| Message ID | 20220601210951.3916598-3-axelrasmussen@google.com |
|---|---|
| State | New |
| Headers | show |
| Series | userfaultfd: add /dev/userfaultfd for fine grained access control | expand |
On Wed, 1 Jun 2022 14:09:47 -0700 Axel Rasmussen <axelrasmussen@google.com> wrote: > To achieve this, add a /dev/userfaultfd misc device. This device > provides an alternative to the userfaultfd(2) syscall for the creation > of new userfaultfds. The idea is, any userfaultfds created this way will > be able to handle kernel faults, without the caller having any special > capabilities. Access to this mechanism is instead restricted using e.g. > standard filesystem permissions. The use of a /dev node isn't pretty. Why can't this be done by tweaking sys_userfaultfd() or by adding a sys_userfaultfd2()? Peter, will you be completing review of this patchset?
On Mon, Jun 13, 2022 at 02:55:40PM -0700, Andrew Morton wrote: > On Wed, 1 Jun 2022 14:09:47 -0700 Axel Rasmussen <axelrasmussen@google.com> wrote: > > > To achieve this, add a /dev/userfaultfd misc device. This device > > provides an alternative to the userfaultfd(2) syscall for the creation > > of new userfaultfds. The idea is, any userfaultfds created this way will > > be able to handle kernel faults, without the caller having any special > > capabilities. Access to this mechanism is instead restricted using e.g. > > standard filesystem permissions. > > The use of a /dev node isn't pretty. Why can't this be done by > tweaking sys_userfaultfd() or by adding a sys_userfaultfd2()? > > Peter, will you be completing review of this patchset? Sorry to not have reviewed it proactively.. I think it's because I never had a good picture/understanding of what should be the best security model for uffd, meanwhile I am (it seems) just seeing more and more ways to "provide a safer uffd" by different people using different ways.. and I never had time (and probably capability too..) to figure out the correct approach if not to accept all options provided. I think I'll just assume the whole thing is acked already from you generally, then I'll read at least the implementation before the end of tomorrow. Thanks,
On Mon, Jun 13, 2022 at 3:29 PM Peter Xu <peterx@redhat.com> wrote: > > On Mon, Jun 13, 2022 at 02:55:40PM -0700, Andrew Morton wrote: > > On Wed, 1 Jun 2022 14:09:47 -0700 Axel Rasmussen <axelrasmussen@google.com> wrote: > > > > > To achieve this, add a /dev/userfaultfd misc device. This device > > > provides an alternative to the userfaultfd(2) syscall for the creation > > > of new userfaultfds. The idea is, any userfaultfds created this way will > > > be able to handle kernel faults, without the caller having any special > > > capabilities. Access to this mechanism is instead restricted using e.g. > > > standard filesystem permissions. > > > > The use of a /dev node isn't pretty. Why can't this be done by > > tweaking sys_userfaultfd() or by adding a sys_userfaultfd2()? I think for any approach involving syscalls, we need to be able to control access to who can call a syscall. Maybe there's another way I'm not aware of, but I think today the only mechanism to do this is capabilities. I proposed adding a CAP_USERFAULTFD for this purpose, but that approach was rejected [1]. So, I'm not sure of another way besides using a device node. One thing that could potentially make this cleaner is, as one LWN commenter pointed out, we could have open() on /dev/userfaultfd just return a new userfaultfd directly, instead of this multi-step process of open /dev/userfaultfd, NEW ioctl, then you get a userfaultfd. When I wrote this originally it wasn't clear to me how to get that to happen - open() doesn't directly return the result of our custom open function pointer, as far as I can tell - but it could be investigated. [1]: https://lore.kernel.org/lkml/686276b9-4530-2045-6bd8-170e5943abe4@schaufler-ca.com/T/ > > > > Peter, will you be completing review of this patchset? > > Sorry to not have reviewed it proactively.. > > I think it's because I never had a good picture/understanding of what > should be the best security model for uffd, meanwhile I am (it seems) just > seeing more and more ways to "provide a safer uffd" by different people > using different ways.. and I never had time (and probably capability too..) > to figure out the correct approach if not to accept all options provided. Agreed, what we have right now is a bit of a mess of different approaches. I think the reason for this is, there is no "perfect" way to control access to features like this, so what we now have is several different approaches with different tradeoffs.
Axel Rasmussen <axelrasmussen@google.com> writes: > I think for any approach involving syscalls, we need to be able to > control access to who can call a syscall. Maybe there's another way > I'm not aware of, but I think today the only mechanism to do this is > capabilities. I proposed adding a CAP_USERFAULTFD for this purpose, > but that approach was rejected [1]. So, I'm not sure of another way > besides using a device node. I take it there's a reason why this can't be done with a security module - either a custom module or a policy in one of the existing modules? That sort of access control is just what security modules are supposed to be for, after all. Thanks, jon
On Jun 13, 2022, at 3:38 PM, Axel Rasmussen <axelrasmussen@google.com> wrote: > On Mon, Jun 13, 2022 at 3:29 PM Peter Xu <peterx@redhat.com> wrote: >> On Mon, Jun 13, 2022 at 02:55:40PM -0700, Andrew Morton wrote: >>> On Wed, 1 Jun 2022 14:09:47 -0700 Axel Rasmussen <axelrasmussen@google.com> wrote: >>> >>>> To achieve this, add a /dev/userfaultfd misc device. This device >>>> provides an alternative to the userfaultfd(2) syscall for the creation >>>> of new userfaultfds. The idea is, any userfaultfds created this way will >>>> be able to handle kernel faults, without the caller having any special >>>> capabilities. Access to this mechanism is instead restricted using e.g. >>>> standard filesystem permissions. >>> >>> The use of a /dev node isn't pretty. Why can't this be done by >>> tweaking sys_userfaultfd() or by adding a sys_userfaultfd2()? > > I think for any approach involving syscalls, we need to be able to > control access to who can call a syscall. Maybe there's another way > I'm not aware of, but I think today the only mechanism to do this is > capabilities. I proposed adding a CAP_USERFAULTFD for this purpose, > but that approach was rejected [1]. So, I'm not sure of another way > besides using a device node. > > One thing that could potentially make this cleaner is, as one LWN > commenter pointed out, we could have open() on /dev/userfaultfd just > return a new userfaultfd directly, instead of this multi-step process > of open /dev/userfaultfd, NEW ioctl, then you get a userfaultfd. When > I wrote this originally it wasn't clear to me how to get that to > happen - open() doesn't directly return the result of our custom open > function pointer, as far as I can tell - but it could be investigated. If this direction is pursued, I think that it would be better to set it as /proc/[pid]/userfaultfd, which would allow remote monitors (processes) to hook into userfaultfd of remote processes. I have a patch for that which extends userfaultfd syscall, but /proc/[pid]/userfaultfd may be cleaner.
On Wed, Jun 01, 2022 at 02:09:47PM -0700, Axel Rasmussen wrote: > Historically, it has been shown that intercepting kernel faults with > userfaultfd (thereby forcing the kernel to wait for an arbitrary amount > of time) can be exploited, or at least can make some kinds of exploits > easier. So, in 37cd0575b8 "userfaultfd: add UFFD_USER_MODE_ONLY" we > changed things so, in order for kernel faults to be handled by > userfaultfd, either the process needs CAP_SYS_PTRACE, or this sysctl > must be configured so that any unprivileged user can do it. > > In a typical implementation of a hypervisor with live migration (take > QEMU/KVM as one such example), we do indeed need to be able to handle > kernel faults. But, both options above are less than ideal: > > - Toggling the sysctl increases attack surface by allowing any > unprivileged user to do it. > > - Granting the live migration process CAP_SYS_PTRACE gives it this > ability, but *also* the ability to "observe and control the > execution of another process [...], and examine and change [its] > memory and registers" (from ptrace(2)). This isn't something we need > or want to be able to do, so granting this permission violates the > "principle of least privilege". > > This is all a long winded way to say: we want a more fine-grained way to > grant access to userfaultfd, without granting other additional > permissions at the same time. > > To achieve this, add a /dev/userfaultfd misc device. This device > provides an alternative to the userfaultfd(2) syscall for the creation > of new userfaultfds. The idea is, any userfaultfds created this way will > be able to handle kernel faults, without the caller having any special > capabilities. Access to this mechanism is instead restricted using e.g. > standard filesystem permissions. > > Signed-off-by: Axel Rasmussen <axelrasmussen@google.com> > --- > fs/userfaultfd.c | 76 ++++++++++++++++++++++++++------ > include/uapi/linux/userfaultfd.h | 4 ++ > 2 files changed, 66 insertions(+), 14 deletions(-) > > diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c > index e943370107d0..8b92c1398169 100644 > --- a/fs/userfaultfd.c > +++ b/fs/userfaultfd.c > @@ -30,6 +30,7 @@ > #include <linux/security.h> > #include <linux/hugetlb.h> > #include <linux/swapops.h> > +#include <linux/miscdevice.h> > > int sysctl_unprivileged_userfaultfd __read_mostly; > > @@ -413,13 +414,8 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason) > > if (ctx->features & UFFD_FEATURE_SIGBUS) > goto out; > - if ((vmf->flags & FAULT_FLAG_USER) == 0 && > - ctx->flags & UFFD_USER_MODE_ONLY) { > - printk_once(KERN_WARNING "uffd: Set unprivileged_userfaultfd " > - "sysctl knob to 1 if kernel faults must be handled " > - "without obtaining CAP_SYS_PTRACE capability\n"); > + if (!(vmf->flags & FAULT_FLAG_USER) && (ctx->flags & UFFD_USER_MODE_ONLY)) > goto out; > - } > > /* > * If it's already released don't get it. This avoids to loop > @@ -2052,19 +2048,33 @@ static void init_once_userfaultfd_ctx(void *mem) > seqcount_spinlock_init(&ctx->refile_seq, &ctx->fault_pending_wqh.lock); > } > > -SYSCALL_DEFINE1(userfaultfd, int, flags) > +static inline bool userfaultfd_allowed(bool is_syscall, int flags) > +{ > + bool kernel_faults = !(flags & UFFD_USER_MODE_ONLY); > + bool allow_unprivileged = sysctl_unprivileged_userfaultfd; > + > + /* userfaultfd(2) access is controlled by sysctl + capability. */ > + if (is_syscall && kernel_faults) { > + if (!allow_unprivileged && !capable(CAP_SYS_PTRACE)) > + return false; > + } > + > + /* > + * For /dev/userfaultfd, access is to be controlled using e.g. > + * permissions on the device node. We assume this is correctly > + * configured by userspace, so we simply allow access here. > + */ > + > + return true; > +} This helper reads a bit weird because potentially it constantly returns "true" for !syscall use case but it's very not obvious.. Would it be cleaner to not pass in the bool at all? Something like (I also un-nested some of the condition checks, hopefully it'll be easier to read): bool userfaultfd_syscall_allowed(int flags) { /* Userspace-only page faults are always allowed */ if (flags & UFFD_USER_MODE_ONLY) return true; /* * The user is requesting kernel fault capabilities. Privileged * users are always allowed even for kernel fault traps. */ if (capable(CAP_SYS_PTRACE)) return true; /* Whether we allow unprivileged users for kernel faults? */ return sysctl_unprivileged_userfaultfd; } Then below... > + > +static int new_userfaultfd(bool is_syscall, int flags) > { > struct userfaultfd_ctx *ctx; > int fd; > > - if (!sysctl_unprivileged_userfaultfd && > - (flags & UFFD_USER_MODE_ONLY) == 0 && > - !capable(CAP_SYS_PTRACE)) { > - printk_once(KERN_WARNING "uffd: Set unprivileged_userfaultfd " > - "sysctl knob to 1 if kernel faults must be handled " > - "without obtaining CAP_SYS_PTRACE capability\n"); > + if (!userfaultfd_allowed(is_syscall, flags)) > return -EPERM; > - } .. we could write it as: if (is_syscall && !userfaultfd_syscall_allowed(flags)) return -EPERM; What do you think? > > BUG_ON(!current->mm); > > @@ -2083,6 +2093,10 @@ SYSCALL_DEFINE1(userfaultfd, int, flags) > refcount_set(&ctx->refcount, 1); > ctx->flags = flags; > ctx->features = 0; > + /* > + * If UFFD_USER_MODE_ONLY is not set, then userfaultfd_allowed() above > + * decided that kernel faults were allowed and should be handled. > + */ Hmm.. why this needs to be added above "released=false"? Did you want to add this (perhaps) above "flags" instead? IMHO when people reading the flags it'll be clear already on how it was handled, the thing is the comment probably hide deep anyway so I'd consider omitting it. The rest looks good to me, thanks. > ctx->released = false; > atomic_set(&ctx->mmap_changing, 0); > ctx->mm = current->mm; > @@ -2098,8 +2112,42 @@ SYSCALL_DEFINE1(userfaultfd, int, flags) > return fd; > } > > +SYSCALL_DEFINE1(userfaultfd, int, flags) > +{ > + return new_userfaultfd(true, flags); > +} > + > +static int userfaultfd_dev_open(struct inode *inode, struct file *file) > +{ > + return 0; > +} > + > +static long userfaultfd_dev_ioctl(struct file *file, unsigned int cmd, unsigned long flags) > +{ > + if (cmd != USERFAULTFD_IOC_NEW) > + return -EINVAL; > + > + return new_userfaultfd(false, flags); > +} > + > +static const struct file_operations userfaultfd_dev_fops = { > + .open = userfaultfd_dev_open, > + .unlocked_ioctl = userfaultfd_dev_ioctl, > + .compat_ioctl = userfaultfd_dev_ioctl, > + .owner = THIS_MODULE, > + .llseek = noop_llseek, > +}; > + > +static struct miscdevice userfaultfd_misc = { > + .minor = MISC_DYNAMIC_MINOR, > + .name = "userfaultfd", > + .fops = &userfaultfd_dev_fops > +}; > + > static int __init userfaultfd_init(void) > { > + WARN_ON(misc_register(&userfaultfd_misc)); > + > userfaultfd_ctx_cachep = kmem_cache_create("userfaultfd_ctx_cache", > sizeof(struct userfaultfd_ctx), > 0, > diff --git a/include/uapi/linux/userfaultfd.h b/include/uapi/linux/userfaultfd.h > index 7d32b1e797fb..005e5e306266 100644 > --- a/include/uapi/linux/userfaultfd.h > +++ b/include/uapi/linux/userfaultfd.h > @@ -12,6 +12,10 @@ > > #include <linux/types.h> > > +/* ioctls for /dev/userfaultfd */ > +#define USERFAULTFD_IOC 0xAA > +#define USERFAULTFD_IOC_NEW _IO(USERFAULTFD_IOC, 0x00) > + > /* > * If the UFFDIO_API is upgraded someday, the UFFDIO_UNREGISTER and > * UFFDIO_WAKE ioctls should be defined as _IOW and not as _IOR. In > -- > 2.36.1.255.ge46751e96f-goog >
On Mon, Jun 13, 2022 at 4:23 PM Jonathan Corbet <corbet@lwn.net> wrote: > > Axel Rasmussen <axelrasmussen@google.com> writes: > > > I think for any approach involving syscalls, we need to be able to > > control access to who can call a syscall. Maybe there's another way > > I'm not aware of, but I think today the only mechanism to do this is > > capabilities. I proposed adding a CAP_USERFAULTFD for this purpose, > > but that approach was rejected [1]. So, I'm not sure of another way > > besides using a device node. > > I take it there's a reason why this can't be done with a security module > - either a custom module or a policy in one of the existing modules? > That sort of access control is just what security modules are supposed > to be for, after all. > > Thanks, > > jon Admittedly I haven't tried proposing a patch, but I suspect there would be pushback against adding an entirely new LSM just for this case, similarly to the reasons the CAP_USERFAULTFD approach was rejected. For existing LSMs, I think SELinux can be used to restrict access to syscalls. But then again, it's fairly heavy weight / difficult to configure, and I suspect migrating production servers which don't use it today would be a nontrivial undertaking. At least to me it seems unfortunate to say, there isn't an obvious "safe" way to use userfaultfd, without enabling + configuring selinux. (That assumes by "safe" we mean, without granting wider-than necessary access to userfaultfd, or without granting uffd-using processes more permissions [root or CAP_SYS_PTRACE] to do their job.) I suspect if we do that then in practice many? most? users will just either run UFFD programs as root, or toggle the sysctl to allow unprivileged UFFD usage.
On Tue, Jun 14, 2022 at 12:09 PM Peter Xu <peterx@redhat.com> wrote: > > On Wed, Jun 01, 2022 at 02:09:47PM -0700, Axel Rasmussen wrote: > > Historically, it has been shown that intercepting kernel faults with > > userfaultfd (thereby forcing the kernel to wait for an arbitrary amount > > of time) can be exploited, or at least can make some kinds of exploits > > easier. So, in 37cd0575b8 "userfaultfd: add UFFD_USER_MODE_ONLY" we > > changed things so, in order for kernel faults to be handled by > > userfaultfd, either the process needs CAP_SYS_PTRACE, or this sysctl > > must be configured so that any unprivileged user can do it. > > > > In a typical implementation of a hypervisor with live migration (take > > QEMU/KVM as one such example), we do indeed need to be able to handle > > kernel faults. But, both options above are less than ideal: > > > > - Toggling the sysctl increases attack surface by allowing any > > unprivileged user to do it. > > > > - Granting the live migration process CAP_SYS_PTRACE gives it this > > ability, but *also* the ability to "observe and control the > > execution of another process [...], and examine and change [its] > > memory and registers" (from ptrace(2)). This isn't something we need > > or want to be able to do, so granting this permission violates the > > "principle of least privilege". > > > > This is all a long winded way to say: we want a more fine-grained way to > > grant access to userfaultfd, without granting other additional > > permissions at the same time. > > > > To achieve this, add a /dev/userfaultfd misc device. This device > > provides an alternative to the userfaultfd(2) syscall for the creation > > of new userfaultfds. The idea is, any userfaultfds created this way will > > be able to handle kernel faults, without the caller having any special > > capabilities. Access to this mechanism is instead restricted using e.g. > > standard filesystem permissions. > > > > Signed-off-by: Axel Rasmussen <axelrasmussen@google.com> > > --- > > fs/userfaultfd.c | 76 ++++++++++++++++++++++++++------ > > include/uapi/linux/userfaultfd.h | 4 ++ > > 2 files changed, 66 insertions(+), 14 deletions(-) > > > > diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c > > index e943370107d0..8b92c1398169 100644 > > --- a/fs/userfaultfd.c > > +++ b/fs/userfaultfd.c > > @@ -30,6 +30,7 @@ > > #include <linux/security.h> > > #include <linux/hugetlb.h> > > #include <linux/swapops.h> > > +#include <linux/miscdevice.h> > > > > int sysctl_unprivileged_userfaultfd __read_mostly; > > > > @@ -413,13 +414,8 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason) > > > > if (ctx->features & UFFD_FEATURE_SIGBUS) > > goto out; > > - if ((vmf->flags & FAULT_FLAG_USER) == 0 && > > - ctx->flags & UFFD_USER_MODE_ONLY) { > > - printk_once(KERN_WARNING "uffd: Set unprivileged_userfaultfd " > > - "sysctl knob to 1 if kernel faults must be handled " > > - "without obtaining CAP_SYS_PTRACE capability\n"); > > + if (!(vmf->flags & FAULT_FLAG_USER) && (ctx->flags & UFFD_USER_MODE_ONLY)) > > goto out; > > - } > > > > /* > > * If it's already released don't get it. This avoids to loop > > @@ -2052,19 +2048,33 @@ static void init_once_userfaultfd_ctx(void *mem) > > seqcount_spinlock_init(&ctx->refile_seq, &ctx->fault_pending_wqh.lock); > > } > > > > -SYSCALL_DEFINE1(userfaultfd, int, flags) > > +static inline bool userfaultfd_allowed(bool is_syscall, int flags) > > +{ > > + bool kernel_faults = !(flags & UFFD_USER_MODE_ONLY); > > + bool allow_unprivileged = sysctl_unprivileged_userfaultfd; > > + > > + /* userfaultfd(2) access is controlled by sysctl + capability. */ > > + if (is_syscall && kernel_faults) { > > + if (!allow_unprivileged && !capable(CAP_SYS_PTRACE)) > > + return false; > > + } > > + > > + /* > > + * For /dev/userfaultfd, access is to be controlled using e.g. > > + * permissions on the device node. We assume this is correctly > > + * configured by userspace, so we simply allow access here. > > + */ > > + > > + return true; > > +} > > This helper reads a bit weird because potentially it constantly returns > "true" for !syscall use case but it's very not obvious.. > > Would it be cleaner to not pass in the bool at all? Something like (I also > un-nested some of the condition checks, hopefully it'll be easier to read): > > bool userfaultfd_syscall_allowed(int flags) > { > /* Userspace-only page faults are always allowed */ > if (flags & UFFD_USER_MODE_ONLY) > return true; > > /* > * The user is requesting kernel fault capabilities. Privileged > * users are always allowed even for kernel fault traps. > */ > if (capable(CAP_SYS_PTRACE)) > return true; > > /* Whether we allow unprivileged users for kernel faults? */ > return sysctl_unprivileged_userfaultfd; > } > > Then below... > > > + > > +static int new_userfaultfd(bool is_syscall, int flags) > > { > > struct userfaultfd_ctx *ctx; > > int fd; > > > > - if (!sysctl_unprivileged_userfaultfd && > > - (flags & UFFD_USER_MODE_ONLY) == 0 && > > - !capable(CAP_SYS_PTRACE)) { > > - printk_once(KERN_WARNING "uffd: Set unprivileged_userfaultfd " > > - "sysctl knob to 1 if kernel faults must be handled " > > - "without obtaining CAP_SYS_PTRACE capability\n"); > > + if (!userfaultfd_allowed(is_syscall, flags)) > > return -EPERM; > > - } > > .. we could write it as: > > if (is_syscall && !userfaultfd_syscall_allowed(flags)) > return -EPERM; > > What do you think? > > > > > BUG_ON(!current->mm); > > > > @@ -2083,6 +2093,10 @@ SYSCALL_DEFINE1(userfaultfd, int, flags) > > refcount_set(&ctx->refcount, 1); > > ctx->flags = flags; > > ctx->features = 0; > > + /* > > + * If UFFD_USER_MODE_ONLY is not set, then userfaultfd_allowed() above > > + * decided that kernel faults were allowed and should be handled. > > + */ > > Hmm.. why this needs to be added above "released=false"? Did you want to > add this (perhaps) above "flags" instead? > > IMHO when people reading the flags it'll be clear already on how it was > handled, the thing is the comment probably hide deep anyway so I'd consider > omitting it. > > The rest looks good to me, thanks. Thanks for reviewing, Peter! Most of these comments look good to me, I'll include them in a v4 after I get back to the office in about a week. > > > ctx->released = false; > > atomic_set(&ctx->mmap_changing, 0); > > ctx->mm = current->mm; > > @@ -2098,8 +2112,42 @@ SYSCALL_DEFINE1(userfaultfd, int, flags) > > return fd; > > } > > > > +SYSCALL_DEFINE1(userfaultfd, int, flags) > > +{ > > + return new_userfaultfd(true, flags); > > +} > > + > > +static int userfaultfd_dev_open(struct inode *inode, struct file *file) > > +{ > > + return 0; > > +} > > + > > +static long userfaultfd_dev_ioctl(struct file *file, unsigned int cmd, unsigned long flags) > > +{ > > + if (cmd != USERFAULTFD_IOC_NEW) > > + return -EINVAL; > > + > > + return new_userfaultfd(false, flags); > > +} > > + > > +static const struct file_operations userfaultfd_dev_fops = { > > + .open = userfaultfd_dev_open, > > + .unlocked_ioctl = userfaultfd_dev_ioctl, > > + .compat_ioctl = userfaultfd_dev_ioctl, > > + .owner = THIS_MODULE, > > + .llseek = noop_llseek, > > +}; > > + > > +static struct miscdevice userfaultfd_misc = { > > + .minor = MISC_DYNAMIC_MINOR, > > + .name = "userfaultfd", > > + .fops = &userfaultfd_dev_fops > > +}; > > + > > static int __init userfaultfd_init(void) > > { > > + WARN_ON(misc_register(&userfaultfd_misc)); > > + > > userfaultfd_ctx_cachep = kmem_cache_create("userfaultfd_ctx_cache", > > sizeof(struct userfaultfd_ctx), > > 0, > > diff --git a/include/uapi/linux/userfaultfd.h b/include/uapi/linux/userfaultfd.h > > index 7d32b1e797fb..005e5e306266 100644 > > --- a/include/uapi/linux/userfaultfd.h > > +++ b/include/uapi/linux/userfaultfd.h > > @@ -12,6 +12,10 @@ > > > > #include <linux/types.h> > > > > +/* ioctls for /dev/userfaultfd */ > > +#define USERFAULTFD_IOC 0xAA > > +#define USERFAULTFD_IOC_NEW _IO(USERFAULTFD_IOC, 0x00) > > + > > /* > > * If the UFFDIO_API is upgraded someday, the UFFDIO_UNREGISTER and > > * UFFDIO_WAKE ioctls should be defined as _IOW and not as _IOR. In > > -- > > 2.36.1.255.ge46751e96f-goog > > > > -- > Peter Xu >
On Jun 14, 2022, at 5:55 PM, Axel Rasmussen <axelrasmussen@google.com> wrote: > ⚠ External Email > > On Mon, Jun 13, 2022 at 5:10 PM Nadav Amit <namit@vmware.com> wrote: >> On Jun 13, 2022, at 3:38 PM, Axel Rasmussen <axelrasmussen@google.com> wrote: >> >>> On Mon, Jun 13, 2022 at 3:29 PM Peter Xu <peterx@redhat.com> wrote: >>>> On Mon, Jun 13, 2022 at 02:55:40PM -0700, Andrew Morton wrote: >>>>> On Wed, 1 Jun 2022 14:09:47 -0700 Axel Rasmussen <axelrasmussen@google.com> wrote: >>>>> >>>>>> To achieve this, add a /dev/userfaultfd misc device. This device >>>>>> provides an alternative to the userfaultfd(2) syscall for the creation >>>>>> of new userfaultfds. The idea is, any userfaultfds created this way will >>>>>> be able to handle kernel faults, without the caller having any special >>>>>> capabilities. Access to this mechanism is instead restricted using e.g. >>>>>> standard filesystem permissions. >>>>> >>>>> The use of a /dev node isn't pretty. Why can't this be done by >>>>> tweaking sys_userfaultfd() or by adding a sys_userfaultfd2()? >>> >>> I think for any approach involving syscalls, we need to be able to >>> control access to who can call a syscall. Maybe there's another way >>> I'm not aware of, but I think today the only mechanism to do this is >>> capabilities. I proposed adding a CAP_USERFAULTFD for this purpose, >>> but that approach was rejected [1]. So, I'm not sure of another way >>> besides using a device node. >>> >>> One thing that could potentially make this cleaner is, as one LWN >>> commenter pointed out, we could have open() on /dev/userfaultfd just >>> return a new userfaultfd directly, instead of this multi-step process >>> of open /dev/userfaultfd, NEW ioctl, then you get a userfaultfd. When >>> I wrote this originally it wasn't clear to me how to get that to >>> happen - open() doesn't directly return the result of our custom open >>> function pointer, as far as I can tell - but it could be investigated. >> >> If this direction is pursued, I think that it would be better to set it as >> /proc/[pid]/userfaultfd, which would allow remote monitors (processes) to >> hook into userfaultfd of remote processes. I have a patch for that which >> extends userfaultfd syscall, but /proc/[pid]/userfaultfd may be cleaner. > > Hmm, one thing I'm unsure about - > > If a process is able to control another process' memory like this, > then this seems like exactly what CAP_SYS_PTRACE is intended to deal > with, right? So I'm not sure this case is directly related to the one > I'm trying to address. > > This also seems distinct to me versus the existing way you'd do this, > which is open a userfaultfd and register a shared memory region, and > then fork(). Now you can control your child's memory with userfaultfd. > But, attaching to some other, previously-unrelated process with > /proc/[pid]/userfaultfd seems like a clear case for CAP_SYS_PTRACE. I agree about CAP_SYS_PTRACE. I just know that if the /dev approach is taken, there would be even more pushback for userfaultfd2. Whatever.
diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c index e943370107d0..8b92c1398169 100644 --- a/fs/userfaultfd.c +++ b/fs/userfaultfd.c @@ -30,6 +30,7 @@ #include <linux/security.h> #include <linux/hugetlb.h> #include <linux/swapops.h> +#include <linux/miscdevice.h> int sysctl_unprivileged_userfaultfd __read_mostly; @@ -413,13 +414,8 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason) if (ctx->features & UFFD_FEATURE_SIGBUS) goto out; - if ((vmf->flags & FAULT_FLAG_USER) == 0 && - ctx->flags & UFFD_USER_MODE_ONLY) { - printk_once(KERN_WARNING "uffd: Set unprivileged_userfaultfd " - "sysctl knob to 1 if kernel faults must be handled " - "without obtaining CAP_SYS_PTRACE capability\n"); + if (!(vmf->flags & FAULT_FLAG_USER) && (ctx->flags & UFFD_USER_MODE_ONLY)) goto out; - } /* * If it's already released don't get it. This avoids to loop @@ -2052,19 +2048,33 @@ static void init_once_userfaultfd_ctx(void *mem) seqcount_spinlock_init(&ctx->refile_seq, &ctx->fault_pending_wqh.lock); } -SYSCALL_DEFINE1(userfaultfd, int, flags) +static inline bool userfaultfd_allowed(bool is_syscall, int flags) +{ + bool kernel_faults = !(flags & UFFD_USER_MODE_ONLY); + bool allow_unprivileged = sysctl_unprivileged_userfaultfd; + + /* userfaultfd(2) access is controlled by sysctl + capability. */ + if (is_syscall && kernel_faults) { + if (!allow_unprivileged && !capable(CAP_SYS_PTRACE)) + return false; + } + + /* + * For /dev/userfaultfd, access is to be controlled using e.g. + * permissions on the device node. We assume this is correctly + * configured by userspace, so we simply allow access here. + */ + + return true; +} + +static int new_userfaultfd(bool is_syscall, int flags) { struct userfaultfd_ctx *ctx; int fd; - if (!sysctl_unprivileged_userfaultfd && - (flags & UFFD_USER_MODE_ONLY) == 0 && - !capable(CAP_SYS_PTRACE)) { - printk_once(KERN_WARNING "uffd: Set unprivileged_userfaultfd " - "sysctl knob to 1 if kernel faults must be handled " - "without obtaining CAP_SYS_PTRACE capability\n"); + if (!userfaultfd_allowed(is_syscall, flags)) return -EPERM; - } BUG_ON(!current->mm); @@ -2083,6 +2093,10 @@ SYSCALL_DEFINE1(userfaultfd, int, flags) refcount_set(&ctx->refcount, 1); ctx->flags = flags; ctx->features = 0; + /* + * If UFFD_USER_MODE_ONLY is not set, then userfaultfd_allowed() above + * decided that kernel faults were allowed and should be handled. + */ ctx->released = false; atomic_set(&ctx->mmap_changing, 0); ctx->mm = current->mm; @@ -2098,8 +2112,42 @@ SYSCALL_DEFINE1(userfaultfd, int, flags) return fd; } +SYSCALL_DEFINE1(userfaultfd, int, flags) +{ + return new_userfaultfd(true, flags); +} + +static int userfaultfd_dev_open(struct inode *inode, struct file *file) +{ + return 0; +} + +static long userfaultfd_dev_ioctl(struct file *file, unsigned int cmd, unsigned long flags) +{ + if (cmd != USERFAULTFD_IOC_NEW) + return -EINVAL; + + return new_userfaultfd(false, flags); +} + +static const struct file_operations userfaultfd_dev_fops = { + .open = userfaultfd_dev_open, + .unlocked_ioctl = userfaultfd_dev_ioctl, + .compat_ioctl = userfaultfd_dev_ioctl, + .owner = THIS_MODULE, + .llseek = noop_llseek, +}; + +static struct miscdevice userfaultfd_misc = { + .minor = MISC_DYNAMIC_MINOR, + .name = "userfaultfd", + .fops = &userfaultfd_dev_fops +}; + static int __init userfaultfd_init(void) { + WARN_ON(misc_register(&userfaultfd_misc)); + userfaultfd_ctx_cachep = kmem_cache_create("userfaultfd_ctx_cache", sizeof(struct userfaultfd_ctx), 0, diff --git a/include/uapi/linux/userfaultfd.h b/include/uapi/linux/userfaultfd.h index 7d32b1e797fb..005e5e306266 100644 --- a/include/uapi/linux/userfaultfd.h +++ b/include/uapi/linux/userfaultfd.h @@ -12,6 +12,10 @@ #include <linux/types.h> +/* ioctls for /dev/userfaultfd */ +#define USERFAULTFD_IOC 0xAA +#define USERFAULTFD_IOC_NEW _IO(USERFAULTFD_IOC, 0x00) + /* * If the UFFDIO_API is upgraded someday, the UFFDIO_UNREGISTER and * UFFDIO_WAKE ioctls should be defined as _IOW and not as _IOR. In
Historically, it has been shown that intercepting kernel faults with userfaultfd (thereby forcing the kernel to wait for an arbitrary amount of time) can be exploited, or at least can make some kinds of exploits easier. So, in 37cd0575b8 "userfaultfd: add UFFD_USER_MODE_ONLY" we changed things so, in order for kernel faults to be handled by userfaultfd, either the process needs CAP_SYS_PTRACE, or this sysctl must be configured so that any unprivileged user can do it. In a typical implementation of a hypervisor with live migration (take QEMU/KVM as one such example), we do indeed need to be able to handle kernel faults. But, both options above are less than ideal: - Toggling the sysctl increases attack surface by allowing any unprivileged user to do it. - Granting the live migration process CAP_SYS_PTRACE gives it this ability, but *also* the ability to "observe and control the execution of another process [...], and examine and change [its] memory and registers" (from ptrace(2)). This isn't something we need or want to be able to do, so granting this permission violates the "principle of least privilege". This is all a long winded way to say: we want a more fine-grained way to grant access to userfaultfd, without granting other additional permissions at the same time. To achieve this, add a /dev/userfaultfd misc device. This device provides an alternative to the userfaultfd(2) syscall for the creation of new userfaultfds. The idea is, any userfaultfds created this way will be able to handle kernel faults, without the caller having any special capabilities. Access to this mechanism is instead restricted using e.g. standard filesystem permissions. Signed-off-by: Axel Rasmussen <axelrasmussen@google.com> --- fs/userfaultfd.c | 76 ++++++++++++++++++++++++++------ include/uapi/linux/userfaultfd.h | 4 ++ 2 files changed, 66 insertions(+), 14 deletions(-)