIO; + if ((addr & 3) || addr > 17*sizeof(int)) + break; + + putreg(child, addr, data); + ret = 0; + break; + + case IA32_PTRACE_GETREGS: + if (!access_ok(VERIFY_WRITE, compat_ptr(data), 17*sizeof(int))) { + ret = -EIO; + break; + } + for (i = 0; i < (int) (17*sizeof(int)); i += sizeof(int) ) { + put_user(getreg(child, i), (unsigned int __user *) compat_ptr(data)); + data += sizeof(int); + } + ret = 0; + break; + + case IA32_PTRACE_SETREGS: + if (!access_ok(VERIFY_READ, compat_ptr(data), 17*sizeof(int))) { + ret = -EIO; + break; + } + for (i = 0; i < (int) (17*sizeof(int)); i += sizeof(int) ) { + get_user(tmp, (unsigned int __user *) compat_ptr(data)); + putreg(child, i, tmp); + data += sizeof(int); + } + ret = 0; + break; + + case IA32_PTRACE_GETFPREGS: + ret = save_ia32_fpstate(child, (struct ia32_user_i387_struct __user *) + compat_ptr(data)); + break; + + case IA32_PTRACE_GETFPXREGS: + ret = save_ia32_fpxstate(child, (struct ia32_user_fxsr_struct __user *) + compat_ptr(data)); + break; + + case IA32_PTRACE_SETFPREGS: + ret = restore_ia32_fpstate(child, (struct ia32_user_i387_struct __user *) + compat_ptr(data)); + break; + + case IA32_PTRACE_SETFPXREGS: + ret = restore_ia32_fpxstate(child, (struct ia32_user_fxsr_struct __user *) + compat_ptr(data)); + break; + + case PTRACE_GETEVENTMSG: + ret = put_user(child->ptrace_message, (unsigned int __user *) compat_ptr(data)); + break; + + case PTRACE_SYSCALL: /* continue, stop after next syscall */ + case PTRACE_CONT: /* restart after signal. */ + case PTRACE_KILL: + case PTRACE_SINGLESTEP: /* execute chile for one instruction */ + case PTRACE_DETACH: /* detach a process */ + ret = sys_ptrace(request, pid, addr, data); + break; + + default: + ret = ptrace_request(child, request, addr, data); + break; + + } + out_tsk: + put_task_struct(child); + out: + unlock_kernel(); + return ret; +} + +typedef struct { + unsigned int ss_sp; + unsigned int ss_flags; + unsigned int ss_size; +} ia32_stack_t; + +asmlinkage long +sys32_sigaltstack (ia32_stack_t __user *uss32, ia32_stack_t __user *uoss32, + long arg2, long arg3, long arg4, long arg5, long arg6, + long arg7, struct pt_regs pt) +{ + stack_t uss, uoss; + ia32_stack_t buf32; + int ret; + mm_segment_t old_fs = get_fs(); + + if (uss32) { + if (copy_from_user(&buf32, uss32, sizeof(ia32_stack_t))) + return -EFAULT; + uss.ss_sp = (void __user *) (long) buf32.ss_sp; + uss.ss_flags = buf32.ss_flags; + /* MINSIGSTKSZ is different for ia32 vs ia64. We lie here to pass the + check and set it to the user requested value later */ + if ((buf32.ss_flags != SS_DISABLE) && (buf32.ss_size < MINSIGSTKSZ_IA32)) { + ret = -ENOMEM; + goto out; + } + uss.ss_size = MINSIGSTKSZ; + } + set_fs(KERNEL_DS); + ret = do_sigaltstack(uss32 ? (stack_t __user *) &uss : NULL, + (stack_t __user *) &uoss, pt.r12); + current->sas_ss_size = buf32.ss_size; + set_fs(old_fs); +out: + if (ret < 0) + return(ret); + if (uoss32) { + buf32.ss_sp = (long __user) uoss.ss_sp; + buf32.ss_flags = uoss.ss_flags; + buf32.ss_size = uoss.ss_size; + if (copy_to_user(uoss32, &buf32, sizeof(ia32_stack_t))) + return -EFAULT; + } + return ret; +} + +asmlinkage int +sys32_pause (void) +{ + current->state = TASK_INTERRUPTIBLE; + schedule(); + return -ERESTARTNOHAND; +} + +asmlinkage int +sys32_msync (unsigned int start, unsigned int len, int flags) +{ + unsigned int addr; + + if (OFFSET4K(start)) + return -EINVAL; + addr = PAGE_START(start); + return sys_msync(addr, len + (start - addr), flags); +} + +struct sysctl32 { + unsigned int name; + int nlen; + unsigned int oldval; + unsigned int oldlenp; + unsigned int newval; + unsigned int newlen; + unsigned int __unused[4]; +}; + +#ifdef CONFIG_SYSCTL +asmlinkage long +sys32_sysctl (struct sysctl32 __user *args) +{ + struct sysctl32 a32; + mm_segment_t old_fs = get_fs (); + void __user *oldvalp, *newvalp; + size_t oldlen; + int __user *namep; + long ret; + + if (copy_from_user(&a32, args, sizeof(a32))) + return -EFAULT; + + /* + * We need to pre-validate these because we have to disable address checking + * before calling do_sysctl() because of OLDLEN but we can't run the risk of the + * user specifying bad addresses here. Well, since we're dealing with 32 bit + * addresses, we KNOW that access_ok() will always succeed, so this is an + * expensive NOP, but so what... + */ + namep = (int __user *) compat_ptr(a32.name); + oldvalp = compat_ptr(a32.oldval); + newvalp = compat_ptr(a32.newval); + + if ((oldvalp && get_user(oldlen, (int __user *) compat_ptr(a32.oldlenp))) + || !access_ok(VERIFY_WRITE, namep, 0) + || !access_ok(VERIFY_WRITE, oldvalp, 0) + || !access_ok(VERIFY_WRITE, newvalp, 0)) + return -EFAULT; + + set_fs(KERNEL_DS); + lock_kernel(); + ret = do_sysctl(namep, a32.nlen, oldvalp, (size_t __user *) &oldlen, + newvalp, (size_t) a32.newlen); + unlock_kernel(); + set_fs(old_fs); + + if (oldvalp && put_user (oldlen, (int __user *) compat_ptr(a32.oldlenp))) + return -EFAULT; + + return ret; +} +#endif + +asmlinkage long +sys32_newuname (struct new_utsname __user *name) +{ + int ret = sys_newuname(name); + + if (!ret) + if (copy_to_user(name->machine, "i686\0\0\0", 8)) + ret = -EFAULT; + return ret; +} + +asmlinkage long +sys32_getresuid16 (u16 __user *ruid, u16 __user *euid, u16 __user *suid) +{ + uid_t a, b, c; + int ret; + mm_segment_t old_fs = get_fs(); + + set_fs(KERNEL_DS); + ret = sys_getresuid((uid_t __user *) &a, (uid_t __user *) &b, (uid_t __user *) &c); + set_fs(old_fs); + + if (put_user(a, ruid) || put_user(b, euid) || put_user(c, suid)) + return -EFAULT; + return ret; +} + +asmlinkage long +sys32_getresgid16 (u16 __user *rgid, u16 __user *egid, u16 __user *sgid) +{ + gid_t a, b, c; + int ret; + mm_segment_t old_fs = get_fs(); + + set_fs(KERNEL_DS); + ret = sys_getresgid((gid_t __user *) &a, (gid_t __user *) &b, (gid_t __user *) &c); + set_fs(old_fs); + + if (ret) + return ret; + + return put_user(a, rgid) | put_user(b, egid) | put_user(c, sgid); +} + +asmlinkage long +sys32_lseek (unsigned int fd, int offset, unsigned int whence) +{ + /* Sign-extension of "offset" is important here... */ + return sys_lseek(fd, offset, whence); +} + +static int +groups16_to_user(short __user *grouplist, struct group_info *group_info) +{ + int i; + short group; + + for (i = 0; i < group_info->ngroups; i++) { + group = (short)GROUP_AT(group_info, i); + if (put_user(group, grouplist+i)) + return -EFAULT; + } + + return 0; +} + +static int +groups16_from_user(struct group_info *group_info, short __user *grouplist) +{ + int i; + short group; + + for (i = 0; i < group_info->ngroups; i++) { + if (get_user(group, grouplist+i)) + return -EFAULT; + GROUP_AT(group_info, i) = (gid_t)group; + } + + return 0; +} + +asmlinkage long +sys32_getgroups16 (int gidsetsize, short __user *grouplist) +{ + int i; + + if (gidsetsize < 0) + return -EINVAL; + + get_group_info(current->group_info); + i = current->group_info->ngroups; + if (gidsetsize) { + if (i > gidsetsize) { + i = -EINVAL; + goto out; + } + if (groups16_to_user(grouplist, current->group_info)) { + i = -EFAULT; + goto out; + } + } +out: + put_group_info(current->group_info); + return i; +} + +asmlinkage long +sys32_setgroups16 (int gidsetsize, short __user *grouplist) +{ + struct group_info *group_info; + int retval; + + if (!capable(CAP_SETGID)) + return -EPERM; + if ((unsigned)gidsetsize > NGROUPS_MAX) + return -EINVAL; + + group_info = groups_alloc(gidsetsize); + if (!group_info) + return -ENOMEM; + retval = groups16_from_user(group_info, grouplist); + if (retval) { + put_group_info(group_info); + return retval; + } + + retval = set_current_groups(group_info); + put_group_info(group_info); + + return retval; +} + +asmlinkage long +sys32_truncate64 (unsigned int path, unsigned int len_lo, unsigned int len_hi) +{ + return sys_truncate(compat_ptr(path), ((unsigned long) len_hi << 32) | len_lo); +} + +asmlinkage long +sys32_ftruncate64 (int fd, unsigned int len_lo, unsigned int len_hi) +{ + return sys_ftruncate(fd, ((unsigned long) len_hi << 32) | len_lo); +} + +static int +putstat64 (struct stat64 __user *ubuf, struct kstat *kbuf) +{ + int err; + u64 hdev; + + if (clear_user(ubuf, sizeof(*ubuf))) + return -EFAULT; + + hdev = huge_encode_dev(kbuf->dev); + err = __put_user(hdev, (u32 __user*)&ubuf->st_dev); + err |= __put_user(hdev >> 32, ((u32 __user*)&ubuf->st_dev) + 1); + err |= __put_user(kbuf->ino, &ubuf->__st_ino); + err |= __put_user(kbuf->ino, &ubuf->st_ino_lo); + err |= __put_user(kbuf->ino >> 32, &ubuf->st_ino_hi); + err |= __put_user(kbuf->mode, &ubuf->st_mode); + err |= __put_user(kbuf->nlink, &ubuf->st_nlink); + err |= __put_user(kbuf->uid, &ubuf->st_uid); + err |= __put_user(kbuf->gid, &ubuf->st_gid); + hdev = huge_encode_dev(kbuf->rdev); + err = __put_user(hdev, (u32 __user*)&ubuf->st_rdev); + err |= __put_user(hdev >> 32, ((u32 __user*)&ubuf->st_rdev) + 1); + err |= __put_user(kbuf->size, &ubuf->st_size_lo); + err |= __put_user((kbuf->size >> 32), &ubuf->st_size_hi); + err |= __put_user(kbuf->atime.tv_sec, &ubuf->st_atime); + err |= __put_user(kbuf->atime.tv_nsec, &ubuf->st_atime_nsec); + err |= __put_user(kbuf->mtime.tv_sec, &ubuf->st_mtime); + err |= __put_user(kbuf->mtime.tv_nsec, &ubuf->st_mtime_nsec); + err |= __put_user(kbuf->ctime.tv_sec, &ubuf->st_ctime); + err |= __put_user(kbuf->ctime.tv_nsec, &ubuf->st_ctime_nsec); + err |= __put_user(kbuf->blksize, &ubuf->st_blksize); + err |= __put_user(kbuf->blocks, &ubuf->st_blocks); + return err; +} + +asmlinkage long +sys32_stat64 (char __user *filename, struct stat64 __user *statbuf) +{ + struct kstat s; + long ret = vfs_stat(filename, &s); + if (!ret) + ret = putstat64(statbuf, &s); + return ret; +} + +asmlinkage long +sys32_lstat64 (char __user *filename, struct stat64 __user *statbuf) +{ + struct kstat s; + long ret = vfs_lstat(filename, &s); + if (!ret) + ret = putstat64(statbuf, &s); + return ret; +} + +asmlinkage long +sys32_fstat64 (unsigned int fd, struct stat64 __user *statbuf) +{ + struct kstat s; + long ret = vfs_fstat(fd, &s); + if (!ret) + ret = putstat64(statbuf, &s); + return ret; +} + +struct sysinfo32 { + s32 uptime; + u32 loads[3]; + u32 totalram; + u32 freeram; + u32 sharedram; + u32 bufferram; + u32 totalswap; + u32 freeswap; + u16 procs; + u16 pad; + u32 totalhigh; + u32 freehigh; + u32 mem_unit; + char _f[8]; +}; + +asmlinkage long +sys32_sysinfo (struct sysinfo32 __user *info) +{ + struct sysinfo s; + long ret, err; + int bitcount = 0; + mm_segment_t old_fs = get_fs(); + + set_fs(KERNEL_DS); + ret = sys_sysinfo((struct sysinfo __user *) &s); + set_fs(old_fs); + /* Check to see if any memory value is too large for 32-bit and + * scale down if needed. + */ + if ((s.totalram >> 32) || (s.totalswap >> 32)) { + while (s.mem_unit < PAGE_SIZE) { + s.mem_unit <<= 1; + bitcount++; + } + s.totalram >>= bitcount; + s.freeram >>= bitcount; + s.sharedram >>= bitcount; + s.bufferram >>= bitcount; + s.totalswap >>= bitcount; + s.freeswap >>= bitcount; + s.totalhigh >>= bitcount; + s.freehigh >>= bitcount; + } + + if (!access_ok(VERIFY_WRITE, info, sizeof(*info))) + return -EFAULT; + + err = __put_user(s.uptime, &info->uptime); + err |= __put_user(s.loads[0], &info->loads[0]); + err |= __put_user(s.loads[1], &info->loads[1]); + err |= __put_user(s.loads[2], &info->loads[2]); + err |= __put_user(s.totalram, &info->totalram); + err |= __put_user(s.freeram, &info->freeram); + err |= __put_user(s.sharedram, &info->sharedram); + err |= __put_user(s.bufferram, &info->bufferram); + err |= __put_user(s.totalswap, &info->totalswap); + err |= __put_user(s.freeswap, &info->freeswap); + err |= __put_user(s.procs, &info->procs); + err |= __put_user (s.totalhigh, &info->totalhigh); + err |= __put_user (s.freehigh, &info->freehigh); + err |= __put_user (s.mem_unit, &info->mem_unit); + if (err) + return -EFAULT; + return ret; +} + +asmlinkage long +sys32_sched_rr_get_interval (pid_t pid, struct compat_timespec __user *interval) +{ + mm_segment_t old_fs = get_fs(); + struct timespec t; + long ret; + + set_fs(KERNEL_DS); + ret = sys_sched_rr_get_interval(pid, (struct timespec __user *) &t); + set_fs(old_fs); + if (put_compat_timespec(&t, interval)) + return -EFAULT; + return ret; +} + +asmlinkage long +sys32_pread (unsigned int fd, void __user *buf, unsigned int count, u32 pos_lo, u32 pos_hi) +{ + return sys_pread64(fd, buf, count, ((unsigned long) pos_hi << 32) | pos_lo); +} + +asmlinkage long +sys32_pwrite (unsigned int fd, void __user *buf, unsigned int count, u32 pos_lo, u32 pos_hi) +{ + return sys_pwrite64(fd, buf, count, ((unsigned long) pos_hi << 32) | pos_lo); +} + +asmlinkage long +sys32_sendfile (int out_fd, int in_fd, int __user *offset, unsigned int count) +{ + mm_segment_t old_fs = get_fs(); + long ret; + off_t of; + + if (offset && get_user(of, offset)) + return -EFAULT; + + set_fs(KERNEL_DS); + ret = sys_sendfile(out_fd, in_fd, offset ? (off_t __user *) &of : NULL, count); + set_fs(old_fs); + + if (!ret && offset && put_user(of, offset)) + return -EFAULT; + + return ret; +} + +asmlinkage long +sys32_personality (unsigned int personality) +{ + long ret; + + if (current->personality == PER_LINUX32 && personality == PER_LINUX) + personality = PER_LINUX32; + ret = sys_personality(personality); + if (ret == PER_LINUX32) + ret = PER_LINUX; + return ret; +} + +asmlinkage unsigned long +sys32_brk (unsigned int brk) +{ + unsigned long ret, obrk; + struct mm_struct *mm = current->mm; + + obrk = mm->brk; + ret = sys_brk(brk); + if (ret < obrk) + clear_user(compat_ptr(ret), PAGE_ALIGN(ret) - ret); + return ret; +} + +/* + * Exactly like fs/open.c:sys_open(), except that it doesn't set the O_LARGEFILE flag. + */ +asmlinkage long +sys32_open (const char __user * filename, int flags, int mode) +{ + char * tmp; + int fd, error; + + tmp = getname(filename); + fd = PTR_ERR(tmp); + if (!IS_ERR(tmp)) { + fd = get_unused_fd(); + if (fd >= 0) { + struct file *f = filp_open(tmp, flags, mode); + error = PTR_ERR(f); + if (IS_ERR(f)) + goto out_error; + fd_install(fd, f); + } +out: + putname(tmp); + } + return fd; + +out_error: + put_unused_fd(fd); + fd = error; + goto out; +} + +/* Structure for ia32 emulation on ia64 */ +struct epoll_event32 +{ + u32 events; + u32 data[2]; +}; + +asmlinkage long +sys32_epoll_ctl(int epfd, int op, int fd, struct epoll_event32 __user *event) +{ + mm_segment_t old_fs = get_fs(); + struct epoll_event event64; + int error; + u32 data_halfword; + + if (!access_ok(VERIFY_READ, event, sizeof(struct epoll_event32))) + return -EFAULT; + + __get_user(event64.events, &event->events); + __get_user(data_halfword, &event->data[0]); + event64.data = data_halfword; + __get_user(data_halfword, &event->data[1]); + event64.data |= (u64)data_halfword << 32; + + set_fs(KERNEL_DS); + error = sys_epoll_ctl(epfd, op, fd, (struct epoll_event __user *) &event64); + set_fs(old_fs); + + return error; +} + +asmlinkage long +sys32_epoll_wait(int epfd, struct epoll_event32 __user * events, int maxevents, + int timeout) +{ + struct epoll_event *events64 = NULL; + mm_segment_t old_fs = get_fs(); + int error, numevents, size; + int evt_idx; + int do_free_pages = 0; + + if (maxevents <= 0) { + return -EINVAL; + } + + /* Verify that the area passed by the user is writeable */ + if (!access_ok(VERIFY_WRITE, events, maxevents * sizeof(struct epoll_event32))) + return -EFAULT; + + /* + * Allocate space for the intermediate copy. If the space needed + * is large enough to cause kmalloc to fail, then try again with + * __get_free_pages. + */ + size = maxevents * sizeof(struct epoll_event); + events64 = kmalloc(size, GFP_KERNEL); + if (events64 == NULL) { + events64 = (struct epoll_event *) + __get_free_pages(GFP_KERNEL, get_order(size)); + if (events64 == NULL) + return -ENOMEM; + do_free_pages = 1; + } + + /* Do the system call */ + set_fs(KERNEL_DS); /* copy_to/from_user should work on kernel mem*/ + numevents = sys_epoll_wait(epfd, (struct epoll_event __user *) events64, + maxevents, timeout); + set_fs(old_fs); + + /* Don't modify userspace memory if we're returning an error */ + if (numevents > 0) { + /* Translate the 64-bit structures back into the 32-bit + structures */ + for (evt_idx = 0; evt_idx < numevents; evt_idx++) { + __put_user(events64[evt_idx].events, + &events[evt_idx].events); + __put_user((u32)events64[evt_idx].data, + &events[evt_idx].data[0]); + __put_user((u32)(events64[evt_idx].data >> 32), + &events[evt_idx].data[1]); + } + } + + if (do_free_pages) + free_pages((unsigned long) events64, get_order(size)); + else + kfree(events64); + return numevents; +} + +/* + * Get a yet unused TLS descriptor index. + */ +static int +get_free_idx (void) +{ + struct thread_struct *t = ¤t->thread; + int idx; + + for (idx = 0; idx < GDT_ENTRY_TLS_ENTRIES; idx++) + if (desc_empty(t->tls_array + idx)) + return idx + GDT_ENTRY_TLS_MIN; + return -ESRCH; +} + +/* + * Set a given TLS descriptor: + */ +asmlinkage int +sys32_set_thread_area (struct ia32_user_desc __user *u_info) +{ + struct thread_struct *t = ¤t->thread; + struct ia32_user_desc info; + struct desc_struct *desc; + int cpu, idx; + + if (copy_from_user(&info, u_info, sizeof(info))) + return -EFAULT; + idx = info.entry_number; + + /* + * index -1 means the kernel should try to find and allocate an empty descriptor: + */ + if (idx == -1) { + idx = get_free_idx(); + if (idx < 0) + return idx; + if (put_user(idx, &u_info->entry_number)) + return -EFAULT; + } + + if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX) + return -EINVAL; + + desc = t->tls_array + idx - GDT_ENTRY_TLS_MIN; + + cpu = smp_processor_id(); + + if (LDT_empty(&info)) { + desc->a = 0; + desc->b = 0; + } else { + desc->a = LDT_entry_a(&info); + desc->b = LDT_entry_b(&info); + } + load_TLS(t, cpu); + return 0; +} + +/* + * Get the current Thread-Local Storage area: + */ + +#define GET_BASE(desc) ( \ + (((desc)->a >> 16) & 0x0000ffff) | \ + (((desc)->b << 16) & 0x00ff0000) | \ + ( (desc)->b & 0xff000000) ) + +#define GET_LIMIT(desc) ( \ + ((desc)->a & 0x0ffff) | \ + ((desc)->b & 0xf0000) ) + +#define GET_32BIT(desc) (((desc)->b >> 22) & 1) +#define GET_CONTENTS(desc) (((desc)->b >> 10) & 3) +#define GET_WRITABLE(desc) (((desc)->b >> 9) & 1) +#define GET_LIMIT_PAGES(desc) (((desc)->b >> 23) & 1) +#define GET_PRESENT(desc) (((desc)->b >> 15) & 1) +#define GET_USEABLE(desc) (((desc)->b >> 20) & 1) + +asmlinkage int +sys32_get_thread_area (struct ia32_user_desc __user *u_info) +{ + struct ia32_user_desc info; + struct desc_struct *desc; + int idx; + + if (get_user(idx, &u_info->entry_number)) + return -EFAULT; + if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX) + return -EINVAL; + + desc = current->thread.tls_array + idx - GDT_ENTRY_TLS_MIN; + + info.entry_number = idx; + info.base_addr = GET_BASE(desc); + info.limit = GET_LIMIT(desc); + info.seg_32bit = GET_32BIT(desc); + info.contents = GET_CONTENTS(desc); + info.read_exec_only = !GET_WRITABLE(desc); + info.limit_in_pages = GET_LIMIT_PAGES(desc); + info.seg_not_present = !GET_PRESENT(desc); + info.useable = GET_USEABLE(desc); + + if (copy_to_user(u_info, &info, sizeof(info))) + return -EFAULT; + return 0; +} + +asmlinkage long +sys32_timer_create(u32 clock, struct compat_sigevent __user *se32, timer_t __user *timer_id) +{ + struct sigevent se; + mm_segment_t oldfs; + timer_t t; + long err; + + if (se32 == NULL) + return sys_timer_create(clock, NULL, timer_id); + + if (get_compat_sigevent(&se, se32)) + return -EFAULT; + + if (!access_ok(VERIFY_WRITE,timer_id,sizeof(timer_t))) + return -EFAULT; + + oldfs = get_fs(); + set_fs(KERNEL_DS); + err = sys_timer_create(clock, (struct sigevent __user *) &se, (timer_t __user *) &t); + set_fs(oldfs); + + if (!err) + err = __put_user (t, timer_id); + + return err; +} + +long sys32_fadvise64_64(int fd, __u32 offset_low, __u32 offset_high, + __u32 len_low, __u32 len_high, int advice) +{ + return sys_fadvise64_64(fd, + (((u64)offset_high)<<32) | offset_low, + (((u64)len_high)<<32) | len_low, + advice); +} + +#ifdef NOTYET /* UNTESTED FOR IA64 FROM HERE DOWN */ + +asmlinkage long sys32_setreuid(compat_uid_t ruid, compat_uid_t euid) +{ + uid_t sruid, seuid; + + sruid = (ruid == (compat_uid_t)-1) ? ((uid_t)-1) : ((uid_t)ruid); + seuid = (euid == (compat_uid_t)-1) ? ((uid_t)-1) : ((uid_t)euid); + return sys_setreuid(sruid, seuid); +} + +asmlinkage long +sys32_setresuid(compat_uid_t ruid, compat_uid_t euid, + compat_uid_t suid) +{ + uid_t sruid, seuid, ssuid; + + sruid = (ruid == (compat_uid_t)-1) ? ((uid_t)-1) : ((uid_t)ruid); + seuid = (euid == (compat_uid_t)-1) ? ((uid_t)-1) : ((uid_t)euid); + ssuid = (suid == (compat_uid_t)-1) ? ((uid_t)-1) : ((uid_t)suid); + return sys_setresuid(sruid, seuid, ssuid); +} + +asmlinkage long +sys32_setregid(compat_gid_t rgid, compat_gid_t egid) +{ + gid_t srgid, segid; + + srgid = (rgid == (compat_gid_t)-1) ? ((gid_t)-1) : ((gid_t)rgid); + segid = (egid == (compat_gid_t)-1) ? ((gid_t)-1) : ((gid_t)egid); + return sys_setregid(srgid, segid); +} + +asmlinkage long +sys32_setresgid(compat_gid_t rgid, compat_gid_t egid, + compat_gid_t sgid) +{ + gid_t srgid, segid, ssgid; + + srgid = (rgid == (compat_gid_t)-1) ? ((gid_t)-1) : ((gid_t)rgid); + segid = (egid == (compat_gid_t)-1) ? ((gid_t)-1) : ((gid_t)egid); + ssgid = (sgid == (compat_gid_t)-1) ? ((gid_t)-1) : ((gid_t)sgid); + return sys_setresgid(srgid, segid, ssgid); +} + +/* Handle adjtimex compatibility. */ + +struct timex32 { + u32 modes; + s32 offset, freq, maxerror, esterror; + s32 status, constant, precision, tolerance; + struct compat_timeval time; + s32 tick; + s32 ppsfreq, jitter, shift, stabil; + s32 jitcnt, calcnt, errcnt, stbcnt; + s32 :32; s32 :32; s32 :32; s32 :32; + s32 :32; s32 :32; s32 :32; s32 :32; + s32 :32; s32 :32; s32 :32; s32 :32; +}; + +extern int do_adjtimex(struct timex *); + +asmlinkage long +sys32_adjtimex(struct timex32 *utp) +{ + struct timex txc; + int ret; + + memset(&txc, 0, sizeof(struct timex)); + + if(get_user(txc.modes, &utp->modes) || + __get_user(txc.offset, &utp->offset) || + __get_user(txc.freq, &utp->freq) || + __get_user(txc.maxerror, &utp->maxerror) || + __get_user(txc.esterror, &utp->esterror) || + __get_user(txc.status, &utp->status) || + __get_user(txc.constant, &utp->constant) || + __get_user(txc.precision, &utp->precision) || + __get_user(txc.tolerance, &utp->tolerance) || + __get_user(txc.time.tv_sec, &utp->time.tv_sec) || + __get_user(txc.time.tv_usec, &utp->time.tv_usec) || + __get_user(txc.tick, &utp->tick) || + __get_user(txc.ppsfreq, &utp->ppsfreq) || + __get_user(txc.jitter, &utp->jitter) || + __get_user(txc.shift, &utp->shift) || + __get_user(txc.stabil, &utp->stabil) || + __get_user(txc.jitcnt, &utp->jitcnt) || + __get_user(txc.calcnt, &utp->calcnt) || + __get_user(txc.errcnt, &utp->errcnt) || + __get_user(txc.stbcnt, &utp->stbcnt)) + return -EFAULT; + + ret = do_adjtimex(&txc); + + if(put_user(txc.modes, &utp->modes) || + __put_user(txc.offset, &utp->offset) || + __put_user(txc.freq, &utp->freq) || + __put_user(txc.maxerror, &utp->maxerror) || + __put_user(txc.esterror, &utp->esterror) || + __put_user(txc.status, &utp->status) || + __put_user(txc.constant, &utp->constant) || + __put_user(txc.precision, &utp->precision) || + __put_user(txc.tolerance, &utp->tolerance) || + __put_user(txc.time.tv_sec, &utp->time.tv_sec) || + __put_user(txc.time.tv_usec, &utp->time.tv_usec) || + __put_user(txc.tick, &utp->tick) || + __put_user(txc.ppsfreq, &utp->ppsfreq) || + __put_user(txc.jitter, &utp->jitter) || + __put_user(txc.shift, &utp->shift) || + __put_user(txc.stabil, &utp->stabil) || + __put_user(txc.jitcnt, &utp->jitcnt) || + __put_user(txc.calcnt, &utp->calcnt) || + __put_user(txc.errcnt, &utp->errcnt) || + __put_user(txc.stbcnt, &utp->stbcnt)) + ret = -EFAULT; + + return ret; +} +#endif /* NOTYET */ -- cgit v1.2.1