From mboxrd@z Thu Jan 1 00:00:00 1970 From: nathan_lynch@mentor.com (Nathan Lynch) Date: Sun, 22 Jun 2014 22:11:57 -0500 Subject: [PATCH v7 8/9] ARM: vdso initialization, mapping, and synchronization In-Reply-To: <1403493118-7597-1-git-send-email-nathan_lynch@mentor.com> References: <1403493118-7597-1-git-send-email-nathan_lynch@mentor.com> Message-ID: <1403493118-7597-9-git-send-email-nathan_lynch@mentor.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Initialize the vdso page list at boot, install the vdso mapping at exec time, and update the data page during timer ticks. This code is not built if CONFIG_VDSO is not enabled. Account for the vdso length when randomizing the offset from the stack. The vdso is placed immediately following the sigpage with a separate install_special_mapping call in arm_install_vdso. Signed-off-by: Nathan Lynch --- arch/arm/kernel/process.c | 13 +++- arch/arm/kernel/vdso.c | 168 ++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 179 insertions(+), 2 deletions(-) create mode 100644 arch/arm/kernel/vdso.c diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c index 40445fb71ac9..39b0d68aa068 100644 --- a/arch/arm/kernel/process.c +++ b/arch/arm/kernel/process.c @@ -520,6 +520,7 @@ extern struct page *get_signal_page(void); int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) { struct mm_struct *mm = current->mm; + unsigned long npages; unsigned long addr; unsigned long hint; int ret; @@ -529,9 +530,11 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) if (!signal_page) return -ENOMEM; + npages = (vdso_mapping_len >> PAGE_SHIFT) + 1; + down_write(&mm->mmap_sem); - hint = vdso_addr(mm, 1); - addr = get_unmapped_area(NULL, hint, PAGE_SIZE, 0, 0); + hint = vdso_addr(mm, npages); + addr = get_unmapped_area(NULL, hint, npages, 0, 0); if (IS_ERR_VALUE(addr)) { ret = addr; goto up_fail; @@ -544,6 +547,12 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) if (ret == 0) mm->context.sigpage = addr; + /* Unlike the sigpage, failure to install the vdso is unlikely + * to be fatal to the process, so no error check needed + * here. + */ + arm_install_vdso(mm, addr + PAGE_SIZE); + up_fail: up_write(&mm->mmap_sem); return ret; diff --git a/arch/arm/kernel/vdso.c b/arch/arm/kernel/vdso.c new file mode 100644 index 000000000000..0cfd25f09adf --- /dev/null +++ b/arch/arm/kernel/vdso.c @@ -0,0 +1,168 @@ +/* + * Adapted from arm64 version. + * + * Copyright (C) 2012 ARM Limited + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +static struct page **vdso_pagelist; + +unsigned long vdso_mapping_len __read_mostly; + +/* + * The vDSO data page. + */ +static union vdso_data_store vdso_data_store __page_aligned_data; +static struct vdso_data *vdso_data = &vdso_data_store.data; + +static int __init vdso_init(void) +{ + unsigned long vdso_pages; + int i; + + if (memcmp(&vdso_start, "\177ELF", 4)) { + pr_err("vDSO is not a valid ELF object!\n"); + return -ENOEXEC; + } + + vdso_pages = (&vdso_end - &vdso_start) >> PAGE_SHIFT; + pr_debug("vdso: %ld code pages at base %p\n", vdso_pages, &vdso_start); + + /* Allocate the vDSO pagelist, plus a page for the data. */ + vdso_pagelist = kcalloc(vdso_pages + 1, sizeof(struct page *), + GFP_KERNEL); + if (vdso_pagelist == NULL) + return -ENOMEM; + + /* Grab the vDSO data page. */ + vdso_pagelist[0] = virt_to_page(vdso_data); + + /* Grab the vDSO code pages. */ + for (i = 0; i < vdso_pages; i++) + vdso_pagelist[i + 1] = virt_to_page(&vdso_start + i * PAGE_SIZE); + + /* Precompute the mapping size */ + vdso_mapping_len = (vdso_pages + 1) << PAGE_SHIFT; + + return 0; +} +arch_initcall(vdso_init); + +/* assumes mmap_sem is write-locked */ +void arm_install_vdso(struct mm_struct *mm, unsigned long addr) +{ + int ret; + + mm->context.vdso = ~0UL; + + if (vdso_pagelist == NULL) + return; + + /* + * Put vDSO base into mm struct before calling + * install_special_mapping so the perf counter mmap tracking + * code will recognise it as a vDSO. + */ + mm->context.vdso = addr; + + ret = install_special_mapping(mm, addr, vdso_mapping_len, + VM_READ|VM_EXEC| + VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC, + vdso_pagelist); + if (ret) { + pr_notice_once("%s: install_special_mapping failed (%d)\n", + __func__, ret); + mm->context.vdso = ~0UL; + return; + } +} + +static void vdso_write_begin(struct vdso_data *vdata) +{ + ++vdso_data->seq_count; + smp_wmb(); +} + +static void vdso_write_end(struct vdso_data *vdata) +{ + smp_wmb(); + ++vdso_data->seq_count; +} + +/** + * update_vsyscall - update the vdso data page + * + * Increment the sequence counter, making it odd, indicating to + * userspace that an update is in progress. Update the fields used + * for coarse clocks and, if the architected system timer is in use, + * the fields used for high precision clocks. Increment the sequence + * counter again, making it even, indicating to userspace that the + * update is finished. + * + * Userspace is expected to sample seq_count before reading any other + * fields from the data page. If seq_count is odd, userspace is + * expected to wait until it becomes even. After copying data from + * the page, userspace must sample seq_count again; if it has changed + * from its previous value, userspace must retry the whole sequence. + * + * Calls to update_vsyscall are serialized by the timekeeping core. + */ +void update_vsyscall(struct timekeeper *tk) +{ + struct timespec xtime_coarse; + struct timespec *wtm = &tk->wall_to_monotonic; + bool use_syscall = strcmp(tk->clock->name, "arch_sys_counter"); + + vdso_write_begin(vdso_data); + + xtime_coarse = __current_kernel_time(); + vdso_data->use_syscall = use_syscall; + vdso_data->xtime_coarse_sec = xtime_coarse.tv_sec; + vdso_data->xtime_coarse_nsec = xtime_coarse.tv_nsec; + vdso_data->wtm_clock_sec = wtm->tv_sec; + vdso_data->wtm_clock_nsec = wtm->tv_nsec; + + if (!use_syscall) { + vdso_data->cs_cycle_last = tk->cycle_last; + vdso_data->xtime_clock_sec = tk->xtime_sec; + vdso_data->xtime_clock_snsec = tk->xtime_nsec; + vdso_data->cs_mult = tk->mult; + vdso_data->cs_shift = tk->shift; + vdso_data->cs_mask = tk->clock->mask; + } + + vdso_write_end(vdso_data); + + flush_dcache_page(virt_to_page(vdso_data)); +} + +void update_vsyscall_tz(void) +{ + vdso_data->tz_minuteswest = sys_tz.tz_minuteswest; + vdso_data->tz_dsttime = sys_tz.tz_dsttime; + flush_dcache_page(virt_to_page(vdso_data)); +} -- 1.9.3