From: luto@amacapital•net (Andy Lutomirski)
To: linux-arm-kernel@lists•infradead.org
Subject: [PATCH v7 8/9] ARM: vdso initialization, mapping, and synchronization
Date: Mon, 30 Jun 2014 14:37:48 -0700 [thread overview]
Message-ID: <53B1D8AC.7060104@mit.edu> (raw)
In-Reply-To: <1403493118-7597-9-git-send-email-nathan_lynch@mentor.com>
On 06/22/2014 08:11 PM, Nathan Lynch wrote:
> 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 <nathan_lynch@mentor•com>
> ---
> 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 <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/err.h>
> +#include <linux/kernel.h>
> +#include <linux/mm.h>
> +#include <linux/slab.h>
> +#include <linux/timekeeper_internal.h>
> +#include <linux/vmalloc.h>
> +
> +#include <asm/barrier.h>
> +#include <asm/cacheflush.h>
> +#include <asm/page.h>
> +#include <asm/vdso.h>
> +#include <asm/vdso_datapage.h>
> +
> +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);
Eek. You're mapping the shared data VM_MAYWRITE. This will cause
bizarre and confusing failures if ptrace pokes at it. You also seem to
be sticking it *before* the vdso in the same vma. This will severely
piss off all the tools that assume that "[vdso]" points to an ELF object.
x86 calls this thing "[vvar]" and sticks it after the vdso. You might
not want to have the complexity of sticking it after the vdso (it's
distinctly nontrivial), but I see nothing wrong with giving it its own
vma just before the vdso. The new _install_special_mapping function
makes it considerably easier to do.
(Don't use any version of x86's code before 3.16-rc3 as a reference.)
--Andy
next prev parent reply other threads:[~2014-06-30 21:37 UTC|newest]
Thread overview: 76+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-06-23 3:11 [PATCH v7 0/9] ARM: VDSO Nathan Lynch
2014-06-23 3:11 ` [PATCH v7 1/9] clocksource: arm_arch_timer: change clocksource name if CP15 unavailable Nathan Lynch
2014-06-23 3:11 ` [PATCH v7 2/9] clocksource: arm_arch_timer: enable counter access for 32-bit ARM Nathan Lynch
2014-06-23 3:11 ` [PATCH v7 3/9] ARM: arch_timer: remove unused functions Nathan Lynch
2014-06-23 3:11 ` [PATCH v7 4/9] arm64: " Nathan Lynch
2014-06-23 3:11 ` [PATCH v7 5/9] ARM: place sigpage at a random offset above stack Nathan Lynch
2014-06-23 3:11 ` [PATCH v7 6/9] ARM: miscellaneous vdso infrastructure, preparation Nathan Lynch
2014-06-30 10:11 ` Arnd Bergmann
2014-06-30 12:27 ` Nathan Lynch
2014-06-30 13:06 ` Arnd Bergmann
2014-06-23 3:11 ` [PATCH v7 7/9] ARM: add vdso user-space code Nathan Lynch
2014-06-28 9:53 ` Russell King - ARM Linux
2014-06-28 10:03 ` Russell King - ARM Linux
2014-06-30 15:56 ` Andy Lutomirski
2014-06-28 15:19 ` Nathan Lynch
2014-06-28 15:26 ` Russell King - ARM Linux
2014-06-28 16:13 ` Nathan Lynch
2014-06-28 18:12 ` Russell King - ARM Linux
2014-06-28 19:45 ` Nathan Lynch
2014-06-28 20:11 ` Russell King - ARM Linux
2014-06-28 21:35 ` Nathan Lynch
2014-06-29 8:34 ` Russell King - ARM Linux
2014-06-29 15:48 ` Nathan Lynch
2014-06-29 16:07 ` Russell King - ARM Linux
2014-06-29 23:04 ` Nathan Lynch
2014-06-30 21:33 ` Andy Lutomirski
2014-06-30 15:59 ` Andy Lutomirski
2014-06-30 16:50 ` Nathan Lynch
2014-06-30 21:29 ` Andy Lutomirski
2014-07-01 9:00 ` Will Deacon
2014-07-01 13:34 ` Nathan Lynch
2014-07-01 14:11 ` Andy Lutomirski
2014-07-01 13:28 ` Nathan Lynch
2014-07-01 14:56 ` Andy Lutomirski
2014-06-23 3:11 ` [PATCH v7 8/9] ARM: vdso initialization, mapping, and synchronization Nathan Lynch
2014-06-30 21:37 ` Andy Lutomirski [this message]
2014-07-01 9:03 ` Will Deacon
2014-07-01 14:11 ` Nathan Lynch
2014-07-01 14:15 ` Will Deacon
2014-07-01 14:17 ` Andy Lutomirski
2014-07-01 17:27 ` Christopher Covington
2014-07-02 14:40 ` Will Deacon
2014-07-02 15:54 ` Andy Lutomirski
2014-07-02 16:18 ` Nathan Lynch
2014-07-02 16:27 ` Will Deacon
2014-07-02 16:47 ` Andy Lutomirski
2014-07-02 17:24 ` Will Deacon
2014-07-02 18:34 ` Andy Lutomirski
2014-07-02 18:54 ` Will Deacon
2014-07-22 0:14 ` Andy Lutomirski
2014-07-22 8:13 ` Will Deacon
2014-07-01 14:01 ` Nathan Lynch
2014-07-01 14:09 ` Andy Lutomirski
2014-07-01 14:14 ` Russell King - ARM Linux
2014-06-23 3:11 ` [PATCH v7 9/9] ARM: add CONFIG_VDSO Kconfig and Makefile bits Nathan Lynch
2014-06-27 8:51 ` [PATCH v7 0/9] ARM: VDSO Jan Glauber
2014-06-27 8:57 ` Russell King - ARM Linux
2014-06-27 9:41 ` Ard Biesheuvel
2014-06-27 9:46 ` Russell King - ARM Linux
2014-06-27 17:01 ` Nathan Lynch
2014-06-28 9:42 ` Russell King - ARM Linux
2014-06-28 9:55 ` Russell King - ARM Linux
2014-06-28 14:49 ` Russell King - ARM Linux
2014-06-30 7:27 ` Arnd Bergmann
2014-06-30 7:43 ` Arnd Bergmann
2014-06-30 18:27 ` Nathan Lynch
2014-06-27 16:00 ` Nathan Lynch
2014-06-30 16:03 ` Andy Lutomirski
2014-06-30 8:12 ` Ard Biesheuvel
2014-07-01 16:34 ` Nathan Lynch
2014-07-01 20:08 ` Ard Biesheuvel
2014-07-01 22:44 ` Russell King - ARM Linux
2014-06-30 14:40 ` Will Deacon
2014-06-30 15:42 ` Nathan Lynch
2014-06-30 21:56 ` Andy Lutomirski
2014-07-01 9:04 ` Will Deacon
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=53B1D8AC.7060104@mit.edu \
--to=luto@amacapital$(echo .)net \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox