fork: do not expose incomplete mm on fork
Subsystems (7)
cover fork: do not expose incomplete mm on fork
During fork we may place the virtual memory address space into an inconsistent state before the fork operation is complete.
In addition, we may encounter an error during the fork operation that indicates that the virtual memory address space is invalidated.
As a result, we should not be exposing it in any way to external machinery that might interact with the mm or VMAs, machinery that is not designed to deal with incomplete state.
We specifically update the fork logic to defer khugepaged and ksm to the end of the operation and only to be invoked if no error arose, and disallow uffd from observing fork events should an error have occurred.
Lorenzo Stoakes (2):
fork: do not invoke uffd on fork if error occurs
fork: only invoke khugepaged, ksm hooks if no error
fs/userfaultfd.c | 28 ++++++++++++++++++++++++++++ include/linux/ksm.h | 10 ++++------ include/linux/userfaultfd_k.h | 5 +++++ kernel/fork.c | 12 ++++++------ 4 files changed, 43 insertions(+), 12 deletions(-)
--
2.46.2
PATCH 1/2 fork: do not invoke uffd on fork if error occurs
Currently on fork we expose the virtual address space of a process to userland unconditionally if uffd is registered in VMAs, regardless of whether an error arose in the fork.
This is performed in dup_userfaultfd_complete() which is invoked unconditionally, and performs two duties - invoking registered handlers for the UFFD_EVENT_FORK event via dup_fctx(), and clearing down userfaultfd_fork_ctx objects established in dup_userfaultfd().
This is problematic, because the virtual address space may not yet be correctly initialised if an error arose.
The change in commit d24062914837 ("fork: use __mt_dup() to duplicate maple tree in dup_mmap()") makes this more pertinent as we may be in a state where entries in the maple tree are not yet consistent.
We address this by, on fork error, ensuring that we roll back state that we would otherwise expect to clean up through the event being handled by userland and perform the memory freeing duty otherwise performed by dup_userfaultfd_complete().
We do this by implementing a new function, dup_userfaultfd_fail(), which performs the same loop, only decrementing reference counts.
Note that we perform mmgrab() on the parent and child mm's, however userfaultfd_ctx_put() will mmdrop() this once the reference count drops to zero, so we will avoid memory leaks correctly here.
--- fs/userfaultfd.c | 28 ++++++++++++++++++++++++++++ include/linux/userfaultfd_k.h | 5 +++++ kernel/fork.c | 5 ++++- 3 files changed, 37 insertions(+), 1 deletion(-)
fs/userfaultfd.c +28 -0
@@ -692,6 +692,34 @@ void dup_userfaultfd_complete(struct list_head *fcs) } } +void dup_userfaultfd_fail (struct list_head *fcs) +{ +struct userfaultfd_fork_ctx *fctx, *n; + +/* +* An error has occurred on fork, we will tear memory down, but have +* allocated memory for fctx's and raised reference counts for both the +* original and child contexts (and on the mm for each as a result). +* +* These would ordinarily be taken care of by a user handling the event, +* but we are no longer doing so, so manually clean up here. +* +* mm tear down will take care of cleaning up VMA contexts. +*/ +list_for_each_entry_safe (fctx, n, fcs, list) { +struct userfaultfd_ctx *octx = fctx->orig; +struct userfaultfd_ctx *ctx = fctx->new; + +atomic_dec (&octx->mmap_changing); +VM_BUG_ON (atomic_read (&octx->mmap_changing) <0 ); +userfaultfd_ctx_put (octx); +userfaultfd_ctx_put (ctx); + +list_del (&fctx->list); +kfree (fctx); + } +} +void mremap_userfaultfd_prep (struct vm_area_struct *vma,struct vm_userfaultfd_ctx *vm_ctx) {
include/linux/userfaultfd_k.h +5 -0
@@ -249,6 +249,7 @@ static inline bool vma_can_userfault(struct vm_area_struct *vma,extern int dup_userfaultfd (struct vm_area_struct *,struct list_head *);extern void dup_userfaultfd_complete (struct list_head *); +void dup_userfaultfd_fail (struct list_head *);extern void mremap_userfaultfd_prep (struct vm_area_struct *,struct vm_userfaultfd_ctx *);
@@ -351,6 +352,10 @@ static inline void dup_userfaultfd_complete(struct list_head *l) { } +static inline void dup_userfaultfd_fail (struct list_head *l) +{ +} +static inline void mremap_userfaultfd_prep (struct vm_area_struct *vma,struct vm_userfaultfd_ctx *ctx) {
kernel/fork.c +4 -1
@@ -775,7 +775,10 @@ static __latent_entropy int dup_mmap(struct mm_struct *mm,mmap_write_unlock (mm);flush_tlb_mm (oldmm);mmap_write_unlock (oldmm); -dup_userfaultfd_complete (&uf); +if (!retval) +dup_userfaultfd_complete (&uf); +else +dup_userfaultfd_fail (&uf); fail_uprobe_end:uprobe_end_dup_mmap ();return retval;
--
2.46.2
PATCH 2/2 fork: only invoke khugepaged, ksm hooks if no error
In reply to Lorenzo StoakesThere is no reason to invoke these hooks early against an mm that is in an incomplete state.
The change in commit d24062914837 ("fork: use __mt_dup() to duplicate maple tree in dup_mmap()") makes this more pertinent as we may be in a state where entries in the maple tree are not yet consistent.
Their placement early in dup_mmap() only appears to have been meaningful for early error checking, and since functionally it'd require a very small allocation to fail (in practice 'too small to fail') that'd only occur in the most dire circumstances, meaning the fork would fail or be OOM'd in any case.
Since both khugepaged and KSM tracking are there to provide optimisations to memory performance rather than critical functionality, it doesn't really matter all that much if, under such dire memory pressure, we fail to register an mm with these.
As a result, we follow the example of commit d2081b2bf819 ("mm: khugepaged:
make khugepaged_enter() void function") and make ksm_fork() a void function also.
We only expose the mm to these functions once we are done with them and only if no error occurred in the fork operation.
--- include/linux/ksm.h | 10 ++++------ kernel/fork.c | 7 ++----- 2 files changed, 6 insertions(+), 11 deletions(-)
include/linux/ksm.h +4 -6
@@ -54,12 +54,11 @@ static inline long mm_ksm_zero_pages(struct mm_struct *mm)return atomic_long_read (&mm->ksm_zero_pages); } -static inline int ksm_fork (struct mm_struct *mm,struct mm_struct *oldmm) +static inline void ksm_fork (struct mm_struct *mm,struct mm_struct *oldmm) { +/* Adding mm to ksm is best effort on fork. */ if (test_bit (MMF_VM_MERGEABLE , &oldmm->flags)) -return __ksm_enter (mm); - -return 0 ; +__ksm_enter (mm); }static inline int ksm_execve (struct mm_struct *mm)
@@ -107,9 +106,8 @@ static inline int ksm_disable(struct mm_struct *mm)return 0 ; } -static inline int ksm_fork (struct mm_struct *mm,struct mm_struct *oldmm) +static inline void ksm_fork (struct mm_struct *mm,struct mm_struct *oldmm) { -return 0 ; }static inline int ksm_execve (struct mm_struct *mm)
kernel/fork.c +2 -5
@@ -653,11 +653,6 @@ static __latent_entropy int dup_mmap(struct mm_struct *mm, mm->exec_vm = oldmm->exec_vm; mm->stack_vm = oldmm->stack_vm; - retval =ksm_fork (mm, oldmm); -if (retval) - goto out; -khugepaged_fork (mm, oldmm); -/* Use __mt_dup() to efficiently build an identical maple tree. */ retval =__mt_dup (&oldmm->mm_mt, &mm->mm_mt,GFP_KERNEL );if (unlikely (retval))
@@ -760,6 +755,8 @@ static __latent_entropy int dup_mmap(struct mm_struct *mm,vma_iter_free (&vmi);if (!retval) {mt_set_in_rcu (vmi.mas.tree); +ksm_fork (mm, oldmm); +khugepaged_fork (mm, oldmm); }else if (mpnt) {/* * The entire maple tree has already been duplicated. If the
--
2.46.2